I GBPT OF ^apg-sa^ EX LIBR1S / . . . ............................. 414 Composition of the limestone .................. ............. , ....... 414 Economics of using slag-limestone mixtures ........................... 414 References on slag-limestone mixtures ................. '. .............. 415 Blast-furnace methods of making cement ................................. 415 Wet methods of preparation ............................................ 416 Disadvantages of wet methods ...................................... 416 Treatment of the marl and clay ..................................... 417 Power and output in wet grinding ................................... 417 Agitating the mix ........ .......... ................... .............. 418 Old-style wet methods ............................. . . . . . ............ 419 Actual equipment of wet-process plants ............ ................... 420 List of references on wet-process plants and methods .................. 421 General crushing practice in cement plants ................................ 422 Necessity for fine grinding. ... ............................ .......... 423 Actual fineness attained in practice ................. ................. 424 Gradual vs. one-stage reduction ...................................... 424 Present day systems ...................... ......................... 425 The omission of separators in cement plants ........................... 426 Advantages of separators .................... . ...................... 426 Disadvantages of separators .................... ..................... 427 Uniformity of product destroyed ................................. 427 Great fineness prevented ................... ..................... 427 CHAPTER XXXV. STANDARD TYPES OF CRUSHING AND PULVERIZING MACHINERY. Classification of grinding machinery ...................................... 429 Class 1 . Jaw crushers .................................................. 430 Class 2. Cone grinders or gyratory crushers ............................... 430 Class 3. Rolls ......................................................... 433 Class 4. Millstones ................................. .................... 437 Class 5. Edge-runners: Dry pans ........................................ 437 Class 6. Centrifugal grinders ........................................... 438 Huntingdon mill ............................................. 438 Griffin mill ............... . ................................. . 440 Kent mill ............... ................................... .443 7. Ball grinders .................... . ............ ................. 447 Kominuter. . .................................... 448 TABLE OF CONTENTS. xix PAGE Ball mills 452 Tube mills. 457 Pebbles for tube mills .463 Class 8. Impact pulverizers 465 Raymond pulverizer 466 Williams mill 466 CHAPTER XXXII. CEMENT BURNING: FIXED KILNS. Classes of fixed or stationary kilns 469 1. Dome or ordinary intermittent kilns 470 2. Dome kilns with drying accessories 471 Johnson kiln . . 471 3. Ring or Hoffmann kiln 472 4. Continuous shaft kilns 474 Dietzsch kiln 474 Aalborg or Schofer kiln. . . . 474 Hauenschild kiln . 475 Schwarz kiln 477 Reference list for fixed kilns 478 CHAPTER XXXIII. THE ROTARY KILN. Early history 480 Summary of burning process 481 Shape and size 482 Feeding coal to the kiln 486 Gas-burners for rotary kilns 489 Kiln linings 490 Actual fuel consumption and output 492 CHAPTER XXXIV. HEAT CONSUMPTION AND HEAT UTILIZATION. Theoretical heat requirements 497 Purposes for which heat is required 497 Heat utilized in evaporation of water 498 Heat utilized in decomposition of clay 498 Heat utilized in dissociation of sulphates 499 Heat utilized in dissociation of carbonates 499 Temperature required for clinkering 499 Heat utilized in heating the mix 500 Total theoretical heat requirements " 501 Heat losses in practice 502 Sources of loss of heat 502 Heat carried out in flue dust 502 XX TABLE OF CONTENTS. PAGE Sources of heat supply 503 Heat supplied by combustion of fuel 503 Heat supplied by chemical combinations 504 Heat derived from the clinker 504 Heat derived from the stack-gases 504 Estimates and tests of heat distribution 505 Newberry's estimates 505 Helbig's estimates \ > . . . . 505 Results of actual tests : 505 Richard's tests 506 Carpenter's tests . 508 Summary of estimates and tests .....';. 509 Heat utilization and economics 509 List of references on heat requirements, etc 510 CHAPTER XXXV. REQUISITES AND TREATMENT OF KILN FUELS. Coal 512 Characters of kiln coals . 512 Analyses of kiln coals 512 References on coal-fields 513 Crushing coal 514 Drying coal 515 Pulverizing coal 515 Power and output in coal grinding 519 Total cost of coal preparation 520 Fire and explosion risks 520 List of references on coal drying, grinding, etc 521 Oil 522 Use of oil in rotary kilns 522 List of references on petroleum 522 Natural gas 522 Use of natural gas in kilns 522 Analyses and thermal value of gas 523 List of references on natural gas 523 Producer gas 524 General note 524 Producer gas from wood and lignite 524 Composition of producer gas 525 Charcoal 526 CHAPTER XXXVI. CLINKER COOLING, GRINDING, AND STORAGE. USE OF GYPSUM. Clinker cooling 527 General methods of clinker cooling 527 Pan conveyors, rolls, and sprinkling . . . . . 528 TABLE OF CONTENTS. xxi PAGE Stationary tower coolers. 528 One-stage rotary cooler. 529 Atlas two-stage rotary cooler 529 Clinker grinding 530 Power and machinery 531 Increase in fineness 531 Actual equipment of various plants 532 Use and effects of gypsum or plaster 534 Form in which the calcium sulphate is used 535 Effect of calcium sulphate on set of cement 538 Effect of calcium sulphate on strength of cement 543 Methods of using gypsum or plaster 545 Analyses of gypsum and plaster actually used 545 Effect of various other salts on set of cement 545 List of references on use of calcium sulphate, chloride, etc 547 Storage and packing 548 Necessity for storage 548 Designs of storage buildings and bins 548 Testing at the mill 549 Packing weights 551 CHAPTER XXXVII. COSTS AND STATISTICS. Costs of manufacture of Portland cement 554 Factors of cost 554 Cost of land and quarries 555 Cost of equipment and erection 555 Total capital required 556 Current administrative expenses 557 Cost of excavating raw materials 557 Total power required ? 557 Cost of coal for kilns and power 558 Cost of gypsum used 558 Distribution and cost of labor 558 Estimates of total cost per barrel 559 Statistics of the American Portland cement industry 561 Total production in the United States, 1870-1904 562 Production by States, 1901, 1902, 1903 562 Production in the Lehigh district, 1890-1903 564 Imports of foreign Portland, 1899-1903 564 CHAPTER XXXVIII. CONSTITUTION, SETTING, PROPERTIES, AND COMPOSITION. Limitations of chemical analyses 566 Constitution and setting properties 566 xxii TABLE OF CONTENTS. PAGE Available methods of investigation 566 Synthetic investigations , 567 Microscopic investigations 568 Theories of constitution 569 Setting properties of Portland cement 572 Replacement of silica by other acids 573 Replacement of alumina by iron oxide. . 573 Replacement of lime by magnesia. ..'..$ 574 Replacement of lime by other bases 574 References on the constitution of Portland cement 575 Composition of Portland cement. . . , 575 Gradual change in composition since 1850 575 Analyses of American Portland cements 576 Standard methods for analysis 576 CHAPTER XXXIX. PHYSICAL PROPERTIES. TESTING METHODS. Physical properties of Portland cement 583 Value of tests for fineness 584 Specific gravity 585 Setting properties. . .- 586 Tensile strength , . . . . 586 Compressive strength 588 Ratio of compressive to tensile strength 589 Modulus of elasticity 591 Sand cement or silica cement 592 List of references on sand cement 596 Effect of heating on Portland cement. 597 Effects of salt and freezing 598 Effects of exposure to sea-water 602 Standard methods of testing, Am. Soc. C. E 603 CHAPTER XL. SPECIFICATIONS FOR PORTLAND CEMENT. New York State Canals, 1896 - 614 Rapid Transit Subway, N. Y. City, 1900-1901 615 Department of Bridges, N. Y. City, 1901 616 Engineer Corps, U. S. Army, 1902. 617 U. S. Reclamation Service, 1904 620 Canadian Society of Civil Engineers 622 Concrete Steel Engineering Company 624 British Standard Specifications 625 American Society for Testing Materials, 1904 629 TABLE OF CONTENTS. Xxiii PART VII. PUZZOLAN CEMENTS. CHAPTER XLI. PUZZOLANIC MATERIALS IN GENERAL. PAGE Definition of puzzolanic materials 632 Natural puzzolanic materials 632 Pozzuolana ...-..- 633 Trass .". . . . Y. 635 Santorin 635 Arenes, etc , . . : 637 Range and average composition of natural puzzolanic materials 638 Natural puzzolanic materials in the .United States 638 Artificial puzzolanic materials. .,.....,.. 639 Burnt clay 639 Blast-furnace slag 639 CHAPTER- XLII. SLAG CEMENT: REQUISITES AND TREATMENT OF THE SLAG. Summary of general methods of manufacture. 641 Composition of the slag 641 Requisite chemical composition 641 Composition of slags actually used. , 642 Selection of slags 644 Granulating the slag 644 Methods of granulating the slag . 645 Effects of granulating the slag 646 Increased hydraulicity due to granulation * . . 647 Desulphurization due to granulation 648 Drying the slag *#v* 649 Types of dryers used 649 Rotary dryers 649 Vertical dryers 652 CHAPTER XLIII. SLAG CEMENT: LIME, MIXING, AND GRINDING. Character and treatment of the lime 653 Composition of the lime 653 Burning the lime 654 Slaking the lime ' 654 Sieving and grinding the lime 655 Mixing and grinding 655 Proportions of lime and slag 655 Calculating the mixture 656 Pulverizing and mixing 657 xxiv TABLE OF CONTENTS. PAGE Regulation of set 658 General practice at various plants 659 Costs of manufacture of slag cement. 663 Production of slag cement 664 List of references on the manufacture of slag cement 664 CHAPTER XLIV. & SLAG CEMENT: COMPOSITION AND PROPERTIES. Identification of slag cements 666 Chemical composition of slag cements. i 666 Elements present 667 Analyses of slag cements 667 Physical properties of slag cements 668 Specific gravity 668 Color 669 Rapidity of set 670 Strength 670 Resistance to mechanical wear. 670 Rates of compressive to tensile strength 670 List of references on properties and testing of slag cements 671 Specifications for slag (puzzolan) cements 672 CHAPTER XLV. SLAG BRICKS AND SLAG BLOCKS. Definition of the two groups 675 Slag bricks 675 Methods of manufacture 676 Practice at various plants 677 Hardening in steam-cylinders 680 Slag blocks 685 INDEX TO SUBJECTS 691 LIST OF ILLUSTRATIONS. PAGE 1. Map of U. S. showing location of gypsum deposits and plaster-plants Opp. 14 2. Map of Kansas, showing location of gypsum deposits 19 3. Map of New York, showing extent of gypsum-bearing formations 20 4. Map of the gypsum deposits of Ohio 22 5. Nipper for coarse crushing of gypsum 34 6. Cracker for intermediate reduction of gypsum 35 7. Stedman disintegrator, opened 36 8. Stedman disintegrator, cage construction 37 9. Construction and setting of gypsum kettles 38 10. Four-flue kettle, with accessories, dismounted ! 39 11. Kettle with sectional bottom 40 12. Section of 100-ton plaster mill, kettle process 42 13. Plan of Electric Plaster Co.'s mill, Blue Rapids, Kansas 45 14. Plan of plaster mill, rotary calciner process 47 15. Cummer rotary calciner for plaster 48 16. Broughton mixer 51 17. Hair-picker 51 18. Effect of sand on tensile strength of plaster 60 18a. Effect of sand on compressive strength of plaster 62 19. Map of the U. S., showing location of principal lime-producing centers . .Opp. 88 20. Plan and section of lime kilns 102 21. Aalborg kiln for lime burning 103 22. Keystone lime kiln 105 23. O'Connell lime kiln 107 24. Eldred lime-hydrating system 125 25. Campbell lime-hydrater 127 26. Schwarz drying and mixing machine. 137 27. Kiln used for burning hydraulic lime 178 28. Section of hydraulic lime plant, Malain, France 181 29. Tensile strength of Lafarge (grappier) cement 186 30. Tensile strength of feebly hydraulic limes 189 31. Tensile strength of plain hydraulic and selenitic limes 192 32. Compressive strength of " " " " 193 33. Map of U. S., showing location of natural-cement plants Opp. 194 34. Tensile strength of various classes of cement 21& xxv LIST OF ILLUSTRATIONS. PAGE :35. Loading cars, Speed natural-cement quarry 219 .36. Crusher for raw rock at Speed's mill 220 37. Sections of kilns used in natural-cement plants 226 .38. Elevation and section of Campbell kiln 227 39. Plan and details of Campbell kiln 227 40. Kilns of the Lawrence natural-cement plant 228 41. Kilns and coal conveyor at .the Speed plant 229 42. Kilns and kiln housing, Speed plant. .?. 230 43. Kilns and loading tracks, Fort Scott, Kansas 233 44. Sections of cracker for natural-cement grinding 239 45. Bturtevant rock-emery mill. ,,. 240 46. Sturtevant rock-emery millstone 240 47. Section of Sturtevant mill 241 48. Berthelet separator 244 49. Installation of Berthelet separating system 245 50. Effect of plaster on set of natural cement ' 265 51. Effect of salt on compressive strength, natural cement 266 52. " " " " tensile " " " 266 53. " " " " " " " " 267 51 Tensile strength of Louisville cement 268 55. " " " Lehigh natural cements 268 56. " '* " Cumberland natural cements 268 57. ' ' " " .Akron and Cumberland natural cements 270 58. " ' ' " natural cements, Cairo Bridge tests 271 .59. ' ' ' " " Rosendale cements 272 63. " " " natural cements, Saulte Ste. Marie 272 -61. Effect of sand on tensile strength, natural cements 273 G2. Tensile tests, Saulte Ste. Marie 274 j63. Map of the U. S., showing location of Portland-cement plants OpP- 294 64. Percentage of total output from different raw materials 303 65. Working a thick limestone bed 311 66. Working heavy horizontal bed of limestone 315 67. Tunnel in cement rock 330 08. Open cut in cement rock 331 69. Pit in heavy shale bed 358 70. Shale pit worked on two levels 358 71. Stripping a flat, shallow bed 368 72. View of typical limestone quarry 369 73. Temporary tracks laid to face. 369 74. Cableway in cement-rock quarry 370 75. "" " " " " ,. 371 76. Shale pit worked in two levels 371 77. Hoisting from quarry worked in levels 372 78. Steam shovel in cement-rock quarry 372 79. Steam shovel in limestone quarry 374 80. Dredge at marl plant. 376 81. Harris system of marl-pumping 377 .82. Marl-drying plant, Hecla P. C. Co 401 LIST OF ILLUSTRATIONS. xxvii PAGE 83. Elevation of stack-drier, Edison P. C. Co 403 84. Installation of kominuters and tube-mills 405 85. Plan of plant, Lawrence Cement Co 406 S6. Plan of plant, Hudson P. C. Co 407 87. External view of Gates crusher 431 88. Sectional view of Gates crusher 432 89. Elevation of rock-crushing system, Edison P. C. Co 434 90. View looking down between rolls, Edison P. C. Co 435 91. Plan and elevation of Edison rolls 346 -92. Dry-pan 438 93. Huntingdon mills at Atlas plant 439 94. Section of Griffin mill 441 95. Interior of Kent mill 444 96. Lindhard kominuter 446 97. " " 447 98. Exterior view of kominuter 451 99. Gates ball-mill 453 100. Transverse section of Gates ball-mill 454 101. Automatic ball-mill feeder 456 102. Gates tube-mill 458 103. Setting of Davidsen tube-mill 461 104. Exterior view of Bonnot tube-mill 463 105. Williams mill, casing opened 466 106. ' ' " , section 467 107. Dome kiln 470 108. Plan and section of Johnson kiln ; 472 1C9. Section of Hoffman kiln 473 110. Plan of Hoffman kiln 474 111. Dieczsch kiln 475 112. Section of Schofer kiln 476 113. Hauenschild kiln 477 114. Schwarz kiln 477 115. Exterior of rotary kiln 480 116. Plan and elevation of rotary kiln 481 117. Driving mechanism of rotary kiln 482 118. Details of 60-foot rotary kiln 483 119. Raw-material storage bin. ; . 484 120. Section of Edison 150-foot rotary kiln 485 121. Coal-burning arrangement for rotary kiln 487 122. " " " " " " 488 123. Kirkwood gas burner for rotary kiln 489 124. " ''.;." applied to kiln 490 125. Arrangement of kiln lining, wet process 493 126. " ' ' ' ' " , dry process 493 127. Coarse-toothed rolls for lump coal 514 128. Smooth rolls for coal crushing 515 129. Section of Smidth ball-mill 517 130. Smidth ball-mill, external view. . . 518 xxviii LIST OF ILLUSTRATIONS. PAGE 131. Tower ccwler, Buckhorn P, C. Co 528 132. Atlas two-stage rotary cooler 530 133. Effect of plaster on strength of Portland cement 536 134. " " " " " " " " 538 135. " " " " setting time of Portland cement 539 136 tf tl lt " " " " " " 540 137. " " " " " " " " " 542 138. " " " " strength of Portlayfd cement 543 139. Concrete-steel bins, Illinois Steel Co. Opp. 548 140. Plan and section of stock house, Hudson 1 ?. C. Co 549 141. Foundation plan " " " " " } 551 142. Plan of bins in stock house, Hudson P. C. Co ." 551 1 43. Effect of fineness on tensile strength 584 144. Effect on strength of regrinding cement 585 145. Effect of temperature on setting-time 586 146. Tensile strength of various classes of cements 587 147. Effect of sand on tensile strength 588 148. Effect of character of sand on tensile strength 589 149. Ratio of compressive tensile strength. 590 150. Strength of sand-cement mortar 593 151. " " " " " 595 152. " " " " " 595 153. Effect of salt, etc., on freezing-point 599 154. " " " " " mortar 599 155. " " " li " " 600 156. " " " " lt " 600 157. " " " " " " 601 158. " " " " " compressive strength 601 159. Effect of granulating slag, compression 647 160. " " " ".tension 647 161. Ruggles-Coles drier , 650 162. Vitry slag-drier 652 163. Elevation and plan of Stewart slag-cement plant 660 164. Effect of hardening in air and water, tensile strength, slag cement 669 165. " " " *<*< tt ^ com p r essive strength, slag cement.. 669 LIST OF TABLES. PAGE 1. Production of structural materials, 1902-1903 1 2. Value of cementing materials produced in the United States, 1900-1903. . . 2 3. Symbols and atomic weights of elements H 4. Names and symbols of principal compounds 12 5. Sizes, capacity, etc., of Stedman disintegrators 36 6. Temperatures in cement-plaster manufacture 44 7. Sizes, capacity, etc., of Broughton mixers 50 8. Analyses of rock gypsum used for plaster 53 9. Analyses of gypsite (gypsum earth) used for plaster 54 10. -Analyses of plaster of Paris 57 11. Analyses of cement-plasters 57 12. Fineness of calcined plasters 58 13. Fineness of plasters tested 59 14. Tensile strength of plasters: effect of sand 59 15. Tensile strength of plasters 61 16. Effect of sand on compressive strength of plasters 63 17. Adhesive strength of plasters 63 18. Effect of retarders on strength of plasters 65 19. Effect of various retarders on rate of set of plasters 66 20. Effect of accelerators on rate of set of plasters 67 21. Tensile strength of Keene's cement 78 22. Tensile strength of American Keene's cement 78 23. Total imports and production of gypsum in U. S., 1900-1903 79 24. Uses of gypsum, 1900-1903 79 25. Gypsum production of the U. S., by uses, 1890-1903 80 26. Gypsum production of the U. S., by States, 1902-1903 82 27. Production and volume of gypsum in U. S., by States, 1890-1901 83 28. Imports of gypsum and plaster, 1900-1903 85 29. Imports of gypsum and plaster, by countries, 1900-1903 86 30. World's production of gypsum, 1893-1903 87 31. Analyses of various molluscan shells 95 32. Analyses of oyster-shells and oyster-shell lime 95 33. Heat and fuel required in burning lime 99 34. Sizes, capacity, etc., of Keystone lime kilns 106 xxix xxx LIST OF TABLES. PAGE 35. Cost of lime manufacture during 1900 in ten States Ill 36. Elements of cost of lime manufacture, in percentages 112 37. Lime production of the U. S., by States, 1902-1903 113 38. Analyses of high-calcium limes 116 39. Analyses of lean limes 118 40. Analyses of magnesian limes 119 41. Tensile strength of magnesium and high-calcium limes 122 42. Tensile strength of lime mortars. "? 123 43.' Sizes, capacity, etc., of Sturtevant crushers 125 44. Capacity, power, etc., of Campbell lime-hydrater 126 45. Sizes, etc., of Jeffrey separators 126 46. Tensile strength of hydrated lime 128 47. Percentage composition of various lime silicates 133 48. Effect of fineness of sand on lime-sand brick 135 49. Comparative tests of magnesian and high-calcium lime bricks 136 50. Effect of percentage of lime on lime-sand brick 138 51. Effect of hardening methods on lime-sand brick 139 52. Comparison of lime-sand bricks and natural sandstone '. 142 53. Physical tests of lime-sand brick 143 54. " " " " " " 143 55. Compressive strength of lirrie-sand brick 144 56. Physical tests of lime-sand brick 144 57. Summary of tests of lime-sand brick 145 58. Summary of tests of clay brick 145 59. Summary of tests of natural sandstones 146 60. Analyses of magnesite from Quebec, Canada 152 61. Production of magnesite in U. S., 1891-1903 153 62. Imports of magnesia and magnesite, 1903 153 63. Analyses of magnesites 153 64. Analyses of calcined magnesite ( = magnesia) 155 65. Analyses of highly magnesian limestones, U. S 157 66. Analyses of magnesia bricks 161 67. Expansion of magnesia bricks on heating 162 68. Compressive strength of Sorel stone 166 69. Composition of ideal hydraulic limestone and hydraulic lime 175 70. Analyses of hydraulic limestones of Teil, France 176 71. Analyses of hydraulic limestones: France and Germany 176 72. Analyses of beds in hydraulic-lime quarries at Malain, France 177 73. Analyses of hydraulic lime before slaking (Teil, France) 179 74. Analyses of hydraulic limes: France, Germany, and England 179 75. Analyses of kiln products, Teil, France 182 76. Analyses of hydraulic limes, after slaking 182 77. Average strength of hydraulic limes 183 78. Tensile strength of hydraulic-lime mortar 183 79. Compressive strength of hydraulic-lime mortar 184 80. Analyses of grappier cements 185 81. Tensile strength of Lafarge grappier cement 186 82. Analyses of feebly hydraulic limestones . 187" LIST OF TABLES xxxi 83. Analyses of feebly hydraulic limes 188> 84. Tensile strength of feebly hydraulic limes 188 85. Compressive strength of feebly hydraulic limes 190- 86. Tensile strength of selenitic limes 191 87. Compressive strength of selenitic limes , 192 88. Analyses of natural-cement rock, Utica, 111 204 89. "" " " " ", Louisville district 204 90. " " " " " , Fort Scott, Kansas 205 91. " " " " ", Cumberland-Hancock, Md 206 92. " " " " " , Mankato, Minn 206 93. " " " " ", Rosendale district, N. Y 208 94. " " " " " , Schoharie Co., N. Y 208 95. ..." " " " " , Central N. Y 209 96. " " " " " , Akron-Buffalo, N. Y 210 97. " " " " " , North Dakota 211 98. " " (t " ",Ohio 211 99. " " " " " , Lehigh district, Pa 212 100. " " " " " , Virginia 212 101. " " " " ", Milwaukee district, Wis 213 102. " " " " " , England 217 103. Fuel consumption in American natural-cement plants 234 104. Power required in grinding natural cement 246 105. Fineness required by various specifications 247 106. ' ' of three brands natural cement 247 107. Estimate of cost of natural-cement manufacture 249 108. Daily cost report of natural-cement plant 250 109. Analyses of natural cements, Ga 253- 110. " " " " , Utica, 111 253 111. " " " " , Louisville district 253 112. " " " " , Ft. Scott, Kans 254 113. " " t{ " , Potomac district, Md 255 114. " " " " , Minnesota 255- 115. " " " " , Rosendale district, N. Y 256 116. " " " " , Central N. Y 258 117. " " " " , Akron-Buffalo district, N. Y 258- 118. " " " " , North Dakota 259 119. " " " " , Lehigh district, Pa 259 120. " " " " , Shepherdst'n-Antietam district, W. Va.-Md . 260 121. " " tl " , Milwaukee district, Wis 260 122. " " natural Portland cements, Belgium 261 123. " " ." cements, England 261 124. " " " " , France : 261 125. " " " " , Germany and Austria 262 126. Specific gravity of American natural cements 263 127. Effect of aeration on setting time of natural cement 264 128. " " plaster " " " " " " t 264 129. " " " " tensile strength of natural cement 265 130. " " sand " " " " " " . 273- xxxii LIST OF TABLES. 131. Compressive tests of natural cements 274 132. ' ' strength of 4-inch natural-cement cubes 275 133. Effect of heating on compressive strength 275 134. Relation of tensile to compressive strength of natural-cement 276 135. " " " " " " " Utica natural cement 276 136. Modulus of elasticity 277 137. Fineness required by various specifications 282 138. Strength " " "~ " -' 283 139. Cement supplied to N. Y. Rapid-transit subway, 1900-1901 283 140. Summary of natural-cement tests, 1897-1900 284 141. Dates of establishment of natural -cement industries in various States. . . . 288 142. Production of natural cement in 1901, 1902, and 1903, by States. 290 143. Total production of various cements in U. S., 1818-1904 291 144. Character of Portland-cement materials 301 145. Production of Portland cement from various materials 304 146. Analyses of hard limestones used in American cement-plants 314 147. " "purechalks " " " " " 321 148. " "clayey " " " " " " 321 149: " " Hudson shale and slate in Pa. and N. J 325 150. " " Trenton limestones 326 151. " " Kittatinny magnesian limestone. . 327 152. " " "cement-rock" 329 153. " " pure limestone used for mixing with cement rock 329 154. Portland-cement production of the Lehigh district, 1890-1902 332 155. Analyses of "cement-rock" materials from the Western U. S 333 156. Fineness of crude marl 340 157. Analyses of marls used in American cement-plants 342 158. " ' ' alkali waste, ammonia process 349 159. ' ' " iron-furnace slags 351 160. " " normal clays used in American cement-plants 355 161. " "limey " " " " " " 353 162. " " normal shales " " " " " 357 163. " "limey " M " " " " \ 359 . 164. " " American roofing slates 364 165. Composition of American roofing slates 364 166. Analyses of slate used for Portland cement, Rockmart, Ga 365 167. Detailed costs of steam-shovel work 373 168. Effect of alumina 388 169. Analyses of raw materials containing phosphoric acid 390 170. Tests of cements containing phosphoric acid 390 171. Composition of actual mixes 394 172. Analyses of fuel ash 398 173. Change in composition during burning. . . . 397 174. Cement mixture and cement, Sandusky 397 175. Fineness of raw mix at various plants 425 176. Sizes, power, etc., of Gates crushers 433 177. Estimated cost of crushing by Gates crusher 433 178. Sizes, power, etc., of Griffin mill . 442 LIST OF TABLES. xxxiii PAGE 179. Siaps, weights, etc., of Gates ball mill. ..,..., 454 180. Analyses of flint pebbles 465 181. " " high-alumina clays used for kiln brick. 491 182. " " " " fire-brick for kilns 491 183. " " low-alumina clays used for kiln brick 492 184. " " low-alumina brick, furnished as kiln brick 492 185. Actual output and fuel consumption at various plants 495 183. Average " " " " " Portland-cement plants 495 187. Heat used in evaporation of water . . 498 188. ll " " dissociation of carbonates per bbl. cement. 500 189. Theoretical heat requirements in B.T.U. per bbl 501 190. Utilization and losses of heat in rotary kilns ....... 502 191. Newberry's estimates on heat distribution in kilns 505 192. Summary of Richards" tests of rotary kilns 507 193. Tests and estimates of heat distribution, B.T.U. per bbl 509 194. Analyses of kiln coals 513 195. " ' l natural gas, Kansas. 523 196. Thermal values of natural gas 523 197. Effect of form of sulphate used 537 198. " " adding various percentages of calcined plasters 540 199. " " calcined sulphate on set of cement 541 200. " " gypsum on setting-time ;', 541 201. " " calcium sulphate on strength of cement. 543 202 tt ttn ti _ 544 203. " " treatment with anhydrous calcium sulphate 544 204. " ' " " li crude gypsum 544 205. ' ' " " " plaster of Paris 544 206. Analysis of gypsum used in cement-plants 545 207. ' ' " calcined plaster used in cement-plants 545 208. Effect of various salts on set of cement 546 209. Capacity of Portland-cement barrels, and weight of contents 552-553 210. Labor costs in cement-plant ' 558 211. " and output in typical plants 559 212. Costs of cement manufacture at Atlas plant. 559 213. Detailed estimates of costs of cement manufacture 560 214. Approximate cost of two-kiln plant 560 215. Estimates of average cost of Portland-cement manufacture 561 216. Total production of Portland cement in U. S 562 217. Production of P. C. in the U. S. in 1901, 1902, and 1903, by States 563 218. Portland-cement production of Lehigh district, 1890-1903 564 219. Imports of hydraulic cement into the U. S., 1899-1903, by countries 564 220. Analyses of Portland cement, 1849-1873 575 221. " " American Portland cement 577 222. Fineness of various American Portlands 585 223. Compressive strength of Portland-cement cubes 589 224 " f( ll " mortar and concrete cubes 590 225. Relation of tensile to compressive strength 591 226. Modulus of elasticity of Portland cement 592 xxxiv LIST OF TABLES. PAGE 227. Comparative tests of Portland cement and sand-cement. ............ ,T . 594 228. Tensile and compressive strength of ' ' 594 229. Compressive strength of silica-cement cubes 590 230. ' ' sand-cement mortars . . . . 597 231. Modulus of elasticity of sand cement .597 232. Effect of heating on compressive strength 598 233. " " alumina 603 234. Analyses of pozzuolana from-Italy. . .. . .^ 633 235. ' ' ' ' ' ' li France. . . C3 1 236. " : " " " the Azores '..,'.... 635 237. " " trass and related materials from Germany. .... ......... 636 238. ' ' santorin ash, from Santorin. . K . i ............ 636 239. " " arSnes, France . . . 637 240. Strength of basaltic dust :."....' 637 241. Average analyses of natural pozzuolanic materials. . ......,..;. ^. ...... 638 242. Strength of lime burnt-clay mortars 639 243. Analyses of slags used for slag-cement 643 244. Strength of granulated and ungranulated slag. . .v. ...... 648 245. Working results of Ruggles-Coles drier 651 246. Analyses of limes used in American slag-cement plants 1 .......... 654 247. Costs of slag-cement manufacture, per barrel. ............ .s, ....... 664 248. Slag-cement production in U. S., 1899-1904. -. .......... 664 249. Analyses of American slag-cements . ...... 667 250. " European " " .......... .668 251. Tensile and compressive strength of slag-cements. ................. 671 252. Crushing strength of indurated slag-bricks 683 253. Porosity of slag-bricks 683 254. Anatyses of slag used for slag-blocks .**;, 689 CEMENTS, LIMES, AND PLASTERS. INTRODUCTION. FEW economic movements, during the past decade, have equaled in importance or interest the marvelous growth of the industries based on the non-metallic structural materials. In this group are included the stone used for building purposes, such clay products as are used in engineering, and, last in order but not in importance, the cementing materials. The production of structural materials in the United States for the years 1902-1903 has been summarized in the following table, for convenience of reference. TABLE 1. PRODUCTION OF STRUCTURAL MATERIALS, 1902-1903. 1902. 1903. Building stone $36 486 392 $36 052 376 Brick, tile, etc 98 042 078 105 526 596 Cementing materials 36,849,943 45 607 436 Total, structural materials. . $171,378,413 $192,186,408 Of the three subgroups of materials quoted in the above table, the third cementing materials is the subject of the present volume. It will therefore be of advantage to consider the financial statistics of this subgroup in somewhat more detail. The following table (Table 2) gives the value of the production in the United States, for the years 1900-1903 inclusive, of the six classes into which the cementing materials may conveniently be divided. On later pages, where these classes are separately discussed, much more detailed statistics are given in regard to eacli product. CEMENTS, LIMES, AND PLASTERS. TABLE 2. VALUE OF CEMENTING MATERIALS PRODUCED IN THE UNITED STATES, 1900-190$ 1900. 1901. 1902. 1903. Plaster of Paris, wall-plaster, etc $1,500,270 $1,325,517 $1,889,190 $3,550,390 Lime 6,970,062 8,597,030 9,573,011 10,105 190 Misrnesia and magnesite . 19,333 43,057 21 362 20 515 Natural cement . 3,728,848 ''3,056,278 4,076,630 3 675 520 Portland cement 9,280,525 12,532,360 20,864,078 27 713 319 Slag cement > 274,208 198,151 425,672 542 502 Total, cementing materials ?21, 773,246 $25,752,393 $36,849,943 $45,607,436 It will be seen that the value of the entire group has more than doubled in these four years. Examination of the figures for each sub- group will show that this increase was by no means equally distributed. The production of Portland cement almost tripled in value from 1900 to 1903; the values of the plaster and slag-cement production about doubled; while lime increased only 50 per cent, and magnesia and the natural cements were practically stationary. The great advance in value of the cementing materials, taken as a group, has been steady and natural. Similar increases can therefore be reasonably expected to occur in future years. Individual years, however, may give less encouraging results; because any general business depres- sion reacts sharply upon building industries first of all. Classification and Relationships of Cementing Materials. It seems desirab ] e, before taking up the various classes of cement materials individually, to devote an introductory section to the con- sideration of the entire group of cementing materials, and to attempt to indicate briefly the relationships that exist between the different classes which compose this group. These relationships, as regards both resemblances and differences, seem to be best brought out by the scheme of classification presented below. This classification was fiist published by the writer in 1902,* in a form differing but slightly from that here given. It is based primarily upon the amount of chemical change caused by the processes of manufacture and use; and secondarily upon the chemical composi- tion of the cementing material after setting. As regard is paid to both technologic and commercial considerations, it would seem to furnish a fairly satisfactory working classification. * Eckel, E. C. The classification of the crystalline cements. American Geol- ogist, vol. 29, pp. 146-154. March, 1902. INTRODUCTION. 3 CLASSIFICATION OF CEMENTING MATERIALS. GROUP I. SIMPLE CEMENTING MATERIALS: including all those cementing ma- terials which are produced by the expulsion of a liquid or gas, through the action of heat, from a natural raw material, and whose setting properties are due to the simple reabsorption of the same liquid or gas, and the reassumption of original composition; the set cement being, therefore, similar in chemical composition to the raw material from which it was derived. SUBGROUP I a. HYDRATE CEMENTING MATERIALS OR PLASTERS: manufac- tured by driving off water from gypsum; setting properties due to the reabsorption of water. SUBGROUP I b. CARBONATE CEMENTING MATERIALS OR LIMES AND MAGNESIA: manufactured by driving off carbon dioxide from limestone or mag- nesite; setting properties due to the reabsorption of carbon dioxide. GROUP II. COMPLEX CEMENTING MATERIALS: including all those cementing materials whose setting properties are due to the formation of entirely new chemical compounds during manufacture or use; the set cement being, therefore, different in chemical composition from the raw mate- rial or mixture of raw materials from which it was derived. SUBGROUP II a. SILICATE CEMENTING MATERIALS OR HYDRAULIC CEMENTS : setting properties due entirely or largely to the formation of silicates during the processes of manufacture or use. SUBGROUP II b. OXYCHLORIDE CEMENTING MATERIALS : setting properties due to the formation of oxychlorides. The various groups and subgroups above noted will now be taken up separately and briefly described, in order that the principles on which the classification is based may be clearly understood. Group I. Simpla Cementing Materials. The products included in the present group include those known as " plasters", "hard-finish cements", "limes", and "magnesia." The material from which the "plasters" and "hard-finish cements" are derived is gypsum, a hydrous calcium sulphate; while the limes are derived from limestone, which is essentially calcium carbonate, though usually accompanied by greater or less amounts of magnesium car- bonate; and magnesia is derived from more or less pure magnesite, a natural magnesium carbonate. On heating gypsum to a certain temperature, the raw material parts readily with much of its water, leaving an almost anhydrous calcium sulphate, known commercially as plaster of Paris. On exposing this plaster to water, it rehydrates, and again takes the composition of the gypsum from which it was derived. 4 CEMENTS, LIMES, AND PLASTERS. In like manner limestone, on being sufficiently heated, gives off Its carbon dioxide, leaving calcium oxide, or "quicklime". This, on exposure to moisture and air carrying carbon dioxide, reabsorbs carbon dioxide and reassumes its original composition calcium carbonate. Magnesite, on being heated, loses its carbon dioxide, leaving magnesium oxide, or magnesia. This, under normal conditions of burning, will reabsorb carbon dioxide, and set in its original form as magnesium carbonate. i* The cementing materials included in this group, therefore, while differing in composition and properties, agree in certain important points. They are all manufactured by heating a natural raw material sufficiently to remove much or all of its water or carbon dioxide; and, in all, the setting properties of the cementing material are due to the fact that, on exposure to the water or carbon dioxide which has thus been driven off, the cement reabsorbs the previously expelled liquid or gas, and reassumes the chemical composition of the raw material from which it was derived. Plaster of Paris, after setting, is not chemically different from the gypsum from which it was derived; while if the sand, added simply to avoid shrinkage, be disregarded, a thoroughly hardened lime mortar is nothing more or less than an artificial lime- stone. The principal points of difference between the two subgroups the plasters and the limes may be briefly noted as follows : Subgroup I a. Hydrate cementing material : plasters. The materials here included are known in commerce as " plaster of Paris", " cement plaster", " Keene's cement", " Parian cement", etc. All of these hydrate cements or plasters are based upon one raw material gypsum. The partial dehydration of pure gypsum produces plaster of Paris. By the addition to gypsum, either by nature or during manufacture, of relatively small amounts of other materials, or by slight variations in the processes of manufacture, the time of setting, hardness, and other important technical properties of the resulting plaster can be changed to a degree sufficient to warrant separate naming and descriptions of the products. Both the technology and the chemistry of the processes involved in the manufacture of the hydrate cements are simple. The mineral gypsum, when pure, is a hydrous sulphate of lime, of the formula CaSO4,2H 2 O, corresponding to the composition: calcium sulphate 79.1%, water 20.9%. As noted later (under the head of Cement Plasters) gyp- sum, as mined, rarely even approximates to this ideal composition, its im- purities often amounting to 25% or even more. These impurities, chiefly INTRODUCTION. * 5 clayey materials and fragments of quartz and limestone, often exercise an appreciable effect upon the properties of the plaster resulting from burning such impure gypsum. On burning gypsum at a relatively low temperature (350-400 F.) much of its water of combination is driven off, leaving a partially dehydrated calcium sulphate. This, when ground, is plaster of Paris, or if it either naturally or artificially contains certain impurities, it is called "cement plaster". When either plaster of Paris or cement plaster is mixed with water, the percentage of water which was driven off during calcination is reabsorbed, and the mixture hardens, having again become a hydrous sulphate 'of lime. The processes involved in the manufacture and setting of the dead- burned plasters and hard-finish plasters are slightly more complicated, but the reactions involved are of the same general type.' Subgroup I b. Carbonate cementing materials : limes and magnesia. The cementing materials falling in the present subgroup are oxides derived from natural carbonates by the application of heat. On exposure, under proper conditions, to any source of carbon dioxide, the cement- ing material recarbonates and "sets". In practice the carbon dioxide required for setting is obtained simply by exposure of the mortar to the air. In consequence the set of these carbonate cements, as com- monly used, is very slow (owing to the. small amount of carbon dioxide which can be taken up from ordinary air) ; and, what is more important from an engineering point of view, none of the mortar in the interior of a wall ever acquires hardness, as only the exposed portions have an opportunity to absorb,, carbon dioxide. From the examination of old mortars it has been thought probable that a certain amount of chemical action takes place between the sand and the lime, resulting in the forma- tion of lime silicates ; but this effect is slight and of little engineering importance compared with the hardening which occurs in consequence of the reabsorption of carbon dioxide from the air. Limestone is the natural ? raw material whose calcination furnishes most of the cementing materials of this group.* If the limestone be an almost pure calcium carbonate, it will, on calcination, yield calcium oxide, or "quicklime". If, however, the limestone should contain any appreciable percentage of magnesium carbonate, the product will be a mixture of the oxides of calcium and magnesium, commercially known as a magnesian lime. A brief sketch of the mineralogic re- lationships of the various kinds of limestone, in connection with the * The subject of magnesia as a cementing material, being too complicated for brief discussion, will not be taken, up here. See Chapters XI and XII. 6 CEMENTS, LIMES, AND PLASTERS. chemistry of lime-burning, will be of service at this point of the dis- cussion. Pure limestone has the composition of the mineral calcite, whose formula is CaCO 3 , corresponding to the composition: calcium oxide 56%, carbon dioxide 44%. In the magnesian limestones part of this calcium carbonate is replaced by magnesium carbonate, the resulting rock there- fore having a formula of the type XCaC0 3 ,YMgCO 3 . This replacement may reach the point at which the rock ffas the composition of the mineral dolomite an equal mixture of the two carbonates, with the formula CaC0 3 ,MgCO 3 , corresponding to the composition: calcium oxide 30.44%, magnesium oxide 21.73%, carbon dioxide 47.83% /' Limestones may therefore occur with any intermediate amount of magnesium carbonate, and the lime which they produce on calcination will carry correspond- ing percentages of magnesium oxide, from 0% to 41.65%. Commer- cially those limes which carry less than 10% of magnesium oxide are, for building purposes, marketable as "pure limes"; while those carry- ing more than that percentage will show sufficiently different properties to necessitate being marketed as "magnesian limes". Aside from the question of magnesia, a limestone may contain a greater or lesser amount oforeal impurities. Of these the most impor- tant are silica (SiC>2) ," alumina (A1 2 3 ). and iron oxide (Fe 2 O 3 ). These impurities, if present in sufficient quantity, will materially affect the properties of the lime produced, as will be noted later under the heads of Hydraulic Limes and Natural Cements. The limes may be divided into two classes: (1) High-calcium limes; (2) Magnesian limes. High-calcium limes. On heating a relatively pure carbonate of lime to a sufficiently high degree, its carbon dioxide is driven off, leaving calcium oxide (CaO), or "quicklime". Under ordinary conditions, the expulsion of the carbon dioxide is not perfectly effected until a temperature of 925 C. is reached. The process is greatly facilitated by blowing air through the kiln, or by the injection of steam. On treating quicklime with water, "slaking" occurs, -'heat being given off, and the hydrated calcium oxide (CaH 2 O2) being formed. The hydrated oxide (slaked lime) will, upon exposure to the atmosphere, slowly reabsorb sufficient carbon dioxide to reassume its original composition as lime car- bonate. As this reabsorption can take place only at points where the mortar is exposed to the air, the material in the middle of thick walls never becomes recarbonated. In order to counteract the shrinkage which would otherwise take place during the drying of the mortar, sand is INTRODUCTION. 7 invariably added in the preparation of lime mortars, and, as noted above, it is possible that certain reactions take place between the lime and the sand. Such reactions, however, though possibly contributing some- what to the hardness of old mortars, are only incidental and subsidiary to the principal cause of setting recarbonation. The presence of impurities in the original limestone affects the character and value of the lime produced. Of these impurities, the presence of silica and alumina in sufficient quantities will give hydraulic properties to the resulting limes; such materials will be discussed in the next group as Hydraulic Limes and Natural Cements. Magnesian limes. The presence of any considerable amount of magnesium carbonate in the limestone from which a lime is obtained has a noticeable effect upon the character of the product. If burned at the temperature usual for a pure limestone, magnesian limestones give a lime which slakes slowly without evolving much heat, expands less in slaking, and sets more rapidly, than pure lime. To this class belong the well-known and much-used limes of Canaan (Conn, ) ; Tuck- ahoe, Pleasantville, and Ossining (N. Y.) ; various localities in New Jersey and Ohio; and Cedar Hollow (Penn.). Under certain conditions of burning, pure magnesian limestones yield hydraulic products, but in this case, as in the case of the product obtained by burning pure mag- nesite, the set seems to be due in part at least to the formation of a hydroxide rather than of a carbonate. Magnesian limestones carrying sufficient silica and alumina will give, on burning, a hydraulic cement falling in the next group under the head of Natural Cements. Group II. Complex Cementing Materials. The cementing materials grouped here as Complex Cements include all those materials whose setting properties are due to the formation of new compounds, during manufacture or use, and not to the mere reas- sumption of the original composition of the material from which the cement was made. These new compounds may be formed either by chemical change during manufacture or by chemical interaction, in use, of materials which have merely been mechanically mixed during manyfacture. Subgroup II a. Silicate cementing materials : hydraulic cements. In the class of silicate cements are included all the materials commonly known as cements by the engineer (natural cements, Portland cement, puzzolan cements), together with the hydraulic limes. Though differing widely in raw materials, methods of manufacture, CEMENTS, LIMES, AND PLASTERS. and properties, the silicate cements agree in two prominent features: they are all hydraulic (though in very different degrees); and this property of hydraulicity is, in all, due largely or entirely to the formation of tricalcic silicate (3CaO,SiC>2). Other silicates of lime, as well as silico aluminates, may also be formed; but they are relatively unim- portant, except in certain of the natural dements and hydraulic limes, where the lime aluminates may be of greater importance than is here indicated. This will be recurred tortn discussing the groups named. The silicate cements are divisible, on technologic grounds, into four distinct classes. The basis for this division is given below. It will be seen that the last named of these classes (ttte puzzolan cements) differs from the other three very markedly, inasmuch as its raw materials are not calcined after mixture; while in the last three classes the raw materials are invariably calcined after mixture. The four classes differ somewhat in composition, but more markedly in methods of manu- facture and in the properties of the finished cements. CLASSES OF HYDRAULIC CEMENTS. 1. Hydraulic limes are produced by burning, at relatively low tem- peratures, a natural siliceous limestone which carries so much lime car- bonate, compared to its content of silica and alumina, that the burned product will contain a considerable amount of free lime (CaO) in addi- tion to the silicates and aluminates of lime that have been formed. In consequence of the relatively large percentage of free lime, the burned masses will slake if water be poured on them, but because much of them is composed of silicates and aluminates of lime, they will slake very slowly and their final product will have hydraulic properties. The hydraulic limes are thus intermediate in composition, methods of manufacture, and properties between the true limes and the cements proper. 2. Natural cements are produced by burning a natural clayey lime- stone, containing from 15 to 40 per cent of silica, alumina, and iron oxide. This burning takes place at a temperature that is usually little if any above that of an orolinary lime-kiln. During the burning the carbon dioxide of the limestone is almost entirely driven off, and the lime combines with the silica, alumina, and iron oxide, forming a mass containing silicates, aluminates, and ferrites of lime. In case the original limestone contained much magnesium carbonate, the burned rock will contain a corresponding amount of magnesia and magnesian compounds. After burning, the burned mass will not slake if water be poured on it. It is necessary, therefore, to grind it quite fine, after which, if INTRODUCTION. 9 the resulting powder (natural cement) be mixed with water, it will harden rapidly. This hardening or setting 'will also take place under water. 3. Portland cement is produced by burning a finely ground artificial mixture containing essentially lime, silica, alumina, and iron oxide, in certain definite proportions. Usually this combination is made by mixing limestone, chalk, or marl with clay or shale, in which case about three times as much of lime carbonate as of clayey materials should be present in the mixture. The burnirfg of this mixture takes place at a high temperature, approaching 3000 F., and must therefore be carried on in kilns of special design and lining. During the burning, com- bination of the lime with the silica, alumina, and iron oxide takes place. The product of the burning is a semifused mass called " clinker", and consists of silicates, aluminates, and ferrites of lime, each being present in certain fairly definite proportions. This "clinker" must be finely ground. After such grinding the resulting powder, which is then Port- land cement, will harden under water. Portland cement is blue to gray in color, with a specific gravity of 3.0 to 3.2. It sets more slowly than natural cements, but soon attains a higher tensile strength. 4. Puzzclan cements. Certain natural and artificial products, such as some volcanic ashes and blast-furnace slags, will show feeble hydraulic properties if finely pulverized. Their hydraulicity is very markedly increased if their powder, instead of being used alone, is mixed with powdered slaked lime. Such mixtures of slaked lime with a relatively feeble hydraulic agent are known as Puzzolan Cements. The puzzolan cements are therefore simply mechanical mixtures of the two ingredients, as the mixture is not burned at any stage of the process. Puzzolan cements are usually light bluish in color, and of lower specific gravity and less tensile strength than Portland cement. They are better adapted to use under water than to use in air, as is explained later in this volume Subgroup II b. Oxychloride cements. In 1853 the chemist Sorel discovered the fact that zinc chloride mixed with zinc oxide united with it to form a very hard cement. Later it was found that a solution of magnesium chloride will unite, in the same manner, with magnesia. The product is, in both cases, an oxy chloride of zinc or magnesium as the case may be. It is obvious that cementing materials of this character cannot well be placed on the market as structural cements, but the property 10 CEMENTS, LIMES, AND PLASTERS. above noted has been taken advantage of in the manufacture of a number of patented artificial stones, the exact methods followed depending upon the particular process that is used. The best known of these artificial stones is probably "Sorel-stone", which is frequently alluded to in engineering text-books. As the subject of oxy chloride cements is intimately connected with that of magnesia cements, the two will be discussed together in Chapters XI to XII. The following table shows the Delation of the various cementing materials which have been briefly discussed in this chapter. For con- venience of reference, the pages on which the different products are described in detail have been added. f Plaster of Paris 1 31-67 Hydrate cements or I Cement plasters J ' plasters ) Hard-finish plasters 76-78 Simple cements. .v'J l Dead-burnt plasters. . ; 68-76 Carbonate cements or limes Complex cements . Silicate cements. [ Magnesia 148-167 { Hydraulic lime 168-193 ] Natural cements 194-293 I Portland cement 294-631 [ Puzzolan cements 632-689 Oxychloride cements. { Sorel-stone, etc 162-167 Chemical and Physical Data Employed in the Discussion of Cementing Materials. In the course of the discussion it will frequently be necessary to use certain chemical and physical data, such as atomic weights, values of various heat-units, etc. These are to be found in most engineering pocket-books, but for convenience of reference the more important data have been placed in the present chapter. Atomic weights of elements. In the following table (Table 3) are given the names, symbols, and atomic weights of all the chemical ele- ments.* The first column of this table contains the name of each element; the second, its chemical symbol or abbreviation; the third, its atomic weight calculated on the basis of the atomic weight of oxygen being 16; the fourth, its atomic weight on the basis of the atomic weight of hydrogen being 1. In use, it is a matter of absolute indifference whether the figures in * Through the courtesy of Prof. F. W. Clarke the writer is enabled to present this table, which contains the corrections for 1905. INTRODUCTION. 11 the third column or those in the fourth are employed: but of course the selection should be consistent throughout. Chemists are about equally divided as to the use of the two series. For many purposes the O = 16 basis is the more convenient, as it gives even figures for several common elements. Such elements as enter into the calculation of cement, plaster, lime, fuel, or slag are given in black-faced type for convenience of reference. TABLE 3. ATOMIC WEIGHTS OF ELEMENTS. Name of Element. Sym- bol. Atomic Weight. Name of Element. Sym- bol. Atomic Weight. O = 16. H = l. O = 16. H = l. Aluminum Al Sb A As Ba Bi B Br Cd Cs Ca C Ce Cl Cr Co Cb Cu Er F Gd Ga Ge. . Gl Au He H In I Ir Fe Kr La Pb Li Mg Mn Hg Mo 27.1 120.2 39.9 75.0 137.4 208.5 11.0 79.96 112.4 132.9 40.1 12. O 140.25 35.45 52.1 59.0 94.0 63.6 166.0 19.0 156.0 70.0 72.5 9.1 197 = 2 4.0 1.008 115.0 126.97 193.0 55-9 81.8 138.9 206.9 7.03 24.36 55-0 200.0 96.0 26.9 119.3 39.6 74.4 136.4 206.9 10.9 79.36 111.6 131.9 39-7 ii .91 139.2 35.18 51.7 58.57 93.3 63.1 164.7 18.9 154.8 69.5 72.0 9.03 195.7 4.0 I. 000 114.1 126.01 191.5 55-5 81.2 137.9 205.35 6.98 24.18 54-6 198.5 95.3 Neodymium. . . . Neon . Nd Ne. Ni N Os Pd P Pt K Pr Ra Rh Rb Ru Sm Sc Se Si Ag Na Sr S Ta Te Tb Tl Th Tm Sn Ti W u V Xe Yb Yt Zn Zr 143.6 20.0 58.7 14.04 191.0 16.00 106.5 31-0 194.8 39-iS 140.5 225.0 103.0 85.5 101.7 150.3 44.1 79.2 28.4 107.93 23 05 87.6 32.06 183.0 127.6 160.0 204.1 232.5 171.0 119.0 48.1 184.0 238.5 51.2 128.0 173.0 89.0 65.4 90.6 142.5 19.9 58.3 13 93 189.6 15-88 105.7 30.77 193.3 38.85 139.4 223.3 102.2 84.9 100.9 149.2 43.8 78.6 28.2 107.1 22.88 86.94 31.82 181.6 126.6 158.8 202.6 230.8 169.7 118.1 47-7 182.6 236.7 50.8 127.0 171.7 88.3 64.9 89.9 Arsron Nickel Arsenic Nitrogen Barium. ... .... Osmium Bismuth. Oxygen Boron Palladium Phosphorus. . . . Bromine Cadmium . . Platinum. Potassium. ..... Praseodymium. . Radium Cspsium o . . Calcium .... Carbon Cerium ..... Rhodium Chlorine .... Rubidium Chromium Ruthenium. .... Samarium Cobalt Columb ium Scandium Selenium Silicon Silver Sodium Copper . Erbium Fluorine Gadolinium Gallium Strontium Sulphur Tantalum Germanium. .... Glucinum Gold Tellurium Helium Terbium Hvdrosfen. . Thallium uyuiugcii. ...... Indium. Iodine Iridium Thorium Thulium Tin Iron Titanium Tungsten Krypton Lanthanum Uranium Vanadium Lead Lithium Xenon Ytterbium Manganese Mercury Yttrium Zinc. Zirconium Molybdenum. . . . 12 CEMENTS, LIMES, AND PLASTERS. Chemical compounds. Certain chemical combinations will be mentioned very frequently in the discussion of cementing materials. The more important of these compounds are listed in the following table, with their symbols and also the name by which they are known when they occur as minerals or as commercial products, TABLE. 4. . V* NAMES AND SYMBOLS OF PRINCIPAL COMPOUNDS. Symbol. Name. Mineral or Commercial Name. CaO Calcium oxide Lime: quicklime CaCOg Calcium, carbonate Calcite 3CaO SiO . Tricalcium silicate CaCOg MgCOg Calcium-magnesium carbonate Dolomite CaH O 2 Lime hydrate Slaked lime CaSO 4 . . . Calcium sulphate Dead-burned plaster CaSO 4 + 2H 2 O Hydrous calcium sulphate Gypsum 2CaSO 4 + H 2 O Plaster of Paris MgO Magnesium oxide Magnesia MgCOo Magnesium carbonate Magnesite SiO Silica . . Quartz Al O, Alumina FeO ' Ferrous oxide* Fe 9 O, Ferric oxide Heat-units. Two heat-units are now in common use the British and the metric. The British thermal unit ( = B.T.U.) is the quantity of heat required to raise the temperature of 1 Ib. of water one degree Fahrenheit when at the temperature of maximum density ( = 39.1 F. or 4 C.). The metric unit ( = calorie) is the quantity of heat required to raise the temperature of 1 kilogram of water one degree centigrade when at the temperature of maximum density. From these definitions the two units may be converted according to the following equations : 1 B.T.U.= .252 calorie. 1 calorie = 3. 968 B.T.U. 1 calorie per kilogram =1.8 B.T.U per poundo Metric conversion tables. Since much of the literature of cementing materials is published in French and German, metric units are frequently employed. In the present volume such units have been converted into American units throughout, but for convenience a few conversion tables are here inserted to cover the more common cases. INTRODUCTION. LENGTH. * 1 inch = 2.54 centimeters. 1 centimeter = .3937 inch. 1 foot = . 3048 meter. 1 meter = 39 . 37 inches = 3. 2808 feet. SURFACE. 1 square inch = 6 . 452 square centimeters. 1 square centimeter = .155 square inch. 1 square foot = . 0929 square meter. 1 square meter = 10 . 764 square feet. VOLUME. 1 cubic inch = 16 . 387 cubic centimeters. 1 cubic centimeter = .061 cubic inch. 1 cubic foot = .02832 cubic meter. 1 cubic yard = . 7645 cubic meter. 1 cubic meter =35.314 cubic feet = 1 . 308 cubic yards. WEIGHT. 1 ounce avoirdupois = 28 . 35 grams. 1 pound = .4536 kilogram. 1 kilogram = 2 . 2046 pounds. CAPACITY. 1 cubic foot = 28. 317 liters. 1 liter =61.023 cubic inches = .03531 cubic foot. 1 gallon = 3. 785 liters. PRESSURE. 1 pound per square inch = . 070308 kilogram per square inch. 1 kilogram per square centimeter = 14 . 223 pounds per square inch. PAKT I. PLASTERS. CHAPTER I. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. THE mineral called gypsum is the raw material which serves as the basis for the manufacture of plaster of Paris, " cement plaster", anc the various related types of plasters. In the present chapter the com- position, properties, varieties, mode of occurrence, origin, and distribu- tion of gypsum will be described in the order named, after which th< methods and costs of quarrying and mining gypsum will be discussed Chemical composition. The mineral gypsum, when absolutely pure is a hydrous sulphate of lime, made up of one molecule of lime sulphate combined with two molecules of water. The chemical formula o gypsum is therefore CaSO 4 +2H 2 O. This, when reduced to percentages of weight, corresponds to the following: n /rt'orv .' i OTT r\\ / L i me sulphate (CaSO 4 ) ...... 79.1% Gypsum (CaSO 4 + 2H 2 O) = | Water ( ^ Q) ^ ..... ...... 2Q g /0 The 79.1% of lime sulphate can, in turn, be considered as being mad< up of 32.6% of lime (CaO), plus 46.5% of sulphur trioxide (S0 3 ) Reduced to its ultimate components, the composition of pure gypsun may therefore be represented as follows: [Lime (CaO) ............... 32.6% Gypsum (CaSO 4 + 2H 2 O) = ^ Sulphur trioxide (SO 3 ) ...... 46.5 [Water (H 2 0) .............. 20.9 100.0 Deposits of gypsum large enough to be worked for plaster are, how ever, rarely even approximately as pure as this. Gypsum as excavatec for a plaster-plant will usually carry varying and often high percentage of such impurities as clay, limestone, magnesian limestone, iron oxide 14 _400 600 CuO mile* FIG. i. [To face p 14. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. 15 etc. Table 8, on pp.ge 53, gives a number of analyses of the gypsum used at various plaster-plants; and a glance at this table will show the kind and amount of impurities which may be expected to occur in commercial gypsum. Varieties of gypsum. Owing to differences in form, texture, color, etc., gypsum presents several varieties, some of which have been given distinct names. The ordinary form in which gypsum occurs in the workable deposits is as massive or rock gypsum. Alabaster is a pure white, fine-grained massive gypsum, occasionally used for statuary, etc. The term selenite is applied to the crystalline, white, almost transparent gypsum which occurs frequently, but in relatively small quantity, scattered through a deposit of massive gypsum. Aside from these various forms of rock gypsum, two less massive forms of the mineral are to be noted as being of commercial importance. In certain Western States and Territories deposits of earthy gypsum, gyp- sum earth, or gypsite occur. These deposits contain an impure, earthy, granular form of gypsum. Deposits of gypsum sands are also found in the West, being dunes or heaps of fine grains of gypsum. Physical properties. Pure gypsum is white and, when in the crystal- line form, translucent. The impurities which it commonly contains usually destroy its translucency and. affect its color, so that the mineral as mined is an opaque, fine-grained mass, varying from white to reddish, gray, or brown in color. Gypsum can be distinguished from most other minerals by its extreme softness, for even when in the crystalline form it can be readily scratched by the finger-nail. When treated with acids it does not effervesce. On heating it loses its water of crystallization and, if previously trans- lucent, becomes a chalky, opaque white. Pure crystalline specimens have a specific gravity * of 2.30 to 2.33. Anhydrite. The mineral anhydrite is closely related to gypsum, as it is an anhydrous lime sulphate, with the formula CaSO4. It there- fore corresponds in composition to the product obtained by heating gypsum so strongly as to drive off all of its water of combination (see pages 31, 32). Anhydrite occurs, but in relatively small amounts, in almost all gypsum deposits. Pure specimens have a specific gravity* of 2.92 to 2.98. Occurrence and origin of gypsum deposits. Rock gypsum occurs in the form of beds, frequently closely associated with beds of rock salt, and almost always interstratified with thin beds of limestone and * Clarke, F. W. Constants of Nature, Part I, pp. 81, 82. 16 CEMENTS, LIMES, AND PLASTERS. thicker beds of red shales. Such gypsum beds may vary greatly in extent as well as in thickness. Beds now worked in different American localities, for example, vary from six to sixty feet in thickness. The gypsum occurring in the beds frequently contains a considerable percent- age of impurities, as is shown by the analyses given in Table 8, page 53. Deposits of rock gypsum have been formed by the gradual evapora- tion, in lake basins or shallow arms of. the sea, of waters carrying lime sulphate in solution. If any natural water be evaporated to a suf- ficient extent, it will deposit the salts which it contains, the order in which the various salts are deposited depending principally upon their relative proportions in the water and their solubility. A normal water, whether from stream, lake, or ocean, will carry as its three commonest constituents lime carbonate, lime sulphate, and sodium chloride. If such a water be evaporated, therefore, deposits of limestone, gypsum, and common salt would result : and, as above noted, these three minerals- are very common associates in gypsum deposits. Gypsum-earth deposits consist of masses of small crystals or grains of gypsum, intermingled usually with much clayey matter, sand, etc. Such deposits occur in depressions, and are supposed to be formed by the evaporation of sjpring-waters which have taken up lime sulphate in solution from underlying beds of rock gypsum, only to deposit it again on reaching the surface and being subjected to evaporation. In certain areas in the West, notably in Arizona and New Mexico, deposits of gypsum sand occur. These deposits are made up of fine grains of gypsum, worn off from outcrops of rock gypsum and carried by the wind to the place of deposition. Geologic distribution of gypsum deposits. Gypsum has a very wide geological range, but the workable gypsum deposits of the United States occur at only a few geological horizons. The Salina group of the Silurian carries large gypsum deposits which are worked in New York, Ontario, Ohio, and Michigan. The Lower Carboniferous carries workable gypsum deposits in Virginia, Michigan, and Montana. Most of the deposits west of the Mississippi occur in rocks of Permian or somewhat later age. Three geological series, therefore, carry almost all of the workable gypsum of the United States. Distribution of gypsum in the United States. The gypsum-producing localities of the United States are indicated on the accompanying map. This map is taken from the publication cited below,* to which the reader * "Gypsum Deposits of the United States," by George I. Adams and others. Bulletin No. 223, U. S. Geological Survey. Washington, D. C. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM 17 is referred for a much more detailed discussion of the subject, and from which most of the descriptive matter given below has been abstracted. East of the Mississippi River, the producing localities are confined to central and western New York, southwestern Virginia, northern Ohio, and two widely separated areas in Michigan; while a large unworked deposit occurs in Florida. West of that river, gypsum deposits are both numerous and widely distributed, and plaster-mills are in operation in fourteen of the Western States and Territories. Brief descriptions of the gypsum resources of the various States are given below, the States being taken up, for convenience of reference, in alphabetical order. Arizona. Gypsum can be obtained in quantity at several localities in southern Arizona, the following being particularly noteworthy: (1) In the Santa Rita Mts., Pima County, southeast of Tucson; (2) in the low hills along the course of San Pedro River, Cochise and Final counties; (3) in the Sierrita Mts., Pima County, south of Tucson; (4) in the foothills of the Santa Catalina Mts., Pima County, north of Tucson; (5) on the Fort Apache Reservation, Navajo County. Of these localities only the fourth, north of Tucson, has as yet been commercially developed. California. In the Tertiary rocks of California gypsum is widely distributed. It is found throughout nearly all the Coast Ranges, par- ticularly south of San Francisco Bay, in the foothills of the Great Valley, and in the valleys of southern California. Deposits are known to occur in the counties of Fresno, Kings, Monterey, Kern, San Luis Obispo, Santa Barbara, Ventura, Los Angeles, San Bernardino, Riverside, and Orange. Colorado. The gypsum-producing localities of Colorado occur at intervals from the northern to the southern border of the State, along the eastern foothills of the Rocky Mountains. "Gypsum has been worked extensively near Loveland: beds hava also been opened on Bear Creek, near Morrison, and eight miles to the southeast, on Deer Creek. Quarries have been developed near Perry Park and in the Garden of the Gods, near Colorado City, and also in the vicinity of Canyon City." Other deposits, as yet unworked, are known to occur in the central and western parts of the State. Florida. An extensive area of gypsum, 6 to 8 feet thick, has been described as occurring about six miles west of Panasoffkee, Fla., on a low-lying area of hummock-land known as Bear Island. The material has not, as yet, been exploited. Iowa. The gypsum of Iowa is confined to a single area of 60 to 70 square miles, near Fort Dodge, Webster County. The material occurs 18 CEMENTS, LIMES, AND PIASTERS. in one bed, which varies from 10 to 25 feet in thickness. It has been extensively worked, eight plaster-mills being now in operation in the district. Kansas. "The gypsum of Kansas consists of extensive beds of rock gypsum and a number of deposits of secondary gypsum, or gypsite. Some of the rock gypsum is suited to the manufacture of the finer grades of plaster of Paris, and the gypsite is particularly adapted for wall and cement plasters., There is a sufficient quantity of the gypsite now known to permit extensive operations for a number of years. Certain of the deposits, however, have shown signs of exhaustion, and have been abandoned. It is probable that others will be discovered, as there is a demand for further development of the industry. The rock-gypsum beds are so vast in their proportions that only those which are favorably situated with respect to transportation facilities will probably be worked. "The area in which gypsum is found is an irregular belt extending northeast and southwest across the State, as indicated on the accom- panying map of Kansas (Fig. 2). It is naturally divided into three districts, which, from the important centers of manufacture, may be named the northern or Blue Rapids area, in Marshall County; the central or Gypsum City areaj in Dickinson and Saline counties; and the southern or Medicine Lodge area, in Barber and Comanche counties. A number of small areas have been developed between these, connecting more or less closely the three main areas. The gypsum is found at Manhattan and north of that city, though not worked. It is worked at Langford, in the southern part of Clay County, and is found near Manchester, in the northern part of Dickinson County. Gypsum is worked near Burns, and has in past years been worked near Peabody and Furley, and large deposits are known near Tampa. Farther south, in Sumner County, a large mill has been operated at Mulvane, and gypsum has been quarried at Geuda Springs. These different localities show an almost continuous belt of gypsum across the State." Michigan. Gypsum is at present worked in two distinct areas in Michigan, while a third locality may prove to be of importance in the future. The two producing areas are (1) in the vicinity of Grand Rapids ; and (2) at Alabaster, near Saginaw Bay. The third, and as yet unex- ploited, area is near St. Ignace, on the Upper Peninsula. Montana. Gypsum is worked for plaster in Cascade and Carbon counties, and is known to occur at many other localities in the State. Nevada. At Moundhouse and Lovelocks, in northwestern Nevada, gypsum deposits have been developed. Large deposits also occur in southern Nevada. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. 19 20 CEMENTS, LIMES, AND PLASTERS. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. 21 New Mexico. Though gypsum is known to occur in quantity at many points, the only commercial development has been at Ancho, where a plaster-mill is now in operation. New York. The gypsum in New York State occurs as rock gypsum interbedded with shales and shaly limestones. Several gypsum beds, separated by shales, usually occur in any given section. They are lenticular in shape, but of such horizontal extent that in any given quarry they are usually of practically uniform thickness. Those that are worked vary from 4 to 10 feet in thickness in most of the quarries, but at Fayetteville a 30-foot bed is exposed. The area in which the gypsum-bearing formations are found as shown in the map, Fig. 3, extends through the central part of the State, the productive portion, of the belt including parts of Madison, Onondaga, Cayuga, Ontario, Genesee, Monroe, Livingston, and Erie counties. The most easterly points at which gypsum has been worked are in Madison County, but the product there is small and is marketed locally for use as land-plaster. In Onondaga County, at Marcellus, Fayetteville, and other points, large quarries are operated, part of the product being calcined and part ground for land-plaster. The quarries near Union Springs, in Cayuga County, produce principally land-plaster, as do those of Phillipsport, Gibson, and Victor, in Ontario County. The gypsum from Mumford, Wheatland, Garbuttville, and Oakfield is used chiefly for calcined plaster. Ohio. "The gypsum deposits of Ohio which are of economic value consist of beds of rock gypsum occurring in the northwestern part of the State. They have been known since the first settlements were made on the northern shore of Sandusky Bay. The expo- sures lie at about the level of the waters of the bay, in some places rising a few feet above it. In addition to the deposits of economic importance, gypsum is found in small pockets and isolated bodies throughout the area of the Salina group, which occurs extensively in northwestern Ohio. The deposits which are worked vary considerably in thickness, ranging from a few inches up to 9 feet. On the north shore of Sandusky Bay, in Portage Township, Ottawa County, 1500 to 2000 acres of land have been thoroughly prospected with a core- drill, and it has been shown that there are from 150 to 200 acres of workable gypsum. On the south shore of the bay, about 2J miles northwest of the town of Castalia, drilling has shown the presence of another area of workable gypsum, but no developments have yet been undertaken. The location of these deposits is shown on the accompanying map, Fig. 4. It is estimated that at the present 22 CEMENTS, LIMES, AND PLASTERS. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. 23 rate of production the known deposits will last about twenty-five years." Oklahoma. Oklahoma occupies a central position in the belt of country which carries extensive gypsum deposits all the way from the northern part of Kansas into central Texas (see Fig. I). Within its borders the number and thickness of the beds appear to be greater than to the north and south. " TJie amount of gypsum appears to be inex- haustible. With perhaps two exceptions, each of the western counties contains enough material to supply the United States for an indefinite length of time, and there are in addition considerable deposits in the eastern part of the Territory." The gypsum in Oklahoma may be con- sidered as occurring in four regions: (1) the Kay County region; (2) the main line of gypsum hills, extending from Canadian County northwest through Kingfisher, Blaine, Woods, and Woodward counties to the Kansas line; (3) the second gypsum hills, parallel with the main gypsum hills, and from 50 to 70 miles farther southwest, which extend from the Keechi Hills, in southeastern Caddo County, northwestward through Washita, Custer, Dewey, and Day counties; and (4) the Greer County region, occupying the greater part of western Greer County and the extreme southeastern corner of Roger Mills County. The deposits in Kay County consist of earthy gypsum, or gypsite. In the other three regions rock gypsum predominates, although there are numerous localities where earthy gypsum occurs in workable bodies. Oregon. Gypsum occurs in Oregon in only one known and exploited locality. This is on the eastern border of the State, near the middle point of the boundary-line, on a ridge dividing Burnt River and Snake River. A plaster-plant Ipcated at Lime uses material from this locality. South Dakota. "In the Black Hills uplift there is brought to the surface an elliptical outcrop of the Red Beds surrounding the high ridges and plateaus of the central portion of the Black Hills. The area is about 100 miles long by 50 miles wide, and the outcrop zone has an average width of 3 miles, except in a few districts where the rocks dip steeply, where it is much narrower. The formation consists mainly of red sandy shales, with included beds of gypsum at various horizons, some of which are continuous for long distances, while others are of local occurrence. The thickness of the deposits varies greatly, but in some districts over 30 feet of pure white gypsum occur, and nearly throughout the outcrop of the formation it contains deposits of suffi- cient thickness and extent as to have commercial value. "The gypsum is a prominent feature about Hot Springs. Here the principal beds occur about 60 feet above the base of the formation and "24 CEMENTS, LIMES, AND PLASTERS. have a thickness of 33J feet, exclusive of the 10-foot parting of shale between them, but this thickness diminishes somewhat northward and rapidly southward/' Texas. "The largest area in Texas containing deposits of gypsum lies east of the foot of the Staked Plains, in northern Texas. The beds have an approximately northeast-southwest strike and extend from Red River to the Colorado in an irregular line, the sinuosities of which are produced by the valleys of the eastward-flowing streams. This belt is a continuation of the deposits in Oklahoma. "In the eastern part of El Paso County, to the east of Guadaloupe rriountains, there is an area of gypsum which extends beyond the bor- der of the State northward into New Mexico. It lies north of the Texas-Pacific Railroad and west of Pecos River. In a few localities this great plain of gypsum is overlain by beds of later limestone and conglomerate. The gypsum is conspicuously exposed along the course of Delaware Creek, a stream rising in the foothills of the Guadaloupe Mountains and flowing eastward into the Pecos. "In the Malone Mountains in El Paso County there is a third area which contains notable deposits of rock gypsum. This locality has the advantage of being situated near the Southern Pacific Railway." Utah. "The more important known deposits occur in the central ;and southern portions of the State, in Juab County, east of Nephi; in Sanpete and Sevier counties, near Salina; in Millard County, at White Mountain, near Fillmore, and in Wayne County in South Wash. They .are all of the rock-gypsum type, except the one near Fillmore, which is in the secondary form of unconsolidated crystalline and granular gypsum blown up from dry lakes into dunes. Deposits are also known in Emery County, about forty miles southeast of Richfield; in Kane County, near Kanab; in Grand County, between Grand River and the La Sal Mountains; in Sanpete County, near Gunnison; in the eastern part of Washington County (?), between Duck Lake and Rockville, and at other places. Recently enormous deposits of gypsum have been reported from Iron County, at points so far from lines of trans- portation, however, as to render their exploitation impracticable for the present." Virginia. All the workable gypsum deposits of Virginia occur in Washington and Smyth counties, in the valley of the North Fork of Holston River. The area within which the known deposits are located is a narrow belt about sixteen miles in length, extending from a short distance southwest of Saltville' to a point about three miles west of Chatham Hill post-office. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. 25 The material occurs as rock gypsum, interbedded with shales and shaly limestones of Carboniferous age. The beds of gypsum average 30 feet in thickness at the localities at which they are now worked. The rocks of the district dip at a high angle, however, usually between 25 and 45, so that certain wells which have been drilled are in the gypsum for long distances, and accordingly immense thicknesses of gypsum have been erroneously reported, because the inclination of the deposits was not taken into account. Near Saltville the dip of the gypsum beds which are worked is toward the northwest; at the mines farther up the valley the dip is to the southeast. The development of the gypsum industry in this area has been governed almost entirely by the transportation facilities. The deposits in the upper valley, though extensive and easily workable, have not been largely exploited, owing to the long wagon-haul necessary. The deposits at Saltville and Plasterco, which are on a branch of the Nor- folk and Western Railroad, have furnished the principal output. Throughout the entire area the dip of the gypsum beds is so high as to require mining, except at the commencement of the working. Wyoming. Though gypsum deposits occur at many localities in the State, only two plaster-plants are at present in operation. These are located at Laramie and Red Buttes respectively. A considerable exten- sion of the Wyoming plaster industry may, however, be expected; for the supplies of gypsum are large and accessible. Canada. Gypsum occurs in New Brunswick, associated with Lower Carboniferious limestones, particularly large deposits being shown near Hillsboro, Albert County. An analysis of a typical sample from Hillsboro is given as no. 25, of Table 8, page 53. The gypsum deposits of Ontario occur in the form of beds, associated with shales and limestones, in the Salina group. The principal exploited deposits are located along the valley of Grand River, from Paris in Brant County to near Cayuga in Haldimand County. Extensive gypsum beds also occur in Devonian limestones along the Moose and French rivers, near James Bay: but these deposits are as yet entirely undeveloped. In Nova Scotia thick beds of gypsum occur near St. John Harbor, Port Bevis, and Baddeck Bay, associated with Carboniferous limestones. An analysis of gypsum from near Baddeck Bay is given as no. 26 of Table 8, page 53. Of the Canadian gypsum deposits, those of New Brunswick and Nova Scotia are of interest to American producers, for they have sup- plied large quantities of crude gypsum to plaster plants located in the 26 CEMENTS, LIMES, AND PLASTERS. United States. Most of this Canadian gypsum is used in plants located in the seaboard cities, but a considerable amount of it is calcined as far inland as Syracuse, N. Y. References on gypsum deposits. The following list, arranged by States, will serve to locate the principal papers on various gypsum deposits. ARIZONA. CALIFORNIA. COLORADO. FLORIDA. IOWA. KANSAS, Blake, W. P. Gypsum deposits in Arizona. Bulletin 223, U. S. Geol. Survey, pp. 100, 1Q1. 1904. Fairbanks, H. W. Gypsum deposits in California. Bulle- tin 223, U. S. Geol. Survey, pp.*119-123. 1904. Grimsley G. P. Gypsum and cement plaster industry in California. Eng. and Mining Journal, June 8, 1901. Anon. Gypsum (localities in California). 12th Ann. Rep California State Mineralogist, pp. 323-325. 1894. Lakes, A. Gypsum and clay in Colorado. Mines and Minerals, vol. 20, Dec., 1899. Lakes, A. Gypsum deposits in Colorado. Bulletin 223, U. S. Geol. Survey, pp. 86-88. 1904. Lee, H. A. Larimer County gypsum. Stone, vol. 21, pp. 35-37. 1900. Day, D. T. Gypsum in Florida. 20th Ann. Rep. U. S. Geol. Survey, pt. 6, pp. 662, 663. 1899. Day, D. T. Gypsum deposits in Florida. Bulletin 223, U. S. Geol. Survey, p. 48. 1904. Keyes, C. R Gypsum deposits of Iowa. Vol. 3, Reports Iowa Geological Survey, pp. 257-304. 1895. Keyes, C. R. Iowa gypsum. Mineral Industry, vol. 4 ; pp. 379-396. 1896. Wilder, F. A. Geology of Webster County. Vol. 12, Reports Iowa Geol. Survey, pp. 63-235. 1902. Wilder, F. A. Gypsum deposits in Iowa. Bulletin 223, U. S. Geol. Survey, pp. 49-52. 1904. Bailey, E. H. S., and Whitten, W. M. On the chemical composition of some Kansas gypsum rocks. Kansas University Quarterly, vol. 6, pp. 29-34. 1897. Crane, W. R. Mining and milling of gypsum in Kansas. Eng. and Mining Journal, Nov. 9, 1901. Grimsley, G. P., and Bailey, E. H. S. Special report on gypsum and gypsum cement plasters. Vol. 5, Reports Kansas Geological Survey. 1899. Grimsley, G. P. Gypsum deposits in Kansas. Bulletin 223, U. S Geol. Survey, pp. 53-59. 1904. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. 27 MICHIGAN. Gregory, W. M. Gypsum in Arenac and adjoining counties. Ann. Rep. Michigan Geol. Surveyfor 1901, pp. 15-18. 1902. Grimsley, G. P. Preliminary report on the gypsum deposits of Michigan. Ann. Rep. Michigan Geol. Survey for 1902, pp. 4-10. 1903. Grimsley, G. P. Gypsum deposits in Michigan. Bulletin 223, U. S. Geol. Survey, pp. 45-47. 1904. Grimsley, G. P. A theory of origin for the Michigan gypsum deposits. American Geologist, vol. 34, pp. 378-387. Dec., 1904. Grimsley, G. P. The gypsum of Michigan and the plaster industry. Vol. 9, pt. 2, Reports Michigan Geol. Sur- vey. 1904. MONTANA. Weed, W. H. Gypsum deposits in Montana. Bulletin 223, U. S. Geol. Survey, pp. 74, 75. 1904. NEVADA. Louderback, G. D. Gypsum deposits in Nevada. Bulletin 223, U. S. Geol. Survey, pp. 112-118. 1904. NEW MEXICO. Herrick, H. N. Gypsum deposits in New Mexico. Bulletin 223, U. S. Geol. Survey, pp. 89-99. 1904. N-EW YORK. Clarke, W. C. The gypsum industry in New York State. Bulletin 11, N. Y. State Museum, pp. 70-84. 1893. Eckel, E. C. Gypsum deposits in New York. Bulletin 223, U. S. Geol. Survey, pp. 33-35. 1904. Lincoln, D. F. Report on the structural and economic geology of Seneca County, N. Y. 14th Ann. Rep. N. Y. State Geologist, pp. 60-125. 1896. Luther, D. D The economic geology of Onondaga County, New York. 15th Ann. Rep. N. Y. State Geologist, vol. 1, pp. 241-303. 1897. Merrill, F. J. H. Salt and gypsum industries of New York. Bulletin 11, N. Y. State Museum, 89 pp. 1893. Parsons, A. L. Recent developments in the gypsum in- dustry in New York State. 20th Ann. Rep. N. Y. State Geologist, pp. 177-183. 1902. Pohlman, J. Cement rock and gypsum deposits in Buffalo. Trans. Am. Inst. Min. Engrs., vol. 17, pp. 250-253. 1889. OHIO. Orton, E. Gypsum or land plaster in Ohio. Vol. 6, Re- ports Geol. Survey Ohio, pp. 696-702. 1888. Peppel, S. V. Gypsum deposits in Ohio. Bulletin 223, U. S. Geol. Survey, pp. 38-44. 1904. OKLAHOMA. Gould, C. N. Oklahoma gypsum. 2d Biennial Rep. Oklahoma Dept. Geology, pp. 75-137. 1902. 28 CEMENTS, LIMES, AND PLASTERS. OKLAHOMA. Gould, C. N. Gypsum deposits in Oklahoma. Bulletin 223, U. S. Geol. Survey, pp. 60-67 1904. OREGON. Lindgren, W. Gypsum deposits in Oregon. Bulletin 223, U. S. Geol. Survey, p. 111. 1904. SOUTH DAKOTA. Darton, N. H. Gypsum deposits in South Dakota. Bulle- tin 223, U. S. Geol. Survey, pp. 76-78. 1904. TEXAS. Hill, B. F. Gypsum deposits in Texas. Bulletin 223, U. S. Geol. Survey, pp. 08-73. 1904. UTAH. Bout well, J. M. Gypsum deposits in Utah. Bulletin 223, U. S. Geol. Survey, pp. 102-110. 1904. VIRGINIA. Boyd, C. R. Gypsum in southwestern. Virginia. Resources of southwest Virginia, 8vo, pp. 104-108. 1881. Eckel, E. C. Salt and gypsum deposits of southwestern Virginia. Bulletin 213, U. S. Geol. Survey, pp. 406-416. 1903 Eckel, E. C. Gypsum deposits in Virginia. Bulletin 223, U. S. Geol. Survey, pp. 36-37. 1904. Stevenson, J. J. Notes on the geological structure of Tazewell, Russell, Wise, Smythe, and Washington counties of Virginia. Proc. Amer. Philos. Soc., vol. 22, pp. 114-161. 1885. WYOMING. Knight, W. C. Gypsum deposits in Wyoming. Bulletin 223, U. S. Geol. Survey, pp. 79-85. 1904. Slosson, E. E., and Moudy, R. B. The Laramie cement plaster. 10th Ann. Rep. Wyoming Agricultural College. 1900. CANADA. Bailey, L. W., and Ells, R. W. Report on the Lower Car- bonaceous belt of Albert and Westmorland counties, New Brunswick. Report Canadian Geological Survey for 1876-77, pp. 351-401 1878. Bell, R., and others. [Gypsum and plaster in Ontario.] Report on the mineral resource.! of Ontario, pp. 119-123. 1890. Fletcher, H. Report of explorations and surveys in Cape Breton, Nova Scotia. Report Canadian Geological Survey for 1875-76, pp. 369-418. 1877. Gesner, A. On the gypsum of Nova Scotia. Quarterly Journal Geological Society, vol. 5, pp. 129, 130. 1849. Gilpin, E. The gypsum of Nova Scotia. Trans. North of England Institute of Mining Engineers, vol. 30, p. 68. 1881. , Nicol, W. Anhydrite in Ontario. Canadian Record of Science, vol. 7, p. 61. 1896. Anon. Nova Scotia gypsum. Canadian Mining Review, March 1896. COMPOSITION, DISTRIBUTION, AND EXCAVATION OF GYPSUM. 29* Excavation and handling of rock gypsum. Deposits of rock gypsum are worked either in open quarries or in mines, the choice depending on the thickness of the bed, its dip, and the amount of stripping neces- sary. Usually work is commenced in an open cut on the outcrop of the gypsum bed. After the entire available face on the property has been opened in this manner, it is necessary to decide whether the work- ings can be most economically driven as underground tunnels or slopes, or by stripping and open-cut work. At the Severance quarries at Fayetteville, N. Y., over 40 feet of shale and limestone stripping is removed, but the total thickness of gypsum beds shown here is 60 feet; and such heavy stripping could not be justified in order to work thinner beds. Under ordinary conditions the cost of quarrying gypsum may range from 20 to 35 cents per ton, as compared with 40 to 60 cents per ton for mining it. In mining, large pillars must be left at frequent intervals,, and timbering is necessary, in addition, for extensive workings. Mining methods. The mining methods practiced at a typical Kansas locality are described * as follows by Crane : "As a rule, there is little or no system employed in laying out the- workings. Main lines of haulage are run as continuations of the sur- face drifts, other openings being run parallel with them on further de- velopment, or run from the foot of a shaft sunk to the workable deposit. On one or both sides of the haulageways rooms are driven, which often run together, thus leaving odd and very irregularly shaped pillars. Long working-faces are often formed, which must be again broken by passages forming pillars for the support of the roof. Usually, how- ever, single rooms, more or less irregular in shape, are opened up and worked until the handling of the product becomes inconvenient, when new and more advantageously placed openings are begun. " The mine in question was opened by an adit, which, beginning on a fairly steep hillside, at a point on a level with the second floor of the mill, extends into the hill for a distance of about 1000 feet. No special attempt was made to align the adit, consequently considerable useless work was done. For the first 400 feet the adit runs approxi- mately north; the next 300 feet shows a marked variation from the north-and-south line. An attempt was then made to rectify the devia- tion by driving a right-angled offset 25 feet in length; the remaining 300 feet was driven approximately parallel with the first 400 feet. * Crane, W. R. The gypsum-plaster industry of Kansas. Eng. and Mining; Journal, p. 442, March 17, 1904. 30 CEMENTS, LIMES, AND PLASTERS. '' Unfortunately, the adit was driven so nearly level as to render drainage very difficult, and much water stands in depressions on the limestone floor. " The adit is lined with rough-hewn oak, walnut, and red-elm timber, except the last 300 feet, which has round timbers of similar material. Three-quarter sets that is, sets with posts and caps only are em- ployed. The posts and caps are 6 feet 2 inches and 6 feet 4 inches long, respectively, both being 8X8 iriches in section. They are spaced 36 inches. The posts stand on a limestone stratum 2 feet in thickness, and therefore require no sills. The sets for the first 700 feet are lagged with 2 X 12-inch oak plank ; the remaining 300 feet' has plank lagging on the caps and pole-lagging on the posts. A single track of 36-inch gauge is laid in the middle of the tunnel for the mine-cars, which are drawn by mule-power. The cars have a capacity of from 800 to 1000 Ibs. gypsum. " The gypsum mined is 8.5 feet thick and is won by shooting it from the face or sides of the rooms, holes being bored by hand-operated post-augers, Hardscop make. The holes are 1.5 inches in diameter and range from 3 to 6 feet deep. Black powder of C grade is usually em- ployed, the charge ranging from 6 to 14 inches per hole. Squibs are employed in firing the charges. The cost of explosive per ton of gypsum extracted is about four cents. A 4 X 6-foot air-shaft connects the end of the adit with the surface, 96 feet above." Working gypsum-earth deposits. Deposits of gypsite or gypsum earth, being purely surface deposits of a soft, granular material, can be worked best by methods entirely different from those used in ex- cavating rock gypsum. The gypsum earth is not only soft, but frequently carries a large percentage of moisture: and as it freezes deeply because of this moisture, the Kansas deposits can be worked only during warm weather. If the gypsum earth is covered by soil or sand, this is stripped. The gypsum earth is then loosened by disk harrows or plows, and taken up by wheeled scrapers. It is then taken to drying-sheds, in order to get rid inexpensively of as much of the water as possible. The cost of working a gypsum-earth deposit, under average conditions, may fall between 10 and 25 cents per ton. CHAPTER II. CHEMISTRY OF GYPSUM-BURNING. MANUFACTURE OF PLASTERS BEFORE taking up the actual methods and details of plaster-manu- facture, it will be of advantage to discuss briefly the chemical and physi- cal principles on which the industry is based. Chemistry of gypsum-burning. Pure crude gypsum is a hydrous sulphate of lime, with a chemical formula CaS04 + 2H 2 0. This corre- sponds to the composition: f fLime(CaO) ....... 32. 6% 1 CaSO +2H O= ] Lime sul P hate (CaSO 4 ) j Sulphur trioxide I =79.1% i (SO 3 ) ........... 46 . 5 [Water (H 2 O). .......... ................. =20.9 100.0 If pure crude gypsum he heated to a temperature of more than 212 F. and less than 400 F., a certain definite portion of the water of combination will be driven off, and the gypsum thus partially de- hydrated will be plaster of Paris. Plaster of Paris has the formula + iH 2 0, corresponding to the composition: Three fourths of the original water of combination have therefore been driven off in the course of the process. Dehydration to this extent can, as above noted, be accomplished at any temperature between 212 F. and 400 F. In actual practice, however, it is found most economical of fuel and time to carry on the process at the highest allow- able temperatures; and 330 to 395 F. may be regarded as the usual limiting temperatures for plaster-manufacture. About 400 F. is a critical temperature, for if gypsum be heated at temperatures much above this, it loses all of its water of combination, becoming an entirely anhydrous sulphate of lime, and useless as a normal plaster. Under certain conditions, however, gypsum burned at temperatures above 400 F. gains valuable properties. Such highly 31 32 CEMENTS, LIMES, AND PLASTERS. burned gypsum products will be considered in Chapter IV, under the head of Flooring and Hard-finish Plasters. Recurring to plasters burned at temperatures lower than 400 F., it may be said that if the gypsum is pure, the resulting plaster will harden or set very rapidly when mixed with water, reabsorbing sufficient water to regain its original comppsition of CaSO 4 + 2H 2 0. Such quick- setting pure plasters are conveniently grouped as plaster of Paris. If, however, the crude gypsum carried a large percentage of impurities, or if certain materials are added to the plaster after burning, the product will set much more slowly. Such slow-setting plasters are of value in structural work, and are marketed under the somewhat misleading name of '" cement plasters". The term is unfortunate, because such "cement plasters" are in no way related to the much better known " hydraulic cements" discussed later in this volume. Using the properties above noted as a basis for classification, the group of plasters may be subdivided as follows: CLASSIFICATION OF PLASTERS. A. Produced by the incomplete dehydration of gypsum, the calcination being carried on at a temperature not exceeding 400 F. 1. Produced by the calcination of a pure gypsum, no foreign materials being added either during or after calcination . . PLASTER OF PARIS. 2. Produced by 1he calcination of a gypsum containing certain natural impurities, or by the addition to a calcined pure gypsum of cer- tain materials which serve to retard the set of the product. CEMENT PLASTER. B- Produced by the complete dehydration of gypsum, the calcination being carried on at temperatures exceeding 400 F. 3. Produced by the calcination of a pure gypsum. . . FLOORING-PLASTER. 4. Produced by the calcination, at a red heat or over, of gyosum to which certain substances (usually alum or borax) have be3n added HARD-FINISH PLASTER. Commercial classification of plasters. In the trade the names above suggested are used quite extensively, biit at times in a careless and indefinite fashion. Calcined plaster commonly means a burned plaster to which no rctarder has been added. If the gypsum from which it was made was pure, the resulting calcined plaster will be a plaster of Paris, as defined above. If the gypsum used was impure, however, the resulting calcined plaster would be a cement plaster, as defined above. Stucco is almost a synonym for plaster of Paris, as it contains no retarder and is made from fairly pure gypsum: but the product handled MANUFACTURE OF PLASTERS. 33 commercially as plaster of Paris is usually more finely ground than stucco and is as white as possible. Wall-plasters are made by adding not only retarder but also hair (or some other fiber) to calcined plaster. Keene's " cement", Parian "cement", etc., are plasters used as hard finishes in buildings. Their properties are due to certain pecu- liarities of their manufacture, for which reference should be made to Chapter IV. In the present chapter the manufacture of plaster of Paris, cement plaster, and wall-plaster will be taken up, and followed by a chapter on the properties of the resulting products. The manufacture and properties of the flooring and hard-finish plasters will be discussed together in Chapter IV. Manufacture of Plaster of Paris, " Cement Plaster", and Wall-plaster. Though plaster of Paris and "cement plasters" are very distinct so far as properties and fields of use are concerned, their processes of manufacture are so similar that they will be treated together in this chapter. It will be recalled that in manufacturing plaster of Paris a pure gypsum is used, so that the product sets very rapidly, while in making cement plasters slowness of set is obtained either by using a naturally impure gypsum or by adding a retarder to the material during or after its manufacture. Aside from this difference, and a slight difference in the calcining temperature, which is usually somewhat lower for plaster of Paris than for cement plaster, the methods employed in making the two products are closely similar. Two operations are necessary in manufacturing both kinds of plaster: the raw material must be properly calcined and finely ground. The grinding may either precede or follow the burning, for the order of the two operations depends largely upon what calcining process is used. If the burning is carried on in kettles, the grinding is usually done first; but if the burning is carried on in ovens or rotating cylinders, the raw material is necessarily or advisably fed in lumps, and the fine grinding, therefore, follows the burning. In the present chapter the subject will be discussed under the following headings: (1) Grinding gypsum and plaster. (2) Calcining by the oven process. (3) Calcining by the kettle process. (4) Calcining by the rotary cylinder process. (5) Addition of retarders and acceleration. (6) Costs of plaster-manufacture. 34 CEMENTS, LIMES, AND PLASTERS. Grinding gypsum and plaster. In American plants using the kettle- calcining process the gypsum is finely pulverized before calcination. This pulverizing is usually accomplished in three stages, though when gypsum earth is used instead of rock gypsum the coarse crushers are dispensed with. The three stages are: (1) The lump gypsum, as quarried, is crushed to 2- to 4-inch size in a Blake, Gates, or other coarse crusher. FIG. 5. Nipper for coarse crushing of gypsum. (Butterworth & Lowe.) (2) The product of the coarse crushers is fed to reducers of the coffee-mill type, which crush it to about } inch or so. (3) The final pulverizing is accomplished in either buhrstone mills, Sturtevant rock-emery mills, or Stedman disintegrators. These reduce the gypsum so that from 55 to 65 per cent will pass a 100-mesh sieve, and it is then ready to be fed to the kettles. A typical series of gypsum-grinding machinery is shown in Figs. 5-8. Fig. 5 shows a " nipper", used for the first coarse reduction. It is a heavy crusher of the jaw type, and when used for gypsum- crushing is usually equipped with corrugated jaws, in order to pre- vent clogging. The machine shown in the illustration has a jaw- opening of 16i"X25f", and a shipping-weight of 10,200 Ibs. A smaller MANUFACTURE OF PLASTERS. 35 nipper, weighing 8100 Ibs. and with a 36"X12" belt pulley, is quoted as having a capacity of 10 to 14 tons per hour, and is listed at $550. The " nipper" is usually followed by the " cracker" (Fig. 6), which is a heavy machine of the familiar toothed spindle type. A cracker FIG. 6. Cracker for intermediate reduction. (Buttcrworth & Lowe.) weighing 8000 Ibs. has a capacity of 12 to 15 tons per hour, and is listed at $850. For the final reduction the Stedman disintegrator, Sturtevant rock- emery mill, or ordinary buhrstones are generally used. The last two machines are described in a later section of this volume (pp. 239, 240), as they are quite extensively used in grinding natural-cement clinker. 36 CEMENTS, LIMES, AND PLASTERS. The Stedman disintegrator (Fig. 7) is composed essentially of totu concentrically placed cages, formed of steel bars. Of these cages, the first and third revolve in one direction, the second and fourth in the opposite. The material to be crushed is fed into a hopper which dis- charges it at' the center of the cages. The gypsum lumps are struck FIG. 7. Stedman disintegrator, 50-inch, heavy pattern; open and slid apart. by the bars of the inner cage, and thrown outward at high velocity. The bars of the second cage, revolving in the opposite direction, strike them with a blow of double force, and this operation is repeated by the bars of the third and fourth cages in succession. TABLE 5. SIZES, CAPACITY, ETC., OF STEDMAN DISINTEGRATORS. Size. Horse- power. Capacity in 10 Hours. Price. Weight, Lbs. 30-inch disint 36- " sgrator, heavy patt ern 6-9 12-18 8t 18 o 10 to 25 ns $300.00 450.00 3,000 5,500 42- ". light 12-18 20 30 500.00 6,000 40- " heavy ' , 20-25 25 35 COO. CO 10,000 44- " i (i i 30-35 40 50 700.00 12,000 50^ " t K i 35-45 60 75 900.00 15,000 MANUFACTURE OF PLASTERS. 37 After being reduced as alcove described, the gypsum is calcined. Usually it is necessary to regrind some of the product which comes from the kettles; and this may be accomplished in any of the fine grinders above noted. When the rotary process is used, it is customary not to pulverize the material until after calcining. As calcined plaster is much easier to grind than crude gypsum, a considerable saving in power and repairs is effected by this difference in practice. FIG. 8. Stedman disintegrator, showing cage construction. Calcining in ovens. In the manufacture of the higher grades of plaster of Paris it is necessary that the material should be calcined with extreme uniformity and at exactly the proper temperature. This uniformity in burning is attainable in ovens, though the process is necessarily expensive in fuel and labor. For these reasons the oven process has not been used in the tJnited States, though it still persists in Europe for certain grades of plasters. Calcining in kettles. The favorite process in the United States, particularly in the plaster-plants of the Middle West, is that in which the calcination is effected in kettles. As noted later in discussing continuous calcining processes (pp. 46-50) the kettle process is slow, low m output, and expensive in fuel. For these reasons it will probably disappear as the continuous rotary calciner becomes perfected; but at present it is still used in the majority of American plaster-plants. The statements above should not be construed as a too sweeping con- 38 CEMENTS, LIMES, AND PLASTERS. demnation of the kettle process, for that process is undoubtedly far superior in economy to its European progenitor, the oven process. FIG. 1. FIG. 2. Section on C-D, Fig. 1 Section on A-B, Fig. 2. FIG. 9. Construction and setting of gypsum-kettles. (Trans. Am. Inst. Min. Engrs.) The following description of the process of calcining plaster in kettles is abstracted, in large part, from an admirable paper * by Wilkinson. In this process the gypsum is ground, and charged into cylindrical * Trans. Am. Inst. Mining Engineers, vol. 27, pp 514 et seq. MANUFACTURE OF PLASTERS. 39 "kettles". Heat is applied both at the bottom of the kettle and by flues passing entirely through the cylinder. A heavy stone or brick masonry support is built for the kettle, in- closing a fire-space in the form of an inverted cone about 4 feet high. At the top of this cone a cast-iron flanged ring is set in the masonry. FIG. 10. Four-flue kettle, with accessories, dismounted. (Butterworth & Lowe.) On this flange is placed the "kettle-bottom", which is an iron casting, concavo-convex in shape, a little less than 8 or 10 feet in diameter, with the convexity placed upward, the rise being 1 foot. This bottom has a thickness of f inch at the edges and 4 inches at the crown. Kettle- bottoms must be made of the best scrap-iron, as ordinary scrap-iron does not last as long as pig. Sheet steel has been tried, but does not serve as well as the best scrap. "The life of a kettle-bottom is terminated by cracking. The cracks can be calked with asbestos cement, but the expense of stoppage and repairing soon overcomes the saving." 40 CEMENTS, LIMES, AND PLASTERS. Within the past few years sectional kettle-bottoms have been in- troduced quite extensively. A kettle of this type is shown in Fig. 11, in which the kettle-bottom is composed of a central circular section and six other sections fitting around it. These sections are made of cast iron. The principal merit of this design is that in case any section of the kettle-bottom burns out, it can be replaced without disturbing the kettle or brickwork. - - r )]aster_:_3_sai _d_. < V FIG. 18. Effect of sand on tensile strength of plasters. (Marston.) These tests appear to show f that (a) Cement plasters and stuccos attain almost their full strength at the end of one week, showing little further gain at three months. (6) The portion of the plaster which passes a 100-mesh sieve is stronger than the coarser portions, and the higher strength of fine plaster is shown better in sand mixtures than when tested neat. (c) The value of fine grinding is further emphasized by the high results shown by the German stucco, which seems to have been the most finely ground of all, though the values of fineness given are not quite consistent. * Iowa Geol. Survey, vol. 12, pp. 232-235. f The conclusions here drawn from these tests are those of the writer. For Professor Marston's conclusions, which do not entirely agree with mine, reference should be made to the original work. COMPOSITION, PROPERTIES, AND TESTS OF PLASTERS. 61 TABLE 15. RESULTS OF TENSILE TESTS OF PLASTERS, ETC. Material. Tensile Strength per Square Inch. Kept in IDay. 7 Days. 28 Days. Stucco, Fort Dodge, Iowa. Air i ( Water Air i ( Water ' i ( Air t ( Water 1 1 Air 1 1 Water 1 1 i ( Air t ( Water i ( Air i ( Water i ( Air < i Water Air Water i ( Air 1 1 Water 226 219 195 208 219 186 175 189 211 208 192 202 215 192 195 131 144 187 214 192 204 188 214 107 131 82 111 227 221 216 226 181 134 183 185 204 210 139 154 188 230 185 170 184 220 172 175 190 301 203 196 170 228 163 193 224 217 207 205 128 175 20 112 236 208 223 201 195 218 196 162 329 438 187 200 379 245 168 209 375 360 180 186 237 437 195 205 483 470 182 148 '348 285 158 163 333 303 151 154 468 461 154 181 465 286 195 215 Kallolite plaster Ft Dodge, Iowa Duncomb plaster, Ft. Dodge, Iowa. . . . . Mineral City plaster Ft Dodge, Iowa Stone plaster, Ft Dodge, Iowa. . . Flint plaster, Ft. Dodge, Iowa ; Acme plaster Acme, Texas Stucco, Baker Stucco Co , Kansas , , ........ I Plaster, Baker Stucco Co., Kansas Compression tests and effects of sand. A valuable series * of ex- periments were carried out during 1899-1900 by Profs. Slosson and Moudy on the compressive strength of plasters, both neat and mixed, with varying properties of sand. Most of the material used for these tests was a cement plaster manufactured at Laramie, Wyo., the tests being made before .the addition of retarder to the plaster. The material was molded into 2-inch cubes and crushed in a Riehle self-registering machine after the cubes had been exposed to the air * Tenth Ann. Rep. Wyoming Agric. and Mech. College, 1900. 62 CEMENTS, LIMES, AND PLASTERS. for one week. "The sand used was obtained from the Laramie River and was composed of sharp-angled granitic fragments. It was sifted through a millimeter sieve." 2tOO JWOO 1800 1600 1100 1200 1000 600 Ibs. 1| 11 11 111 FIG. 18a. Effect of sand on compressive strength of plaster. Adhesive tests of plasters. The adhesive tests in Table 17 were made by Prof. Marston "by taking pieces of No. 2 paving-brick from Des Moines and grinding them on the emery-wheel so as to make approximately 1-inch cubes. Each cube had one face carefully COMPOSITION, PROPERTIES, AND TESTS OF PLASTERS. 63 trued to give a cross-section exactly 1 inch square. These pieces of paving-brick were placed in the cement briquette molds with this true surface exactly at the middle of the mold. The plaster or stucco was placed to fill the other half of the mold, while the half in which the piece of brick was placed was filled with neat Portland-cement mortar ". TABLE 16. EFFECT OF SAND ON COMPRESSIVE STRENGTH OF PLASTERS. Parts by Weight. Number Strength, Lbs. Kind of Plaster. of 2-inch per Tests. Cube. Sq. In. Plaster. Sand. Water. Laramie plaster, no retarder. . . 100 56 4 5435 '1358f 100 12$ 56 2 5430 1357i 100 25 57 10 4575 1143f . 100 50 58ft 5 4378 1094^ ' 100 75 eo 7 4317 1079i 100 100 611 6 3755 938| 100 125 62 1 7 3505 876i 100 150 63A 8 3341 835} 100 1V5 65 6 3523 880i Red Buttes plaster, no retarder 100 50 ? 8945 22361 u a tt it ti 100 100 55 ? 6622 1655^ Agatite plaster, market sample 100 50 ? 3550 8873 100 100 55 ? 2597 649| TABLE 17. ADHESIVE STRENGTH OF PLASTERS. Material. Adhesive Strength per Square Kept in 7 Days. 28 Days. 133 75 115 102 81 84 212 84 80 114 103 117 100 105 95 133 Stucco Ft Dodge Iowa Air Water Air Water Air Water Air Water Air Water Air Water Air t i Water Air Water Air Water 87 87 45 31 52 43 62 72 31 64 26 76 98 83 55 82 63 t i t t t t it Kallolite stucco, Ft Dodge, Iowa tt it a it " plaster, " " . tt (1 11 ti Duncomb " " "... K K a n Mineral City plaster Ft Dodge Iowa n tt t t t t n t i Stone plaster Ft Dodge Iowa Flint " ' l " " it tt ( e ( t i ( Acme plaster Acme, Texas Baker stucco, Kansas t t t i it 1 ' plaster, " ii t t it Plaster Okarche Oklahoma 1 1 e ( it 64 CEMENTS, LIMES, AND PLASTERS. Rate of Set and Hardening. A pure plaster of Paris will normally harden or set in from five to fifteen minutes after having been mixed with water. Plasters made from impure gypsum will be considerably slower setting than this, setting usually in from one to two hours. When plasters are to be used for structural work, they must be either naturally slow-setting, like somtf of the cement plasters, or be made slow-setting by proper treatment. Retarders are therefore used at plaster-plants in preparing their product for the market. Occasion- ally, though rarely, a plaster will be naturally too slow setting for the particular use to which it is to be applied. In this case an accelerator must be used. Theory of the action of retarders and accelerators. As will be noted later, the materials most commonly used as retarders are glue, tankage, and other organic and uncrystallized materials, while accelerators are usually inorganic and crystallized. This fact suggested a theory as to the cause of the action of accelerators and retarders. The theory, as set forth * by its originator, Dr. Grimsley , is as follows : Dr. Grimsley assumes that the set of plaster is due to the presence of a few small crystals which have escaped dehydration during burning and which set the example, so to speak, to the other crystals to form; and, further, that the strength of the set material is due to the formation of a mass of interlacing crystals. The action of retarders and accelerators is therefore explained by assuming that "any substance [added to the water with which the calcined plaster is mixed, or to the dry plaster] which will keep the molecules apart or from too close contact will retard the setting. Such substances are dirt or organic matter that is not of a crystalline character ". On the other hand, the action of accelerators is ascribed to the fact that, being of crystalline character, they induce crystallization in the plaster to which they are added. It is probable that the researches of Rohland on the effect of various substances on the speed of hydration of plasters, Portland cement, etc., are directly applicable to the question on hand. In summing up his conclusions, Rohland decided that substances which increase the solu- bility of the cementing material accelerate its speed of hydration, while substances which decrease the solubility of the cementing material retard its hydration. Materials used as retarders. The materials used as retarders are usually of animal or vegetable origin. Glue, sawdust, blood, and packing- * Vol. 5, Reports Geological Survey Kansas, pp. 167, 168. COMPOSITION, PROPERTIES, AND TESTS OF PLASTERS. 65 house tankage are some of the retarders most commonly used. Many "patent" retarders are also on the market, most of which are based on one or more of the organic materials noted above. In the course of the experiments by Slosson and Moudy, recorded below, an effort was made to obtain a cheap and satisfactory home- grown retarder. With this in view a common western cactus (Opuntia platycarpia) was dried and ground. The common malva (Malvastrum coccineum) was also prepared in a similar manner. The retarders thus made were light green in color, possessed no disagreeable odor, and when used in the proportion of 2 pounds of retarder to a ton of plaster gave excellent results. It is probable that many other plants, dried and ground in similar fashion, would form satisfactory retarders. Effect of retarders on strength of plasters. The following table * contains the results of compression tests by Profs. Slosson and Moudy on plasters containing various amounts of retarder. It appears to show conclusively that the compressive strength (and inferentially the tensile strength) of plasters decreases as the amount of retarder in- creases. TABLE 18. EFFECT OF RETARDERS ON STRENGTH OF PLASTERS. Pounds of Crushing Strength in Pounds. Retarder per Ton of Plaster. 2-inch Cube. Per Square Inch. Laramie plaster 5435 1358.75 2 4065 1016.25 Red Buttes plaster. . . 8945 2236.25 it (i (i 2 7192 1798.00 { ( it 11 4 6480 1620.00 In making the tests the finely ground retarder was accurately weighed and thoroughly mixed with the dry plaster. Water was then added and the mixture, placed on a glass plate, was fashioned into a pat about 4 inches in diameter. For determining the setting time two needles were used, each -rV inch in diameter, one loaded with 'a weight of \ lb., the other with a weight of 4 Ibs. The time when the more lightly weighted needle ceased to make a decided impression on the surface of the plaster pat is reported as " initial set ", and when the more heavily weighted needle was supported without indenting the plaster is 1 reported as " final set ". * Tenth Ann. Rep. Wyoming Agric. and Mech. College. 66 CEMENTS, LIMES, AND PLASTERS. TABLE 19. EFFECT OF VARIOUS RETARDERS ON RATE OF SET. Kind of Plaster. Kind of Retarder. Pounds Retard- er per Ton. Initial Set, Minutes Final Set, Minutes Remarks. Laramie None . . . . ". . i . . . (U ?' ? 1 2 4 6 8 2 4 6 8 1 2 4 6 1 2 2 4 4 6 6 2 2 4 6 2 4 6 6 6 10 2 8 2 8 ? ? 145 280 320 365 520 590 785 350 535 735 755 260 265 280 290 215 255 250 255 250 215 245 275 250 290 245 280 275 290 240 260 30 24 20 60 270 50 285 505 310 240 225 335 720 380 430 655 690 390 645 sio 310 315 320 310 250 315 315 310 300 255 285 320 310 340 295 295 310 375 320 330 50 42 32 105 'so 345 375 305 Cracked a little. Never hardened. Cracked somewhat. Cracked badly. Did not set. Did not set. Soft-cracked. t Market sample. . . . < ( Wyrnore ( i t t i t t n t It Webster Citv tl < it" tt ( et It t ({ tt Swift's, Chicago. . <( (i 1 1 ft 1 1 Swift's, Kansas City n (( t (I ( C ( 1 1 11 ( '. t 11 ( (I 11 I ( t { { ( Cactus tt tt tt tt tt tt tt tt tt tt tt tt i i tt i ( tt (t tt Malva tt tt e t tt Glucose tt Clay (bentonite). . . t( t i None tt Red Buttes stucco 1 1 ft it ti Red Buttes plaster. (i 1 1 1 1 Agatite plaster. . . . i ( 1 1 Wymore (i None Wymore Market sample. . . . Use of accelerators. In making dental plaster and plaster for certain other purposes an extremely rapid set is desirable. For this purpose many crystalline salts are available, common salt being one of the best accelerators known. COMPOSITION, PROPERTIES, AND TESTS OF PLASTERS. 67 TABLE 20. EFFECT OF ACCELERATORS ON RATE OF SET. Kind of Plaster. Kind of Accelerator. Pounds Accelerator Time of Set in Minutes. per Ton. Initial Set. Final Set. Laramie None o 145 225 1 1 Common salt 6 55 150 i ( Sodium sulphate. . . . 6 150 170 1 1 Sodium carbonate 6 145 170 References on properties and tests of plasters. The following brief list covers the principal papers on this subject. Bailey, E. H. S. On the chemistry of gypsum, plaster of Paris, and cement plaster. Vol. 5, Reports Kansas Geological Survey, pp. 134-170. Marston, A. Preliminary tests of stucco and plaster made by the Civil En- gineering Department of Iowa State College. Vol. 12, Reports Iowa Geological Survey, pp. 224-235. 1902. Slosson, E. E., and Moudy, R. B. The Laramie cement plaster. 10th Ann. Rep. Wyoming, College Agriculture and Mechanics, 1900. Anon. Tests for plaster of Paris. Stone, vol. 25, pp. 331-334. 1903. Hardening gypsum and plaster. The following methods of harden- ing an ordinary plaster have been recently recommended : * (1) Two to four per cent of finely ground marshmallow-root are intimately mixed with powdered plaster and the mixture kneaded to a dough with 40 per cent of water. The resulting mass resembles a stiff clay, hardens in about an hour, and finally becomes hard enough to cut, file, or bore. A harder and tougher mass may be obtained by increasing the quantity of marshmallow-root to 8 per cent. Gum, dextrin, or glue may be substituted for the marshmallow-root if more convenient. If the objects are to be exposed to high temperatures shellac may be used. (2) Six parts of gypsum are mixed with one part of freshly slaked lime and the mixture is soaked with a concentrated solution of mag- nesium sulphate. In preparing this mixture, too much gypsum must not be poured into the water, and the mixture must be stirred quickly so that lumps do not form. The smaller the quantity of water used the thicker and firmer is the cement. The porosity caused by the gradual loss of water can be obviated by soaking the objects in a solu- tion of ozocerite or wax in oil of turpentine, varnish, or hot tar, or by coating them with shellac. * Journ. Soc. Chem. Industry, vol. 21, p. 347. CHAPTER IV. FLOORING-PLASTER^ AND H&RD-FINISH PLASTERS THE two groups of plasters to be considered in this chapter agree (1) in being prepared by burning gypsum at a hjgher temperature than is employed in the manufacture of plaster of Paris and " cement " plasters, and (2) in being products which; for plasters, set rather slowly but finally take on great hardness. Because of these last properties, the flooring-plasters and hard-finish plasters are available for certain uses to which ordinary plasters are ill adapted. So much for the resem- blances between the two groups. Their points of difference are, that the flooring-plasters are prepared by simple burning at high tempera- tures, while the hard-finish plasters are produced by a double burning, with the additional use of chemicals. Neither product is made to any extent in the United States, though a considerable quantity of hard-finish plasters are imported every year. The data obtainable as to processes of manufacture are scanty, and the descriptions published are often contradictory, so that it has been difficult to prepare a satisfactory account of these products. It is believed, however, that the descriptions given below contain no errors of importance. Flooring-plasters. The flooring-plasters (" Estrichgips " of German reports) include those plasters made by calcination of a relatively pure gypsum at tem- peratures of 400 F. or higher. In the literature of gypsum and plaster it is often stated that gypsum, burned at temperatures exceeding 400 F., yields a completely dehy- drated product an artificial anhydrite which is entirely valueless as a structural material, because it has completely lost its property of re- combining with water. This statement is, however, erroneous, for plas- ters burned at such temperatures are regularly made and used. They set with extreme slowness, however, and require very fine grinding. Composition of flooring-gypsum. Until very recently no satisfac- tory discussion of this phenomenon had been attempted, and the few published accounts of the manufacturing processes employed were con- 68 FLOORING-PLASTERS AND HARD-FINISH PLASTERS. 69 tradictory as to temperatures reached, composition of product, etc. Fortunately, however, a detailed account * of the chemical changes involved was published during 1903 by Van't Hoff in the Transactions of the Berlin Academy of Sciences. As this paper is practically inac- cessible to the American engineer or manufacturer, a translation f is here appended: "As a complement to the investigations on gypsum and anhydrite we have turned our attention to a kindred product, usually designated by the term hydraulic gypsum or floor-gypsum (Estrichgips). It is obtained by burning natural gypsum, CaS04.2H 2 O. When natural gypsum is worked up into stucco gypsum (CaSO^.H^O, in the process called cooking, the temperature of 120-130 C. is not exceeded; but in the preparation of floor-gypsum higher temperatures are applied. Accordingly the product is free from water, but we are not in this case dealing with dead-burned gypsum, since the capacity to bind water has not yet been lost. However, the time required for setting is much longer than in the case of stucco gypsum ; the latter, as every one knows, hardens in about a quarter of an hour, while the hardening of the floor-gypsum takes place only after some days, and the complete absorption of the amount of water theoretically required may take weeks. " Composition and structure of floor-gypsum. Some indications in literature suggested that in this floor-gypsum we are dealing with a basic sulphate. Accordingly a commercial floor-gypsum was first analyzed; it showed 38.6% CaO, 54.3% SOs, and foreign ingredients. " From this the ratio of CaOiSOs in molecules is found to be 1.01:1. The product therefore is calcium sulphate anhydrite without any notable surplus of lime. The fact that but slight quantities of lime are present was also demonstrated in another way: 1.134 gr. of the substance was boiled for 3^ hours with 75 ccm. of 0.425 normal potassium hydroxide, whereupon everything went into solution except slight impurities. To titrate back, 47.8 ccm. of potassium hydroxide were then required, which indicates a content in free lime corresponding to 125 (75 + 47.8) = 2.2 ccm. in weight 0.026 gr. or about 2 per cent of the whole. " Having ascertained this composition, it only remained to answer the question as to the relation which the floor-gypsum bears to the two known modifications of anhydrous calcium sulphate, namely, the natural and the soluble anhydrite. The difference between these two lies in this, that the former is practically incapable of binding water, that is to say, it does so with extreme slowness, and thus acts like dead- burned gypsum, while the soluble anhydrite takes up water even more rapidly than stucco gypsum does. "A first hint was obtained from the microscopic examination. As shown by this, the floor-gypsum consists for the most part of distinctly * Van't Hoff and Just, G. Der hydraulicshe oder sogennante Estrichgips. Sitzungsberichte der Kgl. Preuss. Akad. der Wissenschaften, 1903, B. I, pp. 249-258. f For this translation the writer is indebted to Dr. Robert Stein, formerly of the U. S. Geological Survey. 70 CEMENTS, LIMES, AND PLASTERS. recognizable needles; these have the form of the crystalline plaster of Paris (CaSO^H^O. As floor-gypsum contains no water, the needles are evidently pseudomorphs after plaster of Paris. > It may be added that stucco gypsum, though consisting essentially of half hydrate is far from showing the crystalline development of the floor-gypsum. "In view of this striking crystalline < condition, we directed our in- vestigation not only to the influence of the burning temperature but also to the possible influence of this factor (crystalline form). "To get an idea of the hardening property, observation is naturally directed first to the process as carried out in practice, and we did not fail to follow it on a small scale. However, we soon felt the need for a process which would enable this hardening to be traced with greater accuracy quantitatively. For this purpose we use'd the change of volume in hardening and the weighing of the quantity of water taken up. "1. Change of volume on hardening. When the practicability of the volume process was first tested in the case of stucco gypsum, a great irregularity manifested itself, which was not observed in the case of floor-gypsum. Hence we might pass those preliminary experiments without mention did they not furnish the explanation of the well-known fact that the hardening of gypsum, though accompanied by contraction, nevertheless may lead to the breaking of the vessels in which it takes place. "The apparatus consisted of a small flask with narrow neck; the sample of gypsum, haying been weighed, is introduced, and then a closed capillary is sealed to it, which at a short distance above the flask bears a lateral branch with stop-cock. By means 6"f this the apparatus is connected with the air-pump, and, after evacuation, is placed in a ther- mostat filled with gypsum-water at 25. The stop-cock is opened and gypsum-water is allowed to flow in. Gpysum-water is used instead of ordinary water, in order to avoid the change of volume accompanying the solution of the gypsum. After closing the stop-cock and breaking off the upper end of the capillary, the reading of the level in the latter may be performed by means of a millimeter scale one minute after the introduction of the gypsum-water. " The following table shows the result of such an observation: Time in Minutes. Height in Millimeters. Time in Minutes. Height in Millimeters. 2 540 38 374 5 522 41 375 8 486 44 376 11 448 47 377 14 408 50 378 17 385 53 380 20 374 56 381 23 371 59 383 26 370 62 387 29 370 65 388 32 371 166 413 35 372 870 438 FLOORING-PLASTERS AND HARD-FINISH PLASTERS. 71 "On the whole, therefore, a contraction takes place, as is required, in fact, by the molecular volumes at 25: Molecular Weight. Specific Gravity. Molecular Volume. CaS0 4 H 2 O. . CaSO 4 2R~O 145.17 172 19 2.75 2 32 52.79 74 22 H 2 O 18 016 997 18 07 Hence for J7= -5.68. However, after a rather large contraction (from 540 to 370) a transient expansion (from 370 to 438) manifests itself. " Although this peculiar behavior suggested the transient formation of an intermediate product, yet all attempts to isolate such a product were in vain. The explanation is found in another way: the solubility of the stucco gypsum (CaSO^-J^O) is much greater than that of normal gypsum (according to Marignac one part of CaS04 in the first form dissolves in 110 parts of water, while of the second form one part dissolves only in 479 parts of water at 24) ; thus the solution first formed is over- saturated with gypsum and precipitates the latter. Now, as shown by direct experiment, this process is accompanied by considerable expansion; in the same apparatus, when 0.54 gram of gypsum was thrown down from the o versa tura ted solution, there was a rise of 78 mm. in the scale, that is to say, quite comparable with the above value (conversely the solution of gypsum was accompanied by a contraction). Undoubtedly connected with this process is the sweating-out, which is occasionally observed in casts of stucco gypsum. While, therefore, the volume method of ascertaining the process of the hardening of stucco gypsum is to be rejected, a satisfactory regularity appears in the case of floor-gypsum, as shown by the following table : Time. Level in mm. Time. Level in mm. 1 minute 405 8 days 6 hours 322 6 minutes 395 9 ' 22 314 11 " 393 11 6 309 16 392 14 301 45 " 390 15 4 297 100 " 388 15 22 295 430 " 383 17 6 292 1 day hours 370 18 7 290 1 " 7 365 18 23 288 3 days ' 349 20 285 4 4 < 342 21 282 5 " ' 33S 22 8 '279 7 "" ' 327 53 243 1 72 CEMENTS, LIMES, AND PLASTERS. "The observed diminution of volume corresponds to expectation, regard being had to the molecular volumes at 25: Molecular Weight. Specific Gravity. Molecular Volume. CaSO 4 136.16 2.97 45 85 CaSO 4 2H O 172 19 2 32 74 22 H 2 0. . 18.016 0.997 18.07 j* Hence for CaS0 4 + 2H 2 = CaS0 4 2H 2 0, J7=-7.77. "The observed diminution of volume satisfied this requirement even quantitatively, inasmuch as 11.63 grams of floor-gypsum caused an expansion of 0.63 c.c., that is to say, 0.054 per gram, while calculation .gives 0.057. It may be added that the water content of the mass formed (20.6 per cent) corresponded to the total transformation into gypsum (20.9 per cent). " With the aid of this method we traced the influence which the burning temperature has on the setting capacity. In particular we investigated the contradictory statements whether the formation of the gypsum capable of setting takes place at a temperature higher than that at which dead-burning ensues or whether with rising tem- perature the binding capacity is gradually lost. This determination is important, for according to the above statement a dead-burned gypsum, and probably also the natural anhydrite, would acquire binding capacity by appropriate burning. Let it be stated at once that, so far as our observations go, the binding capacity decreases regularly with increas- ing burning temperature. " In this respect we had previously found that an anhydrite pre- pared at 100 hardens even more quickly than stucco gypsum. A real burning can only be spoken of above 190, for it is only above that temperature that the water develops out of half-hydrate at a tech- nically utilizable rate. Hence we first of all heated samples of well- crystallized half-hydrate (obtained from gypsum with nitric acid) at 200 and 300 for ten hours each. The volume experiment showed in. the case of the latter sample a decrease of binding capacity: 200 ; 2 . 624 gr. Capillaries : 1 mm. = . 00382 c.c. Time 1 min. 6 min. 26 min. 85 min. 172 min. 23 \ hrs. 46 hrs. 00 Level in mm. 491 481 472 464 463 459 456 456 300; 2 gr. Capillaries: 1 mm. =0.00323 c.c. Time 1 min. 20 min. 28 min. 86 min. 122 min. 7 hrs. 21 hrs. 00 Level in mm. . 506 500 499 496 495 493 487 472 FLOORING-PLASTERS AND HARD-FINISH PLASTERS. 73 " It appears that the gypsum heated at 300 sets more slowly than that which has been heated at 200; after 23 hours, for example, 93 per cent were set in the former case, in the latter after 21 hours only 56 per cent. " As these results rendered it probable that the floor-gypsum is not a product of a temperature higher than the dead-burning temperature, but that dead-burning ensues only after the formation of floor-gypsum, we heated a floor-gypsum for 10 hours at 400. The setting capacity had been thereby considerably decreased: Unheated sample, 2 gr. Capillaries: 1 mm. =0.00398. Time 1 min 8 min 44 min 6 hrs 93 hrs 6 dys 1Q Hv b- CO OS 1>- 1> to o o 00 00 ^1 0, 1C 00 4 s * CO CO O CO to CO OS oT O ^ H/i rH t>* OS b OS CO tO rH O t^ CN^ CO^ o" tc" O CO tC CO o o OS OS -1 w OS O oo 00 1C TH" co" 1> 00 CO iO O5 i ' TH 1C 1C CO O CO TH O o o 00 OS CO OS OS OS OS C-1 CO Tfl tC I> ^ M a oo j* S O CM ^o'H ! ^ ^ StM-rt tT r^ oo^ tc^ CO TH TH CO r^ co co tc i S bt of CO (30 EC* >(!, < 5 ^ 5 CO 00 CO CO rH Tf rH rH CO O rH 1C oo oo tC 00 I> O u C* rH rH rH CO c^ co_ co^ co" i>~ co" "tf rH O CO O5 co" ic" o os tC OS OS 00 rH CO ^ o 2 OS {, ^ 2. ^ g 2 1 t*- OS co * r-T CO" l>" t> os oc H^ 00 CO OS OS 1C CO CO tO 1C l> CO C rH *s !>_ ^ _ o oo" oo a IS^ ^ CO^ CO_ TH^ t>^ CO^ TJH co" i>^ co" to" oo" to" CO TH (N OS T-" CO CO CO 00 b- 1C -1 N 00 CO 8 3 CO rH CO OS CO (N tf b- 00 "T l> O ^ 9 oo CO CO rH rH C *C 04 CO rH OJ rH Tf Tfrl rH OO OI OI rH CO O Oi CO CD C CO -H CO Ol 1C rH O C 1C CO CO CO CD ^ 01 Ol CO i ! t^ co t^ ooa t i> CD co oo o CD co o ^rHO^t- Oi co" S ~ O CO 1> O Oi rH OO O l> O *C >C rH 00 T^I 1 i g CO Ol 00 rH CD 1C C 1C J> !> O4 CO rH rH ii CO CD 00 Ol Ol l> i Blaster ai iris. 1 oToT^ rH O5 CO |> Isisssaa. CO 2 00 CO Oi 01 O rH CO Ol I> CO CM rH i So rH ^ & l lCJ>-I>-OJrHTtliC -^ rHOOOrHCDOCD CO 1* A - G 3"c 2 CO rH 1C CD Tf 1C CO cDr^ooicocoqco -*^ O4 00 CO CD rH CO t^* 1 i 3S S| 51 o Jb-O^cD^Oi 1 02 1 ll g O CDrHCOOlCDOQrH co K Oi 1 1 1 a 6 * S^rHrHCOlCOq Oi 1 1 ii 2 OOrHrHCCCDCD O (N 1C OC O? CO 1C CD *> OOOtCOOarHrH w pq co" a 81 O O GO CO Cb O^ O^ ^O ^"^ g O^rH g oo W Ol" 3 2| |S32|SKSS i o rH 5 cc W i 1 O Ol Oi Ol OI O O ** ^ Oi O !> ^C OQ Oi CO rH TJH" cD"l>TtC oo" co oq i> 1C I rH TJ 1. 1 o : ^^g : ic" CD CD CO" | -^ CD" .1 p 1 02 C tC O D Oi O rH O .2 o o i> o c o ^a O rH Oi Tfl 00 Ol rH 8 1 r c 02 C I> rH O} rH ' O _o ic co oi oo c CO CO O Oi rH rH TV] * o 3 o co" O4 oo"t^ 01 ^ rH rH CO CO 21 Q g o c fe CD . Tfi CO C . rH _ rH . rH O5 . CO i 3 CO CD O CO t^ O 1C 00 Oi ^f 00 O 00 H ;oo o oo ; o Oi H g 1 3 &$ rH 1C rH M flj -o 1 CD .rH O 1C . g a 1 s 03 O IN. 00 >C TfH 00 O 3 CO O CD O Oi O ^ CO Ol 1C CO CD 00 Ol rH Oi CO c 3J CO 1> IT 1 - CO ' O CD 1C 00 C O ^ CO 'CiOO rH rH * 1 I 1-1 OTOQ Oi s $ 1 J N : CD i 02 00 o fe o "3; >, iillllSl I O $ >* 1C CD OJ O 1C CD ^ LCOt^05COrHCO 00 43 5 ^ co co i>- o> r^ Oi i i f : M i ^ rHCD>COC^OO CD i ( ^ rH CO01 rH rH CO rH o | c 1 m 00 p OH p . PH 3 .!.... ! i "Sc * ^ vX 03 .^ ou ,> i 1 5 . * j j i I 1 5 California, Ohio, and "V Colorado and Wyomir Iowa, Kansas, and Te: Michigan New York Oklahoma Other states i 1 i I ! California, Ohio, and A Colorado and Wyomin Iowa, Kansas, and Te: Michigan New York Oklahoma Other states i STATISTICS OF THE GYPSUM AND PLASTER INDUSTRIES. 83 Production by States, 1890-1901. In Table 27 the amounts and value of the gypsum production, arranged by states, is given for the years 1890 to 1901 inclusive. During this period much of the growth in the gypsum industry took place in the Western States; while Texas and Oklahoma entered the list of producers. More recently the gypsum industry has renewed its activity in New York and Michigan. TABLE 27. PRODUCTION AND VALUE OF GYPSUM BY STATES, 1890-1901. State. 1890. 1891. 1892. Quantity. Value. Quantity. Value. Quantity. Value. California Colorado Short tons. 4,249 4,580 20,900 20,250 74,877 32,903 12,748 2,9CO $29,178 22,050 47,350 72,457 192,099 73,093 87,533 7,750 Short tons. 3,000 4,720 31,385 40,217 79,700 30,135 9,123 3,615 $36,360 19,400 58,095 161,322 223,725 58,571 36,586 9,618 Short tons. 1,500 12,000 46,016 139,557 32,394 13,275 1,926 2,600 6,991 $1,500 28,500 195,197 306,527 61,100 49,521 8,640 16,300 28,207 Iowa Kansas Michigan New York Ohio South Dakota. . . . Texas Utah 3,000 5,959 1,992 15,000 22,574 6,200 Virginia 6,350 3,238 20,782 22,231 "Wyoming Total 182,995 574,523 212,846 647,451 256,259 695,492 State or Territory. 1893. 1894. 1895. Quantity. Value. Quantity. Value. Quantity. Value. California Short tons. Short tons. 6 895 $30 4,800 Short tons. 5,158 1,371 13,100 25,700 72,947 66,519 33,587 21,662 6,400 10,750 2,134 5.800 375 ' $51,014 8,281 46,125 36,600 272,531 174,007 59,321 71,204 20,600 36,511 11,484 17,369 2,400 Colorado .... Indian Territory. Iowa 21,447 43,631 124,590 $55,538 181,599 303,921 17,906 64,889 79,958 175 31,798 20,827 1,300 4,295 6,925 1,920 8,106 312 44,700 301,884 189,620 1,820 60,262 69,597 7,500 16,050 27,300 12,225 24,431 1,500 Kansas Michigan Montana New York Ohio 36,126 11,646 65,392 39,884 Oklahoma South Dakota. . . . Texas 5,150 4,011 12,550 13,372 Utah Virginia 7,014 24,359 Wyoming Total 253,615 696,615 239,312 761,719 265,503 807,447 CEMENTS, LIMES, AND PLASTERS. TABLE 27 (Continued). PRODUCTION AND VALUE OF GYPSUM BY STATES, 1896-1901. State or Territory. 1896. 1897. 1898. Quantity. Value. Quantity. Value. Quantity. Value. Arizona Short tons. Short tons. ^ 30 : 351 1,575 10,734 29,430 54,353 94,874 425 33,440 18,592 $250 2,774 10,305 40,050 64,900 189,679 * 193,576 2,3CO 78,684 50,856 Short tons. 30 3,8CO 165 24,733 59,180 93,181 1,123 31,655 21,303 3,150 150 2,740 34,215 2,610 8,378 5,225 $700 24,977 726 45,819 191,389 204,310 7,272 81,969 61,884 12,000 450 9,200 58,130 10,080 23,388 22,986 Califorina 1,452 1,600 8,000 18,631 49,435 67,634 385 23,325 22,634 $11,738 10,547 24,000 34,020 148,371 146,424 1,940 32,812 63,583 Colorado Indian Territory. . Iowa Kansas Michigan Montana New York Ohio Oklahoma Oregon South Dakota. . . . Texas Utah 6,115 16,022 2,866 5,955 200 20,OCO 48,070 13,CCO 17,264 975 8,350 24,454 2,700 6,374 3,3CO 19,240 65,651 13,500 16,899 7,200 Virginia Wyoming Total 224,254 573,344 288,982 755,864 291,638 755,280 State or Territory. 1399. 1900. 1901. Quantity. Value. Quantity. Value. Quantity. Value. Arizona Short tons. 47 2,950 871 12,000 75,574 85,046 144,776 582 52,149 $1,200 14,950 3,904 26,000 296,220 247,690 283,537 3,698 105,533 Short tens. 35 3,280 967 6,500 184,600 48,636 129,654 1,025 58,890 1,000 39,034 18,437 550 2,050 80,622 2,397 11,940 4,845 $900 10,088 5,300 15,000 561,588 150,257 285,119 7,980 150,588 4,805 119,946 60,380 1,710 13,800 192,418 4,984 18,111 24,229 Short tons. 3,550 13,291 63,653 69,390 185,150 119,565 15,930 80,376 15,236 4,103 63,547 $4,200 64,772 160,788 213,260 267,243 241,669 66,031 255,288 45,144 11,663 176,583 California Colorado Indian Territory. . Iowa Kansas Michigan Montana .... New York Ohio 27,205 11,526 550 550 53,773 2,352 11,480 4,804 73,520 36,600 1,895 4,000 125,000 10,240 32,043 21,050 Oklahoma Oregon. South Dakota . . . Texas Utah Virginia . Wyoming Other states . . . . Total ' 486,235 1,287,080 594,462 1,627,203 633,791 1,506,641 Imports of Gypsum and Plaster. Table 28 contains statistics relative to the amount of gypsum, etc., entering the different customs districts. STATISTICS OF THE GYPSUM AND PLASTER INDUSTRIES. 85 oo * II g SMrf 3 o B fi S ft P o ^ o fa / w" 2 o ^2 1 i* g ag S |5 I I - 00 00 I-H Tf CO O O CO 00 i I b- id CO CO iO CO >> g O CO CO CO O TF (N 00 rH O CO O (M O O'* r- rtOft^rUS 1-1 CO O* ) g rH O^ 1^ ^ O t^* "^ O^ O 1 00 CQt**GCQQ* co"^" CO CO CO ' OOOOO50COO COCOCOrHO500r-i CO O O CO1> OCO COI-H COiO~rH 10 CO . CO O O O O5 1C (M c^ 'coojcocooico Tt O5 Tf Oi 00 1> CO O (N CO-^ 1 00CO2, magnesium carbonate contains 47.6 per cent MgO plus 52.4 per cent CO2- The bulk composition of the original limestone may, therefore, be expressed quantitatively as follows: 60 Ibs. lime carbonate- .............. " {aMO ^ CO? 40 " magnesium carbonate = .. ..' 1 \J . V/O The original rock, therefore, carries in 100 Ibs. 33.60 Ibs. of lime (CaO), 19.04 Ibs. of magnesia (MgO), and 47.36 Ibs. of carbon dioxide (C0 2 ). If a rock of this composition be burned, the carbon dioxide will be driven off, as in the case of a pure limestone, but the solid mass remain- ing will consist partly of lime (CaO) and partly of magnesia (MgO). In addition to this difference, a difference in loss of weight is to be noted. In discussing the burning of a pure non-magnesian limestone it was stated that the driving off of the carbon dioxide meant the loss of 44 per cent in weight. In the case of the particular magnesian limestone here discussed it can be seen that the expulsion of the carbon dioxide is equivalent to a loss of 47.36 per cent in weight. Dissociation, in the case of a magnesian limestone, appears to be effected at a somewhat lower temperature * than when a non-mag- nesian limestone is burned, but no accurate data on this point are avail- able. Classification of limes. For commercial purposes limes carrying less than 5 per cent of magnesia can be marketed as pure or high-calcium limes; but those containing over 5 per cent differ so markedly in their properties that it is necessary to class them separately. The groups are, therefore, as follows: GROUP A. High-calcium limes: Limes containing less than 5 per cent of magnesia. The limes of this group differ among themselves according to the amount of silica, alumina, iron, etc., contained. A lime carrying less than 5 per cent of such impurities is a "fat" or "rich" lime, as distinguished from the more impure "lean" or "poor" limes. * Probably about 600-700 C. 98 CEMENTS, LIMES, AND PLASTERS. GROUP B. Magnesian limes: Limes containing over 5 per cent (usually 30 per cent or over) of magnesia. These limes are all slower slaking and cooler than the high-calcium limes of the preceding group, and they appear to make a stronger mortar. They are, however, less plastic or "smooth", and in consequence are disliked by workmen. As com- mercially produced, they usually carry over 30 per cent of magnesia. Methods and Costs of Lime-burning. . Compared with the complicated processes employed in the manu- facture of Portland cement, lime-making is a very simple industry, the only distinctive operation requiring attention being the burning of the limestone. In the present section the types of lime-kilns employed at various localities will be considered, detailed descriptions of several of the more important styles being given. A few brief notes on the utilization of a hitherto practically unused by-product (carbon dioxide) will then be given, after which the costs of lime-manufacture will be considered. Heat requirements in lime-burning. In burning limestone to lim- heat is required for three purposes: (a) Evaporating any water contained in the limestone. (6) Heating the limestone to its dissociation temperature, (c) Driving off carbon dioxide from the lime (and magnesium) carbonate. The water in the limestone, however, aids in the dissociation, so that the first heat requirement may be neglected here. Heating the limestone from the air temperature (say 60 F.) to its point of disso- ciation (about 1300 F. for non-magnesian limestones), assuming that within this range the specific heat of limestone is 0.22, would require, for one ton of limestone 2000 X 0.22 X (1300 -60) = 545,600 B.T.U. For a magnesian limestone, which loses its carbon dioxide at a lower tem- perature, this amount would be considerably reduced. The heat used in the actual dissociation is known quite accurately. One pound of lime carbonate (CaCO 3 ) will require 784 B.T.U. for its dissociation, while one pound of magnesium carbonate would require only 381 B.T.U. Using, then, these data in connection with those quoted in the preceding paragraph, the following tabulated statement can be made concerning the total heat requirements in burning lime- stones of different composition. In actual practice, of course, the fuel consumption is always far in excess of these theoretical requirements. LIME-BURNING. 99 TABLE 33. HEAT AND FUEL THEORETICALLY REQUIRED IN BURNING ONE TON OP LIMESTONE. Composition of Limestone. 100% CaC0 3 . 80% CaCO 3 , 20% MgCO 3 . 50% CaCO s , 50% MgCO 3 . Heat required for heating to dissociation-point 545,600 B.T.U. 1,568,000 " 457,600 B.T.U. 1,406,800 " 369,600 B.T.U. 1,165,000 " Heat required for actual dis- sociation Total heat requirements .... Coal theoretically required (14,000 B.T.U. per lb.): Intermittent kilns 2,113,600 B.T.U. 151 Ibs. 112 Ibs. 1,864,400 B.T.U. 133 Ibs. 101 Ibs. 1,534,600 B.T.U. 110 Ibs. 76 Ibs. Continuous kilns . The above results are to be regarded only as approximations to the truth, because the chemical data on which, the calculations are based are not accurately determined, but the figures suffice to show the great economy in fuel consumption which comes from the use of highly magnesian limestone. Types of lime-kilns. The types of kilns employed in lime-burning may be grouped as follows: Intermittent kilns Continuous kilns xr ' , , ., . . . , C limestone and fuel fed ) Vertical kiln, mixed leed. . . { . ,, , V ( in alternate layers ) Vertical kiln, separate feed . . . limestone and fuel not in contact (3) Ring or chamber kiln (4) (1) (2) (1) Intermittent kilns are those in which each burning of a charge constitutes a separate operation. The kiln is charged, burned, cooled, and the charge is drawn; then the kiln is again charged, and so on. The disadvantages of this intermittent mode of operation are obvious; and kilns of this type are consequently employed only where there is a slight or very irregular demand for the product. Old kilns of this kind can still be seen in farming regions, where charges of lime are burned as the neighborhood demand requires. These primitive kilns were * " rudely constructed of stone, and were located on the side of a hill, so that the top was easily accessible * Blatchley, S. W. 29th Ann. Rep. Indiana Dept. Geology, pp. 225-227, 1904 100 CEMENTS, LIMES, AND PLASTERS. for charging the kiln with stone and the bottom for supplying fuel and drawing out the lime. In charging, the largest pieces of limestone were first selected and formed into a rough dome-like arch with large open joints springing from the bottom of the kiln to a height of five or six feet. Above this arch the kiln was filled from the top with frag- ments of limestone, the larger pieces being used in the lower layers, these being topped off with fragments of smaller size. A wood fire was then started under the dome, the heat being raised gradually to the required degree in order to prevent the sudden expansion and con- sequent rupture of the stones forming the dome. Should this happen, a downfall of the entire overlying mass would take' place, putting out the fire and causing the total loss of the contents of the kiln. After a bright heat was once reached throughout the mass of stone, it was maintained for three or four days to the end of the burning. This was indicated by a large shrinkage in volume of the contents of the kiln, the choking up of the spaces between the fragments, and the ease with which an iron rod could be forced down from the top. The fire was then allowed to die out and the lime was gradually removed from the bottom. It was in this manner that all the lime used in Indiana for many years was burned, and in some localities these temporary intermittent kilns are still in operation. The process of- burning is simple and cheap, the only expense being for blasting the stone and preparing the fuel. Possibly but one or two kilns were necessary to supply a neighbor- hood for a year. These were operated for a week or two when required and remained idle for the remainder of the year. "As the population increased, the demand for lime became greater, and in many places permanent kilns lined with fire-brick were erected. These were the old-fashioned stone ' pot-kilns ' of a quarter century ago. On the inside they were usually circular in horizontal section, tapering slightly, by a curve both up and down, from the circle of largest diameter, which was from 4 to 6 feet above the bottom. A kiln 10 to 11 feet in greatest diameter was 25 to 28 feet high, 5 to 6 feet in diameter at the top and 7 to 8 feet at the bottom. There was an arched opening on one side at the bottom 5 or 6 feet high, through which the wood was introduced and the burnt lime removed. A horizontal 'grating on which the fire was built was usually placed 1 or 2 feet above the bottom. "In all these intermittent kilns there was an enormous loss of heat at each burning, for the quantity of fuel necessary to raise the contents of the kiln and the thick stone and brick walls to the necessary degree of heat had to be repeated each time the kiln was charged. Moreover, LIME-BURNING. IQl the stone nearest the dome arch in the kiln was liable to become in- jured by overburning before the top portions of the charge where tor- oughly calcined." (2) Vertical kilns with mixed feed. In kilns of this type the lime- stone and fuel are charged into the kiln in alternate layers. As the burning progresses burned lime is drawn from the bottom of the kiln, while fresh layers of limestone and fuel are added at the top. The advantages of mixed-feed kilns, as compared with the separate feed-kilns described below, are (a) that they are cheaper to construct, (b) that they are somewhat more economical of fuel, and (c) that they give for the same size of kiln a larger output in the same time. These advantages are partly counterbalanced by the disadvantages to which they are subject, these being (a) that the burned lime is discolored to some extent by its contact with the fuel, (6) that the ashes of the fuel cannot readily be separated from the burned lime, thereby lower- ing the quality of the product, and (c) that a part of the fuel ashes may clinker on the outside of the lumps of lime, preventing even and satisfactory burning. To sum up the advantages and disadvantages: the "mixed-feed" kiln is cheaper both in first cost and in operating expenses ; its product is good enough for most ordinary purposes, but is not as evenly burned or as white as is the product of a "separate-feed" kiln. At a small Pennsylvania lime-plant three vertical mixed-feed kilns are in use. Each of these kilns takes about 24 tons of stone per day, requiring the services of six quarrymen to keep the three kilns supplied. Bituminous slack is used for fuel, the consumption being 26 Ibs. slack per bushel (75 Ibs.) of lime, equivalent to a fuel consumption of 34.7 per cent on the weight of lime produced. This ratio if correct is enormous compared to natural- or even Portland-cement plants, and points to unusually inefficient management. As a general rule, a vertical mixed- feed kiln may be expected to produce lime with a fuel consumption of from 15 to 25 per cent of the weight of clean product. The cause of this apparently high consumption is that so much of the product is usually unfit for use. The Aalborg or Schofer kiln, one of the best types of stationary kilns for cement practice, has been employed in a somewhat modified form for burning lime and hydraulic lime. The lime-kiln of this type is shown in Fig. 21. The limestone is fed in at the charging door B, while the fuel is charged through the chutes //. The mass of limestone in the preheating chamber D is dried, heated, and partly decarbonated before it enters the burning-zone, when the decarbonation is complete. CEMENTS, LIMES, AND PLASTERS. The cooling-chamber C reduces the temperature of the burned lime and incidentally heats the air which passes through it to supply com- bustion. These features make the Aalborg kiln very economical in fuel consumption. Each kiln will turn out 15 to 20 tons per day, and LIME-BURNING. 103 will use 220 to 260 pounds of coal per ton of quicklime burned, equiva- lent to a fuel consumption of 10 to 12 per cent on the weight of the product, which is very close to the theoretical minimum. (3) Vertical kilns with separate feed. Kilns of this type, which are now used at most of the larger lime-burning plants, are equipped with separate fireplaces to carry the fuel, distinct from the body of the kiln. FIG. 21. Aalborg kiln for lime-burning. These fireplaces may be set either in the wall of the kiln, the usual posi- tion when a stone-walled kiln is used, or outside of the kiln-shell. The kiln body proper contains the charge of limestone, while the fuel is fed and burned in these fireplaces or furnaces. The limestone, therefore, does not come into direct contact with the fuel, but only with the hot fuel gases. Other things being equal, kilns of this type could not show quite as high a fuel efficiency as kilns in which the limestone and fuel 104 CEMENTS, LIMES, AND PLASTERS. are charged together in alternate layers. The product, however, is of a much higher grade, for it is not discolored by contact with the fuel, and it contains no fragments of unburned fuel or fuel ashes and clinkers. With average care in feeding and burning, it is probable that at least 90 per cent of the product from a kiln of this type will be a well-burned clean white lime, as compared with the 75 or 80 per cent obtainable from mixed-feed kilns. As the fuel-burning apparatus is entirely dis- tinct from the body of the kiln, the : firing can be kept under better control, so that the percentage of underburned and overburned material in the product should be materially decreased. Kilns of this type are commonly 35 to 50 feet in' height and 5 to 8 feet in inside diameter, with either two or four fireplaces or "furnaces". The Keystone kiln, described in detail below, may be taken as fairly representative of this type of lime-kiln. The Keystone kiln is built on BroomelPs patent by the Broomell, Schmidt & Steacy Company, of York, Pa. Its construction can be clearly seen from Fig. 22, which shows the kiln with a portion of the shell cut away to exhibit the interior, and with the side wall of the fur- nace removed to show its construction. The kiln from top to floor is a heavy steel shell, lined with fire-brick. The base of the kiln below the firing platform is made from very heavy steel plates, reinforced on the inside by numerous stiffening-ribs. The furnaces are carried on steel platforms which extend a sufficient distance in front of the firing-doors to give a convenient working space. In addition to being supported at the inner ends by attachment to the shell of the kiln the steel beams which floor the platforms rest at their outer ends on steel columns. In operating this kiln the flame from the coal (or wood) burned in each furnace is directed through two' large openings in the kiln shell and lining directly against the limestone which fills the kiln. These openings, as well as the kiln shell and the furnaces, are lined with fire- brick. As the lime passes the burning-zone it falls into a " cooling -cone " made of steel plates. This is an inverted hollow frustum of a cone suspended from a heavy cast-iron plate, which in turn is supported by gusset plates riveted to the base of the kiln. The cooling-cone varies from 6 to 6 feet in diameter at top, according to the size of the kiln, and is 7 feet high. The burned and partly cooled lime is drawn from the cone by means of shears or draw-gates at its bottom. These gates are operated by hand- wheels which project outside of the kiln base, thus removing the operator from the dust and heat of the lime. The LIME-BURNING. 105 FIG. 22. Keystone lime-kiln. 106 CEMENTS, LIMES, AND PLASTERS. lime can be discharged into a car run in under the cooling-cone or on the floor. The heated air which ordinarily would accumulate around the cooling-cone is discharged into the ash-pit under the grates, which adds considerably to the efficiency of the furnace. Arrangement can also be made for placing a steam-jet in the hot-air passage so as to pro- vide forced draft if desired. v TABLE 34. DIMENSIONS OF KEYSTONE LIME-KILNS. No. l. No. 2. No. 3. Outside diameter of shell 10 ft ll-i ft 12 ft Inside diameter of lining 5 " 6 " 6*" Total height 38 " 43 " 48 " Output per day, in bushels of 70 Ibs . . . 200-250 250-300 300-350 Many patents have been taken out to cover improvements in various details of the ordinary lime-kiln. One of these patented devices is shown in Fig. 23, where boilers are inserted in the kiln arches so as to utilize the waste heat of the kiln. The boilers, in turn, are used to develop the power needed for running drills, hoists, and other machinery in the quarry and mill. (4) Ring or chamber fains. Chamber kilns of the Hoffman type, though never used in America, are in somewhat extensive use in Europe for both lime and Portland-cement burning. They are described briefly on pages in connection with the burning of Portland cement. When used for burning lime in Europe, a fuel consumption of 400 to 450 Ibs. coal per ton (2000 Ibs.) of burned lime is attained in common practice, while lower consumption can be expected under favorable conditions. The Hoffman kiln is, of course, a great improvement in both economy and quality of product on the old style of vertical kiln, but it is doubtful if it gives better results than the modern kilns now in use in the more important American lime-plants. Gen. Q. A. Gillmore stated * in 1871 that a Hoffman kiln used for lime-burning at Llandulas, Wales, produced about 80 tons of lime per day at the following detailed cost: Cost of quarrying stone, including tools $0 . 31 ^ Charging kiln . lo| Drawing kiln 07 Wages of burners \\ Q7 1 Fuel at SI . 75 per ton '.'.'.'.'. Q. 37* Managing expenses, etc .'.'.' 0.31} Cost of lime per ton $1 . 25 * Gillmore, Q. A. A practical treatise on Coignet-Beton and other artificial stone, pp. 71-72, 1871. LIME-BURNING. 107 LJ t I f ) I I l l < _LJ_L_ / f t 1 I 1 I ( i I FIG. 23. O'Connell lime-kiln. 108 CEMENTS, LIMES, AND PLASTERS. He further estimated that at the current prices (1871) of labor and fuel in the United States, lime could be manufactured in a Hoff- man kiln at about $2 per ton, the following details of cost being given : Cost of quarry and plant \ $20,000 Annual yield of kiln, tons 20,000 Per Ton Lime. Interest on investment ?. $0 .07 Quarrying stone . 65 Charging kiln .20 Drawing kiln 0.15 Wages of burners 0.15 Fuel . 43 Contingent expenses, 20 per cent . 33 $1.98 Utilization of carbonic-acid gas from lime-kilns. During the burn- ing of limestone to lime an enormous amount of carbonic-acid gas (carbon dioxide, 62) is driven off and usually wasted. The extent of this waste may be appreciated when it is recalled that 100 Ibs. of pure limestone would give on calcination 56 Ibs. of quicklime and 44 Ibs. of carbon dioxide. To put the matter in another way, for every ton (2000 Ibs.) of lime made 1571 Ibs. of carbon dioxide are thrown into the atmosphere. During the year 1903, therefore, over one and a half million tons of carbon dioxide were produced and wasted from the lime-works of the United States. Few attempts have been made by lime-manufacturers to utilize this valuable by-product, though the manufacture of carbonic acid, as. an independent industry, has become of great importance. Mr. Henry A. Mather states * that carbon dioxide from lime-kilns "helps make profitable a well-rounded operation in Oldbury, England, the only surviving chemical works using the Leblanc soda process. The salt cake is burned in furnaces, the soda bleached out, the chlorine recovered, and the sludge of partially decomposed carbonaceous matter containing a large percentage of sulphite of lime is treated in closed agitators with carbonic-acid gas obtained from burning lime rock in a closed kiln. The hydrogen disulphide driven off in this operation of precipitating the carbonate of lime is burned in a Klaus kiln, air in proper proportions enters with the gas, and flowers of sulphur is the final product of this part of the operation ". A further utilization of by-products at this plant occurs when the carbonate of lime so pre- * Engineering and Mining Journal, March 39, 1902. LIME-BURNING. 109 cipitated is mixed with clay and burned into Portland cement. An editorial note appended to Mr. Mather's article states that the Har- greaves-Bird alkali-works at Middlewich, England, also use waste car- bon-dioxide gas from lime-kilns. Cost of lime-manufacture. With the exception of a comparatively few large and well-managed lime-plants, lime-manufacture in the United States is not so steadily and economically handled as to give much basis for generalizations concerning costs. The result is that the data obtainable are rarely definite enough to be of much service. The follow- ing is probably as fair a statement of the case as can be made. The principal items to be considered in estimating the cost of lime- manufacture are: (1) Interest on cost of plant and quarry. (2) Cost of quarrying limestone. (3) Cost of fuel for burning. (4) Labor costs, exclusive of quarry. The interest on cost of plant and quarry will vary greatly according to the steadiness with which the plant is operated. This is, of course, true with regard to the same item in the cement industry, but lime- plants are in general subject to greater fluctuations in output. The estimates given below of interest charges per ton of lime are therefore given a very wide limit, but it is believed to be impracticable to place them more definitely. The cost of quarrying is also variable, but within narrower limits. In large, carefully managed quarries located near the kilns, and with stone and stripping so arranged as to admit of cheap extraction, the cost of quarrying the limestone and transporting it to the kiln may fall as low as 25 cents per ton. This cost is attained in Portland-cement quarries in the Lehigh district of Pennsylvania, and in a number of 'natural-cement and lime quarries elsewhere. On pages 378, 379 will be found further details as to cost of quarrying, one of the examples being of the costs at a quarry worked both for Portland cement and for lime. With average skill in locating and managing the quarry, it is probable that the cost of quarrying need never rise above 40 or 45 cents per ton of rock. Allowing for waste and loss by under or over- burning, 2 tons of limestone will be required to make 1 ton of lime. This would give as the probable limits of cost of quarrying 50 to 90 cents per ton of burned lime. Wood is still used for fuel at many lime-kilns, in which case the cost of fuel may be merely nominal or may be very high. When coal is used for fuel in a modern kiln, the coal consumption per ton of burned 110 CEMENTS, LIMES, AND PLASTERS. lime may vary from 300 to 500 Ibs. These limits have been assumed in the estimate below, while the cost of coal has been taken as varying from $2 to $3 per ton. These prices are fairly representative for most of the lime-plants of the country. Labor costs are estimated with a rather liberal maximum limit. The final results of these calculations are shown below. TOTAL COST OF LIME-MANUFACTURE PER TON. Interest on cost of plant and quarry $0 . 05 to $0 . 20 Taxes, minor supplies, etc 0.10" . 25 Cost of quarrying two tons of limestone. ... . 50 " . 90 Cost of fuel for burning 0". 30- " . 75 Cost of labor, exclusive of quarrymen 0.25 " . 80 Total cost per ton of burned lime, in bulk (2000 Ibs.) . $1.20 " $2.90 This corresponds to costs of 4.2 to 10.15 cents per bushel of 70 Ibs. The minimum estimate represents what might be attained by a good modern plant, run steadily and under exceptionally favorable conditions as regards quarrying, fuel, and labor. The maximum estimate could easily be exceeded by the small or unsteadily operated plants. The average cost throughout the entire country is probably in the neigh- borhood of 6 to 8 cents per bushel. General estimates of costs of lime-manufacture by the use of Hoff- man kilns will be found on pages 106 and 108. Actual costs of lime-manfuacture in 1900. In connection with the above estimates of cost it is of interest to compare certain statistics collected by the Census Bureau in 1900 and published in vol. 7, Reports Twelfth Census, pp. 274-277. The tables on the pages cited give total costs of various elements in lime- and cement-manufacture in all the states during 1900. As the figures for lime, natural cement, and Port- land cement are tabulated together, most of the tables are of little value for our present purpose. In the ten states considered below, however, no natural- or Portland-cement plants were in operation during 1900, so that the statistics for these states must necessarily apply only to the lime industry. The data relating to these ten states have ac- cordingly been slightly rearranged and are shown in the following table. As the total quantities of lime produced are not stated, it is impos- sible to reduce the total costs given in the table to costs per ton or bushel of product. A simple calculation, however, enables us to reduce them to percentages of the total cost, so that the relative im- portance of the various elements making up this total can be readily LIME-BURNING. Ill TABLE 35. COST OF LIME-MANUFACTURE DURING 1900 IN TEN STATED. State Arkansas. Con- Iowa Maine Massachu- necticut. setts. Number of plants 5 11 28 20 11 Total capital $53 894 $250 392 $663 830 i 040 007 2). Occurrence and origin of magnesite. Magnesite, when in bodies of workable size, occurs commonly in one of three associations, the methods of origin of the deposits being different in each case. The three types of deposits are: (1) Magnesite occurs most commonly in the form of irregular veins or pockets in serpentine or other magnesian igneous rocks. In this case the magnesite has been formed as a decomposition product aris- ing from the decay of the igneous rock. (2) Magnesite occurs in the form of beds associated with deposits of rock salt, gypsum, etc. In this case the magnesite deposit has un- doubtedly originated by direct deposition of magnesium carbonate from bodies of concentrated saline waters. (3) Magnesite also occurs in the form of beds interstratified with shales, limestones, etc. Magnesite deposits of this type are commonly ascribed to the replacement of the lime (in a limestone) by magnesia carried in by percolating waters. This may be true in some cases, but such deposits may also have originated by direct deposition, as described under (2), above. In most of the workable magnesite deposits noted below, however, the first method of origin is the true explanation. Distribution of magnesite deposits. The magnesite deposits now exploited on a sufficient scale to be of commercial interest occur in Austria, Germany, Greece, Hungary, India, and the United States. Workable deposits also occur in Canada, but as yet have not been suffi- ciently opened up to determine their commerical importance. Foreign localities. The principal Austrian magnesite deposits are near Mittendorf, in Styria, and near Tolsvar, in the province of Minsan, Hungary. The Styrian magnesite averages about 88 per cent mag- nesium carbonate with about 8 per cent of silica, alumina, and iron oxide. The Hungarian product is a purer magnesite, carrying 92 to 95 per cent magnesium carbonate, with 3 or 4 per cent iron oxide. In Germany the deposits now worked occur near Kosewitz and Frankenstein, in Silesia, and are principally worked in connection with 150 CEMENTS, LIMES, AND PLASTERS. the manufacture of carbonic acid. The product will carry about 92 to 94 per cent magnesium carbonate, the principal impurity being 4 to 5 per cent of silica. The principal Grecian deposits are on the island of Euboea, on the east coast of Greece, and also near Corinth. The product is a very pure magnesite, averaging 95 per cent magnesium carbonate. It is low in clayey matter, the principal impurity being 3 to 5 per cent of lime carbonate. The Grecian deposits are worked in primitive fashion by pick and shovel. The mines, or quarries, are usually worked as open cuts. As the rock is broken in the mines it is brought to the sur- face, where the magnesite is sorted out. It is then loaded into small carts and drawn to a narrow-gauge gravity railway, when the mag- nesite is loaded into one-ton cars and sent forward to the shipping port, usually Kymassi or St. Theodore. The cost of producing the mineral is about $3.50 per ton, transportation charges to the seaport about $1.00, and freight to the United States about $2.50 per ton. Magnesite is found in considerable quantity in southern India, about two hundred miles from Madras. Deposits recently exploited extend over 1500 acres. The railroad from Madras to Calicut runs through these deposits, near the center of the magnesite area. The material can be shipped, in any desired quantity, either from Madras on the east coast or from Beypore on the west coast. As described to the present writer by the owner, the magnesite occurs in beds or veins of varying thickness, from a few inches up to several feet, the magnesite beds being separated by bodies of disintegrated material. An analysis o.f this magnesite is given in column 1, Table 63. This was made on a 100-ton sample of crude rock. Another analysis of Indian magnesite, quoted in column 2 of the same table, accompanied a series of specimens exhibited at the St. Louis Exposition in 1904. American localities. The principal American magnesite deposits are in California and in Quebec, Canada. The California deposits are described* as follows: "The principal producing point in California is in the vicinity of Porterville, Tulare County, though a small quantity still comes from Chiles Valley and Pope Valley, Napa County. At Porterville there are several deposits. The main deposit at the opening carries a small vein, but at the end of the 240-foot tunnel the deposit is 40 feet wide, and there are said to be several million tons now in sight. At this place calcining furnaces have been erected and are in operation. The mineral crops out boldly in * Mineral Resources of the U. S. for 1903, pp. 1131-1135. 1904. SOURCES AND PREPARATION OF MAGNESIA. 151 distinct veins, having a general strike northeast and southwest, and there are spurs running in several instances at nearly right angles with the primary veins. On the surface the veins are fom 2 inches to 10 feet wide. They cover an area of over 500 acres. In Pope and Chiles valleys, Napa County, there are somewhat extensive deposits which were formerly worked, but hauling by team to railroad made them more expensive to operate than the mines at Porterville. In Placer County there is a more extensive deposit than elsewhere in California, but it is in an almost inaccessible mountain region where a very costly road would be necessary to get the product out, and the deposit has therefore not been utilized. Near Sanger, Fresno County, 7 miles from Centerville, another deposit is now being opened. A deposit has been discovered also near Walkers Pass, Kern County, but it has never been opened. There are also unutilized deposits near Morgan Hill, Santa Clara County. "The extensive deposits of magnesite on Red Mountain, at a point where Stanislaus, Alameda, and Santa Clara counties join, are now being opened by the American Magnesite Company, of Chicago, which has obtained control of the numerous claims heretofore owned by individuals. None of them have been at all thoroughly prospected as yet, though there are numerous boulders or large croppings, some from 30 to 150 feet wide, supposed to cover extensive beds beneath. The parent company is the American Magnesite Company, organized under the laws of the State of Maine, with Mr. G. Watson French, of Chicago, as president and Mr. H. C. Stillwell, of Fruitvale, Alameda County, CaL, as vice-president and Pacific coast agent; Mr. Charles H. Spinks, of Berkeley, CaL, is to manage the mines. One of the sub- sidiary companies is the Rose Brick Company, which is to manufac- ture magnesite brick, at Oakland, Cal.; the American Carbonic Acid Gas Company is another, of which Mr. John Deere is president and Mr. George A. Wayman, manager. The third corporation is the Plastic Construction Company, of which Mr. Edwin D. Weary, of Chicago, is president. This company controls the American rights for making a fire-proof construction material as well as a patent brick. This factory will also be in Oakland. "The mines of this company are nearly all in Santa Clara County, with a few in Stanislaus, near the Alameda County line. The Ala- meda County supervisors are building a wagon-road from the mines to Livermore, where the railroad is met. There are twenty-seven mining claims in the group, and several are at present being opened. Only a few car-loads for sample purposes have been shipped since the com 152 CEMENTS, LIMES, AND PLASTERS. pletion of the new organization, but it is expected that the properties will shortly be opened on an extensive scale." Recently several large deposits of magnesite have been discovered * in the township of Grenville, Argenteuil County, in the province of Quebec. Large boulders of the mineral were found, and finally the magnesite was found in place. One of these deposits, which showed an outcrop 90 feet long and 20 feet broad, is in the north half of the eighteenth lot of the eleventh range' of Grenville township. Another outcrop, 100 feet wide and traceable for a quarter mile in length, is on the north half of the sixteenth lot of the ninth range. The following analyses of magnesite from these deposits were made by G. C. Hoffmann, and are quoted from the report cited below.* TABLE 60. ANALYSES OF MAGNESITE FROM QUEBEC, CANADA. 1. 2. ' 3. 4. 5. 6. 7. Magnesium carbonate (MgCO 3 ) Calcium carbonate (CaCO 3 ). . . Magnesia (MgO) as silicate. . . . 77.62 16.07 3.50 74.68 18.89 3.71 78.08 15.57 4.18 77.16 10.78 6.14 76.09 16.00 4.29 76.97 13.14 5.87 49.71 30.14 9.17 8. 9. 10. 11. 12. 13. Magnesium carbonate (MgCO 3 ) 75 69 82 72 77 07 85 00 95 50 81 27 'Calcium carbonate (CaCO 3 ) 19 71 12 36 16 28 10 80 tr 13 64 Magnesia (MgO) as silicate 3 08 2 53 3 22 n d n d 3 66 It will be seen that this Canadian magnesite, differs from all the -other magnesites known to commerce, in that it contains a comparatively large percentage of lime carbonate as its principal impurity. Production and imports of magnesite. In the following tables (61 and 62) statistics regarding the domestic production and the imports -of magnesia and magnesite are given. It will be seen that in 1903 the value of the United States production amounted to less than 3 per cent of the total value of magnesia and magnesite used in this country. Analyses of commercial magnesite. As magnesite is simply mag- nesium carbonate, a theoretically pure magnesite would consist of 47.6 per cent magnesia (MgO) and 52.4 per cent carbon dioxide (C0 2 ). De- posits of magnesite, however, rarely yield any considerable amount of material of this degree of purity, and commercial magnesite may con- tain as high as 10 per cent or thereabouts of lime carbonate, silica, .alumina, iron oxide, etc. * Ann. Rep. Canadian Geol. Survey, vol. 13, Report R, pp. 14-19. 1903. SOURCES AND PREPARATION OF MAGNESIA. 153 TABLE 61. QUANTITY AND VALUE OF CRUDE MAGNESITE PRODUCED IN THE UNITED STATES, 1891-1903. Year. Quantity. Value. Year. Quantity. Value. 1891 Short tons. 439 $4,390 1898 Short tons. 1 263 $19 075 1892 1,004 10,040 1899 1,280 18 480 1893 704 7,040 1900 2,252 19 333 1894 1 440 10240 1901 3 500 10 500 1895 2 220 17,000 1902 2 830 8 490 1896 1 500 11,000 1903.. . 3 744 10 595 1897 1,143 13,671 TABLE 62. IMPORTS OF MAGNESITE INTO THE UNITED STATES IN 1903. Quantity. Value. Magnesia : Calcined, medicinal Pounds. 34,586 $4 412 Carbonate of, medicinal 10,569 765 Sulphate of or Epsom, salts 2 392 831 11 326 Magnesite : Calcined not purified 73 534 690 311 396 Crude 36 017 637 150 002 TABLE 63. ANALYSES OF MAGNESITE. 1. 2. 3. 4. 5. Silica (SiO 2 ) 2.20 } 0.30 0.59 46.59 49.63 0.83 0.22 0.30 n. d. 47.35 51.44 0.27 0.30 1.62 n. d. 46.00 51.23 n.d. 0.52 trace 2.25 45.28 51.61 0.34 0.52 0.08 2.46 44.96 51.44 0.54 Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) Lime (CaO) Magnesia (MgO) Carbon dioxide (CO 2 ) ,. Water 6. 7. 8. 9. Silica (SiO 2 ) 1.0 } 3.0-6.0 0.28-1.12 42.84-45.70 n. d. 4.00 4.00 n. d. 41.89 n. d. 0.8 ri.ii 13. 2/ 0.06 45.12 49.72 4.5-5.25 1.5 0.6-0.7 46.0-48.0 46.0-50.0 Alumina (Al^Oa) Iron oxide (F^Os^ I ime (CaO) Magnesia (MgO) Carbon dioxide (CO 2 ) 1. 200 miles from Madras, British India. Private communication. 2. India. Indian Exhibit, World's Fair, St. Louis, 1904. 3. Dept. of Ufa, Southern Urals, Russia. "Mineral Industry," vol. 10, p. 439. 4. Mondondi, Greece. U. S. Consular Reports, No. 168, 1900. 5. Eubrea, Greece. Proc. Inst. C. E., vol. 112, p. 381. 6. Styria, Austria. Proc. Inst. C. E., vol. 112, p. 381. 7. Styria, Austria. Eng. and Mining Journal, March 10, 1900. 8. Minsan, Hungary. Eng. and Mining Journal, March 10, 1900. 9. Frankenstein, Silesia (Germany). Eng. and Mining Journal, March 10, 1900. 154 CEMENTS, LIMES, AND PLASTERS. Effects of heatingmagnesite. If magnesite (MgC0 3 ) be strongly heated, the effect (as with lime carbonate) is to drive off the carbon dioxide (CC^), leaving magnesia (MgO) as a white solid. A curious and technologically important phenomenon connected with the tem- perature employed is to be noted. If the calcination be carried on quickly at a red heat, the magnesia resulting will have a specific gravity of 3.00 to 3.07; while if the calcination is long continued or carried on at a higher temperature the resulting magnesia will be much denser, possessing a specific gravity of 3.61 to 3.80. The technologic importance of the two forms of magnesia lies in the fact that the lightly burned magnesia will slake with water and if then exposed to air will finally recarbonate and harden slowly, just as lime does. The denser, higher-burned magnesia, however, will not take up either water or carbon dioxide from the atmosphere. Another difference of commercial interest lies in the fact that the light form of magnesite possesses a certain amount of plasticity, so that it can be molded into shape under heavy pressure, while the dense form of magnesia is entirely devoid of plasticity. Methods of burning magnesite. For calcining magnesite at low temperature, so as to obtain lightly burned magnesia, kilns closely similar to ordinary lime-kilns are employed in California. The kilns in use at one California magnesia-plant are built in the -form of a frustum of a cone, the broader part downwards. These kilns are about 19 feet in height, 3 feet in diameter at the top, and 7 feet in diameter at the base. Drawing-doors are placed at the base, while draft is obtained by suction, air being drawn through a flue near the top of the kiln. These kilns are charged with coke and magnesite mixed, in about the proportion of 300 Ibs. magnesite to 20 Ibs. of coke. The product is the light form of magnesite, and is probably not entirely decarbonated. This fuel consumption would amount to about 14 per cent on the weight of mag- nesia produced. When the dead-burnt or heavy magnesia is required, the burning must take place at much higher temperatures. This kind of magnesia may be prepared in reverberatory furnaces, in cupolas lined with silicious material, or in highly heated gas-kilns.* The practice in Greece is described f as follows : " At the Greek magnesite mines, until recently roughly built kilns fired by wood were employed for catcining the ore, which required a * Proc. Inst. Civil Engineers, vol. 112, p. 381. 1893. f Engineering and Mining Journal, Feb. 28, 1903 SOURCES AND PREPARATION OF MAGNESIA. 155 large quantity of fuel. In recent years, however, modern shaft cal- ciners have been built and a soft lignite coal is used. When calcined magnesite falls into powder and is apt to choke the lower or cooler portion of the kiln, preventing the access of air and heated gases to the upper portion. The shaft furnaces are constructed to overcome this result. The quantity of fuel required is from 15 to 20 per cent of the weight of magnesite, equivalent to a fuel consumption of 30 to 40 per cent on the weight of magnesia produced. In some cases the cal- cining is done in a double-hearth reverberatory furnace, where the flame is brought into direct contact with the freshly charged magnesite on the upper hearth, the operation being completed on the lower hearth, which is the hotter of the two." Composition of the product. The analyses given in Table 64 will serve to show the composition of the burned product, which naturally varies according to that of the magnesite from which it is made. TABLE 64. ANALYSES OF CALCINED MAGNESITE ( = MAGNESIA). 1. 2. 3. 4. 5. 6. 7. .Silica (SiO 2 ) 0.98 0.16 0.17 0.50 1.2 0.73- 7 98 Alumina (A1 2 O 3 ) 0.10 0.10 2.38 Iron oxide (Fe 2 O 3 ) 5.70 7.40 5.02 6.50 6.90 \ 13.0 0.56- 3.54 Lime (CaO) 1.88 2.66 1.50 1.70 7.3 0.83-10.92 Magnesia (MgO) 91.10 89.36 90.42 90.95 9i.50 77.6 82.46-95.36 arbon dioxide (CO 2 ). . . n. d. n. d. 0.46 n. d. n. d. 1, 2. Burned Hungarian magnesite. Iron Age, Jan. 15, 1903, pp. 20, 21. 3, 4, 5. ' Mineral Industry, vol. 10, p. 439. 6. Styrian (Austrian) magnesite. Proc. Inst. C. E., vol. 112, p. 381. 7. " Grecian magnesite. Proc. Inst. C. E., vol. 112, p. 381. Use of 'magnesite for preparation of carbonic acid, etc. California practice in the manufacture of carbonic acid from magnesite is de- scribed as follows. in a recent report :* " In the manufacture of carbonic-acid gas, the gas is extracted from the jmagnesite by calcining and the remaining calcined material is sold to the manufacturers of wood-pulp paper. The best English coke is used for calcining the magnesite. From one short ton of magnesite, after removing the gas ; they obtain about 1200 Ibs. of residue, which is partly calcined magnesite still carrying some 20 per cent of gas. In the process about 500 Ibs. of gas is obtained when finally compressed * Mineral Resources of the U. S. for 1903, p. 1133. 1904. 156 CEMENTS, LIMES, AND PLASTERS. into liquid form. For every ton of magnesite about 500 Ibs. of coke is burned, and this, containing about 97 per cent of carbon, also fur- nishes considerable gas. The steel cylinders for holding the liquid gas are & inch thick and 5 by 49 inches long, and hold about 25 Ibs. The pressure on the cylinder at 60 F. is about 850 Ibs., a three-stage compressor being used. In shipping the liquid gas through the central valleys and to Arizona the heat in the^cars sometimes runs as high as 145, the pressure being increased thereby. The cylinders containing the liquefied gas are shipped to soda-water manufacturers, ice-fac- tories, refrigerating-plants, breweries, bar-rooms, etq. The cylinders with the liquid gas are shipped all over the Pacific coast, from San Fran- cisco, even the British war vessels stationed at British Columbia using the gas for their refrigerating-plants, The San Francisco carbonic- acid-gas makers use about 1000 tons of crude magnesite annually. As stated, the wood-pulp paper-mills of California and Oregon use the calcined magnesite. They transform it into a sulphite of magnesia and use it as a digester for the wood pulp. To make this sulphite they put the material into a tank and pass sulphurous fumes through it. After being used as a digester they add a little lime and make the ' pearl hardening ' of commerce to be used as a l filler ' for the paper. " Magnesian Limestones as Sources of Magnesia. Highly magnesian limestones, approaching dolomite in composition, may be regarded as possible sources of magnesia. The general char- acters of such limestones are discussed in some detail in earlier chapters of this volume, and reference should be made to pp. 90-91 for data on these points. Occurrence of magnesian limestones in the U. S. Magnesian lime- stones are so widely distributed throughout the United States that no satisfactory summary of their distribution can be given here. On pp. 92-94 is given a list of reports on the limestones of the various states and territories. Reference to these reports will furnish data on the local distribution and composition of magnesian limestones, as well as of other types. Analyses of magnesian limestones. In the following table analyses of a number of highly magnesian limestones from various localities in the United States are presented. It will be seen that -these range from 15 to over 22 per cent in magnesia (MgO), which is about equivalent to a range of from 32 to 45 per cent magnesium carbonate (MgC0 3 ). SOURCES AND PREPARATION OF MAGNESIA. 157 TABLE 65. ANALYSES OF HIGHLY MAGNESIAN LIMESTONES, U. S. 1. 2. 3. 4. 5. Silica (SiO ) 3.24 7.75 48 08 Alumina (A-l^Oo) 0.17 / 1 n on Iron oxide (Fe 2 O 3 ) 0.23 1 1.48 10 02 j 0.20 0.25 Lime (CaO) 29.58 31.00 31 01 31 31 30 46 20 84 16 46 21 79 21 03 21 4g Carbon dioxide (CO 2 ) 45.54 42.47 47.35 46.98 47.58 6. 7. 8. 9. 10. Silica (SiO 2 ) . . ... 73 44 87 20 70 Alumina (A1 2 O 3 ) f 1 22 57 1 ^ 00 /O 95 Iron doxie (Fe 2 O 3 ) I 0.35 [ trace 25 | 0.23 \ 80 Lime (CaO) 32 73 30.73 31 40 30 04 30 50 Magnesia (MgO) . . . . 19 37 20 87 19 95 22 28 20 05 Carbon dioxide (CO 2 ) . .... 46 58 45 85 n d 47 14 45 24 1. Morrisville, Calhoun County, Ala. W, F. Hillebrand. analyst. Bulletin 60, U. S. Geol. Survey, p. 159. 2. S. 16, T. 7, R. 7, E. Talladega County, Ala. J. B. Britton, analyst. Rep. Ala. Geol. Survey for 1875, pp. 149, 150. 3. Inyo Marble Co., Inyo, Calif. 20th Ann. Rep. U. S Geol. Survey, pt. 6, p. 359. 4. East Canaan, Conn. 20th Ann. Rep. U. S. Geol. Survey, pt. 6, p. 370. 5. Canaan, Conn. J. S. Adams, analyst. 20th Ann. Rep. U. S. Geol. Survey, pt. 6, p. 370. 6. Jasper, Ga. W. H. Emerson, analyst. Bulletin No. 1, Ga. Geol. Survey, p. 87. 7. Cockeysville, Md. J. E. Whitfield, analyst. Bulletin 60, U. S. Geol. Survey, p. 159. 8. Ossining, N. Y. H. Ries, analyst. Bulletin 44, N. Y. State Museum, p. 829. 9. Tuckahoe, N. Y. Ledoux, analyst. 20th Ann. Rep. U. S. Geol. Survey, pt. 6, p. 427. 10. Gates, Monroe County, N. Y. D. H. Newland, analyst. Bulletin 44, N. Y. State Museum.. p. 796. Extraction of magnesia from magnesian limestone. Two principal processes have been suggested for extracting magnesia from magnesian limestone. Scheibler process. The mixture of lime and magnesia left by burn- ing magnesian limestone is made into a thick milk by adding suffi- cient water. Into this solution is poured water containing 10 to 15 per cent, by volume, of molasses, and the mixture is mechanically stirred. In a few moments saccharate of lime is formed, which re- mains in solution while the magnesia is precipitated. On putting through a filter-press the magnesia remains behind, while the saccha- rate of lime passes through. The composition of the magnesia sa obtained at a German plant was: Silica (SiO 2 ) ] Alumina (A1 2 O 3 ) Y 1 .47 per cent Iron oxide (Fe 2 O 3 ) J Lime (CaO) ! 2.18 " " Magnesia (MgO) , 95.99 " " The saccharate of lime which passed through the filters is now treated for recovery of its constituents. Carbon dioxide precipitates the lime as carbonate, after which it is filtered and the lime carbonate precipi- 158 CEMENTS, LIMES, AND PLASTERS. tate washed. The filtrate contains the molasses, which can be used over again. In the course of the process a loss of 5 to 10 per cent of molasses occurs. , Closson process. This process is based on the use of magnesium chloride, and is therefore of value at points such as Stassiurt, where that material is obtainable as a cheap by-product. Twenty thousand pounds of magnesium chloride is mixed with the lime-magnesia resulting from the calcination of 3000 Ibs. of magnesian limestone. Water is added to give a thick solution, and mechanical agitation is employed. The result is the formation of lime chloride, and magnesia hydrate. On passing through a filter-press the magnesia hydrate is caught on the filter, while the lime chloride passes through in solution. The hydrate is washed and then burned, giving one ton of magnesia, The magnesia obtained at Horde by this process gave the following composition: Silica (SiO 2 ) ] Alumina (A1 2 O 3 ) . . .- ( 1 .05 per cent Iron oxide (Fe 2 O 3 ) J Lime (CaO) 1.94 " " Magnesia (MgO) 96.90 " " The lime-chloride solution is then treated for recovery. The mate- rial is carried to receptacles like those in which blast-furnace gases are washed, except that revolving wheels stir the chloride, making a thorough mixture of the gases and the liquid. Two of these recepta- cles are placed together back to back. A valve which can be reversed sends the gases to either side and thus keeps up a continuous working. Into these receptacles, together with the lime chloride, is put a quantity of the lime-magnesia resulting from the calcination of magnesian lime- stone. The blast-furnace gases passing through precipitate the lime as carbonate, losing their carbon dioxide in the process, and are thus rendered more combustible. They deposit, besides, a considerable quantity of the solid materials mechanically carried by them and are thus cleaned. ' Magnesium chloride is reformed, remains in solution, and is drawn off and filtered. The entire process shows a loss of 5 to 6 per cent of magnesium chloride. Sea-water and Brines as Sources of Magnesia. Sea-water contains small percentages of different magnesian salts. In the manufacture of table salt from sea-water or salt brines, these magnesian compounds are incidentally concentrated so as to be put in more available form. SOURCES AND PREPARATION OF MAGNESIA. 159 Extraction of magnesia from sea- water.* "Magnesia is made out of sea- water, which contains about 4 Ibs. magnesium chloride per cubic yard of water, on a large scale at Aignes Morts, on the Mediter- ranean coast of France. "The sea- water is pumped into a tank made of masonry, and at the same time milk of lime is pumped in, in the proportion of 1.5 per cent of lime for every 1 per cent of magnesia. From this first tank the liquid flows into two other masonry tanks, when thorough mixing is effected mechanically. It is then filtered into shallow excavations about 1000 feet long and 16 feet wide, in the bottom of which is a bed of clean beach-sand. When enough magnesia has been collected the liquid supply is cut off and the precipitate is allowed to dry. If in summer, it is dried in the sun, taking twenty to thirty days, but in winter artificial drying is necessary." The dried magnesia is then calcined and treated as explained in discussing the burning of magnesite (p. 154), and the manufacture of magnesia bricks (pp. 160-161). References on magnesite, sources of magnesia, etc. Hoffmann, G. C. Magnesite deposits in Quebec, Canada. Ann. Rep. Canadian Geological Survey, vol. 13, Report R, pp. 14-19. 1903. Vlasto, S. J. The magnesite industry [in Europe]. Engineering and Mining Journal, March 10, 1900. .Watts, W. L. [Magnesite in Santa Clara County, California.] llth Ann. Rep. California State Mineralogist, pp. 374-375. 1893. Weiss, N. Magnesite in Hungary. Iron Age, pp. 20-21. Jan. 15, 1903. Yale, C. G. [Magnesite deposits in California.] Mineral Resources U. S. for 1903, pp. 1131-1135. 1904. Anon. Magnesite [in California]. 12th Ann.' Rep. California State Mineralo- gist, p. 328. 1894. Anon. Magnesite in Greece. U. S. Consular Report, No. 168, 1900. * Lock, C. G. W. Economic Mining, p. 331. CHAPTER XH. MAGNESIA BRICKS AND OXYCHLORIDE CEMENTS. t AFTER magnesia (MgO) has been obtained by any of the methods described in the preceding chapter, it is put to use in two quite different ways. As the products differ greatly in both composition and use they will here be discussed separately under the headings of "Magnesia Bricks " and " Oxychloride Cements ". Magnesia Bricks. Magnesia bricks, which are commonly but very erroneously called magnesite bricks in the trade, are largely used as furnace linings, etc., and have also been used to a small extent as linings for Portland-cement kilns. Manufacture of magnesia bricks. In discussing the methods and effects of calcining magnesite it was stated that two different forms of magnesia could be obtained, according to the temperature at which the calcination is carried on. If the magnesite be burned at a light-red heat, the resulting magnesia will have a low specific gravity (3.00 to 3.07), will possess sufficient plasticity to be capable of being molded into shapes, and will gradually absorb water and carbon dioxide from the atmosphere, just as quicklime would do. The result of this absorption is that this form of magnesia will finally become partly recarbonated. If the calcination takes place at a higher temperature, however the resulting magnesia will be heavy, with a specific gravity of 3.61 to 3.80; it will be absolutely devoid of plasticity; and it will not recar- bonate on exposure to the atmosphere. These differences in the physical and chemical properties of the two forms of magnesia are taken advantage of in the manufacture of magnesia bricks. Each contributes certain good qualities to the brick. Magnesia bricks are made of a mixture of the two forms of mag- nesia, in the proportions of four to six parts heavy magnesia to one 160 MAGNESIA BRICKS AND OXYCHLORIDE CEMENTS. 161 part light magnesia. The dense, chemically stable " heavy magnesia " is thus the base of the brick; the light magnesia is added to give plas- ticity to the mixture, enabling it to be molded, and also to harden on exposure to the atmosphere. From 10 to 15 per cent of water is added to this mixture, and the resulting stiff paste is pressed into form in iron molds. The brick will gradually harden on simple exposure to the air, after which it is usually made still more resistant by reburning at a low red heat. Bricks or other objects made in this manner may, if not sufficiently solid for the use for which they are intended, be hardened by dipping into a cold dilute solution of boracic acid in water. After this they should be dried and reburned. Composition of magnesia bricks. TABLE 66. ANALYSES OF MAGNESIA BRICKS. 1. 2. 3. 4. 5. 6. Silica (SiO 2 ) 35 3 45 3 10 3 4 3 2 2 16 Alumina (A1 2 O 3 ) 1 30 JO 98 69 } Iron oxide (Fe 2 O.,) 6 05 7 60 Je.64 151 3 JO. 72 Lime (CaO) 2 10 3 90 3 76 2 8 1 9 4 20 Magnesia (MgO) 91 52 83 00 86 50 87 8 93 88 93 93 Carbon dioxide (CO 2 ) 04 14 1. Brick made from Hungarian magnesite. Mineral Industry, vol. 10, p. 439. Styrian Grecian Trans. Am. Inst. Min. Engrs., vol. 26, p. 268. Mineral Industry, vol. 10, p. 439. Trans. Am. Inst. Min. Engrs., vol. 26, p. 268. Physical praperties of magnesia bricks. The brick * whose analysis is given in column 3 of Table 66 was made in Pittsburg from Styrian magnesite. Its specific gravity was 3.44, equivalent to a weight of 160.9 Ibs. per cubic foot. The brick whose analysis appears in column 6 of the same table was made from Grecian magnesite. Its specific gravity was 3.54, corresponding to a weight of 170.2 Ibs. per cubic foot. Le Chatelier tested two kinds of magnesite bricks (Austrain and Grecian) for expansion with increase of temperature, obtaining the results quoted in Table 67. The expansions given are in millimeters, for a bar 100 mm. in length, and are therefore equivalent to percent- ages. * Trans. Am. Inst. Min. Engrs., vol. 26, p. 268. 162 CEMENTS, LIMES, AND PLASTERS. TABLE 67. EXPANSION OP MAGNESITE BRICKS ON HEATING. (LE CHATELIER.) 200 C. 400 C. 600 C. 800 C. Austrian magnesite brick .... mm. 21 mm. 55 mm. 85 mm. 1 10 Grecian magnesite brick '. < * 25 52 79 1 02 References on magnesite bricks. The following papers contain data regarding the manufacture and properties of magnesia bricks. Bischof, C. On magnesia bricks. Proc. Inst. Civil Engineers, vol. 112, pp. 381-383. 1893. Egleston, T. Basic refractory materials. Trans. Amer. Inst. Mining En- gineers, vol. 4, pp. 455-492. 1876. Pennock, J. D. Laboratory note on the heat-conductivity, expansion, and fusibility of firebrick. Trans. Amer. Inst. Mining Engineers, vol. 26, pp. 263-269. Percy, J. Magnesia crucibles and bricks. Metallurgy, vol. 1, pp. 134-137. 1875. Vlasto, S. J, The magnesite industry. Engineering and Mining Journal, March 10, 1900. Weiss, N. Magnesite in Hungary. Iron Age, Jan. 15, 1903, pp, 20-21. Anon. Magnesite [and magnesia brick]. Mineral Industry, vol. 10, pp. 438- 439. 1902. Oxychloride Cements. In 1853 the chemist Sorel discovered that zinc chloride, when mixed with zinc oxide, united with it to form a very hard cement. Later it was discovered that the same held true of a mixture of magnesium chloride and magnesia. The product in both cases is the same an oxychloride of zinc or magnesium respectively. Chlorides and oxides of several other elements possess this same property, but it has been utilized commercially only in the cases of the zinc and magnesium com- pounds. Of these, zinc oxychloride is extensively used as a stopping by dentists. Magnesium oxychloride, called commonly Sorel cement or magnesia cement, has more important technical uses. SoreFs magnesia cement is made by mixing calcined magnesia with a solution of magnesium chloride of 25 or 30 Baume. If the magnesia has been prepared from magnesite, it usually contains a little residual carbon dioxide (C0 2 ), and though setting very rapidly and giving a very strong cement, cracks are apt to develop during setting. When MAGNESIA BRICKS AND OXYCHLORIDE CEMENTS. 163 made from magnesium chloride (see p. 158) the magnesia is free from carbon dioxide, and though it sets and hardens less rapidly, no cracks appear. The commercial magnesium chloride used in the preparation of Sorel stone, etc., usually contains sulphuric acid. As this acid and its compounds spoil the appearance and the durability of the stone produced, it is eliminated from the magnesium chloride by treatment with barium hydrate or barium carbonate. In practice,* the magnesium chloride is dissolved in water to form a solution of 20 to 25 Eaume, and the barium hydrate or carbonate is added by degrees and carefully stirred until the precipitate of barium sulphate ceases to increase. The amount of reagent required is usually between 6 and 10 per cent of the weight of the magnesium chloride treated. Magnesia cement is used very extensively f as a binder, in connection with briquetting, in the manufacture of artificial building-stones, tiles, grindstones, and emery- and polishing-wheels. Its binding quality is very considerable, and it is very plastic and cheap. A good mixture for this use consists of 25 parts of magnesium chloride (45 per cent solution), 25 parts magnesia (93 per cent MgO), and 50 parts water. About 5 Ibs. of this mixture will serve to cement 95 parts of stone, emery, etc. The resulting blocks are very solid and harden thoroughly within a few hours. Gillmore, in 1871, prepared a report on certain American patented products based upon Sorel cements. As this report is still the only complete discussion of the subject it is reprinted below, almost verbatim. "The several steps in the process, beginning with the raw magne- site, are briefly as follows, viz. : "First. The magnesite is burnt in ordinary lime-kilns, at a dark cherry-red heat, for about twenty-four hours. The result is protoxide of magnesium, which is next ground to fine powder between horizonta' millstones, furnishing what .the Union Stone Company style ' Union cement'. "Magnesite has been procured from various localities. That from Greece, California, Maryland, and Pennsylvania contains about 95 per cent of carbonate of magnesia, the residue being mostly insoluble silicious matter. The burnt product is perfectly white. A magnesite is procured in Canada which contains from 60 to 85 per cent of car- * Journ. Soc. Chem. Industry, vol 21, p. 257. 1902. t Schorr, R. The briquetting of minerals. Eng. and Mining Journal, vol. 74, p. 673. 1902. 164 CEMENTS, LIMES, AND PLASTERS. bonate of magnesia. A variable percentage of iron in the residue gives the cement derived from this stone a reddish tint. "Second. For making stone, the burnt and ground magnesite (oxide of magnesium) is mixed dry in the proper proportion with the material to be united; that is, with powdered marble, quartz, emery, silicious sand, soapstone, or with whatever substance forms the basis of the stone to be imitated or reproduced. "The usual proportions a'rei.for enaery- wheels, 10 to 15 per cent of oxide of magnesium by weight; for building-blocks, such as sills, lintels, steps, etc., 6 to 10 per cent, and for common work for thick walls, less than 5 per cent. K "The dry ingredients are mixed together by hand or in a mill. A hollow cylinder revolving slowly about its axis would answer the pur- pose. "Third. After this mixing they are moistened with chloride of magnesium, for which bittern water the usual refuse of seaside salt- worksis a cheap and suitable substitute. The moistened material is then passed through a mill, which subjects it to a kind of trituration, by which each grain of sand or other solid material becomes entirely coated over with a thin film of the cement, formed by a combination of the chloride with the oxide of magnesium. The bittern water is required to be of the density of from 15 to 30 Baume. The mass on emerging from the mill should be about as moist as molder's clay. The mixing- machine used by the Union Stone Company is an improved pug-mill invented by Mr. Josiah S. Elliott. It is represented as an excellent mill, doing its work thoroughly. "Fourth. The mixture is formed into blocks by ramming or tamping it in strong molds of the required form, made of iron, wood, or plaster, precisely as described in paragraph 24, Report on Beton Agglomere. The block may be taken out of the mold at once and nothing further need be done to it. The setting is progressive and simultaneous through- out the mass, as with other hydraulic cements, and requires from one hour to one day, depending somewhat on the chemical properties of the solid ingredients used, the carbonates as a rule requiring a longer time than the silicates. 11 Building-blocks will bear handling, and may be used when three or four days old, although they do not attain their maximum strength and hardness for several months. Emery-wheels are not allowed to be used in less than four weeks. "This stone so closely resembles the natural stone, whether marble, soapstone, sandstone, etc., from which the solid ingredients are ob- MAGNESIA BRICKS AND OXYCHLORIDE CEMENTS. 165 tained by crushing and grinding, that it is difficult, without the appli- cation of chemical tests, to detect any difference in either texture, color, or general lithological appearance. "Strength. In strength and hardness this stone greatly surpasses all other known artificial stones, and is equaled by few, if any, of the natural stones that are adapted to building purposes. The artificial marble takes a high degree of polish, being in this respect fully equal to the best Italian varieties. "Some trials of 2-inch cubes at the Boston Navy-yard gave the following results, reduced to the crushing pressure upon one square inch: No. 1, crushing strength per square inch 7,187^ Ibs. No. 2, " " 11*5621 " No. 3, " " " " 21,562* " No. 4, " 7,343 J " " In none of these samples did the proportion of the oxide of mag- nesium exceed 15 per cent by weight of the inert material cemented together. This statement is derived from the treasurer of the company. " The principal business of the Union Stone Company up to the present time has been the manufacture of emery-wheels. The great tensile strength of the material may be inferred from the fact that in the proof trials the wheels are made to revolve with a velocity of from 2 to 3 miles per minute at the circumference. They do not usually begin to break until a velocity of from 4 to 5 miles per minute is attained. "From a number of specimens of this stone furnished the writer by the treasurer of the company, who also gave their age and com- position as reported below, comprising coarse and fine sandstone of various shades of color, hones, white and variegated marble, emery- wheels, billiard-balls, concrete building-blocks, etc., some small blocks were prepared and subjected to crushing with the results given in Table 68. "Durability. The proofs of the durability of the Union stone rests upon other evidence than that furnished by severe and prolonged climatic exposure. In Boston, however, building-blocks have resisted two winters, and at the present time appear to be, and doubtless are, harder and stronger than before they were touched with frost. "Dr. C. T. Jackson, State Assayer of Massachusetts, reports upon it as follows: "'I find that the frost test (saturated solution of sulphate of soda) has not the power of disintegrating it in the least. The trial was made by daily immersions of the stone in the sulphate-of-soda solution for a week and allowing the solution to penetrate the stone as much as pos- J66 CEMENTS, LIMES, AND PLASTERS. TABLE 68. COMPRESSIVE STRENGTH OF SOREL STONE. Character of the Inert Materials. Proportion by Weight of Oxide of Magnesium. Age of Blocks. Size of Blocks. Total Crushing Strength. Crushing Strength per Square Inch. 1 Coral sand Per Cent. 12 12 to 15 Not known 15 12 to 13 12 Not known 1 year f* 1 " 2 years 3 " 9 months 2 years Not known 2" X2i"Xl$" lf"X2" Xlf" lf"X2" Xl|" ii"xii"xii" H"x2" xu" lf"X2" Xli" U"Xli"Xl" Lbs. 26,500 20,000 54,000 26,000 23,000 16,000 12,000 Lbs. 6,235 7,272 19,636 11,555 6,133 4,923 7,680 2. Pulverized quartz . 3. Washed flour of emery (a piece of hone) 4. Fine marble 5. Mill-sweepings 6. Marble and sand. . . 7. Marble with colored veneer sible and then to crystallize. From this test it is evident that your stone will withstand the action of frost more perfectly than any sand- stone or ordinary building stone now in use. I see no reason why it will not stand as well as granite/ " PL perfect resistance to the freezing and thawing of one winter may safely be accepted as conclusive evidence of the durability in the open air of an artificial stone of which the matrix is any kind of hydraulic cement. At no subsequent period will it be as likely to fail, from freezing and thawing, as during the first winter. A stone suitable for all kinds of building purposes on land might, however, fail under the solvent action of sea-water. On this head it can be said that magnesian com- pounds are understood to resist the immersion in the sea better than the compounds of alumina or lime. "For these reasons this new stone has, with some exceptions, been limited in its application to articles of small bulk and great comparative value, for which other approved and less expensive artificial stone is either not suitable or of less practical value. Although for architectural ornaments of elaborate design it is perhaps less costly, even now, than granite or marble, it cannot hope to compete successfully for general adoption and use by engineers and architects with the beton agglomere and the softer kinds of natural stone until the market price of the oxide of magnesium is greatly reduced. For the peculiar purposes to which it is adapted, it supplies what has heretofore been felt as a great want, and in this field, which is neither narrow nor unvaried, it has no prominent rival. MAGNESIA BRICKS AND OXVCHLORIDE CEMENTS. 167 "The following formula has been found suitable for window-caps, sills, steps, etc. The quantities specified will make 1 cubic foot of stone. 100 pounds of beach sand, cost $1.00 per ton at the works $0.05 10 of comminuted marble, cost $5.00 per ton at the works 0.02J 10 " of Union cement (oxide of magnesium) o 50 10 " of chloride of magnesium in solution, 20 Baum6 f 02 130 ' ' yielding 1 cubic foot of molded stone $0 . 59 % "The labor, depending somewhat on the design as regards the degree and character of its ornamentation, will vary per cubic foot from 20 to 25 cents, making total cost of 1 cubic foot of finished building-block 79^ to 84 J cents. This price may be reduced 10 to 15 cents per cubic foot by incorporating large pebbles and small cobble-stones during the process of molding. "For foundations and other plain, massive walls, the proportion of cement may be very considerably reduced and the quantity of cobble- stones increased." References on oxychloride cements, Sorel stone, etc. Ebel [Magnesia cement concrete for use in mines]. Zeits. angew. Chemie, vol. 15, p. 44. Abstract in Journ. Soc. Chem. Industry, vol. 21, p. 175. 1902. Gillmore, Q. A. Practical treatise on Coignet-Beton and other artificial stone. 1871. Luhmann, E. Magnesia cement. Chem. Zeitung, vol. 25, Report 345. Ab- stract in Journ. Soc. Chem. Industry, vol. 21, p. 118. 1902. Preussner, L. Magnesia cement. Thonindustrie Zeitung, vol. 25, p. 2115. Abstract in Journ. Soc. Chem. Industry, vol. 21, p. 257. 1902. Schorr, R. The briquetting of minerals. Engineering and Mining Journal t vol. 74, p. 673. Nov. 22, 1902. PART IV. HYDRAULIC LIMES, SELENITIC LIME, AND GRAPPIE.R CEMENTS. CHAPTER XIII. THE THEORY OF HYDRAULIC LIMES. BEFORE taking up the manufacture and properties of the various 'Closely allied products hydraulic limes, selenitic limes, and grappier cements which are to be discussed in this part of the volume, it seems desirable to devote some space to a consideration of the general principles -on which the manufacture and use of these products are based. The materials heretofore discussed in this volume the plasters, common lime, magnesia, etc. have been simple in both composition and action. With the hydraulic limes, however, we take up the first member of a great class of very complex products. All these products possess hydraulic properties. In composition, they further agree in that they all consist essentially of silica, alumina, and lime, with or without magnesia and iron oxide. This group of complex cementing materials includes the hydraulic limes, the natural cements, the Port- land cements, and the puzzolan cements. These four classes are quite distinct commercially, but it is at times difficult to draw the dividing lines between the classes in words. Before defining the class of "hydraulic limes' 7 it will therefore be well to explain the principal criterion which will be employed in drawing up that definition. This criterion is the " Cementation Index", a more satisfactory form of the older " hydraulic index". The "Hydraulic Index". In discussing the classification of cement- ing materials, in the introduction to this volume, the statement is made that the power of setting under water, possessed by the hydraulic limes and cements, is due to the formation of compounds of silica, alu- imina, and lime during the manufacture of the cementing materials 168 THE THEORY OF HYDRAULIC LIMES. 169 in question. This being the case, it is a fair assumption * that the degree of hydraulic activity and the strength of any given cementing material will be related, in some way, to the proportions in which it contains these ingredients (silica, alumina, lime, etc.), and to the man- ner in which they are combined. It is obvious that it would be of great value to both manufacturer and engineer if we could devise some method for quantitatively express- ing this relation between the composition and the hydraulic value of any given sample of cementing material. Several methods of doing this have been suggested and used by various authorities. Of these methods of expression, the one that has come into most general use is based upon the calculation of the " hydraulic index ". The " hydraulic index ", as usually defined, is the ratio between the percentage of silica plus alumina to the percentage of lime (CaO). A hydraulic lime, for example, such as that from Metz (Analysis No. 2, Table 74), containing 18.47 per cent silica, 5.73 per cent alumina, and 68.19 per cent lime would therefore have for its hydraulic index 18.47+5.73 24.20 ~~68l9 = 68T9 (Hydraulic Index). The " hydraulic index ", calculated in this manner, is then used as a basis for classifying cementing materials according to their hydraulic activity. The following grouping, which is substantially that given by Spalding,t is an example of this: Hydraulic Index. Product. Less than 0.10 Common lime, quicklime 0.10 to 0.20 Feebly hydraulic limes 0.20 " 0.40 Eminently hydraulic limes 0.40 " 0.60 Portland cement (if burned at high temperature) 0.60 " 1.50 Natural cements 1.50 " 3.00 Weak natural cements 3.00 Puzzolanas, etc. The " hydraulic index " calculated and used in this fashion is certainly better than nothing, but it possesses defects which render it valueless in dealing with certain classes of cements. These defects arise chiefly from the facts that in calculating the " hydraulic index " (1) no allowance is made for the action of either magnesia or iron oxide, and (2) the assumption is made that silica and alumina are quanti- * Strictly speaking, this statement is based on more than a mere assumption; but as a matter of convenience discussion of the reasons for it will be deferred to later chapters. fSpalding, F. P. "Hydraulic Cement", pp. 8, 31, 38. 170 CEMEiNTS, LIMES, AND PLASTERS. tatively interchangeable, i.e., that 10 per cent of silica will have exactly the same effect as 10 per cent of alumina. These defects have led the writer to abandon the use of the "hydraulic index " and to substitute therefor the index described in the next section as the " Cementation Index ". The Cementation Index. As explained and defined below, the Cementation Index is a natural outgrowth from the formula proposed by Newberry for proportioning' Portlatitl-cement mixtures. The index now proposed differs from that formula in assigning values for the magnesia and iron oxide contained in the cement or lime, a change which is necessary in order to adapt it for use wfth the magnesian natural cements and the puzzolan cements. The proposed index is: (2. 8 x percentage silica) +(1.1 x percentage alumina) Cementation Index. ^ + (.7 xpercentage iron oxide) (rercentage lime)+( 1.4 xpercentage magnesia) EXAMPLE. As an example of the details of calculating the Cementa- tion Index, the hydraulic lime of Metz, whose analysis is given as No. 2 of Table 74, will be used. The essential ingredients of this lime, as given in the quoted analysis, are: . Silica (SiO 2 ) ' 18.47 Alumina ( A1 2 O 3 ) 5 . 73 Iron oxide (Fe 2 O 3 ) - 3.29 Lime (CaO) 68 . 19 Magnesia (MgO) 2.66 Substituting these values in the formula 2.8 percentage silica + 1.1 percentage alumina Cementation Index=-^- +.7 percentage iron oxide rercentage lime +1.4 percentage magnesia we have Cementation Index = (2*X18.47) +(1.1x5.73) +(7X3.29) (68.19) +(1.4X2.66) ^51.716+6.303 + 2.303 68.19+3.724 ^60.322 "71.914 = .839. As will be seen later ,_ this is a very typical value for the Cementa- tion Index of a good hydraulic lime. THE THEORY OF HYDRAULIC LIMES. 171 The use of the Cementation Index, as here stated, involves certain assumptions as to the constitution of hydraulic cementing materials. These are, in order of importance: (1) That in hydraulic limes and cements the hydraulic activity is due to the formation during manufacture of certain compounds of lime and magnesia with silica, alumina, and iron. (2) That the silica combines normally with the lime in such molec- ular proportions as to form the tricalcic silicate, 3CaO.SiO 2 . (3) That the alumina combines with the lime as the dicalcic alu- minate, 2CaO.Al 2 O 3 . (4) That magnesia is, molecule for 'molecule, equivalent to lime in its action. (5) That iron oxide is, molecule for molecule, equivalent to alumina. Of these five assumptions, the first is simply a general statement of conditions which are recognized by everybody as probably existing. The second assumption, likewise, is generally accepted, since it agrees with the views of both Le Chatelier and Newberry. The third, based on Newberry 's experiments and confirmed by those of Richardson, is practically accepted by all American cement chemists, though not by those who follow Le Chatelier. The fourth and fifth assumptions, however, are open to question, and the writer realizes that serious objections may be urged against them. But he also realizes that magnesia and iron must be accounted for in some way, that the assumptions above made are inherently prob- able, and that the resulting " Cementation Index " works out very well in practice. For the present, therefore, the " Cementation Index " will be accepted as a guide in discussing the composition and the char- acteristics of the hydraulic limes. Use of the Cementation Index in classification. The Cementation Index will be used in classifying the various hydraulic products, for it gives information of value concerning the properties of the various products. But it cannot be the' sole basis for classification, because the properties of a hydraulic cementing material will be later -seen to depend not only on its composition, but on the conditions of its manu- facture. A material having a Cementation Index of 1.05 might be, for example, a hydraulic lime, a natural cement, or a cement of the Portland type, depending chiefly on the temperature at which the raw material was burned. In general, however, the possible variation is by no means so wide as this. A material with a Cementation Index of 0.40, for example, could under no possible temperature conditions yield anything but a somewhat weak hydraulic lime. 172 CEMENTS, LIMES, AND PLASTERS. In later chapters, when the separate products are under discussion, their respective Cementation Indexes will be determined and stated. At present we are only concerned with determining the limiting values of this index for the hydraulic limes. As will be seen from following paragraphs, these limits are theoretically very wide, but in actual prac- tice very narrow. Definition of hydraulic Unties. T?he hydraulic limes include all those cementing materials (made by burning siliceous or argillaceous limestones) whose clinker after calcination contains so large a per- centage of lime silicate (with or without lime alumiaates and ferrites) as to give hydraulic properties to the product, but which at the same time contains normally so much free lime (CaO) that the mass of clinker will slake on the addition of water. The commercial advantage of manufacturing a material of this kind is that, while the product has hydraulic properties, yet its clinker will slake and pulverize itself on the simple addition of water, thus avoid- ing the expensive mechanical grinding required by the clinker of natural and Portland cements. The definition, therefore, requires that a material to be called a hydraulic lime must satisfy two conditions: (1) its clinker must con- tain enough free lime to slake with water, and (2) the resulting powder must be capable of setting or hardening under water. These two requisite conditions, in their turn, fix the limits of lime that the clinker may contain. The minimum amount of lime present is obviously determined by the consideration that, after burning, enough free lime (in addition to that combined with the silica, alumina, and iron) must be present in the clinker to reduce the entire mass to powder by the force of its own slaking. The maximum amount of lime, on the other hand, is determined by the commercial condition that no more free lime should be present than is absolutely necessary to accomplish this pulverization, for the free lime, whose slaking powders the mass, is by that same slaking made into an inert, or at least non-hydraulic, material. The desired result the formation of a clinker consisting largely of lime silicates, etc., but also containing sufficient free Ihne to slake readily can be attained in two different ways, which yield products very different in quality. These two methods are: (1) By the calcination, at a medium temperature, of a siliceous or argillaceous limestone having a Cementation Index lying between 0-30 and 1.10. Such a limestone will carry so high a percentage of calcium carbonate (relative to its content of silica, alumina, and iron oxide) as to leave, after most of its silica, etc., have been combined with lime, THE THEORY OF HYDRAULIC LIMES. 173 tfufficient free lime to slake the clinker. Hydraulic limes produced in this fashion are the typical hydraulic limes, and the following chapters will have reference to such materials only. It is possible, however, to produce a hydraulic lime by another method, as above noted. This second and much less satisfactory method is (2) By the calcination, at temperatures too low to permit perfect combination of the silica, alumina, and iron oxide with the lime, of a siliceous or argillaceous limestone (less rich in lime than those employed in the first method) having a Cementation Index of 1.10 to 1.60 or over. In other words, a rock is used which would, under proper conditions of burning, give a good natural cement. If it is burned at too low a tem- perature to effect this, however, the result will be a hydraulic lime, for the clinker will consist partly of silicate and aluminate of lime, together with notable amounts of free lime, free silica, and free alumina. Hydraulic limes produced in this way necessarily carry a very large proportion of absolutely inert material. They are, in fact, simple imperfectly burned natural cements and will not be discussed further in this connection. Reverting to the true hydraulic limes, it has been said above that their Cementation Index may range from 0.30 to 1.10; and it will be seen later that commercial hydraulic limes do occur with indexes a& low as 0.331, while others are as high as 1.06. There is, however, considerable reason for dividing the true hydraulic limes into two groups, the first or eminently hydraulic limes containing those products whose index lies between 0.70 and 1.10; while the second group, or feebly hydraulic limes, contains products whose Cementation Index ranges from 0.70 down as low as 0.30. Commercial as well as- theoretical differences serve to separate the two groups, and for that reason they will be discussed in separate chapters. Curiously enough, each of the two classes has an attendant secondary product to be con- sidered. The eminently hydraulic limes during their calcination pro- duce a by-product (grappiers) which is usually marketed separately as a " grappier cement". The feebly hydraulic limes on the other hand are often treated with sulphuric acid m such a way as to develop new properties, and are then marketed as selenitic limes. In further dis- cussion of the hydraulic limes, therefore, they will be treated as two groups in two separate chapters, covering respectively Chapter XIV. Eminently Hydraulic Limes : Grappier Cements. Chapter XV. Feebly Hydraulic Limes: Selenitic Limes. CHAPTER XIV. EMINENTLY HYDRAULIC LIMES: GRAPPIER CEMENTS. THE hydraulic limes are usually, compared to. Portland or good natural cements, only feebly hydraulic. This fact, taken in connection with the abundance of materials suitable for the manufacture of natu- ral cements, has prevented the introduction of hydraulic-lime manu- facture into the United States, though in Europe the industry is of considerable importance. No hydraulic lime is at present made in this country, nor is there any prospect that the industry will ever be taken up here. A considerable amount of hydraulic lime and grappier cement is, however, annually imported. This is brought about by the fact that these products, being low in iron and soluble salts, are light colored : and do not stain masonry. There is thus a fair market for them for architectural rather than for engineering uses. A promi- nent brand of grappier cement much used in the United States as a " non-staining cement " is called Lafarge. The manufacture and properties of the hydraulic limes and grappier cements will be discussed briefly. This discussion will be practically confined to the practice followed at Teil, France, where the largest and best-known plants are located. Composition of the ideal hydraulic lime. The clinker of an ideal hydraulic lime should, as may be deduced from the considerations set forth in the preceding chapter, satisfy two limiting conditions. On the one hand, it must contain sufficient free lime to disintegr te the entire mass of clinker by the force of its own slaking. On the other hand, no more free lime should be present than is absolutely necessary to effect this disintegration; and no uncombined silica or alumina should be present in the clinker. This ideal condition would be arrived at, according to Le Chatelier,* if we could obtain a clinker containing four equivalents of lime for one of silica. Three of the four equiva- lents of lime would be united with all the silica to form tricalcic sili- cate, while the fourth equivalent of lime would remain free, and would * Trans. Am. Inst. Min. Eng., vol. 22, p. 16. 174 EMINENTLY HYDRAULIC LIMES: GRAPPIER CEMENTS. 175 be sufficient to accomplish the disintegration of the entire mass, through the force produced during its own slaking. Accepting this statement, we can calculate the percentages of the various constituents which should be present in an ideal hydraulic lime, both before and after slaking, and also the composition of the limestone necessary to give, in burning, this ideal product. The results of such a calculation are shown in the following table. TABLE 69. COMPOSITION OF IDEAL HYDRAULIC LIMESTONE AND HYDRAULIC LIME. Hydraulic Lime. Hydraulic Before Burning. Before After Slaking. Slaking. Si0 2 . . 13.20 21.20 19.08 CaO C0 2 H 2 6 1 86.8 | 0.00 78.80 0.00 0.00 70.92 0.00 10.00 100.00 100.00 100.00 Index. . . . 0.753 0.753 0.753 In actual practice, however, it is found that these theoretical com- positions cannot be worked up to advantageously. If, for example, a limestone of the composition given -above (Si(>2 13.2 per cent, CaCOs 86.8 per cent) is burned under the ordinary conditions of hydraulic-lime manufacture, it is found that all the silica does not combine with three fourths of the lime, as is required by the theory. What actually hap. pens is that part of the silica will combine w4th part of the lime to form tricalcic silicate, thus leaving a certain amount of uncombined silica and entirely too much uncombined lime. Any increase in the uncombined lime beyond the amount necessary to cause the clinker to disintegrate by its slaking lessens the hydraulic value of the product. It is therefore necessary, in practice, to modify the ideal compositions, these modifications being in the following directions: (a) Lower lime content. The limestones in actual use, as shown by the analyses quoted in Tables 70 and 71, differ from the ideal hydraulic limestone in carrying from 70 to 80 per cent of lime carbonate in place of the 86.8 per cent of theory. This lowering in the original lime carbonate content of the limestones decreases the amount of uncom- bined lime in the product. (6) Presence of alumina and iron. Even the best hydraulic lime- stones in actual use carry notable amounts of alumina and iron oxide. 176 CEMENTS, LIMES, AND PLASTERS. These constituents act as fluxes,, facilitating the combination of the silica and lime. They also combine themselves with lime to form aluminates and ferrites of lime. These latter salts do not increase the hydraulic value of the product, for they become hydrated and inert during the process of slaking, but their formation disposes of some of the excess of free lime. The effect of these modifications is shown clearly when the Cementa- tion Indexes of the ideal and the various' commercial products are com- puted and compared. Le Chatelier's ideal lime has a Cementation Index of 0.75, while the actual limes whose analyses are given later will average about 0.85. Analyses of a number of commercial hydraulic limes are given in Table 74, page 179. Raw materials: hydraulic limestones. The limestones actually used in the manufacture of hydraulic limes will carry from 70 to 80 per cent of lime carbonate. In hydraulic limestones of the best types, such as are used at Teil, France, the silica will vary between 13 and 17 per cent, while the alumina and iron together rarely exceed 3 per cent. Several analyses of hydraulic limestones are given in Tables 70 and 71. TABLE 70. ANALYSES OF HYDRAULIC LIMESTONES, TEIL, FRANCE. 1. 2 3. 4. Silica (SiO 2 ) ... 12.40 13.75 16 89 14 30 Alumina (A1 9 O.) 60 65 81 70 Iron oxide (Fe 2 O 3 ) 50 trace trace 80 Lime (CaO) 47 49 47 00 45 40 46 50 Magnesia (MgO) n d n. d n d n d Carbon dioxide (CO 2 ) . 37.31 36.93 35.67 36*54 1. Alignole quarry; average of six analyses by Rivot. 2. Gaillant quarry; average of three analyses by Rivot. 3. Tinliere quarry; analysis by Rivot. 4. Lafarge quarry; average of nine analyses by Rivot. TABLE 71. ANALYSES OP HYDRAULIC-LIME ROCKS, FRANCE AND GERMANY. 1. 2. 3. 4. Silica (SiO 2 ) 17 00 11 60 11 03 11 20 Alumina ( A1 2 O 3 ) 1 00 '3 60 3 75 5 30 Iron oxide (Fe 2 O 3 ) 3.0 5 07 4 60 Lime (CaO) 44 80 42 84 43 02 35 50 Magnesia (MgO) 71 1 43 1 34 5 85 Carbon dioxide (CO 2 ) 35 99 35 23 35 27 34 35 1. Senonches, France. Descotils, analyst. 2. Metz, Germany. Berthier, analyst. 3. Hausbergen, Germany. Muspratt, analyst. 4. Plassac, France. Vicat, analyst. Quoted by Zwick, "Hydraulischer Kalk und Portland-Cement", pp. 66, 67. EMINENTLY HYDRAULIC LIMES: GRAPPIER CEMENTS. 177 TABLE 72. ANALYSES OF THE VARIOUS BEDS IN THE HYDRAULIC LIMESTONE QUARRIES AT MALAIN, FRANCE. 1. 2. 3. 4. 5. 6. 7. 8. 9. 14.70 0.60 45.05 0.30 39.35 Silica (SiO 2 ) 7.60 0.75 50.05 0.30 41.30 10.15 0.90 48.05 0.30 40.60 10.30 0.65 48.30 0.30 40.45 0.739 12.30 1.00 46.90 0.25 39.55 15.70 1.10 44.75 0.20 38.25 13.80 0.80 46.30 0.25 38.85 14.20 0.85 45.75 0.40 38.80 14.25 0.75 45.35 0.40 39.25 Alumina (A1 2 O 3 ) \ Iron oxide (Fe O 3 ) J Lime (CaO) Magnesia (MgO) Carbon dioxide (CO 2 ). . . 1 Water J Cementation Index. 10. 11. 12. 13. 14. 15. 16. 17. 18. 6.40 0.45 51.05 0.45 41.65 Silica (SiO 2 ) 14.75 1.05 45.15 0.15 38.90 16.35 1.00 43.85 0.55 38.25 1.05 16.10 0.80 44.20 0.40 38.50 16.80 0.30 44.60 0.40 37.90 14.90 0.80 45.10 0.45 38.75 14.35 0.80 45.55 0.40 38.90 12.45 0.70 46.80 0.45 39.60 14.85 0.75 45.15 0.45 38.80 Alumina (A1 2 O 3 ) \ Iron oxide (Fe 2 O 3 ) J Lime (CaO) Magnesia (MgO) Carbon dioxide (CO 2 ). . . \ Water J Cementation Index Burning. Hydraulic lime is burned in continuous kilns, like common lime. No difference, in fact, exists between the burning of common and of hydraulic limes, so far as the practical operations involved are concerned. The temperature attained in burning is, however, higher in hydraulic lime-kilns than in those burning common lime, and the fuel requirements are correspondingly increased. Beekwith states, for example, that at Teil 100 tons of coal are required to burn stone equivalent to 500 tons of screened lime. This corresponds to a fuel consumption of 20 per cent by weight on the lime production. The temperature and thoroughness of the burning are directly related to the Cementation Index of the lime. The higher the index the less care will be necessary to avoid the presence of too much free lime. A hydraulic lime of index 0.75, for example, would be much more difficult to burn properly than one whose index ran as high as 0.85 or so. In fact, as the index approaches 1.00, the difficulty is, not to avoid free lime, but to keep enough free lime in the product to enable it to slake properly. In Tables 73 and 74 are given the analyses of a number of hydraulic limes, after being burned but before slaking. 178 CEMENTS, LIMES, AND PLASTERS. FIG. 27. Kiln used for burning hydraulic lime, Malain, France. (After Bonnami.) EMINENTLY HYDRAULIC LIMES: GRAPPIER CEMENTS. 179 TABLE 73. ANALYSES OF HYDRAULIC LIME BEFORE SLAKING (TEIL, FRANCE). 1. 2. 3. 4. 5. 6. 7. 8. 9. Silica (SiO 2 ) 20.33 1.00 0.82 77.87 n. d. .753 22.39 1.06 tr. 76.55 n. d. .834 23.60 1.28 tr. 75.12 n. d. .898 22.95 1.12 1.28 74.64 n. d. .899 20.57 1.13 tr. 77.76 0.54 .749 21.7 1.8 0.6 74.0 0.7 .842 26 .07 4.38 tr. 68.94 0.61 1.115 26.4 }3.0 65.16 1.04 1.155 22.59 J2.63 10.84 65.62 1.54 .985 Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) Lime (CaO) Magnesia (MgO) Cementation Index 1. Average of lime burned from Ahgnole quarry rock. 1 Anal b Rivot Quoted b Zwick 3 .. .... .. .. Tinli>? P " "Hydraulischer Kalk und Portland- 1' . .... .. .. ESSS .. .. j Cement", pp. 67, 69. 5. Analysis by Vicat. Quoted by Beckwith, "Hydraulic Lime of Teil", p. 13. 6. Quoted by Stanger and Blount, Mineral Industry, vol. 5, p. 70. 7. Analysis by Vicat. Quoted by Beckwith, "Hydraulic Lime of Teil", p. 13. 8. Analysis by Landrin. Quoted by Thorpe, "Dictionary Applied Chemistry", vol. 1, p. 483. 9. Analysis by Michaelis. Quoted by Schoch, "Mortel-Materialen", p. 73. Two of the above analyses of Teil lime, Nos. 7 and 8, give exceptionally high values for the Cementation Index. The average index of all nine samples is 0.913; if Nos. 7 and 8 be excluded the average is 0.85. TABLE 74. ANALYSES OF HYDRAULIC LIMES, FRANCE, GERMANY, AND ENGLAND. 1. 2. 3. 4. 5. e. Silica (SiO 2 ) 26 77 18 47 17 18 17 75 23 61 24 33 Alumina (A1 2 O 3 ) 1.57 5 73 5 84 8 88 3 89 3 73 Iron oxide (Fe 2 O 3 ) 3 29 6 32 6 18 n d Lime (CaO) 70.54 68.19 68 56 56 01 71 99 71 94 Magnesia (MgO) 1 12 2 66 2 09 9 28 51 n d Cementation Index 1.06 .839 .839 .925 .968 1.00 1. Senonches, France. Descotils, analyst. 2. Metz, Germany. Berthier, analyst. 3. Hausbergen, Germany. Muspratt, analyst. 4. Plassac, France. Vicat, analyst. Quoted by Zwick, "Hydraulischer Kalk und Portland-Cement", pp. 66-67. 5. d'Emondeyille, France. Vicat, analyst. 6. Lyme Regis, England. Quoted by Cummings, "American Cements", p. 35. Slaking. Hydraulic lime, after burning, is a mixture of two distinct compounds. Part of the mass is composed of lime silicate, which would not slake if water were poured on it, but would form a hydraulic cement if finely ground. The remainder of the hydraulic lime consists simply of quicklime, which will slake with water. The result of the mixture of the two ingredients is that if water be poured on a lump of hydraulic lime the portion consisting of quick- lime will rapidly take up the water and slake. In its slaking its expan- sion will break up the entire mass into a fine powder. If this operation be done carefully, with just the proper amount of water, the result will be a fine, dry, white powder, consisting mostly of lime silicate with about one third to one fourth as much of slaked lime. 180 CEMENTS, LIMES, AND PLASTERS. In the earlier days of hydraulic-lime manufacture in France (and even at the present day in England) it was the practice to put the hydraulic lime on the market in lumps, just as it is drawn from the kiln, leaving the work of slaking it to the purchaser. At present, how- ever, the slaking in the French works is done at the lime-plant. The advantages of this method of procedure are that (1) the slaking is done more uniformly and carefully, so that tile value and reputation of the lime is improved, and (2) the lime gains considerably in weight and bulk during slaking, so that the cost of slaking is made up. Slaking should be done with as little water as is '-compatible with thorough slaking. The lime as drawn from the kiln is therefore spread out in thin layers and lightly sprinkled with water. It is then shoveled up into heaps or into bins, where it is allowed to remain for ten days or so. The slaking is completed, while the lime is thus heaped up, by the aid of the steam which is generated. After slaking is completed, the lime remains as a fine powder inter- spersed with lumps (grappiers) of harder material. These lumps con- sist in part of lime silicate and in part of unburned or underburned limestone. It would be desirable if practicable to remove the latter material, as it is, of course, valueless as a cement. The lumps of lime silicate, on the contrary, will, if finely ground, make a good natural cement. This separation is, however, commercially impracticable, and therefore all the grappiers are treated together. The lime after slaking is passed over screens (of about 50-mesh). These screens permit all the slaked lime to pass, but reject the grap- piers. The lime is sent to the packers, while the grappiers are ground finely under millstones. So far as can be learned, a certain percentage of ground grappiers is always added to the lime, in order to increase its hydraulicity. As later briefly noted (p. 185), the grappiers alone are also sold as a cement. The analyses by Durand-Claye, given in Table 75, are quoted in Spalding's "Hydraulic Cements", p. 20, and serve to illustrate the com- position of the various products. In this series analysis No. 1 is of the lime which has completely powdered during slaking and passed through the first sieve, while analysis No. 3 is of the grappiers rejected by this sieve. It will be seen that while the slaked lime has a Cementation Index of 0.992, the grappiers are proportionately less rich in lime (CaO), having an index of 1.63. In order to increase the hydraulic properties of the lime which has passed the sieve, a certain proportion of ground grappiers is aJdded to it. This causes the lime as marketed to have a Cementation Index EMINENTLY HYDRAULIC LIMES : GRAPPIER CEMENTS. 181 o g g a & s? ? I 182 CEMENTS, LIMES, AND PLASTERS. TABLE 75. ANALYSES OF KILN PRODUCTS, TEIL, FRANCE. 1. 2. 3. 4. Silica (SiO 2 ) 23 05 23 95 31 85 43 90 Alumina (A1 2 O 3 ) . .... Iron oxide (i^e 2 O 3 ) . . ......... j 2.75 3.10 4.25 8.20 Lime (CaO) 65. ,75 63 35 55 60 45 25 Magnesia (MgO) 1.50 1 15 1 20 85 Water etc 6 95 8 50 7 10 2 60 Cementation Index 992 1 08 1 63 2 82 of 1.08, as shown by analysis No. 2, which is of the Teil lime in its com- mercial form. During the burning a small percentage of a third compound close in composition to CaO.SiO 2 is formed. This product is not used in either the hydraulic-lime or grappier-cement industries, but is mixed with slaked lime and used in the manufacture of pipe, tile, etc. It is, in fact, an artificial puzzolana, as is seen from its analysis (No. 4), which gives a Cementation Index of 2.82. These analyses by Durand-Claye have been used because they form a complete series. They are not entirely representative, however, of Teil hydraulic lime, as is seen on comparing them with analyses No. 2 and 3 in Table 76, below. These latter analyses give Cementation Indexes of 0.841 and 0.854 respectively, which are considerably lower than of the corresponding analyses of Table 75. TABLE 76. ANALYSES OF HYDRAULIC LIMES, AFTER SLAKING. 1. 2. 3. Silica (SiO 2 ) 22 19 05 18 2 Alumina (A1 2 O 3 ) . 2 1 6 1 2 Iron oxide (Fe 2 O 3 ) 2 55 8 Lime (CaO) 62.0 65 10 60 Magnesia (MgO) 1.5 65 1 32 Sulphur trioxide (SO,) 5 3 n d Carbon dioxide (CO 2 ) 001 r s 00 Water 10 J 12.45 | n d Cementation Index 1 016 841 o &r>i 1 1. Typical hydraulic lime, after slaking. Le Chatelier, Trans. Am. Inst. Min. Engrs., vol. 22, p. 16. 2. Hydraulic lime of Teil, after slaking. Thorpe, Diet. Applied Chem., vol. 1, p. 474. 3. Hydraulic lime of Teil. after slaking. Gillmore, "Limes, Cements, and Mortars", p. 125, Weight and specific gravity. Beckwith states that Teil lime in lumps, before slaking, weighs 36 Ibs. per cubic foot; while slaked EMINENTLY HYDRAULIC LIMES: GRAPPIER CEMENTS. 183 and screened its weight averages about 43 Ibs. per cubic foot. Accord- ing to Schoch,* the hydraulic limes average in specific gravity about 2.9 Tensile and compressive strength. The results given in Table 77 are quoted by Schoch * as being fair averages for hydraulic-lime mortars composed of one part lime and three parts sand; kept for seventy-two hours after molding in a moist atmosphere and the remain- der of the time under water. TABLE 77. AVERAGE STRENGTH OF HYDRAULIC LIMES. (SCHOCH.) Pounds per Square Inch. 7 Days. 28 Days. 1 Year. 64 Ibs. 356 " 100 Ibs. 683 " 299 Ibs. 1920 " Compression These results may be compared with those given in Tables 78 and 79,. which are quoted by Beckwith as the averages of several series of experi- ments carried on at Toulon and Marseilles on hydraulic-lime mortars- composed of about one part lime to two parts of sand. These mortars were made into blocks and kept under salt water the entire time. TABLE 78. TENSILE STRENGTH OF TEIL HYDRAULIC-LIME MORTAR. Time Immersed. Tensil< ; Strength in Pounds oer Square Inch. 1. 2. 3. 4. 5. 6. Average. 45 days 90 " 180 " 1 year . . 31.71 85.06 97.11 123 43 40.38 88.49 106.22 111 63 30.79 83.78 89.16 126.42 30.83 77.68 86.88 122.15 '57i59 86.03 121.30 38.42 83.77 94.86 120 . 94 34.43 79.39 93.38 120 95 2 years 141 06 164 20 152 63 Ratio of compressive to tensile strength. When in use, limes and cements are usually subjected to direct compressive stress only, tensile strains being rarely applied in well-designed and well-built structures. In testing, however, a test for tensile strength is much cheaper and more readily applied than one for compressive strength. The result is, that though limes and cements are almost entirely used in com- * Schoch, C. Die moderne Aufbereitung und Wertung der Mortel-Materialen, p. 74. 184 CEMENTS, LIMES, AND PLASTERS. TABLE 79. COMPRESSIVE STRENGTH OF TEIL HYDRAULIC-LIME MORTARS. Time Immersed. Compressive Strength in Pounds per Square Inch. 1. 2. 3. 4. Average. 45 days 219.59 359.62 593.98 612.91 613.88 191.75 362.41 467.13 591.84 577.33 194.09 355.20 451.34 561.87 573:92 205.04 259.15 504.24 588.99 202.62 334.10 504.17 588.90 588.38 90 " 180 " pression, they are usually tested in tension. For this reason it is desir- able to ascertain, as definitely as possible, the ratio which exists between the compressive and the tensile strength of any type of lime or cement. If this ratio be once determined, a tensile test can thereafter be used to determine the compressive strength of the material. In the present case, the tensile and compressive tests given in Tables 77, 78, and 79 have been compared. The results are sufficiently close to indicate that the compressive strength of a hydraulic-lime mortar mixed in the usual working proportions (1 lime to 2 or 3 sand) will be from five to six times the tensile strength of the same mixture. (The actual average value, given by eight tests, for this ratio was 5.38 to 1.) Proportions for mortars and concretes. The following proportions for making mortars and concretes with hydraulic lime are recommended by Beckwith: (a) Mortar for use in salt water: 10J U. S. bushels (590 Ibs.) of Teil lime to 1 cubic yard of sand, equivalent to one scant measure of lime to two full measures of sand. (6) Mortar for use in fresh water: 9 U. S. bushels (506 Ibs.) of Teil lime to 1 cubic yard of sand, equivalent to 1J measures of lime to 3 measures of sand. (c) Mortar for use in air: 7 U. S. bushels (421 Ibs.) of Teil lime to 1 cubic yard of sand, equivalent to 1 measure of lime to 3 measures of sand. 4d) For concretes the usual proportions are: (1) For use in salt water, 2 measures mortar to 3 measures of broken stone. (2) For use in fresh water, 1 measure mortar to 2 measures of broken stone. EMINENTLY HYDRAULIC LIMES: GRAPPIER CEMENTS. 185 Grappier Cements. Grappier cements are made by grinding finely the lumps of unburned and overburned material which remain when a hydraulic lime is slaked. These lumps, as earlier noted, consist partly of lime silicate and partly of unburned limestone. The value of the resulting grappier cement will depend on the proportions in which these two ingredients occur in the lumps. If lime silicate forms most of the lumps, the grappier cement will be a very satisfactory material, approximating to Portland cement in its properties. If most, or even a large part, of the lumps consist of unburned limestone, however, the grappier cement will be practically worthless. Lafarge cement, well known on the American market as a "non- staining" cement, is a grappier cement of very satisfactory composition made at Teil, France. Composition of grappier cements. TABLE 80. ANALYSES OF GRAPPIER CEMENTS. 1. 2. 3. 4. 5. Silica (SiO 2 ) 26.5 31.85 31.10 27.38 24 65 2.5 f 4.43 2.61 6 55 1.5 >4.25 \2.15 1 02 2 60 Lime (CaO) 63.0 55.60 58.38 58.38 56 30 1.0 1.20 1.09 0.46 90 Alkalies (K O Na 2 O) n d n d 94 n d n d 0.5 n. d. 0.60 0.43 0.35 1 e n f 1.28 n. d. 1 Water r 5.0 7.10 \n. d. n. d. J 8.65 Cementation Index 1 212 1 63 1 560 1 359 1 356 1 Typical grappier cement. Le Chatelier, Trans. Amer. Inst. Min, Engrs., vol. 22, p. 19. 2. Teil grappiers. Analysis by Durand-Claye. Quoted by Spalding, " Hydraulic Cement ", p. 20. 3. Lafarge cement. C. F. McKenna, analyst, 1897. Sales-agents' circular. 4. Lafarge cement. Quoted by E. Duryee, Engineering News, vol. 47, p. 23. Jan. 9, 1902. 6. Malain grappier cement. Quoted by Bonnami, " Fabrication et controle des Chaux Hydrai*. liques ", p. 54. Physical properties of grappier cements. The only data available on the strength, etc., of grappier cements are those contained in the circular issued by the American sales-agents of the Lafarge brand. The tests were conducted in 1897 by Dr. C. F. McKenna. The Lafarge cement gave the following results: Specific gravity, not ignited. " " ignited Loss on ignition 2 . 6 Initial set 4 hours 2 . 7 Final set 10 hours 3 . 83% Fineness 99 . 8% through 50-mesh " 99. 4% through 100- " 186 CEMENTS, LIMES, AND PLASTERS. TABLE 81. TESTS OP TENSILE STRENGTH, LAFARGE CEMENT. (McKENNA.) Composition of Mortar. Tensile Strength in Pounds per Square Inch. 1 Week. 1 IVJonth. 3 Months. 7 Months. 8 Months. 1 Year. 2 Years. Neat cement, 22 \% water. Neat cement, 24% water 330 320 145 465 242 500 298 542 470 645 665 665 1 part cement, 2 parts sand. . . . These results have been plotted diagrammatically, as shown in Fig. 29. TOO 500 400 300 200 100 5 I a a 10 i~ co FIG. 29. Tensile strength of Lafarge (grappier) cement. CHAPTER XV. FEEBLY HYDRAULIC LIMES: SELENITIC LIMES. THE feebly hydraulic limes have been defined in Chapter XIII as in- cluding those products whose Cementation Index ranges between 0.30 and 0.70. This means that in such a product, no matter how high the burning temperature, not over 70 per cent of its total lime (CaO) can be in combination with the silica, etc., while if the Cementation Index, as shown by analysis, falls as low as 0.30, only 30 per cent of the total lime can be so combined, even under the most favorable circumstances. As combination can never be theoretically complete, it is safe to say that in the feebly hydraulic limes only from 20 to 60 per cent of their total lime is combined, the remainder being left free and capable of slaking. A product containing so much free lime and so little in the combined form can obviously possess little hydraulicity or strength. Limes of this class would hardly merit description were it not for the fa"ct that they are the usual type of English hydraulic limes, and that they often serve as a basis for making a product selenitic lime which requires brief attention. TABLE 82. ANALYSES OF HYDRAULIC-LIME ROCKS. 1. 2. 3. Silica (SiO ) 5.00 4.64 7.40 Alumina (Al 6 3 ) f 7.08 2.70 | 4.23 10.85 5.30 Lime (CaO) 48.65 48.27 40.82 Magnesia (MsrO) 1.86 4 52 40.26 37.92 37.06 0.356 0.443 0.581 1. Holywell, England. Muspratt, analyst. 2. Falhagen, Germany. Pasch, analyst. 3- Horb, Wurtemberg. Knauss, analyst. 187 188 CEMENTS, LIMES, AND PLASTERS. TABLE 83. ANALYSES OF FEEBLY HYDRAULIC LIMES. 1. 2. 3. 4. 5. Silica (SiO 2 ) 7.60 11.95 11.00 16.05 8.36 11.60 4.25 3.67 1.92 I 7 nc 0.96 8.52 3.00 3.22 > 7.08 Lime (CaO) 79.09 65.73 78.40 77.29 81.44 7.25 3.93 1.52 3.11 439 0.581 0.440 0.621 0.331 1. Falhagen, Germany. Pasch, analyst. 2. Horb, Wlirtemberg. Knauss, analyst. 3. Fecamp, France. Rivot, analyst. 4. Aberthaw, England. Quoted by Cummings. "American Cements", p. 35. 5. Holy well, England. Muspratt, analyst. Tensile strength. In Table 84 are given the results of tests, on the tensile strength of various English hydraulic-lime mortars, carried out by Grant * about 1880. These tests were made on briquettes hav- ing a cross-section of 2J square inches ; but the results given in Table 84 have been reduced so as to give the strength in pounds per square inch. TABLE 84. TENSILE STRENGTH OF HYDRAULIC-LIME MORTARS. (GRANT.) 1 Lime: 3 Sand. ILime: 4 Sand. 1 Lime : 5 Sand. 1 Lime: 6 Sand. Dry. Wat. Dry. Wet. Dry. Wet. Dry. Wet. Pounds. 50 Pounds. 68 Pounds. 44 Pounds. 57 Pounds. 30 Pounds. 45 Pounds. 21 Pounds* 28 48 95 49 59 32 47 23 27 Lime C 40 81 26 61 21 44 18 34 46 81 40 59 28 45 21 30 Each of the values given in this table represents the average of the results on five specimens tested. All the tests were made one year after the briquettes were molded. The words "wet" and "dry" refer to the fact that half of the briquettes were kept in air and the other half in water during the entire year. .* Proc. Institution Civil Engineers, vol. 62, p. 165. 1880. FEEBLY HYDRAULIC LIMES: SELENITIC LIMES. 189- The results above tabulated are shown diagrammatically in Fig. 30^ It will be noted that the- "wet" briquettes gave results exceeding the "dry" in an average ratio of almost 1.6 to 1. 100 Ibs. 70 80 Ibs. i i FIG. 30. Tensile strength of feebly hydraulic limes. Compressive strength. Tests on the compressive strength of mortars- made * from three English hydraulic limes are given in Table 85. These- tests were made on 6-inch cubes kept in air for one year before testing.. * Proc. Institution Civil Engineers, vol. 62, p. 165. 1880. 190 CEMENTS, LIMES, AND PLASTERS. The results have been reduced to give the values for compressive strength in pounds per square inch. TABLE 85. COMPRESSIVE STRENGTH OF HYDRAULIC-LIME MORTARS. (GRANT.) Composition of Mortar. Kind of Lime. Lime A. Lime B. Lime C. Average. Lime 1, sand and " 1, " " lf .< gravel 6 Pounds per Square Inch. 159 72 81 Pounds per Square Inch. 178 172 179 Pounds per Square Inch. 359 167 133 Pounds per Square Inch. 232 137 131 " 8 " 10 The values given a,re the average of ten specimens tested. Selenitic Lime: Scott's Cement. The cementing material known as Scott's cement, selenitic cement, or selenitic lime consists essentially of lime (CaO) plus a small per- centage of sulphur trioxide (SOs). The lime used as a basis for this cement is always a more or less hydraulic variety, while the sulphur trioxide may be added to it in the form of either plaster of Paris or sulphuric acid. The resulting selenitic lime or Scott's cement shows a markedly higher strength, both in compression and tension, than the lime from which it was made. Manufacture of selenitic limes. In his earlier patents Scott pro- Tided for the manufacture of this product by exposing lime to the fumes of burning sulphur. This was accomplished * "by reheating calcined lump lime in an oven having a perforated floor, beneath which were placed pots of burning sulphur. The sulphurous-acid fumes from the sulphur rose among the red-hot lumps of lime, leading to the formation of calcium sulphite (CaSOs), and this in turn became oxidized into calcium sulphate (CaSO^. The amount of sulphurous acid thus absorbed by the whole bulk of the lime was small, rarely exceeding from 2 to 3 per cent, and of course only the exterior surfaces of the lumps became coated with the sulphur compound; but when the cement was ground, to prepare it for use, the sulphate of lime became evenly dis- tributed throughout the mass. In course of time General Scott found that he could obtain the same results, either by adding sulphuric acid to the water used in preparing * Redgrave, G. R. Calcareous cements, p. 176. 1895. FEEBLY HYDRAULIC LIMES: SELENITIC LIMES. 191 the mortar or by the addition of powdered gypsum or plaster of Paris to the ground lime. It mattered little in what form the sulphuric acid was conveyed to the lime, and many soluble sulphates were found to answer quite as well as the sulphate of lime. Ultimately Scott specified the manufacture of a cement, which he named ' selenitic cement', by the addition of 5 per cent of ground plaster of Paris to calcined hydraulic lime, which was then ground to an impalpable powder and placed in sacks or casks for use". The hydraulic lime used in the manufacture of selenitic lime is appar- ently always one of the feebly hydraulic varieties such as are discussed earlier in the present chapter. Tensile strength of selenitic limes. The following table shows the results of tests * by Grant about 1880 on various selenitic limes. For purposes of comparison tests are also given on two of the limes before the addition of sulphate. The tests were made on briquettes having a sectional area of 2J square inches ; but in the table below the results given are reduced to pounds per square inch. TABLE 86. TENSILE STRENGTH OF SELENITIC LIMES. (GRANT.) 1 Lime: 3 Sand. 1 Lime: 4 Sand. 1 Lime: 5 Sand. 1 Lime: 6 Sand. Dry. Wet. Dry. Wet. Dry. Wet. Dry. Wet. A. Gray lime, not selenitic. . . . 50 128 48 79 123 91 128 68 141 95 131 148 151 204 44 65 49 63 80 59 83 57 139 59 99 129 102 147 30 55 32 44 72 33 71 45 87 47 72 83 77 123 21 40 23 52 58 29 28 65 27 80 74 66 76 A'. " " selenitic B. Lias lime, not selenitic B' " " selenitic C. Selenitic lime D. " " Rugby E. ' ' ' Aberthaw Each of the above results represents the average of the tests of five specimens. The tests were made one year after the briquettes were molded. The words "wet" and "dry" refer to the fact that some of the briquettes were kept in air and others in water during the entire year. These results as to tensile strength are shown diagrammaticaily in Fig. 31. * Proc. Institution Civil Engineers, vol. 62, p. 165. 1880. 192 CEMENTS, LIMES, AND PLASTERS. Compressive strength of selenitic limes. A number of selenitic limes were tested for compressive strength by Grant, the results being given in Table 87. 420 {100 50 1 FIG. 31. Tensile strength of plain hydraulic and selenitic limes. TABLE 87. COMPRESSIVE STRENGTH OF SELENITIC LIMES. (GRANT.) 1 Lime: 6 Sand. 1 Lime: 8 Sand. 1 Lime: 10 Sand. A Gray lime not selenitic 159 72 81 A'. " " selenitic 289 119 127 B. Lias lime, not selenitic 178 172 179 B'. " " selenitic 268 305 159 C. Selenitic lime 414 239 210 D. " . " Rugby. . 577 533 329 E " " Aberthaw 530 339 239 FEEBLY HYDRAULIC LIMES: SELENITIC LIMES. 193 The samples discussed in the above table were made up into 6-inch cubes and kept in air one year before testing. The results in the table have been reduced to pounds per square inch. 300 Ibs. 200 Ibs. 100 Ibs. 1 I FIG. 32. Compressive strength of plain hydraulic and selenitic limes. The gain in strength due to this process of selenitizing is obvious, but it must be recollected that it gives satisfactory results only when employed on feebly hydraulic limes. With common non-hydraulic limes, and with the better grades of hydraulic limes, the results are not commensurate with the extra expense. For American use, therefore, Scott's process has little to commend it, for our good natural cements would leave little field for such a product as selenitic lime PART V. NATURAL CEMENTS. I* CHAPTER XVI. DEFINITION AND RELATIONS OF NATURAL CEMENTS. BEFORE taking up a detailed description of the materials, man- ufacture, and properties of natural cements it will be useful to make some brief general statements concerning the group. In the present chapter, therefore, an attempt will be made to discuss the natural cements as a class, laying emphasis upon the points of resem- blance of the various brands and disregarding for a time their many points of difference. The difficulties which are encountered in such an attempt are greater than the reader, at first sight, may imagine; for few engineers realize what a heterogeneous collection of products is included under the well- known name of " natural cement ". The cause of this lack of knowl- edge is not far to seek. Natural cements are too low in value to be shipped, under ordinary circumstances, far from their point of pro- duction. The natural cement made at any given locality has usually, therefore, a well-defined market area within which it is well known and subject to little competition. The engineer practicing within such an area naturally forms his idea of natural cements in general from what he knows of the brands encountered in his work, and as all the brands from one cement-producing locality are apt to 'resemble one another quite closely, he is likely to conclude that natural cements are quite a homogeneous class, with many points of resemblance and few of difference. The truth is, on the contrary, that there may be far greater differences in strength, rate of set, chemical composition, etc., between the natural cements made in two different localities than between any given brand of natural cement and a Portland cement. This will be brought out clearly in a later chapter, where the compo- 194 100 UNITED SHOWING LOCATION OF NATURAL CEMENT PLANTS FIG. 33. [To face p. 194. DEFINITION AND RELATIONS OF NATURAL CEMENTS. 195 sition and properties of the various natural cements will be discussed in considerable detail. In the present volume the term " natural cements " will be used to include all those cements which are produced by burning, without pre- vious mixing or grinding, a naturally impure limestone rock, i.e., a clayey or argillaceous limestone. As so used the term will include the class of doubtful products commonly known as " natural Portland cements ", a class which is quite largely manufactured in Belgium and France. The reasons for including these " natural Portlands " with the natural cements instead of with the true Portlands are stated in detail in a later section of this volume. The definition of natural cements giver* on a previous page can be restated here to advantage. Definition. Natural cements are produced by burning a natural clayey limestone containing 15 to 40 per cent of silica, alumina, and iron oxide without preliminary mixing and grinding. This burning takes place at a temperature that is usually little, if any, above that of an ordinary lime-kiln. During the burning the carbon dioxide of the limestone is almost entirely driven off, and the lime combines with the -silica, alumina, and iron oxide, forming a mass containing silicates, aluminates, and ferrites of lime. In case the original limestone con- tained any magnesium carbonate the burned rock will contain a corre- sponding amount of magnesian compounds. After burning, the burned mass will not slake if water be poured on it. It is necessary, therefore, to grind it quite fine, after which, if the resulting powder (natural cement) be mixed with water, it will harden rapidly. This hardening, or setting, will take place either in air or under water. Relations of natural cements to others. Natural cements differ from ordinary limes in two very noticeable ways. These are: (1) The burned mass does not slake when water is poured on it. (2) After grinding, natural-cement powder has hydraulic proper- ties, i.e., if properly prepared it will set under water. Natural cements are quite closely related to both hydraulic limes, on the one hand, and Portland cement, on the other, agreeing with both in the possession of hydraulic properties. They differ from hydraulic limes, however, in that the burned natural-cement rock will not slake when water is poured on it. The natural cements differ from Portland cements in the following important particulars: (1) Natural cements are not made by burning carefully prepared 196 CEMENTS, LIMES, AND PLASTERS. and finely ground artificial mixtures, but by burning masses of natural rock. (2) Natural cements, after burning and grinding, are usually yellow to brown in color and light in weight, their specific gravity being about 2.7 to 3.10, while Portland cement is commonly blue to gray in color and heavier, its specific gravity ranging from 3.0 to 3.2. (3) Natural cements are always biuned at a lower temperature than Portland, and commonly at a much lower temperature, the mass of rock in the kiln rarely being heated high enough to even approach the fusing- or clinkering-point. i. (4) In use natural cements set more rapidly than Portland cement, but do not attain such a high ultimate strength. (5) In composition, while Portland cement is a definite product whose percentages of lime, silica, alumina, and iron oxide vary only between narrow limits, various brands of natural cements will show very great differences in composition; while even the same brand, analyzed at different times, will show considerable differences in composition, due to variations in the natural limestones used. Cementation Index. In discussing the hydraulic limes (Chapter XIII) attention was called to the desirability of devising some method of general applicability for comparing the hydraulic activity of various cementing materials. The defects of the old " hydraulic index " were pointed out, and a new and more satisfactory index the Cementation Index was suggested as a substitute. The value of this innovation will appear in the present section, for in dealing with the natural cements such great variations in composition are found that it is absolutely necessary to have some means of comparing such different products. The Cementation Index of any limestone or cement is found by applying the following formula: (2.8 X percentage silica) + ( 1 .1 X percentage alumina) 4- Cementation Index -,= (. 7 X percentage iron oxide) ^^ (Percentage lime) +(1. 4 X percentage magnesia) When this formula is applied to an unburned limestone it must be recollected that the percentages used in the divisor are those of lime (CaO) and magnesia (MgO) respectively, not those of lime carbonate (CaCO 3 ) and magnesium carbonate (MgCO 3 ). Example of calculation. The methods of calculating the Cementa- tion Index of any product may be shown by an example, the Utica natural cement whose analysis appears as No. 1, Table 110, p. 253, being selected "DEFINITION AND RELATIONS OF NATURAL CEMENTS. 197 for this purpose. The five essential ingredients of that cement, as shown by the analysis, are: Silica (SiO 2 ) 19.89 Alumina (A1 2 O 3 ) 11 -61 Iron oxide (Fe 2 O 3 ) 1 .35 Lime (CaO) 29.51 Magnesia (MgO) 20.38 These values are substituted in the following formula: (2.8 X percentage silica) + (1.1 X percentage alumina) + _ . T , (. 7 X percentage iron oxide) Cementation Index =-^ .. . , .., . r-r (Percentage lime) + ( 1 .4 X percentage magnesia) = (2.8X19.89) +(1.1X11.61) + (.7X1.35) (29.51) + (1.4X20.38) ^55.692 + 12.771+0.945 29.51+28.532 ^69.408 " 58.042 = 1.19. As will be later seen, this value is fairly characteristic for many natural cements. Basal assumptions. It has previously been stated (pp. 170, 171) that the applicability of the Cementation Index depends upon the fact that it is the exact equivalent in percentages of a formula which involves the following assumptions: (1) That the hydraulic activity of any material depends on the formation of certain compounds of lime and magnesia with silica, alumina, and iron oxide. (2) That in a hydraulic cement, lime combines with silica in such proportions as to form the tricalcic silicate ( = 3CaO.Si0 2 ); while it combines with alumina in such proportions as to form the dicalcic aluminate (2CaO.Si0 2 ). (3) That in a lightly burned natural cement at least magnesia may be regarded as molecularly interchangeable with lime, though of course the differences in their combining weights -must be allowed for when the calculation is based on percentages. (4) That iron oxide may, in similar fashion, be regarded as molec- ularly interchangeable with alumina. Of these assumptions, the third and fourth may be questioned by other investigators, but it will be seen later that the hydraulicity of 198 CEMEXT3, LIMES, AND PLASTERS. certain well-known products cannot be explained satisfactorily without taking account of the magnesia and iron oxide they contain. Use of the Cementation Index. If the assumptions on which the Cementation Index is founded are well based, it is evident that the hydraulic properties or, rather, the hydraulic possibilities of a prod- uct are indicated by its index. A product whose index falls below 1.00 must necessarily contain free limtf or free magnesia, whatever the temperature at which it is burned, and such a product should therefore be strictly classed with the hydraulic limes, which require slaking before use. It will be seen later, however, that if a product Contains much mag- nesia (say 20 per cent MgO or over) its Cementation Index may fall below 1.00 without demonstrable defects in the cement. This point is taken up on later pages in discussing the actual composition of various natural cements. A product with an index exceeding 1.00 can be burned so as to give complete combination of all its lime and magnesia, leaving none free. As the index increases, the temperature necessary to attain such complete combination decreases, but the hydraulic activity of the product also decreases, until an index exceeding 2.00 indicates a very lightly burned, but also very feeble, cement. Cementation Index of natural cements. The term " natural cement " as used in this volume will cover a very large class of cementing prod- ucts. In the United States the name has become fairly well fixed in use, so that there need be little misunderstanding concerning the limits of the groups. In English and European practice, however, the term " natural cement " has never come into extensive use. It may therefore be necessary to state that, as above defined, it includes the lightly burned but often high-limed cements known to the Euro- pean trade as " Roman cements ", " quick-setting cements ", etc., as well as the so-called "natural Portlands ". The differences in composition between the various cements included in this heterogeneous class naturally give rise to corresponding differ- ences in their cementation index. It may be said for the group taken as a whole that the Cementation Index of natural cements varies between the limits of 1.00 and 2.00, falling below 1.00 only in the case of certain highly magnesian cements, and that most of the natural cements will fall between the narrower limits of about 1.15 to 1.60. This variation of the Cementation Index may be used as a con- venient basis for subdividing the "natural cements " into smaller groups of more homogeneous character. A. Cements with an index between 1.00 and 1.15. These products when burned at sufficiently high temperature are rather slow- DEFINITION AND RELATIONS OF NATURAL CEMENTS. 199 setting and high in tensile strength, including the " natural Portlands " and allied products. If not burned high enough, however, cements of such low index will necessarily contain large amounts of free lime and magnesia. B. Cements with an index between 1.15 and 1.60. These include most American natural cements. As the index is higher than in Class A, it is not necessary to burn these products at so high a temperature. Practically all of the European "Roman" cements will also fall in this subgroup. C. Cements with an index exceeding 1.60. These include the rela- tively low-limed natural cements, which carry so much clayey material that only a light burning is required in order to com- bine all their lime and magnesia. As the index rises above 2.00, the products become feebler in hydraulic properties, until at about 3.00 they can be considered only as artificial pozzuolauas. CHAPTER JXVII. RAW MATERIAL: NATURAL-CEMENT ROCK. Composition of natural-cement rock. The raw rnaterial utilized for natural-cement manufacture is invariably a clayey limestone carrying from 13 to 35 per cent of clayey material, of which 10 to 22 per cent or so is silica, while alumina and iron oxide together may vary from 4 to 16 per cent. It is the presence of these clayey materials which give the resulting cement its hydraulic properties. Stress is often care- lessly or ignorantly laid on the fact that many of our best-known natural cements carry large percentages of magnesia, but it should at this date be realized that magnesia (in natural cements at least) may be regarded as being almost exactly interchangeable with lime, so far as the hydraulic properties of the product are concerned. The presence -of magnesium carbonate in a natural-cement rock is then merely incidental, while the silica, alumina, and iron oxide are essential. The 25 per cent or so of magnesium carbonate which occurs in the cement rock of the Rosendale district, New York, could be replaced by an equivalent amount of lime carbonate, and the burnt stone would still give a hydraulic product. If, however, the clayey portion (silica, alumina, and iron oxide) of the Rosendale rock could be removed leaving only the magnesium and lime carbonates, the rock would lose all of its hydraulic properties and would yield on burning simply a magnesian lime. This point has been emphasized because many writers on the sub- ject have either explicitly stated or implied that it is the magnesium carbonate of the Rosendale, Akron, Louisville, Utica, and Milwaukee rocks that -causes them to yield a natural cement on burning. Even a casual consideration of the subject should have recalled to mind the fact that the Cumberland and Lehigh natural-cement rocks are practically free from magnesium carbonate. A limestone containing sufficient argillaceous matter to make a good natural cement can generally be recognized by the characteristic -clayey odor given forth when breathed on. 200 RAW MATERIAL: NATURAL-CEMENT ROCK. 201 In determining in advance of actual calcination whether or not a given rock will make a good natural cement the Cementation Index will prove of service. This can be calculated, as explained on page 196, from the analysis of the rock. If the value of the Cementation Index is over 2.00, the rock will make only a very weak sort of cement, not worth putting on the market as a new product in face "of competition from older and stronger brands. If, on the other hand, the Cementa- tion Index is less than 1.00, the rock is in most cases unavailable, for after burning it will contain too much free lime and free magnesia to furnish a safe cement. As noted earlier, however, a rock whose index falls between 0.80 and 1.00 can be made into an apparently safe cement if it contains 20 per cent or more of magnesia, by burning at a very high temperature. If the Cementation Index falls between 1.00 and 2.00 it <;an be assumed that a natural cement of good quality can be made from the rock under proper conditions of burning, etc. Within these limits the properties of the cement will vary with the index. A rock with an index of 1.00 to 1.10, for example, will require burning at high temperature, especially if much lime be present (i.e., over 50 per cent CaO). As the index rises, the temperature necessary for burning de- creases. American Natural-cement Rocks. In the following pages analyses of the rocks used at almost all of the natural-cement plants of the United States will be given. Notes on the physical character, geology, and other features of these rocks will also be presented. Clayey limestones of the composition required for natural-cement manufacture are very 'widely distributed, both geologically and geographically, in the United States. There is hardly a State, in fact, in which natural cement of more or less value has not been made at one time or another. In order, however, that a natural-cement industry can become well established in any given locality, certain things are requisite in addition to- the occurrence of a good natural-cement rock. The rock must not only be of the right composition to make a good, sound, and strong cement, but it must be fairly steady in composition, and the beds must be located favorably for cheap extraction of the rock, either by quarrying or by mining. Fuel must also be obtainable at reasonable rates. A good local market and cheap transportation to outside points are necessities. Of the many localities in the United States at which deposits of good natural-cement rock occur, so few possess the commercial advan- 202 CEMENTS, LIMES, AND PLASTERS. tages mentioned above that the important natural-cement-producing districts are correspondingly few. According to the United States Geological Survey Report for 1903 there were 65 natural-cement plants then in operation. Of these 20 were in New York State, 15 in the Louisville district of Indiana and Kentucky, 7 in the Lehigh district of Pennsylvania, 4 in Maryland, 3 in the Utica district of Illinois, 2 each in Georgia, Kansas, Minnesota, Ohio^ Texas, Virginia, and Wisconsin, and 1 each in North Dakota and West Virginia. This suffices to show the extent to which the American natural-cement industry has become concentrated in certain favorable localities. ^ Georgia. Two natural-cement plants are located in northwest Georgia, but they use cement rocks from two different geological forma- tions, and their raw materials and products differ widely in composition. The plant of the Chickamauga Cement Company is located at Ross- ville, Ga., a few miles south of Chattanooga. The raw material used is a thin-bedded slaty limestone of Chickamauga (Ordovician) age, which is 'here exposed over a considerable area. In geologic age, as well as in chemical composition, this rock is quite similar to the cement rock of the Lehigh district of Pennsylvania, but the Georgia deposit is not so thick as in that region. These shaly limestones outcrop at many points in northwest Georgia and northern Alabama, but so far have been utilized for natural cement only at the Rossville plant. The second plant, that of the Howard Hydraulic Cement Company, is working on limestones of quite different character. The Conasauga formation of the Cambrian is described by Dr. C. W. Hayes as being " normally composed at the base of thin limestones interbedded with shales, then of yellowish or greenish clay shales, and at the top of calcareous shales, grading into blue seamy limestones". The Western and Atlantic Railroad, now operated under lease by the Nashville, Chattanooga and St. Louis System, crosses the outcrop of these rocks from above Adairsville to within a mile of Kingston. At one point near the southern end of this belt limestone obtained from beds lying near the top of the Conasauga -formation has long been utilized in the manufacture of natural cement at the plant of the Howard Hydraulic Cement Company, at Cement, Bartow County, Ga., about two miles north of Kingston. In the low ridge east of the railroad at Cement station a section of these Conasauga limestones has been measured by Spencer.* * The Paleozoic group of Georgia, p. 100. RAW MATERIAL: NATURAL-CEMENT ROCK. 203 The series shown, from the top down, was as follows: Feet. "Blue limestone 8 Slaty limestone (cement rock) 4 Blue limestone 6 * Argillaceous limestone 2 * Siliceous limestone (hydraulic) 4 * Siliceous limestone (cement rock) 7 Fine black limestone 12 Earthy limestone 3 Shales When the plant was visited by the writer, in the fall of 1902, the three beds marked with asterisks (*) were being worked for natural cement. Spencer quotes the following analyses, made by W. J. Land: (1) (2) Silica (SiO 2 ) 22. 10 10.00 Alumina (A1 2 O 3 ) 5.45 6.10 Iron oxide (Fe 2 O 3 ) 1 .80 2.00 Lime (CaO) 24.36 30.80 Magnesia (MgO) 12.38 12.42 Carbon dioxide (CO 2 ).. 32.76 37.88 Organic matter 0.15 . 50 Water 1.00 0.30 Cementation Index 1 . 69 . 749 Of these analyses, No. 1 probably represents the composition of the 7-foot bed of cement rock noted in Spencer's section, while No. 2 is probably from the 4-foot bed of hydraulic limestone immediately overlying this. As these beds are both worked for cement, the index of the product lies between those of the two rocks. Illinois. Three natural-cement plants operated by two companies are now working in Illinois, all of them being located near Utica, La Salle County. The rock used is a limestone belonging to the so-called " Lower Magnesian " group of early Western geologists. It is of Ordovician age and underlies the St. Peter's sandstone. The section exposed in the neighborhood of the Utica cement-plants is as follows from the top downward: Thickness. Cement rock 7 feet Limestone 16-22 " Cement rock 6 " Sandstone 2-4 " Cement rock 5 " Of the three cement-rock beds shown in this section, the upper- most bed gives a very quick-setting cement, while the two lower beds furnish products of much slower set. 204 CEMENTS, LIMES, AND PLASTERS. TABLE 88. ANALYSES OF NATURAL-CEMENT ROCK, UTICA, ILLINOIS. 1. 2. 3. 4. 5. Silica (SiO 2 ) 12.22 17.01 1 f 14.15 Alumina (A1 2 O 3 ) i . 9.39 3.35 > 21.00 21.12 1 6.37 Iron oxide (Fe 2 O 3 ) . . 3.90' 2.39 2.00 1.12 2.35 Lime (CaO) . . 24 40 32 85 24 36 23.66 26 32 Magnesia (MgO) 10 43 8 45 14.31 15.22 12.10 Alkalies (K 2 O,Na 2 O) n. d. n d 0.18 n. d. 0.18* Sulphur trioxide (SO 3 ) n. d. 1.81 n. d. n. d. 1.81 / 34.00 35.35 34.70 Water } 38 . 48 34.12 1 3.00 1.07 2.03 Cementation Index 1 21 1 19 1.11 * Far too low: true value is probably over 4 per cent. 1. F. W. Clarke, analyst. Sample collected by E. C. Eckel. 2. C. Richardson, analyst. Brickbuilder, vol. 6, p. 151. July, 1897. 3. Blaney & Mariner, analysts. "Geology of Illinois", vol. 1, p. 151. 4. Blaney, analyst. Trans. Am. Inst. Min. Engrs., vol. 13, p. 180. 5. Average of preceding four analyses. Indiana-Kentucky. The plants of the " Louisville district " are mostly located in Indiana, though one or two mills are in operation on the Kentucky side of the Ohio River. The rock is a fine-grained clayey limestone of Devonian age. In color it varies from light drab to dark or bluish drab when fresh, weathering to a dull buff on long ex- posure. The cement-bed varies from 10 to 16 feet in thickness in the different quarries. As shown by the calculated values of the Cementa- tion Index, the rock varies greatly in composition. TABLE 89. ANALYSES OP NATURAL-CEMENT ROCK, LOUISVILLE DISTRICT, IND.-KY. 1. 2. 3. 4. 5. 6. Silica (SiO 2 ) . . . 9 69 9 80 13 65 15 21 18 33 13 36 Alumina (A1^O 3 ) 2 77 2 03 3 46 4 07 4 98 3 46 Iron oxide (Fe 2 O 3 ) 1.95 1.40 1.45 1 44 1 67 1 58 Lime (CaO) 29.09 29.40 34.55 33 99 30 41 31 49 Magnesia (MgO) 15 69 16 70 7 97 7 57 8 04 11 19 Carbon dioxide (CO 2 ). . . Cementation Index 40.14 0.618 41.49 35.92 35.03 32.76 1 39 37.07 Analyses 1-5 inclusive were made by W. A. Noyes. Quoted by Siebenthal, 25th Ann. Rep. Indiana Dept. Geology and Natural Resources, pp. 380-386. 1. Rock used for "Crown" brand, Hausdale mill, New Albany Cement Company. 2. "Fern Leaf" brand, Ohio yalley mill, Ohio Valley Cement Company. 3. "Diamond" brand, Falls City mill, Union Cement and Lime Co. 4. "Star " -brand, Speed mill, Louisville Cement Company. 5. '* 'Black Diamond", Black Diamond mill, Union Cement and Lime Co. RAW MATERIAL: NATURAL-CEMENT ROCK. 205. Kansas. The natural-cement district of Kansas is located around Fort Scott, where a 44-foot bed of natural-cement rock outcrops. The rock is a dark-colored, fine-grained, compact limestone of Carboniferous age. It extends for a considerable distance throughout the State, but as yet has been worked for natural cement only in the immediate vicinity of Fort Scott. TABLE 90. ANALYSES OF NATURAL-CEMENT ROCK, FORT SCOTT, KANSAS. 1. 2. 3. 4. Silica (SiO ? ) 15 21 17 26 21 80 18 09 4.56 2 05 3 70 3 44 Iron oxide (Fe Oo) n d 5 45 3 10 4 27 36.52 34.45 35.00 35 32 Masjnesia (MffO) 5 07 5 28 3 50 4 62 Carbon dioxide (CO 2 ) 34 27 32 87 33 00 33 38 Cementation Index 1 68 1. Smith, Mineral Industry, vol. 1, p. 49. 2. Brown, "Cement Directory", 2d ed., p. 276. 3. Richardson, Brickbuilder, vol. 6, p. 151. July, 1897. 4. Average of preceding analyses. Maryland. The natural-cement industry of Maryland has been carried on in three separate areas. One of these areas including the old plants at Antietam and Shepherdstown will be described later under the heading of West Virginia-Maryland. The other two areas include respectively the plants at Cumberland and Potomac in Alle- gany County, and that at Round Top or Hancock in Washington County. In both of these areas the limestones used are of the same geologic age, and approximately of the same composition, so that they will here be described together. Analyses are given in Table 90. In geologic age the natural-cement rock of the Cumberland-Han- cock district corresponds closely to that used in the various New York districts, being assigned by geologists to the Salina group of the Silurian. It is a shaly limestone, varying in color from dark bluish gray to dull black. In the Cumberland area it is exposed in four beds of sufficient thickness to be worked, these cement-beds being separated by shales and limestones. The separate beds vary from 6 to 17 feet in thickness. Minnesota. Two natural-cement plants are in operation in Minnesota. One of them is located at Mankato, Blue Earth County, and uses a limestone of the Lower Magnesian (Ordovician) series. The analyses (Table 91) of the raw material used at this plant have been published. 206 CEMENTS, LIMES, AND PLASTERS. TABLE 91. ANALYSES OP NATURAL-CEMENT ROCKS, CUMBERLAND AND HANCOCK, MARYLAND. 1. 2. 3. 4. 5. Silica (SiO 2 ) 19.81 24.74 ] f 28.72 22.07 7.35 16.74 > 27.1 \12.28 12.12 Iron oxide (Fc 2 Oo) ....... 2 -.41 6^0 1.5 5.22 3.36 Lime (CaO) 35.76 23141 36.40 25.54 30.28 2.18 4.09 2.52 1.10 2.47 Alkalies (Na 2 O,K 2 O) n. d. 6.18 0.3 n. d. * n. d. 2.22 n. d. 1.53 * Carbon dioxide (CO 2 ) /22.90 31.38 Water i 31 . 74 \ n. d. n. d. > 24 . 40 27.60 Cementation Index 1 68 3.15 1.62 3.60 2.29 * Data insufficient for averaging. 1. Hancock, Md. C. Richardson, analyst. Brickbuilder, vol. 6, p. 151. .July, 1897. 2. Cumberland, Md. E. C. Boynton, analyst. Quoted by Gillmore, "Limes, Cements, and Mortars", p. 125. 3. Hancock, Md. C. Huse, analyst. Quoted by Gillmore, "Limes, Cements, and Mortars", p. 125. 4. Cumberland, Md. C. Richardson, analyst. Brickbuilder, vol. 6, p. 151. July, 1897. 5. Average of preceding four analyses. TABLE 92. ANALYSES OF NATURAL-CEMENT ROCK, MANKATO, MINN. 1. 2. 3. 4. 5. 6. Silica (SiO ? ) 16.00 12.14 10.10 16.80 8.90 11.80 Alumina (A1 2 O 3 ) 5 85 4 62 2 78 8.76 3 30 3 46 Iron oxide (Fe 2 O 3 ) 2 73 1.84 1.34 tr. 1 02 tr. Lime (CaO) 22.40 22.66 25.96 22.20 24 85 24.64 Magnesia (MgO) 14.99 16.84 14.91 11.99 18.49 16.61 Alkalies (K 2 O Na 2 O) . . . 76 3 52 3 50 4 75 1 53 2 59 Sulphur trioxide (SO 3 ) n d 13 26 22 18 22 Carbon dioxide (CO 2 ) 34 11 39 07 41 29 35 90 41 80 40 85 Cementation Index 1 22 88 69 1 45 58 77 1. C. F. Sidener, analyst, llth Ann. Rept. Minn. Geol. Surv., p. 179. 2-6. Clifford Richardson, analyst. Cement Directory, p. 206. The second plant is said to be located at Austin, Mower Co. If this location be correct, the limestone used is probably of Devonian ANALYSIS OF NATURAL-CEMENT ROCK, AUSTIN, MINN. Silica (Si0 2 ).... l Alumina (A1 2 O 3 ) J D '^ Iron oxide (Fe 2 O 3 ) 2 . 09 Lime (CaO) 27.55 Magnesia (MgO) 13 .80 Sulphur trioxide (SO 3 ) .06 Carbon dioxide (CO 2 ) . 36.84 RAW MATERIAL: NATURAL-CEMENT ROCK. 207 New York. In the State of New York natural cement is now manu- factured in four distinct localities. These are in order of importance: (1) the Rosendale district in Ulster County, (2) the Akron-Buffalo dis- trict in Erie County, (3) the Fayetteville-Manlius district, mostly in Onondaga County, and (4) at Howe's Cave in Schoharie County. The clayey limestones used in these four districts occur in three different but closely related geological formations, all in the Upper Silurian group. The sequence and relation of these formations, from the top downwards, is shown in the following table. Formation. Ulster County. Schoharie County. Onondaga County. Erie County. Manlius limestone (cement rock). ' Worked for cement at Manlius, etc. Absent. Rondout limestone (cement rock). Upper cement- bed of the Rosendale district. Worked for cement at Howe's Cave. Absent. Cobleskill limestone (not used for ce- ment). Bertie limestone (cement rock). Lower cement- bed of the Rosendale district. Present in On- o n d a g a County but rarely used for cement. Worked for cement at Akron and Buffalo. For convenience these districts will be described not in the order of their relative importance but in geographic order, from east to west. The Rosendale district lies entirely in Ulster County, the principal cement-rock quarries being located at East Kingston, Rondout, Rosen- dale, Binnewater, Lawrenceville, and High Falls. Two distinct beds are worked at most of these points, differing in chemical composition as well as in geological age. Darton states * that at Rosendale the lower bed, or dark cement rock, averages about 21 feet in thickness, and the upper, or light cement rock, about 11 feet, the two cement- beds being here separated by 14 or 15 feet of worthless limestone. The lower bed lies directly on the Clinton quartzite, the even upper surface * 13th Ann. Rept. N. Y. State Geologist, vol. 1, 1894, p. 334. 208 CEMENTS, LIMES, AND PLASTERS. of which affords an admirable floor for the galleries. For about 18 inches at the bottom the dark cement rock is too sandy for use. With this exception and a few small layers of chert it is all available. At Whiteport the upper bed is 12 feet thick and the lower 18 feet, while they are separated by 17 to 20 feet of limestone. . TABLE 9.3. f ANALYSES OF NATURAL-CEMENT Rock, ROSENDALE DISTRICT, N. Y. 1. 2. 3. 4. 5. 6. 7. 8. 18.52 6.34 2.63 25.31 12.13 n. d. 0.90 33.31 n. d. 1.43 Silica (SiO 2 ) 10.90 3.40 2.28 29.57 14.04 n. d. 0.61 37.90 n. d. 15.37 9.13 2.25 25.50 12.35 n. d. n. d. 34.20 1.20 18.11 4.64 3.00 24.30 14.26 n. d. tr. 34.01 n. d. 18.76 8.34 1.85 25.96 11.07 n. d. 1.35 32.00 n. d. 21.32 7.39 1.71 23.75 11.07 n. d. 1.90 30.74 n. d. '"21.41 } 10. 09 I 25.80 10.09 n. d. 0.66 J30.93J 23.80 4.17 4.71 22.27 12.09 n. d. 0.90 31.00 n. d. Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) Lime (CaO) Magnesia (MgO) Alkalies (K 2 O,Na.,O) Sulphur trioxide (SO 3 ) . ~ Carbon dioxide (CO 2 ) Water Cementation Index 1. Lawrenceville. J. O. Hargrove, analyst. Letter to writer, Oct. 4, 1900. 2. Rondout. L. C. Beck, analyst. "Mineralogy of N. Y.", p. 78. 3. Lawrenceville. J. O. Hargrove, analyst. Letter to writer, Oct. 4, 1900. 6. " Rosendale district. C. Richardson, analyst. Brickbuilder, vol. 6, p. 151. July, 1897. 7. Lawrenceville. J O. Hargrove, analyst. Letter to writer, Oct. 4, 1900. 8. Average of preceding seven analyses. Northward and northwestward from the Rosendale-Rondout dis- trict no natural-cement plants are to be found until Schoharie County is reached. Here, at Howe's Cave, a single plant has long been engaged in the manufacture of cement from a 7-foot bed of rock. TABLE 94. ANALYSES OF NATURAL-CEMENT ROCK, SCHOHARIE COUNTY, N. Y. 1. 2. 3. Silica (SiO 2 ) .-, 12 89 9 92 1 Alumina (A1 2 O 3 ) . . . } -, -.r f n d > 11.50 Iron oxide (Fe 2 O 3 ) J11.15J n d 1 50 Lime (CaO) 30 90 38 26 31 75 Magnesia (MgO) . . . . 9 38 9 00 14.91 Carbon dioxide (CO 2 ) 34.60 39.96 40.34 Cementation Index. 1 07 1. Bottom of cement-bed, Howe's Cave. Geologist, p. 69. 2. Top of cement-bed, Howe's Cave. C. O. Schaeffer, analyst. 69. C. O. Schaeffer, analyst. 18th Ann. Rept. N. Y. State 18th Ann. Rept. N. Y. State Geologist, p. 3. Howe's Cave. L. C. Beck, analyst. "Mineralogy of New York", p. 79. RAW MATERIAL: NATURAL-CEMENT ROCK. 209 The cement industry in central New York is at present practically confined to Onondaga County, though as a matter of historical interest it may be noted that the first natural cement made in the United States was manufactured in 1818 in Madison County. The natural-cement rock of this central district occurs in two beds which are usually separated by 1 to 4 feet of blue limestone. The upper cement-bed is a little over 4 feet thick at the eastern border of Onondaga County, becoming thinner to the westward until it pinches out entirely in the the Split Rock quarries, but reappearing again at Marcellus Falls, where it is almost 3 feet thick and showing a thick- ness of slightly over 4 feet at Skaneateles Falls. At this last point it is separated from the lower cement-bed only by a shaly parting a few inches thick, so that the two are worked together as practically one bed 9J feet thick. The lower bed is less variable in thickness, ranging from 4 to a trifle over 5 feet. The entire cement series is overlaid by purer limestones, but the cement-rock quarries are usually located at points where these over- lying limestones are thin and can be readily stripped. TABLE 95. ANALYSES OF NATURAL-CEMENT ROCK, CENTRAL NEW YORK. 1. 2. 3. 4. 5. 6. Silica (SiO 2 ) 10 97 10 95 \ f 8 95 11 76 10 66 Alumina (A1 O 3 ) 4.46 5.32 > 13.50 \4.90 2 73 4 35 Iron oxide (Fe 2 O 3 ) 1.54 1.30 1.25 1.75 1.50 1 47 Lime (CaO) 27 51 30 92 25 24 27 35 25 00 27 20 Magnesia (MgO) 16 90 13 64 18 80 16 70 17 83 16 77 Carbon dioxide (CO 2 ) 37 94 38 31 39 80 38 65 39 33 38 81 Water n. d. n. d. 1.41 1.70 1 50 1 53 Cementation Index 0.71 1. Upper cement-bed, E. B. Alvord quarry, Jamesville, Onondaga County. Bull. 44 N. Y. State Mus., p. 806. 2. Lower cement- bed, E. B. Alvord, quarry , Jamesville, Onondaga County. Bull. 44 N. Y. State Mus., p. 806. 3. One and one half miles west of Manlius, Onondaga County. L. C. Beck, analyst. "Mineralogy of New York", p. 81. 4. One and one half miles southwest of Chittenango, Madison County. L. C. Beck, analyst. "Mineralogy of New York", p. 80. 5. Chittenango, Madison County. Seybert, analyst. Trans. Am. Philos. Soc., vol. 2, n. a., p. 229. 6. Average of preceding five analyses. In Erie County natural-cement plants have long been established at Akron and Buffalo. The bed of cement rock used varies in thick- ness from 5 to 8 feet. " It is a firm, fine-grained compact rock of a blue-gray color, weathering to a yellowish white. The Buffalo plant 210 CEMENTS, LIMES, AND PLASTERS. works its cement rock by quarrying methods, stripping off the over-, lying limestones; but the plants at Akron all obtain their raw mate> rial by mining. TABLE 96. ANALYSES OF NATURAL-CEMENT ROCKS, AKRON-BUFFALO DISTRICT, NEW YORK. y . f* l. 2. 3. 4. Silica (SiO 2 ) 9 03 10 68 33 80* 9 85 Alumina/ ( Al 2 Oo) 2 25 \ . ., f 3 96 3 10 Iron oxide (Fe 2 O 3 ) 85 } 4 - 61 i 88 87 Lime (CaO) .... 26 84 25 65 19 93 26 25 Magnesia (MgO) 18 37 17 93 9 17 18 15 Alkalies (K 2 O,Na 2 O) 85 n d n d Sulphur trioxide (SO 3 ) n d n d 50 Carbon dioxide (CO 2 ) 40 33 Water 98 } 25 . 90 Cementation Index 610 * Called "silica, clay, and insoluble silicates". 1. G. Steiger, analyst. Bulletin 168, U. S. Geol. Survey. 2. Lathbury and Spackman, analysts. 3. E. Boynton, analyst. Gillmore, "Limes, Cements, and Mortars", p. 125. 4. Average of analyses 1 and 2. I am informed by Mr. Uriah Cummings, of the Cummings Cement Co., Akron, N. Y., that none of the analyses given in Table 96 are really repre- sentative of the Akron natural-cement rock. The analyses are presented, therefore, subject to this criticism. North Dakota. The single natural-cement plant operating in this state is located about ten miles east of Milton, Cavalier County. The rock used is a soft, chalky limestone of Cretaceous age and outcrops in a bluff several hundred feet high. At present, however, only a 10-foot bed is being worked, by mining. (See Table 97.) Ohio. Small natural-cement plants have been established at vari- ous points in Ohio, those at Defiance and New Lisbon being worthy of some notice. The Defiance plant used a black calcareous shale of Devonian age. If published analyses be Correct (see Nos. 1 and 2 in Table 98) this rock is by far the most argillaceous material used anywhere for this purpose. Pennsylvania. A fairly large production of natural cement has always been maintained in the Lehigh district of eastern Pennsylvania, though at present natural-cement manufacture there is merely inci- dental to the great Portland-cement industry of the district. The analyses given in Table 99 purport to be representative of the rock used at various Lehigh district natural-cement plants. It is hardly RAW MATERIAL: NATURAL-CEMENT ROCK. 211 TABLE 97. ANALYSES OF NATURAL-CEMENT ROCK, NORTH DAKOTA. l. 2. 3. 4. 5. Silica (SiO 2 ) 14 00 16 60 13 10 16 20 16 54 Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) > 6.70 7.10 7.60 7.56 8.20 Lime (CaO) 37 60 35 50 37 80 35 10 35 20 Sulphur trioxide (SO 3 ) 0.58 0.60 n. d. n. d n. d. Sulphur (S) 1 45 1 38 n d n d n d Cementation Index 1 24 6. 7. 8. 9. 10. Silica (SiO 2 ) 14 90 15 24 19.20 17 36 16 00 Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) \ 8.28 7.26 8.90 8.78 7.50 Lime (CaO) 36.90 36.70 32.60 34.90 35.60 Sulphur trioxide (SO 3 ) n. d. 0.40 n. d. n. d. 0.67 Sulphur (S) n d 1 99 n d n d 1 61 Cementation Index 1.92 TABLE 98. ANALYSES OF NATURAL-CEMENT ROCKS, OHIO. 1. 2. 3. 4. 5. Silica (SiO 2 ) 39.95 42.0 16.41 30.60 15 65 Alumina (A1 2 O 3 ) I 20 . 22 < 7.0 5.44 6.8 Iron oxide (Fe 9 Oq) 7 1 3 38 \ 13.00 | 2 5 Lime (CaO) 10 06 9 91 26 05 22 74 38 64 Magnesia (MgO) . . . 2 92 5.81 12 55 7 23 1 62 Carbon dioxide (CO 2 ) Water and organic 124.03J 14.18 14.0 . 34.32 n. d. 25.81 n. d. 32.14 n. d. Cementation Index 1 25 1. Defiance. J. E. Whitfield, analyst. Bull. U. S. Geol. Survey No. 55, p. 80. 2. Defiance. R. C. Kedzie, analyst. Cement Directory. 3. Bellaire. N. W. Lord, analyst. Repts. Ohio Geol. Surv., vol. 6, p. 673. 4. Warnock. Wormley, analyst. Rept. Ohio Geol. Surv., 1870, p. 451. 5. New Lisbon. N. W. Lord, analyst. Rept. Ohio Geol. Surv., vol. 6, p. 673. necessary to say that Nos. 1 and 3 are absolutely unfit for such use. No. 2, on the other hand, is quite satisfactory. I regret that these very untrustworthy analyses are, at present, the only ones available. Texas. The second analysis on p. 212 has been published * as repre- senting the average of the material used in making natural cement by a * 22d Ann. Rept. U. S. Geol. Survey, pt. 3, p. 737. 212 CEMENTS, LIMES, AND PLASTERS. TABLE 99. ANALYSES OF NATURAL-CEMENT ROCK, LEHIGH DISTRICT, PA. 1. 2. 3. 4. Silica (SiO 2 ) 11 62 18 34 27 77 19 24 Alumina (A1 2 O 3 ) Iron oxide (F e 2 O 3 ) ^ j 6.25 7.49 14.29 9.34 Lime (CaO) '. :'. 44 20 37 60 29 94 37 25 Magnesia (MgO) 1 27 1 38 1 55 1 40 Carbon dioxide (CO 2 ) 36 11 31 06 \ Wd,ter n d 3 94 J 26.30 32.47 Cementation Index 843 1 49*- 2 87 1. Siegfried, Pa. Mineral Industry, vol. 1, p. 49. 2. Coplay, Pa. Mineral Industry, vol. 1, p. 49. 3. Lehigh district. Quoted by C. Richardson. Brickbuilder, vol. 6, p. 151. July, 1897. 4. Average of preceding three analyses. Texas natural-cement plant. It is obvious that, if this statement be correct, the product obtained by burning a rock of such composition cannot be a natural cement in any proper use of the term. It would, in fact, be merely a very weak hydraulic lime. 5.77 2 M ANALYSIS OF NATURAL-CEMENT ROCK, TEXAS. Silica (Si0 2 ) ..................................... Alumina (A1 2 O 3 ) ................................. Iron oxide (Fe 2 (X) ........................ ' ....... Lime (CaO) ..... .- ................................ 50.45 Magnesia (MgO) ................................. .28 Virginia. For many years natural cement has been burned near Balcony Falls, Rockbridge County, Va. The rock used is a clayey magnesian limestone of Cambrian age, closely related geologically and technologically to that described below as being used in West Virginia. TABLE 100. ANALYSES OF NATURAL-CEMENT ROCK, VIRGINIA. 1. 2. 3. Silica (SiO 2 ) 17 38 17 21 17 30 Alumina (A1 2 O 3 ) tr 6 18 Iron oxide (Fe 2 O 3 ) } 7 - 80 { 1 62 1 62 Lime (CaO) 34 23 24 85 29 54 Magnesia (MgO) 9 51 16 58 13 05 Carbon dioxide (CO 2 ) 30.40 37 95 34 17 Cementation Index 1 18 1. Balcony Falls. E. C. Boynton, analyst. Gillmore, "Limes, Cements, and Mortars", p. 125 2. Balcony Falls. C. L. Allen, analyst. "The Virginias", vol. 3, p. 88. 3. Average of preceding two analyses. RAW MATERIAL: NATURAL-CEMENT ROCK. 213 West Virginia-Maryland. A wide belt of magnesian limestones of Cambrian age crosses Maryland into the eastern part of West Virginia. Several small natural-cement plants have been established in this district at various times, particularly near Antietam, Md., and Shepherdstown, W. Ya. ANALYSIS OF NATURAL-CEMENT ROCK, ANTIETAM Silica (Si0 2 ) Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ). . . . Lime (CaO) Magnesia (MgO) Alkalies (K 2 O,Na 2 O). . Sulphur trioxide (SO 3 ). Carbon dioxide (CO 2 ). . Water. . MD. (i) 15.97 23. 72 15.6 n. d. 0.71 34 82 Cementation Incfex 1 . 14 (1). Antietam, Md. C. Richardson, analyst. Brickbuilder, vol. 6, p. 151. Wisconsin. Two plants in Wisconsin are engaged in the manufacture of natural cement from a clayey magnesian limestone of Devonian age. These plants are located north of Milwaukee, near the lake. The cement- rock deposit is very thick, compared to most natural-cement proposi- tions, a quarry face 22 feet high being worked by the Milwaukee Cement Company. TABLE 101. ANALYSES OP NATURAL-CEMENT ROCKS, MILWAUKEE DISTRICT, Wis. 1. 2. 3. 4. 5. Silica (SiO ) 17 00 17 56 17 56 16 99 17 28 Alumina (A1 2 O.>) 4 25 1 41 1 40 5 00 3 02 Iron oxide (Fe 2 O 3 ) 1.25 3.03 2.24 1.79 2.21 Lime (CaO) 24.Q.4 25.50 27 14 23.15 25.11 Magnesia (MgO) 11.90 15 45 13 89 16.60 14.46 Carbon dioxide (CO 2 ) 32.46 37.05 36.45 36.47 n. d. 1.18 1. Mineral Industry, vol. 6, p. 95. 2. Trans. Am. Inst. Min. Engrs., vol. 8, p. 507. 3. " " " " " " * 4. " " " " " " " " 5. Average of preceding four analyses. 214 CEMENTS, LIMES, AND PLASTERS. European Natural-cement Rocks. As noted later, the European natural-cement industry dates back to 1796, in which year the manufacture of natural cement was com- menced simultaneously in France and England. At present the indus- try is established in practically everv^ country of Europe, though it is of course subject to severe competition from Portland cement on the one hand and the better class of hydraulic limes on the other. Euro- pean natural cements form two fairly distinct classes, which are there called respectively "natural Portlands " and " Roman cements ": (1) The natural Portlands, which are described on pages 215-217 in detail, are natural cements of low cementation index (1.05 to 1.15 usu- ally), low in magnesia, and burned at fairly high temperatures. In consequence of the combination of their low index and relatively high burning, these products approach true Portland cements in analysis and physical properties, though they necessarily vary considerably from time to time according to the rock from which they are made. The best of these products will pass low-grade Portland tests, and were formerly largely exported to this country, where they were unloaded on the architects and engineers who specified " foreign Portland cement ". The poorer " natural Portlands " are often adulterated with slag or unburned limestone, in order to make their bulk composition agree on analysis with that of true Portlands. While the " natural Portlands " are often useful products, there seems to be no reason for classing them with the true Portlands, for the term Portland is now understood to imply that a very careful and finely ground artificial mixture has been made before burning. (2) The Roman cements form the second class of European natural cements. They are usually cements of moderately high index (1.20 to 1.60), and are also usually but not invariably low in magnesia. They correspond therefore quite closely, so. far as index is concerned, to the best of the American natural cements. In American practice, how- ever, low-magnesia natural cements are quite rare, as can be seen by referring to the tables of analyses on pages 253 to 260, while in Europe high-magnesia cements are very scarce. It may be of interest to call attention here to the fact that Quick- setting artificial cements of high index (1.20 to 1.60), and very lightly burned, have been made at several points in Europe. These products would correspond in every way with our natural cements, except in that they are made from artificial mixtures and are therefore more expensive RAW MATERIAL: NATURAL-CEMENT ROCK. 215 to make. Wherever a good natural-cement rock is obtainable, therefore, these "artificial-natural" cements are driven out of the market. Natural-cement materials of Belgium. " Natural Portland ". cement, as well as Roman cement, is extensively manufactured in cer- tain parts of Belgium. The following analysis is representative of the composition of the rock from which these " natural Portlands " are made: Silica (SiO 2 ) 15.75 Alumina (A1 2 O 3 ) 3.95 Iron oxide (Fe 2 O 3 ) 1 .00 Lime (CaO) 43. 10 Magnesia (MgO) . . . 0.49 Sulphur trioxide (SO 3 ) . 50 Carbon dioxide (CO 2 ) 1 __ Water / d< Cementation Index 1 . 12 It will be seen that if rock of the composition represented by the above analysis could be steadily obtained it would certainly be an excellent natural mixture for a Portland cement. In composition it is admirable, while its index is about that of the average commercial Portland. In practice, however, the variations in composition of the rocks from various parts of the quarry are sufficient to prevent the product from being sufficiently uniform to be considered a Portland cement in our modern use of that term. This will be seen on referring to the analyses of these Belgian products given on page 261. A report* on the Belgian cement industry states: "The most important center for the manufacture of natural and artificial Portland cement in Belgium is the calcareous district of Tournai, in the consular district of Brussels. Some of the quarries in this district date back several centuries, when they were principally worked for building stone and for the manufacture of hydraulic lime. The cal- careous stone of these quarries, which originated the now extensive and important industry of cement-manufacture, extends for many miles in length in apparently inexhaustible quantity. Ordinary lime, best hydraulic lime, slow-setting cement (Portland), and quick-setting cement (Roman) are especial products of these immense quarries. " Natural Portland cement is obtained from calcareous stone, which is carefully analyzed and dosed, calcined in coke-heated kilns, and, after burning, finely pulverized. Before burning, the stone presents a fine, * Roosevelt, R. W. Manufacture of Portland cement in Belgium, 216 CEMENTS, LIMES, AND PLASTERS. close grain and is of a peculiar pasty appearance. Prior to calcina- tion the stone is carefully analyzed, to ascertain the exact quantity of lime, as well as other chemical properties, it may contain. The stone loses about one third of its weight during the process of burning, which also changes it to a brown tinge. When withdrawn from the kiln the cement is put under sherds to thoroughly cool before being ground. After grinding arid before Hbeing barreled it is put into pits .and left undisturbed for two months. " Natural Portland cement was first manufactured in Belgium in 1882 The following are the principal works now engaged in this enter- prise: Compagnie Generate des Ciments Portland de 1'Escaut, Dumon et Cie, and Goblet, Delward et Cie, Tournai; Societe Anonyme de Chercg, Chercg; L' Union Fraternelle, La Franco-Beige, and Dutoit & Tell freres, Calonne; Lemain & Fleury, Lampe et Cie, and Societe Anonyme du Can- aan, Antoing; V ve Alex. Dapsens and Duquesne et Cie, Vaulx; Laurent Delvigne & fils, Gaurany; the United Ghent- Antwerp Portland Cement Works, Ghent. With few exceptions, these establishments have formed a syndicate, under the name of ' Mutualite Commerciale des Ciments Beiges', with headquarters at Tournai. The society sells about 1,200,000 barrels of cement annually. The syndicate has adopted as a trade-mark the figure of a hammer. Any firm, however, of the syndicate having a trade-mark is privileged to use it. For instance, those firms having the well-known ' rhinoceros ', ' trowel ', l sword ', etc., use them in conjunction with the syndicate trade-mark. Inde- pendently of the trade-marks of the manufacturers, important buyers of the Mutualite who have labels enjoying a certain reputation are permitted to affix them on the barrels. It is stated that the principal object of this arrangement by the Belgian manufacturers is to warn and protect purchasers against persons who purchase Roman cement for export without mark or label and unknown and unauthorized by the manufacturers have Portland- cement labels affixed to the barrels at port of shipment. The Mutualite has sent circulars to interested parties in the United States informing them of the measures adopted to prevent this fraud. Hereafter all barrels containing cement manu- factured, shipped, or sold by this syndicate will be fire-marked as fol- lows: First quality, inside and outside of barrel, 'Portland Warranted M. C. B.'; second quality, inside only, ' Deuxieme quality'; third quality, inside only, ' Ciment Romain '. " Roman cement is also manufactured in the Tournai district. It is much cheaper than Portland cement, the selling price being about 50 per cent less than the Portland. It is much employed in Belgium, RAW MATERIAL: NATURAL-CEMENT ROCK. 217 replacing, advantageously, a good hydraulic lime. Manufacturers, however, will not guarantee it, as it is made of refuse stone not suit- able for the manufacture of Portland cement. It has a natural light- yellow color. Cinders are very often added, changing it to a grayish 2.00 4.50 7.27 Cementation Index 1 19 1. Quoted by Cummings. "American Cements", p. 35. "Brockett's Double Star 2. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. 3. "Tests of Mstals, etc., at Watertown Arsenal", for 1897, p. 403. brand. Maryland. The Cumberland and Hancock cements are of partic- ularly high index, carrying large percentages of clayey matter and prac- tically no magnesia. Minnesota. Both brands of Minnesota cements are high in mag- nesia and of very low index COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 255 TABLE 113. ANALYSES OF NATURAL CEMENTS, POTOMAC DISTRICT, MD. (Average index, excluding No. 7, = 1 . 95.) 1. 2. 3. 4. 5. Silica (SiO 2 ) 25 70 28 02 28 30 28 36 28 38 Alumina (A1 O 3 ) 12 28 10 20 10 12 9 85 11 71 Iron oxide (Fe 2 O 3 ) 4 22 8 80 - 4 42 *3 07 2 29 Lime (CaO) 52 69 44 48 49 60 45 04 43 97 Magnesia (MgO) 1.44 1 00 3 76 2 82 2 21 Carbon dioxide (CO 2 ) Water Cementation Index 1 61 2 09 1 70 1 88 1 99 6. 7. 8. 9. 10. Silica (SiO ) 30 02 36 60 29 74 30 22 29 66 Alumina (Al Oo) . 13 55 14 58 8 34 8 38 Iron oxide (Fe 2 O 2 ) 3 00 5 12 4 14 5 38 } 14.76 Lime (CaO) . ~" 44 58 37 50 45 66 39 54 41 96 Magnesia (MgO) . . 2 76 2 73 2 86 3 80 3 19 Carbon dioxide (CO 2 ) . . 1 Water / : 8.13 10.20 7.97 Cementation Index . 2 08 2 95 1 92 2 18 2 11 1. Cumberland, Md. A. W. Dow, analyst. Mineral Industry, vol. 6, p. 96. 2. Hancock, Md. Quoted by Cummings. "American Cements", p. 36. 3. Cumberland, Md. A. W. Dow, analyst. Mineral Industry, vol. 6, p. 96. 4. Hancock, Md. 5. Cumberland, Md. Quoted by Cummings. "American Cements", p. 36. 6. Hancock, Md. A. W. Dow, analyst. Mineral Industry ; voL 6, p. 96. 7. Cumberland, Md. A. W. Dow, analyst. Mineral Industry, vol. 6, p. 96. 8. Hancock, Md. C. Richardson, analyst. Brickbinlder, vol. 6, p. 229. 9-10. Cumberland, Md. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. TABLE 114. ANALYSES OF MINNESOTA NATURAL CEMENTS. 1. 2. 3. 4. 5. 6. Silica (SiO 2 ) 21 36 28 43 16 24 18 59 19 02 27 70 Alumina (A1 2 O 3 ) 3 34 6 71 5 35 9 14 8 96 7 06 Iron oxide (Fe 2 O 3 ) 3 80 1 94 4 71 1 00 1 24 1 86 Lime (CaO) 45 51 36 31 38 53 40 70 41 18 37 00 Magnesia (MgO) 15 02 23 89 22 73 27 00 26 58 22 63 Alkalies (K 2 O,Na 2 O). Sulphur trioxide (SO 8 ) 2.03 1 94 1.80 n d 2.30 n. d n. d. n. d. n. d. 1 27 n. d. 1.23 Carbon dioxide (CO 2 ) /9.26 / 1.75 2.46 Water > 10.00 0.92 1 51 [3.57 1 n d n d Cementation Index 994 1 2C 792 800 816 1 26 1. Mankato. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. 1897. 2. Quoted by Cummings, "American Cements", p. 36. 3. C. F. Sidener, analyst, llth Ann. Rep. Minnesota Geol. Survey, p. 179. 4. Austin. Quoted by Cummings, " American Cements ", p. 36. 5. " Tests of Metals at Watertown Arsenal, 1901. 6. Mankato. " " " 256 CEMENTS, LIMES, AND PLASTERS. New York. The cements of the Rosendale district are the typical American natural cements, with rather high index, and carrying 15 to 20 per cent of magnesia. TABLE 115. ANALYSES OF NATURAL CEMENTS, ROSENDALE DISTRICT, N. Y. l.* 2. 3. 4. 5. 6. Silica (Si0 2 ) Alumina (A1 2 O 3 ) .... 25.91 6 20 27.98 7 28 24.30 7 22 27.75 5 50 30.84 7.75 25.92 Iron oxide (Fe 2 O 3 ) 3.81 1.70 5.06 4.28 2.11 I 9.40 Lime (CaO). ........... Maflsrnesia (MgO) . 34.62 20 92 37.59 15 00 33.70 20 94 35.61 21 18 34.49 17 77 33.18 19 61 Alkalies (K 2 O,Na 2 O). . . . Sulphur trioxide (SO 3 ). . . Carbon dioxide (CO 2 ) Water n. d. n. d. 5.09 2.80 7.96 n. d. } 2.49 n. d. n. d. /n. d. |n. d. tr. 0.5 4.05 n. d. 4.00 n. d. } 3.04 n. d. n. d, 4.40 Cementation Index 1 29 1 33 7. 8. 9. 10. 11. 12. Silica (SiO ) 30 50 30 78 24 42 22 77 29 00 28 91 Alumina (A1 2 O) 6 84 J8 16 / 10 96 Iron oxide (Fe 2 O 3 ) 2 42 | 8.68 1 3 96 J 10.43 10.40 { 4 68 Lime (CaO) 34 38 34.14 36 30 34 54 32 35 34 64 Magnesia (MgO) 18.00 19.61 16.93 21.85 19 92 14 82 Alkalies (K 2 O,Na 2 O). . . . Sulphur trioxide (SO 3 ). . . Carbon dioxide (CO 2 ). . . . Water 3.98 n. d. 1 3.78 1.62 n. d. 3.57 n. d. n. d. /n. d. In d 3.63 1.44 2.84 1 59 n. d. n. d. n. d. 1 n d j 1.80 1.04 4.50 Cementation Index .... 1 74 13. 14. 15. 16. 17. 18. Silica (SiO 2 ) 29 84 27 30 21 73 17 17 27 00 29 98 Alumina (A1 2 O 3 ) I 15.20 [7 14 11 18 [6 88 Iron oxide (Fe 2 O 3 ) Lime (CaO) 35 84 11.80 35 98 4.14 33.77 | 10 . 80 48.28 17.50 35.35 12-50 33.23 Magnesia (MgO) 14.02 18.00 21.20 19.13 14.75 17.80 Alkalies (K 2 O,Na 2 O) .... Sulphur trioxide (SO,). . . Carbon dioxide (CO 2 ). . . Water n. d. 0.93 } 3.73 6.80 n. d. 2.98 2.99 n. d. fn. d. \n. d. tr. 1.20 3.38 n. d. n. d. 1.41 } 4.68 7.10 n. d 3.13 Cementation Index 1 78 1 39 1 19 * See opposite page for references. COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 257 TABLE 115. Continued. 19. 20. 21. 22. 23. Silica (SiO ) 31 28 22 75 25 00 28 71 26 66 Alumina. (A1 2 O 3 ) 1-10* / 13 40 8 93 5 88 11 48 Iron oxide (Fe 2 O 3 ) j il.SO ( 3 30 2 27 3 60 3 02 Lime (CaO) 36 67 37 60 39 30 27 00 38 33 Magnesia (MgO) 14 35 16 65 16 18 30 00 16 41 Alkalies (K 2 O,Na 2 O) n. d n d n d n d n d Sulphur trioxide (SO 3 ) 1 32 n d 1 40 1 30 1 35 Carbon dioxide (CO 2 ) } A nn f 5 00 2 66 3 52 2 75 Water \ 4.27 I 1 36 n d n d n d. Cementation Index . . "F. O. Norton." Private communication. Quoted by Cummings. " American Cements ", p. 35. Quoted by Lewis. Mineral Industry, vol. 6, p. 96. "Beach's." J. O. Hargrove, analyst. Private communication. "Brooklyn Bridge." Quoted by Cummings. "American Cements", p. 35. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. Newark Lime and Cement Co. Quoted by Cummings. "American Cements", p. 35. Newark and Rosendale. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. "Old Newark." Booth, Garrett, and Blair, analysts. Mineral Industry, vol. 6, p. 96. "Lawrence", Rosendale Cement Co. Mineral Resources U. S. for 1883-1884. Lawrenceville cement. A. W. Dow, analyst. Mineral Industry, vol. 6, p. 96. "Hoffmann'', Lawrence Cement Co. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. " " " " Quoted by Cummings. "American Cements", p. 35. " " " " Haas and McGraw, analysts. Engineering News, April 30, 1896. "Hoffmann", Lawrence Cement Co. Mineral Resources U. S. for 1883-1884. Very ex- ceptional analysis. "Rock Lock." C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. Quoted by Cummings. "American Cements", p. 35. "Hudson River." C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. Rondout, N. Y. L. C. Beck, analyst. "Mineralogy of New York", p. 78. " Hoffmann." "Newark and Rosendale." } Tests of Metals, etc., at Watertown Arsenal, 1901. "Norton." The following analysis is of the natural cement made at Howe's Cave by the Helderberg Portland Cement Company: ANALYSIS OF NATURAL CEMENT, SCHOHARIE COUNTY, N. Y. Silica (SiO 2 ) 26.54 Alumina (A1 2 O 3 ) ) c gg Iron oxide (Fe 2 O 3 ) / Lime (CaO) 45.30 Magnesia (MgO) 17.06 Cementation Index . . 1.17 The cements of central New York are of low index, though usually not so low as those of the Akron-Buffalo district. The natural cements of the Akron-Buffalo district carry usually 20 to 25 per cent magnesia and are of very low index. 258 CEMENTS, LIMES, AND PLASTERS, TABLE 116. ANALYSES OF NATURAL CEMENTS, CENTRAL NEW YORK. 1. 2. 3. 4. Silica (SiO 2 ) 20 30 16 56 35 43 24 10 Alumina (A1 2 O 3 ) I 13.67 Iron Oxide (Fe 2 O 3 ) 10.77 9.92 11.45 Lime (CaO) -'47.48 ^ 39.50 33.67 40.22 Magnesia (MgO) . . 18 55 22 27 20.98 20 60 Cementation Index 1.13 1. Brown Cement Co., Manlius, Onondaga County. W. M. Smith, a/ralyst. 20th Ann. Kept* U. S. Geol. Survey, pt. 6, p. 428. 2. Near Chittenango, Madison County. L. C. Beck, analyst. "Mineralogy of New York", p. 80. 3. South of Utica, Oneida County. Gillmore. "Limes, Cements, and Mortars", p. 125. 4. Average of preceding three analyses. TABLE 117. ANALYSES OF NATURAL CEMENTS, AKRON-BUFFALO DISTRICT, N. Y. 1. 2. 3. 4. 5. Silica (SiO 2 ) 17.14 22 62 20.20 22.70 16.48 Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) 7.61 2.00 7.44 1.40 4.40 2.80 7.40 4.40 2.00 Lime (CaO) 36 83 40 68 41 eo 36 31 39 20 Magnesia (MgO) Alkalies (K 2 O,Na 2 O) Sulphur trioxide (SO 3 ) Carbon dioxide (CO 2 ) 25.09 3.64 n. d. n. d. 22.00 2.23 n. d. 3.63 22.24 1.62 2.06 25.72 n. d. n. d. 4.00 26.52 1.85 1.39 Water n d n d } 6 - 90 { n d > 6.80 Cementation Index 0.801 0.874 0.871 0.991 0.686 6. 7. 8. 9. 10. Silica (SiO 2 ) 26 69 20 75 22 94 20 40 23 70 Alumina (A1 2 O 3 ) 7 21 (6 30 6 22 16 70 Iron oxide (Fe 2 O 3 ) 1 30 J 10.02 \2 90 2 56 3 30 Lime (CaO) 43 12 37 54 43 74 40 64 37 00 Magnesia (MgO) Alkalies (K 2 O,Na 2 O) Sulphur trioxide (SO 3 ) Carbon dioxide (CO 2 ) Water 19.55 1.13 n. d. 1.00 n. d. 26.14 2.12 n. d. } 4.58 20.72 n. d. n. d. 1.00 n. d. 25.80 n. d. 2.91 1.47 n d 15.30 n. d. 1.98 2.00 n d Cementation Index 1.18 0.932 1.006 0.856 2.21 Average index, excluding No. 10 = .911 Average CaO. . , =39.666 Average MgO =22 .908 1.73 1. 2. 3. 4. 5. . 8. 9. 10. Union Akron." Haas and McGraw, analysts. Engineering News, April 30, 1896. Newman Akron." Quoted by Cummings. "American Cements", p. 35. Akron Star." C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. 1897. Buffalo Portland." N. Lord, analyst. Reports Ohio Geological Survey, vol. 6, p. 674. 'Buffalo." C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. 1897. Obelisk." Quoted by Cummings. "American Cements", p. 35. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. 1897. Storm King Portland." Tests of Metals, etc., at Watertown Arsenal, 1901, p. Akron Star." Tests of Metals, etc., at Watertown Arsenal, 1901, p. Obelisk." Tests of Metals, etc., at Watertown Arsenal, 1901, p. COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 259 North Dakota. The Pembina cement is a highly satisfactory prod- uct, of high index and low magnesia, much like the Chickamauga cement of Georgia. TABLE 118. ANALYSES OF NATURAL CEMENT, NORTH DAKOTA. 1. 2. 3. 4. Silica (SiO ) 24 62 23 60 23 90 24 72 Alumina (A1 2 O.>) I 15.12 Iron oxide (I 1 e 2 O 3 ) 16.50 15.90 15.00 Lime (CaO) 52 30 51 40 51 40 51 30 Cementation Index 1 61 5. 6. .7. 8. Silica (SiO,) 24 40 24 40 24 06 24 46 Alumina (A1 2 O 2 ) Iron oxide (Fe 2 O 3 ) I 15.26 15.38 15.00 15.30 Lime (CaO) 52 07 51.96 51.96 52.37 Cementation Index 1 58 1-8. Analyses of natural cement, Pembina Cement Co., Milton, N. D. Pennsylvania. The Lehigh district natural cements are low in magnesia. As marketed they are often very badly mixed products. Portland cement is usually added, while adulteration with coke and ground limestone is not unknown. TABLE 119. ANALYSES OF NATURAL CEMENTS, LEHIGH DISTRICT, PA. 1. 2. 3. Silica (SiO ) 18.18 18.28 30.40 10 36 Iron oxide (Fe O 3 ) 9.78 .43 < 2.60 Lime (CaO) 69.18 51.53 52 12 Magnesia (MgO) 1.98 2.07 21 Alkalies (K 2 O,Na 9 O) n. d. 1.50 n d. Sulphur trioxide (SO 3 ) n. d. n. d. 1.24 Carbon dioxide (CO 2 ) n. d. 1 f 3.07 Water n. d. } 16.26 < n d. 1. Quoted by Smith. Mineral Industry, vol. 1, p. 50. 2. 3. "Bonneville Improved." Tests of Metals at Watertown Arsenal, 1901. West Virginia- Maryland. These cements are fairly low in mag- nesia and usually of very high index. 260 CEMENTS, LIMES, AND PLASTERS. TABLE 120. ANALYSES OF NATURAL CEMENTS, SHEPHERDSTOWN-ANTIETAM DISTRICT, W. VA.-Mo. 1. 2. 3. Silica (SiO 2 ) 33.42 10.04 6.00 * 32.79 9.59 0.50 n. d. } 7.66 2.35 36.51 } 9.36 { 34.83 11.33 1.25 1.49 *. 5.13 2.20 33.50 10.44 3.25 29.38 13.37 n. d. 1.15 7.15 2.23 Alumina (A1 2 O 3 ) Lime (CaO) '. i i Magnesia (MgO) Alkalies (K O Na 2 O) Sulphur trioxide (SO 3 ) Carbon dioxide (CO ) Water Cementation Index 1. Shepherdstown, W. Va. Quoted by Cummings. "American Cements", p 35. 2. Shepherdstown, W. Va. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. 3. Antietam, Md. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. Wisconsin. The Milwaukee cements are of quite low index, rang- ing from 1.10 to 1.20, and carry a little more magnesia than do the Rosendale products. TABLE 121. ANALYSES OF NATURAL CEMENTS, MILWAUKEE DISTRICT, Wis. 1. 2. Silica (SiO) 2 23 16 25 00 Alumina (A1 2 O 3 ) 6 33 4 00 Iron oxide (Fe 2 O 3 ) 1 71 2 80 Lime (CaO) 36 08 33 40 Magnesia (MgO) 20 38 22 60 Alkalies (K 2 O Na 2 O) 5 27 2 51 Sulphur trioxide (SO 3 ) Carbon dioxide (CO 2 ) .... n. d. I 7.07 2.59 Water 9.50 Cementation Index 1.09 1.17 1. Quoted by Cummings. "American Cements", p. 35. 2. C. Richardson, analyst. Brickbuilder, vol. 6, p. 229. Belgium. The manufacture and character of the Belgian natural cements have been described in detail on preceding pages. As marketed they are usually cements of low index (1.05 to 1.15) and carry small percentages of magnesia. England. English natural cements are commonly products of high index, carrying much clayey matter, and often containing remarkably high percentages of iron oxide. France. The analyses given in Table 124 represent a peculiarly homogeneous group of natural cements, a fact which is brought out clearly when their Cementation Indexes are calculated and compared. COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 261 TABLE 122. ANALYSES OF "NATURAL PORTLAND" CEMENTS, BELGIUM. 1. 2. Silica (SiO,) 22 17 28 Alumina (A1 2 O 3 ) 4 60 3 Iron oxide (P'e 2 O 3 ) 1 23 2 Lime (CaO) 60 86 62 Magnesia (MgO) 73 6 Sulphur trioxide (SO 3 ) Carbon dioxide (CO 2 ) n. d. 1 46 1.2 1 2 Water 48 n d Cementation Index . . . 1 09 1 32 1. Compagnie Generate des Ciments Portlands de 1'Escaut, Tournai. 2. Dumon et Cie, Tournai. TABLE 123. ANALYSES OF NATURAL CEMENTS, ENGLAND. l. 2. 3. 4, Silica (SiO 2 ) 24.41 31.36 26.12 25 27 Alumina (A1 2 O 3 ) 9.65 5.01 7.92 7 47 Iron oxide (Fe 2 O.>) 18 86 11 74 6 46 9 05 Lime (CaO) 41 36 46 36 56 44 55 65 Magnesia (MgO) 4 73 3 93 1 60 1 59 Cementation Index 1 92 1 48 1. Sheppey. Quoted by Redgrave. "Calcareous Cements", p. 49. 2. Harwich. " " 3. Whitby. " Zwick, p. 74. 4. London. " Germany-Austria. Analyses of a number of German and Austrian natural cements are given in Table 125. This series includes two inter- esting analyses (Nos. 5 and 6) of cements of very low index, carrying high percentages of magnesia. TABLE 124. ANALYSES OF NATURAL CEMENTS, FRANCE. 1.* 2. 3. 4. Silica (SiO ) 24 94 20 50 21 2 21 82 Alumina (Al Oo) 9 00 8 40 6 9 8 88 Iron oxide (Fe 2 O>) 1 16 5 70 13 7 12 47 Lime (CaO) 63 64 52 05 56.6 55 69 Magnesia (MgO) 1 26 0.95 1.1 1.12 Sulphur trioxide (fc>O 3 ) n. d. 2.80 n. d. n. d. Cementation Indet 1 22 1 32 1 32 1 39 * See next page for references. 262 CEMENTS, LIMES, AND PLASTERS. TABLE 124. Continued. 5. 6. 7. 8. Silica (SiO 2 ) 23.50 24.25 22.10 22.61 Alumina, (Al CX) . . . 12.50 10.00 18.21 19.79 Iron oxide (I 1 6 2 O 3 ) 3.00 4.00 tr. tr. Lime (CaO) - 52.50 . 53.50 55.98 51.63 Magnesia (MsrO) ' 2 . 50 ; * 2 50 0.37 0.37 Sulphur trioxide (SO 3 ) . . 1.00 2.00 3.34* 5.60* Cementation Index . 1.46 1.52 1.45 1.63 * Lime sulphate, CaSO 4 . 1. Vassy. Quoted by Cummings. "American Cements", p. 35. Exceptionally high-limed. 2. " " " Bonnami. "Chaux Hydrauliques", etc., p. 54. 3. " " Zwick. "Hydraulischer Kalk und Portland-Cement", p.^ 91. 5. Valbonnais. Slow-setting. Bonnami. "Chaux Hydrauliques", etc., p. ( 54. 6. ' ' Very slow. 7. Porte de France. Quick-setting. Bonnami. "Chaux Hydrauliques", etc., p.^151. 8. " " " Slow -setting. TABLE 125. ANALYSES or NATURAL CEMENTS, GERMANY AND AUSTRIA. 1. 2. 3. 4. 5. Silica (SiO ) 28.83 27.88 24.12 23.66 20.80 Alumina (A1 2 O 3 ) 6.40 6.19 6.47 7.24 5.80 Iron oxide (Fe 2 O 3 ) 4.80 4.64 5.28 7.97 1.50 Lime (CaO) 58.38 56.45 59.10 58.88 47.83 Magnesia (MffO) 5.00 4.84 4.98 2.25 24.26 0.80 6. 7. 8. 9. 10. Silica (SiO ) 22 14 25 00 21 48 26 80 23 67 Alumina (Al 2 Oo) 5 75 9 00 6 45 10 30 8.83 Iron oxide (Fe 2 O 3 ) 3.07 4 39 2 80 1 90 5.92 Lime (CaO) 44.22 58 02 56 73 59 80 58 80 Magnesia (MgO) 17.77 1.08 3.04 0.30 0.73 Alkalies (K O Na O) 4 72 62 70 1 22 Cementation Index . . . 1.02 1 39 1 455 1 34 1. Rudersdorf. Michaelis, analyst. Quoted in Wagner's "Chemical Technology", 13th ed., p. 669. 2. Rudersdorf. Quoted by Cummings. "American Cements", p. 35. 3. ' Zwick. "Hydraulischer Kalk und Portland-Cement", p. 74. 4 Hausbergen. I Michaelis, analyst. Quoted in Wagner's "Chemical Technology", 13th ed. r 5. Tarnowitz. p. 669. 7! Piesting. > Quoted by Zwick. "Hydraulischer Kalk und Portland -Cement", p 74. 8. Haring. \ 10' Pe U rlmoos' Quoted by Schoch. "Mortel-Materialen", p. 74. COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 263 Weight and specific gravity. The specific gravity of American natural cements appears to be greatly underestimated by most engi- neering authorities. In a recent report,* for example, it is stated that "natural cement has a specific gravity of 2.5 to 2.8". In reality very few of our American cements ever fall as low in specific gravity as 2.8, and it would be nearer the truth to say that, as a class, the natural cements range between 2.8 and 3.2. In the following table (126) a number of careful determination are given, selected from various sources so as to cover as many cement districts and brands as possible. TABLE 126. SPECIFIC GRAVITY OF AMERICAN NATURAL CEMENTS. State. Locality. Brand. Authority. Specific Gravity Illinois Utica ...... ? C. Richardson 2 70 Kansas Fort Scott. . . 9 < i ii 2.79 Maryland Cumberland Cumberland Hydraulic Cement Mfg. Co. Phila. Cement Tests, 1897 2.90 tt Cumberland Hydraulic ' Cement Mfg. Co. ii ( i (i 1899 2.846 tt it Cumberland and Poto- mac Cement Co. it it 1 1 1899 2.828 i t( Potomac Watertown Arsenal, 1901 2.94 1 Round Top. ? C. Richardson 2.84 t i: (l ? Phila. Cement Tests, 1899 2.922 Minnesota Austin Watertown Arsenal, 1901 3.15 Mankato. . ii ( < 1901 2.93 ( e i C. Richardson 2 81 New York Rosendale ? 1 1 ( ( 3.04 t < Hoffmann Watertown Arsenal, 1901 3.06 (C Norton it ft 1901 3.03 11 Newark and Rosendale it tt 1901 3.06 Akron. . . . "Storm King Portland" tt t< 1901 3.07 1 1 Obelisk tt tc 1901 3.12 1 1 Star Phila. Cement Tests, 1897 3.17 Pennsy .vania Lehigh ii Co play improved 1897 1899 3.00 2.96 i < t Hercules improved 1897 2.97 i ( t American improved 1897 3.07 t K (i t ( 1899 3.02 i (i Bonneville improved Watertown Arsenal, 1901 2.85 Rapidity of set. Natural cements are normally much quicker- setting than Portlands, but this rapidity of set may be changed by aeration, the use of plaster, etc., to a very remarkable degree. In 1894-95 Sabin tested the effect of aeration on the setting time of natural cement, with the following results :f * Professional Paper No. 28, Corps of Engineers, U. S. A., p. 11. t Report Chief of Engineers, U. S. A. t 1895, p. 2937. 264 CEMENTS, LIMES, AND PLASTERS. TABLE 127. EFFECT OF AFJRATION ON SETTING TIME OF NATURAL CEMENT. (SABIN.) No. Direct from Package. After 19 Days' Aeration. Initial Set, Minutes. Final Set, Minutes. Initial Set, Minutes. Final Set, Minutes. 1 52 50 44 60 101 87 80 72 108 192 110-^ ioo : 100 280 349 1200 1178 1202 . 1256 1247 54 51 48 100 147 130 ' 122 125 202 234 173 164 166 326 306 1241 1233 1227 1221 1216 2 3 4 .... 5 6 7 8 9 10- .. Effects of gypsum or plaster on natural cements. Natural cements are affected by the addition of gypsum in regard to setting time and strength in much the same manner as Portland cement would be. The degree to which these effects are produced, for the same per- centage of gypsum, depends entirely upon the chemical composition of the respective cements. This fact seems to have been entirely over- looked by experimenters, and in consequence the tests which have been made are deprived of much of their value, because the analysis of the cement is rarely included in the report of the test. Experiments on the effect of gypsum on the rate of set have been carried out by Sabin,* and the results are embodied in Table 128, below, and are shown diagrammatically in Fig. 50. TABLE 128. EFFECT OF PLASTER ON SETTING TIME OF NATURAL CEMENT. (SABIN.) Brand. Per Cent Plaster. Setting Time. Initial, Minutes. Final, Minutes. A 1 2 3 6 1 2 3 6 38 106 107 86 42 93 179 302 295 93 543 414 527 671 632 193 439 592 725 698 B .. n Report Chief of Engineers, U. S. A., 1895, p. 2938. COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 265 From this table it will be seen that the maximum retardation of the initial set took place with both brands when 2 per cent of plaster was used. The final set, however, experienced its greatest retardation in both cases when 3 per cent of plaster was employed. FIG. 50.' 5 10 PERCENTAGE OF PLASTER PARIS -Effect of blaster on setting time. (Sabin.) Sabin also tested t the effects of plaster on the tensile strength of both neat and mortar briquettes. The results of these tests are shown in the following table (129) : TABLE 129. EFFECT OF PLASTER ON TENSILE STRENGTH OF NATURAL CEMENT. (SABIN.) Composition. Tensile Strength, Pounds. Per Cent Cement. Sand. Plaster. 7 Days. 6 Months. 1 Year. 1 146 383 1 1 156 398 x 1 2 115 1 323 1 3 312 2 1 6 234 3 1 2 62 374 448 1 2 1 80 312 395 1 2 2 94 355 408 1 2 3 86 *r 131 3 1 2 6 ... 151 3 107 3 1 Surf ace cracks. 2 Swelled and nearly disintegrated. 8 Badly cracked and swelled. * From Johnson's "Materials of Construction ", p. 187. t Report Chief of Engineers, U. S. A., 1896, p. 2857. 266 CEMENTS, LIMES, AND PLASTERS. These tests would appear to show that the addition of even 1 per cent of plaster has injurious effects on the soundness of the cement, and less markedly on its tensile strength. Unfortunately, the analysis of the cement tested is not given, and even its name is suppressed, so that the results are less instructive than they might have been. 1000 FIG. 51.* Effect of salt on compressive strength. (Tetmajer.) I I I HARDENED IN AIR HARDENED IN WATER 21 28 35 42 49 56 62 70 77 84 FIG. 52.* Effect of salt on tensile strength. (Tetmajer.) In general it may be said that the effects of gypsum or plaster will be directly proportional to the percentage of alumina contained in the cement. This statement has never before been explicitly made, * From Johnson's "Materials of Construction", p. 621. COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 267 but it is a necessary corollary from recent studies on the behavior of cements with gypsum and in sea-water. Effect of salt on strength. In laying masonry in freezing weather it has been customary to specify the use of salt in the water used for the mortar. This lowers the freezing temperature of the water, but does not seem to be of any particular benefit in other respects. It decreases quite markedly the tensile and compressive strength of the mortars, even when only a small percentage of salt is added. SSES * SIEVE FIG. 53. 4 6 8 PERCENTAGE OF SALT IN WATER Effect of salt on tensile strength. 10 12* The effect of salt on the strength of natural-cement mortars is shown in Figs. 51 , 52, and 53. Of these Tetmajer's experiments were made on European natural-cement mortars, mixed 1 cement to 3 sand and tested at various ages. Those shown in Fig. 52 were made on Louis- ville and Portland-cement mortars and all tested at six months. Tensile strength. In tensile strength the average natural cement ranks considerably lower than the average Portland. This is particu- larly noticeable when the cements are tested with sand. This general rule as to the relative strength of natural and Port- land cements is well known, but the exceptions to the rule are not fre- quently discussed. It is a fact, however, that certain brands of natu- ral cements are about as strong, either neat or with sand, as the aver- age imported Portland, and there is no reason why a number of natural cements could not be carried up to this grade. The average results of extensive series of tests on various natural cements are given dia- grammatically in Figs. 54, 55, 56, 57, 58, 59, and 60. In Fig. 55 are shown the results of a very large number of tensile * From Johnson's "Materials of Construction", p. 618. 268 CEMENTS, LIMES, AND PLASTERS. TOO FIG. 54.* Tensile strength of Louisville cement, St. Louis Waterworks, 1896. FIG. 55. Tensile strength of Lehigh district natural cements. (Philadelphia tests, 1893, 1894, 1895, 1896.) * From Johnson's "Materials of Construction", p. 570. COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 269 tests, at various ages up to 3 years, on the "improved" natural cements of the Lehigh district of Pennsylvania. The results of a number of tests of natural cements from the Cumberland-Hancock district of Maryland have been averaged and are shown diagramatically in Fig. 56. 400 300 200 1 100 \ ISeaL I, BL. 11 f 1 w t~ c oo I . I I J i "1 I I aaaa aaa FIG. 56. Tensile strength of Cumberland natural cements. (Philadelphia tests, 1894, 1895, 1896.) Tests of natural cements from Akron, N. Y., and Cumberland, Md., during 1897 and 1S98, are shown diagrammatically in Fig. 57. These tests cover ages of 1 day to 6 months. 270 CEMENTS, LIMES, AND PLASTERS. 500 1 FIG. 57. Tensile strength of Akron and Cumberland cements. (Philadelphia tests, 1897, 1898.) COMPOSITION AND PROPERTIES OF NATURAL CEMENTS. 271 300 200 ioo . 300 200 2 100 400 200 100 7 MILWAUKEE CEMENT 86.4/c PASSED NO. 50 SIEVE U' 78.6 r .. but presume from the geological character of that country it may be found in all the country west to Niagara, and probably farther west." Within a few years this proved to be a fact, cement rock being dis- covered in Krie County, in the extreme western part of the State. The first natural cement manufactured in Krie County was made in ivj-l at \Yilliam-ville. the quarry, kiln, anil mill being near the creek. In 1839 Jonathan Delano erected cement works at Falkirk, near Akron, in which he made about 2000 barrels of cement the first year. He furnished the cement for the feeder dam at Tonawanda Creek and for the (Jenesee Valley Canal. In 1843 the business passed into the hands of James Montgomery, who increased the output to 10,000 barrels a year. The business afterwards came into the possession of Enos Newman, a partner of Montgomery, and has been in his family ever since. In 1854, H. Cummings & Son established a natural-cement plant at Akron, which was operated for several years. This plant was succeeded by another, managed by sons of the founder. The Akron plant was sold to the Akron Cement Company in 1871, and the Cummings brothers erected another plant about two miles west of Akron. The first natural cement made within the present limits of Buffalo was manufactured in 1850 by Warren Granger. His plant was located near Scajaquada Creek, just below the Main Street bridge, in what is now Forest Lawn Cemetery. In 1874 Lewis J. Bennett commenced the manufacture of natural cement at Buffalo Plains, near Main Street. This establishment, which has been carried on continuously under the control of the Bennett family, is now incorporated as the Buffalo Cement Company, In following out the history of the western New York cement in- dustry from 1824 to the present time, we have necessarily passed by the inauguration of manufacture in the greatest of all the natural-cement districts the Rosendale region of eastern New York. Third among 288 CEMENTS, LIMES, AXD PLASTERS. the districts in point of age, it soon became first as a producer, and has ever since maintained a high standard in both the quality and quantity of its output. The discovery of cement rock and the commencement of manufac- ture of natural cement in the Rosendale district took place apparently about 1825, though there is considerable uncertainty as to the exact date to be assigned. * TABLE 141. DATES OF ESTABLISHMENT OF NATURAL-CEMENT INDUSTRIES IN VARIOUS STATES. State. Location. Date. California Benicia I860 Connecticut Kensington . 1826 Georgia Howard 1851 it Rossville. . : 1901 Illinois Utica 1838 Indiana-Kentucky Louisville 1829 Kansas Fort Scott 1868 Maryland Round Top 1837 Cumberland 1836 tt Antietam . 1888 Minnesota Mankato 1883 < < Austin 1895 New Mexico Springer 1899 New York Akron, Erie County ... 183 Williams ville, Erie County. 1824 K it Buffalo, Erie County 1850 tt tt Onondaga-Madison Counties 1818 tt tf Rosendale district 1825 it n Howe's Cave . . 1870 North Dakota Pembina 1895 Ohio Defiance 1846 1 1 Barnesville 1858 Pennsylvania Williamsport 1831 < t Lebanon (?) 1825 (?) it Lehigh district 1850 Virginia Balcony Falls 1848 West Virginia Shepherdstown 1829 Wisconsin Milwaukee 1875 The industry, however, had not secured so firm a foothold in the district by 1837 as might be expected, for in 1843 Mather, of the First Geological Survey of New York State, referred to the immediate past as follows: "When making the reconnaissance, soon after the commence- ment of the geolgical survey, the business had but commenced, and there was no cement manufactured on the Rondout except at Lawrenceville, and there but few kilns were in operation. It was not then known to the inhabitants that the cement rock was abundant except at and near these quarries until some of them were then informed of its inexhausti- STATISTICS OF THE AMERICAN INDUSTRY. 289 ble quantities. Even now few are aware of the great extent of the rock and still fewer understand how to trace out the situation of favorably located new quarries." During the six years that had elapsed since 1837, however, the in- dustry seems to have grown rapidly, for in his final report (1843) Mather states that sixteen firms, working sixty kilns, were then operating in the Rosendale district. He estimated the product at 500,000 to 600,000 barrels per year, and notes that about 700 men were employed in the quarries, in the mills, and in handling the cement. After the industry had become established in New York, it was taken up soon in several other States. It is a noteworthy fact, first pointed out by Mr. Lesley, that all these early plants were located along canals, and that in each case the natural-cement rock was discovered through search for a satisfactory mortar material for canal masonry. Statistics of the American Industry. Since within very wide limits of composition any clayey limestone will give a natural cement on burning, it can readily be seen that sat- isfactory natural-cement materials must be widely distributed and of common occurrence. Hardly a State is entirely without limestones sufficiently clayey to be available for natural-cement manufacture. The sudden rise of the American Portland-cement industry, however, has acted to prevent any great recent expansion of the natural-cement industry. It would be difficult to place a new natural cement on the market in the face of competition from both Portland cement and from the older and well-established brands of natural cement. Such new natural-cement plants as have been started within recent years have mostly been located in old natural-cement districts, where the accu- mulated reputation of the district would help to introduce the new brand. The only exceptions to this rule, indeed, were the Pembina plant in North Dakota, the Rossville plant in Georgia, and a plant in the State of Washington. Of these the Pembina plant was established with the intention of making Portland cement, but the raw materials soon proved to be unsuitable, and the plant was converted. The plant in "Washington is located in an area where any kind of cement is readily salable. The Rossville plant was built by an Akron, N. Y., cement manufacturer, to utilize a peculiarly satisfactory natural-cement rock. The following table taken from the annual volume on Mineral Resources, issued by the U. S. Geological Survey, shows the quantity and value of the natural cement produced in the United States in 1901, 1902, and 1903: 290 CEMENTS, LIMES, AND PLASTERS. 3 CM O CO 1C OS O OS COOS I-H CO O O 00 *O I-H iO CM l> CO CO I> <^ t CO CO CO ir I-H '^t | l> (N CO rH 3,675,520 i CO 1 ^isslsi H ii 1 i 1 I-H d 1 Jj 00 ^'CO CNJ CO l> i t COCO -^ CO fQ lO^CN.CNi i $* CO^ CO | o *-* << L* Number of Works. CN CO O (N -^f CM OT-H(Mt>-CN(Ni-HCOOO co" -H 2 i-HCOOSOOI> >O !o OCNCM CO *O CO 00 iO CO CO . rf CM CO CO 1 1-H "rt rH tn ^ O "ri M 5 p < t OS i i 1 Quantity. iCOcOOOO O 'CO OOCO ^OOCMCOO^O l> iOS COOOCO pq co i>- T i Tfi I-H 10 . i> Tt 1 j-* co" ; 1 1 00* | 1 3 a s 1 11 S -a o 2 ^ fc 1*1 & > CM CO iO CM "^ CM OS i i CM CO r-i d rH CM T-4 1 1 1 1 j g &* H fc I> CO O CM >O O CO O T}H 00 CO CO O CO CO CO CO O *-O 00 00 S i o M 3 CO t^ CM t^* *O CO t > * CM CO CM ^f 00 iO OS 1> cO i i CO t> 00 i-H |> T-H T 1 CO rH 8 o co" n jii 1 -/ 1>* ^ O CO C^l O CO CO CO CM -^jiOOOOiOCOO T-H OCO O CO S3 is ^ s SJ -i^ h O g I c3 ^O CO ^O l^ ^O CM CO O "^t 1 00 PQ ^ T^ T-H CO i"H CM ^^ OS Tt^ i i ifll O 1 3 ^Iplfu 'RODU 1^ f MH 2"S S-S.S c- 1 " 1 ft JZ Q'^ S tiC -j Q) g'?' cS ^;^5 C'-'S ^<~ 1^^ "ffl d o o *.9*1 II "sjj JN i! 1 ^ ; : ; ; "o *g ;'S 1 Illlpli iuHiM ^^H^^HHH 8 STATISTICS OF THE AMERICAN INDUSTRY. 291 "The single cement plant in North Dakota has a production which for 1903 has been combined with that of the only plants producing natural cement in Kansas and Texas. The other States stand in the table exactly as the reported productions are given. "The total results of combined productions are placed against those States which contributed the greater proportion of cement to make the entire quantity." The following figures, also taken from the volume on Mineral Resources, issued annually by the U. S. Geological Survey, are of interest in this connection : TABLE 143. TOTAL PRODUCTION OF NATURAL-ROCK, PORTLAND, AND SLAG CEMENT IN THE UNITED STATES, 1818-1904. Year. Natural, Barrels. Portland, Barrels. Puzzolan or Slag, Barrels. 1818 to 1830 300 000 1830 ' ' 1840 1,000,000 1840 " 1850 4,250,000 1850 " 1860 11,000,000 1860 " 1870 16 420 000 1870 " 1880 22 000 000 82000 1880 2 030 000 42000 1881 2 440 000 60000 1882 3 165 000 85000 1883 . . . 4 190 000 90 000 1884 . . 4 000 000 100 000 1885 4 100 000 150 000 1886 . ... 4,186,152 150 000 1887 6,692,744 250 000 1888 6,253,295 250 000 1889 ' 6,531,876 300,000 1890 7,082,204 335,000 1891 7,451,535 454,813 1892 8,211,181 547,440 1893 7,411,815 590,652 1894 7,563,488 798,757 1895 7,741,077 990,324 1896 7,970,450 1,543,023 12,265 1897 8,311,688 2,677,775 48329 1898 8,418,924 3,692,284 150 895 1899 9,868,179 5,652,266 335,000 1900 8,383,519 8,482,020 446,609 1901 7,084,823 12,711,225 272,689 1902 8,044,305 17,230,644 478,555 1903 7,030,271 22,342,973 525,896 1904 . . 4 866,331 26,505,881 303,015 Total 213,998,857 106,114,077 2,573,283 "The figures for natural-rock and Portland cement in this table through the year 1896 are taken from a statement made by Mr. Uriah CEMENTS, LIMES, AND PLASTERS. Cummings, of Akron, N. Y., in his volume entitled ' American Cements, 1898', on page 288." The remainder of the table is compiled from the United States Geological Survey reports on the production of cement. In making a comparison it must of course be borne in mind that the barrel of Portland cement contains 380 Ibs. net, while the natural- cement barrel varies from 240 Ibs. to 300 Ibs. .-'- . v* Prospects of the Industry. Reference to the tables of statistics on preceding pages will show that the natural-cement production of the United States has been prac- tically stationary since 1890. During this period, while the annual American production of Portland cement has advanced from 335,000 barrels to 22,342,973 barrels, the annual production of natural cement has varied between 7,030,271 and 9,868,179 barrels, the lowest pro- dution for any year being that of 1903. In view of these facts and of freely expressed prophecies that the natural-cement industry is gradu- ally nearing its end, it seems desirable to sum up the prospects of the industry from the viewpoint of a disinterested outsider. Engineers, both in text-books and in conversation, bring two charges against the natural cements as a class. Since the fate of the natural-cement industry will be decided finally by the verdict of the engineer who uses the product, it will pay to critically examine these charges. The faults alleged are: (1) lack of strength as compared with Portland cements, and (2) lack of uniformity in both composition and strength. It may as well be admitted that both of these Charges are true as regards the majority of natural cements as now made, but the writer cannot admit either that these faults are universal, or that they are unavoidable. In regard to the first point an advocate of the natural cements could point out that three brands of natural cements are now regu- larly advertised and sold as Portland cements; that they have been tested for use in both State and Federal public works, including canals, locks, dams, and breakwaters; and that neither State nor army engineers seem to have even suspected that they are not Portland cements. This is surely a proof that all natural cements are not so low in strength as ta be readily distinguished from Portlands by ordinary physical tests. So far as lack of uniformity is concerned, attention might be called to several American brands of natural cements whose variation in composition and strength is no more than is shown by the PROSPECTS OF THE NATURAL-CEMENT INDUSTRY. 293 average brar.d of Portland cement. The faults charged against the natural cements are not, therefore, universal. There remains to be considered the second point: whether, when these faults do occur, they are unavoidable and inherent in the idea of a natural cement, or whether, on the other hand, they can be avoided economically. To the present writer it seems obvious that most brands could, with sufficient care, be freed from both faults, and that this improvement could be carried out without raising the cost of manufacture to a profitless point. In regard to chemical control of the raw material and product, it can be said that, with very few exceptions, none is attempted in the American natural-cement industry. The plants of the Lehigh district of Pennsylvania, which are run in connection with Portland-cement plants, are, of course, better off in this respect than the others. Exclud- ing these Pennsylvania plants, there is, to the writer's knowledge, only one American natural-cement plant which employs a chemist. It seems only right that the name of this honorable exception should be published it is the Pembina Cement Company, of North Dakota. Careful analyses of the raw material and the product will, as pointed out earlier in this volume, enable the manager to select the proper burn- ing-point for his rock, and to correct any defects in its composition. Few quarries show rock of such uniform character that it can all be burned at the same temperature, yet at most plants this is jus what is attempted. The product must- also be given more careful treatment after burning, both in regard to seasoning and grinding. At the present day there is no reason why a coarsely ground or unsound natural cement should be put on the market. Economical fine-grinding machines are obtain- able and should be installed. They will soon repay their first cost, and their high power consumption per unit is more than made up by the great increase in the amount and quality of the output. PART VI. PORTLAND CEMENT. CHAPTER XXII. PORTLAND CEMENT: PRELIMINARY STATEMENTS. IN the chapters of the present section the raw materials, methods of manufacture, and properties of Portland cement will be taken up and discussed in turn. In order that the statements made in these chapters and particularly in those on raw materials may be clearly understood, it seems advisable to preface the section with a brief explanation regarding the definition, composition, and constitution of Portland cement. This brief explanation is accordingly given in the present chapter, while in Chapters XXIX and XXXVIII the subject of composition and constitution will be discussed in the greater detail warranted by their importance. Origin of the name " Portland ". In 1824 Joseph Aspdin took out a patent in England on the manufacture of a cement by calcining a mixture of limestone and clay. To the resulting product he gave the name " Portland", in allusion to a fancied (and in reality very slight) resemblance between the set cement and the famous oolitic limestone so extensively quarried for building purposes at Portland, England, and known to all English architects and engineers as " Portland stone". " Portland" cement obtained its name, therefore, because it looked like Portland stone, and not because Portland was the place of its manu- facture. On the contrary, it is not now, and never has been, manu- factured at either Portland, Me., Portland, Ore., or Portland, England. This statement as to the origin of the name may seem unnecessary; but the writer has found many contractors or other useri of Portland cement who believed that the only proper Portland must naturally be made at Portland a belief which is fostered by the myriad of post- offices and villages bearing that name which have sprung up in the wake of the American cement industry. 294 UNITED SHOWING L( PORTLAND CEM9NT PLANTS FIG. 63. [To face p. 2Q4. PORTLAND CEMENT: PRELIMINARY STATEMENTS. 295 Recurring to Aspdin's work, it is to be noted that his original patent did not specify the percentages in which the two raw materials were to be mixed, and that it also omitted any mention of the high tem- perature necessary to secure a good' product. The earliest Portland was probably, so far as its properties were concerned, like one of our poorer low-limed grades of natural cements. These defects were, how- ever, overcome when Aspdin took up the manufacture on a commercial scale, and before 1850 the new product had established its value. In later chapters further details may be found concerning the early history and growth of the industry, both in Europe and in America. Present use of the term "Portland". While there is at present a fairly close general agreement as to what is to be understood by the term " Portland cement", a few points of importance are still open questions. Partly in consequence of this uncertainty, but more largely because of the intense imitativeness of specification-makers, the defini- tions of the term given in specifications and text-books are usually vague and unsatisfactory. It is commonly agreed that the cement mixture must consist essen- tially of lime, silica, and alumina in proportions which can vary but slightly, and that this mixture must be burned at a temperature which will give a semi-fused product a ' ' clinker ' ' . These points must therefore be included in any satisfactory definition. The principal point regard- ing which there is a difference of opinion is whether or not cements made by burning a natural rock without previous mixing and grinding can under any circumstances be considered true Portlands. The ques- tion as to whether the definition of Portland cement should be drawn so as to include or exclude such products is evidently largely a matter of convention; but, unlike most conventional issues, the decision has very important practical consequences. The question at issue may be stated as follows: If we make artificial mixture of the raw materials and a very high degree of burning the criteria on which to base our definition, we must in consequence of that decision exclude from the class of Portland cements certain well-known products manufactured at several points in France and Belgium by burning a natural rock without previous fine crushing or artificial mixture and at a considerably lower tem- perature than is attained in ordinary Portland-cement practice. These "natural Portlands" of France and Belgium have been considered Port- land cements by some of the most critical authorities, though all agree that they are not particularly high-grade Portlands. So that a definition based upon the criteria above named will of necessity exclude from 296 CEMENTS, LIMES, AND PLASTERS. our class of Portland cements some meritorious products. But, on the other hand, such a restricted definition would have decided advantages. There is no doubt that in theory a rock could occur containing lime, silica, and alumina in such uniformly correct proportions as to always give a good Portland cement on burning. Actually, however, such a perfect cement rock is of extremely rare occurrence. As above stated, certain brands of French arid Belgian* " Portland " cements are made from such natural rocks without the addition of any other material; but these brands are not particularly high grade, and in the better Belgian cements the composition is corrected by th$ addition of other material to the cement rock before burning. The following definition of Portland cement is of importance because of the large amount of cement which will be accepted annually under ;the specifications * in which it occurs. It is also of interest as being the nearest approach to an official definition of the material that we have in this country: "By a Portland cement is meant the product obtained from the heating or calcining up to incipient fusion of intimate mixtures, either natural or artificial, of argillaceous with calcareous substances, the calcined product to contain at least 1.7 times as much of lime, by weight, as of the materials which give the lime its hydraulic properties, and to be finely pulverized after said calcination, and thereafter additions or substitutions for the purpose only of regulating certain properties of technical importance to be allowable to not exceeding 2 per cent of the calcined product." It will be noted that this definition does not require pulverizing or artificial mixing of the materials prior to burning. It seems prob- able that the Belgian "natural Portlands" were kept in mind when these requirements were omitted. In dealing with American-made cements, however, and the specifications in question are headed " Speci- fications for American Portland Cement", it is a serious error to omit these requirements. No true Portland cements are at present manu- factured in America from natural mixtures without pulverizing and artificially mixing the materials prior to burning. Several natural- cement plants, however, have placed on the market so-called Port- land cements made by grinding up together the underburned and over- burned materials formed during the burning of natural cements. Sev- eral of these brands contain from 5 to 15 per cent of magnesia, but even if that fact be disregarded, there is no warrant or excuse for con- * Professional Paper No. 28, Corps of Engineers, U. S. A. f p. 30. PORTLAND CEMENT: PRELIMINARY STATEMENTS. 297 Bidering such products as true Portland cements. Nevertheless, there is absolutely nothing in the definition above quoted that would pre- vent their acceptance ; and as a matter of fact at least one of these brands has been submitted, accepted, and used under these specifications. The definition above discussed is fairly typical of those now to be found in American cement specifications. The definition below has recently (1903-1904) been adopted by the Association of German Portland Cement Manufacturers: Portland cement is a hydraulic cementing material with a specific gravity of not less than 3.10 in the calcined condition, and containing not less than 1.7 parts by weight of lime to each one part of silica -f alu- mina + irgn oxide, the material being prepared by intimately grinding the raw ingredients, calcining them to not less than clinkering tem- perature, and then reducing to proper fineness. In view of the conditions above noted, the writer believes that the following definition will be found more satisfactory than those now in use for insertion as a preliminary requirement in cement specifications. " Definition of Portland cement. By the term Portland cement, as used in these specifications, is to be understood the product obtained by finely pulverizing clinker produced by burning to semi-fusion an intimate artificial mixture of finely ground calcareous and argillaceous materials, this mixture consisting approximately of three parts of lime carbonate (or an equivalent amount of lime oxide) to one part of silica, alumina, and iron oxide. The ratio of lime (CaO) in the finished cement to the silica, alumina, and iron oxide together shall not be less than 1.6 to 1, or more than 2.3 to 1." The ratios of lime to silica, alumina, and iron oxide given in the last sentence of the above definition have been determined by exam- ination of a large series of analyses of standard brands of American Portland cements, and it is very unlikely that any good Portland, as at present made, will have a ratio falling outside the limits above given. Occasionally, however, analyses will come very close to these limits, values of 1.64 and 2.29 respectively having been obtained from good brands. A value as low as 1.6 to 1.7 means a low-testing but abso- lutely safe cement. Values above 2.0 will include the high-testing brands. Composition and constitution. Portland cements may be said to tend toward a composition approximating to pure tricalcic silicate (3CaO,SiO2) which would correspond to the proportion CaO 73.6 per cent, SiO2 26.4 per cent. As can be seen, however, from the analyses quoted in Chapter XXXVIII actual Portland cements as at present made 298 CEMENTS, LIMES, AND PLASTERS. differ in composition very markedly from this. Alumina is always present in considerable quantity, forming with part of the lime the dicalcic aluminate (2CaO, A1 2 O 3 ). This would give, as stated by New- berry, for the general formula of a pure Portland, X(3CaO,SiO 2 ), F(2CaO,Al 2 O 3 ). But the composition is still further complicated by the presence of accidental impurities or intentionally added ingredients. These last may be simply adulterants, or they may be added to serve some useful purpose. Calcium sulphate is a type of the latter class. It serves to retard the set of the cement and in small quantities appears to have no injurious effect which would prohibit its use for this pur- pose. In dome kilns, sufficient sulphur trioxide is generally taken up by the cement from the fuel gases to obviate the necessity for the latter addition of calcium sulphate, but in the rotary kiln its addition to the ground cement, in the form of either powdered crude gypsum or plaster of Paris, is a necessity. Iron oxide, within reasonable limits, seems to act as a substitute for alumina, and the two may be calculated together. Magnesium carbonate is rarely entirely absent from limestones or clays, and mag- nesia is, therefore, almost invariably present in the finished cement but in small percentage. Though magnesia, when magnesium car- bonate is burned at low temperature, is an active hydraulic material (see Chapter XII) it does not normally combine with silica or alumina at the clinkering heat employed in Portland-cement manufacture. At the best it is an inert and valueless constituent in the normal Portland * cement ; many regard it as positively detrimental in even small amounts, and because of this feeling manufacturers prefer to carry it as low as- possible. In amounts of less than 3^ per cent to 5 per cent it is cer- tainly harmless and American Portlands from the Lehigh district usually reach well up toward that limit. In European practice it is carried somewhat lower. Cementation Index. In discussing the hydraulic limes and natural cements, use has been made of the Cementation Index, a device which affords an easy means of comparing the hydraulic and other proper- ties of various cements. In dealing with Portland cement, this device * This statement should not be construed to mean that it is impossible to make a good cement of the Portland type, but containing high percentages of magnesia, for this very possibility will be discussed on a later page (p. 348). But such a magnesia Portland will, of necessity, differ quite markedly both in prep- aration and properties from the lime Portlands now in use. PORTLAND CEMENT: PRELIMINARY STATEMENTS. 299 reaches its maximum of efficiency and becomes of great service in every phase of the subject, from the selection of the raw materials and the proportioning of the mix to the valuation of the finished product. In later chapters the basis and determination of the Cementation Index will be found discussed in detail. In the present chapter it is only necessary to state that its value is obtained from the following formula: (2. 8 X percentage silica) -'- (l.lX percentage alumina) -}- T , .7X percentage iron oxide Cementation Index = =r r . Percentage lime (CaO) +1.4 percentage magnesia This formula is applicable to raw materials as well as to cements, but the user must recollect that the first factor in the divisor is based on the percentage of lime (CaO), not of lime carbonate (CaCOs), and similarly with the magnesia. The Cementation Index, determined as above described, is a meas- ure of the degree of basicity of a cement, or the relation of the acid (SiO2,Al 2 O3,Fe 2 O3) to the basic (CaO,MgO) factors in its composition. A high cementation index means a high-limed and low-clayed cement, while a low index would mean the opposite. In Portland cements as at present made the Cementation Index will commonly fall within the limits of 1.00 and 1.20, 1.00 being the ideal index for a Portland. Silica-alumina ratio. The ratio between the silica and the alumina H-iron oxide gives the second ' important index to the character of a cement. For convenience of reference this may be termed the silica- alumina ratio. This ratio, properly speaking, should take into account the different combining weights of the three compounds concerned, and would, therefore, theoretically be found from the formula . . ,., - j 2. 8 X percentage silica Acidity Index = - ; r : r^ . ( 1 . 1 X percentage alumina) + ( . 7 X percentage iron oxide) To the value determined by this formula the term "Acidity Index 7 ' might be very properly applied. But in ordinary practice the per- centage of iron oxide present is so small that the ratio between the silica and the alumina + iron is given correctly enough by simple division, i.e., Percentage silica Percentage alumina + percentage iron oxide* The value thus obtained will be called briefly the silica-alumina ratio (though it considers the iron oxide also) . It may be said that the per- centage of lime being constant, the clinkering temperature decreases 300 CEMENTS, LIMES, AND PLASTERS. with the silica-alumina ratio; while the setting - time and ultimate strength of the cement are in inverse proportion to the values of the ratio. Kinds of material used. Before taking up the detailed discussion of the various raw materials used in the manufacture of Portland cement, some general statements on the kinds and combinations "of raw mate- rials actually in use will probably be^ found serviceable. In order that the value and availability of different raw materials may be estimated, it will be convenient to assume a certain ideal com- position for a cement rock. For the purposes of tljie present chapter this can be done in a sufficiently accurate way by considering that a Portland-cement mixture, when ready for burning , should contain about 75 per cent of lime carbonate (CaC0 3 ), and about 20 per cent of silica (Si02), alumina (A^Os), and iron oxide (Fe20s) together, the remaining 5 per cent or so containing any magnesia, sulphur, and alkalies that may be present. More exact information on these points will be found in Chapter XXIX , where a somewhat detailed discussion of the calculation and composition of Portland-cement mixtures, together with a number of analyses of actual mixtures and cements, will be given. The essential elements which enter into this mixture lime, silica, alumina, and iron are all abundantly and widely distributed in nature, occurring in different forms in many kinds of rocks; and it can read- ily be seen that, theoretically, a satisfactory Portland-cement mixture could be prepared by combining, in an almost infinite number of ways and proportions, many possible raw materials. Obviously, too, we might expect to find perfect gradations in the degree of artificialness of such a mixture, varying from the one extreme where a natural rock of almost absolutely correct composition was used to the other extreme where two or more materials in nearly equal amounts were required to produce a mixture of correct composition. The almost infinite number of raw materials which are theoretically available are, however, reduced to a very few in practice under existing commercial conditions. The necessity for producing the mixture as cheaply as possible rules out of consideration a large number of mate- rials which would be considered available if chemical composition was the only thing to be taken into account. Some materials otherwise suitable are too scarce and consequently too expensive for such use; some are too difficult to pulverize finely and bring into combination. In consequence comparatively few combinations of raw materials are actually in use. PORTLAND CEMENT: PRELIMINARY STATEMENTS. 301. In certain favored localities deposits of argillaceous (clayey) lime- stones or " cement rock" have been found in which the lime, silica,, alumina, and iron oxide exist in so nearly the proper proportions that only a relatively small amount (say 10 per cent or so) of other material, added before calcination, is required in order to make a mixture of correct composition. Certain blast-furnace slags are also close in com- position to the desired mixture, and are used like " cement rock". In the majority of plants, however, most or all of the necessary lime is furnished by one raw material, while the silica, alumina, and iron oxide are largely or entirely derived from another raw material. The raw material which furnishes the lime is usually a natural limestone either a hard limestone, a chalk, or a marl but occasionally an artificial product is used, such as the chemically precipitated lime carbonate which results as a waste or by-product of alkali manufacture. The silica, alumina, and iron oxide of the mixture are usually derived from clays or shales, more rarely from slates. The various raw materials available for use in Portland-cement manufacture differ in composition, physical characters, and origin. As to composition, they may be almost (a) purely calcareous, (6) a mixture of calcareous and argillaceous elements, or (c) almost purely argilla- ceous; as to physical characters they may be (a) hard and massive, like the hard limestones and slates, (6) soft, like the chalks and shales,, or (c) granular or unconsolidated, like the marls, clays, alkali waste, and granulated slag. As to origin, they may be (a) natural, like lime- stones, marls, slates, clays, etc., or (6) artificial, like alkali waste and furnace slag. TABLE 144. CHARACTER OF PORTLAND-CEMENT MATERIALS. Natural. Artificial. Hard. Soft. Unconsolidated. Unconsolidated. Calcareous (CaCO 3 over 75%) Pure hard limestone Pure soft limestone or pure chalk Pure marl Alkali waste Argillo-calcareous (CaCO 3 40 to 75%) Hard clayey limestone (cement rock) Soft limestone or clayey chalk Clayey marl Blast-furnace slag Argillaceous (CaCOg less than 40%) Slate Shale Clay 302 CEMENTS, LIMES, AND PLASTERS. A glance at the tabulation above will show the relative physical and chemical characters of the different raw materials. It is obvious, if 75 per cent of lime carbonate will make a good cement mixture, that any of the materials in the middle line (i.e. ; the Argillo-calcareous group) could be used as a basis and its composition corrected by adding either a purely calcareous material or a purely argillaceous material, as might be necessary. The cement practice in the Lehigh district is an example of this kind of mixing. But the same result could be ob- tained by mixing any one of the materials on the first line of the table (i.e., the Calcareous group) with any one of the argillaceous mate- rials listed in the bottom line. This is the method followed at most plants outside of the Lehigh district. There is really little to choose between the two kinds of mixtures, for the final result is the main thing. In later pages the few differences that do exist are pointed out and the advan- tages and disadvantages of each type are mentioned. In previous papers the writer has grouped, under six- heads, the various combinations of raw materials at present used in the Ignited States in the manufacture of Portland cement. This grouping is as follows : (1) Argillaceous hard limestone (cement rock) and pure limestone. (2) Pure hard limestone and clay (or shale). (3) Soft (chalky) limestone and clay (or shale). (4) Marl and clay (or shale). (5) Alkali waste and clay. (6) Slag and pure limestone. The relative commercial importance of these different combinations is indicated by the figures tabulated on page 304. Examination of the statistics there given, which have been arranged by the writer from figures given in the various volumes on ' ' Mineral Re- sources of the United States", issued by the U. S. Geological Survey, will develop several facts of interest. In the first place it will be seen that the "cement-rock" type of mixture, important because of its use in the Lehigh district, is slowly decreasing in relative importance, having fallen from almost three fourths of the total product in 1898 to only a little over half the total product in 1903. In absolute number of barrels produced per year, it is of course rapidly increasing, but it is no longer the only type of material to be considered. The use of marl as a cement material is also slowly decreasing in relative importance, having reached its point of maximum output in 1899, when it supplied almost one fifth of all the cement made. The hard limestones, on the other hand, have increased steadily in importance PORTLAND CEMENT: PRELIMINARY STATEMENTS. 303 '100 80 60 40 Cement, \ FIG. 64. Percentage of total output produced from different raw materials. 304 CEMENTS, LIMES, AND PLASTERS. -TABLE 145. PRODUCTION OF PORTLAND CEMENT FROM VARIOUS MATERIALS. Type 1. Type 2.* Type 3. Type 4.f Argillaceous Limestone (Cement Rock) Marl and Clay. Soft Limestone (Chalk) and Clay. Hard Limestone and Clay. Year. and Pure Limestone. Per Per * Per Per Barrels. Cent of Barrels. Cent of Barrels. Cent of Barrels. Cent of Total. Total. Total. Total. 1898 2,682,304 74.9 545,372 15.2 39,000 1.1 315,608 8.8 1899 4,010,132 70.9 1,095,934 19.4 88,200 1.6 458,000 8.1 1900 5,919,629 70.3 1,444,797 17.1 184,400 2: -2 874,715 10.4 1901 8,503,500 66.8 2,001,200 15.8 495,752 3.9 1,710,773 13.5 1902 10,923,922 63.6 2,214,519 12.9 372,413 2.2 3,673,790 21.3 1903 12,493,694 55.7 3,052,946 13.6 457,813 2.4 6,338,520 28.3 * Including also the product from alkali waste and clay t Including also the product from slag and limestone. from 1898, when they produced less than one tenth of the total output, to 1903, when they produced almost three tenths. The soft chalky limestones show little increase and are still unimportant producers, though the writer believes that they offer brilliant possibilities. Valuation of Deposits of Cement Materials. Determining the possible value for Portland-cement manufacture of a deposit of raw material is a complex problem, depending upon a number of distinct factors, all of which must be given due consideration. The more important of these factors are: 1. Chemical composition of the material. 2. Physical character of the material. 3. Amount of material available. 4. Location of the deposit with respect to transportation routes. 5. Location of the deposit with respect to fuel supplies. 6. Location of the deposit with respect to markets. Ignorance of the respective importance of these factors frequently leads to an overestimate of the value of a deposit of raw material. Their effects may be briefly stated as follows: 1. Chemical composition. The raw material must be of correct chemical composition for use as a cement material. This implies that the material, if a limestone, must contain as small a percentage as possi- ble of magnesium carbonate. Under present conditions 5 or 6 per cent is the maximum permissible. Free silica in the form of chert, flint, PORTLAND CEMENT: PRELIMINARY STATEMENTS. 305 or sand must be absent, or present only in small quantities say 1 per cent or less. If the limestone is a clayey limestone, or "cement rock", the proportion between its silica and its alumina and iron should pref- erably fall within the limits Al 2 O 3 + Fe 2 O 3 AljsO3 + Fe 2 O3 A clay or shale should satisfy the above equation, and should be free from sand, gravel, etc. Alkalies, sulphides, and sulphates should, if present, not exceed 3 per cent or so. 2. Physical character. Economy in excavation and crushing re- quires that the raw materials should be as soft and as dry as possible. 3. Amount available. A Portland-cement plant running on dry raw materials, such as a mixture of limestone and shale, will use approxi- mately 20,000 tons of raw material per year per kiln. Of this about 15,000 tons are limestone and 5000 tons shale. Assuming that the limestone weighs 160 Ibs. per cubic foot, which is a fair average weight, each kiln in the plant will require about 190,000 cubic feet of limestone per year. As the shale or clay may be assumed to contain considerable water, a cubic foot will probably contain not over 125 Ibs. of dry mate- rial, so that each kiln will also require about 80,000 cubic feet of shale or clay. A cement-plant is an expensive undertaking, and it would be folly to locate a plant with less than a twenty years' supply of raw material in sight. This would require that, to justify the erection of a cement- plant on any property, For each kiln of the proposed plant, there must be in sight at least 3 ,800 flOO cubic feet of limestone and 1, 600 pOO cubic feet of clay or shale. 4. Location with respect to transportation routes. Portland cement is, for its value, a bulky product, and is therefore much influenced by the subject of transportation routes. To locate a plant on only one railroad, unless the railroad officials are financially connected with the cement-plant, is simply to invite disaster. At least two transportation routes should be available, and it is best of all if one of these be a good water route. 5. Location with respect to fuel supplies. Every barrel (380 Ibs.) of Portland cement marketed implies that at least 200 to 300 Ibs. of coal have been used in the power-plant and the kilns. In other words, each kiln in the plant will, with its corresponding crushing machinery, use up from 6000 to 9000 tons of coal per year. The item of fuel cost is therefore highly important, for in the average plant about 30 to 40 306 CEMENTS, LIMES, AND PLASTERS. per cent of the total cost of the cement will be chargeable to coal sup- plies. 6. Location with respect to markets. In order to secure an estab- lished position in the trade, a new cement-plant should have (a) a local market area, within which it may sell practically on a non-competitive basis, and (6) easy access to a larger though competitive market area. All of these factors should receive due consideration in deciding on the erection of a cement-plant. The summary just given is merely in the nature of a preliminary note, for these points are taken up in more detail in later chapters, particularly in those devoted to the vari- ous raw materials and to the costs of manufacture. The prospecting, examining, and sampling of deposits of limestones, marls, clays, and shales will be taken up in the chapters immediately following. CHAPTER XXIII. LIMESTONES. THE Portland-cement materials which are discussed in this and the following chapters (XXIV, XXV) under the names of pure hard lime- stone, chalk, argillaceous limestone, or " cement rock", and marl, agree in that they are all forms of limestones, though they differ sufficiently in their physical, chemical, and economic characters to be discussed separately and under different names. In order to avoid unnecessary repetition, no general discussion of limestone will be presented here, but reference should be made to Chapter VI, where the origin, varie- ties, composition, and properties of limestones are described in detail. In the present chapter these general facts will be briefly summarized, and certain features common to all the types of limestone used in Port- land-cement manufacture will be noted, after which these different types will be separately discussed. Limestones in General. Varieties and origin. Limestones are rocks composed largely or entirely of lime carbonate, or of lime carbonate with magnesium carbo- nate. Though one or both of these carbonates will necessarily be the principal ingredients in the rock, various impurities may. occur. In addition to the chemical differences which are thus caused between different samples or kinds of limestone, they may also differ in their physical characters, or in their methods of origin, or in both of these points. Limestones are primarily formed by the deposition of lime. carbonate from sea- or lake- water which carries this salt in solution. This deposi- tion may be direct, caused by chemical processes, or it may be effected through the agency of living organisms. Travertine and tufa are chem- ically deposited limestones formed by surface waters. Molluscs are able to abstract lime carbonate from sea-water and utilize it in the* forma- tion of tV>pir shells. On the death ^ the animals.. these shells sink to 307 308 CEMENTS, LIMES, AND PLASTERS. the sea-bottom and thus aid in the formation of calcareous deposits. Microscopic organisms acting in this way are the cause of the forma- tion of chalk, as noted later (p. 318). Vegetable life, acting in a more indirect way, appears to be an important agency in the deposition of mar/ (p. 338).. Ordinary limestones may have originated in any of the ways noted above. After their formation, if subjected to sufficient heat and pressure, normal limestones ,-*nay be converted into crystalline limestones or marbles. All the varieties of limestone above named may vary in composition and degree of purity within wide limits. . Composition of limestones. The term limestone is used, in its most general sense, to include all rocks composed largely or entirely of lime carbonate, or of lime carbonate plus magnesium carbonate.* A limestone of ideal purity will of course consist of 100 per cent of these carbonates; but few limestones attain even approximate purity and many are very impure. As the percentage of impurities increases, the limestone becomes more and more clayey or sandy or shaly, until at last the name limestone is no longer applicable. The exact lower limit of the group it would be difficult to fix, because the change is gradual, but probably all would agree that a rock containing less than 50 per cent of carbonates can hardly be called a limestone, but should rather be termed a cal- careous clay or sandstone or shale, as the case may be. In the present volume, therefore, the lower limit in composition of limestones will be accepted as that above noted i.e., 50 per cent of carbonates. As the average composition of a good Portland -cement mixture is about three fourths lime carbonate and one fourth clayey matter, it is obvious that such a composition could be secured either by mixing a pure limestone and a pure clay in the proportions of about three parts limestone and one part clay, or by starting with a clayey limestone carrying, say, 60 to 85 per cent lime carbonate and adding enough clay or pure limestone to bring this percentage up or down to the required 75 per cent. The " cement rock" of the Lehigh district is an example of a highly argillaceous limestone, usually too low in lime carbonate to be a good Portland-cement material of itself and requiring the addi- * When discussing Portland-cement materials, the term "limestone" may be still further restricted so as to entirely exclude the highly magnesian limestones. At present all the Portland cement made is kept as low in magnesia as possible, because of the fear that this ingredient may do some harm to the cement. As a cement carrying over 4 per cent of magnesia (MgO) would be hard to market, a limestone carrying over 6 to 8 per cent of magnesium carbonate (MgCO 3 ) can hardly be classed as a possible Portland-cement material at present. LIMESTONES.- <- 309 tion of a relatively small percentage of pure limestone. At a few Lehigh district quarries, however, the " cement rock 7 ' is a little too high in carbonate, rather than too low, so that it requires the addition of clay and not of limestone. In the present volume the term " cement rock" will be used to cover clayey limestones low in magnesia and carrying from 50 to 80 per cent or so of lime carbonate, while limestones higher than 80 per cent in carbonate will be called for convenience "pure limestones". Impurities of limestone. Whether a limestone consists entirely of calcium carbonate or carries more or less of magnesium carbonate in addition, it may also contain a greater or lesser amount of distinct impurities. From the point of view of the Portland-cement manu- facturer, the more important of these impurities are silica, alumina, iron, alkalies, and sulphur, all of which have a marked effect on the value of the limestone as a cement material. These impurities will therefore be discussed in the order in which they are named above. The silica in a limestone may occur either in combination with alumina as a clayey impurity or not combined with alumina. As the effect on the value of the limestone would be very different in the two cases, they will be taken up separately. Silica alone. Silica, when present in a limestone containing no alumina, may occur in one of three forms, and the form in which it occurs is of great importance in connection with cement -manufacture. (1) In perhaps its commonest form, silica is present in nodules, masses, or beds of flint or chert. Silica occurring in this form will not readily enter into combination with the lime of a cement mixture, and a cherty or flinty limestone is therefore almost useless in cement-manu- facture. (2) In a few cases, as in the hydraulic limestone of Teil, France, a large amount of silica is present and very little alumina, notwith- standing which the silica readily combines with the lime on burning. It is probable that in such cases the silica is present in the limestone in a very finely divided condition, or possibly as hydrated silica, pos- sibly as the result of chemical precipitation or of organic action. In the majority of cases, however, a highly siliceous limestone will not make a cement on burning unless it contains alumina in addition to the silica. (3) In the crystalline limestones (marbles) and less commonly in uncrystalline limestones, whatever silica is present may occur as a com- plex silicate in the form of shreds of mica, hornblende, or other sili- cate mineral. In this form silica is somewhat intractable in the kiln and mica and other silicate minerals are therefore to be regarded as 310 CEMENTS, LIMES, AND PLASTERS. inert and useless impurities in a cement rock. These silicates will flux at a lower temperature than pure silica and are thus not so trouble- some as flint or chert. They are, however, much less serviceable than if the same amount of silica were present in combination with alumina as a clay. Silica with alumina. Silica and alumina, combined in the form of clay, are common impurities, in limestones, and are of special inter- est to the cement manufacturer. The best-known example of such an argillaceous limestone is the cement rock of the Lehigh district of Pennsylvania. Silica and alumina, when present jn this combined form, combine readily with the lime under the action of heat, and an argillaceous limestone therefore forms an excellent basis for a Port- land-cement mixture. Iron. Iron when present in a limestone occurs commonly as the oxide (Fe20 3 ) or sulphide (FeS2); more rarely as iron carbonate or in complex silicate. Iron in the oxide, carbonate, or silicate forms is a useful flux, aiding in the combination of the lime and silica in the kiln. When present as a sulphide in the form of the mineral pyrite it is to be avoided in quantities over 2 or 3 per cent. Alkalies. Soda and potash occur usually in small percentages and most commonly in the looser-textured limestones. It is probable that these alkalies are largely driven off in the kiln, so that they do no particular harm to the cement. If the total amount of alkalies is above 5 per cent, however, a sufficient amount will be carried over into the cement to cause trouble; and raw materials carrying more than 5 per cent of soda and potash together should therefore be looked upon with suspicion, if not absolutely rejected. Sulphur. Sulphur may occur combined with lime as lime sulphate, or combined with iron as the mineral pyrite. In either case it is an injurious impurity, and the presence of over 1 to li per cent of total sulphur should cause the rejection of the raw material. Physical characters of limestones. In texture, hardness, and com- pactness the limestones vary from the loosely consolidated marls through the chalks to the hard, compact limestones and marbles. Paral- lel with these variations are variations in absorptive properties and density. The chalky limestones may run as low in specific gravity as 1.85, corresponding to a weight of, say, 110 Ibs. per cubic foot, while the compact limestones commonly used for building purposes range in specific gravity between 2.3 and 2.9; corresponding approximately to a range in weight of from 140 to 185 Ibs. per cubic foot. From the point of view of the Portland-cement manufacturer these LIMESTONES. 311 variations in physical properties are of economic interest chiefly in their bearing upon two points: the percentage of water carried by the limestone as quarried, and the ease with which the rock may be crushed and pulverized. To some extent the two properties counter- balance each other, for the softer the limestone the more absorbent it is likely to be. These purely economic features will be discussed in more detail in later chapters. Effect of heating on limestone. On heating a non-magnesian lime- stone to or above 750 F., its carbon dioxide will be driven off, leav- ing quicklime (calcium oxide, CaO). If a magnesian limestone be simi- larly treated, the product would be a mixture of calcium oxide and magnesium oxide (MgO). The rapidity and perfection of this decom- position can be increased by passing steam or air through the burning mass. In practice this is accomplished either by the direct injection of air or steam, or more simply by thoroughly wetting the limestone before putting it into the kiln. FIG. 65. Working thick. limestone-bed. If, however, the limestone contains an appreciable amount of silica, alumina, and iron, the effects of heat will not be of so simple a char- acter. At temperatures of 800 C. and upwards these clayey impuri- ties will combine with the lime oxide, giving silicates, alummates, and 312 CEMENTS, LIMES, AND PLASTERS. related salts of lime. In this manner a natural cement will be pro- duced. An artificial mixture of certain and uniform composition burned at a higher temperature will give a Portland cement the details of whose manufacture are discussed in the present section of this book. Pure Hard Limestones. Under this heading are grouped limestones of normal hardness (ex- cluding the soft chalky limestones and the marls) which carry no less than 80 per cent of lime carbonate and less than 6 per cent of magne- sium carbonate. Limestones carrying less than 80 per cent of lime carbonate are described in the next chapter under the heading of Cement Rock. The boundary between the two classes is of course an arbitrary limit, and 80 per cent of CaCOs has been selected for convenience. As a matter of fact, most of the limestones used in cement-plants are much purer than the lower limit above fixed, ranging usually from 90 to 95 per cent of lime carbonate. Soon after the American Portland-cement industry had become fairly well established in the Lehigh district, attempts were made in New York State to manufacture Portland cement from a mixture of pure limestone and clay. These attempts were not commercially suc- cessful, and although their lack of success was not due to any defects in the limestone used, a certain prejudice arose against the use of the hard limestones. In recent years, however, this has disappeared, and a very large proportion of the' American output is now made from mix- tures of limestone with clay or shale. (See page 304 for comparative figures.) This reestablishment in favor of the hard limestone is doubtless due in great part to recent improvements in grinding machinery, for the purer limestones are usually much harder than argillaceous lime- stones like the Lehigh district "cement rock". Composition of hard limestones actually used. In Table 146 analyses of a large number of limestones used at American cement-plants are given. On examination it will be seen that most of these limestones range from 49 to 54 per cent of lime (CaO) and thus represent quite pure rocks, since a theoretically pure limestone composed entirely of lime carbonate (CaCOs) will contain only 56 per cent of lime (CaO), the remaining 44 per cent being carbon dioxide (C02). With few excep- tions the limestones analyzed carry less than 1 per cent of magnesia (MgO). Their sulphur percentages are also low, which appears to be more commonly the case in dealing with a hard limestone than when a soft limestone or marl is in question. The same may be said in regard ;to alkalies. LIMESTONES. 313 In prospectuses and in the reports of "cement experts", analyses of limestones averaging 98 or 99 per cent of lime carbonate are quite common, but in real life a quarry that will steadily turn out limestone 94 per cent pure is about as good as can be hoped for. With a lime- stone of this degree of purity little attention need be paid to the char- acter of the remaining 6 per cent of impurities. But when a limestone carrying 90 per cent or less of lime carbonate (equivalent to about 50 per cent of lime) is in use or under consideration, the character of the impurities becomes of the first importance. Of course objectionable percentages, of sulphur compounds or mag- nesia would be enough to debar a limestone from use, but even when the impurity consists of clayey matter (silica, alumina, and iron oxide) its exact composition is a matter of importance and should be carefully studied. The matter of interest is the ratio given for the formula Percentage silica (SiO 2 ) Percentage alumina (A1 2 O 3 ) + percentage iron oxide (Fe 2 O 3 )' It is to be noted that the importance of this question increases as the limestone becomes less pure. The reason for this is obvious. Sup- pose we are dealing with two limestones of respective composition: A. B. Lime carbonate 95 . 00 80 . 00 Silica 4.00 16.00 Alumina . 70 2 . 80 Iron oxide .30 1 .20 The ratio . ^ -r- will in each case give a value of 4.0; but Al 2 O 3 + Fe 2 O 3 the result to the cement manufacturer will be very different. If he uses limestone A, its silica-alumina ratio is of little importance, for as the limestone is very pure (95 per cent CaCOa) it will require the addi- tion of considerable clay. The silica-alumina ratio of the mix will there- fore be determined by that of the clay, not by the ratio shown by the limestone; and the manufacturer can select a clay which will give what- ever he considers a desirable ratio for the mix. But if he should use limestone B, it would require but little clay, since it is already very clayey; and it would be almost impossible to find a clay sufficiently aluminous to reduce the 77-^- T?~~FT ratio A1 2 (J3 + -b e 2 (J3 much below the 4.0 which is fixed by the limestone. For this reason it may be taken as a safe rule that when a limestone 314 CEMENTS, LIMES, AND PLASTERS. TABLE 146. ANALYSES OF HARD LIMESTONES USED AT AMERICAN CEMENT-PLANTS. Silica (Si0 2 ). Alumina (A1 2 3 ). Iron oxide (Fe 2 3 ). Lime (CaO). Magnesia (MgO). Sulphur trioxide (S0 3 ). Carbon dioxide (C0 2 ). Water. 1 1.21 0.70 0.50 53 . 62 -T 0.44 0.11 42.98 2 0.98 0.58 0.34 54.17 0.13 0.21 42.96 3 1.02 1.91 53.36 0.39 0.12 43.01 4 1.93 2.37 53.15 tr. n.d. 42.46 5 1.46 2.15 53.62 tr. n. d. 42.85 6 2.12 0.28 0.50 54.06 0.77 n. d. 42.34 7 6.06 3.92 49.46 0.91 0.10 39.06 8 8.20 1.30 49.37 0.85 n. d. 39.72 9 7.54 3.43 45.57 4.36 .... 39.57 10 5.06 ' 2.32 48.29 3.66 .... 41.05 11 13.89 2.61 45.91 1.00 .... 36.82 12 5.43 1.43 52.02 1.11 .... 40.24 13 0.74 0.13 52.49 1.87 .... 43.68 14 0.89 0.38 0.25 54.48 0.36 .... 43.40 15 0.87 0.34 0.13 54.68 0.32 .... 43.44 16 0.86 0.29 55.74 0.51 42.76 0.04 17 1.00 0.9 2.0 51.10 1.4 .... 43.5 18 1.19 0.95 1.28 53.13 1.36 tr." 42.66 19 1.1 1.8 51.7 2.0 0.1 43.3 20 0.83 1.07 52.67 1.67 .... 43.19 21 1.54 2.06 52.85 0.65 .... 42.23 22 4.17 1.37 49.34 2.94 .... 41.94 23 0.95 0.50 53.94 0.91 .... 43.38 24 3.12 1.15 52.06 1.07 .... 42.06 25 0.40 0.44 54.87 0.20 .... 43.34 26 0.54 0.42 54.73 0.19 .... 43.22 27 0.24 0.38 55.46 0.26 .... 43.86 28 1.54 0.39 1.04 53.87 0.52 .... 29 3.12 0.93 52.58 0.80 6^24 42.17 30 2.70 1.64 52.18 1.28 0.17 42.39 31 9.72 4.20 0.48 47.11 0.66 32 6.30 3.35 50.25 0.22 33 7.88 4.01 48.10 0.53 34 3.30 1.30 52.15 1.58 0.30 40.98 35 0.06 0.63 1.03 53.86 .... .... 43.20 36 3.53 1.14 54.45 6^44 38.74 37 4.20 1.61 1.90 50 . 66 0.73 6^23 40.60 38 1.30 0.73 1.17 53.34 0.75 0.03 42.72 39 9.46 2.45 2.73 45.70 0.99 1.36 36.98 40 0.56 1.23 0.29 54.45 0.36 tr. 43.17 41 4.50 0.20 1.77 49.31 0.75 0.06 40.54 2.59 42 4.14 0.21 1.77 50.16 0.42 0.20 39.87 2.03 43 2.31 0.24 1.18 52.04 0.43 0.17 41.72 1.65 44 5.52 2.97 49.66 0.78 45 n. d. n. d. n. d. 54.3 0.7 43.63 1-5. Pacific P. C. Co., Suisun, Calif. C. J. Wheeler, analyst. 6. Southern States P. C. Co., Rockmart, Ca. J. F. Davis, analyst. 7. Chicago P. C. Co., Oglesby, 111. Quoted in manufacturers' circular. 8. Marquette C. Co., Oglesby, 111. 20th Ann. Rep. U. S. G. S., pt. 6, p. 544. 9-12. German-American P. C. Works, La Salle, 111. W. E. Pressing, analyst. 13. Lehigh P. C. Co., Mitchell, Ind. F. W. Clarke, analyst. LIMESTONES. 315 carries less than 90 per cent of lime carbonate it should give a value Si0 2 of between 2.25 and 3.0 for the ratio These are comfort- able limits, and will give the manufacturer considerable latitude in his choice of a clay to mix with it. FIG. 66. Working heavy horizontal bed of limestone. Prospecting and examining limestone deposits. The prospector looking for a deposit of good limestone, or the engineer engaged to report on a deposit already located, should both realize that much trouble can be avoided if they will first familiarize themselves with 14-15. Bedford P. C. Co., Bedford, Ind. A. W. Smith, analyst. 20th Ann. Rep. U. S. GeoL Survey, pt. 6, p. 381. 16. Tola P. C. Co., Tola, Kansas. H. N. Stokes, analyst. Bull. 78, U. S. Geol. Survey, p. 124. 17 -18. Tola P. C. Co., lola. Kansas. 19. Kansas P. C. Co., lola, Kansas. 20 -24. Alpena P. C. Co., Alpena, Mich. 25-26. Atlas P. C. Co., Ilasco, Mo. 27. Atlas P. C. Co., Ilasco, Mo. E. Davidson, analyst. 28. Catskill P. C. Co., Smith's Landing, N. Y. 29-30. Helderberg P. C. Co., Howe's Cave, N. Y. Black, analyst. 31-33. Cayuga P. C. Co., Portland Point, N. Y. J. H. McGuire, analyst. 34. Glens Falls P. C. Co., Glens Falls, N. Y. Mineral Industry, vol. 6, p. 97. 35. Ironton P. C. Co., Ironton, Ohio. C. D. Quick, analyst. 36. Alma P. C. Co., Wellston, Ohio. 21st Ann. Rep, U. S. Geol. Survey, pt. 6, p. 402. 37-39. Diamond P. C. Co., Middle Branch, Ohio. E. Davidson. 40. Wellston P. C. Co., Wellston, Ohio. W. S. Trueblood, analyst. 41-44. Crescent P. C. Co., Wampum, Pa. Robertson Bros., analysts. Report Q. Q., Penna. GeoL Surv., p. 107. 45. Virginia P. C. Co., Craigsville, Va. Cement Industry, p. 235. G16 CEMENTS, LIMES, AND PLASTERS. the work that has been done by geologists in the areas under consider- ation. Most States now have geological surveys, and there are few important limestone deposits that have not been located and examined by these organizations or by the Federal survey. Numerous reports * on these subjects have been issued by State or Federal Geological' Sur- veys, and these reports can usually be obtained free or at a merely nominal price on application' to th^ proper officials. If such a report can be obtained covering the area to be examined it will da away with a lot of preliminary work on the part of the prospector or engineer. t. Preliminary Examination. In commencing work, it is desirable to prepare a rough map of the area. For this purpose high accuracy is not required, and a pocket compass or Brunt on compass, with a Locke level, and a small protractor will be the only instruments required. With these a map can be made and plotted on a scale of 50 or 100, feet to the inch, distances being measured by pacing. The location of any natural outcrop, pits, wells, road or railroad cuts, and streams should be shown on the map, and their relative elevations ascertained as exactly as possible. When the rocks are lying almost horizontally, the loca- tions of the outcrops are of far less importance than their elevations. If there are sufficient good exposures of the rock, in either natural or artificial cuts, samples should be collected from these outcrops. The weathered part of the rock should be rejected, care being taken that the samples represent the fresh, undecomposed rock. When the natural exposures are not satisfactory, it will be necessary to secure samples by trenches, pits, or boring. Most' of the limestones with which the cement manufacturer may have to deal occur in beds or layers which are practically horizontal In the Appalachian and other disturbed districts, however, the beds, may be tilted to a considerable angle with the horizontal, and in rare cases they may even be almost vertical. Usually samples from differ- ent parts of the same bed (within reasonable distances of each other) will be very similar in composition; but, on the contrary, two adjoin- ing beds may differ greatly from each other. In sampling, therefore, it is desirable to collect at least one speci- men from each bed or layer, noting the thickness and position of the bed. Even thin beds should not be neglected, for a 4-inch layer of highly magnesian rock might prove a serious drawback to the eco- nomical working of the quarry if its presence were unsuspected. * See reference lists on pp. 92-94. LIMESTONES. 317 When the beds are horizontal or nearly so a stream gorge or road cut may furnish a good idea of the character of the different beds. In default of such an exposure, it will be necessary to sink test pits to the rock, unless it is exposed conveniently at the surface, and then secure samples from various depths by drilling. Whenever possible the diamond-drill is the most satisfactory exploring device, for it is practically an automatic sampler. When the beds are steeply inclined, a trench cut across at right angles to the bedding will expose a series of beds and enable each to be sampled. If the beds are horizontal or nearly so, and the various samples show little difference in composition, such a preliminary examination as is described above may be all that is required. In case the rock- beds dip at high angles, or if folds or faults are suspected, it will be safest to call in a geologist or mining engineer as associate. If the analyses disagree markedly, it will be advisable to undertake a more detailed examination of the area. Detailed Mapping and Sampling. A much more detailed examina- tion is always desirable before the actual erection of the plant is com- menced. Such an examination will decide the best possible location for the quarry, and should also give data which will aid in keeping a uniform mix. For these purposes a contour map, with 1-, 5-, or 10-foot contours, according to the slope, on a scale of 25 feet to the inch, should be care- fully prepared. The area to be examined should be laid out in 25- or 50-foot squares and their corners marked and numbered to correspond to their locations on the map. At least three good points should be selected as permanent bench-marks, far enough away from the pros- pective quarry-site as not to be disturbed by excavation or blasting, and the locations and elevations of these points should be carefully determined and placed on the map. Sampling should now be taken up carefully. For final work this can be done satisfactorily only with the diamond drill. Drill-holes should be put down at every corner of the 50-foot squares. Each 5 feet of the core should be sampled and analyzed separately, to a depth of at least 50 feet. If the rock dips steeply, or if for any other reason a deep, narrow quarry seems probable, the drilling should be continued to 200 feet. If the cores from adjacent bore-holes give closely similar analyses, closer drilling is not necessary. But if two samples taken at the same depth from two adjoining holes show differences of more than 3 per cent in their lime carbonate, or more than 1J per cent in . - . . .. v .... of Ac tioD ft viB be safe to stone in the quarry, and to rulrur ;>>* of limestone per year. His would cnrnryoMJ to a of about 4} feet, over one acre . per year per kim. Chalk and Other Soft Limestones. Cbafl^ propaly speaking, is a pore carbonate of Erne composed of of tfa* flifilg of minute organisms, ****"** windi those of A or^mmfr if ^pinE JUP cspcojui^r t^roffn ^y h^^t \ , j. DC? f'fi^rTsf^ mid. 90i .DDQGSWQMMQB * discussed in this chapter agree not only in having usually originated in this way y but also in being rather soft and therefore readily and cheaply crushed and pulverized. As Portland-cement materials they are therefore almost ideal. One defect, however, which to a saaH extent counterbalances their obvious advantages is the fact that most of these soft, chalky limestones absorb water quite readily. A chalky limestone which in a dry season will not earn* over 2 per cent of n lire as quarried ma}' in consequence of prolonged wet weather show as high as 15 or 20 per cent of water. This difficulty can of course be avoided if care be taken in quarrying to avoid unnecessary exposure to water and, if necessary, to provide facilities for storing a supply of the raw materials during wet seasons. Origin of chalk. The term chalk is properly applied to a fine-grained and usually very pure limestone, formed largely or entirely of the cal- cLrcous shells of microscopic organisms. These shells are chiefly of the minute Foraminifera, though equally small and smaller calcareous particles of various shapes also occur. Calvin describes * a section of chalk from Iowa as follows: * Reports Iowa Geological Survey, vol. 3, p. 224. 1 > LIMESTONES. 319 "In thin sections under the microscope the unbroken shells of Foram- inifera are very conspicuous. They lie in close proximity to each other, and their inflated chambers, filled with crystals of calcite, some- times occupy more than one third the area of the entire field. It is certain that more than one fourth, and in some instances more than one third, of the volume of the chalk is composed of foraminiferal shells still practically entire. The matrix in which the shells are embedded is made up of a variety of objects, the most numerous and the most conspicuous under proper amplification being the circular or elliptical calcareous discs known as coccoliths. The small rodlike bodies to which the name rhabdoliths has been applied are not very common, although their pressure is easily detected with a moderately high-power objective. Mingled with coccoliths and rhabdoliths are numerous fragments that are evidently the debris resulting from comminution of foraminiferal shells. When the chalk is treated with acid there remains a small amount of insoluble matter consisting of clay, fine grains of quartz sand, minute pebbles not exceeding 5 millimeters in diameter, and a very few internal casts of the chambers of Foraminifera. Nearly all the foraminiferal shells have the chambers filled with calcite; a few have these cavities still empty; but in a small number of cases the chambers were filled with an opaque, insoluble mineral, probably silica deeply stained with iron oxide, that remains as perfect internal casts after the shell has been dissolved in acid. The amount and compo-' sition of the residuum varies with- the purity of the chalk. In some samples it scarcely exceeds 1 per cent, hi others it is equal to 10 per cent." Chalk was probably deposited in deep, quiet water little affected by debris from the land. At present material of exactly similar type is being formed hi the deeper portions of the Xorth Atlantic and other oceanic basins. Distribution of chalk and soft limestones. Both the true chalks and the other soft limestones here considered are of comparatively recent geologic age, occurring only in Cretaceous or Tertiary roc*ks. There is also a certain geographic unity apparent, for both types occur only along the Atlantic and Gulf coasts and in the Western States. For detailed information regarding the distribution of these rocks reference should be made to the papers and reports listed on page 322. In the present place only a summary can be given covering the more impor- tant features of the subject. The true chalks occur only in formations of Cretaceous age in certain Southern and Western States. The principal chalk deposits available 320 CEMENTS, LIMES, AND PLASTERS. for use in Portland-cement manufacture occur in three widely separated areas occupying respectively (a) parts of central Alabama and north- eastern Mississippi, (b) southwestern Arkansas and central Texas, and (c) parts of Iowa, Nebraska, North and South Dakota, Colorado, and other States of the Great Plains region. Though the chalk is in all these areas of approximately the. same age and character, the formations containing it have been given different names i.e., the Selma chalk, in Alabama and Mississippi; the Whitecliffs chalk, in Arkansas; the Austin chalk, in Texas; and the Niobrara chalk, in the Great Plains region. In addition to the true chalks, soft limestones of Tertiary age occur in all the Atlantic and Gulf coast States from Virginia to Mississippi inclusive, as well as in California. These are the materials commonly described as "marls" in the older geological reports, though they are in no way related to the fresh-water marls now so largely used in Port- land-cement manufacture, discussed in Chapter XXV. Physical Properties. When dry, the chalks and soft limestones, are commonly considerably lighter than the hard limestones. As noted on a previous page, the chalky limestones may run as low in specific gravity as 1.85, corresponding to a weight of about 110 Ibs. per cubic foot, while the hard, compact limestones in common use range in specific gravity from 2.3 to 2.9, corresponding approximately to a range in weight of from 140 to 185 Ibs. per cubic foot. The low weight above quoted is,- however, exceptional, and the soft limestones may be expected to range between 125 and 150 Ibs. per cubic foot when dry. They are usually very porous, however, and but brief exposure to water will increase their weight and moisture content re- markably. This, indeed, is their single defect from the point of view of the cement manufacturer, for during a rainy season or with a badly drained quarry he may have to handle a material carrying 15 or 20 per cent of moisture. Otherwise they are admirable cement materials, being soft and easily quafried and ground. Composition of chalks and soft limestones used in cement-plants. In composition the chalks and other soft limestones vary from a rather pure lime carbonate low in both magnesia and clayey matter to an impure clayey limestone of about the composition of the Lehigh district cement rock. Magnesium carbonate is rarely present in quantities of over 2 or 3 per cent, but alkalies, sulphur, and phosphoric acid may occur in sufficient percentages to require careful considerations. LIMESTONES. 321 TABLE 147. ANALYSES OF PURE CHALKS USED IN AMERICAN CEMENT-PLANTS. 1. 2, 3. 4. 5. 6. 7 8. Silica (SiO 2 ) 5.33 4.42 6.09 3.83 4.14 2.15 5.77 2.22 Alumina (A1 2 O 3 ) \3.03 f 2.21 ir.oa 3.52 1.20 1 2.31 ri.si 12.72 }2.72 2.14 JO. 92 \0.18 Iron oxide (Fe 2 O 3 ) . . . Lime (CaO) 50.53 53.36 49.24 52.16 51.00 52.48 50.45 54.08 Magnesia (MgO) 0.55 n. d. n. d. 0.14 tr. n. d. 0.28 0.10 Alkalies (K 2 O,Na 2 O). n. d. n. d. n. d. tr. n. d. n. d. n. d. Sulphur trioxide (SO 3 ) n. d. n. d. n. d. 0.20 0.50 n. d. n. d. Carbon dioxide (CO 2 ) Water 50.30 n.<*. n. d. n. d. n. d. n. d. 1.41,64 J39.99 [n. d. n.d. n. d. 40.00 n. d. }42.50 1. Whitecliffs P. C. Co., Whitecliffs, Ark. 18th Ann. Rep., U. S. Geol. Survey, pt. 5, p. 1174. 2-3. Trans. Amer. Institute Mining Engrs., vol. 21, p. 4. Western P. C. Co., Yankton, S. D. C. B. McVay, analyst. 5. " Mineral Industry, vol. 1, p. 52. 6. " " " " " vol. 6, p. 97. 7. Alamo Cement Works, San Antonio, Texas. 22d Ann. Rep. U. S. Geol. Survey, pt. 3, p. 737. 8. Almendares, P. C. Co., Marinao, Cuba. Engineering Record, vol. 49, p. 36. TABLE 148. ANALYSES OF CLAYEY CHALKS USED IN AMERICAN CEMENT-PLANTS. Silica (SiO 2 ) 9.88 7.64 12.13 13 32 Alumina (A1 2 O 3 ) /4.171 Iron oxide (Fe 2 O 3 ) i 6.20 7.62 1 3.28J 8.74 Lime (CaO) 43 19 45 20 42 04 41 41 Magnesia (MgO) 52 59 44 67 Sulphur trioxide (SO 3 ) n d 1 62 n d 27 Carbon dioxide (CO 2 ) 34 49 36 06 33 51 33 26 Water 5 72 1 36 n d n d Examining chalk deposits.* The chalk deposits of most of the States have been carefully mapped by geological surveys, and much time will be saved by procuring and studying the proper reports. These will give general data on the distribution and character of the chalk formations. In examining chalk deposits it is well to recollect that they are always found hi thick and almost horizontal beds. Stream ravines usually give good natural sections, which will serve to give a preliminary idea of the character of the rock. In securing samples, the earth-auger gives satisfactory results in most chalk deposits, because the material is usually soft enough to be penetrated readily by this tool. The principal impurities to be guarded against are nodules of pyrite and grains of sand, both of which are very common in many American chalk deposits. * See pages 315-318 for further notes on this subject. 322 CEMENTS, LIMES, AND PLASTERS. List of references on chalks and soft limestones. The following papers deal largely with the chalky limestones of the United States. For convenience of reference those which consider chiefly the origin and structure of chalk are marked A ; those which describe its distribution in certain States or areas are marked B. B. Branner, J. C. The cement materials of southwest Arkansas. Trans. Am. Inst. Min. Engrs., vol. 27, p. 42-63. 1898. A. Calvin, S. The Niobrara chalk. Proc. Amer. Assoc. Adv. Sci., vol. 43, pp. 197-217. 1895. A. Calvin, S. Composition and origin of Iowa chalk.- Reports Iowa Geolog- ical Survey, vol. 3, pp. 211-236. 1895. A. Dawson, G. M Note on the occurrence of Foraminifera, etc., in the Creta- ceous rocks of Manitoba. Canadian Naturalist, vol. 7, no. 5. 1874. B. Eckel, E. C. Cement materials and cement industries of the United States. Bulletin 243, U. S. Geological Survey. 1905. B. Hill, R. T. A brief description of the Cretaceous rocks of Texas and their economic value. 1st Ann. Report Texas Geological Survey, pp. 103- 144. 1890. A. B. Hill, R. T. Neozoic geology of southwestern Arkansas. Ann. Rep. Arkansas Gecl. Survey for 1888, vol. 2. A. Hill, R. T. The foraminiferal origin of certain Cretaceous limestones. American Geologist, Sept., 1889. B. Smith, E. A Report on the geology of the Coastal Plain of Alabama. Report Alabama Geological Survey, 759 pp. 1894. B. Smith, E. A. Alabama 's resources for the manufacture of Portland cement. Proc. Ala. Industrial and Scientific Society, vol. 5, pp. 44-51. 1895. B. Smith, E. A. The cement resources of central and southern Alabama. Senate Document No. 19, 58th Congress, 1st session. 1903. B. Smith, E. A. Cement resources of Alabama. Bulletin 225, U. S. Geological Survey, pp. 424-447. 1904. B. Smith, E. A. Cement resources of Alabama. Bulletin 8, Alabama Geological Survey. 12mo, 93 pp. 1904. B. Taff, J. A. Chalk of southwestern Arkansas, with notes on its adaptability to the manufacture of hydraulic cement. 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 687-742. 1902. A. Williston, S. W. Chalk from the Niobrara Cretaceous of Kansas. Science, vol. 16, p. 294. 1890. A. Williston, S. W. On the structure of the Kansas chalk. Trans. Kansas Acad. Sci., vol. 12, p. 100. 1890. CHAPTER XXIV. ARGILLACEOUS LIMESTONE: CEMENT ROCK. THE term "cement rock" is here used to include all the very clayey limestones carrying from 50 to 80 per cent of lime carbonate, with correspondingly high percentages of argillaceous matter, and less than 8 per cent of magnesium carbonate. It is evident that an argillaceous limestone low in magnesia, and containing approximately 75 to 77 per cent of lime carbonate and 20 per cent or so of clayey materials (silica, alumina, and iron oxide) would be the ideal material for use in the manufacture of Portland cement; for a rock of this composition would contain within itself, mixed in the proper proportions, all the ingre- dients necessary for the manufacture of a good Portland. Such an ideal rock would require the addition of no other raw material, but when burnt alone would give a good cement. This ideal cement material is, of course, never realized in practice, but certain deposits of clayey limestone approach it very closely in composition. A limestone carrying 70 or 80 per cent of lime car- bonate and 20 to 30 per cent clayey matter will require the addition either of pure limestone or of clay in order to bring it to the desired composition for a Portland-cement mixture. But it will be, of itself, so near to the correct composition that it will need but little of the extra raw material to make it absolutely perfect. Deposits of such "cement rocks" possess important technologic advantages, and have been sought for with great industry. Many such deposits of clayey limestones, low in magnesia, occur in various parts of the United States, but few of them are well located with regard to transportation routes* fuel supplies, and markets. The most important of these argillaceous limestone, or " cement-rock ", deposits is at present that which is so extensively utilized in Portland- cement manufacture in the "Lehigh district" of Pennsylvania and New Jersey, though similar "cement rocks " occur in many other States. As the Lehigh district still produces over half of all the Portland cement manufactured in the United States, its raw materials will be described below in some detail, after which other areas of "cement rock" will be briefly noted. 323 324 CEMENTS, LIMES, AND PLASTERS. Cement Rock of the Lehigh District, Pennsyivania-New Jersey. The "Lehigh district" of the cement manufacturer has been so greatly extended in recent years that the name is now hardly appli- cable. Originally it included merely an area about 4 miles square, located along the Lehigh River partly in Lehigh County and partly in Northampton County, and. containing the villages of Egypt, Coplay, Northampton, Whitehall, and Siegfried. The cement-plants which were early located here secured control of most of the cement-rock deposits in the vicinity, and plants of later establishment have therefore been forced to locate farther away from the original center of the district. At present the district includes parts of Berks, Lehigh, and Northampton Counties, Pa., and Warren County, N. J., reaching from near Reading, Pa., at the southwest, to a few miles north of Stewartsville, N. J., at the northeast. It forms an oblong area about 25 miles in length from southwest to northeast and about 4 miles in width. Within this area about twenty Portland-cement plants are now in operation, and the Portland cement produced in this relatively small district amounts to over half of the entire United States output. Geology of the district. Within the " Lehigh district " three geo- logic formations occur, all of which must be considered in attempting to account for the distribution of the cement materials used here. These three formations are, in descending order, the (1) Hudson shales, slates, and sandstones; (2) Trenton limestone (Lehigh cement rock); (3) Kitta- tinny limestone (magnesian). As all these rocks dip, in general, north- westward, the Hudson rocks occupy the northwestern portion of the district, while the cement rock and magnesian limestone outcrop in succession farther southeast. Hudson shale. This series includes very thick beds of dark-gray to black shales, with occasional thin beds of sandstone. In certain localities, as near Slatington and Bangor, Pa., and Newton, N. J., these shales have been so altered by pressure as to become slates, the quarry- ing of which now supports a large roofing-slate industry. The composition of the typical shales and slates of the Hudson for- mation is well shown by the following analyses (Table 149). The geographic distribution of the Hudson shales and slates in the Lehigh district can be indicated only approximately without the pres- entation of a geologic map of the area. They cover practically all of Northampton, Lehigh, and Berks counties north of a line passing through Martins Creek, Nazareth, Bath, Whitehall, Ironton, Guthsville, Monterey, Kutztown, Molltown, and Leesport. The rocks of the Lehigh district have a general dip to the northwest, ARGILLACEOUS LIMESTONE: CEMENT ROCK. 325 TABLE 149. ANALYSES OF HUDSON SHALE AND SLATE IN PENNSYLVANIA AND NEW JERSEY. 1. 2. 3. 4. Silica (SiO ) Per Cent. 68 62 Per Cent. 68 00 Per Cent. 56 60 Per Cent. * 76 22 Alumina (Al Oo) 12 68 14 40 21 00 Iron oxide (Fe 2 O 3 ) 4 20 5 40 5 65 | 13.05 Lime (CaO) 1 31 2 68 3 42 2 67 Magnesia (MgO) 1 80 1 51 2 30 93 Alkalies 3 73 11 50 n. d. Carbon dioxide (CO,). Water (H 2 O) 2.99 4 47 2.30 2 70 2.20 3 00 n.d. n d * Insoluble. 1. East Bangor, Pa. 20th Ann. Rep. U. S. Geol. Survey, pt. 6, p. 436. 2. 1 mile northwest of Colemanville, N. J. Geology New Jersey, 1868, p. 136. 3. Delaware Water Gap, N. J. Geology New Jersey, 1868, p. 136. 4. Lafayette, N. J. Kept. New Jersey State Geol. for 1900, p. 74. though there are numerous local exceptions to this rule. The lowest beds of the Hudson series, therefore, are those which outcrop along the southern boundary of the formation, as above outlined. These lowest beds carry much more lime and less silica, alumina, and iron than the higher beds whose analyses are given in table 149.. The lowest beds form a natural transition into the underlying cement rock. Trenton limestone. The Lehigh cement rocks, which are equivalent in age to the Trenton limestone beds of New York, are made up of a series of argillaceous limestones. The formation appears to vary in thickness from 150 feet in New Jersey to 250 feet or even more at Naz- areth and on the Lehigh River. Its upper beds near the contact with the overlying Hudson shales are very shaly or slaty black limestones carrying approximately 50 to 60 per cent of lime carbonate and 40 to 50 per cent of silica, alumina, iron, etc. Lower in the formation the per- centage of lime steadily increases, while that of clayey material decreases correspondingly, until near the base of the formation the rock may carry from 85 to 95 per cent of lime carbonate with only 5 to 15 per cent of impurities. This change in chemical composition is accompanied by a change in the appearance and physical character of the rock, which grad- ually loses its slaty fracture and blackish color as the percentage of lime increases, until near the base of the formation it is often a fairly mas- sively bedded dark-gray limestone. Even so, it can usually be readily dis- tinguished from the magnesian Kittatinny limestone, described below, for the cement rock is always darker than the magnesian limestone and con- tains none of the chert beds which are so common in the magnesian rock. The Lehigh cement rock is never nearly so high in magnesia as is the underlying Kittatinny limestone. It does, however, carry con- 326 CEMENTS, LIMES, AND PLASTERS. siderable magnesia (as compared with other Portland-cement materials) throughout its entire thickness, and few analyses will show less than 4 to 6 per cent of magnesium carbonate. The following series of analyses is fairly representative of the lower, middle, and upper beds of the formation. The specimens from the upper beds, near the Hud- son shales, show considerably less lime and more clayey matter than those from the lower parts -of the formation. TABLE 150. ANALYSES OF TRENTON LIMESTONE (LEHIGH CEMENT ROCK). l. 2. 3. 4. 5. Silica (SiO ) Per Cent. 1 83 Per Cent. 5 03 Per Cent. 8 33 Per Cent. 11 90 Per Cent 11 71 Alumina (A1 2 O 3 ) Iron oxide (1 e 2 O 3 ) .CO 51 2.0G 1 23 4.C3 1 32 4.42 1.70 4.36 1 62 Lime (CaO) 53 64 49 73 45 45 44 18 43 47 Magnesia (MgO) .81 1 02 1 34 1 18 1 82 Carbon dioxide (CO 2 ) 43.03 40.19 37.18 3G.01 33 15 6. 7. 8. 9. 10. Silica (SiO 2 ) Per Cent. 11 11 Per Cent. 17 04 Per Cent. 22 71 Per Cent. 19 53 Per Cent. 24 45 Alumina (A1 2 O 3 ) 4 40 6 90 5 84 6 03 5 68 Iron oxide (Fe 2 O 3 ) 1 91 2 13 2 13 1 70 1 57 Lime (CaO) 42 51 37 53 36.50 35 71 35 00 Magnesia (MgO) 2.89 2.17 1.69 3.33 2 21 Carbon dioxide (CO 2 ) 36.57 32.88 30.52 32.73 29.89 Ann. Kept. New Jersey State Geologist for 1COO, p. 95. The specimens whose analyses are given above were mostly from the vicinity of Belvidere, N. J., and, though representative in other respects, seem to have been rather lower in magnesia than the usual run of the Trenton limestone in the Lehigh district. Kittatinny magnesian limestone. Underlying the cement-rock series is a very thick formation consisting of light-gray to light-blue massive- bedded limestone, with frequent beds of chert. These limestones are predominantly highly magnesian, though occasionally beds of pure non-magnesian limestone will be found in the series. The magnesian beds are, of course, valueless for Portland-cement manufacture, but the pure limestone-beds furnish part of the limestone used in the Lehigh district for addition to the cement rock. An excellent example of this is furnished by the quarry near the east bank of Lehigh River, just above Catasauqua. In this quarry most of the beds are highly mag- nesian, and are therefore useful only for road metal and flux; but a ARGILLACEOUS LIMESTONE: CEMENT ROCK. 327 few pure limestone beds occur, and the material from these low-magnesia beds is shipped to a neighboring cement-mill. Numerous analyses of the highly magnesian limestones are available, from which a few typical results have been selected for insertion here. Analyses of the purer limestone, used to add to the cement rock, will be found in the table on page 329. TABLE 151. ANALYSES OF KITTATINNY MAGNESIAN LIMESTONE.* 1. 2. 3. 4. 5. Silica (SiO 2 ) Per Cent. 9.9 Per Cent. 9 9 Per Cent. 8 8 Per Cent. 5.5 Per Cent. 9.8 Alumina (A1 2 O 3 ) I 1 -7 Iron oxiclc (I (1 e 2 Oo) I ^ 1.7 0.8 1.3 3.7 Lime (CaO) 27 6 28 5 29 4 28 2 26 4 Magnesia (MgO) 17 9 17 3 17 8 20 2 15.1 Carbon dioxide (CO 2 ) 41 9 41 5 42 8 44.3 45.0 6. 7. 8. 9. 10. Silica (SiO 2 ) ^er Cent. 4.9 Per Cent. 2 Per Cent. 8.0 Per Cent. 4.1 Per Cent. 16.9 Alumina (Al Oo) \ a _ Iron oxide (Fe 2 Oo) | 6.5 8.4 5.3 1.6 1.0 Lime (CaO) 27 3 32 4 26 3 30 3 28 3 Magnesia (MgO) 14 6 15 5 17 4 18 3 15 3 Carbon dioxide (CO 2 ) 44 8 42 5 41 1 44 1 38 9 From various reports of the New Jersey Geological Survey. 1. Chandlers Island, Sussex County, N. J. 2. Sparta, Sussex County, N. J. 3. Asbury, Warren County, N. J. 4. Oxford Furnace, Sussex County, N. J. 5, 6. Clinton, Hunterdon County, N. J. 7. Pottersville, Somerset County, N. J. 8, 9. Peapack, N. J. 10. Annandale, N. J. While all of the above analyses are from New Jersey localities, the magnesian limestone of the rest of the Lehigh district would give closely similar results. Throughout most of the Lehigh district the practice is to mix a small amount of pure limestone with a relatively large amount of the "cement rock" or argillaceous limestone, in order to bring the lime carbonate content up to the percentage proper for a Portland-cement mixture. As above noted, all of the "cement rock" is derived from the middle part of the Trenton formation, where the beds will run from 60 to 70 per cent of lime carbonate. The pure limestone which is re- quired to bring this material up to the necessary percentage of lime carbonate (75 per cent or so) is obtained either from the lower portion of the Trenton itself or from certain low-magnesia beds occurring in the Kittatinny formation. 328 CEMENTS, LIMES, AND PLASTERS. In the plants located near Bath and Nazareth, however, the practice has been slightly different. In this particular area the cement-rock quarries usually show rock carrying from 75 to 80 per cent of lime car- bonate. The mills in this vicinity, therefore, require practically no pure limestone, as the quarry rock itself is sufficiently high in lime carbonate for the purpose. Indeed, it is at times necessary for these plants to add clay or slate, instead of limestone, to their cement rock, in order to reduce its content of lime carbonate to the required figure. In general, however, it may be said that Lehigh practice is to mix a low-carbonate cement rock with a relatively small amount of pure lime- stone, and analyses of both these materials, as used at various plants in the district, are given below in Tables 152 and 153. Character and composition of the cement rock. The cement rock is a dark-gray to black, slaty limestone, breaking with an even fracture into flat pieces, which usually have smooth, glistening surfaces. As the percentage of lime carbonate in the rock increases i.e., as the lower beds of the formation are reached the color becomes a some- what lighter gray and the surfaces of the fragments lose their slaty appearance. The range in composition of the cement rock as used at various plants is well shown in the first eight columns of the above table. The nearer the material from any given quarry or part of a quarry approaches the proper Portland-cement composition (say 75 to 77 per cent lime carbonate) the less addition of pure limestone will be necessary. In by far the greater part of the district, as above noted, the cement rock is apt to run about 65 to 70 per cent of lime carbonate, therefore re- quiring the addition of a proportionate amount of limestone. Most of the quarries near Bath and Nazareth, however, have been opened on beds of cement rock running considerably higher in lime carbonate and occasionally running so high (80 per cent, etc.) as to require the addition of shale or clay rather than of pure limestone. Character and composition of the pure limestones. The pure lime- stones added to the cement rock are commonly gray and break into rather cubical fragments. The fracture surfaces show a finely granu- lar structure quite distinct in appearance from the slaty cement rock. In composition the limestones commonly used will carry from 90 to 96 per cent of lime carbonate, with rather less magnesium carbonate than is found in the cement rock. All of the cement-plants own and operate their own cement-rock quarries, but most of them are com- pelled to buy the pure limestone. When this is the case only very pure grades of limestone are purchased, but when a cement-plant owns ARGILLACEOUS LIMESTONE: CEMENT ROCK. 329 TABLE 152. ANALYSES OF HIGHLY CLAYEY LIMESTONES: 'CEMENT ROCK". ! . 3 I? |g 1 1| g% g + 3'* .8 J 13 fl I -SB cc q 33 1 3 & 3 S~ o 3 d 1 18.30 6.11 1.85 36.38 2.13 28.96 2. 29 1.51 2 15.97 7.53 2.24 34.34 3.93 32.80 3 17.32 9.11 38.59 2.05 32.55 4 19.62 5.68 39.08 2.35 33.25 5 16.77 6.50 41.37 n. d. n. d. 6 15.73 7.92 39.62 1.81 33.08 1.23 7 19.06 4.44 1.14 38.77 2.02 32.66 8 22.22 7.24 0.92 35.53 2.19 30.29 2. 72 9 19.08 7.92 37.56 1.95 31.62 10 14.20 6.14 41.51 1.56 34.47 11 14.52 6.52 41.17 2.25 34.79 12 15.05 9.02 1.27 39.26 1.90 32.90 1. 46 13 15.20 8. 80 38.70 1. 47 31.99 TABLE 153. ANALYSES OF PURE LIMESTONES USED FOR MIXING WITH CEMENT ROCK. Silica (Si0 2 ). Alumina (A1 2 3 ). Iron Oxide (Fe 2 3 ). Lime (CaO). Magnesia (MgO). Carbon Dioxide (C0 2 ). 3.64 52.93 n. d. n. d. 0. 61 5.56 2 40 50.47 1.00 40.73 3.02 1. 90 51.55 1.46 42.08 1.98 0. 70 53.31 0.97 42.94 2.14 1. 46 52.84 1.05 42.64 4.50 0. 82 51.50 0.66 41.19 3.40 0. 70 52.53 0.61 41.94 1.02 0. 48 54.60 0.53 43.47 0.08 0. 40 54.90 0.61 43.80 6.1 3. 5 47.21 2.35 39.64 its limestone quarry material running as low as 85 per cent of lime carbonate is often used. Quarry practice. In most of the cement-rock quarries of the Lehigh district the rock dips from 15 to 25, usually to the northwest. At a few quarries, particularly in New Jersey, the dip is much steeper. The quarries are opened preferably on a side-hill, and the overlying stripping, which consists of soil and weathered rock, is removed by scrapers or shoveling. The quarry of the Lawrence Cement Company has been extended in its lower levels so as to give a tunnel through which the material is hoisted to the mill. Several other quarries have 330 CEMENTS, LIMES, AND PLASTERS. been carried straight down, until now they are narrow and deep pits, from which the material 'is hoisted vertically. The Bonneville Port- land Cement Company's quarry is an extreme example of this type. FIG. 67. Tunnel, Lawrence Cement Co., Siegfried, Pa. In quarries opened on a side-hill, so as to have a long and rather low working-face and a floor at the natural ground level, the rock is commonly blasted down in benches, sledged to convenient size for handling and crushing, and carried by horse carts to a point in the quarry some distance from the face where the material can be dumped into cars, which are hauled by cable to the mill. Occasionally the material is loaded at the face into small cars running on temporary tracks. The loaded cars are then drawn by horses or pushed by men to a turntable, where they are connected to the cable and hauled to the mill. While these methods seem clumsy at first sight, they are capable of little improvement. The amount of rock used every day in a large ARGILLACEOUS LIMESTONE: CEMENT ROCK 331 mill necessitates very heavy blasting, and this prevents permanent tracks and cableways from being laid near to the working-face. At several quarries the loading into the cars or carts is accomplished by means of steam-shovels. The cement rock seems to be well FIG. 68. Open cut in cement rock. adapted for handling by steam-shovels, but even then much sledging is necessary and the blasting operations are interfered with. Cement production of the district. The importance of these Lehigh district cement-rock deposits is well brought out by Table 154. Probable extension of the industry. As noted in the earlier portion of this chapter, the cement deposits have been developed only from near Reading, Pa., to a few miles west of Stewartsville, N. J. Most of the readily accessible cement land between these points has been taken up by the cement companies or is being held at impossible prices by the owners. Under these circumstances it seems probable that few additional plants can be profitably established in the district now developed, and that the growth of the industry here will be brought about by extending the district. A few notes on the distribution of the same cement-beds in adjoining areas may therefore be of interest to those desiring to engage in the manufacture of Portland cement from materials of the Lehigh district type. 332 CEMENTS, LIMES, AND PLASTERS. TABLE 154. PORTLAND-CEMENT PRODUCTION OF THE LEHIGH DISTRICT, 1890-1902. Year. Lehigh District. Entire United States. Percentage of Total Product Manufactured in Lehigh District. Number of Plants. Number of Barrels. Number of Plants. Number of Barrels. Value. 1890.. 1891 5 5 5 5 7 8 8 8 9 11 15 16 17 19 201,000 248,500 280,840 265,317 485,329 634,276 1,048,154 2,002,059 2,674,304 4,110,132 6,153,629 8,595,340 10,829,922 12,324,922 - 16 17 16 19 24 22 26 29 31 36 50 56 65 75 ^35,500 '. 454,813 547,440 590,652 798,757 990,324 1,543,023 2,677,775 3,692,284 5,652,266 8,482,020 12,711,225 17,230,644 22,342,973 $439,050 1,067,429 1,152,600 1,158,138 1,383,473 1,586,830 2,424,011 4,315,891 5,970,773 8,074,371 9,280,525 12,532,360 20,864,078 27,713,319 60.0 54.7 51.3 44.9 60.8 64.0 68.1 74.8 72.4 72.7 72.6 67.7 62.8 55.2 1892 1893 1894 1895 1896 1897 . . 1898 . . 1899 1900 1901 1902 1903 Northeast of Stewartsville, N. J., the cement-beds outcrop at fre- quent intervals in the Kittatinny Valley all the way across New Jersey and a few miles into Orange County, N. Y. The exact locations of these deposits, with numerous analyses of the cement rocks, are given in the Annual Report of the State Geologist of New Jersey for 1900, pages 41-95. Many detailed maps in this report show the outcrops very precisely. Southwestward from Reading the Trenton beds outcrop in a belt crossing Lebanon, Cumberland, and Franklin counties, Pa., passing near the tows of Lebanon, Harrisburg, Carlisle, and Chambersburg. In Maryland the Trenton rocks occur in Washington County, while in West Virginia and Virginia they are extensively developed. The dis- tribution of these rocks in Virginia is discussed in the papers by Messrs. Bassler and Catlett, cited in the list on page 333. Throughout this southern extension of the Lehigh rocks, the Tren- ton is not everywhere an argillaceous limestone, but it is frequently so, and it is always very low in magnesium carbonate. It is therefore probably safe to say that in southern Pennsylvania, Maryland, West Virginia, and Virginia the Trenton rocks are everywhere good Port- land-cenlent materials, though in some cases they will require pure limestone, and in other places clay, to bring them to proper composition. Cement Rocks in Other States. Limestones sufficiently clayey to be called " cement rocks" are not by any means confined to the Lehigh district, nor even to the imme- ARGILLACEOUS LIMESTONE: CEMENT ROCK. 333 diate vicinity of that fortunate area. As noted in the last chapter, cement rock exactly similar to that used in the Lehigh district occurs in other parts of Pennsylvania, in Maryland, and the Virginias. Similar clayey limestones occur southward, along the Appalachian Valley, through Tennessee and northern Georgia. In all this range, however, they have never been used as Portland-cement materials, though a natural-cement plant was erected a few years ago at Rossville, Ga., a few miles south of Chattanooga, to utilize limestones closely similar to the Lehigh rock in composition. The following analyses show the composition of " cement rocks" used at various Portland-cement plants in the Western States, together with that of the purer limestones used for mixing. TABLE 155. ANALYSES OF "CEMENT-ROCK" MATERIALS FROM THE WESTERN UNITED STATES. Utah. California. Colorado. Cement Rock. Lime- stone. Cement Rock. Lime- stone. Cement Rock. Lime- stone. Silica (SiO 2 ) 21.2 8.0 6.8 3.0 20.06 10.07 3.39 63.40 1.54 7.12 2.36 1.16 87.70 0.84 14.20 5.21 1.73 75.10 1.10 88.0 Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) Lime carbonate (CaCO 3 ) . . . 62.08 3.8 89.8 0.76 Magnesium carbonate (MgCO 3 ). In addition to the "cement rocks" noted in this chapter, it is neces- sary to call atention to the fact that many of the chalky limestones discussed in the following chapter are sufficiently argillaceous to be classed as " cement rocks". Because of their softness, however, all these chalky limestones will be described together. List of references on " cement rock ". Bassler, P. S. Cement materials of the Valley of Virginia. Bulletin 260, IT. S. Geol. Survey, pp. 531-544. 1905. Catlett, C. Cement resources of the Valley of Virginia. Bulletin 225, U. S. Geol. Survey, pp. 457-461. 1904. Eckel, E. C. Cement-rock deposits of the Lehigh district of Pennsylvania and New Jersey. Bulletin 225, U. S. Geol. Survey, pp. 448-456. 1904. Eckel, E. C. Cement materials and cement industries of the United States. Bulletin 243, U. S. Geol. Survey, pp. . 1905. Kiimmel, H. B. Report on the Portland-cement industry in New Jersey. Ann. Rep. N. J. State Geologist for 1900, pp. 9-101. Peck, F. B. The cement-belt of Lehigh and Northampton counties, Pennsyl- vania. Mines and Minerals, vol. 25, pp. 53-57. 1904. CHAPTER XXV. FRESH-WATER MARLS. It MARLS, in the sense in which the term is used in the Portland-cement industry, are fine-grained, friable limestones which have been deposited in the basins of existing or extinct lakes. So far as chemical composi- tion is concerned, marls are practically pure limestones, being usually composed almost entirely of calcium carbonate. Physically, however, they differ greatly from the hard, compact rocks to which the term limestone is more commonly applied, for the marls are granular, loose, non-coherent deposits. These curious physical characters of marls, as compared with ordinary limestones, are due to the peculiar condi- tions under which the former have been deposited. Samples of marl from different localities will on comparison be found to exhibit consider- able variations, and these arise in large part from differences in local conditions during deposition. As explained on a later page, differences of opinion exist as to the exact cause of the formation of marl deposits. The points in con- troversy are of no particular practical importance, and may be dis- regarded in the present brief statement of facts. It may safely be said that marls are deposited in lakes by spring or stream waters carrying lime carbonate in solution. The actual deposition of the marl is in part due to purely physical and chemical causes, and in part to the direct or indirect action of animal or vegetable life. The result in any case is that a calcareous deposit forms along the sides and over the bottom of the lake, this deposit consisting of lime carbonate, mostly in a finely granular form, interspersed with shells and shell fragments. Various uses of the term "marl". A warning to the reader con- cerning other uses of the term "marl" may profitably be introduced here. The meaning above given is that in which the term marl is com- monly used in the cement industry at the present day. But in geological and agricultural reports, particularly in those issued before the Portland- cement industry became prominent in this country, the term marl has been used to cover several very different substances. The following 334 FRESH-WATER MARLS. 335 three uses of the term will be found particularly common, and must be guarded against when such reports are being examined in search for descriptions of deposits of cement materials. 1. In early days the terms " marls" and " marly tes" were used to describe deposits of calcareous shales and often these terms were extended to cover shales which were not particularly calcareous. This use of the term will be found in many of the earlier geological reports issued by New York, Ohio, and other interior States. 2. In New Jersey and the States southward bordering on the Atlantic and Gulf of Mexico the term marl is commonly applied to deposits of soft chalky or unconsolidated limestone often containing consider- able clayey and phosphatic matter. These limestones are of marine origin and not related to the fresh-water marl deposits which are the subject of the present chapter. 3. In the same States, but particularly in New Jersey and Virginia, large deposits of the so-called " green-sand marls" occur. This material is in no way related to the true marls (which are essentially lime carbonates), but consists largely of the iron silicate called glauconite, or green sand, with very small percentages of clayey, calcareous, and phosphatic matter. The three early uses of the term "marl" above noted all agree in that they apply to deposits of marine origin, while the marls of the cement manufacturer are purely fresh-water deposits. Occurrence of marl deposits. Fresh-water marls occur in more or less lenticular or basin-shaped deposits of relatively small size. Their form and local character are both due to the fact that the marls were formed by deposition in lake basins. In many cases these lakes still contain water, and in some instances marl deposition is now in progress. In other cases, however, the lake has entirely disappeared, and the marl- bed now occurs in a swamp or marsh covered with peat or muck. The disappearance of a lake in this fashion must be regarded as part of a very natural and almost invariable cycle of events. The existence of a lake at any point along a drainage system is to be considered a somewhat unnatural and temporary condition, and one which will be removed by natural causes as soon as possible. In the glaciated portion of the United States many lakes were formed at the close of the Glacial period. These lakes were due in some cases to the fact that deposits of sand and clay laid down by the glaciers had filled old valleys and dammed the streams occupying such valleys. In other cases the lake basins were formed by the irregular distribution of these glacial deposits, leaving hollows and depressions which subsequently became illed with water. In either case a lake or pond was formed, and imme- 336 CEMENTS, LIMES, AND PLASTERS. diately a series of natural forces were set in operation which tended to remove this lake. Of the two common methods of lake disappearance, one has but little to do with our present subject, but the other is closely connected with the history of marl deposition. The first method is the gradual deepening of the outlet by the action of its own current, resulting in the draining of. .the lake. The second is the filling up of the lake basin by deposits of sand, cla$, marl, muck, and .peat. Origin of marl deposits. The exact cause of the formation of marl deposits has been the subject of much investigation and discussion, particularly in the past few years, since these deposits have become of so great economic, importance. For details concerning this dis- cussion, reference should be made to the papers listed on pages 34G-347, especially to those by Davis and Blatchley. In the present place only a summary of the main facts and theories in regard to the forma- tion of marl deposits will be given. Marl deposits in their present form and position are due directly or indirectly to glacial action, on which account these deposits occur almost exclusively in that portion of the country which was covered by ice during the Glacial epoch. The glaciers aided in the formation of marl deposits in two ways: first, by furnishing a large supply of finely ground calcareous material from which surface waters could take up lime carbonate in solution, and, second, by forming the lake basins in which these waters deposited their burden of lime carbonate as marl. The processes followed in the formation of marl deposits may be outlined with some confidence, though certain of the steps are still subject to discussion. Rain-water, though theoretically pure before approaching the earth's surface, takes up a considerable percentage of carbon dioxide in pass- ing through the atmosphere. When it reaches the surface, therefore, such rain-water is in reality a very dilute form of carbonic acid (H^COs), and as such is capable of attacking limestone, taking the lime car- bonate (CaCO 3 ) into solution in the form of calcium bicarbonate (CaH 2 (CO 3 )2). Practically all natural water, including the percolat- ing ground-water as well as spring and stream water, is therefore able to change itself with lime carbonate if some source of that compound presents itself. In the present marl districts such a source is not far to seek, for from New York to Michigan limestones cover a large pro- portion of the surface. The glaciers during their advance over this area ground up vast quantities of the surface rocks, leaving the pul- verized debris in the form of deposits of limey gravels and limey clays. Surface waters running through such areas, or underground waters FRESH-WATER MARLS. 337 percolating through beds of limestone or coarse limey clays, will, if charged with carbon dioxide, dissolve and carry off lime carbonate, the exact amount so dissolved being determined by the percentage of CO 2 contained in the water, its temperature, etc. The tendency is for every water to charge itself with its maximum possible amount of calcium bicarbonate, and until it is so charged it will continue to attack and dissolve limestones which it encounters in its course. When the water is almost or quite saturated with calcium bicar- bonate, any increase in temperature or decrease in pressure will cause the deposition of lime carbonate. Reactions of this type have been appealed to as explanations of the formation of marl deposits through the warming or loss of pressure which occurs when spring^ or stream water enters a lake. A recent statement * of this theory is as follows: "This spring-water as it enters the lake is always colder than the waters of the lake itself. The bicarbonate of lime is more soluble in cold water than in warm and a part of the dissolved material is there- fore precipitated in the form of a fine powder soon after the cold stream enters the warmer, still water of the lake. Such precipitation of cal- cium carbonate from cold water as it becomes warm is seen every day in almost every household. The hard water heated in tea-kettles holds while cold a large quantity of bicarbonate of lime in solution. As it becomes warm, much, if not all, of this falls and forms a coating of lime carbonate upon the bottom of the kettle. " Again, if there is a large amount of carbon dioxide in the perco- lating waters the percentage of carbonate of lime held in solution will be increased in proportion. As the spring-water enters the lake and rises to the surface the pressure will be decreased and a part of the carbon dioxide will escape and so cause a precipitation of another part of the carbonate of lime according to the following formula: "CaH 2 (CO 3 ) 2 -CO 2 = CaCO 3 +H 2 O." In support of his belief that the formation of most marl deposits is due to the two causes above outlined, Blatchley urges that most, if not all, of the marl lakes examined in Indiana are fed by subterranean or subaqueous springs, even though they also have streams entering and leaving them, and that "the larger deposits of marl in the lakes are found in close proximity to these springs ". Davis, in studying the Michigan marls, came to the conclusion that the causes above noted would not of themselves account for the * Blatchley, W. S. 25th Ann. Rep. Indiana Dept. Geology, p. 45. 338 CEMENTS, LIMES, AND PLASTERS. majority of marl deposits. His studies * led him to believe that more important effects were due to the action of certain aquatic plants of low type, notably Char a (sonewort) and Zonotrichia, another alga. Plants of higher type are also influential in marl deposition, but to a less degree. Davis has summarized f his views as follows : -'All green plants, whether aquatic or terrestrial, take in the gas (carbon dioxide) through their leaves^ and stems and build the carbon atoms and part of the oxygen atoms of which the gas is composed into the new compounds of their own tissues, in the process releasing the remainder of the oxygen atoms. Admitting these^facts, we have two possible general causes for the formation of the incrustation (of cal- cium carbonate) upon all aquatic plants. If the calcium and other salts are in excess in the water, and are held in solution by free carbon dioxide, then the more or less complete abstraction of that gas from the water in direct contact with plants causes precipitation of the (lime) salts upon the parts abstracting the gas, namely, the stems and leaves. But in water containing the salts, especially calcium bicarbonate, in amounts so small that they would not be precipitated if there were no free carbon dioxide present in the water at all, the precipitation may be considered a purely chemical problem, a solution of which may be looked for in the action, upon the bicarbonates, of the oxygen set free by the plants. Of these bicarbonates, calcium bicarbonate is the most abundant, and the reaction upon it may be taken as typical and expressed by the following chemical equation: CaH 2 (CO 3 ) 2 + = CaCO 3 + C0 2 4-H 2 O+ O, Calcium bicarbonate + oxygen = calcium carbonate + carbon dioxide + water + oxygen in which the calcium bicarbonate is converted into the normal (and very slightly soluble) carbonate by the oxygen liberated by the plants, and both carbon dioxide and oxygen are set free, the free oxygen possibly acting still further to precipitate calcium monocarbonate. It is probable that the plants actually do precipitate calcium carbonate in both these ways (i.e., by abstracting carbon dioxide from the water and by freeing oxygen), but in water containing relatively small amounts of calcium bicarbonate the latter would seem to be the probable method." Professor Davis has further proven that Cham acts in still a third way, abstracting lime salts directly from the water as part of its life processes and depositing them in its tissues. * See papers by Davis, cited in list on pp. 346-347. f Vol. 8, pt. 3, Reports Michigan Geol. Survey, p. 69. FRESH-WATER MARLS. 339 A further possible mode of derivation, which is admittedly the way in which part of all the marl deposits have originated, is through the direct action of molluscs. These animals are especially frequent in limey waters and have the power of abstracting lime salts from the water and utilizing the resulting lime carbonate in the formation of their shells. On the death of the animals their shells sink to the bottom and form an essential portion of any deposit which is in process of forma- tion. In some marl-beds shells amount to an impoitant percentage of the total, but in most cases they will probably constitute less than 5 per cent of the entire mass. The facts so far stated may be summaiized as follows: Spring or stream water, carrying lime carbonate in solution, Deposits it in lakes in the form of marl, this deposition being caused by: (a) Escape of carbon dioxide, owing to decrease in pressure; (6) Supersaturation, owing to rise in temperature; (c) Abstraction of carbon dioxide by plants; (d) Freeing of oxygen by plants, resulting in formation of carbon- ates from bicarbonates ; (e) Direct abstraction and crystallization of lime salts by Chara. (f) Abstraction of limeby molluscs and formation of shell deposits. The formation of a given marl-bed may be due to the operation of any one of these causes, or the cooperation of two or more of them. Geographic distribution of marl deposits. The geographic dis- tribution of marl deposits is intimately related to the geologic history of 'the region in which they occur. Marl-beds are, as indicated in the preceding section, the result of the filling of old lake basins. Lakes are not common except in those portions of the United States which were affected by glacial action, since lakes are in general due to the damming of streams by glacial material or to irregularities in deposition of such material. Workable marl deposits, therefore, are almost ex- clusively confined to those portions of the Unit d States and Ganada lying north of the former southern limit of the glaciers. Marl-bed^ are found in the New England States, where, however, they are seldom of important size, and in New York, large beds occurring in the central and western portions of that State. Deposits are frequent and important in Michigan, and in the northern portions of Ohio, In- diana, and Illinois. Marl-beds occur in Iowa, Wisconsin, and Minnesota, but have not been as yet exploited for cement-manufacture. Extensive marl-beds also occur in Ontario, Quebec, and other Canadian prov- inces. 340 CEMENTS, LIMES, AND PLASTERS. Physical characters of marl. Marl as found in existing lakes may contain as high as 50 per cent of water, while even the dry marl-beds occurring in swamps or marshes will carry 15 to 25 per cent of moisture. This moisture, together with the fine granular character of the marl itself, gives it a sticky, putty-like character. In color pure marl is white, but it usually contains so much organic matter as to give even the better samples a grayish' or yellowish tint, while the more impure marls may be very dark gray in color. Marl usually contains very little sand or grit, though some of its shells and lime carbonate particles may give it a*- gritty feeling when examined. Such shells, etc., can, however, be usually crushed between the fingers, which will serve to distinguish them in the field from sand grains. Though as sticky as clay, marl is markedly lighter in weight, owing to the high percentage of moisture which it contains. The natural fineness of marl is a matter which is of direct interest to the Portland-cement manufacturer, because of its effect on the cost of grinding the raw material. Marls differ quite widely in this regard, some being fine-grained throughout, while others contain considerable percentages of coarse material, including shells, etc. The sieving tests tabulated below * were carried out by Prof. Davis on samples of marl from three Michigan localities, and serve to show the differences in fineness above noted. TABLE 156. FINENESS OF CRUDE MARL. (DAVIS.) 1. 2. 3. Residue on it it ti 1 1 tt 1 1 it tt Passing 20-mesh sieve Per Cent. 32.25 6.06 7.58 2.90 4.81 46.40 Per Cent. 31.52 14.48 12.76 2.56 6.74 31.94 Per Cent, 0.36 3.53 6.51 3.34 6.44 79.82 40- ' ' 60- ' ' 80- ' ' 100- ' ' 100- ' ' 1. Cedar Lake. 2. Littlefield Lake. 3. Michigan P. C. Co., Cold water. The weight of the marl is also a matter of economic interest. A wet marl, as dredged from a lake bottom carrying from 50 to 60 per cent of water, may average about 2000 Ibs. per cubic yard, so that a cubic yard of such material would contain only about 800 to 1000 Ibs. dry marl. A dry marl taken from a well-drained marsh or swamp may run as * Vol. 8, pt. 3, Reports Michigan Geol. Survey, pp. 74-77. FRESH-WATER MARLS. 341 low as 20 per cent of water. Such a marl would then weigh about 2600 Ibs. per cubic yard, and a cubic yard would contain about a ton of pure marl. In dealing with the wet marls a cement-plant may produce from one and a half to three barrels of cement from each cubic yard of marl, while a marsh marl might yield four barrels cement per cubic yard. In estimating the life of a lake marl deposit it will be safest to assume that each cubic yard of marl in place will produce only two barrels of cement. Chemical composition of marl. Marl itself, being a chemical deposit, is almost a pure carbonate of lime. During and after its deposition, how- ever, foreign matter of various kinds is apt to get mixed in with the marl, the principal impurities thus introduced being fine sand, clayey matter, and organic material. Of these the most important, from the cement manufacturer's point of view, is the organic matter. Sand is rarely present in sufficient amount to render the marl unser- viceable, and of the 2 or 3 per cent of sand shown by most marls some is fine enough to pass a 150-mesh sieve and will therefore enter into combination in the kiln. The clay present in marls is principally objec- tionable because of its tendency to increase the percentage of magnesia and sulphur trioxide. Organic matter burns out in the kiln and might therefore be regarded as a harmless impurity. But a high percentage of it in a marl is in reality very objectiona'ble, both negatively, because it lowers the percentage of lime carbonate in the marl, and positively, because it retains moisture with great avidity. It is almost impossible to dry a marl containing much organic matter, and in any semi-dry or dry process this would be a very serious disadvantage. Organic matter in its coarser forms i.e., roots, branches, twigs, etc., interferes greatly with the grinding of the marl, though the larger fragments are usually taken out by a separator early in the reducing process. In the following table (157) are given the analyses of marls used at different American cement-plants, some quoted from published sources and others supplied by the chemists of the plants. A few of the quoted analyses are taken from prospectuses, but in general the analyses are of more satisfactory character. In all cases they are calculated dry, all water below 212 being neglected. The analyses given in this table are mostly not picked analyses, such as are usually quoted in prospectuses, in which the marl rarely carries less than 98 per cent of lime carbonate. On the other hand, some of them are still considerably better than can be expected 342 CEMENTS, LIMES, AND PLASTERS. TABLE 157, ANALYSES OF MARLS USED IN AMERICAN CEMENT-PLANTS. Silica (Si0 2 ). Alu- mina (AlsOs). Iron Oxide (Fe 2 3 ). Lime (CaO). Magnesia (MgO). Alkalies ( S&,. Sulphur oxide (S0 3 ). Carbon Dioxide (COaJ. Organic Matter. 1 2 3 4 5 6 7 '8 9 10 11 12 13 14 15 10 17 18 19 20 21 22 23 24 25 26 1.74 1.78 0.85 0.66 3.80 0.19 0.22 0.77 1.24 0.72 0.06 1.19 1.20 n. d. 0.90 1. 0. 0. n. 0.05 0. 0. 0. 0.24 0. 0.55 0.55 n. d. 0.28 21 86 62 d. 0.07 76 11 80 0.12 80 0.25 0.40 n. d 49.84 49.55 51.04 53 . 17 51.10 51.31 51.56 53.58 50.90 55.12 55.00 52.50 51.15 51.90 50.77 53.50 52.30 52.80 50.98 47.86 53.16 54.40 50.95 52.86 50.62 54.44 l.ft 1.30 1.31 0.47 1.54 1.93 1.26 0.91 1.43 0.44 lYl*6 0.37 0.83 tr. 0.3 1.01 0.18 0.19 0.04 1.50 2.34 0.55 n. d. 2.09 0.30 :::: 1.12 1.58 46. 40.35 42.35 41.82 42.40 46. 43.14 41.54 43.77 43.22 42.51 40.59 41.67 39.89 42.26 41.73 40.26 42.11 42.20 40.03 45. 43.10 01 4.23 2.53 1.50 2.25 2C 4.09 n. d. n. d. n. d. n. d. 0.86 1.54 1.68 0.25 58 1.12 .... G\24. 0.14 i.'si 5.79 n. d. n. d. tr. 0.05 tr. 0.26 0.20 0.66 1.7 2.01 1.65 0.4 0.42 0.26 0.26 6.22 0.14 0.54 1.98 0.26 1.43 0.46 0. 0.2 1. 0. 0.21 I 1.70 1 0. 0. 0. 0. 0.20 | 0. 81 0.2 C8 10 0.01 0.86 36 56 97 20 0.18 44 2.20 0.12 0.10 1-2. Sandusky P. C. Co., Syracuse, Ind. S. B. Newberry, analyst. 25th Ann. Rep. Indiana Dept. Geology, p. 29, 182. 3. W abash P. C. Co., Stroh, Ind. W. R. Oglesby, analyst. 25th Ann. Rep. Indiana Dept. Geology, p. 112. 4. Wabash P. C. Co., Stroh, Ind. 5. Millens P. C. Works, South Bend, Ind. H. H. Hooper, analyst. 25th Ann. Rep. Indiana Dept. Geology, p. 25. 6. Millens P. C. Works, South Bend, Ind. W. A. Noyes, anaylst. 25th Ann. Rep. Indiana Dept. Geology, p. 273. 7. Peninsular P. C. Co., Cement City, Mich. J. G. Dean, analyst. Vol. 8, Mich. Geol. Survey, p. 236. 8. Newaygo P. C. Co, Newaygo, Mich. Lathbury and Spackman, analysts. Manufacturer's prospectus. 9. Newaygo P. C. Co., Newaygo, Mich. Vol. 8, Mich. Geol. Survey, p. 240. 10. Allerman, analyst. Manufacturer's prospectus. 12-13. Wolverine P. C. Co., Coldwater, Mich. H. E. Brown, analyst. Vol. 8, Mich. Geol. Survey, p. 247. 14. Wolverine P. C. Co., Coldwater, Mich. "Cement Industry", p. 78. 15. Bronson P. C. Co., Bronson, Mich. Mineral Industry, vol. 6, p. 99. 16. Iroquois P. C. Co., Caledonia, N. Y. 22d Ann. Rep. N. Y. State Geologist. 17. Millens P. C. Works, Wayland, N. Y. 18-19. Empire P. C. Co., Warners, N. Y. 20. Montezuma, N. Y. First marl used for cement. Mineral Industry, vol. 1, p. 52. 21. American C. Co., Jordan, N. Y. 22. Genesee Wayland P. C. Co., Perkinsville, N. Y. 23. Buckeye P. C. Co., Harper, Ohio. Mineral Industry, vol. 1, p. 52. 24. Castalia P. C. Co., Castalia, Ohio. 25. Imperial P. C. Co., Owen Sound, Ontario. 26. Canadian P. C. Co., Marlbank, Ontario. Rep. Ontario Bureau Mines, 1901, p. 16. FRESH-WATER MARLS. 343 in steady practice. The following data may throw some light on what sort of results are really obtained when a marl deposit is worked con- tinuously. .First, as to water percentages: At a Michigan plant which takes its marl from under about 10 feet of water, an average of ten con- secutive analyses gave the following results. Silica (SiO 2 ) 0.07 Alumina (A1 2 O 3 ) 1 Q lg Iron oxide (Fe 2 O 3 ) / Lime carbonate (CaCO 3 ) 39.80 ( = CaO 22.29%) Organic matter . 59 Water 59.36 The marl in question is, it will be noted, very pure, being low in both clay and organic matter. The point to be noted is the high per- centage of water, each 100 Ibs. of material dredged containing approxi- mately only 40 Ibs. of lime carbonate, with 60 Ibs. of water. It must be borne in mind that this marl was not pumped to the plant, so that this large percentage of water represents only what was unavoidably taken up with the marl during dredging. Second, as to purity: A long series of analyses of marl at another prominent Michigan plant gave the following limits of results. These are calculated on a dry basis, the 55 per cent or so of water which the marl carried when reaching the plant being neglected. Silica (SiO 2 ) 1.40 to 8.60 Alumina (A1 2 O 3 ) 0.55 Iron oxide (Fe 2 O 3 ) .25 Lime (CaO) 54. 60 Magnesia (MgO) 1 .25 Carbon dioxide (CO 2 ) 42.90 Organic matter . 05 1.30 1.54 46.20 2.78 36.30 10.50 It will be seen from these results that the manager should expect to receive material carrying often very high percentages of water and frequently containing a large amount of organic matter and other im- purities. Examining marl deposits. Owing to the nature of the material and the physical conditions under which it occurs, the work of examin- ing and valuing a marl deposit presents certain features of difficulty peculiar to itself. As in any other prospecting work, the two factors which require determination are respectively the extent of the deposit and the composition of the material. 344 CEMENTS, LIMES, AND PLASTERS. When the marl occurs in an old lake-bed overlain by soil or peat, the area can be roughly laid out into blocks or squares of convenient size, borings being made and samples taken at the corners of these squares. A pocket compass, or, better yet, the cheap " drainage level" made t>y Gurley and other instrument-makers, will suffice for laying out the work. A marl deposit covered by water must be handled like any sound- ing proposition. The shore line should*first be roughly surveyed, after which soundings and borings must be made from a raft or boat, the position of each boring being located by bearings taken to fixed points in the shore. A broad steady platform is necessary for the borings. This is afforded either by using a raft with a square opening in its center or by laying planks across the ends of two boats. As the methods employed in determining the thickness and char- acter of marl-beds are of a somewhat special character and are not fully described in current engineering text-books, the writer feels justified in introducing here a very detailed account of these methods by Mr. David J. Hale, of the Michigan Geological Survey. This account is taken almost verbatim from the paper cited below.* In dealing with fairly solid marls not deeply covered with peat or soil, the simple outfit described below has proven very successful when manipulated with care. It is prepared as follows: Weld an ordinary 2-inch auger on a f-inch gas-pipe 2 feet long. Thread the unwelded end of the pipe for coupling. Cut three lengths of pipe each in half or in four pieces if it is desired to carry the outfit long distances. Thread the ends of these sections for coupling. Get couplings enough to couple all together, so as to make a continuous hollow rod with attached auger. Insert a "T" coupling on the handle end, or end farthest from the auger, and pass a rod or stick through this so as to turn the outfit. A better way is to screw into each free end of the "T" a rod or piece of gas-pipe 18 inches long. Usually a pair of Stillson wrenches are needed to untwist the pipe, which becomes jammed during the boring. Three-eighths-inch pipe, as above recommended, will be found to lift out much easier than half-inch, but will not do for deep borings-inch pipe is entirely too light and 1-inch pipe is too difficult to handle. A light, easily handled outfit is a great aid in boring, because the quicker the rod can be driven the less friction there will be to contend with, for the marl particles will not have time to settle after each movement of the rod. * Hale, D. J. The use of marl for cement manufacture. Vol. 8, pt. 3, Reports Mich. Geol. Survey, especially pp. 9-13, 108-110. FRESH-WATER MARLS. 345 In boring, the handle should be twisted around as the rod is shoved down, even though the surface material may be soft enough to permit it to be driven down without twisting, because if the latter method is adopted the surface material first taken up by the auger will cling to it during its descent, and will prevent securing a sample of the marl at lower depth. As each new length of pipe is added, the couplings should be firmly tightened, as neglect of this precaution may mean the loss of most of the outfit through uncoupling while it is being drawn up. This device gives good results when employed on a fairly dense marl not deeply covered by peat or grass, because the auger will clear itself of the surface material on the way down and will retain fairly well the clean sample taken at the bottom. In heavily covered marl-beds, or in dealing with very fluid marls, other boring devices must be tried. A clumsy but efficient type of sampler for this purpose is made as follows: Take a 2-foot length of 1-inch gas-pipe and thread one end for coupling. Screw reducers on the coupled end until the last reducer can take a f-inch or J-inch pipe, any necessary number of lengths of which pipe will form the rod proper. Sharpen the open edge of the 1-inch pipe and fit into it a plug with a shoulder that fits against the rim, allowing the plug to penetrate ' J inch into the open end of the pipe. Sharpen the end of the plug opposite the shoulder and bore a hole lengthwise through the plug. Pass a &-inch iron rod through the plug from the shoulder end and bolt it by screwing a nut upon the end opposite the shoulder, which end should be sharpened so as to penetrate the marl more easily. The end of the rod may be threaded for several inches and a nut screwed on, then pass the end of the rod through the plug and screw the nut tight against the plug. This will hold the plug in place during boring and withdrawing. The rod is inserted in the 1-inch pipe and passed up through* that and the f-inch pipe, from the upper end of which the free end of the rod may project. The rod gives a means of either closing firmly or opening the bottom end of the 1-inch pipe. When boring, the plug is held firmly against the mouth of the pipe by means of the rod, and the whole appara- tus is shoved down into the marl to the desired depth. The pipe is then raised, while the rod is held stationary; the apparatus is held in this position a moment to allow the marl to close in about the rod and then the pipe is lowered until the plug closes it. With the plug held firmly the entire apparatus is raised to the surface, the plug loosened, and the sample of marl taken out of the section of 1-inch pipe. This device can be driven through and withdrawn from a marl-bed covered by peat, 346 CEMENTS, LIMES, AND PLASTERS. grass, etc., and still the marl sample can be preserved from intermixture with these materials. A borer invented and manufactured by Robert G. Hunt & Co. is also noted by Mr. Hale. This consists of a piece of steel about 18 feet long, much the shape of the half of a long gun-barrel split longitudinally. The end which first enters the marl is capped and pointed with steel so that it will penetrate more easily, Ttfhile the other end is provided with a handle for raising the apparatus. The two vertical edges of the barrel are sharpened so as to cut the marl. When the instrument has been driven down to the desired depth it is turned half around, filling the half cylinder with marl for its entire length, and then withdrawn. This gives a perfect sample of the bed from top to bottom. The various devices above described will generally give satisfactory results when operating in moderate depths of water and on any but the most fluid marls. For these latter, as well as for sampling in deep water, special devices are required, which are described by Mr. Hale in the paper cited. But in examing deposits of marl to be used as Portland-cement material the very deep and very fluid marls may be dismissed without sampling. For under present economic conditions such materials could not be profitably used in cement-manufacture. After the results of the borings have been plotted in such a way as to give both depth of water and thickness of marl, the amount of marl available can be calculated quite closely. In making these estimates it will be safest to assume that each cubic yard of marl in the lake will yield 900 pounds of dry marl, or sufficient to make two barrels of cement. Each rotary kiln in the prepared plant will on this basis use about 50 to 60 cubic yards of marl per day, or about 18,000 cubic yards per year. An eight-kiln plant should therefore own about 3,000,000 cubic yards of marl, which would insure a twenty-year supply of raw material. List of references on marls. Of the following papers on marls, those dealing chiefly with the origin of marl deposits are marked A; those describing the deposits of certain States or areas are marked B] and those discussing the technology of marls as cement materials are marked C: A, B. Blatchley, W. S., and Ashley, G. H. The lakes of northern Indiana and their associated marl deposits. 25th Ann. Rep. Indiana Dept. Geology and Natural Resources, pp. 31-321. 1901. A. Davis, C. A. A contribution to the natural history of marl. Journal of Geology, vol. 8, pp. 485-497. 1900. A. Davis, C. A. A remarkable marl lake. Journal of Geology, vol. 8, pp. 498-500. 1900. FRESH-WATER MARLS. 347 A. Davis, C. A. Second contribution to the natural history of marl. Journal of Geology, vol. 9, pp. 491-506. 1901. A. Davis, C. A. A contribution to the natural history of marl. Vol. 8, pt. 3, Reports Michigan Geological Survey, pp. 65-102. 1903. B, C. Eckel, E. C. Cement materials and cement industries of the United States. Bulletin 243, U. S. Geological Survey, 1905. B. Ells, R. Vv 7 . Marl deposits of eastern Canada. Ottawa Naturalist, vol. 16, pp. 59-69. 1902. B. Fall, D. Marls and clays in Michigan. Vol. 8, pt. 3, Reports Michigan Geological Survey, pp. 343-348. 1903. A y B,C f Hale, D. J. Marl and its application to the manufacture of Portland cement. Vol. 8, pt. 3, Reports Michigan Geological Survey, pp. 1-64, 103-190. 1903. A. Lane, A. C. Notes on the origin of Michigan bog-limes. Vol. 8, pt. 3, Reports Michigan Geological Survey, pp. 199-223. 1903. B, C. Lane, A. C. List of marl localities and Portland-cement mills in Michigan. Vol. 8, pt. 3, Reports Michigan Geological Survey, pp. 224-342. 1903. <7. Lathbury, B. B. The development of marl and clay properties for the manufacture of Portland cement. Vol. 8, pt. 3, Reports Michigan Geological Survey, pp. 191-198. 1903. B. Ries, H. Lime and cement industries of New York. Bulletin 44, New York State Museum, pp. 326. 101. A, B. Russell, I. C. The Portland-cement industry in Michigan. 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 629-685. 1902. C. Spackman, H. S. The manufacture of cement from marl and clay. Proc. Engineers' Club of Phila 1 ., April, 1903. Engineering News, vol. 49, pp. 492-494. June 4, 1903. CHAPTER XXVI. ALKALI WASTE: BLAST-FURNACE SL^G. THE two raw materials to be discussed in the present chapter agree in being waste products or by-products of other industries which, because of their chemical composition, can be used in Portland-cement manu- facture. In almost every other respect they differ. Alkali waste is a fine-grained, soft and pure form of lime carbonate. Slag is very hard, coarse-grained, and is composed of lime (CaO), silica, and alumina. Waste products or by-products can, of course, be usually obtained at a low or nominal cost, and on this account both slag and alkali waste assume an importance entirely out of proportion to their other proper- ties. But it must be recollected that as by-products their production and quality depend entirely upon the condition of the industries of which they are wastes, and that no furnace manager or alkali- works superintendent will run his plant solely in or*der to turn out a by-product regular in amount and composition. For this reason it is essential that a cement-plant using a waste product must be closely identified in owner- ship with the furnace or works which furnishes this waste product. Common ownership is practically the only way of insuring a sufficient and regular supply of satisfactory composition. Alkali Waste. \ A large amount of waste material results from the processes used at alkali works in the manufacture of caustic soda. This waste mate- rial is largely a precipitated form of calcium carbonate, and if sufficiently free from injurious impurities, furnishes a cheap source of lime for use in Portland-cement manufacture. The value of the waste for this pur- pose depends largely on the process from which it resulted. Leblanc-process waste. The waste resulting from the Leblanc process carries a very large percentage of sulphur, mostly in the form of lime sulphides, carried over from the pyrites used in the process. A 348 ALKALI WASTE: BLAST-FURNACE SLAG. 349 fairly typical analysis * of Leblanc-process waste is given below and will serve to show the composition of this material: Lime carbonate (CaCO 3 ) 41 .20 Lime sulphate (CaSO 4 ) 2 . 53 Lime hydrate (CaH 2 O 2 ) 8.72 Lime disulphide (CaS 2 ) 5 . 97 Lime sulphide (CaS) 25.79 Sodium sulphide (Na 2 S). 1 .44 Magnesium silicate (MgSiO 3 ) 3.63 Phosphates of iron and alumina 8.91 This material is obviously unfit for use in the manufacture of Port- land cement. Attempts have been made to recover the sulphur con- tained in the waste, but the removal of this constituent is never suffi- ciently perfect to permit the resulting waste to be of use to the cement manufacturer. Ammonia-process waste. The waste from ammonia-process works is, on the contrary, a very pure mass of precipitated lime, mostly in the carbonate form, though some lime hydrate is always present. As pyrite is not used in this process, the sulphur in the waste is commonly well within Portland-cement limits. The magnesia content of the waste may or may not be High, according to the character of the limestone that has been used in the process of soda-manufacture. When a pure limestone low in magnesium carbonate has been used, the waste will be low in magnesia and is then a very satisfactory Portland-cement material. The following analyses are representative of the waste obtained at alkali- plants using the ammonia process. TABLE 158. ANALYSES OF ALKALI WASTE, AMMONIA PROCESS. 1. 2. 3. 4. 5. Silica (SiO 2 ) CO 1.98 1.75 n. d. 0.98 Alumina (A1 2 O 3 ) 1.41 Iron oxide (Fe 2 O 3 ) } 3 - 04 { 1.38 > 0.61 2.20 1.62 Lime (CaO) 53 33 48 29 50 60 52 40 50 40 Magnesia (MgO). 0.48 1.51 5.35 3.75 4.97 Alkalies (K 2 O Na 2 O) 20 64 64 20 0.50 Sulphur trioxide (SO 3 ) n. d. 1 26 n. d. n. d. n d Sulphur (S) n. d. n. d. 0.10 n. d. 06 Carbon dioxide (CO.) 42.43 39.60 41.17 n. d. Water n d 3 80 V 41. 70 < n d n d The analyses given in the above table are of alkali wastes which have at one time or another been used in the manufacture of Portland * Kingzett, C. T. The Alkali Trade, p. 134. 350 CEMENTS, LIMES, AND PLASTERS. cement either in the United States or in England. The effect on the waste when a magnesian limestone is used in the alkali-plant is well shown by analyses 3, 4 and 5, in all of which the magnesia is high for a Portland-cement material. At the only American cement-plant which uses alkali waste the materials (clay and waste) are mixed wet. The waste carries 90 to 95 per cent of lime carbonate/ while th clay used gives the following analysis : Silica (SiO 2 ) 63.54 Alumina (A1 2 O 3 ) \ _ Iron oxide (Fe 2 O 3 ) J Lime (CaO) 1 .66 Magnesia (MgO) 1 .05 Alkalies (K 2 O,Na 2 O) 0.78 Carbon dioxide (CO 2 ) 2.47 Water 7.05 The clay is put through a rotary drier and ground in a dry -pan, after which the waste and clay are mixed in a wet pug-mill and ground in wet tube mills. The mix is made drier than at most of the plants using marl and contains usually about 40 per cent of water. It is not, of course, necessary that wet mixing should be practiced when alkali waste is employed as one raw material. The waste could be dried, though it is possible that its physical properties might render this more difficult than drying limestones or clays. List of references on alkali waste as a cement material. The follow- ing brief list covers the few available references on this subject: Butler, D. B. [Alkali waste used in England.] Portland Cement: its Manu- facture, Testing, and Use, pp. 25-27. 1899. Lathbury, B. B. The Michigan Alkali Company's plant for manufacturing Portland cement from caustic-soda waste. Engineering News, June 7, 1900. Lathbury, B. B., and Spackman, H. S. The Michigan Alkali Company's plant, Wyandotte, Michigan. The Rotary Kiln, pp. 110-119. 1902. Redgrave, G. R. [Use of alkali waste in England.] Calcareous Cements: their Nature and Uses, pp. 182-184. 1895. Blast-furnace Slag. True Portland cements, which must be sharply distinguished from the slag (or puzzolan) cements described in Part VII of this volume, can be made from mixtures which contain blast-furnace slag as one ingredient. In this case the slag is intimately mixed with limestone and the mixture is finely powdered. It is then burned in kilns and the resulting clinker pulverized. ALKALI WASTE: BLAST-FURNACE SLAG. 351 The slags from iron furnaces consist essentially of lime (CaO), silica (SiO 2 ), and alumina (A^Os), though small percentages of iron oxide (FeO), magnesia (MgO), and sulphur (S) are commonly present. Slag may therefore be regarded as a very impure limestone or a very calcareous clay from which the carbon dioxide has been driven off. Two plants are at present engaged in the United States in the manufacture of true Portland cement from slag, and there seems to be no reason why this cheap and satisfactory raw material should not become an important factor in the cement industry of the country. Slags in general. Slags are the fusible silicates formed, during the ' smelting or refining of metals, by the combination of the fluxing materials with the gangue of the ore. The composition of the slag, therefore, will be determined by the composition and relative proportions of the fluxes and the gangue. In general, the slag will contain only those elements which are present in either the gangue or the flux, though it may carry also a percentage, usually small, of the metal which is being smelted or treated. In some processes also the composition' of the slag may be slightly modified through the action of the fuel, from which certain impurities may be taken up. While many elements may occur in slags, those which are of uni- versal or even common occurrence are comparatively few. The slags commonly found in iron metallurgy consist essentially of silica, alu- mina, iron oxide, and lime, with or without magnesia. Alkalies, sul- phur, and phosphorus are also almost invariably present, but in small percentages. The following analyses of slags from various furnaces will serve to give some idea of the range in composition of these products. TABLE 159. ANALYSES OF IRON-FURNACE SLAGS. Silica (SiO 2 ) 30 00 30.72 32 51 32 90 26 88 31 65 Alumina (A1 2 O 3 ) 28 00 16.40 13.91 13.25 24.12 17 00 Iron oxides (FeO,Fe 2 O 3 ) . Lime (CaO) 0.75 32 75 0.43 48 59 0.48 44 75 0.46. 47 30 0.44 45 11 0.65 47 20 Magnesia (MgO) 5 25 1 28 2 20 1 37 1 09 1 36 Lime sulphide (CaS). . . . 1.90 2.16 4.90 3.42 1.86 n. d. Silica (SiO 2 ) 28 35 38 00 31 50 32 20 33 10 Alumina (A1 2 O 3 ) 18 15 10 00 18 56 15 50 12 60 Iron oxides (FeO,Fe 2 O 3 ) 1 50 n d n d n d n d Lime (CaO) 47 40 46.0 42 22 48 14 49 98 Magnesia (MgO) 2.45 n. d. 3.18 2 27 2.45 Lime sulphide (CaS) n d n. d. n d n d n d 352 CEMENTS, LIMES, AND PLASTERS. Slags used as Portland-cement materials. The slags used in Port- land-cement manufacture are iron blast-furnace slags of the more basic types i.e., those in which the lime (CaO) reaches 30 per cent or over. The higher the lime,., up to say 50 per cent, the more valuable the slag for this use. The composition of the slags will usually be controlled, however, by the requirements of the furnaces, not by the needs of the cement-plant. ^ The following shows the range in composition of the slags used at a German Portland-cement plant. ANALYSES OF SLAG USED IN PORTLAND-CEMENT MANUFACTURE. Per Cent. Silica (SiO 2 ). 30 to 35 Alumina (A1 2 O 3 ) 10 "14 Iron oxide (FeO) 0.2 " 1.2 Lime (CaO) 46 "49 Magnesia (MgO) 0.5 " 3.5 Sulphur trioxide (SO 3 ) 0.2 " 0.6 As a Portland-cement material slag possesses one great advantage in addition to its cheapness. This advantage is chemical, and is due to the fact that the lime contained in the slag is present in the form of oxide (CaO), instead of carbonate (CaCOs), as in limestones. It does not require to be decarbonated, and therefore a mixture made up of slag and clay will clinker with less fuel than one consisting of limestone and clay. Opposed to this chemical advantage is a physical disadvantage. If the slag is allowed to cool as it issues from the furnaces, it solidifies into very hard and tough masses much more resistant than the hard- est of limestones. In order to avoid this difficulty, it is the common practice to "granulate" the slag, i.e. to run it direct from the furnace into cold water. This proceeding breaks up the slag into little porous granules -^ to J inch in diameter, and incidentally removes part of the sulphur contained in the slag. But to offset these gains, it intro- duces a large amount of water into the product, so that a granulated slag may carry from 20 to 40 per cent of water, and this greatly increases the cost of drying. As the chemical and physical properties of slag introduce certain interesting features into the manufacture of Portland cement from a limestone-slag mixture, this mixture will be discussed separately on later pages. CHAPTER XXVII. CLAYS, SHALES, AND SLATES. EXCEPT when a very clayey limestone or a slag is one component of a Portland-cement mixture, the silica, alumina and iron oxide necessary for the mix are always supplied in the form of clay, shale, or slate. The materials known respectively as clays, shales, and slates are of practically the same composition and ultimate origin, but differ in their degree of consolidation. Clays are ultimately derived from the decay of older rocks, the finer particles resulting from this decay being carried off and deposited by streams along their channels, in lakes, or along parts of the seacoast or sea-bottom as beds of clay. In chemical composition the clays are composed essentially of silica and alumina, though iron oxide is almost invariably present in more or less amount, while lime, magnesia, alka- lies, and sulphur are of frequent occurrence, though usually only in small percentages. Shales are clays which have become hardened by pressure. The so-called "fire-clays" of the Coal Measures are shales, as are many of the other "clays" of commerce. The slates include those clayey rocks which through pressure have gained the property of splitting readily into thin parallel leaves. Clays. The term clay is applied to fine-grained unconsolidated materials which possess the property of plasticity when wet, while they lose this property and harden on being strongly heated. Being, as explained below, the finer de*bris resulting from the decay of many different kinds of rocks, the clays will naturally differ greatly among themselves in composition, etc. Origin of clays. When rocks of any kind are exposed to atmospheric action, more or less rapid disintegration sets in. This is due partly to chemical and partly to physical causes. It is hastened, for example, by the dissolving out of any soluble minerals that may occur in the rock, by the expansion and contraction due to freezing, and by the action of the organic acids set free by decaying vegetable matter. The 353 354 CEMENTS, LIMES, AND PLASTERS. more soluble ingredients of the rock are usually removed in solution by surface or percolating waters, while the more insoluble portions are either left behind or are carried off mechanically by streams. These relatively insoluble materials when sufficiently fine grained constitute the clays. When they are left as a deposit in the spot where the orig- inal rock disintegrated, they are cajled residual clays; when they are carried off by surface-waters and finally deposited in the sea or along river-beds they are transported or sedimentary clays. A third class of particular interest to the cement manufacturer are the glacial clays, deposited under or in front of the glaciers which formerly covered most of the northern states. Composition of clays. The residual, sedimentary and glacial clays usually differ markedly in composition, owing to the different manner in which they have been deposited. The residual clays, for example, are apt to contain coarse fragments of any very insoluble and hard material which the original rock may have contained. A residual clay arising from the decay of a granite will probably contain frag- ments of quartz; one derived from a limestone may contain chert or flint, as well as masses of undissolved lime carbonate. A sedimen- tary clay, on the other hand, having been transported by water for great distances, has usually lost all its coarser material, and is a fine grained and homogeneous product. The glacial clays, being formed mechanically by the abrading power of the ice, show even less homo- geneity than the residual clays, and are apt to contain much sand, gravel, and pebbles. Clays used in Portland-cement manufacture. For use as Portland- cement materials clays should be as free as possible from gravel and sand, as the silica present as pebbles or grit is practically inert in the kiln unless ground more finely than is economically practicable. In composition they should not carry less than 55 per cent of silica, and pref- erably from 60 to 70 per cent. The alumina and iron oxide together should not amount to more than one-half the percentage of silica, and the composition will usually be better the nearer the ratio Al 2 03+Fe 2 O 3 Si0 2 . = - is approached. o Nodules of lime carbonate, gypsum, or pyrite, if present in any quan- tity, are undesirable; though the lime carbonate is not absolutely injurious. Magnesia and alkalies should be low, preferably not above 3 per cent. The clays actually used in cement plants may be separated, for con- venience, into the normal clays and the limey clays. In this section the CLAYS, SHALES, AND SLATES. 355 dividing line between these classes will be fixed arbitrarily at 5 per cent of lime (CaO) and magnesia (MgO) together, all clays containing over 5 per cent of both oxides being called limey clays, while those carrying less than 5 per cent are termed normal clays. TABLE 160. ANALYSES OF NORMAL CLAYS USED IN AMERICAN CEMENT-PLANTS. Silica (Si0 2 ). Alumina (A1 2 3 ). If 1 1 1 1 1 Sulnhur Trioxide (S0 3 ). |i Water. 1 1 53.30 23.29 9.52 0.35 1.49 n. d. n. d. n. d. n. d. 1.62 2 65.12 19.05 7.66 0.34 0.31 n. d. n. d. n. d. n. d. 2.44 3 63.73 22.12 9.01 2.83 0.21 n, d. n. d. n. d. 2.04 4 53.21 15.91 7.25 1.89 6^99 2.21 0.97 17.21 2.29 5 59.10 25.41 3.61 0.87 1.10 0.31 9.81 2.32 6 58.02 26.12 3.70 1.00 1.55 0.33 9.27 2.22 7 57.25 26.15 3.09 1.10 1.88 0.39 9.30 2.19 8 58.25 18.56 7.35 3.10 1.28 2.35 0.45 8.55 2.25 9 74.29 12.00 4.92 0.41 0.68 2.56 n. d. n. d. n. d. 4.39 10 63.54 24.00 .66 1.05 0.78 2.47 7.05 2.65 11 64.65 24.31 .16 1.92 n. d. n.' d. n. d. n. d. 2.66 12 60.28 27.10 .20 n. d. n. d. n. d. 8.38 2.22 13 60.30 29.78 .98 n. d. n. d. n. d. 8.30 2.03 14 61.40 25.08 .40 n. d. n. d. n. d. 6.60 2.45 15 63.82 25.36 3.42 n. d. n. d. n. d. 6.96 2.52 16 63.07 24.00 3. CO 1.20 n. d. n. d. 8.80 2.63 17 61.92 16.58 7.84 2.01 1.58 3.64 tr. n. d. n. d. 2.53 18 65.68 24.08 2.C1 1.75 n. d. n. d. n. d. n. d. 2.73 19 62.5 20.2 7.5 0.8 1.8 n. d. 0.4 n. d. n. d. 2.26 20 58.90 27.50 4.08 0.79 n. d. n. d. n. d. n d. 2.14 21 59.10 24.01 2.20 2.00 n. d. n. d. n. d. n. d. 2.46 22 64.85 17.98 5.92 2.24 1.40 n. d. n. d. 4.98 2.82 23 58.04 14.63 9.27 1.53 2.02 n. d. 0.37 12 .67 2.33 1,2. Whitecliffs P. C. Co., Whitecliffs, Ark. Trans. Amer. Irjst. Mining Engrs., vol. 21. 3. Santa Cruz, Calif. Mineral Industry, vol. 1, p. 52. 4. Pacific P. C. Co., Suisun, Calif. C. J. Wheeler, analyst, 5. I 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20 21. 22. 23. Bedford P. C. Co., Bedford, Ind. A. W. Smith, analyst. 25th Ann. Rep. Indiana Dept. Geology, p. 328. Wyandotte P. C. Co., Wyandotte, Mich. Peerless P. C. Co., Union City, Mich. Lundteigen, analyst. Peninsular P. C. Co., Cement City, Mich. Catskill P. C. Co., Smiths Landing, N. Y. American Cement Co., Jordan, N. Y Iroquois P. C. Co., Caledonia, N. Y.' Hudson P. C. Co., Hudson N. Y. Heiberg and Roney, analysts. Castalia P. C. Co., Castalia, Ohio. Omega P. C. Co., Jonesville. Mich. Vol. 8, Mich. Geol. Survey, p. 229. Almendares P. C. Co., Marinao, Cuba. Engineering Record, vol. 49, p. 37. 356 CEMENTS, LIMES, AND PLASTERS. TABLE 161. ANALYSES OF LIMEY CLAYS USED IN AMERICAN CEMENT-PLANTS. Silica (SiO 2 ). Alumina (A1 2 O 3 ). 1. HH I I ' 3 . S3 | h- 1 1 3 is 0> M J5 Alkalies (K 2 0,Na 2 0). Sulphur Tri- oxide (S0 3 ). *s II r 1 SiO 2 Al 2 3 +Fe 2 3 1 57.98 18.26 4.57 1.75 1.83 n. d. 1.28 12.08 2 65.99 21.57 6.07 0.47 0.82 n. d. n. d. n. d. n.d. 3 63.30 26.00 1.25 1.25 n. d. n. d. n. d. 6.25 4 63.56 27.32 1.20 1.20 n. d. n. d. n. d. n. d. 5 55.80 30.20 1.16 n. d. n. d. n. d. n. d. n. d. 6 56.30 29.86 n. d. n. d. n. d. n. d. n. d. n. d. 7 60.00 23.26 4.32 0.90 1.12 n. d. n. d. 6.16 8 62.67 19.99 5.46 1.25 0.72 n. d. n. d. n. d. n. d. 9 60.15 19.78 9.10 0.52 0.10 n. d. tr. n. d. n. d. 10 60.97 20.66 6.59 0.65 1.13 n. d. tr. n. d. n. d. 11 55.00 21.79 9.26 n. d. n. d. n. d. n. d. 8.61 12 54.70 31.68 1.15 n. d. n. d. n. d. n. d. n. d. 1.72 13 64.30 33.60 1.46 1.30 n. d. n. d. n. d. n. d. 1.91 14 70.20 26.90 n. d. n. d. n. d. n. d. n. d. n. d. 2.61 15 61.09 19.19 6.78 2.51 0.65 3.16 1.42 5.13 16 61.15 18.47 5.05 0.98 2.26 n. d. 0.91 ' 7.02 1? 58.24 18.56 7.68 0.61 0.24 n. d. n. d. 10.04 18 59.64 19.14 7.59 0.26 2.31 4.33 n. d. 4.71 19 62.10 20.09 7.81 0.65 0.96 n. d. 0.49 20 59.02 29.84 0.68 2.16 n. d. n. d. n. d. n. d. 21 58.44 27.45 1.16 2.23 n. d. n. d. n. d. n.d. 22 58.90 27.25 1.23 2.18 n. d. n. d. n. d. n.d. 23 58.92 27.00 1.41 2.32 n. d. n. d. n. d. n.d. 24 60.24 30.04 2.21 1.55 n. d. n. d. n. d. n. d. 2.01 25 60.02 26.60 2.31 1.62 n. d. n. d. n. d. n. d. 1. Western P. C. Co., Yankton, S. D. Mineral Industry, vol. 6, p. 97. 2. Crescent P. C. Co., Wampum, Pa. 3. Wellston P. C. Co., Wellston, Ohio. W. S. Trueblood, analyst. 4-7. Ironton P. C. Co., Ironton, Ohio. C. D. Quick, analyst. 8-11. Diamond P. C. Co., Middle Branch, Ohio. E. Davidson, analyst. 12-14. Hudson P. C. Co., Hudson, N. Y. Heiberg and Roney, analysts. 15. Alpena P. C. Co., Alpena, Mich. Vol. 8, pt. 3, Reports Mich. Geol. Survey, p. 227. 16. Wolverine P. C. Co., Cold water, Mich. 17. Michigan P. C. Co., Coldwater, Mich. Cement Industry, p. 78. 18. Lehigh P. C. Co., Mitchell, Ind. 26th Ann. Rep. Indiana Dept. Geology, p. 276. 19. Bronson P. C. Co., Bronson, Mich. Mineral Industry, vol. 6, p. 99. 20. Peerless P. C. Co., Union City, Mich. Lundteigen, analyst. 22-25. Cayuga P. C. Co., Portland Point, N. Y. J. H. McGuire, analyst. 358 CEMENTS, LIMES, AND PLASTERS. FIG. 69. Pit in heavy shale-bed. FIG. 70. Shale-pit worked on two levels. CLAYS, SHALES, AND SLATES. 359 The limey shales are almost exclusively shales which occur inter- bedded, in comparatively thin layers, with limestones. Occasionally a limey shale will owe its content of lime almost entirely to the fossil shells it contains, the remainder of the shale being practically free from carbonates. For both of the above reasons limey shales are apt to- be a source of trouble in the practical working of a plant and require considerable care in quarry management to insure that the raw mate- rials are anywhere near uniform in composition from day to day. Fig. 70 shows a shale deposit which consists of three horizontal TABLE 163. ANALYSES OF LIMEY SHALES USED IN AMERICAN CEMENT-PLANTS. 3 . ? 1 jo o il 13 C II |1| o o O | It i 3 h- 1 3 a 1 HS 3 1 53.12 20.60 4.09 4.02 2.24 n. d. 2.15 13.70 2 54.30 19.33 5.57 3.29 2.57 2.36 n. d. n. d. 3 52.74 21.73 12.37 2.01 n. d. 11.27 n. d. 4 54.4 18.2 5.7 7.2 1.8 n. d. 12.3 5 54.18 19.17 6.11 7.05 1.89 n. d. 11.95 6 56.0 22.1 8.0 1.5 tr. 10.7 2.53 7 57.45 20.56 2.78 4.27 3.17 0.35 8.15 8 55.96 22.44 2.80 3.78 3.22 0.74 8.03 9 58.22 17.68 4.48 3.82 2.85 0.43 7.83 10 57.50 21.70 12.19 1.93 n. d. n. d. n. d. 11 57.82 21.76 8.32 1.81 n. d. n. d. n. d. 12 38.84 17.76 21.58 1.78 n. d. n. d. n. d. 13 50.48 8.89 23.74 2.21 n. d. n. d. n. d. 5.68 14 46.54 21.50 11.51 1.88 n. d. n. d. n. d. 15 46.72 22.00 11.82 2.11 n. d. n. d. n. d. 16 56.50 24.50 5.14 1.78 n. d. n. d. n. d. 2.31 17 58.10 26.14 3.34 2.01 n. d. n. d. n. d. 18 58.25 24.18 3.46 2.10 n. d. n. d. n. d. 19 61.94 11.58 3.49 5.92 4.85 .18 n/d. n. d. 20 56.64 12.18 3.59 8.17 4.29 .31 n. d. n. d. 21 61.10 13.91 3.62 6.32 3.91 .31 n. d. n. d. 22 59.36 12.38 3.62 5.63 4.62 .30 n. d. n. d. 23 53.63 24.47 5.94 1.79 n. d. 10.03 2.19 1. Chicago P. C. Co., Oglesby, 111. Manufacturer's circular. 2. Marquette Cement Co., Oglesby, 111. 25th Ann. Rep. U. S. Geol. Survey, pt. 6, p. 544. 3. German-American P. C. Works, La Salle, 111. W. E. Trussing, analyst. 4. lola P. C. Co., lola, Kansas. 5. " 6. Kansas P. C. Co., lola, Kansas. 7. Alpena P. C. Co., Alpena, Mich. 8. " 9. 10-18. Cayuga P. C. Co., Portland Point, N. Y. J. H. McGuire, analyst. 19-22. Bronson P. C. Co., Bronson, Mich. W. H. Simmons, analyst. Vol. 8, Mich. Geol. Survey, p. 239. 23. Virginia P. C. Co., Craigsville, Va. "Cement Industry", p. 235. 360 CEMENTS, LIMES, AND PLASTERS. beds quite different in composition. They are consequently worked as separate benches or levels, temporary tracks being run in on the upper levels, while the main switch tracks are on the lowest level. Examination of clay deposits. Most of the notes in relation to examining limestone deposits presented on an earlier page will apply to the report on a deposit of clay, or shale^ In sampling, however, the earth- auger can be used much more extensively, as the clayey materials are usually soft enough to be bored readily by such means. For valuable notes on the use of the auger, reference should be made to the papers cited below.* List of references on clays and shales. The literature of clays is so extensive that the descriptive papers in the following list have been arranged by States in alphabetical order. GENERAL UNITED STATES. Ries, H. The clays of the United States east of the Mississippi River. Professional Paper No. 11, U. S. Geological Survey, 289 pp. 1903. ALABAMA. Ries, H., and Smith, E. Preliminary report on the clays of Alabama. Bulletin 6, Alabama Geological Survey, 220 pp. 1900. ARKANSAS. Branner, J. C. The cement materials of southwest Arkansas. Trans Am. Inst. Min. Engrs., vol. 27, pp. 42-63. Branner, J. C. The clays of Arkansas. Bulletin No. , U. S. Geological Survey. (In press.) CALIFORNIA, Johnson, W. D. Clays of California. 9th Ann. Report California State Mineralogist, pp. 287-308. 1890. Ries, H. The clay-working industry of the Pacific Coast States. Mines and Minerals, vol. 20, pp. 487-488. 1900. COLORADO. Lakes, A. Gypsum and clay in Colorado. Mines and Minerals, vol. 20. December, 1899. Ries, H. The clays and clay-working industry of Colo- rado. Trans. Am. Inst. Min. Engrs., vol. 27, pp. 336- 340. 1898. FLORIDA. Memminger, C. J. Florida kaolin deposits. Eng. and Min. Journal, vol. 57, 436 pp. 1894. Vaughan, T. W. Fullers' earth deposits of Florida and Georgia. Bulletin 213, U. S. Geological Survey, pp. 392-399. 1903. * Bleininger, A. V. The manufacture of hydraulic cements. Bulletin 4, Ohio Geol. Survey, pp. 102-108. 1904. Catlett, C. The hand-auger and hand-drill in prospecting work. Trans. Amer. Inst. Min. Engrs., vol. 27, pp. 123-129. 1898. Jones, C. C. A geologic and economic survey of the clay deposits of the lower Hudson River Valley. Trans. Amer. Inst. Min. Engrs., vol. 29, pp. 40-83. 1900. CLAYS, SHALES, AND SLATES. 361 GEORGIA. INDIANA. IOWA. KANSAS. KENTUCKY. LOUISIANA. MARYLAND. MASSACHUSETTS. MICHIGAN. MINNESOTA. MISSISSIPPI. MISSOURI. NEBRASKA. NEW JERSEY. Ladd, G. E. Preliminary report on a part of the clays of Georgia. Bulletin 6A, Georgia Geological Sur- vey, 204 pp. 1898. Vaughan, T. W. Fullers' earth deposits of Florida and Georgia. Bulletin 213, U. S. Geological Survey, pp. 392-399. 1903. Blatchley, W. S. A preliminary report on the clays and clay industries of the coal-bearing counties of Indiana. 20th Ann. Rep. Indiana Dept. Geology and Natural Resources, pp. 24-187. 1896. Blatchley, W. S. Clays and clay industries of north- western Indiana. 22d Ann. Rep. Indiana Dept. Geology and Natural Resources, pp. 105-153. 1898. Beyer, S. W., and others. Clays and clay industries of Iowa. Vol. 14, Reports Iowa Geol. Survey, pp. 27- 643. 1904. Prosser, C. S. Clay deposits of Kansas. Mineral Re- sources U. S. for 1892, pp. 731-733. 1894. Crump, H. M. The clays and building stones of Kentucky. Eng. and Min. Jour., vol. 66, pp. 190-191. 1898. Clendennin, W. W. Clays of Louisiana. Eng. and Min. Jour., vol. 66, pp. 456-457. 1898. Ries, H. Report on Louisiana clay samples. Report Louisiana Geological Survey for 1899, pp. 263-275. 1900. Ries, H. Report on the clays of Maryland. Vol. 4, pt. 3. Reports Maryland Geological Survey, pp. 203-507, 1902. Whittle C. L. The clays and clay industries of Massa- chusetts. Eng. and Min. Jour., vol. 66, pp. 245-246, 1898. Ries, H. Clays and shales of Michigan. Vol. 8, pt. 1. Reports Michigan Geological Survey, 67 pp. 1900. Berkey, C. P. Origin and distribution of Minnesota clays. Amer. Geologist, vol. 29, pp. 171-177. 1902. Eckel, E. C. Stoneware and brick clays of western Ten- nessee and northwestern Mississippi. Bulletin 213, U. S. Geological Survey, pp. 382-391. 1903. Wheeler, H. A. Clay deposits of Missouri. Vol. 2, Reports Missouri Geological Survey, 622 pp. 1896. Gould, C. N., and Fisher, C. A. The Dakota and Car- boniferous clays of Nebraska. Ann. Rep. for 1900, Nebraska Board of Agriculture, pp. 185-194. 1901. Cook, G. H., and Smock, J. C. Report on the clay deposits of New Jersey. New Jersey Geological Survey, 381 pp. 1878. 362 CEMENTS, LIMES, AND PLASTERS. NEW JERSEY. NEW YORK. NORTH CAROLINA. NORTH DAKOTA. OHIO. OREGON. PENNSYLVANIA. SOUTH CAROLINA. SOUTH DAKOTA. TENNESSEE. Ries, H., Ktimmel, B., and Knapp, G. N. The clays and clay industries of New Jersey. Vol. 6, Final Reports State Geologist, New Jersey. 8vo., 548 pp. 1904. Jones, C. C. A geologic and economic survey of the clay deposits of the lower Hudson River Valley. Trans. Am. Inst. Min. Engrs.,vol. 29, pp. 40-83. 1900. Ries, H. " Clays of Mew York. Bulletin 35, New York State Museum, 455 pp. 1900. Holmes, J. A. Notes on the kaolin and clay deposits of North Carolina. Trans. Am. Insf,. Min. Engrs., vol. 25, pp. 929-936. 1896. Ries, H. Clay deposits and clay industry in North Caro- lina. Bulletin 13, N. C. Geological Survey, 157 pp. 1897. Babcock, E. J. Clays of economic value in North Dakota. 1st Rep. N. D. Geological Survey, pp. 27-55. 1901. Orton, E. The clays of Ohio and the industries estab- lished upon them. Vol. 5, Reports Ohio Geological Survey, pp. 643-721. 1884. Orton, E. The clays of Ohio: their origin, composition, and varieties. Vol. 7, Reports Ohio Geological Sur- vey, pp. 45-68. 1893. Ries, H. The clay-working industries of the Pacific Coast States. Mines and Minerals, vol. 20, pp. 487- 488. 1900. Hopkins, T. C. Clays of western Pennsylvania. Appen- dix to Ann. Rep. Pennsylvania State College for 1897- 98, 184 pp. 1898. Hopkins, T. C. Clays of southeastern Pennsylvania. Appendix to Ann. Rep. Pennsylvania State College for 1898-99, 76 pp. 1899. Hopkins, T. C. Clays of the Great Valley and South Mountain areas. Appendix to Ann. Rep. Pennsylvania State College for 1899-1900, 45 pp. 1900. Woolsey, L. H. Clays of the Ohio Valley in Pennsyl- vania. Bulletin 225, U. S. Geological Survey, pp. 463- 480. 1904. Sloan, E. A preliminary report on the clays of South Carolina. Bulletin 1, South Carolina Geological Sur- vey, 171 pp. 1904. Todd, J. E. The clay and stone resources of South Dakota. Eng. and Min. Jour., vol. 66, pp. 371. 1898. Eckel, E. C. Stoneware and brick clays of western Tennessee and northwestern Mississippi. Bulletin 213, U. S. Geol. Survey, pp. 382-391. 1903. CLAYS, SHALES, AND SLATES. 363 TEXAS. Kennedy, W. Texas clays and their origin. Science, vol. 22, pp. 297-300. 1893. WASHINGTON. Landes, H. Clays of Washington. Vol. 1, Rep. Wash- ington Geol. Survey, pt. 2, pp. 13-23. 1902. WISCONSIN. Buckley, E. R. The clays and clay industries of Wis- consin Bulletin 7, Wisconsin Geol. Survey, 304 pp. 1901. WYOMING. Knight, W. C. The building stones and clays of Wyoming. Eng. and Min. Jour., vol. 66, pp. 546-547. 1898. Slates. Slate is, so far as origin is concerned, merely a form of shale in which a fine, even, and parallel cleavage has been developed by pressure. In composition, therefore, it will vary exactly as do the shales considered in the last section, and so far as composition alone is concerned, slate would not be worthy of more attention, as a Portland-cement mate- rial, than any other shale. Commercial considerations in connection with the slate industry, however, make slate a very important possible source of cement mate- rial. Good roofing slate is a relatively scarce material and commands a good price when found. In the preparation of roofing slate for the market so much material is lost during sawing, splitting, etc., that only about 10 to 25 per cent of the amount quarried is salable as slate. The remaining 75 to 90 per cent is of no service to the slate-miner. It is sent to the dump heap, and is a continual source of trouble and expense. This very material, however, as can be seen from the analyses quoted below, is often admirable for use in connection with limestone in a Portland-cement mixture. As it is a waste product, it could be obtained very cheaply by the cement manufacturer. Geographic distribution of slates. The principal areas in the United States in which roofing slate is at present quarried are briefly noted below. For more detailed information on the subject, reference should be made to the papers and reports listed on page 366. Beginning in the northeast, slates are extensively quarried in the Brownsville-Monson area in northern Maine, but no satisfactory lime- stones occur in this district The next important slate area lies in western Vermont and eastern New York, a region well supplied with good limestones. In New Jersey and Pennsylvania slates are worked just north of the Lehigh cement-rock belt, as noted in Chapter XXIV. The Peach Bottom slate district, located in southern Pennsylvania and northeastern Maryland, is also important, but is poorly supplied with 364 CEMENTS, LIMES, AND PLASTERS. limestone. Isolated slate districts occur in Virginia, but not near limestone areas. In eastern Tennessee and northwestern Georgia, however, roofing slates and non-magnesian limestones occur in close proximity; and in the Georgia slate district a Portland-cement plant is already in operation. West of the Mississippi, good slates are worked more or less extensively in - Minnesota, Arkansas, Utah, and California. Composition of slates. The composition of a large series of Ameri- can roofing slates from various localities is given in Table 164. TABLE 164 ANALYSES OF AMERICAN ROOFING SLATES. I. 2. 3. 4. 5. 6. 7. Silica (SiO ) 54 24 56 42 60 80 67 70 67 76 59 84 67 61 Alumina (A1 2 O.,) 24 71 24 14 22 00 13 49 14 12 15 02 13 20 Iron oxides (Fe 2 6 3 ,FeO) Lime (CaO) 8.39 5 23 4.46 52 10.50 50 2.75 81 5.52 63 5.96 2 20 6.56 11 Magnesia (MgO) 2 59 2 28 0.70 1.29 2.38 3 41 3 20 Alkalies 2.15 8 68 2.30 4.91 4.82 5.60 5 12 Water and CO 2 (?) 3 88 1 80 9 05 3 61 6 83 3 42 ' 8. 9. 10. 11. 12 13. 14. Silica (SiO 2 ) 56 49 68.62 55.88 58.37 62.71 60 65 58 20 Alumina (A1 2 O 3 ) Iron oxides (Fe 2 O 3 ,FeO) Lime (CaO) 11.59 4.90 5 11 12.68 4.20 2 34* 21.85 9.03 16 21.98 10.66 30 19.40 2.18 1 11 16.87 7.79 1 91 18.83 5.78 4 35 Magnesia (MgO) 6 43 3 76* 1 49 1 20 1 73 2 39 3 51 Alkalies 4.29 3 73 4 10 1 93 4 74 5 98 3 20 Water and CO 2 10 61 4 47 3 39 4 42 4 08 3 63 4 67 1. Monson, Maine. 2. 3. Lancaster, Mass. 4. Hamburg, N. Y. 5. West Pawlet, Vt. * Carbonate. 6. Pawlet, Vt. 7. Poulteney, N. Y. 8. Raceville, N. Y. 9. Bangor, Pa. 10. Peach Bottom, Pa.-Md. belt. 11. Peach Bottom, Pa.-Md. belt. 12. Martinsburg, W. Va. 13. Arvonia, Va. 14. Rockmart, Ga. TABLE 165. COMPOSITION OF AMERICAN ROOFING SLATES. Maximum. Average. Minimum. Silica (SiO,) 68.62 24.71 10.66 5.23 6.43 8.68 60.64 18.05 6.87 1.54 2.60 4.74 0.38 1.47 3.51 0.62 54.05 9.77 2.18 0.00 0.12 1.93 Alumina (A1 2 O 3 ) Iron oxides (FeO,Fe 2 O 3 ) Lime (CaO) Magnesia (MgO) Alkalies (K 2 O,Na 2 O) Ferrous sulphide (FeS 2 ) Carbon dioxide (CO 2 ) Water of combination. . .., Moisture, below 110 C CLAYS, SHALES, AND SLATES. 365 Slates used in cement-manufacture. Only one American Portland- cement plant is at present using roofing slate as one of its raw mate- rials, and this plant is of quite recent construction. It is that of the Southern States Portland Cement Company, and is located about half a mile east of the village of Rockmart, Polk County, Ga. The Port- land cement manufactured here is made from a mixture of pure lime- stone and slate, both of which materials occur in the immediate vicinity of the plant. Hard blue slates, which have been extensively quarried for struc- tural purposes, outcrop on the hills south of Rockmart. These slates are of Ordovician age and have been described as the "Rockmart slates" by Hayes. East of the town the surface rock is the "Chicka- mauga limestone," which here contains beds of pure non-magnesian limestone which have been quarried at several points in the vicinity and burned into lime. The cement company purchased the property of the old Georgia Slate Company, about half a mile southwest of Rockmart, and car- ried on extensive operations with the diamond-drill. The intention was to quarry the slate, sell as slate the portions best suited for that use, and utilize the scrap and waste in the manufacture of cement. The quarries from which the limestone is obtained are located half a mile east of town, near the mill. The president of the cement com- pany is Mr. W. F. Cowhan, who is also connected with the Peninsular Portland Cement Company, of Jackson, Mich., and the National Port- land Cement Company, of Durham, Ontario. TABLE 166. ANALYSES OF SLATE USED FOR PORTLAND CEMENT, ROCKMART, GA. 1. 2. Silica (SiO 2 ) 57 40 58.20 Alumina (A1 2 O 3 ) 23 65 18.83 Iron oxide (CaO) 4 45 5.78 Lime (CaO) 3 23 4 35 Magnesia (MgO) 3 23 3 51 Alkalies (K 2 O,Na 2 O) n d 3.20 Sulphur (S) n d 0.49 Carbon (C) n d 82 Carbon dioxide (CO 2 ) 0.60 Water > 6 . 80 \ 4 07 1. J. F. Davis, analyst. Privately communicated. 2. Slocum and Vandeventer, analysts. 18th Ann. Rep. U. S. Geol. Survey, pt. 5. 366 CEMENTS, LIMES, AND PLASTERS. References on slates. The following papers contain material of interest in connection with the composition, distribution, and structure of slate. Dale, T. N. The slate belt of eastern New York and western Vermont. 19th Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 153-307. 1899. Dale, T. N. The slate industry at Slatjiigton, Pa., and Martinsburg, W. Va. Bulletin 213, U. S. Geol. Survey; pp. 361-364. Dale, T. N. The slate deposits and slate industry of the United States. Bulletin , U. S. Geol. Survey. (In press.) Da vies, D. C. Slate and slate quarrying. 12mo, 181 pp. London, 1899. Eckel, E. C. Slate deposits of California and Utah. Bulletin 225, U. S. Geol. Survey, pp. 417-422. 1904. Eckel, E. C. The chemical composition of American shales and roofing slates. Journal of Geology, vol. 12, pp. 25-29. 1904. CHAPTER XXVIII. EXCAVATING THE RAW MATERIALS. THE excavation of the raw materials is the first step toward the actual manufacture of Portland cement, and the one concerning which least has been published. Local conditions enter into this preliminary phase to such an extent that few general statements can be made con- cerning it. To a large extent, each separate deposit of raw material is an individual proposition to be handled best in a way peculiar to itself. The natural raw materials which are at present used in the American Portland-cement industry are worked by one of three gen- eral methods. These are: (1) quarrying or digging from open pits; (2) mining from underground workings, and (3) dredging from deposits covered by water. Occasionally the cement manufacturer will have an opportunity to choose his general method of working the deposit, in which case his choice will depend partly on the physical characters of the material to be excavated and partly on the topographic and geologic conditions. Usually, however, he will have no opportunity to choose a method, for in any given case one of the methods will be so evidently the only possible mode of handling the material as to leave no room for other considerations. The three different general methods of excavation will first be briefly considered, after which the cost of excavating various raw materials will be discussed in some detail. Quarrying. In the following pages the term " quarrying " will be used to cover all methods of obtaining raw materials from open excavations quarries, cuts, or pits whether the material excavated.be a limestone, a shale, or a clay. Quarrying is the most natural and common method of excavating the raw materials for cement-manufacture. If marl, which is usually worked by dredging, be excluded from consideration, it is probably within safe limits to say that 95 per cent of the raw mate- rials used at American Portland-cement plants are obtained by quarry- 367 368 CEMENTS, LIMES, AND PLASTERS. ing. If marls be included, the percentages excavated by the different methods would probably be about as follows: quarrying, 88 per cent; dredging, 10 per cent; mining, 2 per cent. Stripping. When a limestone is being quarried, the opening should be so located as to give as little stripping as possible in proportion to the available rock; for in this case the stripping is merely dead work, adding greatly to the expense of the product. In dealing with a shale- bed, the "stripping" is usually merely weathered shale and can be used as well as the harder portions of the deposit. FIG. 71. Stripping a flat, shallow bed. A very thick bed of limestone, or a bed of moderate thickness lying almost horizontally, will not give as much stripping per ton of good rock as a thin bed or a bed dipping at a high angle. In handling com- paratively thin earth stripping in flat country, scrapers or excavators may be used (Fig. 71); while at one cement-plant a heavy soil cover on a quarry near a river-bank is removed by hydraulicking. This last process is also used in several large brick-plants. Quarry practice. In most of the quarries for cement rock or lime- stone, the rock is opened up on a low side-hill, so as to give a long work- ing-face with light stripping and as little grade as possible in the work- ings. The rock is blasted down in one or more benches, according to the height of face exposed, and the larger pieces are sledged or reblasted to manageable size by men placed along the working-face. It is then loaded either into one-horse carts or into small cars running on tem- porary tracks laid close up to the face. In the former case the carts are driven to a dump and loaded into cars; in the latter case the cars are drawn by horses or pushed by men to a turntable. This turntable EXCAVATING THE RAW MATERIALS. 369 FIG. 72. View of typical quarry. FIG. 73. Temporary tracks laid to face. 370 CEMENTS, LIMES, AND PLASTERS. is a comparatively fixed affair, located far enough away from the work- ing-face to avoid damage from blasting. Here the cars are attached to a cable or to a locomotive and hauled to the mill. FIG. 74. Cableway in cement-rock quarry. Occasionally an aerial cableway is used for transporting the mate- rial to the mill. This is shown hi the two views of a cement-rock quarry given in Figs. 74 and 75. Working in levels. In quarries containing several beds of rock differing greatly in composition and lying horizontally, tracks are often run in on different levels, so as to insure that each car or cart shall con- tain only one kind of rock. This practice is exemplified in the shale-pit shown in Fig. 76. A similar plan is followed in working the quarry partly shown in Fig. 77, which contains several heavy beds of limestone intercalated with workable but thinner layers of shale. Use of steam-shovels. In a few limestone and cement-rock quarries a steam-shovel is employed to load the blasted rock into the cars, and in shale quarries this use of steam-shovels is more frequent. In cer- tain clay- and shale-pits, where the material is of suitable character, the steam-shovel can be employed to do all the work, both excavating and loading the materials. EXCAVATING THE RAW MATERIALS. 371 FIG. 75. Cableway in cement-rock quarry. FIG. 76. Shale-pit worked in two levels. 372 CEMENTS, LIMES, AND PLASTERS. FIG. 77. Hoisting from quarry worked in levels. FIG. 78. Steam-shovel in cement-rock quarry. EXCAVATING THE RAW MATERIALS. 373 Steam-shovels are in use at the plant of the Purington Paving Brick Company, at Galesburg, 111. Here a bank of firm shale is drilled, shaken with black powder, and then handled entirely by steam-shovel. The following detailed figures of cost have been recently published by Mr. C. W. Purington. The figures cover the handling of 17,422 cubic yards of shale in one month of twenty-six nine-hour days. This shale was dug from a 50-foot bank with a Model 90 Barnhart shovel with 2-yard dipper. It was delivered to twenty 2-yard cars and trammed in two directions (1500 and 2000 feet respectively) to the bottoms of two inclines. It was then hoisted by cable to hoppers placed at an elevation of 20 feet above the track and dumped into the hoppers. TABLE 167. DETAILED COSTS OF STEAM-SHOVEL WORK. (PURINGTON.) Per Month. Per Yard, Cents. Labor Fuel - 1 engineer on shovel $110.00 85.00 52.65 128.85 80.00 46.80 93.60 1 craneman 1 fireman at 22 \ cents per hour 3 trackmen at 17^ cents per hour 1 engineer on locomotive 1 switchman at 20 cents per hour 2 hoistmen at 20 cents per hour Total labor $596.90 78.00 26.00 52.00 0.0343 1 \ tons coal per day at $2.00 per ton for shovel. . \ ton coal per day at $2.00 per ton for locomotive 1 ton coal per aay at $2.00 per ton for two hoists. Total fuel $156. CO 0.0089 Total costs, labor and fuel $752.90 0.0432 These figures do not include charges for superintendence, oil, waste, etc. If these be included, the cost of the steam-shovel work will be about 5 cents per cubic yard. If the cost of drilling and blasting be added, the total cost of handling the shale from the bank to the hop- pers may be about 6 cents per cubic yard. Of this the blasting, dig- ging, etc., amounts to about 4 cents per yard, while the tramming, hoist- ing, and dumping will amount to the remaining 2 cents. In handling shales, steam-shovels are usually very effective exca- vators, for here the material is physically homogeneous and requires only one light blasting to break it into fragments that can be readily handled by the shovel. This is well brought out by the itemized costs given above for the Galesburg shale quarry. 374 CEMENTS, LIMES, AND PLASTERS. When dealing with the Lehigh district cement rock, the steam- shovel is not quite so satisfactory, for much of the rock will require hand sledging after being blasted before it can be conveniently handled by the shovel. The hard limestones are still more intractable, and often require not only sledging but reblasting. This, of course, greatly decreases the financial effectiveness of Jjie shovel, and in many quarries will entirely prevent its use. In shallow quarries vertical seams filled FIG. 79. Steam-shovel handling limestone. with clay, soil, or other wash from the surface may greatly hinder the work of the shovel; and in quarries where rock is so mixed in compo- sition as to require sorting the shovel is worse than useless. Crushing and drying in the quarry. The rock is usually transported directly to the mill just as quarried, except that the larger masses are sledged to convenient size for handling. At a few quarries, however, a crushing-plant is installed at the quarry, and the rock is sent as crushed stone to the mill. Several of the quarries in question sell a certain portion of their product as road metal, which, of course, reduces the cost of the finer material which is sent to the cement-plant. A few plants have also installed their driers at the quarry and dry the stone before shipping it to the mill. This practice, shows a saving in mill space, but otherwise it seems to have little to recommend it. EXCAVATING THE RAW MATERIALS. 375 In the cases where a wet clay or shale is quarried some distance away from the mill the saving in transportation charges, due to dry- ing at the quarry, would, of course, be considerable. Mining. The term "mining" will be used, in distinction from "quarrying", to cover methods of obtaining any kind of raw material by underground workings, through shafts or tunnels. Mining is rarely employed in excavating materials of such low value per ton as the raw materials for Portland-cement manufacture. Occasionally, however, when a thin bed of limestone or shale is being worked, its dip will carry it under such a thickness of other strata as to make mining cheaper than stripping and quarrying for that particular case. Mining is considerably more expensive work than quarrying, but .there are a few advantages about it that serve to counterbalance the greater cost per ton of raw material. A mine can be worked steadily and economically in all kinds of weather, while an open cut or quarry is commonly in a more or less unworkable condition for about three months of the year. Material won by mining is, moreover, always dry and clean. Dredging. The term "dredging" will be here used to cover all methods of excavating soft, wet, raw materials. The fact that the materials are wet implies that the deposit occurs in a basin or depression, and this in turn implies that the mill is probably located at a higher elevation than the deposit of raw material, thus necessitating uphill transporta- tion to the mill. The only raw material for Portland-cement manufacture that is extensively worked by dredging in the United States is marl. Occa- sionally the clay used is obtained from deposits overlain by more or less water; but this is rarely done except where the marl and clay are interbedded or associated in the same deposit. A marl deposit, in addition to containing much water diffused through- out its mass, is usually covered by a more or less considerable depth of water. This will frequently require the partial draining of the basin in order to get tracks laid near enough to be of service. In dredging marl the excavator is frequently mounted on a barge which floats in a channel resulting from previous investigation. Occa- sionally, in deposits which either were originally covered by very little 376 CEMENTS, LIMES, AND PLASTERS. water or have been drained, the shovel is mounted on a car running on tracks laid along the edge of the deposit. The material brought up to the dredge may be transported to the mill in two different ways, the choice depending largely upon the manu- facturing processes in use at the plant. At plants using dome or chamber kilns, or where the marl is" to be drjfed before sending to the kiln, the excavated marl is usually loaded by the shovel on cars and hauled to the mill by horse or steam power. At normal marl-plants using a very wet mixture it is probable that the second ^method of transpor- tation is more economical. This consists of dumping the marl from the excavator into tanks, adding sufficient water to make it flow readily, and pumping the fluid mixture to the mill in pipes. FIG. 80. Dredge at marl-plant. Marl-pumping. The following description of the Harris system of pumping marl from the lake to the mill is taken from the catalogue of the Allis-Chalmers Company: "This method of handling marl by compressed air is now in great favor, and many plants have been and are now being installed in cement works in this country and Canada. "The essential features are given in the diagram herewith, and EXCAVATING THE RAW MATERIALS. 377 comprise the pump-tanks, air-compressor, automatic switch, and piping. There are no floats. There are no air-valves outside the engine-room. "The air is not allowed to escape after doing its work. The expan- sive force of the air is used in the compression cylinder of the com- pressor, thus giving back the greater part of its energy to the compressor. "Suppose the compressor to be in operation with switch set as in --^-^ Water SujjpOy FIG. 81. Harris system of marl-pumping. the figure, the air will be drawn out of the right-hand tank and forced into the left-hand tank, and in so doing will draw marl into the former and force it out of the latter. "The charge of air in the system is so adjusted that when one tank is emptied the other is just filled. At that moment the switch will reverse the pipe connections so that action in the tanks will be reversed. "The switch is adjustable and can be set or regulated while the pump is in operation." For further data, reference should be made to a paper by E. G. Harris, published in "Mines and Minerals," vol. 15, pp. 513-514. May, 1905. 378 CEMENTS, LIMES, AND PLASTERS. Cost of Raw Materials at Mill. The most natural way, perhaps, to express the cost of the raw materials delivered at the mill would be to state it as being so many cents per ton or cubic yard .of raw ^material ; and this is the method followed by quarrymen or miners in general. To the cement manufac- turer, however, such an estimate is not so suitable as one based on the cost of raw materials per ton or barrel of finished cement. Loss on drying, etc. In the case of hard and comparatively dry limestones or shales, it may be considered that the raw mixture loses 33 J per cent in weight on burning. Converting this relation into pounds of raw material and of clinker, we find that 600 pounds of dry raw material will make about 400 pounds of clinker. Allowing something for other losses in the process of manufacture, it is convenient and sufficiently accurate to estimate that 600 pounds of dry raw material will give one barrel of finished cement. These estimates must be increased if the raw materials carry any appreciable amount of water. Clays will fre- quently contain 15 per cent or more of water; while soft, chalky lime- stones, if quarried during weather, may carry as high as 15 to over 20 per cent. A Portland-cement mixture composed of a pure chalky lime- stone and a clay might, therefore, average 10 to 20 per cent of water, and consequently about 700 pounds of such a mixture would be required to make one barrel of finished cement. With marls the loss on drying and burning is much greater. Russell states * that according to determinations made by E. D. Camp- bell, 1 cubic foot of marl as it usually occurs in the natural deposits contains about 47J pounds of lime carbonate and 48 pounds of water. In making cement from a mixture of marl and clay, therefore, it would be necessary to figure on excavating and transporting over 1000 pounds of raw material for every barrel of finished cement. From the preceding notes it will be understood that the cost of raw materials at the mill per barrel of cement will vary not only with the cost of excavation, but with the kind of materials in use. Costs of quarrying or mining. In dealing with hard, dry materials extracted from open quarries near the mills, the cost of raw materials may vary between 8 cents and 15 cents per barrel of cement. The lower figure named is probably about the lowest attainable with good management and under favorable natural conditions; the higher figure * 22d Ann. Rep. U. S. Geol. Survey, pt. 3, p. 657. EXCAVATING THE RAW MATERIALS. 379 is probably a maximum for fairly careful management of a difficulv quarry under Eastern labor conditions. At one Portland-cement plant in the Middle States a 20-foot lime- stone-bed, overlain by 10 to 15 feet of shale and soil, is worked in open cut. When this work was handled by contract, 18 cents per cubic yard was paid for stripping, and 23 cents per ton for quarrying rock. Haulage to the mill, only a quarter of a mile away, cost about 4 cents per ton, owing to local difficulties. Quarry laborers here were worth $1.50 per day, and the following data on actual cost of quarrying were obtained from one of the contractors. The items cover labor and sup- plies for two men for two weeks, during which time the pair got out 300 tons of limestone. 10 Ibs. dynamite at $0. 14 per Ib $1 . 40 I box caps at $0 . 75 per box 0.19 1 keg powder at $1 . 50'per keg 1 . 50 100 feet fuse at $0.45 per 100 feet 0.45 Repairs, sharpening, etc . 60 24 days' labor at $1 . 50 per day 36.00 Cost of quarrying 300 tons limestone $40 . 14 Cost of quarrying, per ton limestone . 13 f " ' ' stripping, ' ' " " 0.07 " " hauling, " " " 0.04 Total cost limestone per ton . $0 . 24 When it is necessary to mine the materials, the cost will be some- what increased. Natural-cement rock has been mined at a cost equiv- alent to 10 cents per barrel of cement; but the figure is attained under particularly favorable conditions. The cost of mining and transpor- tation may reach from this figure up to 20 cents per barrel. Costs of marl-dredging. The costs of dredging and handling marl at several American cement-plants are given below : Plant 1. The marl- and clay-pits are about 3J miles by track from the mill. Both materials are covered by J to 1 foot of earth, but no water. A long cut is made into the deposit, into which cut cars are run on light tracks. These cars, containing about 3000 Ibs. of marl, are loaded by hand. The contract price for loading is 8 cents per car for marl and 14 cents per car for clay, and these prices are equiv- alent to a pay of $3.00 to $4.00 per day of 12 hours for each laborer loading. Two engines are used, one for switching and making up trains at the marl- and clay-pits, the other for hauling the trains to the mill. The total cost for sufficient marl and clay for 1200 barrels cement is: 380 CEMENTS, LIMES, AND PLASTERS. 2 engineers at $1 . 50 S3 . 00 2 firemen at $1 . 25 2 . 50 400 cars marl at $0 .08 32 .00 80 cars clay at $0. 14 11 .20 2000 Ibs. coal at $2 . 40 per ton 2 . 40 Total cost marl and clay for 1200 bbls. cement . . $51 . 10 Cost marl and clay for 1 bb cement .043 Plant %. The marl and clay occur in a swamp half a mile from the mill. The surface material is 2 to 3 feet black loam; this is underlain by 9 feet marl, and this, in turn, by the clay. A dredge with a lo- ll. P. engine and a crew of two men handles the marl, digging enough for 240 barrels cement in ten hours. A smaller dredge with orange- peel bucket, run by one man, handles the clay. One locomotive hauls the material to the plant over tracks laid alongside the excavations. Total costs per day for a 240-barrel plant are as follows : 1 marl-dredge runner $1 . 50 1 leverman, marl dredge 1 . 25 1 clay-dredge runner 1 . 25 1 locomotive engineer 1 . 50 350 Ibs. coal for marl dredge at $2 . 20 per ton . 38 $ 150 Ibs. coal for clay dredge at $2 . 20 per ton . 16 \ 500 Ibs. coal for locomotive at $2 . 20 per ton . 55 Cost of clay and marl for 240 barrels cement ... $6 . 50 Cost of clay and marl for 1 barrel cement 0.027 Plant S. Marl dredged from lake one quarter mile from mill. This is done by contract, the marl being delivered to the mill for 5^ cents per cubic yard. This price is about equivalent to $0.018 per barrel of cement for marl alone. In this case the dredging-plant was bought and installed at .the expense of the company, but the contractor pays all the current expenses, including pay, repairs, coal, etc. Plant 4- Marl dredged from lake one third mile from mill by a dredge operating a 1 ^-cubic-yard orange-peel bucket. The marl is fed through a stone separator and then pumped to the mill on Harris system. Total cost is about as follows: 2 men at $1.50 S3. 00 3 men at $1 . 25 3 . 75 2| tons coal at $2. 40 6.00 Total cost of marl for 500 barrels cement $12 . 75 Cost of marl for 1 barrel cement 0. 025 EXCAVATING THE RAW MATERIALS. 381 The costs above cited show the cheapness with which marl and clay can be excavated. The total costs of raw material at a wet-process plant may therefore range from 3 to 6 cents per barrel of cement. References on quarrying, etc. The following books and papers contain data of interest on methods and costs of quarrying, etc. Crane, W. A. Shale Mine at La Harpe, Kansas. Mines and Minerals, Dec., 1902, pp. 217-218. De Kalb, C. Manual of Explosives. 16mo, 126 pp. Ontario Bureau of Mines, 1900. Foster. C. Le Neol. Elements of Mining and Quarrying. 12mo, 321 pp. 1903. Gillette, H. P. Earthwork and Its Cost. 12mo, 244 pp. New York, 1903. Gillette, H. P. Rock Excavation: Methods and Cost. 12mo, 376 pp. New York, 1904. Green, N. M. [Cost of raw materials in cement manufacture.] Engineering Record, Jan. 23, 1904. Guttmann, O. Blasting. 8vo, 179 pp. London, 1892. Harris, E. G. The Harris System of Pumping with Compressed Air. Mines and Minerals, vol. 25, pp. 513-514, May, 1905. Knight, W. B. Quarrying Limestone at Rockland, Maine. Mines and Min- erals, August, 1899. CHAPTER XXIX. CALCULATION AND CONTROL OF THE MIX. I* IF, as in the present volume, we exclude from consideration the go-called "natural Portlands" (see page 215), Portland cement may be regarded as being entirely an artificial product, obtained by burning to semi -fusion an intimate mixture of pulverized materials, this mix- ture containing lime, silica, and alumina varying in proportion only within certain narrow limits, and by crushing finely the clinker resulting from this burning. If this restricted definition of Portland cement be accepted, four points may be regarded as being of cardinal importance in its manu- facture. These are: 1. The cement mixture must be of the proper chemical compo- sition. 2. The materials of which it is composed must be carefully ground and intimately mixed before burning in order to insure that chemical combination shall take place after calcination. 3. The burning must be conducted at the proper temperature, which varies considerably according to the chemical com- position of the mixture, and the length of time during which it is subjected to the burning process. 4. After burning, the resulting clinker must be finely ground. In this and the succeeding chapters these points will be taken up separately and in some detail. The present chapter deals with the calculations and arrangements necessary for insuring the correctness of the cement mixture. It, there- fore, includes discussions of the theoretical and practical considera- tions which determine the proportions of the mixture. Among these considerations are the theoretical composition and constitution of Port- land cement; the influence of various normal constituents on the prop- erties of the mixture; the influence of fuel ash and other accidental impurities; and the methods of calculating and controlling the mix in actual practice. 382 CALCULATION AND CONTROL OF THE MIX. 383 Theoretical Composition of Portland Cement. During recent years much attention has been paid by various inves- tigators to the constitution of Portland cement. The chemical com- position of any particular sample can, of course, be readily determined by analysis, and by comparison of a number of such analyses, general statements can be framed as to the range in composition of good Portland cements. This subject is discussed further in Chapter XXXVIII, where a large number of analyses are presented. Chemical analyses will determine what ingredients are present, and in what percentages, but other methods of investigation are necessary to ascertain in what manner these ingredients are combined. A summary of the more important practical results brought out by these investiga- tions on the constitution of Portland cement will be given in the present chapter, while in Chapter XXXVIII a more detailed discussion of the problem will be presented, as well as references to the principal papers on the subject. It would seem to be firmly established that in a well-burned Port- land cement much of the lime is combined with most of the silica to form the compound 3CaCO,Si02, tricalcic silicate. To this com- pound is ascribed, in large measure, the hydraulic properties of the cement; and in general it may be said that the value of a Portland cement increases directly as the proportion of 3CaO,SiO2- The ideal Port- land cement, toward which cements as actually made tend in compo- sition, would consist exclusively of tricalcic silicate, and would be there- fore composed entirely of lime and silica in the following proportions: Lime (CaO) 73.6 Silica (SiO 2 ) 26.4 Such an ideal cement, however, cannot be manufactured under present commercial conditions, for the heat required to clinker such a mixture cannot be attained in any working kiln. The oxyhydrogen blowpipe and the electrical furnace will give clinker of this composition; but a pure lime-silica Portland is not possible under present conditions as to burning and grinding on a commercial scale. In order to prepare Portland cement in actual practice, therefore, it is necessary that some other ingredient or ingredients should be pres- ent to serve as a flux in aiding the combination of the lime and silica, and such aid is afforded by the presence of alumina and iron oxide. Alumina (A^Os) and iron oxide (Fe 2 03) when present in notice- able percentages serve to reduce the temperature at which combina- tion of the lime and silica (to form 3CaO,Si02) takes place; and this clinkering temperature becomes further and further lowered as the 384 CEMENTS, LIMES, AND PLASTERS. percentages of alumina and iron are increased. The strength and value of the product, however, also decrease as the alumina and iron increase; so that in actual practice it is necessary to strike a balance between the advantage of ow clinkering temperature and the disadvantage of weak cement, and to thus determine how much alumina and iron should be used in the mixture. Tnis point ^will be further discussed on later pages. It is generally considered that whatever alumina is present in the cement is combined with part of the lime to form the compound 2CaO,Al 2 O3, dicalcic aluminate. It is also held by some, but this fact is somewhat less firmly established than the last, that the iron present is combined with the lime to form the compound 2CaO Fe 2 O 3 . This question of the action of the iron will be later referred to. For the purposes of the present chapter it will be sufficient to say that in the relatively small percentages in which iron occurs in Portland cement it may for convenience be considered as approximately equivalent to alumina in its action. Influence of Normal Constituents on the Cement. Lime, silica, alumina, iron oxide, magnesia, sulphur, and alkalies may be regarded as being normal constituents of any Portland-cement mixture. The three first named are necessary ingredients, while the last two, though undesirable, are rarely entirely absent from the raw materials used. The influence exerted by greater or lesser propor- tions of these seven constituents on the properties of both mixture and finished cement will be discussed in the present chapter. Maximum lime content of mixture. On pages 392-393 New- berry's method of proportioning cement mixtures will be described and exemplified. It should be borne in mind, however, that the New- berry formula there quoted will, if followed, give the maximum lime content that the mixture could bear, providing that the grinding, mixing, and calcination were performed with absolute perfection. As a matter of fact, however, the lime content of the mixture should never be car- ried quite as high as this formula would indicate, for in actual practice the mixing, grinding, and calcination are never theoretically perfect, and in consequence of a perfect combination of all the lime with all the silica and alumina cannot be attained. There will always remain a certain amount of uncombined material. If therefore, the lime in the mixture is carried as high as is theoretically allowable, a certain amount of free lime will occur in the cement. If, on the other hand, the mixture carries less than its proper theoretical percentage of lime, CALCULATION AND CONTROL OF THE MIX. 385 i the cement will, of course, contain some uncombined silica or alumina. A choice must be made, therefore, between the possibilities of having free lime in the product and having uncombined clayey matter. This choice is simple, for the effects on the value of the cement of these two possibilities are very different. Free lime is positively dangerous to the cement, while free clayey materials are merely inert, their only effect being to lower the tensile strength of the product. For this reason, since in practice it is necessary to choose between the two contingencies (free lime vs. free silica and alumina), the lime content of the mixture is always carried lower than theoretical considerations demand. It is to be further noted in this connection that the lime content of Portland cements relatively high in silica may be carried higher than in the case of the more aluminous Portlands. In discussing the constitution of Portland cement in preceding paragraphs it was stated that though lime combines with both silica and alumina, the combin- ing proportions are very different in the two cases. With silica, lime forms the tricalcic .silicate, whose percentage composition is lime 73.6 per cent, silica 26.4 per cent; the lime and silica are therefore com- bined in the proportion of lime 2.8 to silica 1. With alumina, lime forms a less basic compound, the dicalcic aluminate. The percentage composition of this compound is lime 52.3 per cent, alumina 47.7 per cent, corresponding about to the proportion lime 1.1 to alumina 1. It is evident, therefore, that a mixture containing 20 per cent silica and 5 per cent alumina can safely carry more lime than one contain- ing 15 per cent silica and 10 per cent alumina. Since the combination of lime, silica, and alumina becomes more thor- ough in proportion as the mixing, grinding, and burning are better done, higher lime contents can be carried by carefully prepared mixtures than by careless or coarsely ground mixtures ; and in rotary-kiln plants lime may be carried higher than in those using dome kilns. Up to the limit of safety every increase in the percentage of lime in the mixture will cause, other things being equal, an increase in the strength of the cement. This fact is taken advantage of, particularly when a new brand is being placed on the market. The usual method of procedure at such a time is to carry the lime very high, burn very hard, and pulverize very fine. This makes a costly but high-testing cement. As soon as the brand has become well established, the lime content can be dropped to reasonable working limits. Minimum lime content of mixture. The maximum lime content of the mixture is fixed by the considerations set forth in the preceding paragraphs. The minimum lime content, however, will also require 386 CEMENTS, LIME8, AND PLASTERS. some consideration. Low lime will invariably mean low-testing cements, and in the present state of the industry, low-testing cements are not easily marketed. A low-lime content is also the cause, in part, of the "dusting" of clinker in the vertical kiln. Le Chatelier found that the dicalcic silicate (2CaO,SiO2) possesses the property of spontaneously disintegrating on cooling. If the lim^ content of the mixture be carried too low, therefore, the clinker will fall to dust in the kiln, owing to the production of this unstable dicalcic silicate. Magnesia. The question as to the percentage,, of magnesia allow- able in a Portland cement has given rise to serious controversy for many years. In Europe the tendency has been to keep it below 3 per cent; but in this country, largely because of the results attained by Lehigh Valley cements above this limit, 4 or 5 per cent has been con- sidered the allowable maximum. All this discussion was carried on under the idea that magnesia was either inert or positively harmful in a Portland cement. Recent experiments by Prof. Newberry, however, have proven that an entirely satisfactory cement can be made carrying as high as 10 per cent of magnesia, if due care be given to the mixing and burning. This might have been expected, both on theoretical grounds and be- cause of the evidently active nature of magnesia in even the highest- burned natural cements, as pointed out on pages 198-200. At present it seems safe to say that magnesia can be considered equivalent to lime in its action, if due allowance be made for the difference in their com- bining weights. It is therefore theoretically possible to prepare a series of lime-magnesia Portlands, parallel to our present lime Portlands; and it is probable enough that in a few years some move will be made in this direction. But it must be borne in mind that a lime-magnesia Portland will probably differ in important respects from our present lime Portlands, and that it will therefore be inadvisable to group the two types of cement under the same general name. For this reason, in the present volume, the term Portland has been restricted by defini- tion to apply only to cements carrying less than 5 per cent of magnesia (MgO). Silica. It is commonly considered that the ultimate strength of the cement depends in large part upon the amount of calcium trisilicate it contains. Within certain limits, therefore, any increase in the per- centage of silica in the mixture will increase the strength of the cement. On the other hand, an increase in silica will * usually imply a decrease in alumina and iron oxide, and this in turn will cause the cement to be slow-setting (which is an advantage), but hard to clinker. CALCULATION AND CONTROL OF THE MIX. 387 Alumina. To the calcium aluminate of a cement are ascribed the initial setting properties. Decrease in the alumina, therefore, tends to make the cement slower setting, while high alumina affects it in the opposite way. Though it is advisable to carry the alumina as low as possible, so as to secure slowness of set and greater ultimate strength, it is impossible to carry it below a certain minimum, for alumina aids greatly in securing a low clinkering temperature, and a cement very low in alumina will clinker only with great difficulty. Too much alumina, on the other hand, will give a very fusible and sticky clinker, liable to ball in the kiln. Le Chatelier considers that the aluminous compounds present in Portland cement are the direct cause of its destruction by sea-water. His theory to account for this disintegration is as follows: Free lime, liberated during the hardening of the cement, reacts with the mag- nesium sulphate always present in sea-water, to form calcium sulphate. This in turn reacts with the calcium aluminate of the cement to form a sulphaluminate of lime, which swells considerably on hydration and thus disintegrates the cement mass. The extent of the disintegration varies directly with the percentage of alumina present in the cement. Cements containing 1 or 2 per cent of alumina are, for example, prac- tically unaffected by sea-water, while in cements containing as high as 7 or 8 per cent of alumina the swelling and consequent disintegration are very rapid. If the alumina of a cement be replaced by an oxide not reacting with calcium sulphate, the stability of the cement in sea-water is greatly improved. Le Chatelier has demonstrated this by preparing cements in which the alumina was replaced by oxides of iron, chromium, cobalt, etc. All of these were more resistant than an alumina cement to the disintegrating effect of lime sulphate. The best effects were obtained when iron oxide was used, a cement corresponding in composition to 5Si02,Fe 2 O3,17CaO being found to be not only stable in presence of sea-water but to possess excellent mechanical properties. DevaPs researches * on the effect of direct addition of calcium sul- phate to various cements confirm the above theory. Each of the finely ground cements tested was completely hydrated by mixing with 50 per cent of water and storing the mixture under water for three months out of contact with carbon dioxide. The mass was then dried, reground, mixed with half its weight of calcium sulphate and 33 per cent of water, and made up into rods which were kept moist and protected from car- * Abstract in Jour. Soc. Chem. Industry, vol. 21, pp. 971-972. 388 CEMENTS, LIMES, AND PLASTERS. bon dioxide by storage on moistened filter-paper under a glass bell. At the end of three weeks the increase in length of the rods was measured with the following results. TABLE 168. EFFECT OF ALUMINA. Type of Cement. Per Cent of Alumina in Cement. Per Cent of Elongation of the Rods. Slag cement (Vitrv) 15.5 * 27 " (Champignolles) Grappier cement (Besses) . . . 14.5 7 5 16 14 Portland cement 6 2 12 Hydraulic lime (Besses) 4 7 4 It will be noted that the percentage of elongation of the rods varied directly with the percentage of alumina in the cements tested, proving conclusively that the swelling was due to the action of the calcium sulph- aluminate formed during the operation. Iron oxide. Iron oxide, though usually so low as to be negligible in a Portland cement, occasionally is present in considerable percent- ages (4 to 6 per cent). When this is the case, it can only be con- sidered as equivalent to alumina in its action, allowing, of course, for their difference in combining weights. This conclusion is borne out by the" fact that Portland cements practically free from alumina have been made, containing lime, silica, and iron oxide only. Sulphur. Sulphur, when present in a cement mixture, may occur either as a sulphide or sulphate. In the former condition it is usually due to the occurrence of pyrite (iron disulphide, FeS 2 ) either in the lime- stone or in the clay. When present as a sulphate, it is usually in the form of gypsum (hydrous calcium sulphate, CaS04 + 2H 2 0). In the rotary kiln, which usually has an abundantly oxidizing flame, it is probable that any calcium sulphate present is dissociated (CaSC>4 =^CaO + S0 3 ) and the sulphur trioxide carried off, as this dissociation occurs at a temperature much lower than that reached in clinkering. If the flame is not sufficiently oxidizing, however, and because of imper- fect draft this condition is likely to occur in vertical kilns, any lime sulphate present will be reduced to the sulphide form. Alkalies. Small percentages of soda and potash are usually present in the mixture, due mostly to their presence in the clay or shale. Alka- lies have been regarded as detrimental, as inert, and as beneficial; and much discussion has taken place on the subject, based mostly on purely theoretical considerations. CALCULATION AND CONTROL OF THE MIX. 389 In experimenting with various methods for analyzing Portland cement, Hillebrand encountered the question of loss of alkalies during burning, which he discusses * as follows: "Long before the last of the sulphur trioxide is expelled alkali begins to volatilize, and it is easy to remove all or nearly all in this manner. The alkali is volatilized as oxide and may be collected in quantity on the under side of the crucible lid. At the intense temperature of the rotary-kiln furnace this action must play an important part, and to it is to be attributed the great loss of alkali noted by me in the cement of 1901, as compared with the raw mix from which it was made, an obser- vation which is repeated in the present case and must be general in cement-burning . ' ' Phosphorus. Phosphorus, combined with lime in the form of lime phosphate, frequently occurs in notable percentages in limestones, par- ticularly in the soft, chalky limestones and " marls" of the Southern States. In analyses this will be reported as phosphoric acid or phos- phorus pent oxide (^2^5), when it is determined at all. Few com- mercial analysts, however, would look for it in a cement material, and it is therefore rarely reported. Late in 1903 samples of a "marl" and clay from a Southern State were sent to a leading testing laboratory to obtain a decision on their value as cement materials. Three different burnings of cement were made from the raw materials in various mixtures, and the resulting cements gave the tests shown in Table 172, below. In addition to these generally poor results the chemists reported that the cement, for a week or so after setting, was so soft that it could be readity rubbed off by the hand. The various defects in the cements were ascribed by the laboratory experts to the presence in the marls of notable per- centages of phosphoric acid. The matter was referred to me by the Southern company, and at my request Prof. Clifford Richardson exam- ined microscopically several thin sections of the clinker which had been made in the laboratory tests. He reported that the raw mix had been very coarsely ground and the clinker underburned. The raw materials, as analyzed at the laboratory, showed the results given in Table 169. Two samples of marl were tested and one of clay. Of the three samples of cement made up from these materials and tested as below (Table 170), Cements A and B were made by mixing Marl 1 and clay in different proportions, while Cement C was made from a mixture of Marl 2 and the same clay. *Journ. Amer. Chem. Soc., vol. 25, p. 1200. 1903. 390 CEMENTS, LIMES, AND PLASTERS. TABLE 169. ANALYSES OF RAW MATERIALS CONTAINING PHOSPHORIC ACID. Marl 1. Marl 2. Clay. Silica (SiO 2 ) 9.02 3.88* 1 10 9.99 2.05 1.20 45.82 0.80 37.99 1.23 *. 38.96 22.60 5.82 16.44 0.32 16.02 Alumina (A1 2 O 3 ) Lime (CaO) : 45.78 0.75 38.87 Magnesia (MgO) Volatile (CO 2 , etc ) Phosphorus pentoxide (P 2 O 5 ) * Including about 1 per cent TABLE 170. TESTS OF CEMENTS CONTAINING PHOSPHORIC ACID. Cement A. Cement B. Cement C. Composition : Silica (SiO 2 ) 22 20 21 87 24 26 Alumina (A1 2 O 3 ) 10 23 f 6 84 7 97 Iron oxide (Fe 2 O 3 ) 2 64 2 60 3 22 Lime (CaO) 63 83 64 85 58 74 Magnesia (MgO) 1 11 1 30 1 21 Phosphorus pentoxide (P 2 O B ).. See A1 2 O 3 2.50 3.82 Per cent plaster added U% 2% 14% Fineness : Passing 50-mesh sieve 100.0 100.0 100 100- " " . . 96.3 98.8 94.0 < < 200- ' " " 76 80 71 Setting time initial 1 hr 10 min 1 hr 25 min 12 min final 5 hrs min 7 hrs 10 min 18 min Tensile strength: neat, 1 day 56 Ibs. 49 Ibs. 173 Ibs 1 ' 7 days 510 " 531 " 213 " " 28 " 754 " 754 " 340 " 1-3 7 " 180 " 166 " 72 " ' ' 28 " .... 327 " 280 " 80 " t Including about 2 per cent P2Os- Influence of intentionally added fluxes. At a number of plants working on materials or mixtures which are naturally difficult to fuse, experiments have been made on the reduction of the clinkering tem- perature by the addition of fluxing materials. Experiments of this kind are usually taken up in the early stages of the manufacturer's experi- ence. They rarely outlast the first year of actual practice, because he then begins to realize that it is difficult enough to secure a homo- geneous and uniform mixture of two ingredients without going to the extra trouble of adding a third material. Occasionally, however, the CALCULATION AND CONTROL OF THE MIX. 391 fluxing mania persists, and in a few rare cases it may be entirely justi- fiable. Fluorspar, sodium carbonate, and other alkali salts are the favorite materials for use as fluxes. It is certainly true that the addition of a very small percentage of some of these salts will decrease materially the difficulty of clinkering a cement mixture. Any other effect they may have on the cement, however, is either negatively or positively harmful; and in all cases their use can be avoided and equally good burning results obtained by a slightly increased fineness of grinding of the raw materials. The direct addition of iron oxide as a flux, a practice which is followed by at least one large American plant, is somewhat different from the use of fluorspar or alkalies. The iron oxide decreases the clinkering temperature very materially and gives a slower setting product than would an equal percentage of alumina. Adding it separately to the mixture is, however, a difficult matter to arrange. The more natural course to pursue would be to look for another source of clay supply, attempting to find a clay sufficiently high in iron to obviate the necessity for adding iron oxide separately. Calculating Mixtures of Untried Materials. When absolutely untried raw materials are being tested for the first time, the experimental mixture must be solely on the basis of their analyses, as developed in the formula given below or in some similar device. After the plant has once started, more empirical methods of calculating the mix are used, as set forth in a later section (pp. 393-394). Cementation Index. Recalling the discussion on page 383 of the theoretical constitution of Portland cement, it is evident that the ideal cement (and therefore the cement mixture) should contain its various ingredients in such percentages that the following compounds can be formed: 3CaO, Si0 2 , 2CaO.Al 2 O 3 , 2CaO.Fe 2 3 , 3MgO.SiO 2 , 2MgO.Al 2 O 3 , 2MgO.Fe 2 O3. These conditions are satisfied if the formula below, called for convenience the Cementation Index, gives a value of unity. In this formula the chemical equivalents above noted have been changed into percentages. (2. 8 X per cent age silica (Si0 2 ) ) + (1.1 X percentage alumina, A1 2 3 ) + (.7 X percentage iron oxide, Fe 2 O 3 ) (Percentage lime, CaO) + (l. 4 X percentage magnesia, MgO) When the value given by this formula falls below 1.0 the cement must necessarily contain free lime or free magnesia; when it rises above 1.0, the cement must necessarily be lower in lime than is theo- 392 CEMENTS, LIMES, AND PLASTERS. retically possible. The aim of the manufacturer, therefore, is to get a cement whose Cementation Index is on the safe side (i.e., over 1.0), but not too much so. Use of the formula in proportioning mixtures. The use cf a similar formula in calculating mixtures to be made from untried materials has been well described by Prof; Newberry. The discussion here presented differs from his only in the fact that -the magnesia and iron are allowed for, a correction which now seems necessary. Following this rule, the various steps in the proportioning of a cement mixture are given below in sufficient detail to be*' readily followed. OPERATION 1. Multiply the percentage of silica in the clayey ma- terial by 2.8, the percentage of alumina by 1.1, and the percentage of iron oxide by 0.7; add the products; subtract from the sum thus obtained the percentage of lime oxide in the clayey material plus 1.4 times the percentage of magnesia and call the result n. OPERATION 2. Multiply the percentage of silica in the calcareous material by 2.8, the percentage of alumina by 1.1, and the percentage of iron oxide by 0.7; add the products and subtract the sum from the percentage of lime oxide plus 1.4 times the percentage of magnesia in the calcareous material, calling the result m. OPERATION 3. Divide n by m. The quotient will be the number of parts of calcareous material required for one part of clayey material. Example. Assuming that materials of the following composition are in use the operation would be as follows: Clay. Limestone. Silica (SiO 2 ) 62.2 2.4 Alumina (A1 2 O 3 ) 16.1 2.0 Iron oxide (Fe 2 O 3 ) 4.2 0.3 Lime (CaO) 1.6 50.2 Magnesia (MgO) 1.2 1.5 Sulphur trioxide (SO 3 ) 1.7 0.6 Alkalies (K 2 O,Na 2 O) 0.8 0.4 Water, carbon dioxide, etc 12.2 42 .6 Operation (1). Clay. Silica X2. 8 = 62. 2X2. 8 = 174. 16 Alumina Xl.l = 16.1X1.1 = 17.71 Iron oxide X0.7= 4.2X0.7= 2.94 194.81 Lime Xl.0= 1.6X1.0= 1.6 Magnesia XL 4= 1.2X1.4= 1.68 3.28 194. 81-3. 28 = 191. 53 = n. CALCULATION AND CONTROL OF THE MIX. 393 Operation (2). Limestone. Silica X2.8 = 2.4 X2.8 = 6.72 Alumina Xl.l = 2.0X1.1= 2.20 Iron oxide X0.7= 0.3X0.7= 0.21 9.13 Lime Xl.O = 50.2Xl .0 = 50.2 Magnesia Xl.4= 1.5X1.4= 2.10 52.30 52. 30-9. 13 = 43. 17 =m. Operation (3). -j^- = 4 44= parts of limestone to be used for each part of clay, by weight. It must be recollected that the value given by the above formula represents the highest amount of lime theoretically possible under the best possible conditions of fine grinding and thorough burning. Even in the best-run plants these conditions cannot be attained in practice, and in a trial run either in a test kiln or in an actual plant it is foolish to attempt to reach this limit. The limestone shown by the formula should therefore be reduced in order to get safe results. A reduction of 10 per cent will probably be satisfactory. In the ex- ample given above this would work out as follows: 4.44= parts limestone (to 1 of clay) allowed by formula 0.44 = 10% reduction for safety 4.00 = parts limestone (to 1 of clay) to be actually used Calculating Mixtures in Current Work. After a plant has once gotten into good working order, and as long as the same raw materials are in use, the calculation of the mix becomes a much simpler affair. Two general methods are in use: At most plants the percentage of carbonates in the mix is made the criterion. If good results have been attained with mixtures carry- ing 78 to 80 per cent total carbonates (CaCOs + MgCOs), the aim of the chemist is simply to keep the mix within the'se limits. The calcu- lation in this case is simply a matter of arithmetic which does not re- quire explanation. The other method is to keep a fixed ratio between the total insoluble matter and the total carbonates. This ratio will naturally be different at each plant, but will always be fairly constant at any one plant. In a well-known and admirably managed marl-plant the marl is analyzed after being pumped into tanks at the mills, and the clay on 394 CEMENTS, LIMES, AND PLASTERS. its arrival at the mill. Four determinations are made on each sample of marl and three on the clay. These are: Marl. 1. Percentage of water; 2. Weight per cubic foot; 3. Percentage of insoluble matter; 4. Percentage of carbonates. Clay. 1. Percentage of water; ' 2. Percentage of insoluble matter; 3. Percentage of carbonates. From these determinations the mix is proportioned in such a way that the ratio Carbonates Insoluble matter shall fall within certain numerical limits. At the plant in question, which runs a high-testing cement which is also very high in silica, the above formula is made to give a value of 4.2. In the majority of plants it would fall about 3.0 to 3.4, Composition of mixture. The cement mixture ready for burning will commonly contain from 74 to 77.5 per cent of lime carbonate, or an equivalent proportion of lime oxide. Several analyses of actual cement mixtures are given in the following table. The ratio of silica to alumina plus iron for ordinary purposes should be about 3:1, for the cement becomes quicker setting and lower in ultimate strength as the percentage of alumina increases. If the alumina percentage be carried too high, moreover, the mixture will give a fusible, sticky clinker when burned, causing trouble in the kilns. TABLE 171. COMPOSITION OF ACTUAL, MIXES. Silica (SiO 2 ) 14.77 12 85 15.18 11.8 13.52 V , OK f 4.92 Iron oxide (Fe 2 O 3 ) 1 - 35 l 1 21 > 6.42 8.2 6.56 Lime (CaO) 43 03 42 76 42 97 41.8 42.07 1.74 1.02 n. d. 0.8 2.07 Carbon dioxide (CO 2 ) 35 61 34 71 n d n d 35 31 Water n d n d n d n d. n d Silica (SiO 2 ) 13 46 13 85 12.62 14.94 12. 9? n d 1 f 2.66 4.8c Iron oxide (Fe 2 O 3 ) n d J 7.20 6.00 t 1.10 1.77 Lime (CaO) 41 25 41 40 42 26 42.34 42.30 Magnesia (MgO) . . i . . . n d n d 2 67 2.21 2.08 Carbon dioxide (CO 2 ) 1 34 . 86 [36.10 35.68 35.49 Water 36.42 \n d n d n. d. CALCULATION AND CONTROL OF THE MIX. 395 Methods of control. The chemist having determined the standard of composition which he wishes to maintain in the mix, several different methods of maintaining this standard are possible. Theoretically, of course, the best of these methods is: (1) Both raw materials are analyzed as they arrive at the mill; the mix is made according to these analyses; after grinding the mix is analyzed as a check, and if seriously incorrect is corrected by the addition of the necessary ingredients. This method is actually prac- ticed at some plants, but in general one or the other of its two elements is gradually dropped out, so that most plants approach one of the two following extremes in practice. (2) The raw materials are analyzed, either by borings in the quarry or by an arrival at the mill, and the mix made in accordance with these analyses. The mix may be analyzed occasionally as a check, but no serious attempt is made to correct it. In this method the entire reli- ance is placed on the analyses of the raw materials. With hard, dry, raw materials varying little in composition the plan works well. In dealing with marls, etc., the third plan is most used. (3) The raw materials are mixed without analysis in approximately correct proportions, according to previous experience, and the ground mix is analyzed and brought up to proper composition (standardized) by the addition of whichever raw material proves to be deficient. In this method the correction of the mix is a regular part of the procedure. For convenience the mix is usually made always a little low in the same con- stituent, so that only one tank or bin of raw material needs to be kept on hand for standardizing. The following blank order shows how this is arranged in actual practice under the chemist's direction: CLAY ORDER. Date Tank No requires one hopper of clay for each inches of marl. Slurry tank No Changes in Composition During Manufacture. In theory the cement produced should correspond in composition to the mixture from which it is made. In practice it is found that, in 396 CEMENTS, LIMES, AND PLASTERS. addition to the expected loss of water, carbon dioxide, and other vola- tile components, the cement has suffered other changes which prevent it from having the exact composition calculated from the mixture. During the process of burning, the clinker has taken up a certain amount of material from the fuel ashes, the kiln linings, or the gases produced in the kiln. The changes in , composition thus caused will be briefly discussed. The change in composition during burning is almost inevitably in the direction of raising the Cementation Index of the cement, i.e., making it more clayey. This is due to the fact that the impurities picked up during burning are all of a clayey character, the kiln linings and the fuel ash being predominantly composed of silica and alumina. To partly counterbalance these additions of clayey matter, it is prob- able that the dust blown out of the kiln is more clayey than the rest of the mix; but this is not sufficient in amount to avail much against the combined influence of the fuel ash and the kiln lining. Of the two factors the fuel ash is by far the most important, because the kiln bricks are pretty steadily covered by a skin of clinker. The variation in composition of the ash derived from different types of fuel is shown by the following analyses made by Candlot.* TABLE 172. ANALYSES OF FUEL ASH. Composition of Ash from Anthracite. Gas Coke. SiO 2 . . 40.10 42.80 4.70 8.10 0.90 1.23 29.30 19.63 14.64 10.64 2.70 13.72 ALO, FeA ' CaO MgO SO 3 The differences between the calculated and actual compositions of a cement are well illustrated by the example given below. In this case a marl and clay of determined composition were mixed in a known ratio. The composition which a cement made from this mixture should show was calculated and is given in column 3, while the composition of the cement actually resulting is given in column 4. For these data the writer is indebted to Prof. S. B. Newberry ; who carried out the test in question. * Bonnami. Fabrication et Controle des Chaux Hydrauliques et des Ciments, p. 58. CALCULATION AND CONTROL OF THE MIX 397 TABLE 173. CHANGE IN COMPOSITION DURING BURNING. Raw Materials. Finished Product. Marl. Clay. Calculated. Actual Silica (SiO 2 ) 1.16 0.75 0.75 49.44 2.04 46.40 57.08 10.01 5.37 8.32 5.22 14.00 22.20 5.02 2.85 65.79 4.06 n. d. 0.974 22.42 5.68 3.22 62.24 3.22 n. d 1.068 Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) Lime (CaO) Magnesia (MgO) Loss (H 2 O,CO 2 , etc ) Ceme 1 tation Index This point is also illustrated by the following analyses of raw mix and cement from the Syracuse plant , analyzed by Hillebrand.* TABLE 174. CEMENT MIXTURE AND CEMENT, SANDUSKY. Mix. Cement. Silica (SiO 2 ) 13 51 21 93 Alumina (Al O 3 ) 3 32 5 68 Titanic oxide (TiO 2 ) 18 31 Iron oxide (Fe 2 O 3 ) 1 43 2 35 Lime (CaO) 40 84 62 92 Magnesia (MgO) 75 1 10 Potash (K 2 O) 79 61 Soda (Na 2 O) 22 29 Sulphur (S) 16 09 Sulphur trioxide (SO 3 ). . . . Carbon dioxide (CO 2 ) 1.43 n d 1.53 1 73 Water 4 20 1 40 Cementation Index 1.014 1 075 It will be seen that in both these experiments the Cementation Index of the cement has been raised considerably by the amount of silica and alumina taken up during calcination. * Jour. Amer. Chem. Soc., vol. 25, p. 1186. 1903. 396 CEMENTS, LIMES, AND PLASTERS. addition to the expected loss of water, carbon dioxide, and other vola- tile components, the cement has suffered other changes which prevent it from having the exact composition calculated from the mixture. During the process of burning, the clinker has taken up a certain amount of material from the fuel ashes, the kiln linings, or the gases produced in the kiln. The changes in . composition thus caused will be briefly discussed. The change in composition during burning is almost inevitably in the direction of raising the Cementation Index of the cement, i.e., making it more clayey. This is due to the fact mat the impurities picked up during burning are all of a clayey character, the kiln linings and the fuel ash being predominantly composed of silica and alumina. To partly counterbalance these additions of clayey matter, it is prob- able that the dust blown out of the kiln is more clayey than the rest of the mix; but this is not sufficient in amount to avail much against the combined influence of the fuel ash and the kiln lining. Of the two factors the fuel ash is by far the most important, because the kiln bricks are pretty steadily covered by a skin of clinker. The variation in composition of the ash derived from different types of fuel is shown by the following analyses made by Candlot.* TABLE 172. ANALYSES OF FUEL ASH. Composition of Ash from Anthracite. Gas Coke. SiO 2 40.10 42.80 4.70 8.10 0.90 1.23 29.30 19.63 14.64 10.64 2.70 13.72 ALO, FeLo* CaO MgO SO,. . The differences between the calculated and actual compositions of a cement are well illustrated by the example given below. In this case a marl and clay of determined composition were mixed in a known ratio. The composition which a cement made from this mixture should show was calculated and is given in column 3, while the composition of the cement actually resulting is given in column 4. For these data the writer is indebted to Prof. S. B. Newberry. who carried out the test in question. * Bonnami. Fabrication et Controle des Chaux Hydrauliques et des Ciments, p. 58. CALCULATION AND CONTROL OF THE MIX 397 TABLE 173. CHANGE IN COMPOSITION DURING BURNING. Raw Materials. Finished Product. Marl. Clay. Calculate!. Actual Silica (SiO 2 ) 1.16 0.75 0.75 49.44 2.04 46.40 57.08 10.01 5.37 8.32 5 22 14.00 22.20 5.02 2.85 65.79 4.06 n. d. 0.974 22.42 5.68 3.22 62.24 3.22 n. d 1.068 Alumina (A1 2 O 3 ) .... Iron oxide (Fe 2 O 3 ) Lime (CaO) Magnesia (MgO) Loss (H 2 O,CO 2 , etc.) Ceme 1 tation Index This point is also illustrated by the following analyses of raw mix and cement from the Syracuse plant, analyzed by Hillebrand.* TABLE 174. CEMENT MIXTURE AND CEMENT, SANDUSKY. Mix. Cement. Silica (SiO 2 ) 13.51 21 93 Alumina (Al O 3 ) 3 32 5 68 Titanic oxide (TiO 2 ) Iron oxide (Fe 2 O 3 ) 0.18 1 43 0.31 2 35 Lime (CaO) ... 40 84 62 92 Magnesia (MgO) . . . 75 1 10 Potash (K 2 O) . . . 79 61 Soda (Na 2 O) 22 29 Sulphur (S) 16 09 Sulphur trioxide (SO 3 ). . . . Carbon dioxide (CO 2 ) 1.43 n d 1.53 1 73 Water 4 20 1 40 Cementation Index 1 014 1 075 It will be seen that in both these experiments the Cementation Index of the cement has been raised considerably by the amount of silica and alumina taken up during calcination. * Jour. Amer. Chem. Soc., vol. 25, p. 1186. 1903. 400 CEMENTS, LIMES, AND PLASTERS. 30 per cent in fresh clays. The chemically combined water will depend largely on the composition of the clay ; and may vary from 5 to 12 per cent. The hygroscopic or mechanically held water of clays can be driven off at a temperature of 212 F. ; while the chemically combined water is lost only at a low red heat. The total water, therefore, to be driven off from clays may range frorn^G to 42 per cent, depending on the weather, the drainage of the clay-pit, and the care taken in pre- venting unnecessary exposure to moisture of the excavated clay. The average total amount of moisture will probably be about 15 per cent. In dealing with shales, the mechanically held water will rarely rise above 10 per cent, and can commonly be kept well below that limit. An additional 2 to 7 per cent of water will be carried by any shale in a state of chemical combination. At a few plants marl is used with clay in a dry process. As noted elsewhere the marls as excavated carry usually about 50 per cent of water. Marl presents a more difficult problem than do the other raw materials, because the vegetable matter usually present in marls is extremely retentive of water. It will be seen, therefore, that cement materials may carry from 1 per cent to 50 per cent of water when they reach the mill. The aver- age throughout the country would probably fall close to 5 per cent if the marls are excluded. In a dry process it is necessary to remove practically all of this water before commencing the grinding of the materials. One reason for this is that fine pulverizing cannot be eco- nomically or satisfactorily accomplished unless absolutely dry material is fed to the grinding machinery. Another reason, which is one of con- venience rather than of necessity, is that the presence of water in the raw materials complicates the control of the cement mixture. Methods and costs of drying. The type of dryer used at most cement- plants is a cylinder approximately 5 feet in diameter and 40 feet or so in length, set at a slight inclination to the horizontal and rotating on bearings. The wet raw material is fed in at the upper end of the cylinder, and it moves gradually toward the lower end, under the influence of gravity, as the cylinder revolves. In many dryers angle irons are bolted to the interior in such a way as to lift and drop the raw material alter- nately, thus exposing it more completely to the action of the heated gases and materially assisting in the drying process. The dried raw material falls from the lower end of the cylinder into an elevator boot and is then carried to the grinding-mills. The drying-cylinder is heated either by a separate furnace or by waste gases from the cement-kilns. In either case the products of PREPARING THE MIXTURE FOR THE KILN. 401 402 CEMENTS, LIMES, AND PLASTERS. combustion are introduced into the cylinder at its lower end, are drawn through it, and escape up a stack set at the upper end of the dryer. The dryer above described is the simplest and is most commonly used. For handling the small percentages of water contained in most cement materials it is very efficient, but for dealing with high percent- ages of water, such as are encountered* when marl is to be usd in a dry process, it seems probable that double-heating dryers will be found more economical. This type is exemplified by the Buggies-Coles . dryer, a detailed description of which is given in the section on slag cements, p. 649. In this dryer a double cylinder is employed. The wet raw material is fed into the space between the inner and outer cylinders, while the heated gases pass first through the inner cylinder and then, in a reverse direction, through the space between the inner and outer cylinders. This double-heating type of dryer is employed in almost all of the slag- cement plants in the United States, and is also in use in several Port- land-cement plants. When vertical kilns were in use, drying-floors and drying-tunnels were extensively used, but at present they can be found only in a few plants, being everywhere else supplanted by the rotary dryers. At the marl-plant of the ill-fated Hecla Portland Cement Company, which is shown in Fig. 82, rotary kilns were actually used as driers, because of the extreme difficulty encountered in properly drying this material in a drier of ordinary type. In the Edison plant a stationary vertical tower drier is used for the cement rock and limestone. The Edison stack drier shown in Fig. 83 is described as follows in a recent article * in the Iron Age: The chimney surmounting this flue is used only when starting a fire, the gases of combustion ordi- narily passing directly to the dryer stack to rise through the falling stream of rock and thoroughly dry it. The baffle-plate system is such that the fall of a piece of rock from the lowest screen to the bottom of the dryer requires 26 seconds. From above the baffles near the top of the stack the gases are drawn out by an 80-inch exhaust- fan, driven by a 50-horse-power motor, and are passed through a dust-settling chamber on their w r ay to the atmosphere. A 12- inch screw conveyor returns the collected dust to the bottom of the dryer stack and replaces it in the system. The baffle-plates of the * The Iron Age, Dec. 24, 1903, p. 5. PREPARING THE MIXTURE FOR THE KILN. 403 upper sections of the stack are arranged to slide longitudinally in their slots, reciproca- ting motion being provided by a motor- driven system of rocker arms sliding suc- cessive rows of plates in opposite directions at the rate of 20 cycles per minute. By this action clogging of possibly damp rock is pre- vented until it has fallen far enough to be dried sufficiently to have no such tendency. The shear-pin principle, used at this plant for driving the crushing rolls, is also applied in a modified form to the baffle-shakers. The rock-dryer is 8X8 feet in plan section, 40 feet high, and has a capacity of 3000 tons per day, the same as the crusher plant. The performance of the dryer stack is very efficient; the fuel consumption is small, the percentage of moisture in the crushed rock is reduced from 3 or 4 per cent to within 1 per cent, and the gases leave at a temperature scarcely above 212. A blower equipment is pro- vided for increasing the furnace draft when necessary. The cost of drying raw materials will depend on the cost of fuels, the percentage of water present in the wet material, and the efficiency of the dryer. Dryers are usually arranged and located so as to require little attention, and the labor costs of drying are therefore slight. Even under the most un- favorable conditions 5 Ibs. of water can be expected to be evaporated for each pound of coal used, while a good dryer will usually evaporate 7 or 8 Ibs. of water per pound of coal. Marls containing much organic matter are notably more reten- tive of moisture than any other raw material, and a marl-drying proposition is therefore apt to be expensive. For a f , . , . . iull description of a most elaborate and 10-C ia 83. Elevation of stack drier. Edison plant. (En- gineering NewsT) 404 CEMENTS, LIMES, AND PLASTERS. unsuccessful installation for marl-drying, reference should be made to the paper cited below.* Grinding and Mixing. Part at least of the reduction is usually accomplished before the materials are dried, but for convenience; the subjects have been separated in the present chapter. General methods. Usually the limestone or cement rock is passed through a crusher at the quarry or mill before being sent to the drier; and occasionally one or both of the raw materials is still further reduced before grinding, but the principal part of the grinding process always takes place after the material has been dried. After drying, the two raw materials may either be mixed imme- diately or each may be separately reduced before mixing. Automatic mixers, of which many slightly different types are in use, give a mix- ture in the proportions determined upon by the chemist. The further reduction of the mixture is usually carried on in two stages, the material being ground to, say, 30-mesh in a ball mill, kom- minuter, Griffin mill, etc , and finally reduced in a tube mill. At a few plants, however, single-stage reduction is practiced in Griffin or JBuntington mills, while at the Edison plant at Stewarts ville, N. J., the reduction is accomplished in a series of rolls. The majority of plants use either the Griffin mill and tube mill or the ball mill and tube mill; and there is probably little difference in the total cost of operating these two combinations. The former combination (Griffin + tube) is commonly considered to require less power, but more repairs than the latter; but even this can hardly be regarded as an established fact. The ball mill has never been quite as much of a success as its companion, the tube mill, and has been replaced at a number of plants by the kominuter. Plans of actual plants. Plans of several actual plants have been inserted for the purpose of illustrating the brief statement made above. The plant of the Lawrence Cement Company, of Siegfried, Pa., published by courtesy of Messrs. Lathbury and Spackman, is given in Fig. 85. The materials used here are cement rock and limestone. These are separately crushed in Gates crushers and dried in rotary driers, after which they are mixed and reduced in Williams mills and tube mills. * Plant and buildings of the Hecla Portland Cement and Coal Co. Engineering News, vol. 51, pp. 243-245. 1904. PREPARING THE MIXTURE FOR THE KILN. 405 In Fig. 84 the raw side of an ideal mill is presented, showing a very compactly installed layout of kominuters and tube mills for an output of 3500 barrels per day. FIG. 84. Installation of kominuters and tube mills. The plant of the Hudson P. C. Co., a typical modern dry-process plant, is shown in Fig. 86, reproduced by courtesy of Engineering News. 406 CEMENTS, LIMES, AND PLASTERS. In the article * accompanying this figure, the raw side of the mill is described as follows: "Following the course of the material step by step, it will be seen that the loaded cars from the quarry come into the mill at the east end at an elevation of 12 feet above the crusher-room floor, which is itself FIG. 85. Plan of plant, Lawrence Cement Co., Siegfried, Pa. (Lathbury and Spackman.) elevated 13J feet above the main mill floor, and that they dump through the track onto the crusher-room floor. Flush with this floor are the tops of three rotary crushers, two for crushing limestone and one for crushing shale. The two limestone crushers are run by a 45-H.P. electric motor and the shale crusher by a 22-H.P. electric motor. From the crushers the stone is delivered, shale and limestone separately, into four rotary driers, each of which is operated by a 5-H.P. electric motor. From the driers the stone passes separately to the ball mills for the first grinding. These ball mills are of the Krupp type, and there are five of them, each operated by a 50-H.P. electric motor. From the ball * Engineering News, vol. 50, pp. 70, 71. July 23, 1903. PREPARING THE MIXTURE FOR THE KILN. 407 CEMENTS,' LIMES, AXD PLASTERS. mills the shale powder is delivered to a set of two bins and the lime- stone powder to a set of five bins. These bins are so constructed as to discharge automatically into a double elevator, whence the materials ,are discharged into a double hopper over a tandem automatic weigh- ing-machine, which weighs out the proper proportion of each material. 'The two products are then mi^ed thoroughly by being conveyed together by elevator E and conveyors 9^ and 9 -to the steel bins feeding the tube mills. There are six of these tube mills and they are driven in groups by a 75-H.P. electric motor. " The tube-mill discharges feed onto a screw conveyor 10, thence to the elevator EE, and thence to screw conveyor 11, which discharges into two groups of stock bins. Screw conveyors 13 running under- neath these bins take the material right and left to the elevator F, which feeds the right and left screw conveyors 12 that discharge into the kiln feed bins. There are ten of these bins and each one feeds one rotary kiln." Actual Equipments of Dry-process Plants. The present-day practice in dry-process plants is shown better by the following data on the actual equipments of a number of these plants than by any amount of general statements on the subject. Reference should also be made to Chapter XXXI, where general crushing practice is discussed. Plant No. 1. Uses limestone and shale. Limestone Shale 1 Gates crusher 1 Gates crusher Plant No. Plant No. 3. 1 Mosser drier 1 kominuter 4 tube mills 4 kilns Limestone and shale. Limestone Shale 1 crusher 1 rotary drier 1 Williams mill 2 tube mills 2 kilns Limestone and shale. Limestone 1 crusher 1 rotarv drier Shale 1 tunnel drier 1 crusher 1 dry-pan to 30-mesh 2 Raymond pulverizers 3 kilns, 50 feet PREPARING THE MIXTURE FOR THE KILN. 409 Plant No. 4. Limestone and shale. Limestone Shale 2 Alton crushers 1 dry-pan to 8-mesh 1 Bonnot drier 1 rotary drier 10 sets rolls 2 tube mills 9 sets Sturtevant emery mills 4 intermittent tube mills 4 kilns Plant No. 5. Uses fairly hard limestone, with shale. Limestone Shale 1 Gates crusher, coarse rock sold; 1 disintegrator screenings used in cement plant 2 Bonnot rotary driers 2 Bonnot rotary driers 3 kominuters to 20-mesh 4 tube mills to 92% through 100-mesh 8 kilns Plant No. 6. Limestone and shale. Limestone Shale 1 Gates crusher 1 Gates crusher 1 rotary drier 3 ball mills 4 tube mills 6 kilns Plant No. 7. Uses hard limestone and shale. Limestone 2 Austin crushers 2 Bonnot driers 2 Krupp ball mills Shale 1 Sturtevant crusher 2 Bonnot driers 1 Bonnot ball mill Plant No. 8. 5 tube mills 10 kilns Limestone and shale. Limestone Shale 1 Gates crusher, No. 5 1 rotary drier 3 Williams mills 3 tube mills 6 kilns Plant No. 9. Limestone and shale. Limestone 2 crushers 2 rotary driers 4 ball mills 6 tube mills 8 kilns Shale 410 CEMENTS, LIMES, AND PLASTERS. Plant No. 10. Limestone and cement rock. Limestone Cement rock 2 crushers 2 rotary driers 6 ball mills 6 tube mills .10 kilns . 1 . v* Plant No. 11. Limestone and shale. Limestone Shale 1 crusher 1 disintegrator 1 rotary drier 1 rotary drier , 2 kominuters 2 Davidsen tube mills 6 kilns, 60 feet Plant No. 12. Uses marl and clay in a dry process. Marl Clay 1 rotary drier 1 rotary drier 2 tube mills 3 kilns List of references on dry-process plants and methods. The follow- ing papers describe plants using the dry process or details connected with that process. Haight, H. New works of the William Krause & Sons Cement Co., Martins Creek, Pa. Engineering Record, March 31, 1900. Cement Industry, pp. 107-116. 1900. Humphrey, R. L. Plant of the Buckhorn Portland Cement Co. [W. Va.]. Engineering News, vol. 50, pp. 408-414. Nov. 5, 1903. Lathbury, B. B., and Spackman, H. S. Portland Cement Co. of Utah [Salt Lake, Utah]. The Rotary Kiln, p. 127. 1902. Lathbury, B. B., and Spackman, H. S. Rotary plant of the American Cement Co., Egypt, Pa. The Rotary Kiln, pp. 66-71. 1902. Lathbury, B. B., and Spackman, H. S. Plant of the Lawrence Cement Co., Siegfried, Pa. Engineering Record, May 12, 1900. Cement Industry, pp. 117-131. 1900. The Rotary Kiln, pp. 96-109. 1902. Lathbury, B. B., and Spackman, H. S. Alsen's American Portland Cement Works, West Camp, N. Y. The Rotary Kiln, pp. 52-65. 1902. Lewis, F. H. The Vulcanite Portland Coment Company's Works, Vulcanite, N. J. Engineering Record, May 6, 1899. Cement Industry, pp. 96-106. 1900. Lewis, F. H. The Portland-cement plant of the Coplay Cement Co., Coplay, Pa. Engineering Record, Dec. 18, 1897. Cement Industry, pp. 20-32. 1900. Lewis, F. H. Mechanical equipment of a modern Portland-cement plant. Mineral Industry, vol. 11, pp. 88-119. 1903. PREPARING THE MIXTURE FOR THE KILN. 411 Meade, R. K. The plant of the Northampton Portland Cement Co., Pa. En- gineering Record, vol. 48, pp. 182, 183. Dec. 5, 1903. Meyer, H. C. New works of the Coplay Cement Co., Coplay, Pa. Engineering Record, Feb. 27, 1900. Cement Industry, pp. 69-77. 1900. Meyer, H. C. The works of the Nazareth Portland Cement Co., Nazareth, Pa. Engineering Record, Dec. 16, 1899. Cement Industry, pp. 85-95, 1900. Meyer, H. C. The Whitehall Portland Cement Works, Cementon, Pa. En- gineering Record, Sept. 15, 1900. Cement Industry, pp. 142-150. 1900. Stanger, W. H., and Blount, B. [The Atlas Portland- cement plant, Pa.] Proc. Inst. Civil Engineers, vol. 145, pp. 57-68. Vredenburgh, W. The Virginia Portland Cement Co. 'a Works, Craigsville, Va. Engineering Record, July 28, 1900. Cement Industry, pp. 132-141. 1900. Anon. The Almendares Portland Cement Works, Cuba. Engineering Record, vol. 49, pp. 36-38. Jan. 9, 1904. Anon. Edison Portland Cement Company [N. J.]. Iron Age, Dec. 24, 1903. Anon. The works of the Edison Portland Cement Co., near Stewartsville, N. J. Engineering Record, vol. 48, pp. 796-802. Dec. 26, 1903. Anon. The Edison Portland Cement Works at New Village, N. J. Engineer- ing News, vol. 50, pp. 555-559. 1903. Anon. Alsen's American Portland Cement Works, N. Y. Engineering Record, vol. 47, pp. 10-13. Jan. 3, 1903. Anon. Plant of the Hudson Portland Cement Co., at Hudson, N. Y. En- gineering News, vol. 50, pp. 70-71. July 23, 1903. (2) Methods Used with Slag-Limestone Mixtures. While the manufacture of Portland cement from a mixture of slag and limestone is similar in general theory and practice to its manu- facture from a limestone-clay or other dry raw materials, certain inter- esting differences occur in the preparation of the mixture. In the following paragraphs the general methods of preparing mixtures of slag and limestone for use in Portland-cement manufacture will first be noted, after which certain processes peculiar to the use of this par- ticular mixture will be described separately. General methods. After it had been determined that the puzzo- lan cement made * by mixing slag with lime without subsequent burning of the mixture was not an entirely satisfactory structural material, attention was soon directed toward the problem of making a true Portland cement from such slag. The blast-furnace slags com- monly available, while carrying enough silica and alumina for a cement mixture, are too low in lime to be suitable for Portland cement. Addi- *See Part VII. 412 CEMENTS, LIMES, AND PLASTERS. tional lime must be added, usually in the form of limestone, the slag and limestone must be well mixed and the mixture properly burned. The general methods for accomplishing the proper mixture of the mate- rials vary in details. It seems probable that the first method used in attempting to make a true Portland cement from slag was to dump the proper proportion of limestone, broken into small lumps, into molten slag. The idea was that both mixing and calcination could thus be accomplished in one stage; but in practice it was found that the result- ing cement was variable in composition and always l^ow in grade. This method has accordingly fallen into disuse, and at present three different general processes of preparing the mixture are practiced at different European and American plants. 1. The slag is granulated, dried, and ground, while the limestone is dried and ground separately. The two materials are then mixed in proper proportions, the mixture is finely pulverized in tube mills, and the product is fed in a powdered state to rotary kilns. 2. The slag is granulated, dried, and mixed with slightly less than the calculated proper amount of limestone, which has been previously dried and powdered* To this mixture is added sufficient powdered slaked lime (say 2 to 6 per cent) to bring the mixture up to correct composition. The intimate mixture and final reduction are then accom- plished in ball and tube mills. About 8 per cent of water is then added, and the slurry is made into bricks, which are dried and burned in a dome or chamber kiln. 3. Slag is granulated and mixed, while still wet, with crushed lime- stone in proper proportions. This mixture is run through a rotary calciner, heated by waste kiln gases, in which the temperature is suffi- cient not only to dry the mixture but also to partly powder it and to reduce most of the limestone to quicklime. The mixture is then pulverized and fed into rotary kilns. Of the three general processes above described the second is unsuited to American conditions. The first and third are adapted to the use of the rotary kiln. The third seems to be the most economical, and has given remarkably low fuel consumption in practice, but so far has not been taken up in the United States. Certain points of manufacture peculiar to the use of mixtures of slag and limestone will now be described. Composition of the slag. The slags available for use in Portland- cement manufacture are of quite common occurrence in iron-producing districts. Those best suited for such use are the more basic blast- furnace slags, and the higher such slags run in lime the more available PREPARING THE MIXTURE FOR THE KILN. 413' they are for this use. The slags' utilized will generally run from 30 to 40 per cent lime. The presence of over 3 per cent or so of magnesia in a slag is, of course, enough to render its use as a Portland-cement material inadvisable; and on this account slags from furnaces using dolomite (magnesian limestone) as a flux are unsuited for cement- manufacture. The presence of any notable percentage of sulphur is also a drawback, though, as will be later noted, part of the sulphur in the slag will be removed during the processes of manufacture. Granulation of slag. If slag be allowed to cool slowly, it solidifies into a dense, tough material, which is not readily reduced to the requisite fineness for a cement mixture. If it be cooled suddenly, however, as by bringing the stream of molten slag into contact with cold water, the slag is " granulated", Le., it breaks up into small porous particles. This granulated slag or "slag sand" is much more readily pulverized than a slowly cooled slag; its sudden cooling has also intensified the chemical activity of its constituents so as to give it hydraulic properties, while part of the sulphur contained in the original slag has been removed. The sole disadvantage of the process of granulating slag is that the product contains 20 to 40 per cent of water, which must be driven off before the granulated slag is sent to the grinding machinery. In practice the granulation of the slag is effected by directing the stream of molten slag direct from the furnace into a sheet-iron trough. A small stream of water flows along this trough, the quantity and rate of flow of the water being regulated so as to give complete granulation of the slag without using an excessive amount of water. The trough may be so directed as to discharge the granulated slag into tanks or into box cars, which are usually perforated at intervals along the sides so as to allow part of the water to drain off. Drying the slag. As above noted, the granulated slag may carry from 20 to 40 per cent of water. This is renewed by treating the slag in rotary driers. In practice such driers give an evaporation of 8 to 10 pounds of water per pound of coal. The practice of slag-dry- ing is very fully described in Part VII of this volume, pages 649-652, where figures and descriptions of various driers are also given, with data on their evaporative efficiency. As noted earlier in this article, one of the methods of manufacturing Portland cement from slag puts off the drying of the slag until after it has been mixed with the limestone, arid then accomplishes the drying by utilizing waste heat from the kilns. Kiln gases could, of course, be used anyway in the slag-driers, but it so happens that they have not been so used except, in plants following the method in question. 414 CEMENTS, LIMES, AND PLASTERS. Grinding the slag. Slag can be crushed with considerable ease to about 50-mesh, but notwithstanding its apparent brittleness it is diffi- cult to grind it finer. Until the introduction of the tube mill, in fact, it was almost impossible to reduce this material to the fineness neces- sary for a cement mixture, and the proper grinding of the slag is still an expensive part of the process, as^compared with the grinding of limestone, shales, or clay. Composition of the limestone. As the slag carries all the silica and alumina necessary for the cement mixture, the limestone to be added to it should be simply a pure lime carbonate. The limestone used for flux at the furnace which supplies the slag will usually be found to be of suitable composition for use in making up the cement mixture. Economics of using slag-limestone mixtures. The manufacture of a true Portland cement from a mixture of slag and limestone presents certain undoubted advantages over the use of any other raw materials, while it has also a few disadvantages. Probably the most prominent of the advantages lies in the fact that the most important raw material the slag can usually be ob- tained more cheaply, than an equal amount of natural raw material could be quarried or mined. The slag is a waste product, and a trouble- some material to dispose of, for which reason it is obtained at small expense to the cement-plant. Another advantage is due to the occur- rence of the lime in the slag as oxide, and not as carbonate. The heat necessary to drive off the carbon dioxide from an equivalent mass of limestone is therefore saved when slag forms part of the cement mixture^ and very low consumption is obtained when slag-limestone mixture is burned. Of the disadvantages, the toughness of the slag and the necessity for drying it before grinding are probably the most important. These serve to partly counterbalance the advantages noted above. A third difficulty, which is not always apparent at first, is that of securing a proper supply of suitable slag. Unless the cement-plant is closely connected in ownership with the furnaces from which its slag supply is to be obtained, this difficulty may become very serious. In a season when a good iron market exists the furnace manager will naturally give little thought to the question of supplying slag to an independent cement-plant. The advantages of the mixture, however, seem to outweigh its dis- advantages, for the manufacture of Portland cement from slag is now a large and growing industry in both Europe and America. Two Port- land-cement plants using slag and limestone as raw materials have been PREPARING THE MIXTURE FOR THE KILN. 415 established for some time in this country, several others are in course of construction at present, and it seems probable that in the near future Alabama will join Illinois and Pennsylvania as an important producer of Portland cement from slag. References on slag-limestone mixtures. (The more important articles are preceded by an asterisk.) Eckel, E. C. Preparation of slag limestone mixtures. Municipal Engineerhig, vol. 25, pp. 227-230. 1903. Hughes, 0. J. D. Portland cement from slag. U. S. Consular Reports, No. 1700, July 18, 1903. * Jantzen. Utilization of blast-furnace slag. Stahl und Eisen, vol. 23, pp. 361-375. 1902. Journ. Iron and Steel Inst., 1903, No. 1, pp. 634-637. Kammerer. Von ForelPs process for the production of Portland cement from basic slag. Stahl und Eisen, vol. 19, p. 1088. 1899. Journ. Soc. Chem. Industry, vol. 19, p. 48. * Lathbury, B. B., and Spackman, H. S. The Clinton Cement Company's plant, Pittsburg, Pa. The Rotary Kiln, pp. 82-85. 1902. May, E. Slag (Portland) cement. Stahl und Eisen, vol. 18, pp. 205-211. 1897. Journ. Iron and Steel Inst., 1898, No. 1, pp. 461-464. Schiele, F. Manufacture of Portland cement from slag at Lollar, Germany. Proc. Inst. Civ. Engrs., vol. 145, pp. 119-120. 1901. Steffens, C. Portland cement from slag in Germany. Stahl und Eisen, vol. 20, pp. 1170-1171. 1900. Journ. Iron and Steel Inst., 1901, No. 1, pp. 439- 440. Von Forell, C. Patent Portland cement from slag. Journ. Soc. Chem. Indus- try, vol. 19, p. 50. 1899. *Von Schwarz, C. The utilization of blast-furnace slag. Journ. Iron and Steel Inst., 1900, No. 1, pp. 141-152. Engineering News, Sept. 27, 1900. Engineering Record, June 2, 1900. * Von Schwarz, C. Portland cement manufactured from blast-furnace slag. Journ. Iron and Steel Inst., 1903, No. 1, pp. 203-230. (3) Blast-furnace Methods of Making Cement. Attempts have been made to manufacture Portland cement by mixing the raw materials without grinding and burning the mixture to a state of complete fusion in a kiln resembling a blast-furnace in de- sign and action. -The Hurry and Seaman patents covering a method of this type are described as follows:* Raw materials containing carbonate of lime, silica, and alumina are mixed with carbonaceous fuel, the combustion of which is supported * Journal Soc. Chem. Industry, vol. 21, p. 1079. 1902 416 CEMENTS, LIMES, AND PLASTERS. by a blast of air supplied through tuyeres, and a pressure about 10 to 20 Ibs. above that of the atmosphere is maintained in the furnace, whereby the materials are melted, the molten cement being afterward drawn off, cooled, and pulverized. The carbon dioxide derived from the carbonate of lime is reduced to carbonic oxide by the incandescent fuel, and in this atmosphere any oxide of iron in the raw materials is said to be reduced to metallic iron, which sinks and can ,thus be sepa- rated from the molten cement, whereby a superior product is obtained. The carbonate of lime may be preliminarily calcined and the carbon dioxide introduced together with air into the calcining furnace, where it is reduced and then again burned to carbon dioxide. The increased pres- sure is maintained either by arranging the height of the kiln so that the combustion gases formed in the lower part are prevented from escaping freely by the height of the mass of materials above or by a throttle- valve arranged in the outlet at the top of the kiln. Von Forell has taken out foreign patents on processes of quite similar type. (4) Wet Methods of Preparation. Wet methods of preparing Portland-cement mixtures date back to the time when millstones and similar crude grinding contrivances were in use. With such imperfect machinery it was almost impossible to grind dry materials fine enough to give a good Portland-cement mix- ture. The advent of good grinding machinery has practically driven out wet methods of manufacture in this country, except in dealing with materials such as marls, which naturally carry a large percentage of water. Two plants in the United States do, it is true, deliberately add water to a limestone-clay mixture; but the effect of this practice on the cost sheets of these remarkable plants is not encouraging. In preparing cement mixtures from marl and clay, a few plants dry both materials before mixing. It seems probable that this practice will spread, for the wet method of mixture is inherently expensive. At present, however, almost all marl-plants use wet methods of mixing, and it is therefore necessary to give some space to a discussion of such methods. Certain points regarding the location, physical condition, and chem- ical composition of the marls and clays used in such mixtures have impor- tant effects upon the cost of the wet process. As regards location con- sidered on a large scale, it must be borne in mind that marl deposits of workable size occur only in the Northern States and in Canada. In consequence the climate is unfavorable to continuous working through- out the year, for the marl is usually covered with water, and in winter PREPARING THE MIXTURE FOR THE KILN. 417 it is difficult to secure the material. In a minor sense location is still an important factor, for marl deposits necessarily and invariably are found in depressions; and the mill must, therefore, just as necessarily be located at a higher level than its source of raw material, which involves increased expense in transporting the raw material to the mill. Glacial clays, which are usually employed in connection with marl, commonly carry a much larger proportion of sand and pebbles than do the sedimentary clays of more southern regions. The effect of the water carried by the marl has been noted in an earlier paper. The material as excavated will consist approximately of equal weights of lime carbonate and of water. This on the face of it would seem to be bad enough as a business proposition; but we find that in practice more water is often added to permit the marl to be pumped up to the mill. On the arrival of the raw materials at the mill the clay is often dried, in order to simplify the calculation of the mixture. The reduction of the clay is commonly accomplished in a disintegrator or in edge-runner mills, after which the material is further reduced in a pug-mill, suffi- cient water being here added to enable it to be pumped readily. It is then ready for mixture with the marl, which at some point in its course has been screened to remove stones, wood, etc., so far as possible. The slurry is further ground in pug-mills or wet-grinding mills o/ the disk type, while the final reduction takes place commonly in wet-tube mills. The slurry, now containing 30 to 40 per cent of solid matter and 70 to 60 per cent of water, is pumped into storage-tanks, where it is kept in constant agitation to avoid settling. Analyses of the slurry are taken at this point, and the mixture in the tanks is corrected if found to be of unsatisfactory composition. After standardizing, the slurry is pumped into the rotary kilns. Owing to the large percentage of water contained in the slurry the fuel consumption per barrel of finished cement is 30 to 50 per cent greater and the output of each kiln corre- spondingly less than in the case of a dry mixture. This point will, how- ever, be further discussed in a later chapter. At a plant working a rather stiff slurry carrying only 40 per cent of water two Bonnot tube mills, using 25 H.P. each, handled together 4000 cubic feet of slurry in twenty-four hours, equivalent, to a production of 300 or more barrels of cement per day. This is; equivalent to a power consumption of a little less than 4 H.P. hours per barrel of finished cement. The clay contained in the slurry had been passed through a dry-pan and the slurry was then mixed and ground to some extent in a pug-mill. 418 CEMENTS, LIMES, AND PLASTERS. A Bonnot 22' X 5' tube mill used at a marl-plant on not very wet slurry ground about 20 to 30 barrels per hour, taking 30 H.P. in doing so. This slurry had not been previously treated except by passing it through a stone separator, so that the total power for grinding raw material at this plant was from 1 H.P. to 1J H.P. per barrel cement. At another plant a Bonnot 16-foot tube mill ground 12 barrels per hour raw wet mix, taking 15 to 20 'H.P. in doing so, the marl having previously been passed through a stone separator and pug-mill and the clay through dry-pans. Other plants report slightly different results with wet-tube mills. To sum up, from all data it seems that the preparation of a wet mix in tube mills will usually require from 1 H.P. hour to 2 H.P. hours per barrel cement. The power and product, however, will vary greatly with the percentage of water in the mix, as well as with the hardness of the particular marl and clay employed. The highest power consumption per barrel was shown by a plant which required 3.9 H.P. hours per barrel for preparing its raw mate- rials for the kiln. The marl used at this plant is unusually hard, and the mixture is made with less water than usual. This gives a fairly high kiln efficiency (100 barrels per day per kiln with a fuel consump- tion of 160 Ibs. coal per barrel), but it largely increases the work to be done by the grinding machinery on the raw side. Several of the grinding-mills used at this plant are, in addition, very inefficient types, and to this combination of unfavorable conditions is to be ascribed the high-power consumption on the raw side of the plant. All of the tanks containing slurry must be provided with some appli- ance for agitating the mixture or otherwise the heavier portion would settle at the bottom of the tanks, leaving fairly clear water above. Three different methods of agitating are in use at various marl-plants: 1. A vertical central shaft equipped with long arms or paddles; 2. A horizontal shaft crossing the tank a little above its bottom and fitted with screw blades; 3. The injection at intervals of jets of compressed air. Any of these three devices gives fairly good results, but none of them seems entirely satisfactory to the managers of the plants in which they are installed. The first two use an unexpectedly large amount of power. It may be of interest, for comparison with the above description of the wet process with rotary kilns, to insert a description of the semi- wet process as carried on a few years ago at the dome-kiln plant of the Empire Portland Cement Company of Warners, N. Y. The plant has been remodeled since that date, but the processes formerly followed PREPARING THE MIXTURE FOR THE KILN. 419 are still of interest, as they resulted in a high-grade, though expensive, product. At the Empire plant the marl and clay are obtained from a swamp about three fourths of a mile from the mill. A revolving derrick with clam-shell bucket was employed for excavating the marl, while the clay was dug with shovels. The materials are taken to the works over a private narrow-gauge road, on cars carrying about three tons each, drawn by a small locomotive. At the mill the cars were hauled up an inclined track, by means of a cable and drum, to the mixing floor. The clay was dried in three Cummer "salamander" driers, after which it was allowed to cool, and then carried to the mills. These mills were of the Sturtevant " rock-emery " type, and reduced the clay to a fine powder, in which condition it was fed, after being weighed, to the mixer. The marl was weighed and sent directly to the mixer, no pre- liminary treatment being necessary. The average charge was about 25 per cent clay and about 75 per cent marl. The mixing was carried on in a mixing pan 12 feet in diameter, in which two large rolls, each about 5 feet in diameter and 16-inch face, ground and mixed the materials thoroughly. The mixture was then sampled and analyzed, after which it was carried by a belt conveyor or two pug-mills, where the mixing was completed and the slurry formed into slabs about 3 feet long and 4 to 5 inches in width and height. These on issuing from the pug-mill were cut into a number of sections, so as to give bricks about 6 inches by 4 inches by 4 inches in size. The bricks were then placed on slats, which were loaded on rack cars and run into the drying tunnels. The tunnels were heated by waste gases from the kilns and required from twenty-four to thirty-six hours to dry the bricks. After drying, the bricks were fed into dome kilns, twenty of which were in use, being charged with alternate layers of coke and slurry bricks. The coke charge for a kiln was about four or five tons, and this produced 20 to 26 tons of clinker at each burning, thus giving a fuel consumption of about 20 per cent, as compared with the 40 per cent or so required in the rotary kilns using wet materials. From thirty-six to forty hours were required for burning the charge. After cooling, the clinker was shoveled out, picked over by hand, and reduced in a Blake crusher, Smidth ball mills, and Davidsen tube mills. 420 CEMENTS, LIMES, AND PLASTERS. Actual Equipment of Wet-process Plants. Plant No. 1. Uses a hard limestone and a clay in a wet process. Limestone Clay 2 crushers 4'ballmilig 4 wet paddle mills 4 wet tube mills 10 kilns Plant No. 2. Uses a fairly hard limestone and a snale in a wet process. Limestone Shale 1 Gates crusher to 3^ inches 2 Williams mills 3 small Gates crushers to J inch 3 rotary driers 21 Griffin mills to 90% through 100-mesh 3 pug-mills 15 slurry tanks 21 kilns Plant No. 3. Marl and clay. Marl Clay 1 rotary drier 1 dry-pan 1 pug-mill, 40% water 2 Bonnot tube mills 3 kilns, 60 feet Plant No. 4- Marl and clay. Marl Clay 1 stone separator 1 dry-pan 2 pug-mills 4 Bonnot tube mills 5 kilns Plant No. 5. Marl and clay. Marl Clay 1 stone separator Tanks 1 wet tube mill Tanks 3 kilns Plant No. 6. Marl and clay. Marl Clay 2 wet-pans 3 Abbie mills 2 wet ball mills j 3 wet tube mills 13 kilns PREPARING THE MIXTURE FOR THE KILN. 421 Plant No. 7. Marl and shale. Marl Shale 1 stone separator 2 dry-pans 1 pug-mill 6 Bonnot tube mills, 16 feet 14 kilns Plant No. 8. Marl and day. Marl Clay 1 stone separator. 1 pug-mili 1 pug-mill 3 tube mills 6 kilns Plant No. 9. Marl and clay. Marl Clay 1 stone separator | 2 wet-pans 4 tube mills 14 kilns Plant No. 10. Marl and clay. Marl Clay 1 stone separator 1 rotary drier 1 pug-mill 4 tube mills 6 kilns, 70 feet 4 kilns, 60 feet Plant No. 11. Marl and shale. Marl Shale 1 stone separator 1 rotary drier 1 Williams mill 1 pug-mill 2 tube mills 9 kilns, 70 feet List of references on wet-process plants and methods. The following papers, mostly descriptive of individual wet-process plants, contain sufficiently detailed data on methods and machinery to be worth refer- ring to as sources of further information. For convenience of reference, the plants have been separately named, though at the cost of some space. Grimsley, G. P. A new Portland-cement mill in the gas fields of Kansas [lola]. Engineering and Mining Journal, Feb. 16, 1901. Lathbury, B. B. The Michigan Alkali Company's plant for manufacturing Portland cement from caustic-soda waste. Engineering News, June 7, 1900. 422 CEMENTS, LIMES, AND PLASTERS. Lathbury, B. B., and Spackman, H. S. The Michigan Alkali Co.'s plant, Wyan- dotte, Mich. The Rotary Kiln, pp. 110-119. 1902. Lathbury, B. B., and Spackman, H. S. Detroit Portland Cement Co., Fenton, Mich. The Rotary Kiln, pp. 86-95. 1902. Lathbury, B. B., and Spackman, H. S. Wabash Portland Cement Co., Stroh, Ind. The Rotary Kiln,, pp. 128-133. 1902. Lathbury, B. B., and Spackman, H. S. Alma Portland Cement Co., Welleston, Ohio. The Rotary Kiln, pp. 44-51. 1902. Lathbury, B. B., and Spackman, H. S. Castalia Portland Cement Co., Cas- talia, Ohio. The Rotary Kiln, pp. 78-81. 1902. . Lathbury, B. B., and Spackman, H. S. Beaver Portland Cement Co., Marl- bank, Ontario. The Rotary Kiln, pp. 74-77. 1902. Lathbury, B. B., and Spackman, H. S. Plant of the Aalborg Portland Cement Fabrik, Aalborg, Denmark. The Rotary Kiln, pp. 42-43. 1902. Lewis, F. H. The plant of the Bronson Portland Cement Co., Bronson, Mich. Engineering Record, April 30, 1898. Cement Industry, pp. 33-44. 1900. Lewis, F. H. The plant of the Michigan Portland Cement Co., Coldwater, Mich. Engineering Record, Feb. 25, 1899. Cement Industry, pp. 78-84. 1900. Lewis, F. H. The Empire Portland-cement plant, Warners, N. Y. Engineer- ing Record, July 16, 1898. Cement Industry, pp. 45-51. 1900. Lewis, F. H. The Buckeye Portland-cement plant, near Bellefontaine, Ohio. Engineering Record, Oct. 15, 1898. Cement Industry, pp. 52-59. 1900. Lewis, F. H. Western Portland Cement Co.'s plant, Yankton, S. D. Engi- neering Record, Nov. 19, 1898. Cement Industry, pp. 60-68. 1900. Russell, I. G. The Portland-cement industry in Michigan. 19th Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 629-685. 1902. Simmons, W. H. [Plant at Bronson, Michigan.] Proc. Inst. Civ. Engrs., vol. 145, pp. 120-121. 1901. Stanger, W. H., and Blount, B. [Plant at Bronson, Mich.] Proc. Inst. Civ. Engrs., vol. 145, pp. 68-69. 1901. Anon. Plant and buildings of the Hecla Portland Cement and Coal Co., Mich. Engineering News, vol. 51, pp. 243-245. 1904. General Crushing Practice. The crushing and grinding practice in cement-making is so important a part of the manufacture, and so closely connected with the financial success or failure of the plant, that it requires special consideration, both as to general methods and in regard to the special types of machinery employed. In the present section certain features will be discussed which are common to both the wet and dry methods of preparation, while the following chapter will be devoted to a description of the various standard types of crushing and grinding machinery. PREPARING THE MIXTURE FOR THE KILN. 423 Necessity for fine grinding. The necessity for very fine grinding of the raw mixture, if a sound and volume-constant cement is to be obtained, was early stated by Newberry,* and the value of such fine grinding has been recently expressed in the quantitative form by the experiments of Professor Campbell, f To secure a sound and volume-constant cement it is necessary that the raw mixture be very finely ground. Other things being equal, the finer the grinding of the raw mixture the better will be the resulting cement. The degree of fineness necessary to secure a given grade of cement will depend upon: (a) The percentage of lime in the mixture. The higher the percentage of lime in the mixture, the finer the raw mixture must be ground, because the chances of getting an unsound or expensive cement will increase as the percentage of lime rises, and this tendency will have to be coun- teracted by greater fineness of grinding. (6) The carefulness with which the materials have been mixed* The more careful and thorough the mixture, the less care need be be- stowed upon the grinding, and vice versa. (c) The character of the raw materials. This point, which has been emphasized by Newberry,* is of great importance. When a very pure limestone or marl is mixed with a clay or shale, the grinding must be much finer than in the plants (such as those in the Lehigh district of Pennsylvania) where a highly argillaceous limestone (" ce- ment rock") is mixed with a comparatively small quantity of purer limestone. In the latter case the coarser particles of the argillaceous limestone will be so near in chemical composition to the proper mixture as to do little harm to the resulting cement, even if both the grinding and the mixing should be incompletely accomplished, while in the former case, where a pure limestone or marl is mixed with clay or shale, both of the constituents are very different in composition from the proper mixture, and coarse particles will therefore be highly injurious to the cement. (d) The duration of the burning. In the old-fashioned dome kilns, where the mixture was exposed to the action of heat for a week or more, the duration of the burning compensated in some degree for the lack of thoroughness in grinding or mixing. In modern rotary kilns, how- ever, in which the mixture is burned for only an hour or so, this aid can- not be counted on, and both grinding and mixing must therefore be done more carefully. * 20th Ann. Rept. U. S Geol. Survey, pt. 6, p. 545. f Journ. Am. Chem. Soc., vol. 25, p. 40 et seq. 424 CEMENTS, LIMES, AND PLASTERS. Actual fineness attained. After its final reduction, and when ready for burning, the mixture will usually run from 90 to 95 per cent through .a 100-mesh sieve. In the plants of the Lehigh district the mixture is rarely crushed as fine as when limestone and clay are used. New- berry * has pointed out in explanation for this that an argillaceous limestone (cement rock) mixed with # comparatively small quantity of purer limestone, as in the Lehigh plants, requires less thorough mix- ing and less fine grinding than when a mixture of limestone and clay (or marl and clay) is used, for even the coarser particles of the argilla- ceous limestone will vary so little in chemical composition from the proper mixture as to affect the quality of the resulting cement but little should either mixing or grinding be incompletely accomplished. A very good example of typical Lehigh Valley grinding of raw mate- rial is afforded by a specimen examined f by Prof. E. D. Campbell. This .specimen of raw mix ready for burning was furnished by one of the best of the eastern Pennsylvania cement-plants. A mechanical analysis of it showed the following results. Mesh of sieve 50 100 200 Per cent passing 96.9 85.6 72.4 Per cent residue 3.1 14.4 27.6 The material, therefore, is so coarsely ground that only a trifle over -85 per cent passes a 100-mesh sieve. Bleininger has recently published the results of a series of tests for fineness, made on the raw mixtures used by various plants. The results are given in Table 175. Gradual vs. one-stage reduction. This question is now of little more than theoretical interest, as almost all cement-plants seem to Tiave given the same decision in regard to it Until within the past few years, examination of a number of cement-plants would have developed the fact that two radically different systems of reduction were in use. Reference is made to the "gradual" and "one-stage" methods. In those earlier days many plants, after preliminary treatment in a coarse crusher, completed the entire process of further reduction in a Griffin mill or ball mill, for it must be recollected that the latter mill was introduced earlier than its companion, the tube mill, and was at first expected to do the work now done by both. I * 20th Ann. Rep. U. S. Geol. Survey, pt. 6, p. 545. t Jour. Amer. Chem. Soc., vol. 25, p. 39. PREPARING THE MIXTURE FOR THE KILN. 425 TABLE 175. FINENESS OF RAW Mix AT VARIOUS PLANTS. (BLEININGER.) J > *jf si I-H 0> eir75 tt PI PS 111 I 1 Raw Materials. *i J3 1 02 0,^3 y c 8 S OOQ N OJGO-* ^ GOO 2sl 3=3 1 4) J2 T5g T3S ^ u c -5 cj * goo fe s | o IS 1 9 Cd 1 rt i S r .2"' (2) sa & Mills. 1 Limestone and shale Dry Emery 2. CO 9.89 4.85 16.93 8.96 8.70 48.68 2 Cement rock and limestone Dry Griffin 16.38 11.57 4 75 16.21 13.29 7.61 30 17 Cement rock and limestone Dry Tube 3 03 7 42 3 68 23 77 17 52 10 26 34 33 4 Limestone and clay Wet 7.40 9.56 2.48 17.72 8.96 8.83 45.05 5 Marl and clay Wet 3 04 5 50 5 M 20.31 12 63 9 61 44 72 6 tt it < Wet 30 46 4 23 ? 17 6.73 10.61 9 31 47 06 7 1C ( { (I Wet 2 48 5 23 ? 47 16.14 14.22 12 01 47 46 S ( C (I t C Wet 26.74 6.99 2.13 10.52 9.77 7.67 36.18 At present , however, the " one-stage " type of fine-reduction prac- tice has fallen into disuse. It has been gradually modified out of exist- ence, so that cement-mills now are all using a gradual reduction system. The change from " one-stage" to "gradual" has been effected in two opposite ways, according to whether Griffin or ball mills were in use at the original plant. In plants using the Griffin (or Huntingdon) mill, it has finally become the general practice to place either a fine gyratory crusher or a set of fairly fine rolls between the coarse crusher and the Griffin mill, thus insuring that the material fed to the Griffin mill shall not be over J or even J inch in size. In plants using ball mills, the change has taken place at the latter end of the series by put- ting in a tube mill after the ball mill. The ball mill, instead of taking a coarse product from the crusher and reducing this product to its ulti- mate fineness, has been made an intermediate reducer, the final reduc- tion taking place in the tube mill. Present-day systems. Present-day practice in the reduction of dry raw materials may therefore be summarized as in the following two schedules. The two schemes differ from each other mainly in the mills used. TYPE A. GRIFFIN (OR HUNTINGDON) MILL TYPE. 1. Gyratory crusher to about 1\ inches 2. Crusher or rolls " " i to inch 3. Griffin or Huntingdon mill " " 85 to 95% through 100-mesh TYPE B. BALL (OR WILLIAMS) MILL TYPE. 1. Gyratory crusher to about 2| inches 2. Kominuter, ball mill or Williams mill ' ' " 20- to 30-mesh 3. Tube mill " " 85 to 95% through 100-mesh 426 CEMENTS, LIMES, AND PLASTERS. The omission of separators in cement-plants. Practically every writer who has discussed the crushing practice at American Portland- cement plants has noted and deplored the absence of separators. In spite of this general unity of opinion in the subject, only a few plants, to the writer's knowledge, are now equipped with any separators, and none have as many as might "be used. **In view of the fact that Ameri- can Portland-cement practice, so far as crushing methods are concerned, is to-day far ahead of crushing practice in ore-treatment works, this apparent disregard of one great principle of general, crushing practice seems to require some explanation. It will not do merely to assume- that separators are omitted because the advantages to be gained from their use are not understood by designers and managers of cement- plants, for that amounts to charging a peculiarly expert clan of mechan- ical experts with gross ignorance. In the opinion of the writer, the gen- eral omission of separators is entirely justified by certain conditions peculiar to the process of Portland-cement manufacture, and the general use of separators would be a serious error in this business. As this opinion may not be generally accepted, and as the grounds on which it is based have never been discussed in print, the advantages and dis- advantages of separators may be worth discussing in some detail. This is particularly necessary because the writer realizes, and freely admits, that the views which he holds may be proven to be based on incorrect premises. In that case a relatively brief series of experiments, which could be carried on in any cement-mill, would be of great service to all interested in the technology of the cement industry. Advantages of separators. In crushing any material, if the only things to be arrived at are low cost of crushing per ton of product and high tonnage of product per hour, one fact may be regarded as firmly established; that is, that it is an error to feed to any machine (of the series of reducers employed) material fine enough to go to some machine further along in the reduction process. Beginning with the first or coarsest crusher of the series, while most of its product will only be fine, a certain part of its product will be sufficiently fine to be passed on to the third or even fourth reducer. If this be done the product per hour of the series will be greatly increased, each machine will be working on material of fairly uniform size, and the cost of crushing per ton will be greatly reduced. Applying this to the Portland-cement industry it is probably safe to say that in a series of reducers (consisting for example of a coarse crusher, a fine crusher, a Williams or ball mills and tube mills) the product per hour of the series could be increased at least 50 per cent PREPARING THE MIXTURE FOR THE KILN. 427 by simply placing separators between each of the steps of the series. The cost of crushing per ton of product would incidentally be decreased, but not in quite the same ratio. Disadvantages of separators. In view of the enormous gain in out- put that would be secured by the use of separation, as noted in the preceding paragraph, it is evident that the disadvantages attending their use must also be very great, for otherwise every cement-mill would now be fully equipped with them. The principal disadvantages attending the use of separators in a Portland-cement plant are two : 1. Separators destroy the uniformity of the product. 2. Separators prevent the attainment of very great fineness. These two objections will be discussed in the order named. It is probable that the second is in actual practice the more important. Uniformity of product destroyed by separators. Throughout the entire course of cement manufacturing processes, from the moment the raw materials enter the will until the finished cement is packed the object of the manufacturer is to attain a product as nearly homo- geneous as possible. The quality of the cement depends in large part on the extent to which his attempts to secure absolute uniformity have been successful. In the opinion of the present writer, one great objec- tion to the introduction of separators into cement practice is that their use will tend to destroy this uniformity. The action of separators is based commonly on one of two principles. They take advantage either (a) of differences in the weight of particles or (6) in differences of size of particles. In the first case, the separator, other things being equal, will take out both the finer particles and the particles of lowest specific gravity. In dealing with raw mixtures they will tend to separate the clay particles from the limestone particles. Separators depending for their action on the differences in size between the particles will take out the finer, which, other things being equal, will, of course, be particles of the constituent which is most readily pulverized. Similar effects will be observed if separators be used in the progress of clinker-grinding, for the lighter-burned, more readily ground portions will be separated from the rest and the uniformity of the product will be destroyed. Great fineness prevented by separators. If in a plant not using sepa- rators the finest grinders be adjusted to give a product of which 95 per cent passes a 100-mesh sieve, a very large proportion of that prod- uct (say 70 to 75 per cent) will pass a 200-mesh sieve. If separators were installed throughout the plant, and the same adjustment of the fine grinders maintained, a product passing the same 95 per cent through 428 CEMENTS, LIMES AND PLASTERS. a 100-mesh sieve would probably pass not over 60 per cent through 200-mesh. As it is the percentage of very finely ground particles which gives the chief value to the cement, the use of separators on the clinker- grinding side of a cement-plant would seem to be inadvisable. In pul- verizing raw materials the same objection can be made, though it is not so important as in grinding the finished product. The following papers contain data on the subject of separators: Eckel, E. C. Some of the reasons why separators are not used in Portland- cement works. Engineering News, vol. 51, p. 344.*. April 7, 1904. Fraser, G. H. Recent results obtained with the Kent mill as a fine grinder. Cement, vol. 6, pp. 74-79. May, 1905. Humphrey, R. L. The plant of the Buckhorn Portland Cement Co. Engi- neering News, vol. 50, pp. 408-414. Nov. 5, 1903. Michaelis, W., jr. Air separation of cement. Cement and Engineering News, vol. 15, p. 2. Oct., 1903. CHAPTER XXXV. STANDARD TYPES OF CRUSHING AND PULVERIZING MACHINERY IT has seemed advisable to devote a chapter to the description of various standard types of crushing and pulverizing machinery. In selecting the particular machines to be described in this chapter, the writer has attempted to include those, and only those, which are known to him as being in satisfactory operation at one or more American cement- plants. It is possible that in endeavoring to exclude worthless types some meritorious machines may have been unintentionally neglected; but it is safe to say that over 95 per cent of the cement made in this country is crushed and pulverized by machines described in the follow- ing pages. In describing the various types of crushers and pulverizers, the drawings of the mills and the data relative to their construction and mechanical operation have been taken mostly from trade catalogues or from descriptions prepared by the manufacturers. The data on output, fineness and power lequired were, however, in almost all cases obtained from managers of cement-mills, and are believed to be entirely reliable. Classification of Grinding Machinery. So many types and varieties of crushing and pulverizing machinery are now on the market that it is difficult, from a single description, to form much of an idea of the relation of any given one of these machines to any of the others. To aid in this, the machines described in the following pages have been grouped under eight classes, according to their general methods of action. This grouping is as follows: Class 1. JAW CRUSHERS; material crushed between two jaws which approach and recede BLAKE CRUSHER. Class 2. CONE GRINDERS ; material crushed by the revolution of a toothed cone or spindle within a toothed cup GATES CRUSHER, CRACKERS. Class 3. ROLLS; material crushed between two or more plain, fluted, or toothed cylinders revolving in opposite directions ROLLS. 429 430 CEMENTS, LIMES, AND PLASTERS. Class 4. MILLSTONES; material crushed between two flat or grooved discs, one of which revolves. MILLSTONES, BUHRS, STURTEVANT EMERY MILLS, CUMMINGS MILL. Class 5. EDGE-RUNNERS; material crushed in a pan, under a cylinder turning on a horizontal axis and gyrating about a vertical axis. EDGE-RUNNERS, DRY-PAN. Class 6. CENTRIFUGAL GRINDERS;' material crushed between rollers and an annular die, against which the rollers are pressed by centrifugal force. HUNTINGDON MILL, GRIFFIN MILL, NAROD MILL, CLARK PULVERIZER. Class 7. BALL GRINDERS; material crushed by balls or p'ebbles rolling freely in a revolving horizontal cylinder. KOMINUTER, BALL MILL, TUBE MILL. Class 8. IMPACT PULVERIZERS; material crushed by a blow in space delivered by revolving hammers, bars, cups, or cages. WILLIAMS MILL, RAYMOND PULVERIZER, STURTEVANT DISINTE- GRATOR, STEDMAN DISINTEGRATOR, CYCLONE PULVERIZER. Class i. Jaw Crushers. In this familiar type of coarse crushing machines, of which the Blake crusher is both the oldest and best-known representative, the material is fed between two powerful jaws, and crushed by their near approach to each other. The principle upon which these machines work is well adapted for breaking stone, but is not serviceable for finer reduction. The Blake crusher in its various forms, and the Dodge and other de- vices of the movable-jaw type, are therefore machines suitable for first reduction only. For detailed descriptions of the Blake crusher, and of the modifi- cations in which the same principle has appeared, reference should be made to the valuable paper cited below.* In mining and metallurgical practices jaw crushers are commonly used, but in the cement industry they have almost entirely given place to the gyratory crushers described in the next section. Class 2. Cone Grinders; Gyratory Crushers. Under this heading are grouped the crushing machines in which the material is crushed between a toothed or grooved cone or spindle and the grooved cup within whicK it revolves. The Gates crusher, the Mosser crusher, and the McEntee and other " crackers" are here * Blake, W. P. The Blake stone and ore-breaker: its invention, forms, and modifications, and its importance in engineering industries. Trans. Amer. Inst. Mining Engineers, 1903, pp. 988-1031. CRUSHING AND PULVERIZING MACHINERY. 431 included. The " crackers' 7 have been described quite fully on page 239, in dealing with the crushing practice in natural-cement plants, and the Mosser crusher is of essentially the same design as these crackers, though built for heavier work. The Gates crusher, shown in view in Fig. 87, and in section in Fig. 88 ; is probably the most extensively used machine of its type. In re- FIG. 87. External vie\\ of Gates crusher. gard to its power requirements, capacity, etc., its manufacturers state: " In estimating power to drive our breakers we have provided for running an elevator and screen also. But it must be borne in mind that no close estimate can be made to cover all sorts of rock and ore; and further, it should be observed - that it requires much more power per ton to break rock to J inch than is required to break it to an inch. The estimates given in Table 176 are intended to cover the ordinary macadam breaking. For fine breaking add liberally to the power. Long experience has demonstrated the reliability of the follow- ing general rule, applicable to breaking the hardest rock to 2^-inch ring, viz.: The Gates breaker will not require over one horse-power per ton of rock broken per hour.'' 432 CEMENTS, LIMES, AND PLASTERS. The estimate (see Table 177) of the cost of crushing with a Gates crusher is based by Professor Richards on the results of a number of FIG. 88. Sectional view of Gates crusher. The names of the several parts designated by numbers in the above illustration may be found in the following table: 1. Bottom plate 11. Bevel-pinion 24. Octagon step 2. Bottom shell 12. Band-wheel 25. Main shaft 3. Top shell 13. Break-pin hub 26. Upper ring nut 4. Bearing-cap 14. Break-pin 27. Lower ring nut 5. Oil-cellar cap 15. Oil-bonnet 28. Steel step 6. Spider 16. Dust-ring . 29. Lighter screw 7. Hopper 17. Dust-cap 30. Lighter screw, jam nut 8. Eccentric 18. Head 31. Counter-shaft 9. Bevel-wheel 19. Concaves 33. Oiling-chain 10. Wearing-ring 22. Chilled wearing-plates mill tests at ore-treatment plants. As the ore handled at these mills was, in general, harder than the raw materials used in Portland-cement manufacture, allowance should be made for this feature. CRUSHING AND PULVERIZING MACHINERY. 433 TABLE 176. SIZES, POWER, ETC., OF GATES CRUSHERS. Dimen- Size Engine Size. Dimen- sions of Each Receiving Opening Inches. Dimen- sions of Receiving Openings Combined Inches. Weight of Breaker, Lbs. Capacity per Hour, in Tons of 2000 Lbs., Passing 2i- inch Ring, According to Character of Rock or Ore. Smallest Size Prod- uct can be Made at One Break, Inch. sions of Driving- pulley, Inches. evolutions of Driving-pulley. Recom- mended to Drive Breaker, Elevator, and Screen. Indicated Horse- 1 3 ^ fc &, power. F 2X 6 2X 12 650 a 8 2f 700 1 to 11 H 7X18 7X 54 11,800 8 1 24 *8 4 600 A LU -L ^ 10 15 OD 4X15 4X 30 3,550 2 to 4 2 16 6 500 4 5 ID 5X18 5X 36 5,5CO 4 8 | 20 7 475 8 10 2D 6X21 6X 42 8,000 6 12 1 24 8 450 12 15 3D 7X22 7X 45 14,000 10 20 H 28 10 425 20 25 4D 8X27 8X 54 21,000 15 30 32 12 400 25 30 5D 10X30 10 X 60 30,000 25 40 if 36 14 375 30 40 6D 11X36 11 X 72 42,000 30 60 2 40 16 350 40 60 7D 14X45 14 X 90 63,000 75 125 2| 44 18 350 75 125 7|K 14X45 14 X 90 67,000 75 125 2i 44 18 350 75 125 8D 18X63 18X126 94,000 125 200 3 48 20 350 100 150 8L 18X63 18X126 91,500 125 200 3 48 20 350 100 150 9K 21X48 21 X 144 155,000 300 500 4 56 20 300 100 150 TABLE 177. ESTIMATED COST OF CRUSHING BY GATES CRUSHER.* Number of crusher 4X30 72 3 $375 0.021 0.169 0.541 5.556 0.971 0.308 2 6X42 216 9 $760 Cost in Ce 0.021 0.114 0.541 1.852 0.971 0.308 4 8X54 540 22 $1800 ;nts per Toi 0.021 0.108 0.541 0.741 0.971 0.308 6 11X72 1080 45 $3300 i Crushed. 0.021 0.099 0.541 0.370 0.971 0.308 " 8 18X126 3000 125 $7000 0.021 0.076 0.541 0.133 0.971 0.308 Size of mouth in inches Tons crushed in twenty-four hours. . Horse-power used List price of crusher Cost of oil interest and depreciation ' ' of power " " labor. . " wear " " repairs Total 7.566 3.807 2.678 2.310 2.050 Class 3. Rolls. In the group of crushing machines here considered the material is crushed between two or more cylinders which revolve in opposite direc- tions. These cylinders may be plain-surfaced, longitudinally fluted, or toothed, according to the character of work they are expected to perform. As the fineness of the product is regulated chiefly by the * Richards, R. H. Ore Dressing, vol. 1, p. 50. 1903. 434 CEMENTS, LIMES, AND PLASTERS. FIG. 89. Elevation of rock-crushing system, Edison plant. (The Iron Age.") CRUSHING AND PULVERIZING MACHINERY. 435 closeness with which the cylinders are set, it is obvious that we have here a type of machine very different from the jaw crushers and cone grinders previously discussed, for rolls can be used for either coarse or fine grinding. By far the most extensive use of rolls in the cement industry is made at the Edison plant at Stewartsville, N. J., where all the crush- ing and grinding on both the raw and clinker sides of the mill is accom- plished hi rolls. At other plants rolls are used, if at all, only for crush- ing clinker, direct from the kilns, to a size that can be economically handled by Griffin, Huntingdon or ball-mills. FIG. 90. View looking down between 5-foot rolls, Edison plant. (Engineering News.) At the plant of the Edison Portland Cement Company the raw materials are crushed and pulverized in a series of rolls of special design. The elevation of this series is shown in Fig. 89, while Fig. 90 is a view looking downward on one of the coarser sets of rolls. In Fig. 91 the plan and elevation of the fine-grinding rolls of the Edison plant are presented. With the exception of the Edison plant, however, rolls are rarely or never used for fine grinding in Portland-cement practice on either the raw or clinker sides of the mill. They are commonly used in coal crushing, particularly when lump or run-of-mine coal is bought (see pp. 514, 515) ; and in many plants they are used as a first reducer on 436 CEMENTS, LIMS, AND PLASTERS. i^ .,, * 4-10 CRUSHING AND PULVERIZING MACHINERY. 437 clinker. For this last purpose they are specially well adapted, because they can handle hot clinker with less injury than any other form of grind- ing machinery. Class 4. Millstones. This class of crushing machinery includes those types in which the material is ground between two flat or grooved discs, one of which revolves. These discs are usually set horizontally, though they may also be arranged vertically. The class includes the millstones and buhr- stones proper, and several patented devices such as the Sturtevant rock-emery mill and the Cummings mill. As these machines are exten- sively used in grinding natural-cement clinker, they have been already described under that head (pp. 239-243). In Portland-cement practice they are now rarely used, though in the old-style wet-process plants millstones were almost invariably adopted. At one small Portland-cement plant seven runs of 48-inch upper- runner French buhrstones were employed in taking J-inch clinker from the rolls and reducing it to about 20-mesh. Each run of stones required 8 H.P. and handled about two barrels of clinker per hour. A sieve test showed the following results: Mesh of sieve 20 30 50 100 Per cent passed 98 98 92 65 Per cent residue 2 4 8 35 The seven mills required about one man's time for dressing. The particular economy of the installation lay in the fact that the mills cost only about $10 per run, having been purchased from old flour-mill plants. New mills of similar type would cost about $250 each. Class 5. Edge-runners. The class of edge-runner mills includes those types in which the material is crushed in a pan under a cylinder, the cylinder haying two motions turning on itself on a horizontal axis and also revolving as a whole about a fixed vertical axis. The familiar " dry-pans" and "wet-pans" of brick-plants are here included, a modern dry-pan of improved type being shown in Fig. 92. In the Portland-cement indus- try, particularly in wet-process plants, dry-pans are much used in grind- ing the clay or shale. Elsewhere they are of little service. Clay is ground in a dry-pan, at one marl-plant, at the rate of 6 tons per hour, the dry-pan requiring about 30 H.P. for operation. This 438 CEMENTS, LIMES, AND PLASTERS. power figure seems unusually high, for at other plants the dry-pan doing similar work is usually estimated as taking about 15 H.P. A dry-pan used on fairly hard shale at another wet-process cement- plant takes fragments up to 4"XS" in size and reduces them to about 8-mesh. It requires about 20 H.P. and gives a product of about 6 tons per hour. <* FIG. 92. Dry-pan. (Allis-Chalmers Co.) Class 6. Centrifugal Grinders. In the more typical examples of this class of mills the material is crushed between a horizontal die ring and one or more vertically hung rollers which are held against the ring by centrifugal force. Such typical examples among cement-grinding machinery are the Huntingdon and Griffin mills. A somewhat different arrangement exists in the Kent mill, which is included here for convenience. In this mill the ring is set vertically and the three rolls horizontally. Only one of the rolls is positively driven, the other two rolls and the ring taking motion from the driven roll. The .mills of this class have been used as one-stage reducers at a number of cement-plants, taking raw material or clinker in ^- to 1-inch sizes and reducing it to 85 to 95 per cent through a 100-mesh sieve, and they have accomplished this task with a fair degree of efficiency. Of late years, however, the tendency has been very markedly in favor of lightening this task by putting in tube mills for the final reduction. Huntingdon mill. The Huntingdon mill as used in the manufac- ture of Portland cement is a slightly modified form of the mill of the same name used in gold-ore treatment. The rights of the Huntingdon CRUSHING AND PULVERIZING MACHINERY. 439 mill for the cement industry are controlled by the Atlas Portland Cement Company. The mill has in consequence been used only in the plants of that company, where it seems to have given good satisfaction. In the Huntingdon mill three heavy rollers are suspended from a circular horizontal head, the attachment being such as to allow free radial swing to the rollers. The rapid revolution of the head causes the rollers to diverge, swinging outward, and being pressed by centrif- FIG. 93. Huntingdon mills at Atlas plant. (Atlas Portland Cement Co.) ugal force against a horizontal annular die ring of steel. The material fed in is pulverized between this fixed ring and steel-head rings attached to the bottom end of each roller. The mill, therefore, differs from the Griffin mill, chiefly in the fact that the single roller of the latter mill is revolved positively by power applied directly to its upper end, while in the Huntingdon mill the individual rollers are not positively rotated. A Huntingdon mill requires about 40 H.P. If fed with material varying from \ to 1 inch in size, its output on clinker will be about 8 barrels per hour, ground so as to pass 92 per cent through a 100-mesh sieve. On raw material its output will vary from 15 to 25 barrels per 440 CEMENTS, LIMES, AND PLASTERS. hour, ground to pass 93 per cent through a 100-mesh sieve. For grind- ing clinker to a fineness of 90 or 92 per cent on 100-mesh, the Hunting- don mill has given good satisfaction, but for the greater fineness now required by many specifications, it is probable that the finishing work can be done more economically with the tube mill. The rate of grind- ing clinker, quoted above, is -equivalent to an expenditure of 5 H.P. hours per barrel of cement, which is about the same as the work done by the Griffin mill under similar conditions. Griffin mill. If we disregard the enormous Atla$. plants, the Griffin mill is by far the most extensively used of the class of centrifugal grind- ers. It is shown in section in Fig. 94. On reference to this figure it will be seen that the power is received by a pulley (17) running horizontally. From this pulley is suspended the shaft (1) by means of a universal joint (9), and to the lower ex- tremity of this shaft is rigidly secured the crushing roll (31), which is thus free to swing in any direction within the case. This case consists of the base, or pan (24), containing the ring, or die (70), against which the roll (31) works, and upon the inner vertical surface of which the pulverizing is done. In dry pulverizing, this pan x or base (24), has a number of openings through it downward, outside of the ring, or die, which lead into a pit, or receptacle, from which it is delivered by a conveyor. Upon this base is secured the screen frame (44), which is surrounded with a sheet-iron cover (45) (in the wet mill this cover is not used), and to the top of which is fastened a conical shield (25), open at the apex, through which the shaft works. The cut shows the pulverizing roll attached to the lower end of the shaft (1), and just above the roll is the fan (7), which is used in the dry mill, but not in the wet. On the under side of the roll are shown shoes, or plows (5), which are used in both, and varied in shape accord- ing to the nature of the work to be done. The pulley (17) revolves upon the tapered and adjustable bearing (20), which is supported by the frame composed of the standards (23). Two of these standards (23a) are extended above the pulley to carry the arms (22), in which is secured the hollow journal pin (12). Within the pulley is the universal joint from which the shaft (1) is suspended. This joint is composed of the ball, or sphere (9), with trunnions attached thereto^ These trunnions work in half boxes (11) which slide up and down recesses in the pulley-head casting (16). The joint in the pulley is enclosed by means of the cover (13), thus keeping the working parts away from all dust and grit. CRUSHING AND PULVERIZING MACHINERY. 441 The lubricating oil is supplied for all parts needing it through the hollow pin (12). The roll is revolved within the die in the same direction that the shaft is driven, but when coming in contact with the die it travels around the die in the opposite direction from that in which the roll is revolv- 16 17 FIG. 94. Section of Griffin mill. ing with the shaft, thus giving the mill two direct actions on the material to be ground. There is a pressure by centrifugal force of 6000 Ibs. brought to bear on the material being pulverized between the roll and die, the united actions being very effective in their combination. 442 CEMENTS, LIMES, AND PLASTERS. When a quantity of the material to be reduced has been fed into the mill sufficient to fill the pan as high as the shoes, or plows, on the lower side of the roll, they work in it, stir it up, and throw it against the ring, so that it is acted upon by the roll; and when fairly in oper- ation the whole body of loose material whirls around rapidly within the pan, and, being brought between* the roll and die, is crushed, and all that is sufficiently fine passes at once through the screen above the die, the coarser portion falling down to be acted upon again. The universal joint, by which the shaft is connected with the pulley, allows perfect freedom of movement to the roll, so that it can safely pass over pieces of iron, steel, etc., such as are usually found in all rock to be pulverized, without damage to the mill. The fan attached to the shaft above the roll draws air in at the top of the cone, forcing it through the screens and out into the discharge, thus effectually keeping all dust within the mill. In working dry the screen which surrounds the pulverizing chamber is of much coarser mesh than the delivered product; for instance, a 16-mesh screen delivers a product over 90 per cent of which will pass a 60-mesh screen. Two sizes of the Griffin mill are made: TABLE 178. SIZES, POWER, ETC., OF GRIFFIN MILL. 30-inch Mill. 36-inch Mill. Extreme height of mill above foundation Extreme width of mill 8 ft. 2J ins. 5 " 3 " 8 ft. 7$ ins. 6 " 3 " Height from top of foundation to center of pulley . Weight complete 6 " 4$ " 10,500 Ibs 6" 6i " 14,500 Ibs Speed of pulley 190 to 200 rev 135 to 150 rev per min. 15 to 25 H P. per min. 25 to 30 H.P. Diameter of pulley 30 ins 40 ins Diameter of roll 18 to 20 ins 22 and 24 ins Diameter of ring or die . . . 30 ins 36 ins Depth of contact surface on roll . 6 " 6 " Weight of ring or die . 260 Ibs 408 Ibs Weight of tire 100 " 175 " Pressure of roll against die 6000 " 8000 " When running on clinker which has been previously crushed to about J-inch size the Griffin mill will handle from 5 to 10 barrels per hour, using 25 to 30 H.P. For clinker-grinding the mill is usually equipped with 30- or 32-mesh screens, giving a product of about 95 per cent through a 100-mesh, and 70 to 80 per cent through a 200-mesh. In grinding raw materials, 24- or 28-mesh screens are used, which, CRUSHING AND PULVERIZING MACHINERY. 443 however, give a product practically equal in fineness to the 30- or 32- mesh screens used in clinker-grinding. With these screens the mill will turn out 2 to 3 tons of raw material (equivalent to 8 to 10 barrels) per hour, taking slightly less power than when running on clinker. The repair costs of a Griffin mill were stated, at a plant which has always used these mills extensively, to vary with the material crushed in about the following ratio: Repair costs on clinker : repairs on raw mix : repairs on coal. 11 : li : 1 Kent mill. The Kent mill is a comparatively untried machine, but deserves mention here because of the favorable results reported for it by Professor Newberry, and G. H. Fraser (see p. 468). The Kent mill is shown in Fig. 95, the casing being broken out, and one fixed check-ring B being partly broken away to show the feed chutes A, the free revolving ring C, the three crushing-rolls G, and the bottom discharge outlet F. Referring to the interior view, A is the feed-chute ; which enters the casing at opposite sides above one of the three rolls G and feeds into the angle between this roll and the ring C. The three rolls G (of which one is driven) are within and support the ring C, being drawn yieldingly against its concave inner face at three points by stiff springs acting against the bearing yokes carrying the shafts of the rolls. These yokes slide in lugs on the casing and pull outward according to the adjustment of the springs by their screws. The convex faces of the rolls fitting the concave inner face of the ring hold it in position side- ways, so that the ring always tracks on the rolls, but it is also checked against too much side play by fixed check-rings B fastened on the inside of the casing at a slight distance from the edges of the free ring, so as to leave a free space D between in which the free ring can play and through which the fine material may escape to the discharge chamber E f which surrounds the ring C. and at its lower part meets the discharge outlet F. The rings B are cut away for a space above the outlet F. The fixed rings can be easily replaced if ever worn out. In operation, the free ring is cushioned by the rolls and held cen- trally, both axially and laterally, thereby, but can yield to pass a hard substance or to cushion unequal thrusts, and can play sideways between the fixed rings B to equalize variations of charge between any roll and the ring. The driven roll drives the ring by contact with its inner face, the other rolls being passive and free to revolve by contact with the ring. The rolls and ring run on each other at like surface speeds. The charge streams in between the ring and one roll, passes the latter, 444 CEMENTS, LIMES, AND PLASTERS. and is carried on the ring past the other rolls, being held against the ring by centrifugal force, due to the speed of revolution, until the par- ticles are reduced to the mesh desired, when they fly off of their own lightness and float inward and fall by gravitation, or pass outward between the free and fixed rings, into the chamber E and down there FIG. 95. Interior of Kent mill. through to the outlet F. On entering the mill the charge is set in rapid revolution therein, and is kept revolving until all is reduced. This action is the same whether the charge is great or small. The charge cushions the operation of the mill, the greater the feed the less being the noise and vibration. Gravity doors giving an intermittent feed close the casing against escape of dust. These open every time the weight fed on them is enough to tilt them. The discharge flows out in a large continuous stream of the size for which the mill is adjusted. CRUSHING AND PULVERIZING MACHINERY. 445 Straight belts from any shafting, either directly above or beneath, or at any angle from the mill, are run to the pulleys at each end of the shaft of the driven roll, the speed required being 180 to 200 revolutions- for the driving-pulleys. No quartered belts are needed. 6 ft. 6 ins. 6" 6" 6 " 2 " 9 " 7500 Ibs. Height of mill over all Width of mill over all Length of mill over all Distance from floor to center of driving-pulleys . Weight, complete Pulley speed 180 to 200 rev. per min. Power required (according to fineness and ma- terial) from 15 to 35 H.P. Diameter of pulleys 30 or 36 ins. Width of belts 8 ins. Diameter of rolls 14 ' ' Diameter of ring (inside) 36 lt Width of contact surface on rolls and ring. ... 7 " Weight of ring 800 Ibs. Pressure of each passive roll against ring to 20,000 Ibs. Pressure of driven roll against ring 1000 " 21,000 " LIST PRICES. Style F "Mining" Mill, complete, f.o.b., N. Y . $2500 " " " ' " with separator and 30-foot elevator. . . 3000 '"'.' " " " crusher, separator and 30-foot elevator 3500 For the following data on the performance of the Kent .mill when handling cement clinker in actual practice the writer is indebted to Prof. S. B. Newberry: Working as an intermediate reducer on clinker, taking it from the kilns and grinding it to 20-mesh, the Kent mill has given as high as 50 barrels per hour, taking 24 H.P. in doing so. Repairs may aggregate $180 per year per mill. The following table shows the results of several sieve tests of the product from a Kent mill used in this way at a time when it was giving 43J barrels cement per hour. Mesh of Sieve. Per Cent Passing. 20 100 99.5 53.0 35.0 98.5 49.5 35.0 98.0 48.0 33.0 200 When used as a one-stage pulverizer, taking clinker from the kilns and reducing it to finished cement, a 60-mesh screen being used on the Columbian separator, the Kent mill gave a product of 11.6 barrels cement per hour. A sieve test of this product showed that 97.7 per 446 CEMENTS, LIMES, AND PLASTERS. CRUSHING AND PULVERIZING MACHINERY. 447 cent of it would pass a 100-mesh sieve and 77.7 per cent passed a 200- mesh. Class 7. Ball Grinders. The mills of Class 7 include all those in which the raw material is ground by flint, iron, or steel balls rolling freely in a horizontal or NOTE: 4"sq. holes around anchor-bolts to be left open imtil mill is adjusted on foundation. Scale of Feefi 12345 7 8 9 10 11 12 13 FIG. 97. Foundation plan, Lindhard kominuter, No. 66. slightly inclined revolving cylinder. Three fairly distinct types are included under this definition, examples of which types are described below under the respective headings of kominuters, ball mills, and tube mills. Of these three types, the kominuter is represented by only one machine, the Lindhard kominuter; but ball and tube mills are made by many manufacturers, the different makes differing slightly in design but often greatly in value. 448 CEMENTS, LIMES, AND PLASTERS. Kominuter. The Lindhard kominuter, like the ball mill, is a drum- shaped device, about as high as broad, suspended by a shaft passing through the heads and resting on bearings. Outside of this drum is a coarse perforated cylindrical plate, and this in turn is surrounded by a screen frame fitted with wire screening cloth. The material is fed to the kominuter, as in the ball mill, through an opening near the shaft at the head end of the mill. It passes from this to the other end, being ground meanwhile by the balls, because its only place of exit is at the tail end of the kominuter. The discharge material passing thus out of the inner drum, returns in opposite direction over the perforated plates, the coarser particles being returned to the inner drum by curved pipes. Of the material which succeeds in passing the perforated plate the finer particles go also through the final exterior screen and are sent to the tube mills. The coarser particles, rejected by the exterior screen, are returned to the inner drum by two other curved pipes. This arrangement, while giving a regular and fine product, prevents excessive wear on the wire-cloth screens, since most of the coarser par- ticles do not reach them, being rejected by the perforated plate. The kominuter was designed to overcome certain disadvantages which have become apparent in machines of the ball-mill type. As stated by the manufacturers of the kominuter, the ball mill had the following defects: I. Insufficient screen area to prevent materials already ground fine enough from returning to the mill to be acted on again. II. The gradual closing of the holes in the perforated grinding plates under the peening action of the balls, resulting in a daily de- crease in the output from the ball mill. III. The impracticability of lining the drum with the multiples of a single templet small enough to be handled by one man; that is, the lining is either a set of plates few of which are to the same templet or else a single plate to each face of the polygon, made so large as to re- quire much power to handle. IV. The lack of means of easy adjustment to allow for the widely varying grindability of materials, as well as for the desired fineness. The manufacturers present the following comparative statement, designed to show the manner in which these defects have been overcome by the kominuter. CRUSHING AND PULVERIZING MACHINERY. 449 BALL MILL. A polygon. Plates, angles, bars, bolts, and rivets machine-fitted and having only the strength of an angular construc- tion, held together with bolts and rivets. Perforations in the grinding-plates ex- tend from inlet end to the opposite end, and the material escapes through these throughout the whole length of the grinding surface. The peening grad- ually closes these perforations, thus con- stantly reducing the capacity. Injudicious feeding will pack the ma- terial between the perforated exits and the screens, and stop the sieving action. It is frequently necessary to remove the screens to again put the mill into service. Choking damages the screens. Best obtainable information shows about 40 H.P., operating under normal conditions. Maximum charge 4400 to 4500 Ibs. Take the ball mill at 100 per cent per ton of balls. The grinding-plates form in part the support for the screens. The life of the lining is governed bv the nature of the material ground. Eighteen months seems to be a safe average. As, however, the grinding-plates form a principal part of the ball mill, the re- lining of the mill involves the discon- necting of many of the main parts arid takes a gang of men several days. The cost of relining is further increased by the fact that the plates must be partially finished, it being only possible therefore to use in them material which can be readily machined. Since the grinding-plates form also the supporting plates, care must be exer- cised not to let them wear down too far or the balls will fall through the screens and dust casing. A drum boiler. KOMINUTER. like a section of a steam- The material travels from the inlet end to the opposite end across SOLID grinding-plates and escapes through large openings am the action of the bs protected from The kominuter can be intentionally overfed until the screens are completely choked. Running the kominuter with- out feed for 20 min. will clear the screens. The inside screen is usually coarse and strong. It serves to limit the amount of material passing to the outside final screen, and at the same time protects the final screen from excessive wear. Operating under normal conditions, between 40 and 45 H.P. With maxi- mum charge of balls and materials about 45. Maximum charge 6600 Ibs. The kominuter has given 125 per cent per ton of balls, as compared with 100 per cent from the ball mill, showing a capacity of about 85 per cent more than that of the ball mill. The screen device derives its support wholly from the drum itself, and is in no way connected with the grinding plates or with any part subject to heavy wear. The life of the lining is at least equal to that of the ball mill, but as the komi- nuter will grind at least 50 per cent more than a ball mill of equal size the- wear per ton of finished product is ob- viously very much smaller. The relining of the mill is done with much greater ease and at much less cost, as all plates can be handled by one man and merely passed through the manhole, no disconnecting of the principal parts, of the mill being necessary. The grinding-plates are not machined, which reduces the cost and permits the use in the plates of the material best suited for the purpose, irrespective of its- suitability for machining. Since the drum itself forms the sup- porting plate, the grinding plates can be worn out completely Consequently very little metal is thrown away in the re- placement of a lining. 450 CEMENTS, LIMES, AND PLASTERS, BALL MILL. Rejected materials return to the mill by gravity through perforated plates at each angle of the polygon. These plates extend the length of the mill. Assuming one third of the interior area to be cov- ered with balls and material, this pro- portion of the return plates is of 'course under an outward pressure. Another third of the interior area is moving up- ward, and through this portion alone the rejected material must pass back into the mill. As the remaining third is moving downward, a portion of the ma- terial may move downward with it and thus remain on the screen. Consequently a considerable proportion of rejected ma- terial is carried around indefinitely at the periphery. If the mill is overfed, the insufficiency of the return openings results in choking the screens and limit- ing, if not altogether stopping, the screen action. The discharge area in the ball mill is limited by the size and number of the holes in the perforated grinding- plates. The ball mill is not supplied with means for enlarging or reducing this dis- charge area. If the ball mill is fed with a material relatively fine and difficult of reduction, too large an amount will pass the perforations, overloading the screens and leaving too little material in the drum to be acted upon by the balls. Here the exit area is too large, and it cannot be readily reduced. On the other hand, if the ball mill is fed with a ma- terial easy of reduction, the balls will reduce more than the perforations and sieves can pass, resulting in an accumu- lation of an excess of finished material in the grinding-drum. Here the dis- charge area is too small, and it cannot be enlarged. A very large percentage of material which passes through the perforated grinding-plates has been ground small enough to pass the screens, but owing to the limited screen area it is returned to the mill and the power for regrind- ing is wasted. Twenty to twenty-five per cent of the material is thus reground unnecessarily. KOMINUTER. All materials fed to the kominuter must pass the full length of the drum under the grinding action of the balls, and is discharged through ports at the outlet end. The conical shape of the screens forces the material to move from vihe outlet end back to the inlet end, Hvhere the material rejected by the screens is caught by return buckets and returned to the inlet end of the drum by gravity. The buckets empty at the center of the mill around the shaft and above the level of the balls. It will thus be seen that it is impossible for any of the material to reenter the screens with- out repassing the grinding chamber, as may easily happen in the ball mill. On the kominuter adjustments of the feed, of the discharge ports, and of the screens, are easily and quickly made. The total area of the discharge ports is greater than is required by the maxi- mum charge of balls. It is an easy mat- ter to insert a sufficient number of port closers, or covers, to allow the discharge of only sufficient material to utilize the extreme capacity of the sieves. With these three adjustments, the highest possible efficiency upon a given material may be promptly and easily secured. The material discharged from the grinding-drum must pass over screens extending from the outlet end to the in- let end. The material refused by the screen in this travel is automatically re- turned to the inlet end of the drum. About one half of one per cent only of such returned material would have passed the screen. A kominuter running on raw mix a fairly soft, shaly limestone gave a product of 15,046 Ibs. per hour. The feed was coarse, varying from 4 inches down to dust, and would probably average 1J inches. The product gave the following sieve test: CRUSHING AND PULVERIZING MACHINERY. 451 Mesh of Sieve. Per Cent Passing. Per Cent Residue. 10 99.5 0.5 20 95.0 5.0 30 75.5 24.5 50 57.5 42.5 80 45.5 54.5 100 43.5 56.5 Working on Lehigh district raw material, a kominuter has given over 14,000 Ibs. per hour. On a harder mix, at a plant in the middle west, a kominuter gave 13,812 Ibs. per hour. FIG. 98. Exterior view of kominuter. (F. L. Smidth & Co.) A kominuter running on a mixture of limestone and shale, of which the limestone had been passed through a crusher to, say, 1 inch, and the shale through a disintegrator, reduced about 18,000 Ibs. per hour. The product on a number of sieve tests gave the following residues on a 100-mesh sieve: 54.5%, 55.2%, 59%, 52.4%, 43%, 45%. 452 CEMENTS, LIMES, AND PLASTERS. A test run on clinker of a kominuter recently installed gave a product of 60 barrels per hour. The kominuter was fitted with 14- mesh screen and the product gave a residue of 55 per cent on a 50-mesh sieve. Another plant using fresh clinker got a product of 35.3 bbls. per hour from one kominuter. This product gave the following results on a sieve test: Mesh ot Sieve. Per Cent Passing. Per Cent Residue. ^ 20 95.5 4.5 50 56.6 43.4 100 31.3 68.7 200 20.0 80.0 At a third plant clinker averaging J-inch size was handled by the kominuter at the rate of 15,114 Ibs. per hour. The product gave the following sieve results: Mesh of Sieve. Per Cent Passing. Per Cent Residue. 10 99.5 0.5 20 95.0 5.0 30 72.5 27.5 50 50.5 49.5 80 37.5 62.5 100 34.5 65.5 Ball mills. Ball mills are now made by a number of manufacturers, the types most commonly found at the Portland-cement industry being the Smidth, Gates, and Krupp, and in wet-process plants, the Bonnot. Of these the Smidth ball mill is being gradually replaced by the Lind- hard kominuter, and is now recommended by its manufacturers only for coal-grinding. It will therefore be omitted here and described on pages 516-519. The wear of steel balls in ball mills working on clinker will vary usually between the limits of T V and ^ Ib. per barrel cement. As the balls at eastern points cost about $120 per ton, the cost per barrel will therefore range between J and ^ cent. Gates ball mill. This mill is described by its makers as follows: "The Gates ball mill consists of two circular side plates provided with inwardly projecting and eccentrically located shelves. The side plates have rigidly attached to them at their centers hubs which are mounted on a heavy shaft which revolves in dust-proof bearings. One CRUSHING AND PULVERIZING MACHINERY. 453 of the hubs has suitable openings through which the material is auto- matically fed by the Gates patent feeder. "Resting on the inwardly projecting shelves and reaching from one side plate to the other and bolted thereto are the wearing-plates. These are eccentrically arranged so that one plate passes behind the next one, thus producing a step and also providing an opening through FIG. 99. Gates ball mill with feeder attached (Allis-Chalmers Co.), showing grind- ing-plates in position and all other screws removed. which residues from the screens are returned to the mill. The tum- bling of balls and material due to revolving the drum rapidly reduces the material to fine grit and powder, the steps serving to greatly in- crease the beating action of the balls against the material. "When partially reduced, the ground product falls through aper- tures in the wearing-plates onto the first screen. The rejections from this screen are promptly returned to mill through the openings between the overlapping plates; the screened material falls upon the second screen and the rejections from it are likewise returned to mill, while the fines go to the outside finishing screen, and what passes through falls into the dust-proof housing and is removed by conveyor or other 454 CEMENTS, LIMES, AND PLASTERS. means for final pulverization; the residues join those of the other screens and with them are returned to the mill and subjected to further grind- ing action of the balls until fine enough to pass the outside or finishing screen. FIG. 100. Transverse section of Gates ball mill. (Allis-Chalmers Co.) TABLE 179. SIZES, WEIGHTS, ETC., OF GATES BALL MILLS. Size, Numbers. Weight without Charge Balls. Weight Charge of Balls. Capacity on Portland- cement Clinker to 20-Mesh. Power Required. 7 29,500 Ibs. 3000 Ibs. 12 to 16 bbls. per hour 30 to 40 8 41,100 " 4500 " 18 to 24 " " " 40 to 50 NOTE. From 100 to 120 per cent additional power is required momentarily in starting the above machines. When pulverizing to pass all through 20-mesb from 30 to 40 per cent will pass a 100-mesh sieve. CRUSHING AND PULVERIZING MACHINERY. 455 Krupp ball mill. This- mill is described as follows by its agents: "The Krupp ball mills are made in twelve different sizes, but for cement purposes in this country the size known as No. 8 has been uni- formly adopted. "The mill comprises essentially a cylindrical grinding-drum mounted on a hammered steel shaft running through it and provided with a dust- proof sheet-iron casing. The main shaft is carried on bearings secured to heavy bedplates bolted to a masonry foundation. The grinding- drum is composed of overlapping cast-steel plates (grinding-plates). The sides or head-walls of the drum are of wrought iron, secured to the main shaft by means of naves and lined with steel side-plates. "The grinding-plates are strengthened on one half and bent inwards, and in the other half holes are bored and a series of blades or scoops (return scoops) of sheet iron, directed outwards, is provided. At those places where the scoops overlap, channels are left extending right across the whole breadth of the drum and partly stopped by low pro- tection sieves arranged radially on the scoops. " The drum is surrounded externally by a fine cylindrical sieve located at a certain distance from the grinding-plates. This sieve is constructed of a number of frames connected together and secured to angle-iron rings, forming the outside limit of the sides of the drum. " The sieve frames are supplied in wrought iron. The sieves are covered by a phosphor-bronze or steel-wire gauze; the mesh is adapted to the required degree of fineness of the product desired. "To protect the fine sieves, a coarse inner or fore sieve is arranged between the outer sieve and the grinding-plates. This inner sieve is constructed of finely perforated steel plates secured to the lateral flangeb of the grinding-plates and to the back of the scoops; in front of the latter and across the whole breadth of the drum apertures are left. " The drum contains a charge of Krupp special forged steel balls of various sizes and of a certain specified weight. "The material is fed into the drum through a hopper secured to the nave at the front end of the drum. The nave itself is designed so as to form two peculiar slanting spokes which act as a screw conveyor and prevent the balls from being thrown out of the mill. " The lower part of dust casing of the mill serves also for collection and discharge of the product ground and is provided with an outlet closed by a slide. "In the larger mills a rectangular aperture is provided at the top of the dust casing for the purpose of connecting the mill with an air- shaft arranged above. The connection is made by means of a canvas 456 CEMENTS, LIMES, AND PLASTERS. hose secured to the dust casing by a wooden frame and clamping-screws. The current of air created in the shaft carries off the moist vapors .arising during the grinding process, prevents dust flying from the feed- hopper, and likewise prevents the mill and material being ground from becoming unduly warm. In order that the current of air may be regu- lated according to the weather it is Advisable to fit a damper in the shaft. FIG. 101. Automatic ball-mill feeder. (Smidth & Co.) " Those parts of the mill which are subject to wear and tear, i.e., the balls, grinding, and side-plates, are made of extremely resistant well- tested material, so that the wear and tear is reduced to a minimum. "The various parts of the mill may be readily renewed. For this purpose a manhole closed by a wrought-iron cover is provided in the side of the drum opposite to the hopper, so. as to admit of easy access to the interior. CRUSHING AND PULVERIZING MACHINERY. 457 "The No. 8 mill will take pieces up to 6 inches in diameter, but the feeding device usually used is designed to take material up to 2J inches only, which is as large as required for cement purposes in this country. "In grinding, the balls in consequence of the peculiar arrangement of the grinding-plates, on rotation of the drum not only fall on the plate and roll forwards, but also fall over one another, whereby the mate- rial is rapidly broken up and finely crushed. After being thus ground for a sufficient time, the material falls through the apertures of the grinding-plates onto the inner or fore sieves, from this onto the fine sieves and through this finally, completely ground, into the lower part of the dust casing, whence it is discharged through the outlet. The material not yet sufficiently ground, and retained by the inner or the fine sieve is carried back by the return scoops into the drum for further grinding. This method of working gives rise to very little dust, as the material being ground as soon as it is sufficiently finely crushed, is immediately discharged automatically." A Krupp 5'X21' ball mill, run on J-inch clinker, gave a product of 7914 Ibs. per hour. This product showed sieve results as follows: Mesh of Sieve. Per Cent Passing. Per Cent Residue. 10 100.0 0.0 20 99.5 0,5 30 84.5 15.5 50 57.0 43.0 80 39.0 61.0 100 36.0 64.0 Tube mills. Unlike the ball mill, the tube mill has been an unquali- fied success as a fine grinder. Several manufacturers have placed these mills on the market, those most generally used being the Davidsen (F. L. Smidth & Co.), the Gates (Allis-Chalmers Company), the Krupp, and the Bonnot. The latter is rarely used in dry-process plants, but is common in wet plants. The following description of the tube mill, published by the manu- facturers of the Davidsen type, contains much of interest on the group in general. "Davidsen tube mill. The tube mill is simple, effective, and eco- nomical in both operation and maintenance. It consists of a wrought tube mounted as a shaft by the attachment of dome-shaped ends so formed as to make shafts, which rest in bearings at both ends. A large gear attached to the tube and a pinion attached to the pulley-shaft make the actuating device. The tube is lined with stone or chilled 458 CEMENTS, LIMES, AND PLASTERS. iron, as may be required, and the tube is about one-half filled with flint balls. The enormous grinding surface thus provided permits of a very slow speed of operation 27 turns per minute. The shaft at the feed end is hollow and a screw conveyor carries in the material. By simply regulating the feed any degree of fineness, even to impalpable powder, may be attained. ^ " Careful experiments have proved that for coarse grinding in a non- continuous mill, such as an Alsing drum, the highest efficiency is obtained from a charge consisting of a large mass of the material to be ground with a relatively small amount of flint balls, and that the highest effi- ciency in fine grinding results from the charge of a small mass with a FIG. 102. Gates tube mill. (Allis-Chalmers Co.) large number of flint balls. Take, for example, a charge in the Alsing drum of a small proportion of balls to a large proportion of material. Up to a certain degree of coarseness the grinding will proceed rapidly with such a charge, but beyond that point to secure a high degree of fineness would require a proportionately very much longer working of the materials. Consequently any non-continuous working machine for pulverizing material is both unsatisfactory and expensive. " In the tube mill this principle has been taken advantage of; the flint balls bear a progressive ratio to the mass of material best adapted to the conditions demonstrated in the original experiments. "The material to be ground is fed in at the center of one end of the tube mill, and is delivered, ground, at the periphery of the other end. The material consequently travels in a vertical line from the center to the circumference of the mill, and in a horizontal line from one end to the other. The tube mill being mounted horizontally, a CRUSHING AND PULVERIZING MACHINERY. 459 longitudinal section, if taken when the mill is at rest, would show the material in the shape of an acute-angled triangle, with point at delivery end and base at the inlet; the transverse section at the inlet will, there- fore, show a large amount of material to a small amount of flint balls, and a transverse section at outlet end would show the mill half full of flint balls, with but a small amount of material extremely finely ground. Between these two points there is a progressive disposition of the balls and material, but at the same time the flint balls are evenly distributed throughout the tube. Applying the principle enunciated above it will be seen that the coarse grinding is accomplished where the proportion of material is great in relation to the flint balls, and that the mere fact of delivery at the periphery of the other end compels an automatic and gradual adjustment of relations between balls and materials, so that at the delivery end a transverse section would show the mill half filled with balls with but a small portion of material subject to the grinding action. " The speed at which the tube revolves tends to carry the mixture of balls and material to a certain height in the mill from which point balls and material fall together rolling and grinding as they seek the bottom of the mill; the action might be likened to the action of water at the crest of a wave. This sort of action can only take place when the mill has been half filled with the flint balls; if less than a proper charge of flint balls is used the whole mass has an inclination to only slide upon the inner surface of the tube, and none of the turbulent wave-crest action suggested takes place. Of course, much the best grinding is accomplished through keeping the whole mass of material grinding and pounding. "The tube is lined with iron, specially made tiles, or a natural stone which we have named 'Silex.' These are set with Diamond cement made only for this purpose. Any bricklayer can readily lay up a lining. "The life of a lining depends upon the character of the materials ground on it. With ordinary care all linings should last at least a year. In most cases linings have served eighteen and twenty-four months. Experience with Silex indicates that it will serve two or three years of continuous grinding. "Experiment has proven that for grinding, flint balls are positively the most economical medium. The tube mill has therefore been de- signed for and proportioned for the use of flint balls. "The tube mill for grinding slurry or wet materials is designed to use steel balls instead of flints, and is proportionately heavier in con- struction. 460 CEMENTS, LIMES, AND PLASTERS. "It is absolutely necessary that the tube mill should be filled up to the center line with the flint balls; there is, consequently, no interior shaft. The tube revolves upon stub shafts, which are firmly anchored to the dome-shaped ends, the one at the inlet end being hollow, and through which the material is fed. "The fact that flint balls in a tube, mill represent an extraordinarily large grinding surface, makes it possible to run the mill at a slow speed so that all the power required is actually made use of in the grinding process itself and nothing is wasted in maintaining a high speed of ma- chinery, which effects nothing in the grinding process itself. "The fineness of the output is regulated by the speed at which the material is fed into the machine. As every particle of the material fed must pass under the grinding action of the entire charge of balls, a thor- ough and uniform grinding is bound to be the result ; in fact, in practice it is found that the uniformity of the output is so great that it is unneces- sary that sieves should be used, and no provision is made in the machine by which they can be used. There is, of course, nothing to prevent the materials being sieved, if it is desired, but this is absolutley useless when there is regularity of feeding. At a slow rate of feeding any required fineness can be obtained with any grindable material. The requirements for grinding Portland-cement clinkers are ordinarily that no more than 12 to 15 per cent residue shall be caught on a No. 200 sieve; when the tube mill is fed with coarsely ground product it will turn out from 8 to 16 barrels per hour of Portland-cement ground to the required fineness. However, these figures do not represent fully the value of the grinding capacity of the tube mill. Careful scientific investigation has proved that the fine product from the tube mill, which will pass sieve No. 200, contains 50 per cent more overfine particles than is the case in the out- put from any other existing grinding-machine. "The foundation may consist of any material which 'will support the dead weight and enable the journals to be held firmly enough to withstand the tension of the driving-belt. In practice, however, it is, of course, preferable that the pedestal foundations should be deep enough in the earth to resist the action of frost, and should be made of permanent material, such as brick, stone or concrete. "In the matter of repairs in this machine, as in others, the advan- tage of slow speed over high speed is practically noticeable. This prin- cipal wear is, of course, upon the flint balls, and by merely dropping in a few at stated intervals, to maintain the charge at its original bulk, the principal results of wear are made good at once. "The flint balls, or pebbles, are a natural product; they are found CRUSHING AND PULVERIZING MACHINERY 462 CEMENTS, LIMES, AND PLASTERS. in certain parts of Europe and are comparatively inexpensive and of extraordinary hardness. It is the combination of these qualities that makes them preferable to" the steel balls, even to the extent of building the mill four times as large as would be required with the steel balls. These flint balls are sold in four different sizes. The mill is originally charged with all four sizes - in a set ^proportion, but replacements are always made with the largest size, as in practice the smaller pebbles seek the outlet end, and replacements are made at the inlet end. "On average Portland-cement clinker the pebbles wear at the rate of about 1 Ib. to 30 barrels of product. "Little or no harm follows the introduction into the mill of sub- stances foreign to the intended supply, such as steel or iron. When such substances reach the interior of the mill they either become grind- ing mediums and remain in the mill to accomplish that work, or else they are ground by the action of the mill and pass out with the product. This result is so unlike that following the introduction by accident of a piece of hard material into crushers, rollers, buhrstones or like ma- chines, where the grinding surface is of metal, emery or dressed stone, that this one fact recommends it promptly to the observing purchaser. "Two sizes of the Davids en tube mill are made: "No. 16. Requires 60 H.P. and a floor space of 36 feet by 13 feet, and turns 25 revolutions per minute. "No. 12. Requires 27 H.P. and a floor space of 29 feet by 11 feet, and turns 27 revolutions per minute. "The tube mill is a pulverizer, not a coarse grinder. "The ' ' grindability' of materials, even of the same class, varies so widely that it is impossible to give more than a general idea of the ca- pacity of the tube mill. As an index, however, it may be stated that the No. 12 tube mill will pulverize from 3500 to 6500 Ibs. of Port- land-cement clinker per hour, 95 per cent of which will pass 10,000 meshes per square inch, such clinker having been preliminarily ground to pass No. 20 or No. 30 sieve. "The No. 16 has about two and one quarter times the capacity of the No. 12." Krupp tube mill. The Krupp tube mill is described by its agents as follows: "The Krupp tube mill is a cylinder or drum 5 feet in internal diam- eter and 22 feet long between the heads of the mill. This mill is so designed as to admit either silex or iron lining being installed. "The feeding mechanism, the most accessible and most easily ad- justed feeder on the market, is so designed that in making a change CRUSHING AND PULVERIZING MACHINERY. 463 neither the mill nor the feeding mechanism is stopped so that an abso- lutely uninterrupted grinding action takes place. The main trunnion bearings are so designed as to admit of their being rebabbitted with the least possible lost of time and amount of labor. "The mill is driven by means of a split-spur gear and pinion, the former being just scant of 10 feet in outside diameter. The discharge device is the Krupp patent cone discharge and includes all the advan- tages claimed for what u known as the peripheral discharge and also separates worn-out pebbles, and eliminates all the dust. The heads of the mill are of the strongest possible type, being conical in form. "Every part of the mill is easily accessible, and all parts are made of the strength required for mills to be used in American cement prac- tice. "The capacity of the above mills, when working in battery, varies widely with the character of the material, varying as much as from 15 barrels per hour on a hard Portland-cement clinker to 60 barrels on a natural-cement clinker, the fineness of the product being 95 to 96 per cent through 100-mesh, 75 to 80 per cent through 200-mesh. FIG. 104. Exterior view of Bonnot tube mill. (Bonnot & Co.) "The capacity on raw rock varies from 8 to 10 tons per hour, same fineness as mentioned for clinker. "The capacity of the tube mill on coal varies from 3 to 4J tons per hour, 90 to 95 per cent through a 100-mesh sieve." Pebbles for tube mills. For use in tube mills flint pebbles are the most satisfactory grinding materials. These are obtained chiefly from Greenland, Norway, Denmark, France, and England. Very little is in print concerning the flint industry, two recent papers of interest being listed below.* * Hill, R. T. Flint, an ancient industry. Eng. and Mining Journal, Nov. 7, 464 CEMENTS, LIMES, AND PLASTERS. Concerning the French flints, Mr. Thackara writes as follows: "By the action of the sea on the base of the chalk cliffs, which form the coast-line of a portion of the Department of Seine Inferieure, frag- ments of the rock are detached. Those which are composed of the flint found in the cliffs, on account of their hardness, are not reduced to sand by the trituration .arising from the movement of the waves or tidal currents, and become what '^are known as sea flint pebbles. These are gathered on the beaches between Havre and St. Valery-sur- Somme, a distance of a little over 100 miles. Those which are nearly spherical in shape are carefully selected and are used in the Alsing sys- tem of cylinder grinding, which is becoming so generally employed for pulverizing cement, chemical and pharmaceutical products, etc. The others are bought by the potteries for making ordinary porcelain ware after being calcined, ground into a fine powder, and mixed with china clay. "According to the official custom-house statistics, there were 13,592 tons of flint pebbles exported from France during 1900, valued at $39,348. The value of the declared exports of these stones from France to the United States for the fiscal year ended June 30, 1900, was $16,743, of which $3849 were shipped from Havre, $4458 from Boulogne, and $8436 from Dieppe. "The prices of the flint pebbles for use in the potteries range from 5s. 3d. ($1.27) to 12s. ($2.92) per ton f. o. b. in bulk at Fecamp, St. Va- lery-en-Caux, Dieppe, T report, St. Valery-sur-Somme, and Havre, accord- ing to quality and to the port from which they are shipped. For the selected pebbles the prices vary from 35s. ($8.52) to 42s. ($10.21) f. o. b., packed in barrels or bags, packing included. "The rate of freight from Havre or Dieppe to New York averages 10 francs ($1.93) per ton of 1000 kilograms (2204.6 pounds). "French flint pebbles are shipped to England, Scotland, Norway, Sweden, Russia, Spain, Japan, and the United States. In the Baltic ports they have to compete with the pebbles exported from Denmark. Germany is now using silica sand from the river Rhine for pottery purposes, which replaces the flint pebbles. The French pebbles also have to compete with those collected on the English coast at Newhaven, Shoreham, and Rye, with the chalk flints shipped from London, and with the Greenland selected pebbles." For the following analyses of tube-mill pebbles, from lots furnished by various importers, I am indebted to the chemists named below. 1903. Thackara, A. M. Export of sea flint pebbles from France. U. S. Consular Report No. 1231. Jan. 6. 1902. CRUSHING AND PULVERIZING MACHINERY. . 465 TABLE 180. ANALYSES OF FLINT PEBBLES. 1. 2. 3. 4. 5. 6. 7. Silica (SiO 2 ) 97.16 95.20 95 00 93 05 91 50 90 20 87 00 Alumina (Al 2 Os) Iron oxide (Fe 2 Oa) JO. 64 3.40 3.40 6.97 2.96 10.10 13.30 Lime (CaO) 22 1 35 2 92 n d n d 1. Greenland. " Dana" brand. H. S. Turner, analyst. 2. Havre, France, light. Heiberg and Roney, analysts. 3. " : dark. 4. St. Valerien, France: no. 1. " 5., " " > : no. 2. " 6. Norway: light. " " " " 7. " : dark. From these analyses it will be seen what large variations in com- position occur in different kinds of flint pebbles. Other things being equal, the pebbles highest in silica should give the best results, while lime is a particularly injurious impurity. The high cost of flint pebbles for mills situated in the middle and western United States has led to many attempts to secure a domestic substitute for the expensive imported pebbles. Rounded pebbles of shape and character suitable for this use occur only on the shores of great lakes or along the beds of mountain streams. A California mill secures its supply from the American river, where rounded granite pebbles occur in quantity. These pebbles, gathered by Chinamen, cost less than $5.00 per ton at the mill, and are about half as durable as imported flints. For grinding 3000 barrels of cement, 800 Ibs. of granite pebbles were used up, as against 400 Ibs. of imported flint pebbles. Flint occurs in several formations in America, but in no case do these formations outcrop along the shore, so that the flint can be obtained only in rough angular masses. Along the north shore of Lake Superior hard quartzite pebbles are said to occur in quantity, and this district may furnish a supply for the American cement trade if any manufac- turer cares to investigate the matter. Class 8. Impact Pulverizers. The impact pulverizers include all those types of grinding-machines in which the material is broken by a blow, in free space, delivered by a series of rapidly revolving hammers, bars, cups or cages. This group therefore includes the Williams mill, the Raymond pulverizer, the Stur- tevant and Stedman disintegrators, the Cyclone pulverizer, and many other less well-known devices. The Stedman disintegrator is exten- sively used in crushing gypsum or plaster and natural-cement clinker, 466 CEMENTS, LIMES, AND PLASTERS. and has, therefore, been already figured and described on pages 36, 37. Of the other impact pulverizers, the Williams mill is the only one ex- tensively used in Portland-cement manufacture. Raymond pulverizer. Two Raymond pulverizers, one with three rollers and one with four, are in use at one plant, taking a dry lime- stone-and-shale mixture from dry-pahs at about 30-mesh and reduc- ing it to 93 per cent through a 100-mesh sieve. Each of these mills requires about 85 H.P., and delivers about 3| tons of product (=11J- barrels cement). This seems to be a remarkably high figure for power, being equivalent to about 7J- H.P. hours per barrel, and it is probable that the pulverizer is not being run up to its true efficiency. \ \ \ \ X X , X X FIG. 105. View of Williams mill, casing opened. Williams mill. The Williams mill is shown in view, with its casing opened, in Fig. 105, and in section in Fig. 1C6. It will be seen that it crushes by the blows of a series of hammers, rapidly revolving about a horizontal central axis. The following record of an actual working test, communicated to me by the chemist of a plant using the Williams mill on its raw mate- rials, will serve to show the percentages on different sieves that are CRUSHING AND PULVERIZING MACHINERY. 467 produced by this mill. At the plant in question the materials used are hard limestone and shale. The raw materials are run through a Gates crusher, which gives a product averaging 1J inches in size. FIG. 106. Section of Williams mill. Three Williams mills, using about 18 H.P. each, take the product from this crusher and reduce it to give the following residues: Mesh of sieve 20 50 100 200 Per cent residue 25 45.1 60.9 69.5 Per cent passing 75 54 . 9 39 . 1 30 . 5 The three Williams mills in use handle sufficient raw mix to give a production of about 1000 barrels per day. From these mills the material is fed to three tube mills, which take up 60 to 70 H.P. each, and complete the reduction so that 92 to 93 per cent passes a 100-mesh sieve. Another test, for the record of w r hich I am indebted to the manu- facturers of the Williams mill, was carried on in a plant also using lime- stone and shale, crushed to about 1J inch size. Mesh of sieve 20 50 100 200 Per cent residue 26.8 51.3 68.7 76.0 Per cent passing 73.2 48.7 31.3 24.0 468 CEMENTS, LIMES, AND PLASTERS. The writer has seen the Williams mill in operation at a number of Portland-cement plants, working on various raw materials and also on cement clinker. Its results when operating on shales, slates or thin bedded slaty limestone are remarkably good, and the machine appears to be particularly well adapted for such materials. References on crushing machinery. The following books and papers contain matter of interest in this connection, Richard's work being, of course, by far the most important: Blake, W. P. The Blake stone and ore-breaker; its ^invention, forms and modifications. Trans. Amer. Inst. Mining Engrs., vol. 33, pp. 988- 1031. 1903. Fischer, H. The operator of a tube mill. Engineering and Mining Journal, Nov. 17, 1904, pp. 791-793. Fraser, G. H. Recent results obtained with the Kent mill as a fine grinder. Cement, vol. 6, pp. 74-79. May, 1905. Hutchinson, W. S. The plotting of sizing tests. Transactions Amer. Inst. Mining Engineers, 1904. Lesley, R. W. The manufacture of cement. Trans. Am. Soc. C. E., vol. LIV, part B, pp. 89-130. 1905. Richards, R. H. Ore Dressing. 2 vols, 8vo, pp. 1236. 1903. Schvverin, M. Notes on some regrinding machines. Engineering and Min- ing Journal, Mch. 10, 1904, pp. 403-404. Anon. The pebble tube mill in metallurgy. Electrochemical and Metal- lurgical Industry, vol. 3, pp. 41-42. Jan., 1905. CHAPTER XXXII. CEMENT BURNING: FIXED KILNS THE preceding chapters have been devoted to a discussion of the raw materials for Portland-cement manufacture, and to the processes and methods of preparing a mixture of these materials for the kiln. In the present and following chapters the next stage of the industry will be taken up that of burning the raw mix into cement clinker. Fixed or Stationary Kilns. The earliest type of kiln used in Portland-cement manufacture was a simple vertical bottle-shaped kiln closely similar to those used in the burning of lime and natural cements. This was largely succeeded by improved types of stationary kilns in Germany and France, while in the United States the rotary kiln has become standard. Though stationary kilns are now very rare in American practice they have some undoubted advantages in localities where fuel is expensive and labor is cheap. As American engineers may soon have to consider the pos- sibility of manufacturing cement in Central and South America, where these fuel and labor conditions are fulfilled, it has been considered advisable to discuss the improved type of stationary kilns in some detail. A list of references to the more important papers on the sub- ject is also given at the end of the chapter. In order that the relationships of the various types of fixed or sta- tionary kilns may be clearly understood, it will be well to group them in classes according to the general principles on which their construc- tion and operation are based. Four such groups can be formed: 1. Dome or intermittent kilns. 2. Dome kilns with drying accessories. 3. Ring or Hoffmann kilns. 4. Continuous shaft kilns. These classes will be described in the order named. 469 470 CEMENTS, LIMES, AND PLASTERS. I. Dome or Ordinary Intermittent Kilns. All intermittent kilns will, for convenience, be here termed dome kilns, though the term is properly restricted to intermittent kilns of one particular shape. FIG. 107. Dome kiln The dome or bottle-shaped kiln is the original form on which most fixed kilns are based. As shown in Fig. 1C7 it is practically the shape of the older lime-kilns, differing usually in having a somewhat greater height for a given diameter. The type shown in the figure, which is the ordinary English form, is perhaps 9 to 12 feet in diameter at its widest portion, 15 to 18 feet from its base to this widest zone, and 25 to 35 feet in total height. This kiln is usually charged at several levels, one charging door being located a little below its widest point, and others being opened in the truncated cone which serves as a chimney. In German practice these kilns assumed a form nearly like that of the blast-furnace. The bo-dy of the German dome kiln is usually CEMENT BURNING: FIXED KILNS. 471 a cylinder, 9 to 12 feet wide and 25 to 30 feet high. This is surmounted by a truncated-cone chimney, often high so that the total height of the kiln may be 35 to 75 feet. Candlot states that at some German plants kilns 22 feet in diameter and 100 feet in height were used, each of which kilns would turn out 400 tons (metric) of cement for each run. Dome kilns are charged with fuel and mix, the latter in the form of bricks, in alternate layers, the proportions varying principally with the height of the kiln and the wetness of the bricks of mix. When the kiln is full the charging doors are closed and luted with fire-clay, and the lowest layer of fuel is ignited. As the burning progresses the entire mass settles, owing to the loss in fuel and carbon dioxide. The kiln may now be refilled to its former level, but nothing is drawn from it until the burning is complete, which may take from one to two weeks. Candlot states that the production of a dome kiln varies from J to 1 ton of clinker for each cubic meter of burning space, and that from 23 to 30 Ibs. of fuel are required per 100 Ibs. of clinker, the latter quantity varying according to whether anthracite, gas-coke, or oven-coke is employed. The labor cost of charging, drawing, and picking clinker from the dome kiln may vary from 30 to 50 cents per ton of cement, equivalent to about 5 to 10 cents per barrel. 2. Dome Kilns With Drying Accessories. The first and simplest improvement on the primitive dome kiln was to provide each kiln with a drying tunnel. The kiln thus improved was still intermittent, but the drying tunnel gave a certain fuel economy, particularly when very wet mixes were employed. The principal type of this class of kiln is the Johnson kiln. Johnson kiln. The Johnson or chamber kiln was apparently the first English improvement on the simple dome kiln. It consists essen- tially of a dome kiln roofed over at the top, and with a long horizontal passage, semi-circular in section, opening into the kiln near the top and leading to a stack. The wet slurry is placed in the horizontal passage and dried by the hot gases passing through it irom the kiln to the stack. The slurry when dry must be shovelled up and charged into the kiln by hand. Various modifications of the Johnson kiln have been suggested and used in English plants,* most of them depending for extra economy on passing the hot gases under as well as over the slurry to be dried. The Johnson kiln, with its different modifications, may be considered Proc. Institution Civil Engineers, vol. 62, pp. 74-76. 1880. 472 CEMENTS, LIMES, AND PLASTERS. essentially as combinations of old-style dome kilns and drying-floors. They utilize waste heat for drying the slurry; and are, therefore, more -economical in fuel consumption than is the single-dome kiln. They are all based on intermittent working of the kiln, however; and in all, the dried slurry must be charged into the kiln by hand. Six Johnson kilns were installed irf 1890 at the plant of the Western Ctiimuey SECTION Gate Chimney Drying_space. for .Slurry 1 , , jR Kiln FIG. 108. Plan and section of Johnson kiln. (Engineering News.) Portland Cement Company, Yankton, S. D., but have recently been Teplaced by rotaries. I believe that similar kilns were used in the first plant at Whitecliffs, Ark. 3. Ring or Hoffmann Kilns. The Hoffmann or ring kiln has been used quite extensively in Ger- many for burning Portland cement, lime, and bricks, but has never come into favor in either England or the United States. It consists essentially of a number of chambers arranged in a circle or ellipse around a central stack. Three flues lead from each chamber to (1) the cen- tral stack, (2) the chamber preceding it in the series, and (3) the chamber following it in the series. Each of these flues may be closed at will by the insertion of a partition of sheet iron. Each chamber also has a door opening to the outside of the kiln and used for charging and drawing. Assuming that the kiln is entirely empty (a condition which could occur only in firing up a newly built kiln), the operations would be as iollows: Each chamber would be loaded with bricks of dried slurry .stacked up as in a brick kiln. Slack or other fine coal is fed in at the I CEMENT BURNING: FIXED KILNS. 473 top of the chambers and one chamber is fired. All of the flues in the kiln leading to the stack are closed except one, i.e., the flue from the chamber behind the one which has been fired. All the inter-chamber flues are open except one, i.e., the flue between the fired chamber and FIG. 109. Section of Hoffmann kiln. the one immediately behind it in the series. The result of this arrange- ment is that the hot gases from the fired chamber pass in turn through each of the other loaded chambers until they arrive in the chamber immediately behind the fire, when they are passed into the central stack. The waste heat from the fired chamber is therefore utilized to the fullest extent in heating up all the other chambers. When the slurry in the fired chamber is converted into clinker this is allowed to cool. The chamber is then temporarily cut off from the rest of the series by closing its flues and the clinker is drawn. In the meantime the chamber next to it has been fired. The empty cham- 474 CEMENTS, LIMES, AND PLASTERS. her is recharged and the flues to the central stack and from the chamber behind it are opened, thus making the newly filled chamber the end term of the series. FIG. 110. Plan of Hoffmann kiln.' As noted, the slurry charged into a Hoffmann kiln is necessarily in the form of bricks. The expense of partly drying the slurry and mold- ing it into bricks must, therefore, be charged against the kiln. Taken as a whole the system is low in fuel consumption, but high in labor cost, especially since skilled labor is required for all the operations. Usually one chamber is loaded and one drawn each day. The out- put per kiln per day will, therefore, depend on the size of the chambers. 4. Continuous Shaft Kilns. Die tzsch kiln. In 1884 the Dietzsch kiln was first used in cement- manufacture, and its advantages soon became known. It has been in use at several American plants, and in the matter of fuel consumption is, perhaps, the best type of kiln that can be employed. Dietzsch kilns are built in pairs, back to back, as shown in Fig. 111. They are 60 to 75 feet h gh, and consist of a cooling chamber at the base D, a fire-chamber or "creuset" C, and a preheating chamber A. It will be seen that these three parts of the kiln are not all in one ver- tical alignment, but that the axis of the preheating chamber, though parallel to the axis of the main kiln, is off to one side some distance, so that the two portions of the structure communicate by a horizontal passage B. Aalborg or Schofer kiln. The Aalborg kiln, soon introduced in European cement practice after the success of the Dietzsch kiln had proven the possibility of economical continuous kilns, has been used at several American plants in a more or less modified form. The kiln is shown in section in Fig. 112. It will be seen that it is essentially the same as the Dietzsch, except that the preheating chamber, the burning space, and the cooling chamber are all in the same vertical line. This change, slight in appearance, economizes considerably in labor, for the charge descends of itself, without the rehandling necessary in the Dietzsch kiln. The mix is introduced through the charging opening A, while the coal is charged through the shutes (shown in the figure about an inch below A). CEMENT BURNING: FIXED KILNS. 475 In European practice Candlot states that an Aalborg kiln will turn out 10 to 15 tons of clinker per day, with a fuel consumption of 280 Ibs. coal per ton of product. FIG. 111. Dietzsch kilns. Hauenschild kiln. The Hauenschild kiln is a simple cylinder, charged at the top with both fuel and mix. It differs from a cylindrical lime- kiln only in having two distinct walls, with a space between. This annu- lar space is used either for drying the mix or for heating the air to be supplied to the kiln. The result is that the interior lining is kept fairly cool, so that the charge does not clinker in masses against the walls, 476 CEMENTS, LIMES, AND PLASTERS, FIG. 112. Section of Schofer kiln. CEMENT BURNING: FIXED KILNS. 477 which is the principal defect encountered in running a vertical kiln* continuously. FIG. 113. Hauenschild kiln. Schwarz kiln. In a recent paper on the manufacture of Portland cement from a mixture of slag and limestone C. von Schwarz describes a kiln used at a German cement plant. This kiln, here called the Schwarz kiln, is shown in partial section in Fig. 114. It is described as follows : Each kiln consists, in its essential part of a series of rings, each 1 inch to 1J inches in thick- ness, 8J feet inner diameter, and 18 inches in height. These rings are provided outside with ribs, r, Fig. 112, and placed in such a way, one above the other, that the vertical ribs cover one another, thus forming little vertical channels c, c, c all around, in which the air circulates from below to the top, like in a chimney, thus continually cooling the cast-iron rings from the s. SECTION a-b outside, and preventing them from getting over- FIG. 114. Partial plan i j .m ? i i i- and section of Schwarz, heated. The materials to be burnt are in direct kiln. (Engineering contact with the cast-iron rings, no lining of any News). kind being provided for. 'There are 18 such rings, put one above the 478 CEMENTS, LIMES, AND PLASTERS. other, the upper rings where the greatest heat occurs being hooped at the joints. The top of each kiln is provided with a cone and a chimney made of sheet iron, 3 feet in diameter and 30 feet in height. The cone has four charging doors, which can be closed by sheet-iron covers as soon as the charging is done. At a depth of 12 feet from the top the inner diameter of the kiln is lessened to nearly half its inner horizontal section, and on this zone is provided with a double row of tuyeres to admit compressed air, this arrangement having for its object to burn any carbpjiic oxide or carbu- retted hydrogen gas arising from below as completely as possible, as well as to concentrate the heat exactly where it is required, viz., on the place where the formation of the clinker is to take place. Compressed air is also introduced from below in two places. The pressed air is produced by a ventilator, the pressure being } inch to 1} inches of water. One charge consists of 100 bricks and 65 to 70 Ibs. of coke as fuel; one third of the coke could be rep'aced, if necessary, by anthracite or other small coal. As a rule, four kilns are arranged in one set, being provided with a common elevator and a common platform, for all four kilns together. They are surrounded by a scaffolding made of angles and tees, on which the staircase to mount the platform is fixed. At the same time cor- rugated galvanized sheets are riveted on this scaffolding all round, in order to prevent unequal cooling of the furnaces outside in case of rain, wind, or snow. The principal advantage of a kiln of this description is that, owing to the continuous and regular cooling from outside, the fritted clinker cannot clog the interior of the furnace, thus ensuring a regular and continuous working of the furnace. The ribs at the same time give strength, and prevent the cast-iron rings from warping. Each fur- nace produces about 25 tons of well-burnt clinker, equal to as much finished cement, in twenty-four hours. Reference list for fixed kilns. The design, construction, and opera- tion of vertical or stationary kilns of various types are discussed in many books and papers on Port land -cement manufacture. The most satisfactory of these discussions are included in the following annotated list of references on the subject: Butler, D. B. Portland Cement: its manufacture, testing, and use. 1899. Chapter IV of this volume, pp. 71-102, includes descriptions of the dome kiln, Johnson kiln, Batchelor kiln, Dietzsch kiln, and Hoffmann kiln. The discussion of the Johnson kiln and its modifications is par- ticularly valuable. CEMENT BURNING: FIXED KILNS. 479 Candlot, E. Ciments et chaux hydrauliques. 1898. Fixed kilas of various types are well described on pp. 53-71, inclusive. Lewis, F. H. The Candlot oscillating grate for cement kilns. Engineering Record, May 21, 1898. Description of a grate devised to improve draft and prevent balling in shaft kilns. Schoch, C. Die moderne Aufbereitung und Wertung der Mortel-Materialen. 1896. Pages 124-157 of this volume contain descriptions of various improved types of fixed kilns. Those of the Hoft'ma;nn > Dietzsch, Stein, Hanenschild, and Schofer kilns are particularly valuable. Scott, H. S. D., and Redgrave, G. R. The manufacture and testing of Portland cement. Proc. Inst. Civ. Engrs., vol. 62, pp. 67-86. 1880. The Johnson kiln and i.ts modifications are described in considerable detail on pp. 74-76. Stanger, W. H., and Blount, B. The rotatory process of cement-manufacture. Proc. Inst. Civ. Engrs., vol. 165, pp. 44-136. 1901. Valuable data on the design, construction, and results obtained from various types of fixed kilns will be found on pp. 44, 48, 81, 82, 99, and 100. Von Schwarz, C. The utilization of blast-furnace slag. Journal Iron and Steel Institute, 1900, No. 1, pp 141-152. 1900. The Schwarz kiln is de- scribed with figures. Zwick, H. Hydraulischer Kalk und Portland-Cement. 1892. Pages 148- 184 are devoted to discussions of kilns and burning practice. The Hoffmann ring kiln is described in great detail. CHAPTER XXXIII. THE ROTARY KILN. IN the early days of the Portland-cement industry a simple vertical kiln, much like that used for burning lime and natural cement, was used for burning the Portland-cement mixture. These kilns, while fairly efficient so far as fuel consumption was concerned, were expen- sive in labor, and their daily output was small. In France and Ger- many they were soon supplanted by improved types, but still stationary and vertical, which gave very much lower fuel consumption. Kilns of these types have been discussed in the preceding chapter. In America, however, where labor is expensive while fuel is comparatively cheap, an entirely different style of kiln has been evolved. This is the rotary kiln. With the exception of a very few of the older plants, which have retained vertical kilns, all American Portland-cement plants are now equipped with rotary kilns. FIG. 115. Exterior view of rotary kiln. (Bonnot & Co.) The rotary kiln as at first used in cement-manufacture was adapted to dry materials only, while gas or oil were used as fuel. A long series of experiments and improvements have perfected a burning process in which finely pulverized coal is used as fuel, while wet mixtures can now be fed directly to the kiln. The present condition, in which the rotary kiln is adapted to the use of several different types of fuel, and to all kinds of Portland-cement mixtures, has been attained only through long and earnest effort on the part of American cement-manufacturers. 480 THE ROTARY KILN. 481 The history of the gradual evolution of the rotary is of great interest, but as the subject cannot well be taken up here, reference should be made to the papers cited below,* which contain the details of this his- tory, accompanied in many cases by illustrations of early forms of rotary kilns. Summary of burning process. As at present used, the rotary kiln Is a steel cylinder, about 5 to 7 feet in diameter; its length for dry materials is 60 to 150 feet, while for wet mixtures an 80-foot or even longer kiln is commonly employed. This cylinder is set in a slightly inclined position, the inclination ^^ ~~*j FIG. 116. Plan and elevation of 60-foot rotary kiln. (Engineering News) being approximately one half inch to the foot. The kiln is lined, ex- cept near the upper end, with very resistant fire-brick, to withstand both the high temperature to which its inner surface is subjected and also the destructive action of the molten clinker. * Duryee, E. The first manufacture of Portland cement by the direct rotary kiln process. Engineering News, July 26, 1900. Eckel, E. C. Early history of the Portland-cement industry in New York State. Bulletin 44, New York State Museum, pp. 849-859. 1901. Lesley, R. V/. History of the Portland-cement industry in the United States. 8vo, 146 pp. Philadelphia, 1900. Lewis, F. *H. The American rotary kiln process for Portland cement. Cement Industry, pp. 188-199. New York, 1900. Matthey, H. The invention of the new cement-burning method. Engineering and Mining Journal, vol. 67, p. 555, 705. 1899. Smith, W. A. Manufacture of cement, 1892. Mineral Industry, vol. 1, pp. 49-53. 1893. Stanger, \V. H., and Blount, B. The rotatory process of cement-manufacture. Proc. Institution Civil Engineers, vol. 145, pp. 44-136. 1901. Editorial. The influence of the rotary kiln on the development of Portland-cement manufacture in America. Engineering News, May 3, 1900 482 CEMENTS, LIMES, AND PLASTERS. The cement mixture is fed in at the upper end of the kiln, while fuel (which may be either powdered coal, oil, or gas) is injected at its lower end. The kiln, which rests upon geared bearings, is slowly revolved about its axis. This revolution, in connection with the inclination at which the cylinder is set, gradually carries the cement mixture to the lower end of the kiln."' In the Bourse of this journey the intense heat generated by the burning fuel first drives off the water and car- bon dioxide from the mixture and then causes the lime, silica, alumina, and iron to combine chemically to form the partialjy fused mass known as "cement clinker". This clinker drops out of the lower end of the kiln, is cooled so as to prevent injury to the grinding machinery, and is then sent to the grinding mills. Shape and size. The rotary kilns in use at various plants differ considerably in both shape and size. FIG. 117. Driving mechanism of rotary kiln. (Engineering News.) As to shape, the simplest and commonest form is that of a cylin- der a straight tube of equal diameter throughout. At many plants, however, the kilns are wider at the lower or discharge end than at the stack end. This is usually accomplished by means of a reducing sec- tion near the middle of the kiln, so that a kiln of thi3 type would con- sist really of a lower section 6 feet in diameter and about 30 feet long, .an upper section 5 feet in diameter and 30 feet long, and an intermediate reducing section in the shape of a frustum of a cone. The theory on which this arrangement is nominally based is that the gases, cooling each exhauster being of sufficient capacity to operate all injectors, the other exhauster being held in reserve in case of accident to the first. THE ROTARY KILN. 487 The blast-fans are so arranged that the reserve fan could be placed in service in case of accident to the operating fan before kilns have time to cool to any great extent; after the kiln has cooled or when starting the kiln for burning clinker ; it requires several hours to heat DETAIL ELEVATION OF FEEDING MECHANISM DETAIL PLAN OF VARIABLE SPEED MECHANISM FIG. 121. Coal-burning arrangement, International Portland Cement Co. (B. F. Sturtevant Co.) kiln to a proper temperature for the complete combustion of the pul- verized coal. The blast-fans in .most cases, draw their supply of air directly from the kiln-room. In some plants, however, the inlet to blast-fans are connected with the clinker-coolers or the other parts of the 488 CEMENTS, LIMES, AND PLASTERS. THE ROTARY KILN. 489 cement apparatus, which have a tendency to heat large volumes of $ir by means of the galvanized steel- ducts, thus insuring air at a much higher temperature for use as blast to the injector. The air-blast of fans is distributed to the several injectors by means of galvanized steel piping J, usually located directly in back of injectors. The pulverized coal as it drops into injector is taken into suspen- sion by air-blast supplied to injector through blast-piping J and fed through wrought-iron feed*pipes K into discharge end of kiln. The feed-pipe usually enters kiln about 3 to 12 inches below the center. FIG. 123. Kirkwood gas-burner for rotary kiln. In extremely large kilns two feed-pipes are usually used so that cur- rent of air containing the pulverized coal can. be blown into different points of the kiln at the same time. The volume of air admitted to injector is regulated by blast-gate L. The length of the flame in the kiln is dependent upon the velocity of the aj r as it leaves the feed-pipe. Gas-burners for rotary kilns. Natural gas is at present utilized as a kiln fuel at several Kansas plants. An Ohio plant when last visitecj 490 CEMENTS, LIMES, AND PLASTERS. was running some of its kilns on natural gas and some on producer gas. At all these plants the Kirkwood burner, shown in Figs. 123, 124, is used in supplying gas to the kiln. The Kirkwood gas-burner is manufactured by Tate, Jones & Co., of Pittsburg, Pa., under patents granted to R. G. Kirkwood in 1896. It consists of two concentric cylindrical casings, which are bolted to- gether, forming an annular chamber. A large number of small pipes are set to form a spiral series, passing from one side to the other of the annular chamber. These pipes are provided with t a number of fine holes, and a nozzle caps the entire outfit. Gas is introduced through an opening into the annular chamber, from which it passes ' into the small pipes, issues through the holes in these pipes in a great number of fine jet" nd mixes with the air which is blown through the burner, Air Inlet Forced Draught 2 "Kirkwood Patent Natural Gas Burner FIG. 124. Kirkwood gas-burner applied to rotary kiln. (Tate, Jones & Co.) thus securing a proper mixture of air and gas to be burned at the noz- zle. A cast-iron plate filled with asbestos cement is provided for bolt- ing to the front of the kiln to receive the end of the burner, and the back of the burner has a projection into which the air-blast pipe is in- troduced. These burners are about 18 inches in diameter and 5 feet in length, and are designed to work efficiently with a gas pressure of from 3. to 4 ounces per square inch. Kiln linings. Three materials have been used for kiln linings; cement clinker, alumina brick, and magnesia brick. A fourth will probably be introduced in the near future bauxite brick. Of these lining materials, the use of alumina brick may be con- sidered to be the standard American practice. In the following tables analyses of these products are given. Table 181 contains analyses of clays used in the manufacture of high-alumina kiln brick, while analyses of the brick are given in Table 182. Tables 183 and 184 contain analyses of low-alumina clays and the resulting brick, which have been supplied for rotary-kiln linings at several plants. THE ROTARY KILN. 491 TABLE 181. ANALYSES OF HIGH-ALUMINA CLAYS USED FOR KILN BRICK. 1. 2. 3. 4. Silica (SiO 2 ) 4 338 44 52 43 05 40 30 Alumina (A1 2 O 3 ) . . . . , 40 35 40 81 44 60 45 00 Iron oxide (Fe 2 O 3 ) 85 1 03 2 60 n d Lime (CaO) 88 62 40 n d Magnesia (MgO) 23 55 20 n d Carbon dioxide (CO 2 ) Water ) 13.41 12.11 9.00 n. d. 5. 6. 7. 8. Silica (SiO 2 ) 40 80 42 -71 44 00 43 52 Alumina (A1 2 O 3 ) 49 00 38 88 42 12 42 18 Iron oxide (Fe 2 O 3 ) n d 3 36 86 42 Lime (CaO) n. d. 13 0.24 25 Magnesia (MgO) n d 00 10 16 Carbon dioxide (COy) 1 A Water f n. d. 15.19 14.20 14.31 1, 2. Olive Hill, Carter County, Ky. Analyses from Stowe-Fuller Co.'s catalogue, p. 25. 3, 4, 5. Hayward, Carter County, Ky. Ironton Fire-brick Co. 6. Carter County, Ky. Chas. Taylor's Sons. F. W. Clarke, analyst. Specimen selected by E. C. Eckel. 7. Lock Haven, Pa. P. L. Hobbs, analyst. Stowe-Fuller Co.'s catalogue, p. 26*. 8. " " , " Crowell and Peck, analysts. Stowe-Fuller Co.'s catalogue, p. 26. TABLE 182. ANALYSES OF HIGH-ALUMINA FIRE-BRICK FOR KILNS. 1. 2. 3. 4. Silica (SiO 2 ) . . . 54 86 52 64 49 70 54 03 Alumina (A1 2 O 3 ) I 42.74 40 45 Iron oxide (Fe 2 O 3 ) 44.84 47 . 86 < 3.47 Lime (CaO) 1 30 84 0.80 0.31 Magnesia (MgO) 62 0.88 0.80 tr. 5. 6. 7. 8. 9. Silica (SiO 2 ) 55.22 54.38 58.90 51.95 56.44 Alumina (Al O 3 ) 41 51 41 72 36.30 45.01 35.81 Iron oxide (Fe 2 O 3 ) 2 84 3.20 2.01 4.79 Lime (CaO) 1.00 1.60 0.04 n. d. Magnesia (MgO) tr. tr. 0.29 n. d. 1. 2. "Tyrone" brick, Harbison-Walker Co. H. S. Turner, analyst. 3. Kentucky Fire Brick Co. H. S. Turner, analyst. 4. Ironton Fire Brick Co. F. W. Clarke, analyst. Specimen selected by E. C. Eckel. 5. " " " " Analysis quoted by manufacturers. 6. 7. Christy Fire Brick Co. 8. "Munro" brick, Stowe-Fuller Co. P. L. Hobbs, analyst. Catalogue, p. 70. 9. Stowe Fuller Co. E. Davidson, analyst. 492 CEMENTS, LIMES, AND PLASTERS. TABLE 183. ANALYSES OF LOW-ALUMINA CLAYS USED FOR KILN BRICK. Silica (SiO 2 ) Alumina (A1 2 O 3 ) Iron oxide (Fe 2 O 3 ) . . . Lime (CaO) Magnesia (MgO) Alkalies (K 2 O,Na 2 O). Carbon dioxide (CO 2 ). 55.0 30.0 tr. tr. tr. 2.0 vydi wii U.IVJA.IU.C \\^\s 2 J t -lo r\ Water ! J 13 ' 56.02 28.26 2.18 2.04 0.95 n. d. 10.50 TABLE 184. ANALYSES OF LOW-ALUMINA BRICK, FURNISHED AS KILN BRICK. i. 2. 3. 4. 5. Silica (SiO 2 ) 62 58' 63 94 61.20 62 92 72 71 Alumina (A1 2 O 3 ) 25 62 30.14 29 05 30 47 22 24 Iron oxide (Fe 2 O 3 ) 4 76 3.70 5.55 4.61 4 47 Lime (CaO) 6 05 2.20 n. d. n. d. 94 Magnesia (MgO) 85 tr n d n d 42 The manner in which these bricks are set in lining kilns is shown in Figs. 125 and 126. Actual fuel consumption and output. In the following chapter the question of heat requirements and heat distribution in the rotary kiln will be discussed in considerable detail. At present it is only necessary to state that in burning a dry mixture to a clinker, practi- cally all of the heat consumed in the operation will be that required for the dissociation of the lime carbonate present into lime oxide and carbon dioxide. Driving off the water of combination that is chem- ically held by the clay or shale, and decomposing any calcium sulphate (gypsum) that may be present in the raw materials, will require a small additional amount of heat. The amount required for these purposes is not accurately known, however, but is probably so small that it will be more or less entirely offset by the heat which will be liberated during the combination of the lime with the silica and alumina. We may, therefore, without sensible error regard the total heat theoretically required for the production of a barrel of Portland cement as being that which is necessary for the dissociation of 450 Ibs. of lime carbonate. With coal of a thermal value of 13,500 B.T.U. per pound, burned with only the air-supply demanded by theory, this dissociation would require about 25J Ibs. of coal per barrel of cement, a fuel consumption of only 6J per cent on the weight of cement produced. In actual practice, however, the heat required for cement produc- THE ROTARY KILN. oa "* "' ^ I CL ~ fe: o s R F ?ii 3 ^ 5- ^ o w g 8 g. H * 494 CEMENTS, LIMES, AND PLASTERS. tion is immensely greater than that demanded by theory. This is due to the fact that heat is wasted or lost in various ways during the process of burning in the rotary kiln. The more important losses of heat occur from the fact that the stack-gases and clinker are usually discharged at high temperatures; that the air-supply injected into the kiln is always greater, and usually much-' greater, ^han that theoretically necessary; and that much heat is lost by radiation from the exposed surface of the kiln. Sixty-foot rotary kilns are nominally rated + at a production of 200 barrels per day per kiln. Even on dry materials and with good coal, however, such an output is rarely attained. Normally a 60-foot kiln working on a dry mixture will produce from 140 to 180 barrels of cement per day of twenty-four hours. In doing this, if good coal is used, its fuel consumption will commonly be from 120 to 140 Ibs. of coal per barrel of cement, though it may range as high as 160 Ibs., and, on the other hand, has fallen as low as 90 Ibs. An output of 160 barrels per day, with a coal consumption of 130 Ibs. per barrel, may therefore be considered as representing the results of fairly good practice on dry materials. With longer kilns, however, much better results are obtained, as will be noted later. In dealing with a wet mixture, which may carry anywhere from 30 to 70 per cent of water, the results are more variable, though always worse than with dry materials. In working a 60-foot kiln on a wet material, the output may range from 80 to 140 barrels per day, with a fuel consumption of from 150 to 230 Ibs. per barrel. Using a longer kiln, partly drying the mix, and utilizing waste heat will, of course, improve these figures materially. When oil is used for kiln fuel, it may be considered that one gallon of oil is equivalent in the kiln to about 10 Ibs. of coal. The fuel con- sumption, using dry materials, will range between 11 and 14 gallons of oil per barrel of cement; but the output per day is always some- what less with oil fuel than where coal is used. Natural gas in the kiln may be compared with good Pennsylvania or West Virginia coal by allowing about 20,000 cubic feet of gas as equivalent to a ton of coal. This estimate is, however, based upon too little data to be as close as those above given for oil. The figures given in Table 185, below, are believed to be entirely reliable. They are of interest as showing what can actually be expected from kilns under average management, as distinguished from the expectations which embellish company prospectuses and the reports of "cement experts." With the exception of A, B, and J, the mills THE ROTARY KILN. 495 here listed are good average plants. Mill results A and B are from one of the best of the Lehigh district plants, while J is perhaps the best of all marl-plants. Excluding these three, it will be seen that the production per kiln per day is considerably lower, and the fuel con- sumption much higher, than is usually allowed for. TABLE 185. ACTUAL OUTPUT AND FUEL CONSUMPTION AT VARIOUS PLANTS. Mill. Materials. Process. Per Cent of Water in Mixture. Length of Kiln, Feet. Output per Day, Barrels. Coal per Barrel, Pounds. A Cement rock Dry 60 225 105-113 B it t ( 80 260 95-100 c tt 60 160-180 109-175 D tt (i 150 350 62-75 E F Limestone and shale. .... 1 1 1 1 a 0. 60 60 170 ? 160 105 G (( ( ( 60 ? 122 H I J it tt ii tt Marl and clay ..... ..... Wet 60 60 60 110 185 170 135 130 135 ? K Marl and shale 50 70 145 ? L 1 1 tt 1 1 60 60 85 ? M 1 1 1 1 t < 50 60 120 ? N Marl and clay 30 60 125 173 o 1C t t (I 35 60 100 160 p It It 11 65 60 80 180-210 Q Limestone and clay 35 60 ICO 200-220 R t < it 30 60 140 180 Using the figures above given as a basis, Table 185 has been con- structed to give some idea of what may fairly be expected from kilns of various length working on different raw materials. TABLE 186. AVERAGE OUTPUT AND FUEL CONSUMPTION. Process. Raw Materials. Length of Kiln, Feet. Output per Kiln per Day, Barrels. Coal Consumption per Kiln per Barrel, Pounds. Range. Aver- age. Range. Aver- age. Wet < . Dry i Marl and clay 60 80-90 110* 60 60 80 150* 60-140 80-150 135 150-200 180-250 225-300 375 85 100 135 160 200 260 375 150-250 140-220 150 90-170 85-160 85-120 65 200 160 150 130 115 110 65 < 1 1 tt Limestone and clay Cement rock and limestone. . t i 1 1 tt' it it 1 1 Based on only one plant. 496 CEMENTS, LIMES, AND PLASTERS. The differences in composition between Portland-cement mixtures are very slight if compared, for example, to the differences between various natural cement rocks. But even such slight differences as do exist exercise a very appreciable effect on the burning of the mix- ture. Other things being equal, any increase in the percentage of lime in the mixture will necessitate, a higher temperature in order to get an equally sound cement. A mixture which will give a cement carry- ing 59 per cent of lime, for example, will require much less thorough burning than would a mixture designed to give a cement with 01 per cent of lime. With equal lime percentages, the cement carrying high silica and low alumina and iron will require a higher temperature than if it w r ere lower in silica and higher in alumina and iron. But, on the other hand, if the alumina and iron are carried too high, the clinker will ball up in the kiln, forming sticky and unmanageable masses. CHAPTER XXXIV. HEAT CONSUMPTION AND HEAT UTILIZATION. AN investigation of the ways in which the heat supplied to the kiln is utilized and wasted is a matter of both theoretical and practical importance. It can readily be seen that until some idea can be gained of the relative importance of the different causes of loss of heat, little can be done to prevent this -waste or to utilize the heat so dispersed. An exact knowledge of the distribution of the total heat supplied to the kiln would therefore be of great service to the manufacturer. In the present chapter the writer has attempted to present such data on this subject as are available, and to discuss them in such a way as to bring out the relations of the various factors in the problem of heat distribution. Attention is drawn, whenever necessary, to any doubts as to the accuracy of the data employed. Theoretical Heat Requirements. In order that a raw mixture shall be converted into cement clinker in the kiln, sufficient heat must be applied to bring about the necessary physical and chemical changes. The purposes for which this heat is required are: (1) Evaporation of the water of the mix. (2) Decomposition of the clay. (3) Dissociation of sulphates. (4) Dissociation of carbonates. (5) Heating the mix to clinkering point. Of these five requirements, it is to be noted that the first four are for accomplishing chemical changes, and that the heat supplied for these purposes is entirely absorbed in doing chemical work. This is not true with regard to the fifth requirement the heating of the mix for the heat used for this purpose, after it has once served its purpose, still remains as sensible and therefore utilizable heat. Most of it, in fact, passes out in the clinker. In a perfect kiln the only heat required would be that sufficient to accomplish the first four operations in the above list, for in a theo- retically perfect burning device there would be no loss by radiation, 497 498 CEMENTS, LIMES, AND PLASTERS. the stack-gases would be cold, and the clinker heat would be utilized. In actual practice, however, a very large amount of heat is carried out with the stack-gases, radiated from the exposed surfaces of the kiln, and carried out in the hot clinker. Heat utilized in evaporation of water. It is obvious that any water contained in the charge must be evaporated, and the steam thus formed must be raised to the temperature of the stack-gases. It is here that the great difference in economy between the dry and wet methods of mixing is shown. In the dry method the total water (mechanically held and combined) contained in the charge will rarely rise above 2^ per cent, of which about 2 per cent may be combined in the clay and ^ per cent held mechanically. The products in the dry process, when working with a 60-foot kiln, issue from the stack at a temperature of about 1500 F. = 815 C. When a longer kiln is employed, and the trend of present practice seems to be in the direction of 100-foot or even longer cylin- ders, the stack temperatures will be correspondingly reduced. With a 100-foot kiln it seems probable that they can be kept down at least to 1000 F. In the wet process, on the other hand, the charge usually contains about 60 per cent of water, though in a few plants this is kept down to 30 or 40 per cent. The stack temperatures are, however, much lower than m the dry process, ranging from about 800 F. with a 60-foot kiln to 450 or so in a 100-foot kiln. This partly counterbalances the loss of heat due to the high percentage of water. Using these data as a basis, Table 187 has been prepared to show the amount of heat required for simply evaporating the water from three different types of mixture, in kilns of two different lengths. TABLE 187. HEAT USED IN EVAPORATION OF WATER. Process. Kiln Length, Feet. Stack Temperature. Per Cent Water. Pounds Water per Barrel. B. T. U. Used per Barrel. Dry 60 1500 F 2* 15 23 367 y " 100 1000 F 2* 15 21 079 Wet . 60 800 F 30 272 365 650 < ( 100 450 F 30 272 336 630 < < 60 800 F 60 900 1 209 870 i 100 450 F 60 900 1 113 840 Heat utilized in decomposition of clay. An unknown, though prob- ably small, amount of heat is required to dissociate the clayey portion of the mix. No exact data on this point are known to the writer, but HEAT CONSUMPTION AND HEAT UTILIZATION. 499 the amount so utilized will probably be covered if we estimate all the water which is really chemically combined with the clay as being mechan- ically held water. This course has been followed in the present esti- mates. On this assumption, even a dry mix will carry about 2J per cent of water, and this amount has been accordingly allowed for in the previous paragraph and in Table 187. Heat utilized in dissociation of sulphates. A certain amount of heat is taken up in dissociating any lime sulphate (gypsum) present in the raw mix. Newberry has taken this as requiring 1890 B.T.U. per pound of SOs. In marl plants the percentage of sulphates present may rise to notable quantity, but in most other plants they are negligible. In the present discussion the assumptions will be made that the average dry mix carries 0.3 per cent of sulphur trioxide, and that the usual wet mix may carry 1 per cent. The total amount of heat required for the dissociation of sulphates will therefore be: Dry mix 600 lbs.X0.3%Xl890= 3,402 B.T.U. per barrel Wet mix " " X 1.0% X 1890 = 11, 340 " " " Heat utilized in dissociation of carbonates. The most important heat requirement by far is that for the dissociation of the carbonates of the charge. The values assumed by Richards for the dissociation requirements of the two carbonates are: Liberation of 1 kilo CO 2 from CaCO 3 = 990 calories. ' 1 " CO 2 " MgCO 3 = 407 " These are referred to Berthelot. They correspond respectively to the two values of: Dissociation of 1 pound CaCO 3 requires 584 B.T.U. " 1 " MgC0 8 " 381 " These values will be accepted in the following calculations for the sake of uniformity, though Ostwald * quotes from Thomsen a value corresponding to 765 B.T.U. for the dissociation of 1 Ib. of lime carbonate. If this latter value were accepted, the quantities given in the table below (188) should be reduced about 2J per cent. Other values for these dissociation constants have been quoted by various authorities, with a much wider range, but for the present purpose those first noted will be satisfactory enough. Temperature required for clinkering. Widely differing statements have been made as to the temperature required in order to clinker the average Portland-cement mixture. Carpenter, in testing the Cayuga plant noted later, determined * Lehrbuch der allgemeinen Chemie, vol. II, pt. 1, p. 272. 500 CEMENTS, LIMES, AND PLASTERS, TABLE 188. HEAT USED IN DISSOCIATION OF CARBONATES PER ^BARREL CEMENT. Percentage MgO in Mixture. Percentage of Lime (CaO) in Mixture. 40%. 41%... 42%. 43%. 44%. " 45%. . : 1% -.-'. i 2% 3% 4% B.T.U. 333,234 : 338,362 342,758 347,154 351,550 B.T.U. 342,144 347',272 351,668 356,064 360,460 B.T.U. 351,054 - .356,182 360,578 364,974 369,370 B.T.U. 359,964 - 365,092 369,488 373,884 378,280 B.T.U- 368,874 374,002 378,398 * 382,794 387,190 B.T.U. 377,784 382,912 387,308 391,704 396,100 : the kiln temperature by optical methods. The temperature in the j kiln when .working under best conditions, as determined by the "Noel j optical pyrometer, varied from 2250 F. near the discharge' 'end 1 to 2950 F. about 20 feet from the lower end, and about 1800 F. at -ihe : upper end. The temperature in the burning zone seemed to average about 2850 F;, and the temperature of the entire kiln on thejnside seemed - to average nearly 2500 F. - -For ordinary purposes of calculation, it may be assumed that 1400 1500 C., or 2500-2700 F., is about the necessary temperature* in rotary ' -kilns under present 'conditions for an average mixture'.' 'Variations in the composition of the mixture would, of 'course; change the clinker- ing point for a low-limed high-alumina mix will clinker at a consider- ably, lower temperature than will a mix high in lime and silica. It is also true that to a certain degree longer exposure 'to the heat will be equivalent in effects to higher temperature. In stationary kilns, for example, where the charge may be exposed for days to the heat, the requisite temperature is much less than in the modern rapid 'practice with the rotary kiln. Heat utilized in heating the mix. One of the important uses of the kiln heat is in simply heating the mix up to the point at which it : will clinker. Fortunately, this can be determined with sufficient ac- curacy for all practical purposes. Whatever the percentage of the water present, the dry portion "of the mix will be about 600 Ibs. for each barrel of cement. This 600 Ibs. of material must be raised from the temperature of the air say 60 F. to about 1300 F. At this latter temperature the carbon dioxide, ' sulphur trioxide, etc., will have been driven off; and this will reduce ..the weight to 380 Ibs. This 380 Ibs. of quicklime and clay must now be raised to a temperature of about 2600 F., at which clinkering. will take place. HEAT CONSUMPTION AND HEAT UTILIZATION. 501 Assuming that the above data are substantially correct, and that the specific heat of the mix is ab'out 0.22, the heat required for the simple,, heating of the mix to the clinkering point can be calculated ass follows: '". Per Barrel B.T.U. 600 Ibs. mix heated 60 to 1300, spec, heat 0.22 = 1260X0.22X600 = 166,320 380 " " " 1300 to 2600, " " 0.22 = 1300 X0*;22x 380 -108,680 Total heat required, B,T.U., ."..........;... 275,000 This estimate is probably aboypjwhal is actually required, for the temperatures, weights and specific ..heat have all been taken on the safe side. The actual heat requirements are probably close to 250,000 B.T.U. for this part of the operation. In running an actual test of a kiln this quantity could tie checked roughly by the amount of, heat contained in the clinker as it leaves the kiln. In other words, a barrel of, clinker carries out ..with it almost as much heat as. was required to clinker the raw mix for that barrel. >As pointed out on a previous ... page (p. 497), the. heat/ required for bringing the mix up to "the clinkering point" is not "utilized -in causing chemical changes, .and can therefore be utilized. again. In this" 'respect it differs from the neat required for dissociating the carbonates and sulphates,- decomposing the clay, etc. for in ' these _ cases' the' heat is absorbed in doing chemical work and' cannot be regained. For this reason it will be convenient to omit, .from the total .thermal re- quirements, the heat used in heating the mix up to the clinkering point; and to consider it rather in its- outgoing form as heat carried '/put by the clinker. Total heat requirements. The data given in preceding paragraphs may now be conveniently summed up as in the table below. The basis for the various figures may be seen by referring back to the upper pages. TABLE 189. THEORETICAL HEAT REQUIREMENTS IN B.T.U. PER BARREL. Process Dry Dry Semi-wet Wet' Length of kiln .......'....... 60 ft. 100 ft. 60 ft. 106 ft. Water in mix ' . 2\7 2 % 30% *i --60%- Stack-gases 1500 F. 1000 F. 800 F. -450-F; Evaporation of water B.T:U. 23,367 B.T.U. 21,079 B.T.U. 365,650 B.T.U. 1,113,840 Dissociation of sulphates . , 3,402 3402 11,340 11,340 Dissociation of carbonates 369,488 369,488 369,488 369,488 Total heat required 396,257 393,969 746,478 I r 494,668 Coal theoretically necessary, Ibs. per bbl. . . Coal actually used, Ibs. per bbl 28 120 28 90- 53 ' 46Q : '>: 107 150 502 CEMENTS, LIMES, AND PLASTERS. Heat Losses in Practice. In practice with the rofary kiln, there are a number of distinct sources of loss of heat, which result in a fuel consumption immensely greater than the theoretical requirements given above. The more important of these sources of loss are the following: (1) The kiln gases are 'discharged at a temperature much above that of the atmosphere, ranging from 300 F. to 2000 F., according to the type of materials used and the length of the kiln. % (2) The clinker is discharged at a temperature varying from 200 F. to 2500 F., the range depending as before on materials and length of the kiln. (3) The air-supply injected into the kiln is always greater, and usually very much greater, than that required for the perfect combus- tion of the fuel, and the available heating power of the fuel is thereby reduced. (4) Heat is lost by radiation from the ends and exposed surfaces of the kiln. (5) The mixture in plants using a wet process carries a high per- centage of water, which must be driven off. It is evident, therefore, that present-day working conditions serve to increase greatly the amount of fuel actually necessary for the pro- duction of a barrel of cement above that required by theory. The extent of these losses, compared with the amount of heat ac- tually used, can be seen from* the following comparison of various estimates and tests, all relating to a 60-foot kiln on dry material: TABLE 190. UTILIZATION AND LOSSES OF HEAT IN ROTARY KILNS. Richards. Carpenter. Helbig. Newberry. Eckel. Average. Total heat supplied to kiln Per Cent. 100 00 Per Cent. 100 . 00 Per Cent. 100.00 Per Cent. 100 . 00 Per Cent. 100 00 Per Cent. 100 00 Heat utilized 19 75 23 . 43 25.56 25.54 23.59 23 57 Heat lost in clinker .... 10 72 14 09 12 01 15.47 13 07 Heat lost in stack-gases . Heat lost by radiation. . . Minor heat losses 72.46 -4.46 1 54 47.42 15.07 00 50.24 4.88 7 31 43.62 15.38 00 76.41 53.43 7.72 2 21 Heat carried out in flue-dust. A considerable amount of fine dust is carried out of the kiln by the hot gases. This flue dust, deposited wherever the air current is checked, may amount to from J to 3 per cent of the total amount of mix charged to the kiln. HEAT CONSUMPTION AND HEAT UTILIZATION. 503 The composition of the flue-dust is a matter of considerable indus- trial importance. It is composed of the lighter and finer particles of the cement mix and the ash, plus a certain amount of material deposited from the stack-gases. This last factor includes in some cases a large percentage of alkali salts, whose recovery has been suggested as a profit- able by-product. (See p. 510.) Sources of Heat-supply. To counterbalance the heat utilized and the heat wasted, as above noted, heat is always supplied to the kiln from two sources, and occa- sionally from two other sources. The invariable sources of supply are: (1) A large and well-known supply is derived from the combustion of the fuel fed to the kiln. (2) A smaller and very poorly defined supply is obtained from exothermic chemical combinations which take place in the kiln during clinker ing. Supplies from these two sources are necessarily received in every kiln. In addition, however, heat may be supplied from (3) Regeneration of the clinker heat. (4) Utilization of the heat in the stack-gases. Heat supplied by combustion of fuel. The most important source of the heat supplied to the kiln is, of course, the burning of the fuel injected into it. This can be estimated accurately enough, for any given kiln, if the composition of the coal and the amount of coal used per barrel of cement are known. It must be borne in mind, however, that any defects in the coal-feeding arrangements, or deficiencies in the fineness of coal grinding, should not properly be charged against the efficiency of the kiln, but against the efficiency of the superintendent. In calculating the heat supplied to the kiln by combustion of fuel the assumption is always made that the coal is ground as fine as is econom- ically possible, and that the injecting apparatus gives perfect combus- tion. Actually we know that neither of these assumptions is ever quite justified and that in some mills both are very incorrect. If the best bituminous coal from western Pennsylvania or West Virginia be used, a theoretical heating value of 14,000 B.T.U. per Ib. may be assumed : but the coals used in practice often fall very far short of this. Such a coal, used at the rate of 120 Ibs. per barrel of cement, would give a heat supply of 1,680,000 B.T.U. per barrel. This is prob- ably about equal to the average practice with 60-foot kilns on a dry mixture of limestone and clay. With longer kilns, under specially 504 CEMENTS, LIMES, AND PLASTERS favorable circumstances, a fuel consumption of 90 Ibs. per barrel may be expected, corresponding to a heat supply of 1,260,000 B.T.U. per barrel. With the wet process a fuel consumption of 160 Ibs. per barrel is rather better than the average. This corresponds to a heat supply of 2,240,000 B.T.U. per barrel. These three estimates have therefore be used in making up the summary ^able. Heat supplied by chemical combinations. It is undoubtedly true that a considerable quantity of heat must be liberated when the lime and magnesia combine, at the clinkering temperature, with the silica, .alumina and iron oxide, and that in this way considerable heat is added to that derived from the fuel. Unfortunately, however, we have no very definite knowledge as to the exact chemical combinations which take place during clinkering, and lacking such knowledge any estimate of the amount of heat thus liberated must be considered as merely a wild guess. Both Helbig and Richards, in the papers previously cited, have quoted Berthelot on this point as giving the following data for the heat .liberated during this combination. 1 kilogram lime (GaO) liberates. 530 calories 1 ' ' magnesia (MgO) liberates. .... 827 ' ' , These figures, changed into English measures, are: 1 pound lime (CaO) liberates . . 954 B.T.U. 1 " magnesia (MgO) liberates. . , 1489 : rt For convenience these figures might be adopted in discussion, but both the reader and experimenter must bear in mind that they represent very doubtful assumptions, and are accepted merely because no better data are obtainable. In the present discussion of the subject no esti- mate of this type will be used. Heat derived from the clinker. A large part of the heat carried out in the hot clinker may be used to heat the incoming air. In Carpenter's experiments a little less than half of the clinker heat was thus utilized, but other experimenters have claimed 80 to 90 per cent efficiency for various types of clinker-heat regenerators. The amount of heat thus returned to the kiln might therefore vary from 90,000 to 175 ; 000 B.T.U. per barrel of cement. Heat derived from the stack-gases. Heat may also be taken from the stack-gases and used to heat either the raw material or the air- supply. Usually, however, stack-gas heat when utilized is used in the power department of the mill, rather than in the kiln. HEAT CONSUMPTION AND HEAT UTILIZATION. 505 Estimates and Tests of Heat Distribution. Various estimates of the heat requirements of cement-manufacture have been presented by different authors, and several actual tests have been made of heat distribution in the rotary 'kiln. The principles on which these calculations are based have been discussed in the preceding pages, and the estimates and tests in question will now be presented for comparison. Newberry's estimates. Some years ago Prof. Newberry published a discussion of the question of fuel consumption which leaves little to be desired even in spite of recent changes in rotary practice. His Tesuits are summarized in Table 191. . . ,,. t . TABLE 191. NEWBERRY'S ESTIMATES ON HEAT DISTRIBUTION IN KILNS. .- Vertical Kiln. Rotary Dry Pro'cess. Rotary Wet Process. B.T.U. Per Cent. B.TrUv Per Cent. B.T.U. I Per Cent. Evaporation of water. . . . . . 14,498 11,340 344,250 3.7 2,8 88.9 20,832 11,340 344,250 . 228,000 . ! 24, 480 3.3 1.8 54.7 ; 36.3) 3V&J; -827,424 11,310 344,250 .213,312 59.3 0.8. 24.6 15.3 Liberation of sulphates. .... Dissociation of carbonates. . Heating iof mix and clinker Heating of CO 2 and SO 3 from mix ' 16,646 4.6 Total B.T.U. required Lbs. coal required per bbl. , theoret. air-s'ply 'Lbs. coal required per bbl., 50% excess air. . Lbs. coal actually used in practice; 386,734 31.0 32.1 42-46 100,0 ! 628,902 66.9 82 2 100,0 1,396,326 120.0 128. 150-160 100.0 110-120 Helbig's estimates. Very recently Mr. A. B. Helbig has discussed this question, but only incidentally to a subject of more importance, i.e., the utilization of waste heat. Mr. Helbig's figures, slightly rearranged for convenience of comparison, will be found in Table 193 on page 509. Results of actual tests. It might be supposed that actual tests of the thermal efficiency of the rotary kiln could be readily made,, and that the results of these tests would afford data of great value to the manufacturer. To a certain extent this is true, but, unfortunately, the results afforded by such tests require interpretation, and this in turn requires that certain chemical constants so called by courtesy should be employed as bases. These constants are, for example, the heat of dissociation of the carbonates and the sulphates, of the decomposition of clay, of the formation of lime silicates and aluminates, etc. The 506 CEMENTS, LIMES, AND PLASTERS. error into which most experimenters fall is to assume that these con- stants are quite accurately known. As a matter of fact, even the sim- plest of them the heat of dissociation of lime carbonate is given a variation of almost 50 per cent by different chemists of about equal standing; while the heats of formation of the silicates, etc., are much less certain constants. * In reporting and discussing actual tests, or in reading the reports of such tests, it must, therefore, be borne in mind that the assumptions which are necessarily made are based, in large p#rt, on determinations of more than questionable accuracy. Two such tests have been recently published, by Richards and Car- penter respectively, and are summarized below. Richards' tests. Prof. J. W. Richards tested a 60-foot rotary at the Dexter Portland Cement Company plant, Nazareth, Pa. The cement mixture and the resulting clinker are said to have had the following compositions : Raw Mix. Clinker. Silica (SiO 2 ) 13.38 21 .27 Alumina (ALO.) \ A n . / 6 . 42 Iron oxide (Fe 2 O 3 ) J 1 3. 18 Lime (CaO) 41.96 66.70 Magnesia (MgO) 1 .53 2.43 Carbon dioxide (CO 2 ) 34 . 65 Water 0.43 The clinker " analysis " must evidently have been calculated from the raw mix, and not obtained by direct analysis. A proximate analysis of the kiln coal bituminous stack from Fair- mount, W. Va. gave: Volatile matter 38. 10 Fixed carbon 53 . 24 Ash 8.06 Moisture . 60 "The following ultimate composition of the coal was assumed from average analyses of coal from that region of similar proximate com- position. Carbon 73 . 60 Hydrogen 5 . 30 Nitrogen 1 . 70 Sulphur 0.75 Oxygen 10 . 00 Moisture . 60 Ash.. 8.05 HEAT CONSUMPTION AND HEAT UTILIZATION. 507 "The kiln turns out an average of 3635 Ibs. of clinkered cement per hour from 5980 Ibs. of material fed to it, producing 200 Ibs. of flue- dust, equal to 3.35 per cent of the weight of mixture charged. The coal used averages 110 Ibs. per barrel of cement produced." The temperature of the clinker falling out of the lower end of the kiln was measured by the Le Chatelier pyrometer, and determined to be 1200 C. = 2192 F. The temperature of the waste gases, deter- mined 4 feet below the top of the stack, was 820 C. or 1508 F. The sensible heat in the clinker (leaving the kiln at 1200 C.) was determined by a calorimeter as 290 kilogram calories per kilo = 522 B.T.U. per pound. The waste gases in the stack analyzed as follows: Carbon dioxide 10 . 2 Oxygen 11.8 Carbon monoxide 0.2 Sulphur dioxide not determined Water Nitrogen It will be noted that part of these preliminary data were deter- mined by direct experiment, while others apparently are " averages/' or otherwise of less value than experimental results. This, unfortu- nately, throws doubt upon some of the results obtained, as noted below. At the close of his paper, after making the necessary calculations, Mr. Richards summarized his results. This summary, recalculated to calories and B.T.U. per barrel, is presented below. TABLE 192. SUMMARY OF RICHARDS' TESTS OF ROTARY KILNS. Heat-units per Barrel. Calories. B.T.U. Per Cent. HEAT SUPPLY. Theoretical heating power of the fuel 395,000 1,567,460 84 7 Heat of combination of the clinkering materials. . . . 71,410 283,355 15.3 HEAT DISTRIBUTION. Heat carried out by hot clinker 466,410 50025 1,850,815 198,499 100.0 10 7 __ . ( in necessary products 170,000 674,560 36 1 Heat in waste gases | | n J^SS P e tc 168,000 666,625 36 Heat in the flue-dust 1,056 4,190 0.2 Loss bv imperfect combustion .... 6,124 24,300 1 3 Evaporation of water of charge 723 2,869 2 Dissociation of the carbonates 10,814 42,910 2 3 Loss by radiation etc (by difference) . 59 668 236 862 12 8 466,410 1,850,815 100.0 508 CEMENTS, LIMES, AND PLASTERS. In regard to Richards* 'results it may be said that the use of such a large excess of air is not normal practice, either in the Lehigh district in general or at the Dexter plant in particular. It is further doubtful whether a kiln run so wastefully as this one appears to have been could make good cement with a fuel consumption as low as 110 Ibs per barrel. These questions throw doubt. on the Calculated loss of heat in the waste gases. The amount allowed for dissociation of the carbonates is appar- ently only about one-tenth of what should be allowed, owing to an arithmetical error. When this error is corrected, ^he "loss of heat by radiation" is made a minus quantity. In Table 193 below, this cor- rection has been made, but Richards' estimates as to waste gases are left unchanged. Carpenter's tests. Prof. R. C. Carpenter tested two rotary kilns at the plant of the Cayuga Portland Cement Company, near Ithaca, N. Y. The test was made primarily to determine the efficiency, not of the kilns, but of a boiler designed to utilize their waste heat. The coal used in the kilns was Westmoreland (Pa.) slack of the following composition and heating value. Moisture , 2 . 19 Volatile matter 32 . 9 Fixed carbon 54 . 66 Ash 10.25 B.T.U. per pound 14,022 At the time of test the two kilns were taking together 1889 Ibs, coal per hour, producing 21.2 barrels of clinker, equal to a coal con- sumption of 89.1 Ibs. per barrel. This low fuel consumption is attained in part by the use of waste heat from the clinker as shown in the table below. Carpenter's paper, as originally published, contained a number of serious typographic errors, which the author has kindly corrected on the copy sent to me. In the table below I have therefore made use of these corrected results, so that the second column of this table (193) will be found to differ considerably from that given in the original. It will be seen that Richards' results, when corrected for the carbon- ate requirements, leave no room for radiation losses. For my own detailed estimates on most of these points, the r is referred back to pages 498, 501. HEAT CONSUMPTION AND HEAT UTILIZATION. 50$ TABLE 193. TESTS AND ESTIMATES OF HEAT DISTRIUBTION, B.T.U. PER BBL. Richards. Carpenter. Helbig. Newberry. Heat from combustion of coal ' ' drawn from clinker cooler 1,567,460 1,247,641 96,273 992,000 149 348 1,474,000 " derived from chemical com- bination 283,355 132,456 240 770 1 Total heat supplied 1,850,815 1,476,370 1 382 118 1 474 000 Heat used in evaporation of water . . " *'-." dissociation of sul- phates 2,869 14,302 11,233 3,214 4,500 20,832 11 340 tt tt dissociation of car- bonates 362 524 320 253 345 625 344 250 Heat discharged in clinker * 198 499 207 999 165 942 228 000 " " " stack-gases, nec- essary products 674 560 {462 940 Heat discharged in stack-gases, ex- cess air 666 625 700,093 694,400 180 000 Heat discharged as CO (imperfect combustion) 24 300 Heat discharged in flue dust 4 190 14 134 by radiation, etc 222 490 f 67 456J 226 638? Total heat distributed 1,933,567 1,476,370 1,295,271 1,474,000 * Roughly equivalent to the heat necessary to bring the mix up to the clinkering point. A By difference. By calculation. Does not check, owing in part to temperature allowances. Heat Utilization and Economics. Much of the heat carried out by the clinker and the stack-gases is : recoverable with some ease, while that lost by radiation from the kiln is not so readily utilized. Helbig and Carpenter have described methods of waste-heat utilization in the papers cited on p. 511, to which refer- ence should be made for further details. Carpenter, in discussing his Cayuga tests, notes that "at the time of the test that portion of the air not supplied by the coal-feeding appa- ratus was forced by a special blower through the hot clinker and thence into the kiln. By this regenerative action about 80 per cent of the enter- ing air was heated to 480 F., restoring to the two kilns about 2,000,000 B. T. U. per hour, or about 7 per cent of the heat produced by the com- bustion of the coal. The regenerator, while distinctly economical, made the clinker elevators difficult to keep in order and tended to deliver dust into the kiln-room; it also took up valuable room and after a. few months of use was abandoned. The test shows, however, the: 510 CEMENTS, LIMES, AND PLASTERS. value of conserving the waste heat from the clinker by heating the entering air." As the clinker of the two kilns during this test carried out 4,409,540 B.T.U. per hour, and the clinker regenerator returned 2,041,000 of this, its efficiency was 46.3 per cent. At a German Portland-cement plant where the stack-gases are used in drying the raw materials, a large. amount of very fine dust settles from the stack-gases in the drying chambef. This dust has been exam- ined f by Seger and Kramer, and found to consist of 43.65 per cent of insoluble and 56.35 per cent of soluble matter. The insoluble matter gave: silica 31.4 per cent, alumina 14.7 per cent, iroh oxide 4.9 per cent, lime 36.8 per cent, magnesia 1.3 per cent, loss on ignition 0.9 per cent. The soluble portion consisted of potassium sulphate 61.1 per cent, and potassium carbonate 38.9 per cent. Calculating these proportions to percentages of the total dust, we have: Silica (SiO 2 ) 13.71 Alumina (A1 2 O 3 ) , 6.42 Iron oxide (Fe 2 O 3 ) 2. 14 Lime (CaO) 16.06 Magnesia (MgO) . 57 Potash carbonate 34 . 43 Potash sulphate 21 . 92 Loss on ignition 4 . 76 The composition of this stack-dust has directed attention to the possibility of utilizing it as a source of potash, and both American and foreign patents have been taken out to cover possible processes for this purpose. List of references on heat requirements. The following list con- tains the principal papers dealing with this phase of cement-manufac- ture. Those marked with an asterisk are restricted mainly to a dis- cussion of clinkering temperatures, etc. * Bleininger, A. V. Manufacture of hydraulic cements. Bulletin No. 3, Ohio Geological Survey, 1904. * Campbell, E. D. Some preliminary experiments upon the clinkering of Portland cement. Journ. Amer. Chemical Soc., vol. 24, pp. 969-992, Oct., 1902. * Campbell, E. D., and Ball, S. An experiment upon the influence of the fineness of grinding upon the clinkering of Portland cement. Journ. Amer. Chem. Soc., vol. 25, pp. 1103-1112, Nov., 1903. * Campbell, E. D. Further experiments on the clinkering of Portland cement and on the temperature of formation of some of the constituents. Journ. Amer. Chem. Soc., vol. 26, pp. 1143-1158, Sept., 1904. f Journ. Soc. Chem. Industry, vol. 23, p. 661. June 30, 1904. HEAT CONSUMPTION AND HEAT UTILIZATION. 511 Carpenter, R. C. A test of a process for utilizing waste heat from rotary cement-kilns. Sibley Journal of Engineering, March, 1904. Helbig, A. B. The efficiency of waste-gas boilers in connection with rotary cement-kilns. Engineering News, vol. 53, pp. 163-166, Feb. 16, 1905. Newberry, S. B. Fuel consumption in Portland-cement burning. Cement and Engineering News, July, 1901. Richards, J. W. The thermal efficiency of a rotary cement-kirn. Cement, vol. 5, pp. 30-35, 1904. CHAPTER XXXV. REQUISITES AND TREATMENT OF KILN FUELS. THE usual fuel in rotary-kiln practice is pulverized bituminous coal. Oil, natural gas, and producer gas are, however, used at several plants, while charcoal has recently been suggested for use in a projected Arizona mill. These fuels will be discussed in the order named. Coal. Character of kiln coals. In order to be suitable for use in rotary kilns the coal must be of the bituminous type, and preferably a gas- coal. Coals high in fixed carbon and low in volatile matter, while giving high temperatures, will not burn properly when pulverized and blown into the kiln, for they are slow to ignite. The anthracite and semi- bituminous coals are, therefore, ruled out, though they can be used in small quantities mixed with gas-coal, if the mixture be pulverized fine enough. For economic reasons the kiln coal should run as low in ash as pos- sible. The ash not only lowers the heating value of the coal, but it interferes with the composition of the mix, for much of it is always taken up by the cement during burning. The presence of sulphur, in amounts of over 1J per cent, is also technologically a defect, and if the sulphur averages over 2 per cent it is advisable to look up a better coal. As shown by the analyses below, the better coals actually used range in composition about as follows: Volatile matter 30%-40% Fixed carbon 50%-60% Sulphur 0%- 1|% Ash. 5%- 8% Analyses of kiln coals. The following table (194) of analyses of kiln coals is fairly representative of the various types of coal actually in use in rotary-kiln plants. 512 REQUISITES AND TREATMENT OF KILN FUELS. 513 TABLE 194. ANALYSES OF KILN COALS. 1. 2. 3. 4. 5. Volatile matter 32 90 38 10 31 38 35 41 35 26 Fixed carbon 54.66 53 24 58 23 56 15 56 33 n. d. n. d. n. d. 1 30 1 34 Ash 10.25 8.06 9.42 6 36 7 06 2.19 0.60 1.03 2 08 1 35 6. 7. 8. 9. 10. Volatile matter 39 52 39 37 31 87 37 44 38 00 Fixed carbon 51 69 55 82 51 05 53 72 51 72 Sulphur . 1 46 42 n d n d n d Ash 6 13 3 81 5 22 5 50 5 38 Moisture 1 40 1 00 11 86 3 334 4 90 1. Westmoreland,, Pa., slack, used at Cayuga Cement Co., Portland Point, N. Y. R. C. Car- penter. Cement, vol. 5, p. 1904. 2. Fairmount, W. Va., slack, used at Dexter Portland Cement Co., Nazareth, Pa. J. W. Richards. Cement, vol. 5, p. 30. 3. Fairmount, W. Va., slack, used at Alpha Portland Cement Co., Alpha, N. J. F. E. Walker, analyst. 4. West Virginia slack, used by Wolverine Portland Cement Co. at Coldwater, Mich. H. E. Brown, analyst. 22d Ann. Rep. U. S. Geol. Sur., pt. 3, p. 675. 5. West Virginia slack, used by Wolverine Portland Cement Co. at Quincy, Mich. H. E. Brown, analyst. 22d Ann. Rep. U. S. Geol. Sur., pt. 3, p. 675. 6. West Virginia slack, used by Peninsular Portland Cement Co., Cement City, Mich. J. G- Dean, analyst. 22d Ann. Rep. U. S. Geol. Sur., pt. 3, p. 675. 7. Pennsylvania slack, used by Omega Portland Cement Co., Jonesville, Mich. 22d Ann. Rep. U S. Geol. Sur., pt. 3, p. 675. 8. Ohio coal, used by Wellston Portland Cement Co., Wellston, Ohio. W. S. Trueblood, analyst. 9. 10. Ohio coal, used by Ironton Portland Cement Co., Ironton, Ohio. C. D. Quick, analyst. References on coal-fields. The following reports contain data on the distribution and character of American coals. Ashley, G. H. The coal deposits of Indiana. 23d Rep. Indiana Dept. Geology and Natural Resources, pp. 1-1573. 1899. Ashley, G. H. The eastern interior coal-field (Illinois and Indiana). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 265-306. 1902. Bain, H. F. The western interior coal-field (Iowa, Missouri, Kansas). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 333-366. 1902. Diller, J. S. The Coos Bay coal-field, Oregon. 19th Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 309-376. 1898. Haseltine, R. M. The bituminous coal-field of Ohio. 22d Ann. Rep. U. S, Geol. Survey, pt. 3, pp. 215-226. 1902. Hayes, C. W. The coal-fields of the United States (summary). 22d Ann, Rep. U. S. Geol. Survey, pt. 3, pp. 7-24. 1902. Jayes, C. W. The southern Appalachian coal-field (Ala., Ga., Tenn., Ky., Va.). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 227-264. 1902. Lane, A. C. The northern interior coal-field (Michigan). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 307-332. 1902. 514 CEMENTS, LIMES, AND PLASTERS. Smith, G. 0. The Pacific Coast coal-fields (Oregon, Washington, California). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 473-514. 1902. Storrs, L. S. The Rocky Mountain coal-fields (Mont., Wyo., Colo., Utah, N. Mex.). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 415-472. 1902. Taff, J. A. The southwestern coal-field (Ind. Terr., Ark., Texas). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 367-414. 1902. White, D. The bituminous coal-field dtf Maryland. 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 201-214. 1902. White, D., and Campbell, M. R. The bituminous coal-field of Pennsylvania. 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 127-^200. 1902. White, J. C. Report on coal (of West Virginia). Vol. 2, Reports W. Va. Geol. Survey, pp. 81-725. 1903. Woodworth, J. B. The Atlantic Coast Triassic coal-fields (Virginia, North Carolina). 22d Ann. Rep. U. S. Geol. Survey, pt. 3, pp. 25-54. 1902. Crushing. Coal may be bought in the shape of slack, lump or run- of-mine. In the former case no preliminary crushing is required, for the slack can be readily handled by ball mills, Griffin mills, or Williams mills. When slack is bought, therefore, it is sent direct to the drier and then to the fine-reducing mills. But when lump or run-of-mine are purchased the coal can profitably be crushed before being sent to the drier. FIG. 127. Coarse, toothed rolls for lump coal. (Allis-Chalmers Co.) In such cases the preliminary crushing; seems to be accomplished most effectually by rolls. Figs. 127 and 128 show rolls adapted to this kind of work, both being made by the Allis-Chalmers Co. The rolls shown in Fig. 127 are very coarsely toothed, and are intended for use on large lump or run-of-mine coal. They are 24"X30" in size, and can conveniently reduce large lump to about 1- or 2-inch size. In Fig, REQUISITES AND TREATMENT OF KILN FUELS. 515 128 a set of 24" X 18" plain-faced disintegrating rolls are shown. These will handle coal up to say, 4 to 6 inch size, and reduce it economically to | or J inch. Finer than this it is hardly profitable to go, for J-incn coal is readily dried and is of convenient size for either ball, Griffin, or Williams mills. Drying coal. Coal, as bought, may carry as high as 15 per cent of water in winter or wet seasons; usually, it will run from 3 to 8 per cent. To secure good results from the crushing machinery it is neces- sary that this water should be driven off. For coal drying, as for the drying of raw materials, the rotary drier seems best adapted to FIG. 128. Rolls for coal-crushing. (Allis-Chalmers Co.) American conditions. Several types of these driers are discussed on pp. 400 and 649. It should be said, however, that in drying coal it is inadvisable to allow the products of combustion to pass through the cylinder in which the coal is being dried. This restriction serves to decrease slightly the possible economy of the drier, but an evap- oration of 6 to 8 pounds of water per pound of fuel coal can still be counted "on with any good drier. The fuel cost of drying coal containing 8 per cent of moisture, allowing $2 per ton for the coal used as fuel, will therefore be about 3 to 4 cents per ton of dried product. Pulverizing coal. Though apparently brittle enough when in large lumps, coal is a difficult material to pulverize finely. For cement-kiln use, the fineness of reduction is very variable. The finer the coal is pulverized the better results will be obtained from it in the kiln, and the poorer the quality of the coal the finer it is necessary to pulverize it. The fineness attained in practice may, therefore, vary from 85 per cent through a 100-mesh sieve to 95 per cent or more through the same. At one plant a very poor but cheap coal is pulverized to pass 98 per cent through a 100-mesh sieve, and in consequence gives very good results in the kiln. 515 CEMENTS, LIMES, AND PLASTERS. Coal-pulverizing is usually carried on in two stages, the material being first crushed to 20- or 30-mesh in a Williams mill or ball mill and finally reduced in a tube mill. At many plants, however, the entire reduction takes place in one stage, Griffin or Huntingdon mills being used. Descriptions of the WiHiams, Grjtffin, and Huntingdon mills, and of several makes of ball mills and tube mills, will be found in Chapter XXXV. The Smidth ball mill, however, was not described in that chapter, and as it is recommended by its manufacturers for use as an inter- mediate reducer on coal, its general make-up can properly be noted here. The following description of the Smidth ball mill is given s by its makers : "The illustration (Fig. 129) shows the internal arrangement of the type A ball mill. This machine has a through-shaft, with journals running in bearings at both ends. "The machine consists of a drum with two strong end-plates c, between which the curved drum-plates d are fixed. As will be seen these drum-plates do not form a cylinder, but one end of each is set a few inches toward the center, forming steps. The balls and materials tumble over these steps when the drum revolves, and by this the pound- ing action of the balls is considerably increased, the steps at the same time allowing the residue from the sieve to be caught and re-enter the drum. The curved drum-plates are protected on the inside by thick steel plates e which are divided in several sections, so that each sec- tion can be separately renewed. The lining plates are fixed by means of bolts. The end-plates c are also lined with thick plates d. "The grinding-plates have rows of perforations / through which the crushed material constantly falls on the slotted-steel screen-plates g. Whatever is fine enough to pass these screen-plates falls on the inner sieves I, which are coarse and strong sieves, and that which passes through these falls on the finishing-sieves k. The material passing the finishing-sieves is collected in the lower hopper-shaped part of the dust-casing m surrounding the drum, and from this it either falls to a conveyor or elevator, or may be dropped direct into the con- tainers. The outlet of the dust-casing is provided with a slide-gate. "The residue from the sieves, coarse and fine, is carried- up with the mill until it falls through large holes in the curved part / of the slotted- steel screen-plates, and, together with the residue from the slotted-steel screen-plates themselves, falls into the drum again through the steps. "In one of the end-plates a manhole is fitted and in the dust-cas- ing a door corresponding to the manhole, giving access to the interior REQUISITES AND TREATMENT OF KILN FUELS. 517 518 CEMENTS, LIMES, AND PLASTERS. of the mill. The dust-casing is made in three parts , a small section being easily removable for brushing the sieves. The top half may be removed for repair of the drum. "The illustration below (Fig. 130) shows the type B machine, which is the same as type A ; except that the shaft, instead of going through the feed-hopper, is stopped short inside the drum and there fixed in a ' spider' the external part of which runs on a roller bearing. By means of this arrangement lumps up to 10 inches each way can be fed into the mill, and the feed opening itself at the same time is reduced in diameter, which allows a heavier charge of balls to be used. "The very large capacity of the Davidsen tube mill as a pul- verizer created the demand for a large ball mill of sufficient capacity to feed the tube mill with coarse material. This demand has been met by the type C ball mill, in the construction of which the through- shaft of type A and the 'spider'' of type B have been omitted, giving a clear feeding opening of 10 inches without interference. The omis- sion of the shaft and the 'spider' have made it possible to use a much larger charge of steel balls and at the same time lower the cost of repairs by avoiding the necessity for replacement of either shaft or 'spider'. FIG. 130. Smidth ball mill. (F. L. Smidth& Co.) Type B, showing roller bearing as used on type C. " Up to this time there has been but one size of the type C constructed, namely, No. 7. This has become the popular form and size for Port- land-cement work. One No. 7 ball mill and one No. 12 tube mill can be brought together as a grinding unit through placing on the ball-mill REQUISITES AND TREATMENT OF KILN FUELS. 519 screen-frames wire cloth of such mesh as shall be shown by the grinda- bility of the material to pass the proper capacity for pulverization in the tube mill. No. 7 ball mill requires at the maximum 18 horse-power and uses a floor space of 11 feet 4 inches by 15 feet 7 inches. Like the tube mill, the ball mill is a slow-speed machine, making 22 revolutions per minute. "The parts most subjected to wear are the grinding-plates, the side linings, the gear, and the pinion. Depending entirely upon the char- acter of the material, the grinding-plates should give a wear of from six to eighteen months. The side linings do not wear so rapidly. As the gear and pinion are operated at slow speed the wear on these is relatively small. Gears are seldom replaced short of one or two years and pinions have a life of from six to twelve months. Of course gears and pinions wear in proportion to the attention given to lubrication and protection from dust." Power and output in coal-grinding. There is probably greater diversity in coal-grinding practice than in the grinding of either raw material or clinker. Grinding-machines of many different types are in use; the coal reaches the plant in various sizes from slack to lump, and is ground to different finenesses. All this makes it difficult to esti- mate closely on the power requirements and output of the coal-grind- ing mill, but the following data may be of use in this connection. A Williams mill employed at an Illinois cement-plant, working on Illinois coal from the dryer and preparing it for the tube mills, ground six tons of coal per hour to the following fineness. Mesh of sieve 20 50 100 200 Per cent residue 6.9 43 . 3 76 . 2 87 . 3 Per cent passing. 93.1 56.7 23.8 12.7 If the results of this test be compared with those given on p. for the same mill ..working on raw materials, it will be seen that coal is very readily crushed to 20-mesh, and quite easily to 50-mesh. But the percentages of coal passing the 100-mesh and 200-mesh sieves respectively are very much lower than the percentages of raw mix passing the same sieves. A Griffin mill grinding coal from rolls or small crushers will reduce about two tons per hour to a fineness of 95 per cent through a 100- mesh screen, taking 25 to 30 H.P. in doing so. Slack coal for three kilns is ground at one plant in a single No. 16 Davidsen tube mill, the product being about three tons per hour. The above data show a considerable variation in the power required 520 CEMENTS, LIMES, AND PLASTERS. for coal-grinding, the performances quoted being equivalent to power consumptions of from 5 to 15 H.P. hours per ton of coal ground. Mr. C. D. Bartlett, in discussing * this question, states that a good mill will handle slack at the rate of four tons per hour, crushing it to 80-mesh and using about 25 H.P. in doing so. For kiln use, of course, the fine- ness must be considerably greater than this. Total cost of coal preparation. *The total cost of crushing (if neces- sary), drying, and pulverizing coal, and of conveying and feeding the product to the kiln, together with fair allowances ^or replacements and repairs, and for interest on the plant, will probably range from about 20 to 30 cents per ton of dried coal for a four-kiln plant. This will be equivalent to a cost of from 3 to 5 cents per barrel of cement. While this may seem a heavy addition to the cost of cement-manufacture, it should be remembered that careful drying and fine pulverizing enable the manufacturer to use much poorer, and therefore cheaper, grades of coal than could otherwise be utilized. The coal used at American plants costs from 80 cents to $2.50 per ton delivered at the mill, according to the quality of the coal and the location of the mill. In the West, where good coal is far more expensive than stated, oil is used in its place. It is probably safe to say that if a plant is so located that coal will cost over $4 per ton, and no oil or gas is obtainable, the rotary kiln is too expensive for use. Under such fuel conditions it is probably best to install stationary kilns of one of the improved designs described in Chapter XXXII. This is particularly the case if a wet mix be used in the kilns. Fire and explosion risks. The coal-handling end of the plant is subject to two quite distinct, though related, kinds of risks from explosion and fire respectively. Precautions must be taken to guard .against both of these dangers. Explosions may occur when finely divided powdered coal is given free access to air. In order to keep as little powdered coal on hand as possible, the coal-mill is usually run so as to just supply the kilns. This has some inconveniences, but it lessens the risk. During grinding care must be taken to prevent the use of exposed lights or even motors, which are apt to spark, in the coal-pulverizing building. The methods of supplying coal to the kiln should give as little access to air as possible. Separation by blowing is, of course, inadmissible, as was emphasized by the fatal results at the Edison plant in 1903. * Journ. Assoc. Engineering Societies, vol. 31, pp. 44-48. 1903. REQUISITES AND TREATMENT OF KILN FUELS. 521 In addition to the risk of explosion from coal-dust, there is always the chance that coal stored in bulk will heat up and cause a disastrous fire. The following statement regarding coal storage has been recently published * by F. M. Griswold, General Inspector to the Home Insurance Company, of New York: "The quantity stored in any one pile, heap, pocket, or bunker should in no case exceed 1,500 tons. When a greater quantity must be stored there should be a clear space of at least 5 feet between the piles, and that space should be maintained absolutely free for ventilation and disper- sion of gases from the mass. "No accumulation of coal of 1500 tons or less should be piled in excess of 12 feet in height, when trimmed off, or squared, but where such accumulation is delivered from dump-cars on a trestle over 12 feet in height, the extreme height of the pile formed by the natural run of the coal as dumped may be 15 feet, but not more. "Where coal is stored under shelter, there should be perfect ventila- tion, to facilitate escape of gas by circulation of the atmosphere. "Wet coal, especially that wetted by snow and ice, should be dis- posed for immediate use; if its storage be necessary, it should be placed at the top of the pile and be spread out as thinly as practicable, in order to expedite drying. "All accumulations of coal, large or small, should be 'rod-tested' with frequency and regularity, in order to discover any tendency toward dangerous heating, the danger-point being set at about 160 F. If that terrfperature be reached, the exact locality of increasing heat may be determined by inserting an iron pipe, into which a self-registering ther- mometer can be lowered, allowing it to remain for sufficient time to record the full intensity of the heating." List of references on coal drying, grinding, etc. Bartlett, C. D. The burning of pulverized coal. Journ. Assoc. Engineering * Societies, vol. Cl, pp. 44-48. 1903. Doane, A. 0. The spontaneous ignition of coal. Engineering News, vol. 52, p. 141. Aug. 18, 1904. Frazier, W. H. Fire hazards in Portland-cement mills. New York Journal of Commerce, April, 1901. Griswold, F. M. Specifications for storage of bituminous coal. Engineering and Mining Journal, vol. 77, p. 725. 1904. Engineering News, vol. 52, pp. 409-410. Nov. 10, 1904. * Engineering and Mining Journal, vol. 77, p. 725. 1904. 522 CEMENTS, LIMES, AND PLASTERS. Lathbury, B. B., and Spackman, H. S. The Lathbury and Spackman coal- drier. The Rotary Kiln, pp. 150-151. 1902. Anon. Powdered fuel for boiler-furnaces at the Alpha Cement Co.'s works, Alpha, N. J. Engineering News, 1897. Anon. A new system for burning powdered coal. Engineering News, vol. 48, p. 548. Dec. 25, 1902. , Oil. Petroleum was early used in New York and Pennsylvania as a fuel for rotary kilns, but was gradually supplanted by powdered coal. At present no Eastern plants use oil as fuel. In the West, however, where good gas coals are unobtainable at reasonable prices, oil is now in use at four Portland-cement plants. From 11 to 14 gallons of oil are required in Western practice to burn a barrel of cement ; a safe estimate is that one barrel of oil (42 gal- lons) will burn three barrels of cement. Oil may, therefore, be com- pared with coal, in the rotary kiln, on the basis of 1 gallon of oil being equal in effect to 10 Ibs. of coal. List of references on petroleum. The papers on petroleum contained in the following list are of interest either as containing discussions of the fuel value of petroleum, or as describing certain oil fields whose product is at present utilized in Portland-cement manufacture. Eldridge, G. H. The Florence oil field, Colorado. Trans. Amer. Inst. Mining Engrs., vol. 20, pp. 442-462. 1892. Eldridge, G. H. The petroleum fields of California. Bulletin 213, U. S. Geological Survey, pp. 306-321. 1903. Fenneman, N. M. The Boulder, Colorado, oil field. Bulletin 213, U. S. Geolog- ical Survey, pp. 322-332. 1903. Peckham, S. F. Petroleum in southern California. Science, vol. 23, pp. 74- 75. 1894. Anon. Fuel oil on the Pacific Coast. Engineering and Mining Journal, Dec. 20, 1902. Natural Gas. Use of natural gas in kilns. Natural gas is at present used as a kiln fuel in several Kansas plants and at one in Ohio. As a kiln fuel it is satisfactory enough, giving as much results per B.T.U. as does a good coal. Apparently, however, a gas-fired kiln cannot be pushed as hard as a kiln using coal, though the data are insufficient to give any decisive evidence on this point. A recent report on a Western cement proposition states that the natural gas to be used in the kilns REQUISITES AND TREATMENT OF KILN FUELS. 523 has been contracted for at the rate of 3 cents per thousand feet. This is about equivalent to coal at 90 cents per ton. Analyses and thermal value. The following analyses, made * by Prof. E. H. S. Bailey, will serve to give some idea of the composition of natural gas from a number of Kansas localities. TABLE 195. ANALYSES OF NATURAL GAS, KANSAS. lola. Inde- pend- ence. Cherry- vale. Coffey- ville. Paola. Ossawat- omie. Hydrogen (H) 00 0.00 0.00 0.00 00 00 Oxvgen (O) 45 trace 0.22 0.12 45 trace Nitrogen (N) 7.76 3.28 5.94 2.21 2 34 60 Carbon monoxide (CO) 1.23 0.33 1.16 0.91 1.57 1 33 Carbon dioxide (CO 2 ) 90 44 22 00 33 22 Ethvlene series (C 2 H 4 , etc.) 0.00 0.67 0.00 0.35 0.11 22 Marsh-gas (CH 4 ) 89 . 66 95.28 92.46 96.41 95.20 97.63 TABLE 196. THERMAL VALUES OF NATURAL GAS. State. Field. B.T.U. per Cubic Foot. State. Field. B.T.U. per Cubic Foot. Indiana Anderson 1021 Pennsylvania East Liberty 592 i Kokomo 1030 Grapeville 823 < Marion 1024 (i 891 ( ( Muncie 1019 (( Harvey . . 990 Kentucky . . Louisville 939 n Leechburg 1073 New York . . . Olean 1071 t ( St Joe 1170 < i (i West Bloomfield 998 West Virginia Fairmount 1137 Ohio Findlay ... . 1100 (i it Lurnberport . . . 1131 1 1 i ( 1020 K (t Morgantown . . . 1143 1 1 Fostoria 1016 ti n Shinnston 1141 1 t St- Mary's 1028 n fi 1 1 1144 Pennsylvania Cherrv Tree 840 tt n 1065 Creighton 1025 List of references on natural gas. Of the following papers, those marked with an asterisk are of interest as discussions of the fuel value of natural gas, while those unmarked contain data on its utilization in the lola district. Adams, G. I., and others. Economic geology of the lola quadrangle, Kansas. Bulletin 238, U. S. Geological Survey, 83 pp. 1904. * "Mineral Resources of Kansas for 1897 ", p. 52. 524 CEMENTS, LIMES, AND PLASTERS. Crane, W. R. Natural gas in steam production (in Kansas). Mines and Minerals, vol. 24, pp. 154-156. Nov., 1903. Grimsley, G. P. A new Portland- cement mill in the gas-fields of Kansas. Engineering and Mining Journal, Feb. 16, 1901. Bailey, E. H. S. Natural gas and coal oil in Kansas. Kansas University Quarterly, vol. 4, pp. 1-1 4. 1895. * Bownocker, J. A. Occurrence and exploitation of petroleum and natural gas in Ohio. Bulletin 1, 4th Series, Ohio Geol. Survey, 1903, p. 125. * Ford, S. A. Fuel value of Pittsburg gas. American Manufacturer, supple- ment, April, 1886. , * Howard, C. D. Composition and fuel value of West Virginia gas. Vol. la, Reports West Virginia Geol. Survey, pp. 553-556. 1904. * Orton, E. Preliminary Report upon Petroleum and Natural Gas (in Ohio). 1887, pp. 53-54. * Phillips, F. C. The chemical composition of natural gas. Report I, 2d Geol. Survey Penna., pp. 787-827. 1887. * Phillips, F. C. The chemical composition of natural gas. Vol. la, Reports Wes^ Virginia Geological Survey, 1904, pp. 513-552. * White, I. C. The composition of natural gas. Vol. la, Reports West Virginia Geological Survey, 1904,^ pp. 513-557. Producer-gas. Producer-gas has been used in rotary kilns at three American plants at least. Two of these plants report that their best fuel consumption, when producer-gas was used, was equivalent to 220 to 240 Ibs. coal per barrel of cement. The third plant, however, has recently experi- mented with the Swindell gas-producer, and reports that a really eco- nomical fuel consumption is attained. Producer-gas from wood or lignite. A cement-plant located in a district where wood, lignite, or poor coal were the only natural fuels would probably get good results by utilizing these fuels in the gas-pro- ducer. A recent installation of this type at a Mexican copper-plant has utilized wood, bituminous coal, and anthracite. This plant, in respect to the use of wood, is described t as follows : "The plant has been operated with wood and with coal, both anthra- cite and bituminous. The down-draught principle of the producers, by which all gases pass out through the bottom of the fire, has proved thoroughly efficacious in producing fixed gases from all kinds of fuel used at Nacozari. The water from the scrubber rarely shows even that trace of tar which manifests itself by an iridescent film on the sur- face of the water in the lower scrubber tank. f Langton, J. The power-plant of the Moctezuma Copper Co., Mexico. Trans. Amer Inst. Mining Engineers, Oct., 1903. REQUISITES AND TREATMENT OF KILN FUELS. 525 " The use of producer-gas made from wood alone is the most novel feature of the plant. No guiding experience was found for this process; but, with the desire to utilize as far as possible the limited local wood- supply, the gas-producer plant was selected with the object, among other things, of determining the advisability of using, if not wood alone, at least a considerable admixture of wood with bituminous coal. The most obivous difficulty to be feared arose from the large proportion of condensible distillates yielded by wood, and the danger that some portion of these might be imperfectly fixed in passing through the pro- ducer. The trouble from tar deposited in the gas apparatus and pipes would be serious, and even a small quantity of tar in the gas itself is. a fertile source of trouble at the engine-valves. Unless a permanent gas could be made from wood, this fuel would be unavailable. " The first care, therefore, was to insure that there should be a bed of charcoal on the grate sufficient to form an adequate fixing-zone. To obtain this the producers were filled about 5 feet deep with cordwood sawn in blocks about 6 inches long and the contents blown with a slow fire for four or five hours before the gas was turned into the holder. The gas, as it proved, was turned into the holder too soon. At first it contained some tar, and it was not until after three hours' operation that the charcoal accumulated in sufficient quantity, so that the pro- ducers delivered fixed permanent gases to the holder. "The character of the gas produced from different fuels is shown by the following averages from a series of analyses made by A. Sand- berg, of Lund, during the final trials of the gas-making plant, which extended from February 16 to March 20, 1901. AVERAGE COMPOSITION OF GAS FROM DIFFERENT FUELS. Fuel. Components. CO. H. CH 4 . CuH 2 N. CO 2 . Anthracite coal 21.6 20.32 13.27 12.68 13.08 20.97 1.86 2.35 2.61 0.18 0.2 0.28 8.1 7.66 15.96 Bituminous coal Wood Fuel. Components. Calorific Value. B.T.U. per Cubic Foot at60F. and 29.9 Inches Barometric Pressure. O. N. Anthracite coal 0.2 0.04 0.11 55.38 55.35 46.80 131.1 133.27 140.22 Bituminous coal Wood 526 CEMENTS, LIMES, AND PLASTERS. "As compared with coal-gas, the proportions of CO and H in wood- gas are reversed, and the percentage of CC>2 is doubled. With the typical compositions (see page 525) the wood-gas shows less tendency than the coal-gas to pre-ignition in the engines. This effect appears to be due to the large proportion of CO2, but the reason why this is so is not apparent." ^ Charcoal. The use of charcoal as a rotary kiln fuel has been seriously sug- gested recently by at least one engineer for a plant located where no other type of fuel was obtainable at reasonable rates. It is, however, a matter of serious doubt whether this material could be brought to ignite properly before it reached the stack. CHAPTER XXXVI. CLINKER COOLING, GRINDING, AND STORAGE. USE OF GYPSUM. THE clinker, issuing hot from the rotary kilns, must be very finely ground in order to convert it into cement. This involves cooling the clinker previous to grinding; otherwise the hot clinker would be difficult to handle both in transportation and in the pulverizing ma- chinery. A third requisite of the process is that either the clinker or the ground cement must be seasoned, in some way, in order to slake any free lime that may be present. Modern clinker invariably contains some free lime, and while its effects may be masked by the free use of gypsum, it is advisable to give it as much opportunity as possible to slake and become inert. In the present chapter, therefore, the subjects of clinker-cooling, clinker-grinding, the use of gypsum and cement storage will be taken up. Clinker-cooling. General methods of clinker-cooling. Methods of clinker-cooling vary exceedingly in their processes and effectiveness. At one extreme might be placed the device, used at one plant only, of receiving the clinker from each kiln in a shute which passes through the wall of the kiln build- ing and deposits the clinker in a heap on the ground outside. This is, of course, a remarkably simple process, mechanically, but as it involves hand-labor to an alarming extent it is hardly probable that any other American plants will take it up. At the other extreme is, decidedly, the Atlas two-stage cooling system. Omitting the crude device first mentio'ned above, clinker-cooling systems may be roughly grouped as follows: (1) Pan conveyors, rolls, and sprinkling. (2) Stationary tower coolers. (3) One-stage rotary coolers. (4) Two-stage (Atlas) rotary coolers. These methods will be briefly discussed in the order named. 527 528 CEMENTS, LIMES, AND PLASTERS. Pan conveyors, rolls, and sprinkling. At a number of plants the hot clinker is caught, as it drops out of the kiln, in pan conveyors. As it passes along in these it is sprinkled with fine jets of water, and at some point of its progress is passed through a pair of rolls. This method therefore, contains all the elements of any cooling system, and in a very simple form. It is not adapted to utilize the heat of the clinker however, and the product even after ^sprinkling and passing the rolls is too hot to be sent immediately to the grinding-mills. The simplicity of the method is therefore counterbalanced by a loss of heat and rela- tively high amuont of hand-labor. Stationary tower coolers. Many plants use stationary coolers in the form of towers. The Mosser cooler, shown in section in Fig. 131, is a good example of this type. The cooling installation at the Buck- horn plant is described as follows by Mr. Humphreys in Engineering News : * FIG. 131. Tower cooler, Buckhorn Portland Cement Co. (Engineering News.) "Each pair of kilns discharges through a fire-brick-lined shute into the boot of a single-chain open elevator. "As the clinker falls into the buckets of this elevator it is sprayed * Humphreys, R.L. The plant of the Buckhorn Portland Cement Co. Engi- neering News, vol. 50, pp. 408-411. Nov. 5, 1903. CLINKER COOLING, GRINDING, AND STORAGE. 529 with water. The elevator dumps the clinker into a cooler built by Wm. F. Mosser & Son. There are three of these coolers, each 32 feet high, 8 feet in diameter, having a cast-iron blast-pipe running through the center, with sheet-steel conical shields every 5 feet, extending to within 10 inches of the shell of the cooler. "Under this shield are holes in the blast-pipe, through which a constant flow of fresh air is maintained by means of a fan, the air passing out of the cooler through holes in its shell the latter having conical shields on the inside just above these openings. "The heat of the clinker is absorbed in the vaporization of the water and is removed by the current of air which passes through the thin stream of clinker moving through the cooler between the two shields. "The coolers rest on a cast-iron plate, supported by foundations 4 feet high, in a pit about 20 feet below the kiln-room floor. Running under these coolers are belt conveyors which receive the cooled clinker (drawn from four openings in each cooler) and carry it to the boot of an elevator, which discharges it through an opening in the wall between the kiln room and the clinker ball-mill department onto a storage floor." One-stage rotary cooler. The next step in clinker-cooling devices is the use of rotary coolers. These are simply rotary driers, reversed in action, and require no special description here. Atlas two-stage rotary cooler. By far the most satisfactory of cool- ing devices is the two-stage rotary cooler employed by the Atlas Port- land Cement Company. It is, so far as the writer knows, the only cooling system which really cools the clinker to a handling temperature and does so quickly and economically. The cooling system at the main Atlas plant was described by Stanger and Blount in 1901 as follows: "The clinker drops from the burning cylinder into a second rotating cylinder, about 30 feet long and 3 feet in diameter, revolving about six times as fast as the burning cylinder. This is lined with fire-brick, and through it passes a current of air which goes to feed the flame of burning coal-dust. The greater part of the sensible heat in the clinker is thus saved and utilized. The clinker, still moderately hot, falls on to three crushing rolls contained in a housing and moistened by a spray of water. As shown in the figure a pair of kilns with their accompany- ing first cooling cylinders converge so as to deliver the clinker onto these rolls and from this point a single secondary cooling apparatus serves this pair of kilns. The object of the rolls is to crush large lumps of clinker which may have been formed by the aggregation of a num- ber of small fragments adhering together when plastic in the burning 530 CEMENTS, LIMES, AND PLASTERS. cylinder. These lumps being built up of small pieces loosely stuck together differ entirely from the tough hard masses formed in a fixed kiln fed with blocks or bricks of raw material, and are readily broken up to the size of a hazelnut. The warm moist clinker passes down a third rotating cylinder 60 feet long by 5 feet in diameter, lined with hard cast-iron plates provided with shelves so as to toss and tumble the pieces as they creep down. Air is drawn in through this cylinder by means of a chimney which also carries off the water vapor from the housing of the rolls. It is intended that the clinker shall emerge ELEVATION rm r -1 i , .j T" Ih Burning Cylinder { 1 i^tjii)a|H f uJ Hi nji rp j 1 | Burning Cylinder { ! .r~-^ " *- ij L il T1 I U- Cruahing Rolla PLA.N FIG. 132. Atlas rotary two-stage coolers. (Engineering News.) from the end of the last cooling cylinder, in a slightly moist condition, and to ensure this, regulation of the water at the rolls is supplemented by a small jet at the end of the last cooler." This system is shown in Fig. 132, taken from the paper * below cited. Clinker-grinding. After cooling sufficiently to be workable, the clinker passes to the clinker-grinding department of the mill. The problem before this department is to reduce large quantities of an intensely hard and semi- vitrified material to finely ground cement at the lowest cost possible. This reduction is now usually accomplished in two or three stages. * Stanger, W. H., and Blount, B. The rotary process of cement-manufacture. Proc. Inst. Civil Engineers, vol. 145, pp. 57-68. 1901. See especially p. 62 for coolers. CLINKER COOLING, GRINDING, AND STORAGE. 531 Somewhere in the process it is necessary to provide for the addition of a certain comparatively small percentage of gypsum or plaster, in order to bring the setting properties of the cement up to commercial requirements. Though this addition is commonly made during the grinding process, it will be discussed later in the chapter. The power allowed and machinery installed for pulverizing the clinker at a Portland-cement plant using the dry process of manu- facture are very closely the same as that required for pulverizing the raw materials for the same output. This may seem, at first sight, improbable, for Portland-cement clinker is much harder to grind than any possible combination of raw materials; but it must be remembered that for every barrel of cement produced about 600 Ibs. of raw materials must be pulverized, while only a scant 400 Ibs. of clinker will be treated, that the large crushers required for some raw materials can be dis- pensed with in crushing clinker, and that the raw side rarely runs full time. The raw material side and the clinker side of a dry-process Port- land-cement plant are, therefore, usually almost or exactly duplicates. The difficulty, and in consequence the expense, of grinding clinker will depend in large part on the chemical composition of the clinker and on the temperature at which it has been burned. The difficulty of grinding, for example, increases with the percentage of lime carried by the clinker, because of the higher burning which has been necessary, and a clinker containing 64 per cent of lime will be very noticeably more resistant to pulverizing than one carrying 62 per cent of lime. So far as regards burning, it may be said in general that the more thor- oughly burned the clinker the more difficult it will be to grind, assuming that its chemical composition remains the same. The tendency among engineers at present is to demand more finely ground cement. While this demand is doubtless justified by the results of comparative tests of finely and coarsely ground cements, it must be borne in mind that any increase in fineness of grinding means a decrease in the product per hour of the grinding-mills employed, and a conse- quent increase in the cost of cement. At some point in the process, therefore, the gain in strength due to fineness of grinding will be counter- balanced by the increased cost of manufacturing the more finely ground product. The increase in the required fineness has been gradual but steady during recent years. Most specifications now require at least 90 per cent to pass a 100-mesh sieve; a number require 92 per cent; while a few important specifications require 95 per cent. Within a few years it is probable that almost all specifications will go as high as this. 532 CEMENTS, LIMES, AND PLASTERS. The following description of the clinker-grinding side of the plant of the Hudson Portland Cement Co. has recently appeared in Engineer- ing News. The layout of this plant is shown in Fig. 85, p. 407: "The kilns are 60 feet long and 6 feet in diameter, and each is driven by a 7^-H.P. electric motor. Fig. 4 is a view of the under side of the kilns showing the rolls on w r hich they^un. The immense size of these modern cement burners is excellently indicated by this illustration. From the kilns the clinker is run into rotary coolers. There are five of these, one for each pair of kilns. In these coolers tjie clinker is cooled by a current of air which is blown in at the forward ends and passes out of the rear ends into the trunk mains whence the air-pipes to the kilns branch off. The coal is thus blown into the kilns by heated air. From each cooler the clinker drops into a clinker pit and from these pits the elevators G hoist it to the bins of the five Krupp ball mills. "From the ball mills the underground screw conveyor 14 and the elevator H take the powder to the double hopper where the final adjust- ment of proportions is made if necessary, and thence the conveyor 15, the elevator HH, and the conveyor 16 take the powder to the feed-bins of the seven tube mills, for final grinding. The discharge from the tube mills is taken by the conveyor 16 J to conveyor 26, which leads to the storage-bins. "The course of the coal through the drier and grinders in the coal- grinding room and thence by conveyor 24, elevator L, and conveyor 26 to the powdered coal-bins for the kilns can be readily traced from the drawings, and need not be explained further here." Actual practice. The following data relate to the machinery actually used on the clinker-grinding side of a number of American plants and will serve to give a good idea of present practice in that line: Plant No. 1. Dry process : 3 kilns, about 450 barrels per day. 1 small jaw-crusher; 2 ball mills; 2 tube mills. Plant No. 8. Dry process : 4 kilns, about 700 bbls. per day. 1 rotary cooler; - 2 Smidth ball mills to 8 mesh; 1 Bonnot ball mill to 30 mesh; 4 tube mills. CLINKER COOLING, GRINDING, AND STORAGE. 533 Plant No. 3. Wet process: 3 kilns, about 240 bbls. per day. 1 crusher; 1 set rolls; 7 run millstones; 2 tube mills. Plant No. 4. Wet process: 13 kilns, about 1300 bbls. per day. 4 ball mills; 4 tube mills. Plant No. 5. Wet process : 3 kilns, about 350 bbls. per day. 3 Smidth ball mills, No. 7; 3 Davidsen tube mills, No. 12. Plant No. 6. Dry process: 6 kilns, about 1200 bbls. per day. 1 set rolls; 5 Bonnot ball mills; 5 Bonnot tube mills. Plant No. 7. Wet process : 5 kilns, about 550 bbls. per day. 1 rotary cooler; 2 Smidth ball mills, No. 7; 2 tube mills. Plant No. 8. Wet process : 14 kilns, about 1200 bbls. per day. 1 set rolls; 15 Griffin mills. Plant No. 9. Wet process : 6 kilns, about 600 bbls. per day. 3 ball mills; 3 tube mills. Plant No. 10. Wet process : 14 kilns, about 1600 bbls. per day. 2 rotary coolers; 6 ball mills; 6 tube mills. Plant No. 11. Wet process: 10 kilns, about 1100 bbls. per day. 4 ball mills; 4 tube mills. Plant No. 12. Wet process : 9 kilns, about 1300 bbls. per day. 1 roll-crusher; 10 Griffin mills; 3 Griffin mills. Plant No. 13. Dry process : 3 kilns, about 525 bbls. per day. 1 crusher; 2 ball mills; 2 tube mills. 534 CEMENTS, LIMES, AND PLASTERS. Plant No. 14. Dry process: 10 kilns, about 1600 bbls. per day. 7 ball mills; 7 tube mills. Plant No. 15. Dry process : 8 kilns, about 1300 bbls. per day. 4 ball. mills; 6 tube mills, f* Plant No. 16. Dry process : 6 kilns, about 1000 bbls. per day. 4 ball mills; 4 tube mills. Plant No. 17. Dry process: 6 kilns, about 1200 bbls. per day. 1 kominuter and 3 ball mills; 4 tube mills. Plant No. IS. Wet process: 21 kilns, about 3000 bbls. per day. 32 Griffin mills. Plant No. 19. Dry process : 10 kilns, about 1700 bbls. per day. 5 rotary coolers; 5 ball mills; 6 tube mills. Plant No. 20. Wet process: 10 kilns, about 1400 bbls. per day. 1 cracker; 2 kominuters; 5 tube mills. Plant No. 21. Dry process : 3 kilns, about 450 bbls. per day. 1 Williams mill; 2 tube mills. Plant No. 22. Dry process: 4 kilns, about 700 bbls. per day. 1 cracker; 7 Griffin mills. Plant No. 23. Dry process : 2 kilns, about 300 bbls. per day. 2 ball mills; 2 tube mills. Use and Effects of Gypsum or Plaster. The high-limed clinker now produced in the rotary process is nat- urally very quick-setting. In order to retard its set sufficiently to pass commercial requirements, sulphate of lime, in the form of gypsum or plaster, is now universally employed. This substance, when added in quantities up to 2 to 3 per cent, retards the set of the cement pro- portionately, and also increases somewhat its tensile strength in short CLINKER COOLING, GRINDING, AND STORAGE. 535 time tests. In larger quantities, its retarding influence becomes less, and finally negative, while a decided weakening of the cement is notice- able. The more theoretical part of the discussion, relating to the form in which the sulphate is applied, and the influence of various percentages of sulphate on the set and strength of the cement, will be first presented : after which the actual methods of application, with analyses of gypsums and plasters used in practice, will be discussed. Form in which calcium sulphate is used. The requisite calcium sulphate may be added to the cement in one of three forms: as crude gypsum, as calcined plaster, or as dead-burnt (anhydrous) plaster. For a full description of the manufacture and properties of these three products the reader is referred to Part I of this volume. In the pres- ent place their essential characters can be briefly stated as follows: Crude gypsum is a natural hydrous sulphate of lime, corresponding to the formula CaSO4+2H 2 O, and to the composition calcium sulphate 79.1 per cent, water 20.9 per cent. Calcined. plaster, or plaster of Paris, is obtained by heating gypsum at temperatures of 350-400 F., the result being that three fourths of the combined water is driven off. The resulting plaster has the formula CaSC^ + JH^O, corresponding to the composition calcium sulphate 93.8 per cent, water 6.2 per cent. If gypsum be calcined at temperatures much above 400 F., all of its combined water will be expelled, leaving dead-burnt or anhydrous plaster, which is simply CaSO*. Considerable discussion has been aroused over the question, which of these three forms of calcium sulphate is the more advantageous for use: but few satisfactory series of experiments are on record in regard to this point. A misleading statement often made is that plaster of Paris, because of its greater chemical activity, will naturally be much more effective than gypsum, weight for weight. The fallacy involved in this statement is revealed when it is considered that the calcium sulphate added to the cement has absolutely no effect until the mixture is gauged with water; and that this addition of water will naturally reconvert the plaster immediately into the hydrous lime sulphate, gypsum. Any argument based on relative chemical activity, so-called, is therefore fallacious. The results of a few recorded experiments, on the comparative effects of the various forms of calcium sulphate, on the set and strength of the cement, will be given below: after which the conclusions which may be drawn from these experiments and from commercial conditions and actual practice will be summarized. 536 CEMENTS, LIMES, AND PLASTERS. Neat, 52 weeks 1 2 Per- cent of plaster " t IG. 133. Effect of plaster on strength; different ages and compositions. (Dyckerhoff.) CLINKER COOLING, GRINDING, AND STORAGE. 537 Nihoul and Dufossez, in the course of the experiments described on page 540, tested the comparative effect of calcium sulphate in four different forms i.e., as crude gypsum, as calcined plaster, as anhydrous plaster and as chemically precipitated calcium sulphate. Their con- clusions were: (1) that with the precipitated calcium sulphate and cal- cined plaster the retardation of set is proportional to the amount of sulphate added: and, (2) that with crude gypsum this is true only when less than 2 per cent of gypsum is employed, larger percentages causing acceleration rather than retardation of set. Lewis has carried out a short series of experiments on the influence of calcium sulphate on the strength of the cement, applying the cal- cium sulphate in three different forms gypsum, plaster of Paris, and anhydrous plaster. The results, given in the table below, are not de- cisive: but seem to show a somewhat greater regularity of effect when plaster of Paris or anhydrous plaster are used than when gypsum is employed. TABLE 197. EFFECT OF FORM OF SULPHATE USED. (LEWIS.) Amount Added. Tensile Strength, 7 Days Neat. Tensile Strength, 7 Days, 3:1. Anhydrous Sulphate. Plaster of Paris. Crude Gypsum. Anhydrous Sulphate. Plaster of Paris. Crude Gypsum. per cent 444 444 589 651 729 524 247 254 444 196 196 212 215 225 165 66 63 196 179 194 179 li 2 647 673 541 533 593 3 4 663 293 148 127 5 6 To summarize the matter: The active retarding agent is the sul- phur trioxide present in the gypsum or plaster. As anhydrous plaster and plaster of Paris both contain somewhat higher percentages of SOa than gypsum, they will exercise a proportionally greater retarding effect, weight for weight, than will gypsum. But for ordinary practice this slight advantage is immensely counterbalanced by the fact that gypsum costs usually less than half as much as either of the plasters: and for ordinary practice, therefore, gypsum is the only form of cal- cium sulphate that can be considered available. In certain plants, however, where the sulphate is added after the cement has been ground, it is necessary to use plaster of Paris; because gypsum as bought is ground too coarsely to add to a finely pulverized cement. 538 CEMENTS, LIMES, AND PLASTERS. Effect of calcium sulphate on set of cement. Experiments on the effect of setting time of the addition of gypsum or plaster are fairly numerous. Unfortunately, such records mean very little, unless they are accompanied by sufficient data, as to the chemical composition, 3 4 Per cent of Plaster. FIG. 134. Effect of plaster on tensile strength of Portland cement. (Lewis.) fineness, etc., of the cement; to enable some idea to be formed con- cerning the general type of cement tested. Experiments on low-limed cement can not be fairly compared with those carried out on high-limed cements; and cement made in stationary kilns behaves differently from the usual product of the rotary. CLINKER COOLING, GRINDING, AND STORAGE. 539 2 3 Plaster, in per cents FIG. 135. Effect of plaster on setting-time of Portland cement. (Dyckerhoff's tests; Nihoul's tests.) 540 CEMENTS, LIMES, AND PLASTERS. In the experiments * of Nihoul and Dufossez a commercial Portland cement of the following composition was used: Silica (SiO 2 ) 22.80 Alumina (Al 2 Oj,) 7 . 79 Iron oxide (Fe 2 O 3 ) 1 - 27 Lime (CaO) ".' . ' ? 65.80 Magnesia (MgO) . 59 Carbon dioxide (CO 2 ) 1 .36 Water 0.20 It is to be noted that this cement, though probably made in a sta- tionary kiln, is very high-limed and correspondingly quick-setting. It can therefore be considered as closely similar to the average rotary clinker. FIG. 136. f Effect of plaster on setting-time. (Sab in.) TABLE 198. EFFECT OF ADDING VARIOUS PERCENTAGES OF CALCINED PLASTER. (NlHOUL AND DUFOSSEZ.) Composition. Pure cement .................. Cement with 1 per cent plaster. - C ( it it H (t If Initial Set. Final Set. Hours. Minutes. Hours. Minutes, 8 10 40 7 51 33 13 20 32 50 10 12 * Journ. Soc. Chem. Industry, vol. 21, pp. 859-860. 1902. tFrom Johnson's "Materials of Construction", p. 187. CLINKER COOLING, GRINDING, AND STORAGE. 541 TABLE 199. EFFECT OF CALCINED SULPHATE ox SET OF CEMENT. (DYCKERHOFF.) Per Cent Plaster Added. Setting-time, Hours. Minutes. 20 1* 3 30 1 10 2 14 Results obtained by Dyckerhoff * are given in the table above. It is unfortunate that no analysis of the cement experimented on is- obtainable, in view of the remarkably great retardation effected by very small percentages of sulphate. A very unusual set of results, obtained in experiments, by Messrs. Kniskern and Gass, has recently been published f by Prof. R. C. Car- penter. Clinker was procured from a cement-plant in unground form and ground in the laboratory, being mixed with various percentages of gypsum. The results are as follows : TABLE 200. EFFECT OF GYPSUM ON SETTING-TIME. (KNISKERN AND GASS.) Per Cent Gypsum, Initial Set, Minutes. Final Set, Minutes. Per Cent Gypsum. Initial Set, Minutes. Final Set, Minutes. 2 52 4 28 45 | 6 87 4$ 22 40 i| 80 157 5 27 59 2 24 114 5J 20 78 2i 29 79 6 19 37 3 30 69 6| 22 40 3* 27 72 18 59 These results are shown diagrammatically in Fig. 137 and compari- son of this curve with those of Figs. 135 and 136 will show their unique- character. The maximum effect was obtained with 1J per cent of gypsum, and a rapid decrease in effect was shown when 2 per cent or more was used. Unfortunately no analysis is given of the cement experimented on, so that we cannot judge whether or not there is any reason for these curious results. * Proc. Inst. Civ. Engrs., vol. 62, p. 156. 1880. t Engineering News, vol. 53, pp. 13-14. Jan. 5, 1905. 542 CEMENTS, LIMES, AND PLASTERS. 2345 Per cent of plaster FIG. 137. Effect of plaster on setting-time of Portland cement. Kniskern and Gass, 1905. CLINKER COOLING, GRINDING, AND STORAGE. 543 Effect of calcium sulphate on strength of cement. In addition to retarding the set of the cement, gypsum or plaster exerts an interesting influence on its strength and, in some cases, its soundness. 1.0 2.0 3.0 PERCENTAGE OF PLASTER PARIS FIG. 138.* Effect of plaster on strength of 1:3 mortar. (Tetmajer.) TABLE 201. EFFECT OF CALCIUM SULPHATE ON STRENGTH OF CEMENT. (DYCKERHOFF.) Percent- age Gypsum. Time of Set. Hrs. Min. Neat Cement. 1 Cement : 3 Sand. Week. 4 Wks. 12 Wks. 25 Wks. 52 Wks. 1 Week. 4 Wks. 12 Wks. 26 Wks. 52 Wks. 1 2 20 3 30 10 14 323 315 375 425 405 456 508 543 518 572 568 688 620 623 695 718 700 650 780 805 115 142 159 180 168 212 238 263 238 339 311 305 302 353 368 375 360 390 384 410 * From Johnson's "Materials of Construction", p. 187. 544 CEMENTS, LIMES, AND PLASTERS. TABLE 202. EFFECT OF CALCIUM SULPHATE ON STRENGTH OF CEMENT. (GRANT.) Per Cent SO 3 Added. Neat Cement. 1 Cement : 1 Sand. 7 Days. 30 Days. GO.Days. 90 Days. v* 7 Days. 30 Days. 60 Days. 90 Days. i 313 305 500 503 536 567 605.4 618.0 106.8 129.2 159.2 226.6 188.4 259.6 266.8 255.2 TABLE 203. EFFECT OF TREATMENT WITH ANHYDROUS CALCIUM SULPHATE. (LEWIS.) Tensile Strength. Per Cent Sulphate Added. 7 Days: Neat. 7 Days: 3:1. 444 Ibs. 196 Ibs. 2 647 " 4 663 " 148 " 5 293 " 127 " TABLE 204. EFFECT OF TREATMENT WITH CRUDE GYPSUM. (LEWIS.) Per Cent Tensile Strength. Gypsum Added. 7 Days: Neat. 7 Days: 3:1. 444 Ibs. 196 Ibs. 2 673 " 3 541 " 179 " 4 533 " 194 " 5 593 " 179 " TABLE 205. EFFECT OF TREATMENT WITH PLASTER OF PARIS. (LEWIS.) Tensile Strength. Per Cent Plaster Added. 7 Days: Neat. 7 Days: 3:1. 444 Ibs. 196 Ibs. U 589 " 212 " 2 651 " 215 " 3 729 " 225 " 4 524 " 165 " 5 247 " 66 " 6 254 " 63 " CLINKER COOLING, GRINDING, AND STORAGE. 545 Methods cf using gypsum. From what has been said on pre- ceding pages, it is evident that in Portland-cement manufacture either gypsum or burned plaster may be used to retard the set of the cement. As a matter of fact, gypsum is -the form almost uni- versally employed in the United States. This is merely a question of cost. It is true that to secure the same amount of retardation of set it will be necessary to add a little more of gypsum than if burned plaster were used; but, on the other hand, 'gypsum is much cheaper than burned plaster. The addition of the gypsum to the clinker is usually made before it has passed into the ball mill, kominuter, or whatever mill is in use for preliminary grinding. Adding it at this point secures much more thorough mixing and pulverizing than if the mixture were made later in the process. At some of the few plants which use plaster instead of gypsum, the finely ground plaster is not added until the clinker has received its final grinding and is ready for storage or packing. Analyses cf gypsum used. The following analyses will serve to illustrate the composition of the crude gypsum and of the calcined plaster used at different American Portland-cement plants. TABLE 206. ANALYSES OF GYPSUM USED IN CEMENT-PLANTS. Silica (SiO 2 ) 32 86 ] (n. d. Alumina (A1 2 O 3 ) \ 5 10 7.26 n. d. Iron oxide (Fe 2 O 3 ) | 0.87 0.36 J n. d. Lime (CaO) 31 94 30.84 n d. n. d. 35.8 Magresia (MgO) 52 Sulphur trioxide (SO 3 ) 44 93 43 60 40 20 43 20 43 6 Carbon dioxide (CO 2 ) Water 20 95 21 68 n d n d 20 3 TABLE 207. ANALYSES OF CALCINED PLASTER USED AT CEMENT-PLANTS. Silica (SiO 2 ) 83 1 10 n d Alumina (A1 O 3 ) > 0.46 0.32 n. d. Lime (CaO) 38.58 37 87 37 26 Magnesia (M^O) 0.50 0.73 1 11 Sulphur trioxide (SO 2 ) 54.12 53 26 50 50 Carbon dioxide (CO 2 ) n. d. n d 3 40 Water 5 61 6 32 5 50 Effect of various salts on set of cement. Experiments have been made on the use of various other salts sulphates, phosphates, chlorides, etc. by different chemists. Few of the results thus obtained are of any 546 CEMENTS, LIMES, AND PLASTERS. practical importance, for most of the salts experimented with are too costly for use. TABLE 208. EFFECT OF VARIOUS SALTS ON SET OF CEMENT. (NmouL AND DUFOSSEZ.) f Initial Hrs. Set. Min. Final Set. Hrs. Min. Pure ceme Cement w >nt ith t ( 2 8 ' 8 7 3 3 8 1 5 6 13 32 5 4 18 3 2p 2 2 2 2 er c( jnt calcium sulphate ' strontium sulphate ' barium sulphate ' calcium phosphate ' ' aluminate ' precipitated silica Recent experiments on the use of solid chloride of lime as a retarder have been published by Carpenter. These were carried out by the same experimenters, and probably on the same cement which, when treated with gypsum, gave the erratic results noted on p. 541. Carpenter sum- marizes * these experiments as follows: "Messrs. Kniskern and Gass, in the Sibley Laboratory, ground differ- ent percentages of chloride of calcium (CaCl 2 ) with cement clinker and afterwards made pats, using in each case simply enough water to give the material its normal consistency for this purpose. Their results show that the chloride of calcium had great effect in retarding the time of setting and exerted the greatest effect when about 0.5 per cent by weight of the chloride of calcium was employed. On account of the water required, 1 per cent of the chloride of calcium would correspond approximately to gauging with a solution of 30 grams per liter in the previous experiments quoted. CaCl 2 GROUND DRY WITH THE CLINKER. Per Cent of CaCl 2 . Per Cent of Water. Initial Set, Minutes. Final Set, Minutes. 0.0 29.8 . 2 52 0.5 34.1 115 274 1.0 29.8 160 272 1.5 26.4 167 234 2.0 25.4 127 212 2.5 26.4 103 180 3.0 26.4 45 182 3.5 26.4 97 185 4.5 28.6 63 150 5.0 29.8 73 160 5.5 29.8 76 84 6.0 29.8 68 145 * Engineering News, vol. 53, pp. 13-14. Jan. 5, 1905 CLINKER COOLING, GRINDING, AND STORAGE. 547 "The experiments quoted indicate that chloride of calcium added in small percentages either to the ground clinker as a powder or mixed with the water for gauging has an important effect in extending the time of setting of Portland cement, and so far as the investigations which are accessible show it does not have any detrimental effect on the permanent strength and hardness. "Chloride of calcium is a deliquescent material which rapidly ab- sorbs moisture, and it is possible that if ground dry with the Port- land-cement clinker, even to the amount of J per cent, it would cause the material to gather dampness and thus have a bad effect. The chloride of calcium solution can be added readily by adding it to the water used in gauging, since it dissolves with extreme rapidity. The experiments indicate that the set can be controlled by using less than i per cent, which would be something less than 2 Ibs. to the barrel of Portland cement. Investigations are still necessary for determining whether the effect of chloride of calcium added to the cement before grinding is permanent in its effects, and whether if ground with the cement clinker it would avert any detrimental effect." List of references on use of calcium sulphate, chloride, etc. Candlot. Ciments et chaux hydrauliques. Carpenter, R. C. Recent experiments with materials which retard the activity of Portland cement. Engineering News, vol. 53, pp. 13-14. Jan. 5, 1905. Deval, L. Composition of sulpho-aluminate of lime (in hydraulic cements). Bull, de la Soc. d'Encourag. ITnd. National, vol. 5, pp. 49-54. 1900. Abstract in Jou- n. Soc Chem. Industry, vol. 19, pp. 247-248. Deval, L. Action of sulphate of lime on cements. Bull, de la Soc. d'Encourag. I'lnd. National, no. 101, p. 784-787. 1901. Abstract in Journ. Soc. Chem. Industry, vol. 21, p. 257. Deval, L. Influence of calcium sulphate on cements. Thonindustrie Zeitung, vol. 26, p. 913-915. 1902. Abstract in Journ. Soc. Chem. Industry, vol. 21, pp. 971-972. Lewis, F. H. Specifications for Portland cements and cement mortars. Proc. Engrs Club, Phila., vol. 11, pp. 310-346, 1894. Ljamin, X Abnormalities in the initial setting of cement. Thonindustrie Zeitung, vol. 26, pp. 874-876. 1901. Abstract in Jcurn. Sec. Chem. Industry vol. pp. 972-973. Nihoul, E., and Dufossez, P. Note en ohe retardation of se+tine: of Portland cement. Bull. Scient. de 1'Assoc. des Eleves des Ecoles speciales de Liege, no. 3, Abstract in Journ. Soc. Chem. Industry, vol. 21, pp. 859- 860. Rohland, P. Hydration of Portland cement. Zeits. angew. Chemie, vol. 16, pp. 1049-1055. 1903. Abstract in Journ. Soc. Chem. Industry, vol. 22, pp. 1244-1245. 548 CEMENTS, LIMES, AND PLASTERS. Storage, Packing, and Market. Necessity for storage. A twofold necessity exists for large storage space at a modern cement-plant. The cement will in many cases be improved by storage, particularly if it can be so stored that air will gain access to the mass. Aeration in tile storage building is, however, rarely possible; and in consequence the tendency now is in the direction of aerating or slaking the clinker before grinding. The main reason for storage still remains prominent. It is caused by the fact that while the average mill runs twelve months in a year, the cement-selling period, in most of the United States, is practically confined to six months or even less. This of course necessitates very extensive storage facilities enough to hold at least three months' output of the mill, and preferably to hold six months' production. This means that for each kiln in a dry- process plant, storage space for at least 20,000 barrels should be provided. As Portland cement dumped from a conveyor will pile up so as to weigh about 90 to 100 Ibs. per cubic foot, the storage space above stated (20,000 barrels) would be equivalent to about 80,000 cubic feet for each kiln in the plant. This is the minimum of space that can be given with safety, and an allowance of 150,000 cubic feet per kiln would be much better for the average plant in the Middle or Eastern States. In the South and West conditions are different, and much less storage space is required. Designs of storage buildings and bins. In Figs. 139, 140, 141, and 142 are given plans of several recently erected storage buildings and bins. Those shown in Fig. 139 are concrete-steel bins erected for the Portland-cement plant of the Illinois Steel Co. The stock-house and bins shown in Figs. 140, 141, and 142 are those of the Hudson Portland Cement Co. In a description of that plant accompanying these figures * the following data are given: "The stock- house is a structure 410 feet long and 105 feet wide, having as founda- tions a series of concrete walls founded on piles. It contains three groups of twenty bins each. The bins are of wood, their walls being laid up solid of 2"X10" and 2"X8" plank laid flat and spiked together. The total capacity of the bins is 200,000 barrels of cement. Cement is con- veyed to the stock-house from the mill by means of conveyor 26, which discharges into the boot of elevator 0. This elevator discharges into the transverse screw conveyor 27, which spouts onto two belt conveyors * Engineering News. vFRSHTY OF OALMW **t _ SECTIONAL PLAN FIG. 139. Concrete-steel bins Monier Roof Platen VERTICAL SECTION tool Co. (Engineering News.) [r^ /ac^ /. 548. CLINKER COOLING, GRINDING, AND STORAGE. 549 running lengthwise of the building over the bins. From these belts the cement can be deflected into any one of the 60 bins." CROSS SECTION OF MACHINERY BUILDING AND STOCK HOUSE V - I^T^I \ ^^ - y|ii TrusB^ ea^Q *\~ ElevatoKiaVT'CuBii Elevator 12"x7"Cup8 c,A4S&|gU-l | I rfra afej i b g i . i LONGITUDINAL SECTION OF STOCK HOUSE, SHOWING CONVEYORS AND ELEVATORS, rf Sprocket ator DETAIL OF PASSAGE WAY BETWEEN STOCK HOUSE AND MACHINERY BUILDING PART PLAN FIG. 140. Plan and sections of stock-house, Hudson Portland Cement Co., showing conveying machinery. (Engineering News.) Testing at the mill. The contents of each bin should be sampled as soon as it is filled, and the usual physical tests made on these samples, supplemented by analysis if need be. If the cement is weak or unsound, it is far better in every way to detect it at the mill than to run the risk of having it rejected at the work. The results of these tests are then forwarded to the purchaser with each shipment, typical blanks for this purpose being shown on page 550. Of recent years it has become the fashion, in specifications for cement for important work, to require that the purchaser should be represented at the cement-mill by an inspector during the manufacture and ship- ment of the cement covered by the specifications. This practice, if 550 CEMENTS, LIMES, AND PLASTERS. Sn | .I 1 1 1 | 1 .s ^s '. S V* pq . . . : . ijj 1 B 1 2 1 4 .:: "o ^ ? 5 03 s ^ 1 Q i> i -4- o- "^3 ^ EH' t- 1 G a o -f o| p t p l C 02 4 ' ^ &," C OJ a; CO .-, H-PtNv.s.'^ ^ S ^ t 'T ^^M c^ C^I ^O i-^ t-C ^^ a t^ "" ^ ^P fcC x ^ XN * -x%s ~^' "*^ ~*^ ^ C > G) IJ ||f| : : If! a ,T 2 S i S C -4- E ji i c R -1-2 -I a 3 e c 51 f far ^ S rT MM 9 c rt -V i g 2-? M ii 1 ? ::! n Vi M s j O -0) __ C 3 ^ C 3 C ; c d i fi ( SH 1 O > r : :J X ^ '- C 2 e:~ ^ g 2 g '-5 'SJ ~^- : - 3.= o.= 5 i = :f g- CLINKER COOLING, GRINDING, AND STORAGE. 551 taken up in a proper spirit by both sides, will prevent many difficulties and misunderstandings. If the duties of such an inspector are to be properly carried out, however, he should possess a somewhat intimate knowledge of the processes of cement manufacture. The position is, moreover, one of extreme delicacy, and it offers abundant opportunity for bribery on the one side and blackmail on the other. For these reasons it is obvious that a mill-inspector must be a man of special training and qualifica- tions and unimpeachable integrity. As a matter of fact, such men are rarely obtainable for this position, and the inspector is usually a young engineer, with but little knowledge of the points which it is necessary to guard against. Packing. According to the recently issued specifications of the American Society for Testing Materials, quoted later, Portland cement should be packed in bags of 94 Ibs. net weight, four, of which make a barrel of 376 Ibs. net. Several other important specifications require a barrel of 375 Ibs. net. In ordinary calculations it is often assumed for C.L. of Conveyor^ FIG. 141. Foundation plan of cement stock-house, Hudson Portland Cement Co. Platform Platform 20 9^ KM'9%- FIG. 142. Plan of bins in cement stock-house, Hudson Portland Cement Co. convenience that a barrel of Portland cement weighs 400 Ibs. gross or 3SC Ibs. net. The following table (209), published some years ago by Mr. Sanford Thompson,* gives the results of a series of actual tests on the weight, size, etc., of cement barrels. * Thompson, Sanford E. Weights of Portland cement and capacity of cement barrels. Engineering News, Oct. 4, 1900. For a more detailed discussion of the subject, reference should be made to Taylor and Thompson's "Concrete, Plain and Reinforced", 1905, pp. 216 et seq. 552 CEMENTS, LIMES, AND PLASTERS. TABLE CAPACITY OF PORTLAND-CEMENT BARRELS AND Number Barrels Tested ; Results Aver- aged. Brand. Height between Heads, Feet. Average. Capacity Barrel between Heads, Cu. Ft. Depres- sion, Cement below Head, Feet. Volume of De- pression, Cu. Ft. 0.171 0.059 0.096 0.093 0.039 0.235 0.148 Diam- etej of Barrel, Feet. Hori- zontal Area, Sq. Ft. 6 6 3 5 1 5 5 Giant 2.19 2.08 2.07 2.01 2.13 2.12 2.01 .430 .403 .412 .407 .38 1.437 1.455 1.605 1.546 1.571 1.554 1.496 1.622 1.662 3.495 3.219 3/249 3.123 3.186 3.446 3.327 0.12 0.04 0.07 0.07 0.03 0.17 0.10 Alsen Savior's Dyckerhoff Fiske Lion * Atlas . . . Aalborg f Final averages 2.09 1.418 1.579 3.292 0.09 0.120 * Hanover. t Denmark. Partial averages. TO be compared only At present a very large proportion of American cement is marketed in sacks or bags, packing in wood being almost entirely confined to cement inteijded for export or coastwise shipment. Foreign cement, of course, reaches the American market in barrels, but imported cements are becoming of less consequence each year. When cement is packed in wood, the barrels are usually charged for at a rate sufficient to give a slight profit to the packing department of the mill. Bags and sacks are also charged extra, with a rebate for re- turned bags. In figuring the cost of cement manufacture it is there- fore unnecessary to include the actual cost of the packages, but the labor and power used in the packing-house must be figured in. CLINKER COOLING, GRINDING, AND STORAGE. 553 209. WEIGHT OF CONTENTS. (HOWARD A. CARSON.) Volume of Cement per Barrel. Net Weight of Cement per Barrel at Dumping. Weight per CubicFoot. Weight Barrel, Pounds. Packed, Cu. Ft. Loose, Cu. Ft. Shaken, || Cu. Ft. Before, Pounds. After, Pounds. Packed, Pounds. Loose, Pounds. Shaken, Pounds. Sifted, Pounds. 3.347 3.161 3.152 3.031 3.147 3.211 3.206 4.173 4.192 4.052 3.989 4.270 3.754 4.058 3^522 3.695 3.432 3.598 381.0 374.2 387.0 373.2 378.0 377.4 370.7 371 A 378.0 376.9 370.2 113.81 118.45 122.75 123.16 120.11 117.54 115.71 91.38 89.20 94.24 93.18 88.52 100.49 91.40 29.0 24.3 22.7 25.6 22 21.1 23.3 105.54 102.29 109.45 102.94 90]6 80.3 3.179 4.070 3.562 377.4 374. 1 118.79 92.63 105. 06 85. 4 24.0 with like brands. II Box rocked over bar. Note. Paper weighs about 1 Ib. CHAPTER XXXVII. COSTS AND STATISTICS Costs of Manufacture of Portland Cement. To the popular mind the manufacture of Portland cement is, though fccuiewhat mysterious in process, so remarkably cheap as to be a very attractive investment. This opinion is carefully fostered for selfish purposes by many so-called "cement experts" and " cement engineers", men of a class comparable with the "mining experts" who infest the Western mining districts and differing in degree only, not in kind, from the more humble green-goods man of the East. Michigan, especially, has been the prey of this class of cement experts, for in no other State has so much money been sunk by comparatively poor people in the erection of unprofitable Portland-cement plants. Many of these com- panies have been reorganized and are now on a firm basis, but the amount of money thrown away by ignorant investors, misled by lying estimates and prospectuses, has been enormous. The truth of the situation appears to be, on the contrary, that there is a very small margin of profit for most of the plants now engaged in Portland-cement manufacture. It is probable, indeed, that the per- centage of profit In the lime or natural-cement industries is far greater than in the average Portland- cement piant. In the following pages some estimate of the cost of Portland-cement manufacture under various conditions will be given. Factors of cost. The total cost of manufacture of Portland cement may be subdivided as follows: A. Fixed charges. (a) Interest on cost of land and quarries; (6) Interest on cost of construction and erection; (c) Interest on reserve capital invested; (d) Allowance for sinking fund; (e) Insurance and taxes. 554 COSTS AND STATISTICS, 555 B. Current administrative expenses. (a) Salaries of administrative officers; (b) Expenses of sales and advertising department. C. Current factory costs. c 1. Cement materials; (a) Raw materials \ 2. Coal; (.3. Gypsum; (b) Supplies and repairs; (c) Labor; (d) Mill-office and laboratory expenses. Cost of land and quarries. The cost of land, including under this head both the land on which to locate the plant and also the land con- taining the supplies of raw material, is naturally the most > variable item of the* lot. Much will depend on the manner in which the nego- tiations are conducted. When marl lands are hunted with the aid of a brass band and a press bureau prices as high as $40 per acre (contain- ing about 9 feet of workable marl) have been paid. On the other hand, lands located on transportation routes and underlaid by 40 feet of lime- stone and more than enough shale have been purchased as farming land at about $5 per acre, the seller concealing from the guileless purchaser the fact that the soil was too thin for good farming. In the Lehigh district, where little cement land remains unbought, prices varying from $100 to $500 per acre have been quoted, but I have no means of determining the true average value of land in this district. Examination of a number of prices paid for land by cement-plants shows that, considering both area and workable thickness, the price paid usually falls between J and ^ cent per cubic yard of raw material. Land for the plant itself may be expensive, for good mill-sites are scarce. In this connection it may be well to recall some estimates previously made on the amount of raw material required by a plant. It was stated that each kiln of a dry-process plant will use about 190,000 cubic feet of limestone per year, equal to a thickness of 4^ feet over an acre, plus 80,000 cubic feet of shale or clay, equivalent to an acre 2 feet thick. A wet-process plant will use about 450,000 cubic feet of marl (measured in place) plus about 45,000 cubic feet of clay or shale. In buying quarry lands for a new plant, at least a twenty-year supply of workable raw material should be secured. Cost of equipment and erection. Exclusive of the cost of land, the cost of equipping and erecting a good plant will usually fall within the limits of $50,000 to $80,000 per kiln. These limits may seem wide, but 536 CEMENTS, LIMES, AND PLASTERS. it is difficult to make a closer general estimate. Two plants recently built cost as follows: 4-kiln plant, total cost. . . $287,000 Cost per kiln $71,750 6-" " " " ... 373,000 " " " 62,167 For small plants of 2 to 6 kilns each such costs would not be excep- tional. For larger plants it" is to be^ remembered that cost per kiln decreases with increase in the number of kilns. The following table of average costs will serve to exemplify this and may be of use as a basis for general estimates : 2-kiln plant $70,000 to $80,000 per kiln 4-" " 60,000" 70,000 " " 6-" " 50,000" 60,000 " " 8-" " or over 45,000" 50,000 " " The distribution of this total throughout the plant may be shown by the following estimate for cost of construction of a 6-kiln dry- process plant: Quarry, track, and trestle $15,000 Crusher and mixing 22,100 Raw and clinker mills 63,640 Kiln mill 53,080 Coal mill 15,920 Power-plant 63,000 Stock-house 25,000 Office, laboratory, etc 10,000 $267,740 Total capital required. The amount of capital required to prop- erly float a cement proposition is considerably in excess of the costs of land, construction, etc. The principal causes of this condition are: (a) It is within bounds to say that the average cement-plant will not produce a normal cement at a normal cost for a considerable period (varying from 3 to 6 months or even longer) after the plant h first put in operation. Both the machinery and the personnel of the plant will require numberless (though individually small) alterations before good work can be accomplished. The plant must be carried through this profitless and expensive period entirely on its reserve capital. (6) It is becoming more and more the fashion among engineers to judge a cement by its past record, and to refuse bids from a plant not possessing a record of success in actual work. Even after the plant is working normally, therefore, steady sales cannot be counted on for some time. The intervening time can, of course, be devoted to filling large stock-houses, but this brings in no ready money to the plant. (c) Cement is sold on comparatively long time, while many of the COSTS AND STATISTICS. 557 expenses of the plant must be paid in cash. This is particularly the case with regard to quarry and mill labor, an item which alone may amount in a six-kiln plant to from $4000 to $6000 per month. For these reasons it is advisable to make a very liberal allowance for the working capital required. A reserve amounting to $20,000 to $25,000 per kiln will probably be found sufficient to cover most cases, though in commencing work on absolutely new materials a consider- ably larger reserve will be found advisable. Adding this ncces~ary reserve to the amount required for the purchase of land and the actual erection of the plant, and for ordinary small plants, it will be found that the total working capital necessary will be in the neighbor- hood of $100,000 per kiln. Current administrative expenses. This group of expense factors- includes the salaries of the administrative officers of the plant and also the expenses of the sales department, including advertising. It is obviously one of the most variable factors in the cost of cement-man- ufacture as between different plants. Even at the same plant it may vary widely from time to time, because its total amount does not depend directly upon the amount of cement sold. In a large plant during a prosperous year the costs chargeable to this group of expenses may fall as low as 5 cents per barrel of cement produced, while in a small plant running undertime they may easily rise as high as 15 cents per barrel. Cost of excavating raw materials. This point has been covered quite fully on preceding pages. In the present place it will only be necessary to summarize this information by stating that the total cost of raw materials, delivered at the mill, per barrel of cement produced, will usually fall within the following limits : Cost per Barrel. Cement rock and limestone. . Pure limestone and clay. . . . Marl and clay or shale. a. Limestone on property . . $0 . 07 to $0 . 10 b. Limestone purchased ... .15 a. Quarried 06 b. Mined 10 a. Both on property 03 b Clay purchased 08 .20 .15 .20 .05 .15 Total power required. The total power developed in a Portland- cement plant is not far from 1 H.P. for each barrel of cement turned out per day. As most of the power is required continuously, it is &afe to figure on a requirement of 20 to 30 H.P.-hours for each barrel of cement. This will include the power used in grinding and mixing the raw materials in running conveyors, kilns, etc., and in grinding the clinker. At most American plants all this power is derived from the direct consumption of coal under boilers. At several plants, however, water 558 CEMENTS, LIMES, AND PLASTERS. power is used with electric transmission; in at least one plant natural gas is used, partly in gas-engines and partly under boilers, and in a few plants part of the power is derived from the waste gases of the kilns. Cost of coal for kilns and power. The total amount of coal used per barrel of cement, both under the boilers, in driers, and in the kilns, may vary from 170 Ibs. in a-'Lehigh-dis.trict plant using long kilns to 300 or 350 Ibs. in a wet-process plant. Even this latter figure is exceeded at times, for one wet-process plant recently reported a total coal consump- tion of over 400 Ibs. per barrel. Cost of gypsum used. The amount of gypsum or plaster used in a large plant is quite an item, as 8 to 14 Ibs. may be used per barrel of cement. Crude gypsum in small lumps may be bought as low as 80 cents or so per ton when the cement-plant is located in a gyp- sum r producing district. On the other hand, burned plaster plus freight may cost as high as $6 to $10 per ton for a plant not so near the source of supply. Assuming that on the average 10 Ibs. of gypsum or plaster are used in each barrel of cement, these prices would make the cost per barrel of cement range from J cent to 5 cents. At most of the plants in this country the gypsum cost will probably range between 1 and 2 cents per barrel of cement. Distribution and cost of labor. The following statement* of the distribution and amount of labor in a six-kiln Lehigh-district plant has recently been published. The number of men given are those actually employed in one twenty-four-hour day (two shifts), and the cost per barrel has been based on an output of 1200 barrels per day. TABLE 210. LABOR COSTS IN CEMENT-PLANT. Location. Number of Men. Labor Cost per Barrel. Stone house Raw mill . 4 6 $0.005 0075 Coal mill 6 01 Kiln-room 8 015 Clinker mill 6 0075 Boiler-room 4 0075 Engine-room 4 0075 Yard o'ano: 13 015 Repair gang Packing-house 13 30 f .0225 04 94 $0.1375 * Boilleau and Lyon. Cost of building and operating a Portland-cement plant. Municipal Engineering, vol. 26, pp. 391-395. June, 1904. f Contract work (estimated). COSTS AND STATISTICS. 559 The total labor costs given in the footing of the above table may profitably be compared with those of two large and admirably managed plants elsewhere. This is done in the following summary: TABLE 211. LABOR AND OUTPUT IN TYPICAL PLANTS. Output per Day. Process. Number of Men. Barrels per Man. Cost per Barrel. 1-00 bbls. Drv 94 12.78 $0.1375 2800 ' ' 11 159 17.61 .1216 1GOO " Wet 142 11.27 .135 It is to be noted that the second of these plants uses three shifts per day, but its mechanical equipment is so perfect that this does not show in 'the labor costs. Estimates of Total Cost per Barrel. Disregarding the remarkably low estimates of cost to be found in prospectuses, several estimates of better quality have been published within the past few years. These relate to manufacture carried oh under very different conditions, two applying to the Lehigh district, with cheap labor and fairly cheap fuel, while a third estimate is for a plant located in perhaps the worst possible spot in the country for fuel, labor, and freight. To these published estimates has been added a table prepared by the present writer. A partial account * of costs at the Atlas plant was given by Messrs. St anger and Blount a few years ago. This is as follows: TABLE 212. COSTS OF CEMENT MANUFACTURE AT ATLAS PLANT. Cost per Ton, English Money. Cost per Barrel. s. d. 1 95 $0 076 Fuel for power 3 125 Fuel for kiln 3 125 Labor 2 082 Repairs, lubricants, etc 2 11* .124 Superintendence laboratory and mill -office 1 If 048 13 11 .58 The following estimate, f recently published by Boilleau and Lyon, is on the basis of a 2000-barrel plant (ten 60-foot kilns) located in the Lehigh district. * Proc. Institution Civil Engineers, vol. 145, pp. 65-66. 1901. f Municipal Engineering, vol. 26, p. 394. June, 1904. 560 CEMENTS, LIMES, AND PLASTERS. TABLE 213. DETAILED ESTIMATES OP COSTS OF CEMENT MANUFACTURE (2000-BARREL PLANT), Labor : Cost Per Barrel. Quarry $0.05 Stone house 00? Mill building 01$ Kiln-room ., 01$ Engine- and boiler-room /f 01$ [ $0. 15 Fuel mill .' 01 , Yard gang 01$ Repair gang 02^ Miscellaneous OQ J . Raw materials : Coal : 22$ \ 24 Gypsum Olf J Supplies : Repair parts 04 ") Lubricants 02 Y .09 Miscellaneous 03 J General mill account. Packing and shipping 04 1 Qg Office force 02 Fixed charges: Interest on cost of plant 07 Sinking fund 05 Depreciation 05 .23$ Administration and selling expenses 06$ J Total $0.77f Estimates have been published by Mr. E. Duryee, covering the cost of building a small plant and making cement at a very inaccessible point in Arizona. The raw materials are located quite near the plant, but fuel, labor, and freight are all very high. These estimates are as follows : TABLE 214. APPROXIMATE COST OF TWO-KILN PLANT, DAILY CAPACITY 300 BARRELS. (DURYEE.) Crusher $2,000 Mill for disintegrating clay 500 Rotaiy clay drier 1 ,500 Elevators and conveyors for raw materials 1 ,000 Storage-bins for raw materials 1 ,000 Mills for grinding raw materials 10,000 Two rotary kilns, linings, and stacks 15,000 Mills for grinding cement 10,000 300-H. P. electric motors and step-down transformers. 9,000 Conveyors and elevators for cement 1,500 Cost of grading and erecting machinery 10,000 Shafting and pulleys, belts, and setting up 5,000 Buildings and cement-bins 10,000 Office, laboratory, and laboratory equipment 1,500 Freight. ... 10,000 Plans, specifications, and superintendence 3,000 $91,000 COSTS AND STATISTICS. 561 "With such a mill, and using charcoal for burning cement, the cost of manufacture would be approximately as follows, allowing for power only the proportion necessary for maintaining and operating the electric plant provided": Per Barrel. Labor and superintendence $0 . 70 Raw materials 30 Fuel for burning 90 Power (maintenance and operation only) 05 Repairs and sundries 05 $2.00 This estimate is of interest in connection with the possible erection of plants at various points in Spanish America, where similar conditions may be encountered. The following general estimates have been prepared embodying the data presented on pages 554 to 559. It will be seen that they apply, except for the second and fifth cases, to very small plants, and that the totals are correspondingly high. TABLE 215. ESTIMATES OF AVERAGE COST OF PORTLAND-CEMENT MANUFACTURE. Materials { Lime- ( stone i Cement rock and 1 Marl and Marl and Marl and Process and clay [ Dry limestone Dry clay Dry clay Wet clay Wet Number of kilns 4 7 S 7 4 4 8 Size of kilns Output per day . . CO feet 700 bbls 80 feet 2000 bbls 60 feet 600 bbls 60 feet 400 bbls 100 feet 1100 bbK Cement materials. . . . 08 08 03 03 03 Power coal ] f Drier coal ^ at $2. . \ Kiln coal J Labor 0.08 0.02 0.12 12 0.08 0.01 0.10 10 0.09 0.07 0.14 16 0.13 0.00 0.20 20 0.11 0.00 0.15 15 Supplies, etc Office and laboratory. . Admin, and sales Interest, etc . . . 0.15 0.05 0.08 16 Jll 0.03 0.05 12 0.15 0.05 0.10 20 0.16 0.05 0.13 26 0.12 0.04 0.09' 20 0.86 0.68 0.99 1.16 0.89 Statistics of the American Portland-cement Industry. Statistics relative to the growth of the Portland-cement industry in the United States will be found serviceable for many purposes, and certain data on this point have been accordingly incorporated in the present chapter. Total production and growth of the industry. The following table has been prepared to illustrate the growth of the American Portland- 562 CEMENTS, LIMES, AND PLASTERS. cement industry, from its inception to the present day, in regard to number of plants, annual output, and value of product. The data used in compiling this table are taken in large part from the annual report on mineral resources issued by the United States Geological Survey. For the years prior to 1890 the figures are to be regarded as merely rough approximations to the, truth, but in later years they may be accepted as practically correct. : TABJ.E 216. TOTAL PRODUCTION OF PORTLAND CEMENT IN THE UNITED STATES. Years. Barrels. Value. Plants. 1870-1880.. 1880 82,000 42,000 $246,000 126,000 1881 60,000 150,000 1882 85,000 191,250 1883 90,000 193,500 1884 100,000 210,000 1885 150,000 292,500 1886 150,000 292,500 1887 1888 .... 250,000 250,000 487,500 487,500 1 889 300,000 500,000 1890 335,500 704,050 16 1891 1892 454,813 547,440 967,429 1,153,600 17 16 1893 590 652 1,158,138 19 1894 1895 1896 . . . 798,757 990,324 1,543,023 1,383,473 1,586,830 2,424,011 24 22 26 1897 2,677,775 4,315,891 29 1898 3,692,284 5,970,773 31 1899 5,652,266 8,074,371 36 1900 8 482 020 9,280,525 50 1901 12,711,225 12,532,360 56 1902 17,230,644 20,864,078 65 1903 .... 22,342,973 27,713,319 71 1904 26,505,881 23,355,119 75 106,114,077 $124,660,717 Production of Portland cement by States, 1901-1903. In the table on the page opposite, which is slightly revised * from one published by L. L. Kimball in the " Mineral Resources of the U. S. for 1903", the Portland-cement production of the separate States is given for the years 1901, 1902, and 1903. In such States as have but a single plant their production is com- bined with that of another State, in order that the separate figures of any plant shall not be revealed. In this table the Portland-cement * The revision is the correction, in the third column from the right, of the number of plants in 1903, "which in the original was evidently incorrect. COSTS AND STATISTICS. 563 PRODUCTION OF PORTLAND CEMENT IN THE UNITED STATES IN 1901, 1902, AND 1903, BY STATES. i i os" co i-T i i TF 00 l> CD ^ CO C^ O O O^ CO C^ CO ri O^ ^^ Oi C^ CO 27,713,319 le single plants in Alabama and Georgia. " " " Missouri and South Dakota, plant " Utah, only Portland-cement plant in Kansas. " Virginia, single plant in South Dakota. ^^^(M^H (N coo COM O !> r-i CO OS CO 1-1 O CO 00 00 "7) 00 "t 1 I-H i i CO IO 1-1 CO i i (M CO OS O CO I-H ioi>i itociosoc^^o co (M^O^O^OS^OO cq^CD^t^ t^ iO i-H i-H i 1 i 1 C^l i 1 OS 22,342,973 Number of Works. i i i i CO I-H i * COi-H COCN 1 a oo i-H r-H i i iO coo Tfl , 1 OS O O_i-^ ilS 1 co" I-H" o" o" TJH" co" CO (N 00 CO CO CO. ofi-T o" oo o o" li OS 00 1> I> O O O CO CO 1> 00 1 s * CO ^t 1 O OS O O I-H O O O i i oo I-H rfi 10 oo C^l O CO O >o" Tp" O iO O t> CO CO CO (N i-T co" 1 i-H * "S 1 jj 00 O 00 O O C^ 00 (MO r-( OS rf t>- 00 o o" 00 00 *O (N I-H Tos"T-H uf i 1 1 i oo os os CO O CO O i-i 5 c Number of Works. i-H i 1 i-H ^ (M I-H O 10 ^ :E 61 E^.sVsl)S,s2 ~. Q : :-S> 53 C O S fcJD'o 5 3 .-< 0^^ 02 r-i 02 C 564 CEMENTS, LIMES, AND PLASTERS. product in 1903 of the only plant in Alabama which produces that variety of cement is combined with the product of the plants in Georgia, Virginia, and West Virginia. The plants in Missouri and Arkansas have their products combined; those in Kansas and Texas, .and those in Utah, South Dakota, and Colorado also show combined products, and in each case the result^ is given in connection with the State which was the largest contributor to the total product. Production in the Lehigh district, 1890-1903. The following table has been prepared by the writer to illustrate the position held by the Lehigh district of Pennsylvania-New Jersey: TABLE 218. PORTLAND-CEMENT PRODUCTION OF THE LEHIGH DISTRICT, 1890-1903. "Year. \ Lehigh District. Entire United States. Percentage of Total Product Manufactured in Lehigh District. Number of Plants. Number' of Barrels. Number of Plants. Number of Barrels. Value. 1890 . . 1891 1892 1893. 1894 .... 5 5 5 5 7 8 8 8 9 11 15 16 17 17 201,000 248,500 280,840 265,317 485,329 634,276 1,048,154 2,002,059 2,674,304 4,110,132 6,153,629 8,595,340 10,829,922 12,324,922 16 17 16 19 24 22 26 29 31 36 50 56 65 71 335,500 454,813 547,440 590,652 798,757 990,324 1,543,023 2,677,775 3,692,284 5,652,266 8,482,020 12,711,225 17,230,644 22,342,973 $439,050 1,067,429 1,152,600 1,158,138 1,383,473 1,586,830 2,424,011 4,315,891 5,970,773 8,074,371 9,280,525 12,532,360 20,864,078 27,713,319 60.0 54.7 51.3 44.9 60.8 64.0 68.1 74.8 72.4 72.7 72.6 67.7 62.8 55.1 1895 1893 1897 1898...... 1899 1900 1901 1902 1903 Imports of cement, 1899-1903. The imports of cement into the United States in 1899 to 1903 inclusive, by countries, were as follows: TABLE 219. IMPORTS OF HYDRAULIC CEMENT INTO THE UNITED STATES IN 1899, 1900, 1901, 1092, AND 1903, BY COUNTRIES. Country. 1899. 1900. 1901. 1902. 1903. United Kingdom Barrels. 199,633 Barrels. 267,921 Barrels. 37,390 Barrels. 79,087 Barrels. 146,994 Belgium 624,149 826,289 303,180 615,793 737,576 Trance 15,649 32,710 11,771 14,922 14,865 'Germany . . . 1,193,822 1,155,550 555,038 1,259,265 1,377,414 Other European coun- tries British North America. . Other countries 68,348 4,398 2,389 75,827 4,517 23,869 19,077 6,066 6,808 17,956 3,611 4,153 27,415 4,421 9,265 Total 2,108,388 2,386,683 939,330 1,994,787 2,317,950 COSTS AND STATISTICS. 565 "The figures used in compiling this table are those which show the total imports. In 1903 England stands third in the list of foreign coun- tries which sent cement to America. From 1871 to 1876 nearly all importations of foreign cement were from England. In the four years following Germany gradually assumed an important place as rival, and in 1882, while England sent one half the cement exported to the United States, Germany sent three fourths of the remainder. Ten years later Germany was the leading foreign country sending cement to America, and since theii has held that position/' CHAPTER XXXVIII. CONSTITUTION, SETTING PROPERTIES, AND COMPOSITION OF PORTLAND CEMENT. Limitations of chemical analyses. An ordinary chemical analysis of a specimen of cement will determine what elements are present in the cement, and in what percentages these various constituents are represented. The comparison of a long series of such analyses, such as is presented later in this chapter, will enable certain conclusions to be drawn as to the probable limits of composition of good Portland cements; and an analysis of a single cement may show that it contains undesirable ingredients or that inert material is present in undesirable quantity. But these methods of investigation fail to give the least information concerning the real constitution of Portland cement as distinguished from its composition; they give no information whatever as to the man- ner in which the various elements are combined among themselves. They fail, moreover, to give any clue to the reason why certain mixtures give good cements, while others give weak or unsound products; and they afford no explanation of the "hydraulic" or setting properties which the powdered clinker possesses. It is evident, therefore, that other methods of investigation must be adopted, since even the most careful chemical analysis fails to aid us in this line of research. Constitution and Setting Properties. Available methods of investigation. Two distinct methods of in- vestigation are available microscopic and synthetic. The first has been applied with great success by geologists to the study of the igneous rocks, and as cement clinker is practically an arti- ficial (though very basic) igneous rock, the microscope can be used successfully in its examination. By grinding normal clinker of known analysis down to thin transparent slices, the microscope is able to detect certain constituents common to all good clinkers. The next step, of course, is to determine the composition of these different constitu- ents, and here the synthetic method is applicable. 566 CONSTITUTION, ETC., OF PORTLAND CEMENT. 567 In synthetic work pure lime, silica, alumina, etc., are mixed in certain definite proportions and burned to a clinker. The hydraulic properties of this clinker can be examined by powdering part of it and testing the resulting cement. Examination under the microscope will fix certain optical characters peculiar to each clinker composition, and the data thus obtained can be used to determine the constituents of commercial-cement clinker, as noted in the preceding paragraph. Synthetic investigations. Richardson has recently summarized the results of his own studies and those of previous observers as follows : "The preparation of synthetic silicates and aluminates which might exist in Portland cement was carried out to a certain extent by Le Cha- telier and the Newberrys, but in neither case were these compounds characterized completely, especially as to their optical properties. This has been done by the writer within the last two years, and the optical properties and other characteristics of the following definite silicates and aluminates have been determined. " Monocalcic silicate (Si0 2 CaO) : A crystalline substance of high optical activity and little or no hydraulic properties. Specific gravity 2.90. "Dicalcic silicate (SiO 2 2CaO, or more probably 2SiO 2 4CaO): A definite crystalline compound of high optical activity and of very little hydraulic activity except in the presence of carbonic acid, but setting slowly in water, generally lacking volume constancy. Specific gravity 3.29. "Tricalcic silicate (Si0 2 3CaO) : A definite crystalline silicate of low optical activity and corresponding in this respect with alit. Its hy- draulic activity is not great, but greater than that of dicalcic silicate. If fused and reground it sets slowly like Portland cement. Specific gravity 3.03. "Three definite silicates of calcium, therefore, appear to exist, the two more basic ones being strongly differentiated from each other by their optical activity. "Monocalcic aluminate (A^OaCaO): This aluminate is a crystal- line substance of high optical activity, but is not sufficiently basic to permit of its existence in a material of such basic character as Portland- cement clinker. Specific gravity 2.90. "Tricalcic dialuminate (2Al 2 3 3CaO) : This aluminate is one of highly crystalline character and of great optical activity, making it readily recognizable. Specific gravity 2.92. "Dicalcic aluminate (Al 2 O32CaO): A substance crystallizing from a state of fusion in dendritic forms having no optical activity and being, 568 CEMENTS, LIMES, AND PLASTERS. therefore, iso tropic. This differentiates this alumina te very sharply from the preceding one and makes the identification of the two materials very easy. Specific gravity 2.79. "Tricalcic alumina te (A^OaSCaO): This aluminate crystallizes from the fused condition in elongated octahedra. It is isotropic and it might at first be assumed that it was not a definite compound, but merely the dicalcic aluminate crystallizing out of a magma of indefinite composi- tion. It has been shown, however, by further investigations too lengthy to go into at this point, to be undoubtedly a definite aluminate. Specific gravity 2.91. "Definite compounds of iron and lime and alumina and magnesia have also been shown to exist, but their consideration here is unneces- sary, as the constitution of Portland cement can be better discussed, theoretically, by a study of clinker, into which these elements do not enter. "Among the theories advanced as to the constitution of Portland cement there are those which assume the presence of certain so-called silico aluminates, such as 2Si0 2 , 2Al 2 Os, GCaO, and others of less basic form. All of these proposed compounds have been prepared by the writer and found not to be definite chemical compounds nor to corre- spond in any way with any of the mineral entities found in industrial clinker. They are in fact only solid solutions, of aluminates in silicates, of indefinite structure." Microscopic investigations. Le Chatelier, Tornebohm, and Richard- son have studied both industrial and synthetic clinker under the micro- scope, and the results of these preliminary studies have thus been sum- marized by the last-named investigator: "By this method of study Le Chatelier, and, at the same time inde- pendently of him, Tornebohm identified in Portland-cement clinker four distinct mineral constituents which Tornebohm described as fol- lows, naming them Alit, Belit, Celit, and Felit: "Alit is the preponderating element and consists of colorless crys- tals of rather strong refractive power, but of weak double refraction. By this he means that alit in polarized light between crossed nicol prisms has insufficient optical activity to produce more than weak bluish gray interference colors. " Celit is recognized by its deep color, brownish orange. It fills the interstices between the other constituents, being the magma or liquid of lowest freezing-point out of which the alit is separated. It is strongly double refractive, that is to say, gives brilliant colors when examined between crossed nicol prisms. CONSTITUTION, ETC., OF PORTLAND CEMENT. 569 " Belit is recognized by its dirty green and somewhat muddy color and by its brilliant interference colors. It is biaxial and of high index of refraction. It forms small round grains of no recognized crystalline character. " Felit is colorless. Its index of refraction is nearly the same as that of beiit and it is strongly double refractive. It occurs in the form of round grains, often in elongated form, but without crystalline outline. Felit may be entirely wanting. " Besides these minerals an amorphous isotropic mass was detected by Tornebohm and Le Chatelier. It has a very high refractive index. " Tornebohm adds ttoe important fact that a cement 4 per cent richer in lime than usual consists almost entirely of alit and celit." Theories of constitution. Until recently Portland-cement clinker was commonly assumed to be a mixture of two or more definite chemi- cal compounds, and the principal points at issue between various inves- tigators were: (1) the exact formulas for these compounds, and (2) the proportion in which they must exist to give a good Portland cement. The two theories, in this regard, that have made the most impression upon modern cement practice are those presented respectively by Le Chatelier and Newberry. Recently, however, Richardson has formu- lated a theory of entirely different type. These three explanations of the constitution of Portland clinker will, therefore, be described briefly. Le Chatelier, speaking of Portland-cement clinkers, states * that;: " Examined in thin plates under the microscope they are formed of tricalcie silicate in crystals, with very feeble double refraction, embedded in a crystalline ground-mass of silico-alumina ferrites of lime. These are the two essential elements of Portland cement. If the lime is in excess, aluminate of lime is first formed; then for a still greater excess, ferrite of lime, and finally free lime. If, on the other hand, the lime is deficient in quantity, a dicalcic silicate is formed, recognizable by the spontaneous crumbling of the burnt pieces of cement. When the mixture is imperfect or the burning insufficient, the reactions jemain incomplete, and although the average composition may be suitable, there is a simultaneous production of free lime and aluminate of calcium with dicalcic silicate. In a Portland cement of normal com- position the proportion of lime, according to the chemical formulas of the compound, should be greater than that determined by the follow- ing formula: Si0 2 -Al 2 3 -Fe 2 03^ Cao >3 ' W * Trans, Anier. Inst. Mining Engineers, vol. 22, pp. 3-52. 570 CEMENTS, LIMES, AND PLASTERS. in which CaO, Si0 2 , A1 2 O 3 , Fe 2 O 3 represent not the equivalent weights but the number of equivalents of these substances present; that is to say, the quotients of the weights of the substances divided by their equivalent weights.* This proportion of lime must never, on the other hand, reach the relation indicated by the following formula: (2) which corresponds to the exclusive formation of aluminate of calcium. It is necessary, by reason of the inevitable imperfection of the mix- ture, to keep always well below this limit, beyond which there will remain uncombined lime. In the use of this formula, magnesia should be added to the lime and sulphuric acid to the denominator after dividing its number of equivalents by 3." On a later page Le Chatelier states that Portland cements of good quality would give a value between 3.5 and 4.0 for formula 1, and between 2.5 and 2.7 for formula 2. The Newberrys, working on synthetic cements prepared from pure raw materials, obtained results differing from those of Le Chatelier in one important particular. They agreed with him that the lime and silica combined in the form of the tricalcic silicate 3CaO.Si02; but in regard to the lime-alumina compound they decided that it was present as the dicalcic aluminate 2CaO.Al 2 Os instead of in the tricalcic form given by Le Chatelier. These results gave, as the general formula for a pure Portland, *(3CaO.SiO 2 ) +2/(2CaO.Al 2 3 ). No allowance is made for magnesia, as the experimenters decided that it could not give a hydraulic product if present in Portland cement; while iron is neglected because of the small percentage in which it usu- ally occurs. Richardson, working both with the microscope and with synthetic preparations, has evolved a theory of much ingenuity and complexity by treating the investigation as a study in solid solutions. For the details of this remarkable and important work reference should be made to his original papers. t In the present place only a brief sum- mary of his principal conclusions can be given. He believes that the two principal constituents of a good Portland- * For a table of combining weights, see p. 11 of this volume, t See list on p. 575. CONSTITUTION, ETC., OF PORTLAND CEMENT. 571 cement clinker are the materials identified under the microscope by Tornebohm and named alit and celit; that alit is a solid solution of tricalcic aluminate (SCaO.A^Os) in tricalcic silicate (CaO.Si0 2 ), while celit is a solid solution of dicalcic aluminate pCaO.A^Os) in dicalcic silicate (2CaO.SiO2). "Having determined that alit and celit are solid solutions of alu- minates in silicates, the aluminates being present in less than an amount sufficient to make a saturated solution of aluminate in the silicate, it becomes of interest to consider how these solutions are formed during the conversion of a raw mixture or of a mixture of pure chemicals into a clinker. It would be simple to understand this if fusion took place in its formation, but this does not happen, the material is only sintered. If two gases are brought together they diffuse into each other with very great rapidity. If two liquids are poured one upon the other in layers without mixing, they diffuse more slowly. If solids are brought into contact it would be naturally assumed that diffusion would cease. Experi- ments of Robert-Austen have shown that molecular mobility in solids exists, since when carefully polished surfaces of gold and lead are brought into contact and left under pressure for some months, at the ordinary temperatures, gold is diffused into the lead and the lead into the gold for an appreciable distance. Mixtures of the components which would produce a fusible wood metal when subjected to pressure at ordinary temperature become converted into this alloy. Anhydrous sulphate of soda and carbonate of barium also diffuse when brought into close contact with the formation of barium sulphate and carbonate of soda. It is not difficult to understand, therefore, how at a temperature of 1650 C. the particles of silica, alumina, and lime may diffuse below the melting-point of the resulting clinker to form a Portland cement, and the fact that such a clinker is stable depends not only on its com- position, but upon the fact that the diffusion has been complete, even in material which is only sintered. Sintering, therefore, may be defined as diffusion at a temperature below the melting-point of the compo- nents or of the resulting solid solution. That diffusion under such con- ditions is surprisingly rapid is seen by placing a particle of ferric oxide on the surface of white Portland-cement clinker, and then submitting it to a moderately high temperature. The rapid diffusion of iron through the white clinker can readily be noticed by the color which spreads through the mass. It is evident that the higher the temperature the more rapid the diffusion until it becomes very rapid on fusion. From this it may be concluded that the length of time during which it is necessary to expose any mixture of silica alumina and lime to a tern- 572 CEMENTS, LIMES, AND PLASTERS. perature is a function of the temperature, and should be longer, the lower the temperature." Setting properties of Portland cement. The theory which has been quite generally accepted as explaining the setting of Portland cement was that advanced by Le Chatelier. He considered that the aluminate of lime, in contact with water, hydjated and hardened like plaster, according to the equation: 3CaO. A1 2 O 3 + 12H 2 O = 3CaO. A1 2 O 3 . 12H 2 0. i. To this action of the aluminate was ascribed the initial set of the cement. The later hardening was ascribed, however, to the decomposition of the lime silicate. In contact with water it sets, dividing so as to give hydrated monocalcic silicate crystallizing in microscopic needles, and calcium hydrate crystallizing in large hexagonal plates : 3CaO.SiO 2 + water = CaO.SiO 2 .2JH 2 + 2(CaO.H 2 O) . In general this theory, has been accepted. Richardson, however, has recently modified* this theory in an im- portant way. He considers that the setting of Portland cement is due to the decomposition of the silicates and aluminates of the clinker by the action of water, producing lime hydrate (Ca 2 H 2 2 ) in a peculiarly active form. "On the addition of water to a stable system made up of the solid solutions which composed Portland cement a new component is intro- duced which immediately results in a lack of equilibrium, which is only brought about again by the liberation of free lime. This free lime the moment that it is liberated is in solution in the water, but owing to the rapidity with which it is liberated from the aluminate, the water soon becomes supersaturated with calcic hydrate, and the latter crystallizes out in a network of crystals which binds the particles of undecomposed Portland cement together. From the characteristics of the silicates and aluminates it is evident that the latter are acted upon much more rapidly than the silicates, and it is to the crystallization of the lime from the aluminates that the first or initial set must be attributed. Subsequent hardening is due to the slower liberation of lime from the silicates. If the lime is liberated more rapidly than is possible for it to crystallize out from the water, expansion ensues and the cement is not volume constant." * Richardson C. The setting or hydration of Portland cement. Engineering News, vol. 53, pp. 84-85. Jan. 26, 1905. CONSTITUTION, ETC., OF PORTLAND CEMENT. 57$ He further notes that of the two constituents of the clinker the celit. is almost inert, being usually unattacked by the water, while the alit furnishes most of the lime needed for the setting effect. As the celit is a solution of dicalcic salts (2CaO.SiO2 + 2CaO.Al203), while the alit is a solution of tricalcic compounds (3CaO.Si02 + 3CaO.Al 2 O 3 ), the lower- limed cements are, therefore, the less hydraulic. This agrees with experience. This theory differs from Le Chatelier's in that it considers setting as due only indirectly to the presence of silicates and aluminates. "The strength of the Portland cement after setting is due entirely to the crystallization of calcium hydrate under certain favorable conditions, and not at all to the hydration of the silicates or the aluminates, since in this act of hydration nothing can take place which would tend to bind these silicates and aluminates together." The formation of lime silicates and aluminates during clinkering is on this theory only a convenient way of securing indirectly a very active lime hydrate, which is itself the real cementing material. Replacement of silica by other acids. Various oxides of the silica group have been substituted by Richardson in his series of synthetic cements. Titanic oxide (TiO2), stannic oxide (SnO2), and plumbic oxide (PbO) have been so used. "The ground clinker in each case has been found to set rapidly, although the resulting test pieces were not volume constant, the temperature obtainable in our furnace being evidently insufficient to bring about a thorough combination between these oxides and lime. Cements have been made in which phosphoric acid (P20s) has been substituted for silica." Replacement of alumina by iron oxide. Some difference of opinion appears concerning the extent to which the alumina of a Portland ce- ment may be replaced by iron oxide. This problem was taken up by the Newberrys in the classic researches before cited. They prepared mixtures of pure iron oxide and calcium carbonate in such proportions as to correspond to the formula 2CaO.- Fe 2 03, which in percentages is equivalent to CaO 41.3 per cent, Fe20s 58.7 per cent. "On burning, the material fused to a black slag, which yielded a brown color on grinding. Mixed with water to a paste, this powder showed no heating, and did not set or harden in air or cold water. A part placed in steam however, after setting one day in air, hardened rapidly, and after several hours in boiling water showed no cracking and appeared very hard. From this experiment it appears that lime and iron oxide readily combine, yielding a product which is constant in volume, though it shows no hardening properties in the cold." The 574 CEMENTS, LIMES, AND PLASTERS. Newberrys cany the experiments further, making a silica-, iron oxide, lime mixture entirely free from alumina. This was made to correspond to the formula (3CaO.Si0 2 ) + X (2CaO.Fe 2 3 ), and contained about 7 per .cent of Jron oxide. On burning this gave a black, fusible clinker. When powdered this was dark-gray, and gave a slow-setting hard and sound cement. Their final conclusions were, that though "iron oxide evidently combines with lime in the same manner as alumina," the amount of iron oxide present in ordinary clays is so small that "it is quite unneces- sary, in working with ordinary clays, to take the iron oxide into con- sideration in calculating the amount of lime required." In view of the manufacture of cements containing appreciable per- centages of iron oxide, it seems advisable to take this constituent into consideration in proportioning mixes, and this has accordingly been done in the formula given earlier in this volume. Replacement of lime by magnesia. The possibility of this replace- ment has been flatly denied by some of our leading authorities on cement chemistry, while it has been maintained, but less confidently, by others. To the present writer it seems certain that magnesia is absolutely in- terchangeable with lime,- due regard being paid to their differences in atomic weight. It is only necessary to adduce the example of the high-burned natural cements, such as the Akron, to make it clear that a cement containing 15 to 20 per cent of magnesia can be made at almost clinkering temperature. Recent experiments by Newberry seem to con- firm this conclusion. It is to be noted, however, that a Portland cement carrying high percentages of magnesia will necessarily differ considera- bly from our present-day lime Portlands. It is even probable that the differences in physical and technical properties will be so great that it will be necessary to market such magnesia Portlands under some dis- tinct trade-name. Replacement of lime by other bases. Magnesia is not the only base that can replace, either partly or entirely, the lime of a normal Portland-cement clinker. Other alkaline earths can be so substituted, as was proven in the course of Richardson's recent experiments. He describes * this phase of his work as follows: "Clinkers have been made in which baryta (BaO) and strontia (SrO) are the bases. They must be burned at a very much higher temperature than similar clinkers containing lime. In powder these * Engineering News, vol. 53, p. 85. Jan. 26, 1905. CONSTITUTION, ETC., OF PORTLAND CEMENT. 575 (barium and strontium cements) possess strong hydraulic properties, and are volume constant in water for a few days, but owing to the greater solubility in water of barium and strontium hydrate than of lime hydrate, the material after setting is much more readily attacked by water than is lime cement, strontium hydrate being about twice as soluble as calcium hydrate, and barium hydrate about eight times as soluble." References on the constitution of Portland cement. The following brief list will serve as an introduction to the mass of literature on this subject: Bonnami, H. Fabrication et contiole des chaux hydrauliques et des ciments. 8vo, 276 pp. Paris, 1888. Le Chatelier, H. Tests of hydraulic materials. Trans. Amer. Inst. Mining Engineers, vol. 22, pp. 3-52. 1894. Newberry, S. B. and W. B. The constitution of hydraulic cements. Journ. Soc. Chem Industry, vol. 16, pp. 887-894. 1897 Richardson, C. The constitution of Portland cement. Cement, vols. 3, 4, 5. 1903-1905 Richardson, C. The constitution of Portland cement from a physico-chemical standpoint. 12mo, 20 pp. Long Island City, N. Y., 1904. Richardson, C. The setting or hydration of Portland cement. Engineering News, vol. 53, pp. 84-85. Jan. 26, 1905. Composition of Portland cements. The chemical composition of Port- land cements has been changing slowly in one direction since 1850. This is well brought out by the analyses of old Portland cements given in the following table, when compared with the analyses of modern Port- lands given in Table 221. TABLE 220. ANALYSES OF PORTLAND CEMENT, 1849-1873. 1. 2. 3. 4. Silica (SiO ) 18 60 22 23 23 72 18 60 11.30 7 75 7 36 4 75 Iron oxide (Fe 2 O 3 ) 17.90 5 30 5 05 5 60 Lime (CaO) 49.80 54 11 54 40 58 50 0.70 0.75 86 2 55 Alkalies (K 2 O,Na 2 O) n. d. 1.76 2.62 1 70 Sulphur trioxide (SO 3 ) n. d. 1.00 1 12 2 10 Carbon dioxide (CO 2 ) n d 2 15 2 80 50 Water n d 1 00 96 50 1. Manufactured about 1873 by I. C. Johnson & Co., England. Reports Vienna Exposition, vol. 4, pt. D, p. 35. 2. Manufactured about 1849 in England. Analyzed by Pettenkofer. Proc. Institution Civil Engineers, vol. 62, p. 77. 3. Manufactured about 1873 in England. Analyzed by Feichtinger. Reports Vienna Exposition, vol. 4, pt. D, p. 37. 4. Manufactured about 1873 in Austria. Analyzed by Wagner. Reports Vienna Exposition, vol. 4, pt. D, p. 37. 576 CEMENTS, LIMES, AND PLASTERS. From inspection of the above table it will be seen that old Portlands were very low-limed products. Some, in fact, were too low in lime to be considered, at the present day, as falling in the Portland class. Composition of American Portland cements. Table 221, contain ing a large series of analyses of American Portland cements, has been compiled by the writer from various, sources. About half of the analyses contained in it . have already been published in different books and periodicals, while for the remainder the writer is indebted to the chemists of the various plants. Standard Methods of Analysis. The following methods of analysis are those suggested by a com- mittee of the New York section, Society of Chemical Industry, consist- ing of W. F. Hillebrand and Clifford Richardson. For exact work it is desirable that these methods be closely followed. They are not intended for use in making the rapid determinations which are necessary for the control of the mix when the plant is in operation. A method of rapid analysis has recently been published by several members of the Lehigh section, American Chemical Society, which is probably well adapted for use in the Lehigh cement district; but it is doubtful if it is worth while attempting to formulate standard methods for rapid analysis, since the requirements vary so much at the different plants. Method Suggested for the Analysis of Limestones, Raw Mixtures and Portland Cements.* Solution. One-half gram of the finely powdered substance is to be weighed out and, if a limestone or unburned mixture, strongly ignited in a covered platinum crucible over a strong blast for fifteen minutes, or longer if the blast is not powerful enough to affect complete con- version to a cement in this time. It is then transferred to an evapora- ting dish, preferably of platinum for the sake of celerity in evapora- tion, moistened with enough water to prevent lumping, and 5 to 10 c.c. of strong HC1 added and digested, with the aid of gentle heat and agita- tion until solution is completed. Solution may be aided by light pres- sure with the flattened end of a glass rod. f The solution is then evapo- rated to dryness, as far as this may be possible on the steam-bath. * Eng. News, 50, p. 60. Eng. Record, 48, p. 49. Cement, Sept., 1903. t If anything remains undecomposed it should be separated, fused with a little Na 2 CO 3 , dissolved and added to the original solution. Of course, a small amount of the separated non-gelatinous silica is not to be mistaken for undecom- posed matter. CONSTITUTION, ETC , OF PORTLAND CEMENT. 577 TH coi-i IO CO CC O OOOO^ (M O5 O r ~ l S<~S CO T^O i I CQ CO GO CM iC^OcOt^- W 1>O l> OO 00 1C 1> l> Tf CO l>l>OOOi 1> 00 X -V C !-= O 3 ^' c "o .3 ; c aj 3 : ! ISS?: 8,181'afti . 6S 578 CEMENTS, LIMES, AND PLASTERS. tO O CM CM _J CO GO O CO rfi CO to to to O5 "* OC O5 i i O O O CO t> OOT-* CM CO T-H CM " ^ !> i~H to to OS OS t^* 1-"^ CM t^ 00 CO ^ 00 to to ^ TH CO ^O ^^ to T (i li (i (CCMr IT li I i ( i i i I TH OOOi li 'i IT lOrHrH TH rHi-Hi I ^; C 1 ^^ ^^ ^^ rv %** ^ * '"^ Cs CT-I CT^ C^l W T 113 o o ~* ._. \.-^ \.-*/ \^ i. ~ \i> **^ %i- ^^* v_ i "^ \j- \^^/ r CO to O O !^ CM i^ CO CO r-H O5 CO CO -^ CM i i ^ to Tf 00 CM CO X O i ' GO CM CO CO ^ O5 00 O * i i 1> CO tO CO -^ CMCMi-HrHrH CCOi-ii-ii-i CMi-MCMi-iCMOrHO CJCOCMCMi-lOCOTHCJ CM CM CM t^ CO Oi CO TH CO O *C O CM OOO Oi 1-1 X to o: - CX) co co OCO 1> CO iO GO C5 >O O CO GO CO i-> CO GO CO GO i-H CO 00 O O5O5 CO^rH COCMOO5O5 tO 00 tO O CO CO "* CM T CM rH J> O O O5 CM CO CO O rH 00 tO !>. l> CO CO CO O CO 00 O5 t>. COCO t^ CO ^i GO GO -^f ^ O ^f 00 O O CO 00 O5 CO O tO O i i -* CO O l>- CO O5 CM 00 to O "f i i O5 1-1 O ^ C5Tt< 00 OO5 to COOO Oi-i CO(N t>00 CO O O5 COOOiN Oi O5 O-* Oi O CO (N O (M CO O O t^ ^ ^ Tf Tji 1C O CO O OJ Tfi t>- 00 CO 00 00 t> CO *O CO ^ ^ - c^ i> co Oi rr os 580 CEMENTS, LIMES, AND PLASTERS. Silica. The residue, without further heating, is treated at first with 5 to 10 c.c. of strong HC1. which is then diluted to half strength or less, or upon the residue may be poured at once a larger volume of acid of half strength. The dish is then covered and digestion allowed. to go on for ten minutes on the bath, after which the solution is filtered and the separatecj silica washed -thoroughly with water. The filtrate is again evaporated to dryness, the residue, without further heating, taken up with acid and water, and the small amount of silica it con- tains separated on another filter-paper. The papers containing the residue are transferred wet to a weighed platinum crucible, dried, ignited, first over a Bunsen burner until the carbon of the filter is completely consumed, arid finally over the blast for fifteen minutes and checked by a further blasting for ten minutes or to constant weight. The silica, if great accuracy is desired, is treated in the crucible with about 10 c.c. of HF1 and four drops of H 2 SO4 and evaporated over a low flame to complete dryness. The small residue is finally blasted for a minute or two, cooled, and weighed. The difference between this weight and the weight previously obtained gives the amount of silica.* A1 2 3 and Fe 2 O 3 : The filtrate, about 250 c.c. from the second evaporation for Si0 2 , is made alkaline with NH 4 OH after adding HC1, if need be, to insure a total of 10 to 15 c.c. strong acid, and boiled to expel excess of NH 3 , or until there is but a faint odor of it, and the pre- cipitated iron and aluminum hydrates, after settling, are washed once by decantation and slightly on the filter. Setting aside the filtrate, the precipitate is dissolved in hot dilute HC1, the solution passing into the beaker in which the precipitation was made. The aluminum and iron are then precipitated by NH 4 OH, boiled and the second precipi- tate collected and washed on the same filter used in the first instant. The filter-paper, with the precipitate, is then placed in a weighed platinum crucible, the paper burned off and the precipitate ignited and finally blasted 5 minutes, with care to prevent reduction, cooled and weighed as Al 2 3 + Fe 2 03.t Fe 2 3 : The combined iron and aluminum oxides are fused in a platinum crucible at a very low temperature with about 3 to 4 grams of KHS04, or, better, NaHSO4, the melt taken up with so much dilute H 2 S04 that there shall be no less than 5 grams absolute acid and enough water to effect solution on heating. The solution is then evaporated and eventually heated till acid fumes come off copiously. After cooling * For ordinary control jvork in the plant laboratory this correction may, perhaps, be neglected; the double evaporation never. f This precipitate contains TiO 2 , P 2 O 5 , Mn 3 O 4 . CONSTITUTION, ETC., OF PORTLAND CEMENT. 581 and redissolving in water the small amount of silica is filtered out, weighed, and corrected by HF1 and H 2 SO4.* The filtrate is reduced by zinc- or preferably by bydrogen sulphide, boiling out the excess of .the latter afterward while passing C0 2 through the flask, and kitrated with permanganate. f The strength of the permanganate*, solution should not be greater than .0040 grains Fe^Oa per c.c. ; .\, j CaO: To the combined filtrate from |Jie Al 2 03rjr f .IfeQ3* precipitate a few drops of NH4OH are added, and the solution brougj^t to boiling. To the boiling solution 20 c.c. of a saturated solution of ammonium oxalate is added, and the boiling continued until the precipitated CaC 2 C>4 assumes a well-defined granular /orm. It is 'then allowed to stand for 20 minutes, or until the precipitate has settled, and then filtered and washed. The precipitate and filter are placed wet in a platinum crucible, and the paper burned off over a small flame of a Bunsen burner. It is then ignited, redissolved in HC1, and the solution made up to 100 c.c. with water. Ammonia is added in slight excess, and the liquid is boiled. If a small amount of A1 2 Q3 separates, this is filtered out, weighed, and the amount added to that found in the first determina- tion, when greater accuracy is desired. The lime is then reprecipitated by ammonium oxalate, allowed to stand until settled, filtered, and washed, t weighed as oxide by ignition and blasting in a covered crucible to constant weight, or determined with dilute standard per- manganate. AlgO: The combined filtrates from the calcium precipitates are acidified with HC1, and concentrated on the steam-bath to about 150 c.c., 10 c.c. of saturated solution of N"a(NH4)HP04 are added, and the solution boiled for several minutes. It is then removed from the flame and cooled by placing the beaker in ice- water. -After cooling, NH 4 OH is added drop by drop with constant stirring until the crystal- line ummonium magnesium orthophosphate begins, to form, and then in moderate excess, the stirring being continued for several minutes. It is then set aside for several hours in a cool atmosphere and filtered. * This correction of A1 2 O 3 , Fe 2 O 3 , for silica should not be made when the HF1 correction of the main silica has been omitted, unless that silica was obtained by only one evaporation and nitration. After two evaporations and nitrations 1 to 2 mg. of SiO 2 are still to be found with the Al 2 O 3 .Fe 2 O 3 . t In this way only is the influence of titanium to be avoided and a correct result obtained for iron. J The volume of wash-water should not be too large, vide Hillebrarid. The accuracy of this method admits of criticism, but its convenience and rapidity demand its insertion. 582 CEMENTS, LIMES, AND PLASTERS. The . precipitate is redissolved in hot dilute HC1, the solution made up to about 100 c.c. ; 1 c.c. of a saturated solution of Na(NH4)HP04 added, and ammonia drop by drop, with constant stirring until the precipitate is again formed as described and the ammonia is in moderate excess. It is then allowed to stand for about 2 hours when it is filtered on a paper or a Gooch crucible, ignited, cooled, and weighed as Mg 2 P 2 O7. K 2 O and Na2O: For the determination of the alkalies, the well- known method of Prof. J. Lawrence Smith is to be followed, either with or without the addition of CaCOs with NHCU. 80s: One gram of the substance is dissolved in 15 c.c. of HC1, filtered and residue washed thoroughly.* The solution is made up to 250 c.c. in a beaker and boiled. To the boiling solution 10 c.c. of a saturated solution of BaCL 2 is added slowly, drop by drop, from a pipette and the boiling continued until the pre- cipitate is well formed, or digestion on the steam-bath may be sub- stituted for the boiling. It is then set aside overnight, or for a few hours, filtered, ignited, and weighed as BaS04. Total sulphur. One gram of the material is weighed out in a large platinum crucible and fused with Na 2 C0 3 and a little KN0 3 , being care- ful to avoid contamination from sulphur in the gases from source of heat. This may be done by fitting the crucible in a hole in an abestos board. The melt is treated in the crucible with boiling water and the liquid poured into a tall, narrow beaker, and more hot water added until the mass is disintegrated. The solution is then filtered. The filtrate contained in a No. 4 beaker is to be acidulated with HC1 and made up to 250 c.c. with distilled water, boiled, the sulphur precipitated as BaS0 4 and allowed to stand overnight or for a few hours. Loss on ignition. Half a gram of cement is to be weighed out in a platinum crucible, placed in a hole in an asbestos board so that about three fifths of the crucible projects below, and blasted 15 minutes, prefer- ably with an inclined flame. The loss by weight, which is checked by a .second blasting of 5 minutes, is the loss on ignition. Note. Recent investigations have shown that large errors in results are often due to the use of impure distilled water and reagents. The analyst should, therefore, test his distilled water by evaporation and his reagents by appropriate tests before proceeding with his work. * Evaporation to dry ness is unnecessary, unless gelatinous silica should have separated and should never be performed on a bath heated by gas, vide Hillebrand. CHAPTER XXXIX. PHYSICAL PROPERTIES: TESTING METHODS. THE utilization of Portland cement does not properly come within the province of this volume, as it is already covered by several excellent books. An extensive and readily accessible literature has been created on the subject of testing methods and testing results; but most of this literature is more important to the professional cement-tester than to the cement-manufacturer or cement-user. In the present chapter the subject of testing will necessarily be considered, but merely incidentally. Stress will be laid, on the other hand, on the general properties which Portland cement develops in use, and attention will be directed to the chemical and physical agencies which operate to disintegrate, or weaken, or destroy the cement, or the structures hi which it is used. Physical Properties of Portland Cement. Portland cement 13 at present used for many different purposes, and the use to which it is applied seems to be rapidly increasing. Under such circumstances it is necessary to supply a product well-fitted to withstand the various disintegrating agencies to which it may be sub- jected. In its ordinary uses, hi heavy masonry for example, the cement will be subjected to . compressive stresses, but rarely to tensile. When used as a paving material it will encounter transverse stresses and severe abrasion. As a lining material its imperviousness will be tested. In other places, as in gun emplacements for example, it may be sub- jected to severe and often-repeated shocks. To these physical agencies of disintegration or destruction, are added chemical agents, which are at times of paramount importance. Works exposed to sea-water, for example, are subject to purely chemical attack which must be guarded against so far as possible. The situation might be summed up by stating that cement may fail through defects in its manufacture (internal agencies), or through 583 584 CEMENTS, LIMES, AND PLASTERS. the purely external agencies, and that these agencies may be either physical or chemical. This brief outline will serve to give some idea of the wide scope which might be given to a discussion of the properties of Portland cement. , ALL PASQED NO. lOC^SIEVE . o o ' o o _ ' > / 87$ P ^SSED MO. 100 SIEVE -S / / /* p ASSED TO^A-rfD STOPP ED ON 100 SIE VE / / ?> SSED 5 AND STOPPl D ON ' SIEV FIG. 143.* Variation of tensile strength with fineness. Value of fineness tests. The reason for testing the fineness of a cement depends on the facts that (a) the strength of the cement, and particularly its tensile strength when mixed with sand, increases with the fineness, and (6) the soundness of the cement may be improved by fine grinding. The second point is one that concerns the manu- facturer more than the user, because an unsound cement will usually fail to pass other tests and will therefore be rejected. The increase in strength consequent on increased fineness is well shown in Figs. 143 and 144, both showing the results of tests on 1:3 mixtures, the tests of Fig. 143 having been made at four months while those in Fig. 144 are at various ages. The value of fine grinding is evident, and engineers are constantly raising the standard of fineness in specifications. Unfortunately, how- ever, they fail to make proper use of this fine cement after they have paid extra for. getting it. They insist, for example, in obtaining cement which will pass 92 or 95 per cent through a 100-mesh sieve, and then use it in the same sand mixtures that they would if it were an English cement passing perhaps 85 per cent through 100-mesh. The actual fineness of a number of typical American Portlands is shown very exactly in the tests given in Table 222. * From Johnson's " Materials of Construction", p. 409. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 585 TABLE 222. FINENESS OF VARIOUS AMERICAN PORTLANDS. (BLEININGER.) Diam- Diam- Diam- Residue Residue on Residue on eter between eter between eter between Finer Total Coarser Brand. Reduced on. 80-mesh Sieve. 120- mesh Sieve. 200- mesh Sieve. 0.008 and 0.002 0.002 and 0.0002 0.0003 and 0.0007 Last Size. than 200- mesh. i Inches. Inches. Inches. 1 Tube miU. . 7.07 14.56 4.45 22. CS 19.29 7.36 24.69 25.98 2 ( i 9.01 15 35 5.09 21.50 20.53 7.06 21.52 29.45 3 e e 12.12 15.05 7.61 21.11 16.85 5.91 21.37 34.78 4 ( e 14.11 14 57 7.82 22.43 13.95 7.81 19.32 36.49 5 e t 3 84 17 64 5 1C 25 27 12 56 10 16 20 42 26 58 6 Gr ffin mill. . 3^06 15.41 8.24 28 . 56 16.48 12.74 15.51 26.72 7 < c 9.43 16.91 6.37 25.52 12.18 9.20 18.49 32.71 8 e t 5 00 15 42 14 52 27 30 19 10 9 22 14 88 29.53 9 i e 4.40 11.35 5.13 23 . 79 21.30 10.01 24.01 20.89 10 e ( 4.18 13.30 5.07 22.63 14.64 12.31 28.79 22.55 Specific gravity. The specific gravity of a Portland cement is a property which is of no importance of itself to the engineer. The reason for determining it is in order to rule out underburned or adul- 400 V 3000 2000 1000 COMPRESSION 10 15 20 25 5 AGE IN DAYS 10 15 20 25 30 FIG. 144.* ^Effect on strength of regrinding cement. (Tetmajer.) terated cement. The specific gravity of a well-dried sample of Port- land cement will rarely fall below 3.10; while that of a natural cement, a slag cement, or a Portland adulterated with slag w r ill rarely rise above * From Johnson's " Materials of Construction ", p. 411. 586 CEMENTS, LIMES, AND PLASTERS. 3.00. Some few American natural cements do, however, show a higher specific gravity, as can be seen from the table on page 263. Setting properties. A certain minimum time of initial and final set is usually specified, for the convenience of the workmen. This is regu- lated by the use of gypsum or plaster at the plant, a practice whose effects have been discussed -in detail jji Chapter XXXVI. CURVE OF TIME BEFORE SETTING IS COMPLETED. u .8 IN MIN zm: 30 S5 40 45 50 C. \ CURVE OF TIME BEFORE SETTING BEGINS. co|\ 1.5 IN MIN 6 10 15 20 25 30 35 40 45 50 C. FIG. 145.* Effect of temperature on setting time. The effect of temperature on the setting of Portland cement is well shown in Fig. 145. It will be noted that the setting is much slower at low than at high temperatures, within the limits of the experiments. Tensile strength. The tensile strength of a cement is of very little importance or interest of itself, because cements are rarely subjected intentionally to tensile strains. But in practice the tensile test is the * From Johnson's "Materials of Construction," p, 616. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 587 Age of briquettes. Fid 146. Tensile strength of various classes of cements. (Philadelphia tests, 1899.) 588 CEMENTS, LIMES, AND PLASTERS. most commonly applied of all tests, this action being based on the assumption that the ratio between compressive and tensile strength for all Portland cements is quite uniform, and that therefore varia- tions in tensile strength will indicate corresponding (though much greater) variations in compressive strength. This assumption is to a large extent correct, and for all practical purposes may be considered satisfactory. The question as to the ratio existing between the two types of strength will be taken up on a later page (p. 589.) In Fig. 146 the results of a large series of tests on various classes of cement are shown diagrammatically. The cements tested included American and foreign Portlands, foreign "natural Portlands", and Amer- ican natural cements, and the comparative results are quite represen- tative. 600 r 400 30 AGE IN WEEKS FIG. 147.* Effect of proportions of sand on tensile strength. The three points of most general interest in connection with tests of tensile strength are (a) the decrease in tensile strength with increase of percentage of sand, (6) the increase in strength with increased age, and (c) the variation in strength due to differences in the character of the sand. Two of these points are illustrated in Figs. 147 and 148, while all three are constantly discussed in engineering publications. Compressive strength. The compressive strength of a cement or concrete is a matter of direct practical importance, for these materials are rarely subjected to any other type of strain when used in actual * From Johnson's " Materials of Construction", p. 571. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 580 work. Compressive tests, however, require the use of heavy testing machines, and are therefore not adapted for field or ordinary office tests. (See Tables 223 and 224.) 800 i 2400 6200 A A 2468 PROPORTIONS OF SAND TO 1 CEMENT 10 FIG. 148.* Effect of character of sand on tensile strength. TABLE 223. COMPRESSIVE STRENGTH OF PORTLAND-CEMENT CUBES, WATERTOWN ARSENAL. Brand. Per Cent Water. Compressive Strength, Pounds per Square Inch. 7 Days. 1 Month. 3 Months. 8580 5870 6310 6980 8170 8180 7720 5930 7730 6810 7630 5510 4660 Alpha 25 25 26.8 18 22.5 25 30 22.5 25 30 25 29.2 26.7 6010 3490 4280 5780 : 5960 6320 6340 4620 5560 5030 5630 3510 2750 7340 5370 5590 5990 7080 6750 6850 5180 5980 5620 6640 4940 4030 Atlas Lehigh n Star with plaster ( ' (C It I (( ( C 1 without plaster ... . . ... ( a (i I 11 (C Whitehall Alsen Josson . . . . t Report on Tests of Metals, etc., at Watertown Arsenal, 1902, pp. 369-376. Ratio of compressive to tensile strength. For a given age and mixture, the ratio between the compressive and tensile strength of a Portland-cement mortar is practically fixed. The ratio increases with inceasing age, and also increases with increasing proportions of sand. * From Johnson's " Materials of Construction", p 581. 590 CEMENTS, LIMES, AND PLASTERS. In Fig. 149 are plotted the curves, by Johnson, resulting from compari- son of a large series of tests by Tetmajer on 1:3 mixtures. TABLE 224. COMPRESSIVE STRENGTH OP PORTLAND-CEMENT MORTAR AND CONCRETE CUBES, WATERTOWN ARSENAL. Brand. Composition. v* Age. Size of Cube, Inches. Compres- sive Strength per Square Inch. Cement. Sand. Stone. Years. Months. Days. Atlas *. . 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1J 2 2* 3 3* 4 2 3 2 2 2 3 4 6 4 4 4 5 2 2 2 2 2 2 2 3 5 5 5 5 5 5 5 0* 1 2 3 20 19 16 17 15 13 12 7 7 12 22 25 2 6 6 6 6 6 6 6 12 12 12 * 12 12 12 11,330 10,390 9,520 8,110 6,140 6,280 5,230 1,303 1,053 2,615 3,392 4,135 3,758 * * * * * * * * Alpha * Vulcanite f . . t.- Grant f * Report on Tests of Metals, etc., at Watertown Arsenal, 1902, pp. 512-514. t Ibid., 1900, pp. 1105-1111. TENSILE STRENGTH ^ 00 o o - .^- ^ ^^ ^\ c^ ~^"~ f^ ^ ^"' V? ^ / / * / $ ' /i 7 '// I . i * J 2 4 6 8 10 13 AGE IN MONTHS FIG. 149. t Ratio of compressive to tensile strength. (Johnson.) J From Johnson's " Materials of Construction", p. 419. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 591 In practical use, it may be assumed that at the end of a year the average Portland-cement mortar will have a compressive strength about ten times as great as its tensile strength. In Table 225 are given the results of a series of tests carried out at the Watertown Arsenal.* The tensile tests were made on the usual briquettes, the compressive tests on 2-inch cubes, and each average given is the result of ten tests. The cement used was the Peninsular brand, giving the following results for fineness and specific gravity. Per cent of fineness: Retained on 98 X ICO sieve 4 . 95 Passed by 98X100 sieve; retained on 174X182 bolting-cloth 19.75 Passed by 174X 182 bolting-cloth 75.30 Specific gravity: As taken from barrel 3 . 20 After mixing with 22 per cent water, setting 7 days in air, regrinding, and heating to a constant weight at 110 C . . 2.81 A chemical analysis of the cement is also given, but is evidently erroneous and therefore will not be quoted here. TABLE 225. RELATION OF TENSILE TO COMPRESSIVE STRENGTH. (WATERTOWN ARSENAL.) Per Cent _f Ages in Tensile Strength, Pounds per Square Inch. Compressive Strength, Pounds per Square Inch. Ol Water. Air, Days. Water, Days. Maximum. Minimum. Average. Maximum. Minimum. Average. 20 1 221 177 196 801 654 717 20 7 393 301 354 3430 2700 3040 20 28 641 487 566 4370 3490 3990 20 1 6 835 653 780 4830 3770 4250 20 1 27 952 857 906 8280 5730 7370 22 1 209 156 189 670 530 595 22 7 . . 482 303 392 3680 3010 3260 22 28 518 421 457 4310 3030 3760 22 1 6 724 502 666 5370 3620 4720 22 1 27 1010 782 866 7810 5360 6870 25 1 . . 223 148 190 450 398 430 25 7 475 301 402 3210 2120 2610 25 28 552 393 450 3550 2630 3130 25 1 6 388 251 329 4440 3360 3880 25 1 27 807 696 758 8740 6310 7580 Modulus of elasticity. The determinations of the modulus of elas- ticity given in the following table were made at the Watertown Arsenal. * Report on Tests of Metals, etc., at Watertown Arsenal during 1902, p. 511. 1903. 592 CEMENTS, LIMES, AND PLASTERS. TABLE 226. Brand. Composition. Weight per Cubic Foot, Pounds. Age. Mos. Dys. Ultimate Strength, Pounds per Square Inch. Pounds per " Square Inch. Alpha Neat 135.5 f* 1 ' / #(500-2000) =3,000,000 \ "(500-3000) =3,030,000 ii ft [(500-2000) =3,488,000 135.5 7 %if 8530 1 (2000-4000) =3,279,000 [ (4000-6000) =2,963,000 it it 137.3 1 ;..:-.. (500-2000) =3,061,000 f (500-2000) =4,545,000 i ( It 137 . 3 -5 V... 9260 \ (2000-4000) =4,255,000 .. . .. [ (4000-6000) =3,846,000 Atlas ({ 134.7 2J . . (500-2000) =2,326,000 it ft 134.7 64 5450 / (500-2000) =2,479,000 \j 2, \ (2000-4000) =2,581,000 Lehigh. . . . It 129.2 1 26 5800 J (500-2000) =2,500,000 \ (2000-4000) =2,353,000 ( ( 1 cement : 1 sand 133 1 26 3420 (500-2000) =2,778,000 I C Neat 135.9 1 26 7540 / 7? (500-2000) =4,348,000 \ (2000-3000) =4,444,000 f (500-2000) =3,571,000 Peninsular . it 135.3 2 13 6710 (2000-3000) =3,448,000 - .-* .. I (3000-4000) =3,125,000 f (500-2000) = 3,846,000 it { ( 138.0 2 14 6720 \ (2000-3000) =3,571, 000 I (3000-4000) =3,509,000 (i 1 cement : 1 sand 133.4 2 13 4200 / (500-2000) =2,941,000 \ (2000-3000) =2,439,000 Sand cement. Sand cement, or silica cement, is the name given to the product made by grinding up together Portland cement with an equal or greater quantity of sand, limestone, or other chemically inert substance. Description of the making and properties of sand cement is not properly part of a discussion of the manufacture of Portland cement, but rather a matter for the engineer to consider in connection with the uses of cement. For this reason the question will be touched on very briefly. It is found that if Portland cement be mixed with an equal quantity of sand or limestone and the mixture ground very finely in a tube mill, the resulting product (sand cement) will show a strength almost or quite as great as the Portland cement from which it was made, not- withstanding the fact that the sand cement consists only half of Port- land cement. When Portland cement is very expensive, economies are, therefore, possible in this line. The gain in strength is due entirely to the extra fineness given by the extra grinding. The sand does not enter into chemical combina- PHYSICAL PROPERTIES OF PORTLAND CEMENT. 593 tion with the cement in any way, for ground limestone will give as good results as ground quartz. 150 200 AGE IN DAYS 300 350 FIG. 150.* Strength of sand-cement mortar. The first figure on each curve de- notes parts of Portland cement; the second, parts of ground sand; the third, parts of uiiground sand. The tests quoted in Table 227 were made at Albany in the labora- tory of the State Engineer. Iron Clad is a Portland cement of high grade, while Victor is the sand cement made from it by grinding Iron Clad with limestone. * From Johnson's "Materials of Construction", p. 579. 594 CEMENTS, LIMES, AND PLASTERS. TABLE 227. COMPARATIVE TESTS OF PORTLAND CEMENT AND SAND CEMENT. Cement. Fineness. Setting-time, Minutes. Tensile Strength, Pounds. 50-Mesh. lotf-Mesh Initial. Final. 7 Days. 28 Days. 1897. Iron Clad, 1:3 100 100 94f 96 51 35 122 79 170 198 274 265 Victor, 1:3 1898. Iron Clad, 1 : 3 99 100 94i 96 27 41 81 89 189 184 277 272 Victor, 1:3 1899. Iron Clad, 1:3 Victor, 1:3 100 100 98 100 45 60 94 158 207 178 311 264 TABLE 228. TENSILE AND COMPRESSIVE STRENGTH OF SAND CEMENTS. (SMITH.) Name of brand Citadel Ensign Jubilee Sand cement composed of 1 cement 1 cement 1 cement 1 sand 1 sand 6 sand Fineness: Passing 100-mesh. . , 99.8 99.4 99 7 " 120- " 99 3 98 4 " 180- " . . 99 3 Neat sand cement: Tension, 1 week 332 ' ' 4 weeks 475 1 ' 4 months 810 340 " 6 " 780 540 Compression, 4 weeks 3837 Sand cement 1, sand 1 : Tension, 1 week . . . 300 ' ' 2 weeks 379 Compression, 1 week. . . 2800 Sand cement 1 , sand 2 Tension 1 week 184 ' ' 2 weeks 215 Compression, 1 week . . . 1225 Sand cement 1, sand 3: Tension, 1 week .... 135 189 ' ' 2 weeks 201 " 4 " 141 ' ' 2 months Compression, 1 week . . . 4 weeks. . . 135 470 687 900 Brickbuilder, vol. 6, p. 281. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 595 2600 2000 cc. ui o. 1500 z 1 z Q 1000 i 500 \ TENSION \ \ PERCENTAGE TOTAL8AND AND CEMENT FIG. 151.* Strength of sand-cement mortar with varying proportions of sand. 500 400 300 200 100 f IG. 152. f Variation in strength of sand-cement mortar when the total proportion of sand is constant, but the relative proportions of ground and unground sand are variable. * From Johnson's " Materials of Construction ", p. 580. t Ibid., p. 581 596 CEMENTS, LIMES, AND PLASTERS. TABLE 229. COMPRESSIVE STRENGTH OF SILICA-CEMENT CUBES. (WATERTOWN ARSENAL.) \ Dimensions of Cube. Compressive Strength. Per Cent Water. Age. Height, . -i*- Surface, Com- pressed Area Total Per Square Average Pounds ' Inches. Inches. Square Inches. Pounds. Inch, Pounds. per Square Inch. 28* 7 days 4.05 4.00X4.06 16.24 19,950 1228 1 28| i i 4.00 4.09X4.08 16.68 22,600 1355 28* ( ( 4.03 3.98X4.13 16.44 20,900 1271 1 1300 28| ( t 3.96 4.03X4.07 16.40 19,980 1218 28*. i { 3.96 4.06X4.10 16.64 23,500 1412 j 28| 1 month 3.96 4.15X4.04 16.77 27,100 1616 28* ' l 3.97 4.02X4.18 16.80 30,600 1821 1 28* 1 1 3.96 4.04X4.07 16.44 33,500 2038 \ 1790 28* t ( 3.98 4.04X4.06 16.40 27,900 1701 28| i ( 4.00 4.13X4.00 16 . 69 29,500 1768 j 28* 3 months 3.97 4.01X4.10 16.44 32,800 1995 j 28| 1 1 3.99 4.05X4.05 16 V 40 34,100 2079 28*. 1 1 3.96 4.08X4.09 16.69 34,500 2067 j- 2110 28*. ( ( 3.98 4.01X4.18 16.76 29,200 2339 28^ e t 3.96 4.05X4.10 16.61 34,500 2077 j 28^ 12 months 3.96 4.03X4.19 16.89 33,600 1990 28| < < 3.98 4.07X4.10 16.69 39,900 2390 > 2190 18 8 days 4.00 3.92X4.13 16.19 47,100 2910 1" 18 i ( 4.06 3.99X4.00 15.96 53,100 3330 18 tt 4.08 4.00X3.95 15.80 53,400 3380 3050 18 t ( 4.02 4.06X3.86 15.67 43,600 2780 18 '< 3.99 3.98X4.08 16.24 46,500 2860 18 1 month 4.00 4.03X3.98 16.04 54,900 3420 18 i e 4:00 4.02X4.08 16.40 60,600 3700 1 18 1 1 4.08 4.01X4.00 16.04 65,800 4100 \ 3470 18 t ( 4.05 4.01X3.98 15.96 39,500 2480 18 i ( 4.07 4.00X4.06 16.24 59,400 3660 j 18 3 months 4.08 3.98X4.05 16.12 76,100 4720 J 18 1 1 4.00 4.09X4.03 16.48 70,500 4280 18 t ( 4.08 4.00X4.02 16.08 73,600 4580 j. 4470 18 ( t 3.98 4.05X4.06 16.44 70,600 4290 \ Report of Tests of Metals, etc., at Watertown Arsenal for 1902, pp. 376-377. 1903. List of references on sand cement. The following papers are of interest in this connection: Butler, M. J. Silica Portland cement. Canadian Engineer, March, 1899. Klein, O. H. Report on concrete foundations for pavements. 8vo, 58 pp. New York, 1903. (Much criticism of sand cements.) Reeves, H. E. The effect of grinding mixed sand and cement. Technograph, May, 1896. Smith, C. B. Sand cement. Brickbuilder, vol. 6, p. 280. 1897. (Tests of three Canadian brands. Important paper.) PHYSICAL PROPERTIES OF PORTLAND CEMENT. 597 TABLE 230. COMPRESSIVE STRENGTH OF SAND-CEMENT MORTARS. (WATERTOWN ARSENAL.) Brand. Composition. Per Cent Age. Average Strength, Pounds per Square Inch. Months. Days. In Air. In Water. Silica Neat li 1 cement, 1 sand t ( ii t f 1 1 1 cement, 2 sand it t ( t { 1 1 1 cement, 3 sand t i i i a v 3 Y 3 1 3 1 3 7 6. 5 4* 1670 2070 2420 942 1460 1610 386 424 850 130 219 306 1880 2830 3110 1090 1920 2340 424 708 1120 : 132 360 571 . Report of Tests of Metals, etc., at Watertown Arsenal, 1902, p. 443. TABLE 231. MODULUS OF ELASTICITY OF SAND CEMENT. (WATERTOWN ARSENAL.) Weight Brand. Mortar. per Cu. Ft. Time. Mos. Dys. Ultimate Strength. ET Pounds per Square Inch. Pounds. Silica. . . Neat 117.8 1 28 2520 f #(100-1000) =1,607,000 1 #(1000-2000) = 1,205,000 tl ( t ' 116.3 1 28 2400 f #(100-1000) = 1 ,475,000 \ #(1000-2000) = 1,117,000 ( ( I cement, 1 sand 126.9 1 29 1200 #(100-1000) =1,286,000 ( ( 1 cement, 2 sand 122.4 1 28 618 #(100-500) = 909,000 ( ( 1 cement, 3 sand 120.8 1 27 404 #(100-400) = 632,000 Report of Tests of Metals, etc., at Watertown Arsenal, 1902, pp. 498-500. References on sand cement (Continued). Anon. The manufacture and use of sand cement. Engineering News, April 16, 1896. Anon. Le Silico-Portland on silico-cement. La Revue Technologique, Jan. 25, 1898. Anon. The hydraulic experiment station of Cornell University. Engineering News, vol. 41, pp. 130-133. March 2, 1899. (Description of use of sand cement.) Effect of heating. The effect of high temperatures on cements or concretes is, in these days of fireproof construction, a matter of con- siderable interest to architects and engineers. In 1902 a series of tests along this line were carried out at Watertown Arsenal, some of which are summarized in Table 232, below. 598 CEMENTS, LIMES, AND PLASTERS. These tests were made on 2-inch cubes of neat Portland cement, all being crushed at a period of 1 year, 1 month, and 16 days after making. "The cubes for this series were prepared and set in air or in water for a period of one year to a year and a half before they were heated, and intervals ranging from four days to nearly four months intervened between the time of heating and the^time of testing. " The heated cubes were gradually raised to the temperatures recorded, and slowly cooled in dry sawdust or powdered asbestos. The time of heating was one hour, and the maximum temperature was main- tained for one hour. " Cubes which were set in water were dried off on a radiator for twenty- four hours before heating in the muffle to the temperatures recorded. " During heating some of the cubes developed fine cracks, at first faintly shown, which enlarged after a few hours or days had elapsed. In other cases the cracks appeared more promptly. Among those which were heated to the higher temperatures of the series, which ranged from 200 to 1000 F., there were cubes so badly cracked as to be unsuitable for testing." TABLE 232. EFFECT OF HEATING ON COMPRESS: VE STRENGTH. (WATERTOWN ARSENAL.) Brand. Per Cent Water. Heated to. Compressive Strength, Pounds per Square Inch. Aloha 25 not heated 9167 fT 25 200 F. 8830 ti 25 300 F. 7920 tt 25 400 F. 9190 it 25 500 F. 9400 tt 25 600 F. 9000 et 25 700 F. 8217 tt 25 800 F. 8730 tt 25 900 F. 6060 Dyckerhoff 29 not heated 5017 29 600 F 4347 1 1 29 700 F 3483 1 1 29 800 F 4280 Report of Tests of Metals, etc., at Watertown Arsenal, 1902, pp. 459-460 Effects of salt and freezing. The use of cement or concrete in build- ings constructed during very cold weather has led to a long series of experiments, designed to determine the effects of using salt and other anti-freezing agents in the water used in mixing the mortar. The results of a number of such tests are shown diagrammatically in Figs. 153 to 158 inclusive. The results as to strength are rather con- PHYSICAL PROPERTIES OF PORTLAND CEMENT, 599 tradictory, but it seems probable that any addition of salt will decrease the ultimate tensile and compressive strength of the mortar in which 20 40 60 80 X PERCENT SOLUTION FIG. 153.* Effect on the freezing-point of cement of various proportions of glycerine, alcohol, and salt. (Tetmajer.) 5 10 15 20 PERCENTAGE OF SALT FIG. 154. f Effect of salt on mortar, 1 cement: 2 sand, made in freezing weather. (Sabin.) it is used, but that for the lower percentages of salt this injurious effect may be slight enough to be safely disregarded. *From Johnson's " Materials of Construction ", p. 615. t Ibid., p. 617. 600 CEMENTS, LIMES, AND PLASTERS. GOO to 500 i 400 "6 8 10 AGE IN MONTHS FIG. 155.* Effect of salt on Portland-cement mortar, 1 cement: 2 sand, made in freezing weather. (Sabin.) AGE "IN MONTH'S FIG. 156.f Effect of salt on tensile strength of mortars, 1 cement :1 sand and 1 cement: 4 sand. Those left in air remained frozen almost sixty days. Those out in water were first frozen in air for three days. * From Johnson's ' Materials of Construction ", p. 618. f Ibid., p. 619. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 601 7 14 21 28 35 42 49 56 63 70 77 84 AGE IN DAYS TENSION FIG. 157.* Effect of salt on tensile strength of Portland-cement mortar, (Tetmajer.) 4500 2000 1500,; 14 21 28 35 42 49 56 63 70 AGE IN DAYS FIG. 158.* Effect of salt on compressive strength of Portland-cement mortar. (Tetmajer.) * From Johnson's Materials of Construction ", p. 620. flbid. 602 CEMENTS, LIMES, AND PLASTERS. Effects of exposure to sea-water. Portland cement is not entirely satisfactory in its resistance to exposure to salt water, though in part this is often due to the use of porous mixtures which permit access of the water to the interior of the block of cement or concrete. The use of richer mixtures, or at least of a richer mixture for the surface of the block, will do away with many of tlj difficulties encountered. Aside from this, two methods of improvements have been advocated. One is to make the cement more resistant of itself by making it of such a chemical composition as will show the maximum resistance to the effects of salt water. This is the method of Le Chatelier, discussed below. The second method is, to add to the cement, trass, slag, or other puzzolanic material, in order that the lime liberated by the cement during hardening may be taken up and combined with the trass. Le Chatelier considers that the aluminous compounds present in Portland cement are the direct cause of its destruction by sea-water. His theory, to account for this disintegration, is as follows : Free lime, liberated during the hardening of the cement, reacts with the mag- nesium sulphate always present in sea-water, to form calcium sulphate. This in turn reacts with the calcium aluminate of the cement to form a sulphaluminate of lime, which swells considerably on hydration and thus disintegrates the cement mass. The extent of the disintegration varies directly with the percentage of alumina present in the cement. Cements containing 1 or 2 per cent of alumina are, for example, practi- cally unaffected by sea-water; while in cement containing as high as 7 or 8 per cent of alumina the swelling and consequent disintegration are very rapid. If the alumina of a cement be replaced by an oxide not reacting with calcium sulphate, the stability of the cement in sea-water is greatly improved. Le Chatelier has demonstrated this by preparing cements in which the alumina was replaced by oxides of iron, chro- mium, cobalt, etc. All of these were more resistant than an alumina cement to the disintegrating effect of lime sulphate. The best effects were obtained when iron oxide was used, a cement corresponding in composition to 5SiO2,Fe 2 3 ,17CaO being found to be not only stable in presence of sea-water, but to possess excellent mechanical properties. DevaPs researches * on the" effect of direct addition of calcium sul- phate to various cements confirm the above theory. Each of the finely ground cements tested was completely hydrated by mixing with 50 per cent of water, and storing the mixture under water for three months out of * Abstract in Journ. Soc. Chem. Industry, vol. 21, p. 971-972. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 603 contact with carbon dioxide. The mass was then -dried, reground, mixed with half its weight of calcium sulphate and 33 per cent of water, and made up into rods, which were kept moist and protected from carbon dioxide by storage on moistened filter paper under a glass bell. At the end of three weeks the increase in length of the rods was measured, with the following results: TABLE 233. EFFECT OF ALUMINA. Type of Cement. Per Cent of Alumina in the Cement. Per Cent of Elongation of the Rods. Slag cement (Vitry) 15 5 27 Slag cement (Champignolles) 14 5 16 Grappier cement (Besses) .... 7 5 14 Portland cement 6 2 12 Hydraulic lime (Besses) 4 7 4 It will be noted that the percentage of elongation of the rods, varied directly with the percentage of alumina in the cements tested, proving conclusively that the swelling was due to the action of the calcium sulphaluminate formed during the operation. STANDARD METHODS OF TESTING, AM. SOC. C.E. A Committee appointed by the American Society of Civil Engineers to examine methods of making cement tests, offered a progress report early in 1903, and a brief supplementary report in 1904. These reports have been combined and are presented below. Standard Methods of Cement Testing. SAMPLING. 1. Selection of sample. The selection of the sample for testing is a detail that must be left to the discretion of the engineer; the number and the quantity to be taken from each package will depend largely on the importance of the work, the number of tests to be made and the facilities for making them. 2. The sample shall be a fair average of the contents of the packing; it is recommended that, where conditions permit, one barrel in every ten. be sampled. 3. All samples should be passed through a sieve having 20 meshes per linear inch in order to break up lumps and remove foreign material; 604 CEMENTS, LIMES, AND PLASTERS. this is also a very effective method for mixing them together in order to obtain an average. For determining the characteristics of a ship- ment of cement, the individual samples may be mixed and the average tested; where time will permit, however, it is recommended that they be tested separately. 4. Method of sampling. Cement inj^arrels should be sampled through a hole made in the center of one of the staves, midway between the heads, or in the head, by means of an auger or a sampling iron similar to that used by sugar inspectors. If in bags, it should be taken from surface to center. CHEMICAL ANALYSIS. 5. Significance. Chemical analysis may render valuable service in the detection of adulteration of cement with considerable amounts of inert material, such as slag or ground limestone. It is of use, also, in determining whether certain constituents, believed to be harmful when in excess of a certain percentage, as magnesia and sulphuric anhy- dride, are present in inadmissible proportions. While not recommending a definite limit for these impurities, the committee would suggest that the most recent and reliable evidence appears to indicate that mag- nesia to the amount of 5 per cent, and sulphuric anhydride to the amount of 1.75 per cent, may safely be considered harmless. 6. The determination of the principle constituents of cement silica, alumina, iron oxide and lime is not conclusive as an indication of quality. Faulty character of cement results more frequently from imperfect preparation of the raw material or defective burning than from incor- rect proportions of the constituents. Cement made from very finely ground material, and thoroughly burned, may contain much more lime than the amount usually present and still be perfectly sound. On the other hand, cements low in lime may, on account of careless preparation of the raw material, be of dangerous character. Further, the ash of the fuel used in burning may so greatly modify the composition of the product as largely to destroy the significance of the results of analysis'. 7. Method. As a method to be followed for the analysis of cement, that proposed by the Committee on Uniformity in the analysis of Ma- terials for the Portland Cement Industry, of the New York Section of the Society for Chemical Industry, and published in the Journal of the Society for January 15, 1902, is recommended. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 605 SPECIFIC GRAVITY. 8. Significance, The specific gravity of cement is lowered by under- burning, adulteration and hydration, but the adulteration must be in considerable quantity to affect the results appreciably. 9. Inasmuch as the differences in specific gravity are usually very small, great care must be exercised in making the determination. 10. When properly made, this test affords a quick check for under- burning or adulteration. 11. Apparatus and method. The determination of specific gravity is most conveniently made with Le Chatelier's apparatus. This con- sists of a flash of 120 cu. cm. (7.32 cu. inches) capacity, the neck of which is about 20 cm. (7.87 inches) long; in the middle of this neck is a bulb, above and beolw which are two marks ; the volume between these marks is 20 cu. cm. (1.22 cu. inches). The neck has a diameter of about 9 mm. (0.35 inch), and is graduated into tenths of cubic centimeters above the bulb. 12. Benzine (62 Baume naphtha), or kerosene free from water, should be used in making the determination. 13. The specific gravity can be determined in two ways: (1) The flask is filled with either of these liquids to the lower mark and 64 grs. (2.25 ozs.) of powder, previously dried at 100 C. (212 F.) and cooled to the temperature of this liquid, is gradually introduced through the funnel (the stem of which extends into the flask to the top of the bulb), until the upper mark is reached. The difference in weight between the cement remaining and the original quantity (64 grs.) is the weight which has displaced 20 cu. cm. 14. (2) The whole quantity of the powder is introduced, and the level of the liquid rises to some division of the graduated neck. This reading plus 20 cu. cm. is the volume displaced by 64 gr. of the powder. 15. The specific gravity is then obtained from the formula: . . Weight of cement Specific gravity = f r7- J , Tr , . Displaced Volume 16. The flask, during the operation, is kept immersed in water in a jar, in order to avoid variations in the temperature of the liquid. The results should agree within 0.01. 17. A convenient method for cleaning the apparatus is as follows: The flask is inverted over a large vessel, preferably a glass jar, and shaken vertically until the liquid starts to flow freely, it is then held still in a vertical position until empty, the remaining traces of cement can be 606 CEMENTS, LIMES, AND PLASTERS. removed in a similar manner by pouring into the flask a small quantity of clean liquid and repeating the operation. 18. More accurate determinations may be made with the pic- nometer. FINENESS. 19. Significance. It is generally accepted that the coarser particles in cement are practically inert, and it is only the extremely fine powder that possesses adhesive or cementing qualities. The more finely cement is pulverized, all other conditions being the same, thfe more sand it will carry and produce a mortar of a given strength. 20. The degree of final pulverization which the cement receives at the place of manufacture is ascertained by measuring the residue re- tained on certain sieves. Those known as the No. 100 and No. 200 sieves are recommended for this purpose. 21. Apparatus. The sieves should be circular, about 20 cm. (7.87 inches) in diameter, 6 cm. (2.36 inches) high, and provided with a pan, 5 cm. (1.97 inches) deep, and a cover. 22. The wire cloth should be woven (not twilled) from brass wire having the following diameters: No. 100, 0.0045 inch; No. 200, 0.0024 inch. 23. This cloth should be mounted on the frames without distortion; the mesh should be regular in spacing, and be within the following limits : No. 100, 96 to 100 meshes to the linear inch. No. 200, 188 to 200 " " " 24. Fifty grams (1.76 oz.) or 100 gr. (3.52 oz.) should be used for the test, and dried at a temperature of 100 C. (212 F.) prior to sieving. 25. Method. The Committee, after careful investigation, has reached the conclusion that mechanical sieving is not as practicable or efficient as hand-work, and, therefore, recommends the following method: 26. The thoroughly dried and coarsely screened sample is weighed and placed on the No. 200 sieve, which, with pan and cover attached, is held in one hand in a slightly inclined position, and moved forward and backward, at the same time striking the side gently with the palm of the other hand, at the rate of about 200 strokes per minute. The operation is continued until not more than one tenth of 1 per cent passes through after one minute of continuous sieving. The residue is weighed, then placed on the No. 100 sieve and the operation repeated. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 607 The work may be expedited by placing in the sieve a small quantity of large shot. The results should be reported to the nearest tenth of 1 per cent. NORMAL CONSISTENCY. 27. Significance. The use of a proper percentage of water in making the pastes * from which pats, tests of setting and briquettes are made, is exceedingly important, and affects vitally the results obtained. 28. The determination consists in measuring the amount of water required to reduce the cement to a given state of plasticity, or to what is usually designated the normal consistency. 29. Various methods have been proposed for making this determina- tion, none of which has been found entirely satisfactory. The Com- mittee recommends the following: 30. Method. Vicat needle apparatus. This consists of a frame bear- ing a movable rod, with the cap at one end, and at the other end the cylinder, 1 cm. (0.39 inch) in diameter, the cap, rod, and cylinder weigh- ing 300 gr. (10.58 oz.). The rod, which can be held in any desired position by a screw, carries an indicator, which moves over a scale (graduated to centimeters) attached to the frame. The paste is held by a conical, hard-rubber ring, 7 cm. (2.76 inches) in diameter at the base, 4 cm. (1.57 inches) high, resting on a glass plate about 10 cm. (3.94 inches) square. 31. In making the determination, the same quantity of cement as will be subsequently used for each batch in making the briquettes (but not less than 500 gr.) is kneaded into a paste, as described in Para- graph 58, and quickly formed into a ball with the hands, completing the operation by tossing it six times from one hand to the other, main- tained 6 inches apart; the ball is then pressed into the rubber ring through the larger opening, smoothed off, and placed on a glass plate (on its large end) and the smaller end smoothed off with a trowel; the paste, confined in the ring, resting on the plate, is placed under the rod bearing the cylinder, which is brought in contact with the surface and quickly released. 32. The paste is of normal consistency when the cylinder penetrates to a point in the mass 10 mm. (0.39 inch) below the top of the ring. Great care must be taken to fill the ring exactly to the top. *The term "paste" is used in this report to designate a mixture of cement and water, and the word "mortar" a mixture of cement, sand, and water. CEMENTS, LIMES, AND PLASTERS. 33. The trial pastes are made with varying percentages of water until the correct consistency is obtained. 34. The Committee has recommended, as normal, a paste, the con- sistency of which is rather wet, because it believes that variations in the amount of compression to which the briquette is subjected in mould- ing are likely to be less with such a paste. 35. Having determined .in this manner the proper percentage of water required to produce a neat paste' of normal consistency, the proper percentage required for the sand mortars is obtained from an empirical formula. 36. The Committee hopes to devise such a formula. The subject proves to be a very difficult one, and, although the Committee has given it much study, it is not yet prepared to make a definite recom- mendation. TIME OF SETTING. 37. Significance. The object of this test is to determine the time which elapses from the moment water is added until the paste ceases to be fluid and plastic (called the "initial set"), and abo the time required for it to acquire a certain degree of hardness (called the "final" or "hard set"). The former of these is the more important, since, with the commencement of setting, the process of crystallization or hardening is said to begin. As a disturbance of this process may produce a loss of strength, it is desirable to complete the operation of mixing and moulding or incorporating the mortar into the work before the cement begins to set. 38. It is usual to measure arbitrarily the beginning and end of the setting by the penetration of weighted wires of given diameters. 39. Method. For this purpose the Vicat needle, which has already been described in Paragraph 30, should be used. 40. In making the test, a paste of normal consistency is moulded and placed under the rod, as described in Paragraph 31; this rod, bear- ing the cap at one end and the needle, 1 mm. (0.039 inch) in diameter, at the other, weighing 300 gr. (10.58 oz.). The needle is then carefully brought in contact with the surface of the paste and quickly released. 41. The setting is said to have commenced when the needle ceases to pass a point 5 mm. (0.20 inch) above the upper surface of the glass plate, and is said to have terminated the moment the needle does not sink visibly into the mass. 42. The test pieces should be stored in moist air during the test; this is accomplished by placing them on a rack over water contained PHYSICAL PROPERTIES OF PORTLAND CEMENT. 609 in a pan and covered with a damp cloth, the cloth to be kept away from them by means of a wire screen; or they may be stored in a moist box or closet. 43. Care should be taken to keep the needle clean, as the collection of cement on the sides of the needle retards the penetration, while cement on the point reduces the area and tends to increase the penetration. 44. The determination of the time of setting is only approximate, being materially affected by the temperatures of the mixing water, the temperature and humidity of the air during the test, the percentage of water used, and the amount of moulding the paste receives. STANDARD SAND. 45. The Committee recognizes the grave objections to the standard quartz now generally used, e3pecially on account of its high percentage of voido, the difficulty of compacting in the moulds, and its lack of uniformity; it has spent much time in investigating the various natural sando which appeared to be available and suitable for use. 46. For the present, the Committee recommends the natural sand from Ottawa, 111., screened to pass a sieve having 20 meshes per linear inch and retained on a sieve having 30 meshes per linear inch; the wires to have diameters of 0.0165 and 0.0112 inch, respectively, i.e., half the width of the opening in each case. Sand having passed the No. 20 sieve shall be considered standard when not more than 1 per cent passes a No. 30 sieve after one minute continuous sifting of a 500 gr. sample. 47. The Sandusky Portland Cement Company, of Sandusky, Ohio, has agreed to undertake the preparation of this sand, and to furnish it at a price only sufficient to cover the actual cost of preparation. \ FORM OF BRIQUETTE. 48. While the form of the briquette recommended by a former Committee of the Society is not wholly satisfactory this Committee is not prepared to suggest any change, other than rounding off of the corners by curves of J-inch radius. MOULDS. 49. The moulds should be made of brass, bronze, or some equally non-corrodible material, having sufficient metal in the sides to pre- vent spreading during moulding 610 CEMENTS, LIMES, AND PLASTERS. 50. Gang moulds, which permit moulding a number of briquettes at one time, are preferred by many to single moulds; since the greater quantity of mortar that can be mixed tends to produce greater uni- formity in the results. 51. The moulds should be wiped with an oily cloth before using. MIXING. 52. All proportions should be stated by weight; the. quantity of water to be used should be stated as a percentage p_f the dry material. 53. The metric system is recommended because of the convenient relation of the gram and the cubic centimeter. 54. The temperature of the room and the mixing water should be as near 21 C. (70 F.) as it is practicable to maintain it. 55. The sand and cement should be thoroughly mixed dry. The mixing should be done on some non-absorbing surface, preferably plate glass. If the mixing must be done on "an absorbing surface it should be thoroughly dampened prior to use. 56. The quantity of material to be mixed at one time depends on the number of test pieces to be made; about 1000 gr. (35.28 oz.) makes a convenient quantity to mix, especially by hand methods. 57. The Committee, after investigation of the various mechanical mixing-machines, has decided not to recommend any machine that has thus far been devised, for the following reasons: (1) The tendency of most cement is to "ball up" in the machine thereby preventing the working of it into a homogeneous paste; (2) there are no means of ascertaining when the mixing is complete with- out stopping the machine, and (3) the difficulty of keeping the machine clean. 58. Method. The material is weighed and placed on the mixing table, and a crater formed in the center, into which the proper per- centage of clean water is poured; the material on the outer edge is turned into the crater by the aid of a trowel. As soon as the water has been absorbed, which should not require more than one minute, the operation is completed by vigorously kneading with the hands for an additional 1J minutes, the process being similar to that used in kneading dough. A sand-glass affords a convenient guide for the time of kneading. During the operation of mixing, the hands should be protected by gloves, preferably of rubber. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 611 MOULDING. 59. Having worked the paste or mortar to the proper consistency, it is at once placed in the moulds by hand. 60. The Committee has been unable to secure satisfactory results with the present moulding-machines; the operation of machine mould- ing is very slow, and the present types permit of moulding but one briquette at a time, and are not practicable with the pastes or mortars herein recommended. 61. Method. The moulds should be filled at once, the material pressed in firmly with the fingers and smoothed off with a trowel with- out ramming; the material should be heaped up on the upper surface of the mould, and, in smoothing off, the trowel should be drawn over the mould in such a manner as to exert a moderate pressure on the excess material. The mould should be turned over and the operation repeated. 62. A check upon the uniformity of the mixing and moulding is afforded by weighing the briquettes just prior to immersion, or upon removal from the moist closet. Briquettes which vary in weight more than 3 per cent from the average should not be tested. STORAGE OF THE TEST PIECES. 63. During the first twenty-four hours after moulding, the test pieces should be kept in moist air to prevent them from drying out. 64. A moist closet or chamber is so easily devised that the use of the damp cloth should be abandoned if possible,, Covering the test pieces with a damp cloth is objectionable, as commonly used, because the cloth may dry out unequally, and, in consequence, all the test pieces are not maintained under the same condition. Where a moist closet is not available, a cloth may be used and kept uniformly wet by immer- sing the ends in water. It should be kept from direct contact with the test pieces by means of a wire screen or some similar arrangement. 65. A moist closet consists of a soapstone or slate box; or a metal- lined wooden box, the metal lining being covered with felt and this felt kept wet. The bottom of the box is so constructed as to hold water, and the sides are provided with cleats for holding glass shelves on which to place the briquettes. Care should be taken to keep the air in the closet uniformly moist. 66. After twenty-four hours in moist air, the test pieces for longer periods should be immersed in water maintained as near 21 C. (70 F.) 612 CEMENTS, LIMES, AND PLASTERS. as practicable; they may be stored in tanks or pans, which should be of non-corrodible material. TENSILE STRENGTH. 67. The tests may be made on any standard machine. A solid metal clip is recommended. This clip is to be used without cushion- ing at the points of contact with the test specimen. The bearing at each point of contact should be J inch wide, and tfee distance between the centers of contact on the same clips should be 1J inches. 68. Test pieces should be broken as soon as they are removed from the water. Care should be observed in centering the briquette in the testing-machine, as cross-strains, produced by improper centering, tend to lower the breaking strength. The load should not be applied too suddenly, as it may produce vibration, the shock from which often breaks the briquette before the ultimate strength is reached. Care must be taken that the clips and the sides of the briquettes be clean and free from grains of sand or dirt, which would prevent a good bear- ing. The load should be applied at the rate of 600 Ibs. per minute. The average of the briquettes of each sample tested should be taken as the test, excluding any results which are manifestly faulty. CONSTANCY OF VOLUME. 69. Significance. The object is to develop those qualities which tend to destroy the strength and durability of a cement. As it is highly essential to determine such qualities at once, tests of this char- acter are for the most part made in a very short time, and are known, therefore, as accelerated tests. Failure is revealed by cracking, check- ing, swelling, or disintegration, or all of these phenomena. A cement which remains perfectly sound is said to be of constant volume. 70. Methods. Tests for constancy of volume are divided into two classes: (1) normal tests, or those made in either air or water main- tained at about 21 C. (70 F.), and (2) accelerated tests, or those made in air, steam, or water at a temperature of 45 C. (115 F.) and upward. The test pieces should be allowed to remain twenty-four hours in moist air before immersion in water or steam, or preservation in air. 71. For these tests, pats, about 7J cm. (2.95 inches) in diameter, 1} cm. (0.49 inch) thick at the center, and tapering to a thin edge, should be made, upon a clean glass plate [about 10 cm. (3.94 inches) square], from cement paste of normal consistency. PHYSICAL PROPERTIES OF PORTLAND CEMENT. 613 72. Normal test. A pat is immersed in water maintained as near 210 C. (70 F.) as possible for twenty-eight days, and observed at intervals; the pat should remain firm and hard and show no signs of cracking, distortion, or disintegration. A similar pat is maintained in air at ordinary temperature and observed at intervals. 73. Accelerated test. A pat is exposed in any convenient way in an atmosphere . of steam, above boiling water, in a loosely closed vessel, for three hours. 74. To pass these tests satisfactorily, the pats should remain firm and hard, and show no signs of cracking, distortion, or disintegration. 75. Should the pat leave the plate, distortion may be detected best with a straight-edge applied to the surface which was in contact with the plate. 76. In the present state of our knowledge it cannot be said that cement should necessarily be condemned simply for failure to pass the accelerated tests, nor can a cement be considered entirely satisfactory simply because it has passed these tests. CHAPTER XL. SPECIFICATIONS FOR PORTLAND CEMENT. VARIOUS specifications for Portland cement have been collected for insertion in the present chapter. These are of interest partly for com- parison and partly to show the growth of intelligent treatment of this subject. The specifications of the American Society for Testing Ma- terials will, it is probable, become the standard in this country. New York State Canals, 1896. The mortar and grout will be made of the best quality of Portland or natural hydraulic cement, as may be directed, and clean, sharp sand, in such proportions and made and used in such manner as may be re- quired by the engineer. No cement shall be used in any part of the masonry until the State engineer shall have examined, tried, and approved the same. It must be delivered in tight casks or bags, as the division or resident engineer may direct, and thereafter be properly protected from the weather. The engineer to direct in what manner the sand shall be screened and worked, and washed, if necessary. When considered necessary by the engineer, salt shall be used with the mortar in such manner and proportions as he may direct. Special directions shall be given by the engineer as to the delivery of cement and as to the time and facilities required for testing it previous to its use in the work. No cement will be used except in compliance with these directions. All facilities required by the engineer for securing tests must be afforded by the contractor. All cement must be stored in substantial water-proof structures from the time of delivery till used. All cement offered for use in any work will be sampled by an agent of the State Engineer's Department. Samples will be collected immedi- ately on delivery of cement at site of work, and contractors will promptly notify the engineer of the receipt of cement, in order that no delay may be had in the sampling thereof. All samples will be forwarded to the 614 SPECIFICATIONS FOR PORTLAND CEMENT. 615 cement-testing office in Albany, and will be subjected to the following tests, and any cement failing on either of them will be rejected, though the further right is reserved to reject any and all cements the qualities of which have not become well known through prior use in State work or elsewhere. Portland cement must be of the best quality and of such fineness that 95 per cent of the cement will pass through a sieve of 2500 meshes to the square inch, and 90 per cent through a sieve of 10,000 meshes per square inch. Portland cement when mixed neat and exposed one day in air and six days in water shall withstand a tensile strain of not less than 400 Ibs. to the square inch, and when mixed in the ratio of 3 Ibs. clean, sharp sand to 1 Ib. of cement and exposed one day in air and six days in water, it shall withstand a tensile strain of not less than 125 Ibs. per square inch. Rapid-transit Subway, New York City, 1900-1901. Fineness. Ninety-eight per cent shall pass a No. 50 sieve and 90 per cent a No. 100 sieve. Tensile strength. At the end of one day in water after hard set, 150 Ibs. neat; at the end of seven days, one day in air, six days in water, 400 Ibs. neat; at the end of twenty-eight days, one day in air, twenty-seven days in water, 500 Ibs. neat. When mixed 2 to 1 with quartz sand: At the end of seven days, one day in air, six days in water, 200 Ibs.; at the end of twenty-eight days, one day in air, twenty-seven days in water, 300 Ibs. Chemical analyses. Chemical analyses will be made from time to time, and cement furnished must show a reasonably uniform com- position. Soundness. Tests for checking and cracking and for color will be made by molding, on plates of glass, cakes of neat cement about 3 inches in diameter, \ inch thick in the center, and with very thin edges. One of these cakes when set per ectly hard shall be put in water and examined for distortion or cracks, and one shall be kept in air and examined for color, distortion, and cracks. Another cake shall be allowed to set in steam for twenty-four hours and then put in boiling water for twenty-four hours. Another cake shall be allowed to set hard in dry air for twenty-four hours and then put in boiling water for twenty-four hours. Such cakes should at the end of the tests still adhere to the glass and show neither cracks nor distortion. A briquette, in like man- ner, should be allowed to set hard in dry air for twenty-four hours, 16 CEMENTS, LIMES, AND PLASTERS. then boiled for twenty-four hours, be kept for five days in water, and show 350 Ibs. tensile strength. Department of Bridges^ New York City, 1901. (106) That all cement used on this work must be the best quality of imported or American Portland cement, manufactured by works of established reputation for furnishing a high-grade and uniform product. Cement must show a chemical analysis satisfactory to the engineer. (107) That briquettes of neat cement exposed to air for twenty- four hours and then immersed in water for six days must have a tensile strength of at least 400 Ibs. per square inch. (108) That briquettes of mortar mixed in proportion of one part of cement to two and one half parts of dry sand, by weight, exposed to the air for twenty-four hours and then immersed in water for six days, must have a tensile strength of not less than 180 Ibs. per square inch. (109) That cement mast be ground so fine that 90 per cent of it will pass through a sieve of 10,000 meshes per square inch. (110) That pats of neat cement set in the air and then immersed in boiling water for twenty-four hours must show no checks or cracks. (111) That cement must be sufficiently fresh to have lost no strength from age, but it must not be so fresh as to be "hot" and quick-setting. Neat cement at temperature of 70 F. must not take an initial set in less than thirty minutes, nor its final set in less than one hour. (112) That the contractor must provide adequate storage and enough cement ahead to enable seven-day tests to be made before cement has to be used. (113) That the contractor must furnish every reasonable facility to the inspectors for drawing samples of cement, and not less than ten days (holidays and Sundays excluded) must elapse between time of drawing the samples and using the cement. (114) That cement must at all times be protected from dampness, .air-currents, or other source of injury. (115) That the laboratory tests given above are not final. Should the engineer at any time deem any lot of cement damaged or question- .able in any respect, the same shall be rejected, although it may pre- viously have met other tests. (116) That cement must be delivered on the work in barrels of 375 Ibs. net weight, or in sacks of 94 Ibs. net weight. SPECIFICATIONS FOR PORTLAND CEMENT. 617 Engineer Corps, U. S. Army, 1902. (1) The cement shall ^be an American Portland, dry and free from lumps. By a Portland cement is meant the product obtained from the heating or calcining up to incipient fusion of intimate mixtures, either natural or artificial, or argillaceous with calcareous substances, the calcined product to contain at least 1.7 times as much of lime, by weight, as of the materials which give the lime its hydraulic proper- ties, and to be finely pulverized after said calcination, and thereafter additions or substitutions for the purpose only of regulating certain properties of technical importance to be allowable to not exceeding 2 per cent of the calcined product. (2) The cement shall be put up in strong, sound barrels well lined with paper, so as to be reasonably protected against moisture, or in stout cloth or canvas sacks. Eacn package shall be plainly labeled with the name of the brand and of the manufacturer. Any package broken or containing damaged cement may be rejected or accepted as a fractional package, at the option of the United States agent in local charge. (3) Bidders will state the brand of cement which they propose to furnish. The right is reserved to reject a tender for any brand which has not established itself as a high-grade Portland cement and has not for three years or more given satisfaction in use under climatic or other conditions of exposure of at least equal severity to those of the work proposed. (4) Tenders will be received only from manufacturers or their author- ized agents. (The following paragraph will be substituted for paragraphs 3 and 4 above when cement is to be furnished and placed by the contractor: No cement will be allowed to be used except established brands of high-grade Portland cement which have been made by the same mill and in successful use under similar climatic conditions to those of the proposed work for at least three years.) (5) The average weight per barrel shall not be less than 375 Ibs. net. Four sacks shall contain one barrel of cement. If the weight as deter- mined by test weighings is found to be below 375 Ibs. per barrel, the cement may be rejected, or, at the option of the engineer officer in charge, the contractor may be required to supply, free of cost to the United States, an additional amount of cement equal to the short- age. 618 CEMENTS, LIMES, AND PLASTERS. (6) Tests may be made of the fineness, specific gravity, soundness, time of setting, and tensile strength of the cement. (7) Fineness. Ninety-two per cent of the cement must pass through a sieve made of No. 40 wire, Stubbs gauge, having 10,000 openings per square inch. (8) Specific gravity. The specific gravity of the cement, as deter- mined from a sample which has been carefully dried, shall be between 3.10 and 3.25, (9) Soundness. To test the soundness of the*- cement, at least two pats of neat cement mixed for five minutes with 20 per cent of water by weight shall be made on glass, each pat about 3 inches in diameter and J inch thick at the center, tapering thence to a thin edge. The pats are to be kept under a wet cloth until finally set, when one is to be placed in fresh water for twenty-eight days. The second pat will be placed in water which will be raised to the boiling-point for six hours, then allowed to cool. Neither should show distortion or cracks. The boiling test may or may not reject at the option of the engineer officer in charge. (10) Time of setting. The cement shall not acquire its initial set in less than forty-five minutes and must have acquired its final set in ten hours. (The following paragraph will be substituted for the above in case a quick-33tting cement is desired: The cement shall not acquire its initial set in less than twenty nor more than thirty minutes, and must have acquired its final set in not less than forty-five minutes nor in more than two and one-half hours.) #.- The pats made to test the soundness may be used in determining the time of setting. The cement is considered to have acquired its- initial set when the pat will bear, without being appreciably indented,, a wire -fa inch in diameter loaded to weigh J Ib. The final set has been acquired when the pat will bear, without being appreciably indented, a wire -fa inch in diameter loaded to weigh 1 Ib. (11) Tensile strength. Briquettes made of neat cement, after being kept in air for twenty-four hours under a wet cloth and the balance of the time in water, shall develop tensile strength per square inch as follows : After seven days, 450 Ibs.; after twenty-eight days, 540 Ibs. Briquettes made of 1 part cement and 3 parts standard sand, by weight, shall develop tensile strength per square inch as follows: After seven days, 140 Ibs.; after twenty-eight days, 220 Ibs. SPECIFICATIONS FOR PORTLAND CEMENT. 619 (In case quick-setting cement is desired, the following tensile strength shall be substituted for the above: Neat briquettes : After seven days, 400 Ibs,; after twenty-eight days, 480 Ibs. Briquettes of 1 part cement to 3 parts standard sand: After seven days, 120 Ibs. ; after twenty-eight days, 180 Ibs.) (12) The highest result from each set of briquettes made at any one time is to be considered the governing test. Any cement not show- ing an increase of strength in the twenty-eight-day tests over the seven- day tests will be rejected. (13) When making briquettes neat cement will be mixed with 20 per cent of water by weight, and sand and cement with 12J per cent of water by weight. After being thoroughly mixed and worked for five minutes, the cement or mortar will be placed in the briquette mold in four equal layers, and each layer rammed and compressed by thirty blows of a soft brass or copper rammer three quarters of an inch in diameter (or seven tenths of an inch square, with rounded corners), weighing 1 Ib. It is to be allowed to drop on the mixture from a height of about half an inch. When the ramming has been completed, the surplus cement shall be struck off and the final layer smoothed with a trowel held almost horizontal and drawn back with sufficient pressure to make its edge follow the surface of the mold. (14) The above are to be considered the minimum requirements. Unless a cement has been recently used on work under this office, bidders will deliver a sample barrel for test before the opening of bids. If this sample shows higher tests than those given above, the average of tests made on subsequent shipments must come up to those found with the sample. (15) A cement may be rejected in case it fails to meet any of the above requirements. An agent of the contractor may be present at the making of the tests, or, in case of the failure of any of them, they may be repeated in his presence. If the contractor so desires, the en- gineer officer in charge may, if he deem it to the interest of the United States, have any or all of the tests made or repeated at some recognized standard testing laboratory in the manner herein, specified. All ex- penses of such tests to be paid by the contractor. All such tests shall be made on samples furnished by the engineer officer from cement ac- tually delivered to him. 620 CEMENTS, LIMES, AND PLASTERS. U. S. Reclamation Service, 1904. 1. Definition. The cement shall be high-grade Portland cement, By the term Portland cement is to be understood the material obtained by finely pulverized clinker produced by burning to semi-fusion an inti- mate mixture of finely ground calcareous and argillaceous materials. 2. Composition. It must be of normal composition, in which the proportion of the sum of calcium oxide and alkalies to the sum of the silica, alumina, and ferric oxide must not be less than 1.7 to 1 nor more than 2.2 to 1. It shall not contain over 3 per cent of magnesia nor 2J per cent of sulphate of lime. But in certain cases where such amounts of these substances are objectionable the engineer in charge may specify lower percentages. Its freedom from uncombined lime shall be deter- mined as in article 12. The question of adulteration may be determined either by chemical analyses or by inspection of the process at the factory. 3. Bids. Bids will be received only from manufacturers or their authorized agents, and the name of the brand offered shall in all cases be stated. 4. Weight per barrel or sack. The average weight per barrel shall not be less than 375 Ibs. net. Four sacks shall contain 1 barrel of cement. If the weight as determined by test weighings is found to be below 375 pounds per barrel, the contractor may be required to supply, free of cost to the United States, an additional amount of cement equal to the shortage. 5. Barrels ; damaged cement. If the cement is delivered in barrels, the barrels shall be strong and lined with paper, and the cement shall be free from lumps. Any package that is broken or that contains dam- aged cement may be rejected by the United States agent in local charge. 6. Sampling. Samples of cement are to be taken from the barrels or sacks with a sampling-tube in such manner as to secure fair average of the packages. They are to be taken from every tenth barrel or for- tieth sack and numbered, and the packages from which they are taken to be sealed and corresponding numbers attached for future identifi- cation. The quantities taken are to be kept separate and tested separately. Where the results of tests indicate variation in the quality of the cement, additional barrels or sacks will be sampled and tested. 7. Aeration and testing. No cement shall be shipped until at least sixty days after its manufacture, except that in case of an emergency, and with the approval of the engineer in charge, a shorter time may be allowed, but if the cement shows indications of unsoundness, a longer SPECIFICATIONS FOR PORTLAND CEMENT. 621 time may be required. The contractor shall keep in storage, in sacks or barrels, such stocks of cement as the engineer shall require, free of expense to the United States, for sampling and testing during a period of twenty-eight days. 8. Shipment. The engineer shall give notice in writing to the con- tractor of the approximate requirements for cement shipments and of dates for sampling. In all cases the contractor shall be responsible for the delivery of the cement in good condition at the place of consignment. 9. Factory inspection. The Government engineer, or his author- ized agent, shall at all times have liberty to inspect the materials, process of manufacture, and daily laboratory records of analyses and tests at the cement works. 10. Fineness. Ninety-five per cent by weight must pass through a No. 100 sieve having 10,000 meshes per square inch, the wire to be No. 40 Stubbs wire gauge; and 75 per cent by weight must pass through a No. 200 sieve having 40,000 meshes per square inch, the wire to be No. 48 Stubbs wire gauge. 11. Specific gravity. The specific gravity of the cement shall not be less than 3. 12. Soundness. Pats are to be made of neat mortar of normal consistency. The pats are to be molded on glass plates. They are to be circular in shape, 3 inches in diameter, J inch thick in the center, arid drawn to a thin edge at their circumference, and are to be kept under a wet coth, or in a moist atmosphere, until finally set. One pat is to be put in water, the temperature of which is to be raised to the boiling- point and kept at that point for six hours. If the pat softens, cracks, warps, or disintegrates, the cement is unsound. 13. Time of setting. The cement shall not acquire its initial set in less than forty-five minutes, and must acquire its final set within twelve hours. The pats made to test the soundness may be used in determin- ing the time of setting. The cement is considered to have acquired its initial set when the pat will bear, without being appreciably indented, a wire T V inch in diameter loaded to weigh one fourth pound. The final set has been acquired when the pat will bear, without being appre- ciably indented, a needle -fa inch in diameter loaded to weigh one pourd. 14. Making briquettes. In making briquettes, neat cement mortar of normal consistency will be used. The mortar will be thoroughly mixed with a trowel and kneaded into the molds with the thumbs, a blunt stick, or a plunger. Six briquettes will be made from each sample. In making sand briquettes, the proportions shall be one part by weight of cement to three parts of standard crushed quartz sand and about 622 CEMENTS, LIMES, AND PLASTERS. half as much water as is used for neat briquettes. Six briquettes will be made from each sample. 15. Tensile strength. The neat briquettes prepared as specified above shall stand a minimum tensile strain per square inch as follows : For one day in air and six days in water 450 Ibs. For one day in air and twenty-seven days in water. . . 550 " ' f* The sand-mortar briquettes, prepared as specified above, shall stand a minimum tensile strain per square inch as follows : After one day in air and six days in water . 175 Ibs. After one day in air and twenty-seven days in water . . 225 ' ' 16. Requirements. The above are to be considered the minimum requirements. The neat tests are to be considered of less value than those of sand and cement. The twenty-eight-day tests must always be higher than the seven-day tests. A cement may be rejected which fails to meet any of the above requirements. Canadian Society of Civil Engineers.* The whole of the cement is to be well-burned pure Portland cement, of the best quality, free from free lime, slag dust, or other foreign mate- rial. (1) Fineness. The cement shall be ground so fine that the residue on a sieve of 10,000 meshes to the square inch shall not exceed 10 per cent of the whole by weight, and the whole of the cement shall pass a sieve of 2500 meshes to the square inch. (2) Specific gravity. The specific gravity of the cement shall be at least 3.09, and shall not exceed 3.25 for fresh cement, .the term " fresh" being understood to apply to such cements as are not more than two months old. (3) Tests. The cement shall be subjected to the following tests: (a) Blowing test. Mortar pats of neat cement thoroughly worked shall be troweled upon carefully cleaned 5-inch by 2J-inch ground- glass plates. The pats shall be about J inch thick in the center and worked off to the sharp edges at the four sides. They shall be covered with a damp cloth and allowed to remain in the air until set, after which they shall be placed in vapor in a tank in which the water is heated to a temperature of 130 F. After remaining in the vapor six hours, * Proposed Canadian standard specifications for Portland cement. Cement, vol. 4, pp. 98-99. May, 1903. SPECIFICATIONS FOR PORTLAND CEMENT. 623 including the time of setting in air, they shall be immersed in the hot water and allowed to remain there for eighteen hours. After removal from the water the samples shall not be curled up, shall not have fine hair cracks, nor large expansion cracks, nor shall they be distorted. If separated from the glass, the samples shall break with a sharp, crisp ring. (b) Tensile est (neat cement). Briquettes made of neat cement mixed with about 20 per cent of water, by weight, after remaining one day in air, in a moist atmosphere, shall be immersed in water, and shall be capable of sustaining a tensile stress of 250 Ibs. per square inch after submersion for two days, 400 Ibs. per square inch after sub- mersion for six days, 500 Ibs. per square inch after submersion for twenty-seven days. The tensile test shall be considered as the aver- age of the strength of five briquettes, and any cement showing a decrease in tensile strength on or before the twenty-eighth day shall be rejected. (Sand and cement.) The sand for standard tests shall be clean quartz, crushed so that the whole shall pass through a sieve of 400 meshes to the square inch, but shall be retained on a sieve of 900 meshes per square inch. The sand and cement shall be thoroughly mixed dry, and then about 10 per cent of their weight of water shall be added, when the briquettes are to be formed in suitable molds. After remaining in a damp chamber for twenty-four hours the briquettes shall be immersed in water, and briquettes made in the proportion of one of cement to three of sand, by weight, shall bear a tensile stress of 125 Ibs. per square inch after submersion for six days, and 200 Ibs. per square inch after submersion for twenty-eight days. Sand and cement briquettes shall not show a decrease in tensile strength at the end of twenty-eight days or subsequently. (4) The manufacturer shall, if required, supply chemical analyses of the cement. (5) Packing. The cement shall be packed either in stout air- and water-tight casks, carefully lined with strong brown paper, or in strong air- and water-tight bags. (6) The manufacturer shall give a certificate with each shipment of cement, stating (1) the date of manufacture; (2) the tests and analyses which have been obtained for the cement in question at the manufacturer's laboratory; (3) that the cement does not contain any adulteration. 624 CEMENTS, LIMES, AND PLASTERS. Concrete-steel Engineering Company.* No cement will be allowed to be used except established brands of high-grade Portland cement which has been in successful use under similar conditions to the work proposed for at least three years, and has been seasoned or subjected to aeration for at least thirty days before leaving the factory. All cement shall be dry and free from lumps, and immediately upon receipt shall be stored in a dry, well-covered, and ventilated place thoroughly protected from'- the weather. If required the contractor shall furnish a certified statement of the chem- ical composition of the cement and of the raw material from which it is manufactured. The fineness of the cement shall be such that at least 90 per cent will pass through a sieve of No. 40 wire, Stubbs gauge, having 10,000 openings per square inch, and at least 75 per cent will pass through a sieve of No. 45 wire, Stubbs gauge, having 40,000 openings per square inch. Samples for testing may be taken from every bag or barrel, but usually for tests of 100 barrels a sample will be taken from every tenth barrel. The samples will be mixed thoroughly together while dry, and the mixture be taken as the sample for test. Tensile tests will be made on specimens prepared and maintained until tested at a temperature not less than 60 F. Each specimen will have an area of 1 square inch at the breaking section and after being allowed to harden in moist air for twenty-four hours will be immersed and maintained under water until tested. The sand used in preparing test specimens shall be clean, sharp, crushed quartz retained on a sieve of 30 meshes per lineal inch, and passing through a sieve of 20 meshes per lineal inch. In test speci- mens of one cement and three sand, no more than 12 per cent of water by weight shall be used. Specimens prepared from a mixture of one part cement and three parts sand, parts by weight, shall after seven days develop a tensile strength of not less than 170 Ibs. per square inch, and not less than 240 Ibs. per square inch after twenty-eight days. Cement mixed neat from 20 per cent to 25 per cent of water to form a stiff paste shall, after 30 minutes, be appreciably indented by the end of a wire inch T V in diameter loaded to weigh J Ib. Cement made into * The specifications from which this section is taken were published in Cement, vol. 4, pp. 105-108, May, 1903. They are for concrete-steel structures on the Melan, Thacher, and Von Emperger patents. SPECIFICATIONS FOR PORTLAND CEMENT. 625 thin pats on glass plates shall not crack, scale, or warp under the fol- lowing treatment: Three pats will be made and allowed to harden in moist air at from 60 to 70 F. ; one of these will be placed in fresh water for twenty-eight days, another will be placed in water which will be raised to the boiling-point for six hours and then allowed to cool, and the third is to be kept in the air of the prevailing outdoor temperature. British Standard Specifications.* Quality and preparation. (1) The cement is to be prepared by intimately mixing together calcareous and argillaceous materials, burn- ing them at a clinkering temperature and grinding the resulting clinker. No addition of any material is to be made after burning, except when desired by the manufacturer, and if not prohibited in writing by the consumer, in which case calcium sulphate or water may be used. The cement, if watered, shall contain not more than 2 per cent of water, whether that water has been added or has been naturally absorbed from- the air. If calcium sulphate is used, not more than 2 per cent calcu- lated as anhydrous calcium sulphate of the weight of the cement shall be added. Sampling and preparation for testing and analysis. (2) As soon as the cement has been bulked at the maker's works, f or on the works in connection with which the material is to be used, at the consumer's option, samples for testing are to be taken from each parcel, each sample consisting of cement from at least twelve different positions in the same heap, so distributed as to insure, as far as is practicable, a fair average sample of the whole parcel, all to be mixed together and the sample for testing to be taken therefrom. (3) Before gauging the tests, the sample so obtained is to be spread out for a depth of 3 inches for twenty-four hours, in a temperature of 58 to 64 F. (4) In all cases where consignments are of 100 tons and upwards samples selected as above from each consignment, either at the maker's works or after delivery at the works where the cement is to be used, are to be sent for expert testing and for chemical analysis. In no case is cement so tested and analyzed to be accepted or used unless * British standard specifications for Portland cement. Engineering News, vol. 53, pp. 227-228. March 2, 1905. t Should the consumer desire to stipulate for any special quantity, the size of the heap should be stated. 626 CEMENTS, LIMES, AND PLASTERS. previously certified in writing by the consumer to be of satisfactory quality. Payment for such tests and analyses to be made by the con- sumer, the manufacturer supplying the cement required for the same free of charge. When consignments of less than 100 tons have to be supplied, the maker shall, if required, give certificates for each delivery, to the effect that such cement complies with the* terms of this standard specification, with regard to quality, tests, and chemical analyses, no payment being made by the consumer for such certificate nor for the making of such tests and analyses. (5) Should it be deemed more convenient by "the consumers that the samples for testing should be taken at the makers' works before delivery, the latter are, in that event, to afford full facilities to the inspector who may be appointed by the consumers to sample the cement as he may desire at the makers' works, and subsequently to identify each parcel as it may be dispatched, with that sampled by him. No parcel is to be sent away unless a written order has been previously received by the makers from the said consumer to the effect that the material in question has been approved. Fineness and sieves. (6) The cement shall be ground to comply with the following degrees of fineness, viz.: The residue on a sieve 76X76 = 5776 meshes per square inch is not to exceed 5 per cent. The residue on a sieve 180X180 = 32,400 meshes per square inch is not to exceed 22J per cent. The sieves are to be prepared from standard wire; the size of the wire for the 5776 mesh is to be .0044 inch and for the 32,400 mesh .0018 inch. The wire shall be woven (not twilled), the cloth being carefully mounted on the frames without distortion. Specific gravity. (7) The specific gravity of the cement shall be not less than 3.15 when sampled and hermetically sealed at the makers' works, nor less than 3.10 if sampled after delivery to the consumer. Chemical composition. (8) The cement is to comply with the follow- ing conditions as to its chemical composition. There shall be no excess of lime, that is to say, the proportion of lime shall be not greater than is necessary to saturate the silica and alumina present. The percent- age of insoluble residue shall not exceed 1.5 per cent; that of mag- nesia shall not exceed 3 per cent, and that of sulphuric anhydride shall not exceed 2.5 per cent. Tensile tests. (9) The quantity of water used in gauging shall be appropriate to the quality of the cement, and shall be so proportioned that when the cement is gauged it shall form a smooth, easily worked SPECIFICATIONS FOR PORTLAND CEMENT. 627 paste that will leave the trowel cleanly in a compact mass. Fresh water is to be used for gauging, the temperature thereof, and of the test- room at the time the said operations are performed, being from 58 to 64 F. The cement gauged as above is to be filled, without mechanical ram- ming, into molds ; each mold resting upon an iron plate until the cement has set. When the cement has set sufficiently to enable the mold to be removed without injury to the briquette, such removal is to be effected. The said briquettes shall be kept in a damp atmosphere and placed in fresh water twenty-four hours after gauging and kept there until broken, the water in which the test briquettes are submerged being renewed every seven days and the temperature thereof maintained between 58 and 64 F. Neat tests. (10) Briquettes of neat cement are to be gauged for breaking at seven and twenty-eight days, respectively, six briquettes for each period. The average tensile strength of the six briquettes shall be taken as the accepted tensile strength for each period. For breaking, the briquette is to be held in strong metal jaws, the briquettes being slightly greased where gripped by the jaws. The load must then be steadily and uniformly applied, starting from zero, increasing at the rate of 100 Ibs. in twelve seconds. The briquettes are to bear on the average not less than the following tensile stresses before breaking: 7 days from gauging 400 Ibs. per square inch of section. 28 days from gauging 500 Ibs. per square inch of section. The increase from seven to twenty-eight days shall not be less than: 25% when the 7-day test falls between 400 to 450 Ibs. per square inch. 20% when the 7-day test falls between 450 to 500 Ibs. per square inch. 15% when the 7-day test falls between 500 to 550 Ibs. per square inch. 10% when the 7-day test falls between 550 Ibs. per square inch or upwards. Sand tests. (11) The cement shall also be tested by means of briquettes prepared from one part of cement to three parts by weight of dry standard sand, the said briquettes being of the shape described for the neat-cement tests; the mode of gauging, filling the molds, and breaking the briquettes is also to be similar. The proportion of water used shall be such that the mixture is thoroughly wetted, and there shall be no superfluous water when the briquettes are formed. The cement and sand briquettes are to bear the following tensile stresses: 7 days from gauging 120 Ibs. per square inch of section. 28 days from gauging 225 Ibs. per square inch of section. 628 CEMENTS, LIMES, AND PLASTERS. The increase from seven to twenty-eight days shall not be less than 20 per cent. The standard sand referred to above is to be obtained from Leighton Buzzard. It must be thoroughly washed, dried, and pass through a sieve of 20X20 meshes per square inch, and must be retained on a sieve of 30X30 meshes per square inch, tjae wires of the sieve being .0164-inch and .0108-inch respectively. Setting-time. (12) There shall be three distinct gradations of set- ting-time, which shall be designated as " quick", "inedium", and "slow".* Quick. The setting-time shall not be less than ten minutes or more than thirty minutes. Medium. The setting-time shall not be less than half an hour or more than two hours. Slow. The setting-time shall not be less than two hours or more than five hours.* The temperature of the air in the test-room at the time of gauging and of the water used is to be between 58 and 64 F. The cement shall be considered as "set" when a needle having a flat end -f$ inch square, weighing in all 2i Ibs., fails to make an im- pression when its point is applied gently to the surface. Soundness. (13) The cement shall be tested by the Le Chatelier method; and is in no case to show a greater expansion than 12 milli- meters after twenty-four hours' aeration and 6 millimeters after 7 days' aeration. Note. The apparatus for conducting the Le Chatelier test consists of a small split cylinder of spring brass or other suitable metal of 0.5 millimeter (.0197 in.) in thickness, 30 millimeters (1.1875 inches) inter- nal diameter, and 30 millimeters high, forming the mold, to which on either side of the split are attached two indicators 165 millimeters (6.5 inches) long frorn the center of the cylinder, with pointed ends. In conducting the test the mold is to be placed upon a small piece of glass and filled with cement gauged in the usual way, care being taken to keep the edges of the molds gently together while this operation is being performed. The mold is then covered with another glass plate, a small weight is placed on this, and the mold is immediately placed in water at 58 to 64 F. and left there for twenty-four hours. The distance separating the indicator points is then measured and the mold placed in cold water, which is brought to the boiling-point in * When a specially slow-setting cement is required the minimum time of setting shall be specified. SPECIFICATIONS FOR .PORTLAND CEMENT. 629 15 to 30 minutes and kept boiling for six hours. After cooling, the distance between the points is again measured; the difference between the two measurements represents the expansion of the cement, which must not exceed the limits laid down in this specification. (14) The tests and analyses hereinbefore referred to shall in no case relate to a larger quantity of cement than 250 tons sampled at one time. Acceptance. (15) No cement is to be approved or accepted unless it fully complies with the foregoing conditions. American Society for Testing Materials, 1904. GENERAL OBSERVATIONS. 1. These remarks have been prepared with a view of pointing out the pertinent features of the various requirements and the precautions to be observed in the interpretation of the results of the tests. 2. The committee would suggest that the acceptance or rejection under these specifications be based on tests made by an experienced person having the proper means for making the tests. 3. Specific gravity. Specific gravity is useful in detecting adultera- tion or uriderburning. The result of tests of specific gravity are not necessarily conclusive as an indication of the quality of the cement, but when in combination with the results of other tests may afferd valuable indications. 4. Fineness. The sieves should be kept thoroughly dry. 5. Time of setting. Great care should be exercised to maintain the test pieces under as uniform conditions as possible. A sudden change or wide range of temperature in the room in which the tests are made, a very dry or humid atmosphere, and other irregularities vitally affect the rate of setting. 6. Tensile strength. Each consumer must fix the minimum re- quirements for tensile strength to suit his own conditions. They shall, however, be within the limits stated. 7. Constancy of volume. The tests for constancy of volume are divided into two classes, the first normal, the second accelerated. The latter should be regarded as a precautionary test only, and not infallible. So many conditions enter into the making and interpreting of it that it should be used with extreme care. 8. In making the pats the greatest care should be exercised to avoid initial strains due to molding or to too rapid drying out during the first twenty-four hours. The pats should be preserved under the most 630 CEMENTS, LIMES, AND PLASTERS. uniform conditions possible, and rapid changes of temperature should be avoided. 9. The failure to meet the requirements of the accelerated tests need not be sufficient cause for rejection. The cement may, however, be held for twenty-eight days and a retest made at the end of that period. Failure to meet the requirements #t this time should be considered sufficient cause for rejection, although in the present state of our knowl- edge it cannot be said that such failure necessarily indicates unsound- ness, nor can the cement be considered entirely satisfactory simply because it passes the tests. STANDARD SPECIFICATIONS FOR CEMENT. 1. General conditions. All cement shall be inspected. 2. Cement may be inspected either at the place of manufacture or on the work. 3. In order to allow ample time for inspecting and testing, the cement should be stored in a suitable weather-tight building having the floor properly blocked or raised from the ground. 4. The cement shall be stored in such a manner as to permit easy access for proper inspection and identification of each shipment. 5. Every facility shall be provided by the contractor and a period of at least twelve days allowed for the inspection and necessary tests. 6. Cement shall be delivered in suitable packages with the brand and name of manufacturer* plainly marked thereon. 7. A bag of cement shall contain 94 Ibs. of cement net. Each barrel of Portland cement shall contain 4 bags, and each barrel of natural cement shall contain 3 bags, of the above net weight. 8. Cement failing to meet the seven-day requirements may be held awaiting the results of the twenty-eight-day tests before rejection. 9. All tests shall be made in accordance with the methods proposed by the Committee on Uniform Tests of Cement of the American Society of Civil Engineers, presented to the society Jan. 21, 1903, and amended Jan. 20, 1904, with all subsequent amendments thereto. 10. The acceptance or rejection shall be based on the following requirements : PORTLAND CEMENT. 18. Definition. This term is applied to the finely pulverized product resulting from the calcination to incipient fusion of an intimate mixture of properly proportioned argillaceous and calcareous materials, and to SPECIFICATIONS FOR PORTLAND CEMENT. 631 which no addition greater than 3 per cent has been made subsequent to calcination. 19. Specific gravity. The specific gravity of the cement, thoroughly dried at 100 C., shall be not less than 3.10. 20. Fineness. It shall leave by weight a residue of not more than 8 per cent on the No. 100, and not more than 25 per cent on the No. 200-sieve. 21. Time of setting. It shall develop initial set in not less than thirty minutes, but must develop hard set in not less than one hour, nor more than ten hours. 22. Tensile strength. The minimum requirements for tensile strength for briquettes one inch square in section shall be within the following limits, and shall show no retrogression hi strength within the periods specified : NEAT CEMENT. Age. Strength. 24 hours in moist air 150 200 Ibs. 7 days (1 day in air, 6 days in water) 450 550 ' ' 28 days (1 day in air, 27 days in water) 550 650 " One Part Cement, Three Parts Sand: 7 days (1 day in moist air, 6 days in water) 150 200 u 28 days (1 day in moist air, 27 days in water) 200 300 ' ' 23. Constancy of volume. Pats of neat cement about three inches in diameter, one half inch thick at the center, and tapering to a thin edge, shall be kept in moist air for a period of twenty-four hours. (a) A pat is then kept in air at normal temperature and observed at intervals for at least 28 days. (6) Another pat is kept in water maintained as near 70 F. as prac- ticable, and observed at intervals for at least 28 days. (c) A third pat is exposed in any convenient way in an atmosphere of steam, above boiling water, in a loosely closed vessel for five hours. 24. These pats, to satisfactorily pass the requirements, shall remain firm and hard and show no signs of distortion, checking, cracking or disintegration. 25. Sulphuric acid and magnesia. The cement shall not contain more than 1.75 per cent of anhydrous sulphuric acid (SOs), nor more than 4 p'er cent of magnesia (MgO). PART VIL PUZ20LAN CEMENTS. CHAPTER XLI. PUZZOLANIC MATERIALS IN GENERAL. PUZZOLANIC materials include all those natural or artificial materials which are capable of forming hydraulic cements on being simply mixed with lime, without the use of heat. Many materials possess this property, but relatively few have ever attained to sufficient commercial impor- tance to be discussed here. In composition the puzzolanic materials are largely made up of silica and alumina, usually with more or less iron oxide; some, as the slags used in cement-manufacture, carry also notable percentages of lime. As might be inferred from this compo- sition, most of the puzzolanic materials possess hydraulicity to a greater or less degree of themselves, but the addition of lime usually greatly increases their hydraulic power. The term puzzolan, here adopted for this group of cementing mate- rials, is a corruption of the adjective form of the name pozzuolana. It has no particular etymological excuse for existence, but will be accepted in this volume for the sake of uniformity, as it seems to have been adopted by various authorities in the United States. Natural Puzzolanic Materials. Natural puzzolanic materials are quite widely distributed, though they have never attained much commercial importance, save in Europe. As regards their origin, they are of two classes: In the first class may be included all those which are the direct products of volcanic action, the material being a fine volcanic ash or dust deposited either on the slopes of the volcano or carried by the wind to lakes or streams in which the ash is deposited. This group includes the more active puzzo- lanic materials, its chief representatives being pozzuolana proper, san- 632 PUZZOLANIC MATERIALS IN GENERAL. 633 torin, tosca, tetin and trass. It may be noted that in origin materials of this class resemble closely the granulated slags used in slag-cement manufacture both volcanic ashes and granulated slags being due to the processes of (1) fusion of a silico-aluminous material, and (2) rapid cooling of the resulting product by ejection into air or immersion in water. The second class includes a number of less important (because less active) hydraulic materials, such as arenes, psammites, etc., which are materials resulting from the decay of certain igneous rocks. The principal natural puzzolanic materials will be discussed sepa- rately, in the following order: Pozzuolana (tosca, tetin), trass, san- torin, arenes. Pozzuolana. Pozzuolana derives its name from the little town of Pozzuoli, located a few miles west of Naples, at which point the material was first obtained 'by the Greek colonists, and at a later date by the Romans. The material has also been exploited at other points near Rome and Naples. TABLE 234. ANALYSES OF POZZUOLANA FROM ITALY. Si0 2 A1 2 3 FegOa CaO MgO * K 2 Na 2 O H 2 1 58.58 "~22T74 4.06 1.37 2 52.66 14.33 10.33 7.66 3.86 4.13 7.03 3 44.5 15.0 12.0 8.8 4.7 1.4 4.0 9.2 4 63.18 19.8 5.68 0.35 5 60.91 21.28 4.76 1.90 0.00 4.37 6.23 6 44.0 10.5 29.5 10.0 tr. 1.00 2.5 7 44.5 15.75 16.3 8.96 tr. 11.0 3.5 8 46.0 16.5 15.5 10.0 3.0 4.0 5.0 9 44.5 15.5 12.5 9.5 4.4 10.27 3.33 10 39.0 14.0 13.0 18.0 3.0 11.0 11 56.31 15.23 7.11 i.74 1.36 6.54 2.84 6.12 1. Pozzuolana, Rome. Stanger and Blount, Mineral Industry, vol. 5, p. 71. St. Paul's Caves. Thoyn, Diet. App. Chem., 3d ed., vol. 1, p. 475. Civita Vecchia. Berther, Anal. Gillmon, Limes, Cements, and Mortars, p. Naples. Stanger and Blount, Mineral Industry, vol. 5, p. 71. . Stengel, Anal. Zervas, School of Mines Quart, vol. 18, p. 230. Vesuvius. Brown. Thorpe, Diet. App. Chem., 3ded., vol. 1, p. 475. " " 3ded., vol. 1, p. 475. Dark gray. 3d ed., vol. 1, p. 475. Light gray. " 3d ed., vol. 1, p. 475. 10. Lava, Vesuvius, 1868. Thorpe, Diet. App. Chem., 3d ed., vol. 1, p. 475. 11. Tuff, Monte Nuova. Merrill, Rocks, Rock Weathering and Soils, p. 141. Most of the Italian pozzuolana is obtained from small open cuts, or pits, though some of these workings are now of great depth. Those of Trent aremi, for example, are about 600 feet deep. The various 634 CEMENTS, LIMES, AND PLASTERS. deposits differ greatly in the quality of the materials obtained from them. Care should therefore be exercised in selecting a spot for exploita- tion, and sorting of the material dug would be advisable in order to keep the product of uniformly high grade. After extraction the ma- terial is screened and ground. In addition it is occasionally slightly roasted, which process increases its ^hydraulic properties. Carelessness, both in the mining and in the later preparation of the pozzuolana, has brought the Italian article somewhat into disrepute among Euro- pean engineers. In consequence it is losing ground with respect both to pozzuolana from the Azores and to trass from Rhenish Prussia. Pozzuolana is also obtained at a number of localities in southeastern France. These localities occur mostly in three areas : (1) in the Auvergne Mountains, lying in the Departments of Puy de Dome and Cantal; (2) in the Mountains du Vivarais, between Haute Loire and Ardeche; and (3) in the Department of PHerault, near the Gulf of Lyons. TABLE 235. ANALYSES OF POZZUOLANA FROM FRANCE. 1. 2. 3. 4. 5. 6. 7. Silica (SiO 2 ) Alumina (A1 2 O 3 ) 47 ..9 47.1 46.05 17 48.0 35.09 17 65 30.73 11 63 38.50 18 35 Iron oxide (Fe 2 O 3 ) .... Lime (CaO) J34.2 8 2 39.0 | 7 20.55 8 55 K 4 { 8 10 16.82 4 26- 24.92 3 73 14.90 8 70 Magnesia (MgO) Alkalies (K 2 O,Na 2 O) . . Water (H 2 O) 3.9 2.6 3 2 tr. 4.7 2 2 tr. 6.35 1 6 tr. 4.8 2 4 3.17 n. d. 19 06 2.49 n. d. 19 02 tr. 7.30 7.75 1. Auvergne Mountains, black. Thorpe, Diet. App. Chem., vol. 1, p. 475. 2. " reddish-brown. " 3. " " brick-red. 4. Gravenydre. 5. Vivarais Mountains, gray. Vicat, analyst. 6. brown. 7. Department of 1'Herault, brown. Vicat, analyst. Pozzuolana has been shipped from San Miguel and Terceira in the Azores, to Portugal for over a hundred years, and has been used with very satisfactory results in many important buildings, harbor works, etc. The Azores pozzuolana varies in color from yellowish to brownish, and sometimes to grayish. It is frequently so fine-grained as not to require screening or grinding before use. A reddish colored v.ariety from the same islands is termed tetin. A similar ash, locally called "tosca", is obtained from Teneriffe, one of the Canary Islands, and shipped to Spain for use as a cementing material. PUZZOLANIC MATERIALS IN GENERAL. TABLE 236. ANALYSES OP POZZUOLANA FROM THE AZORES ISLANDS. 635 1. 2. 3. Silica (SiO 2 ) 60 90 54 70 57 73 11.14 20 50 13 81 Iron oxide (Fe 2 O 3 ) 12 78 6 30 12 02 Lime (CaO) 2 57 2 20 3 74 Magnesia (MgO) 1 45 1 70 1 73 Potash (K 2 O) 2 64 3 21 Soda (Na 2 O) 2 74 > 2.20< 2 76 Water (H r O) 5 78 12 40 4 66 1. From St. Miguel. tetin. Zervas, analyst. School of Mines Quarterly, vol. 18, p. 230. 2. pozzuolana. Chateau, 3. " Terceira. Zervas, Volcanic materials of a type somewhat different from normal pozzuo- lana occur on Tile Bourbon, a French island lying about 400 miles east of Madagascar. ANALYSIS OP VOLCANIC ASH, ILE BOURBON. Silica (SiO 2 ) 25.67 Alumina (A1 2 O 3 ) 16.33 Iron oxide (Fe 2 O 3 ) 40.00 Magnesia (MgO) tr. Water (H 2 O) 17.00 Trass. Trass is a pale yellowish to grayish rock, rough to the feel, composed of an earthy or compact pumiceous dust mixed with fragments of pumice, trachyte, carbonized wood, etc. It is, so far as origin is concerned, an ancient volcanic mud. Trass occurs along the Rhine, in Rhenish Prussia, from Koln on the north to Coblenz on the south. The towns of Brohl, Kruft, Plaidt, and Andernach, all located north- west of Coblenz and within fifteen miles of that city, are prominent points in connection with the trass industry. A series of analyses of trass and related products is given in Table 237. Santorin. The island of Santorin, or Thera, is one of the most southeasterly of the islets of the Grecian Archipelago, lying in the Cyclades group. An ash called in commerce "santorin ", derived from the volcano of the same as a cementing material. name, is quite extensively shipped for use 636 CEMENTS, LIMES, AND PLASTERS. TABLE 237. ANALYSES OP TRASS AND RELATED MATERIALS FROM GERMANY. SiO 2 A1 2 3 Fe 2 3 CaO MgO K 2 Na 2 O H 2 O 1 46.25 20.71 5.48 2.15 1.00 6. 30~~ 9.25 2 46.6 20.6 12.0 3.0* 5.0 12.8 3 48. S4 18.95 12.34 5.41 '2i42 0.37 3.56 11.94 4 53.07 18.28 3.43 1.24 1.31 4.17 3.73 12.78 5 53.58 19.11 9.24 3.21 0.30 4.84 1.87 7.50 6 54.0 16.5 6.1 4.0 0.7 JO.O 7.0 7 55.28 17.34 3.90 3.17 0.87 4.70 3.80 10.63 8 57.0 16.0 5.0 2.6 1.0 7.0 1.0 9.6 9 57.5 10.1 3.9 7.7 1.1 6.4 12.6 10 58.32 20.88 4.15 2.19 1.10 3.91 4.11 5.87 11 61.10 12.70 10.20 8.10 1.90 2.10 2.10 1.40 12 59.40 22.70 2.50 3.10 0.80 3.50 2.80 4.80 13 60.49 19.95 9.37 3.12 1.43 3.40 1.33 14 62.83 21.55 4.11 0.72 0.42 3.35 3.02 4.19 15 66.39 17.74 4.97 0.53 0.47 3.05 1.94 4.89 16 67.60 11.30 5.20 8.20 2.80 0.60 0.50 3.10 1. Trass. Rhenish. Thorpe, Diet. App. Chem., 3d ed., vol. 1, p. 475. 2. Dutch. Thorpe, p. 475. 3. Andernach. Thorpe, p. 475. 4. Plaidt. von Decken, Anal. Zirkel, Lehrbuch der Petrographie, 1894, vol. 3, p. 678* 5. Zervas, Anal. Zervas, School of Mines Quart., vol. 18, p. 230. 6. Andernach. Chatoney and Rivol, Anal. Zirkel, p. 678. 7. Kruft. Mengerschausen, Anal. Zervas, p. 320. 8. Brohl. Berthier, Anal. Gillmore, Limes, Cements, and Mortars, p. 125. 9. Andernach. Chatoney and Rivol, Anal. Zirkel, p. 678. 10. Brohl. Bruhus, Anal. Zirkel, p. 678. 11. " Kyll, Anal. Zervas, p. 320. 12. Trachyte tuff. Siebengebirge, Kyll, Anal. Zervas, p. 320. 13. " " Laacher See. Merrill, Rocks, Rock Weathering and Soils, p. 141. 14. " " Siebengebirge. Bischof, Anal. Zirkel, p. 675. 15. " " von der Marck, Anal. Zirkel, p. 675. 16. Leucite tuff. Weibern. Kyll, Anal. Zervas, p. 320. TABLE 238. ANALYSES OF SANTORIN ASH, FROM SANTORIN. l. 2. 3. 4. Silica (SiO 2 ) . . 72 84 71.44 63.07 66 37 Alumina (AlgO^ 12 26 9.87 15.67 13 72 Iron oxide (Fe 2 O 3 ) 4.35 3.84 8.73 4.31 Lime (CaO) 2.55 2.64 3.83 2.98 1.58 1.84 1.93 1.29 Potash (K 2 CN 1 28 1 86 1 87 2 83 Soda (NaO) 2 65 3 74 3 86 4 22 Water (H 2 O) 2 25 4 61 1.14 4 06 1. Pumiceous portion. Feichtinger, analyst. Thorpe, Diet. App. Chem., vol. 1, p. 476. 2. Fine ash. 3. Obsidian particles. 4. Average sample. " " School of Mines Quarterly, vol. 18, p. 230. PUZZOLANIC MATERIALS IN GENERAL. 637 Arenes, etc. The materials called "arenes" by early French writers on cement technology are sands and residual material derived from the decay of various igneous rocks, and particularly from the decay of the more basic rocks, such as trap, basalt, etc. Such materials will naturally vary greatly in composition and properties, but all of them agree in possessing feeble hydraulicity. - For present-day commercial purposes they are practically worthless. TABLE 239. ANALYSES OF ARENES, FRANCE. 1. 2. 3. 4. 5. Silica (SiO 2 ) 38.54 60 33 42 10 38.50 60 30 Alumina (AlgO^ 20.00 21 43 23 . 65 29.40 23.70 Iron oxide (Fe 2 O 3 ) 12.00 8 57 22.47 18.10 10.30 Lime (CaO) 8 OO 1 1 ... ( tr 2 00 tr Magnesia (MgO) n d \ 6 . 69 < 2 50 Alkalies (K 2 O Na 2 O) n d n d 1 28 1 50 3 20 i Lime carbonate (CaCO 3 ). 1. Saint Astier, Department Dordogne. Vicat, analyst. 2. Brest. Vicat, analyst. 3. Saint Servan. Vicat, analyst. 4. 5. Chateaulin. As might be inferred from the examples given, natural materials: showing slightly hydraulic properties are not cf rare occurrence. With the exception of trass, santorin and pozzuolana proper, these materials are rarely sufficiently hydraulic to be of service as bases for puzzo- lan cements or mortars. The feebly hydraulic materials have, how- ever, a practical value which may be noted briefly here. It is that, owing to the fact that they are hydraulic, they can be profitably sub- stituted in places where they occur for common sand in mortar. Chelius has tested * the fine material remaining after the crushing of basalt in an ordinary stone-crusher. This fine material (dust and screenings) gave the following results as compared with normal sand; TABLE 240. STRENGTH OF BASALTIC DUST. Tension. Compression. 28 Days. 90 Days. 1 Year. 28 Days. 90 Days. 1 Year. 35.8 67.8 Cement and normal sand Cement and basalt fines Lime and normal sand 6^3 7.7 20.9 43.6 237.7 320.8 8.5 11.1 14.2 44.9 Lime and basalt fines * Journ. Soc. Chem. Ind., vol. 19, p. 826. 838 CEMENTS, LIMES, AND PLASTERS. Part of the superiority, as shown by these tests, of the basalt dust to normal sand is probably due to purely physical causes. In part, however, it is probably due to the fact that the finely crushed basalt acted as a puzzolanic material. Range and average composition of natural puzzolanic materials. From the separate tables of analyses given in preceding paragraphs the following table of average analyses have been prepared: TABLE 241. AVERAGE ANALYSES OF NATURAL PUZZOLANIC MATERIALS. Pozzuo- lana, Italy. Pozzuo- lana, France. Pozzuo- lana, Azores. Trass, Ger- many. San- torin. Average Natural Puzzo- lanic Material. 9 7 3 11 1 31 Silica (SiO 2 ) 50 98 41.91 57.78 53.78 66.37 51.08 Alumina (A1 2 O 3 ) . . 15 55 16 16 15 15 17 38 13 72 16 30 Iron oxide (Fe 2 O 3 ) 14 41 19 30 10 37 6 89 4 31 11 13 Lime (CaO) 7 39 6 93 2 84 3 89 2 98 5 46 Magnesia (MgO) 1 96 1 37 1 63 1 17 1 29 1 50 Alkalies (K 2 O,Na,O) 6 63 5 15 4 52 6 82 7 05 6 21 Water (H 2 O) 5 09 7 89 7 61 9 22 4 06 7 64 Puzzolanic materials in the United States. Volcanic ash and other materials which may be expected to show puzzolanic action occur ex- tensively in the western United States, but few tests appear to have been made of their hydraulic properties. Mr. J. S. Diller, in a recent description* of the mineral resources of the Redding district of California, has noted that a "tuff, bordering the northern end of the Sacramento Valley, is very like the trass of the Rhine Valley. This is especially true of that on Stillwater, near the Copper City road, or east of Millville, and at a number of points on the western side of the Sacramento Valley. The limestone and the tuff are at several places within a few miles of each other, and there is reason to believe that a good quantity of hydraulic cement may be made from them within convenient reach of the railroad. This matter is of impor- tance in the construction of large dams for irrigation or water-power in the Redding region. Similar volcanic products occur in Arizona, and have been used locally as puzzolanic materials. * Bulletin 225, U. S. Geological Survey, p. 177, 1904. PUZZOLAXIC MATERIALS IN GENERAL. 639 Artificial Puzzolanic Materials. Blast-furnace slag is by far the most prominent of the artificial poz- zuolanic materials. Other artificial materials have, however, been used for this purpose, burnt clay being one of the better known of these minor products. Burnt clay. The following recent note * is of interest in the present connection. "Mortar composed of lime and burnt clay was used extensively in constructing the Asyut Barrage completed in 1902, across the Nile, and described in detail in a paper by Mr. George Henry Stephens, M. Inst. C. E., to the institution of Civil Engineers on March 15, 1904. After being burnt the clay was ground and passed through a 100-mesh sieve. The best results were had with a clay burnt to a light terra-cotta color as compared with clay burned brick-red and clay burned dark red to purple. The ground clay was mixed with slaked lime and sand was added to form a mortar. The following are the results of long- time tensile tests made with various mixtures moulded into standard briquettes kept in water after 12 hours in air": TABLE 242. STRENGTH OF LIME BURNT-CLAY MORTARS. Mixture by Volume. Age 1 Year. Pounds per Square Inch. Age 2 Years. Pounds per Square Inch. 3 clay 1 Maximum 400 Maximum 410 2 lime 1 clay J Average of 52 samples \ Maximum 272 305 Average of 56 samples. . Maximum 320 350 1 lime f clay / Average of 38 samples. (Maximum 239 320 Average of 25 samples. . Maximum 291 376 1 lime sand Average of 37 samples. 259 Average of 55 samples. . 280 Blast-furnace slags. Slags, according to the general use of that term, are the fusible silicates formed during metallurgical operations by the combination of the fluxing materials with the gangue of the ore. The composition of the slag, therefore, depends upon the character and relative proportions of the gangue and the fluxes. The slag will, in general, contain only those elements present in either gangue or flux; though it may contain also a percentage, usually small, of the metal which is being reduced, and its composition may, in some processes, * Engineering News, vol. 53, p. 177. Feb. 16, 1905. 640 CEMENTS, LIMES, AND PLASTERS. be slightly modified by the presence of the elements taken up from the fuel. The slags or " cinders" obtained in refining the metals differ from the normal slags in that they may contain a very appreciable percentage of metal, sufficient in many cases to justify further treat- ment of the slag in order to recover its metallic contents. As this utili- zation of such slags is entirely a mejfcallurgical operation, they will not be further discussed in the present volume. While many elements may occur in slags, those which are of universal or even common occurrence are relatively few. fhe slags most com- monly formed are silicates, consisting essentially of silica, oxides of the alkaline elements, and certain metallic oxides, these last, with the ex- ception of alumina, being usually present in small quantity only. In certain metallurgical operations, however, the percentage of metallic oxides may rise so as to make them important ingredients in the slag. According to the processes, ores or fluxes used, slags may also contain more or less phosphoric anhydride, sulphur and fluorine. The particular use, or uses to which the slag from any given furnace may be most profitably put, will depend upon several factors. When considering possible utilizations, the most important factor will gen- erally be found to be the chemical composition of the slag. It is true that, for certain uses, as for example highway macadam and railroad ballast, the physical condition of the slag is of rather more importance than its chemical composition; but the two utilizations named are among the less profitable, and are only to be considered when the slag cannot be disposed of more profitably. Local conditions, under which head may be grouped questions of furnace management, possible markets, and transportation routes and charges, will be found to be of great economic importance. These factors are, however, too variable to be dis- cussed in the present volume, with one exception. The exception noted is the effect of slag utilization upon the general furnace management. The furnace manager who is endeavoring to profitably utilize his slag will often find it necessary to consider how far he may economically go in changing details of his main process in order to increase the value of his by-product. This is particularly the case where the slag is used for cement. Blast-furnace slags of certain types have been used extensively in Europe, and to a less extent in the United States, in the manufacture of slag cement. The following chapters will therefore be devoted to a discussion of the materials, manufacture and properties of slag cements. CHAPTER XLII. SLAG CEMENT. REQUISITES AND TREATMENT OF THE SLAG. SLAG cement is at present by far the most important member- of the group of puzzolan cements, so that its manufacture will be described in some detail. Summary of general methods of manufacture. Slag cement is com- posed of an intimate mechanical mixture of slaked lime and granu- lated blast-furnace slag of suitable chemical composition; both mate- rials being finely pulverized before, during or after mixing. The process of manufacture includes the granulating and drying of the slag, the slaking of the lime, the mixing of these materials, and the grinding of the resulting cement, together with any means which may be employed for the regulation of the setting time of the cement. These different factors hi the manufacture will be described in the order named above. In the present chapter, the character and treat- ment of the slag will be taken up. Composition of the Slag. Requisite chemical composition of slag. The slag used in cement- manufacture must be a basic blast-furnace slag. Tetmajer, the first investigator of slag cements, announced as the results of his experi- ments (a) that the hydraulic properties of the slag increased with the proportion of lime contained in it, and that slags in which the ratio CaO was so low as to approach unity were valueless for cement-manu- facture; (6) that, so far as the alumina content of the slag was con- cerned, the best results were obtained when the ratio c< .^. 3 gave a value of 0.45 to 0.50; and (c) that with any large increase of alumina above the amount indicated by this value of the alumina-silica ratio, the ten- dency of the cement to crack (when used in air) was increased. Prost, at a later date, investigated the subject, using for experi- 641 642 CEMENTS, LIMES, AND PLASTERS. ment several commercial slags and also a series prepared from pure CaO, Si0 2 , and A1 2 03. He decided that the hydraulic properties (both as regards rapidity of set and ultimate strength) of the slag increased as the proportions of lime and alumina increased; and failed to find any indication that a high alumina content causes disintegration. His best results were obtained .from slags having the compositions re- spectively of 2SiO 2 , A1 2 O 3 , 3CaO, arid 2SiO 2 , A1 2 O 3 , 4CaO. Mahon, in 1893, made a series of experiments to determine the value (for cement-manufacture) of a large series of the slags produced by the furnaces of the Maryland Steel Company: and found that the slags giving the best results were two having respectively the following com- positions : (1) Si0 2 , 30%; A1 2 3 , 17%; CaO, 47.5%; S, 2.38%; and (2) Si0 2 , 25.3%; A1 2 3 , 20.1%; CaO, 48%; MgO, 3.28%; S, 2.63%. The ratios of c ,.^ and c< .^ 3 , calculated for these slags are: At the close of the experiments Mahon recommended that slags be used slightly higher in alumina than those above quoted. Composition of slags actually used. The specifications under which slag from the furnaces is accepted by the cement department of the Illinois Steel Company are: (1) Slag must analyze within the following limits: Si0 2 + Al 2 3 not over 49%; A1 2 O 3; from 13 to 16%; MgO, under 4%. (2) Slag must be made in a hot furnace and must be of a light-gray color. (3) Slag must be thoroughly disintegrated by the action of a large stream of cold water directed against it with considerable force. This contact should be made as near the furnace as is possible. A series of over 300 analyses of slags used by this company in their slag (puzzolan) cement, show the following range in composition: Si0 2 , 29.60 to 35.60%; A1 2 O 3 , and Fe 2 O 3 , 12.80 to 16.80%; CaO, 47.99 to 50.48%;' MgO, 2.09 to 2.81%. The requirements of the Birmingham Cement Company as to the chemical composition of the slags used for cement are: that the lime content shall not be less than 47.9 per cent; that the silica and lime together shall approximately amount to 81 per cent; and that the alumina and iron oxide together shall equal from 12 to 15 per cent. SLAG CEMENT. REQUISITES AXD TREATMENT OF THE SLAG. 643 TABLE 243. ANALYSES OF SLAGS USED FOR SLAG CEMENT. Compo- nents. SiO 2 .. $&.: CaO. . MgO. . CaSO' SO, . Middlesboro, England. Harzb'g, Germany (used at Bruns- wick). Belgium. Bilbao, Spain. Choindez, Switzerland. 30.00 28.00 0.75 32.75 5.25 1.90 31.50 18.56 42^22 3.18 30.72 16.40 0.43 48.59 1.28 2.16 32.51 13.19 0.48 44.75 2.20 4.90 32.90 13.25 0.46 47.30 1.37 3.42 38.00 10.00 26.88 24.12 0.44 45.11 1.09 1.80 27.33 23.81 O.C3 45.83 0.92 1.34 0.17 1.67 0.87 2G.24 24.74 0.49 46.83 0.88 0.59 0.32 1.78 0.93 46.00 0.45 2.21 0.44 1.34 0.59 tr. 1.58 0.53 0.60 1.37 0.43 1.13 1.44 0.41 0.50 1.68 0.89 S ' MnO 2 . CaO. . Si0 2 . . A1A . SiO 2 . . 6.60 }l.09 JO. 93 1.21 0.27 Compo- nents. Saulnes, France. Marnaval (used at Donjeux). Pont-a- Mousson. Chicago, 111. SiO 2 . A1 2 3 . FeO. . CaO. . MgO. . S 31.65 17.00 0.65 47.20 1.36 31.50 16.62 0.62 46.10 28.35 18.15 1.5.0 47.40 2.45' 1 ^0 28.00 19.5 45.0 1.61 O.C9 32.00 22.0 4.00 + MgO 42.00 See FeC 1.31 O.C8 32.20 15.50 48.14 2.27 1.49 0.48 33.10 12.60 49.98 2.45 1.51 0.38 31.80 14.80 49.74 2.29 1.56 0.46 34.30 14.76 48.11 2.66 1.40 0.43 MnO,. CaO. . SiO 2 . . A1 2 O 3 . Si0 2 . . 0.85 }l.49 J0.53 1.46 0.52 1.67 O.C4 Analyses of a number of slags used in slag-cement manufacture are shown in Table 243. The analyses of foreign slags are quoted from various reliable authorities and the analyses of the Illinois Steel Com- pany slags have been selected from a large series published in the report of the U. S. Army Board of Engineers to show the extreme ranges of the different elements. The ratios ~.~ O1U2 , Al 2 Os , and . have been calculated for each slag and are shown in this table. From these data it can be seen that the ratio of alumina to silica is carried v$ry high at Choindez, and is rather low at Chicago, rela- tively to most of the European plants. It must be remembered, how- 644 CEMENTS, LIMES, AND PLASTERS. ever, that one reason for carrying a high alumina-silica ratio does not apply at Chicago, as there rapidity of set is gained by the use of the Whiting process. Taking these two plants as representative of the best European and American practice, the average of the analyses given shows the ratios actually used, to be: Choindez, Switzerland, 1.71, = 0.90; and Chica, 111., - -l. These results may be compared with the theoretical ratios advised by Tetmajer, Prost, and Mahon, and discussed on, a previous page of the present Chapter. Selection of slags. The erection of a slag-cement plant in connection with any given furnace is not justified, unless a sufficient amount of the slags usually produced will fall within slag-cement requirements, as these requirements have been outlined above in the section on chem- ical composition of the slag. In a large plant it will usually be easy to secure a constant supply of slag of proper composition without inter- fering with the proper running of the furnaces. In a small plant, how- ever, or in one running on a number of different ores, such a supply may be difficult to obtain. These points, of course, should be settled in advance of the erection of the cement-plant. In the case of any given furnace running on ores and fluxes, which are fairly steady in composition and proportions, the selection of the slag used for cement-making may be largely based on its color, checked if necessary by rapid determinations of lime. The darker-colored slags are generally richest in lime, except when the depth of color is due to the presence of iron; the lighter-colored slags are usually higher in silica and alumina. Candlot states further* in this connection that the slag issuing at the commencement and toward the end of a discharge should be rejected because of the air-chilling which attends its slow movement. Granulating the Slag: Methods and Effects. Assuming that a slag of proper composition has been selected, the first step in the actual manufacture of slag cement will be the " granu- lation " of the molten slag. Granulation is the effect produced by bringing molten slag into contact with a sufficient amount of cold water. The physical effect of this proceeding is to cause the slag to break up into porous particles ("slag sand")- Granulation has also certain chemical effects, highly important from an economic point of view, which will be discussed later. * Ciments et chaux hydrauliques. SLAG CEMENT. REQUISITES AND TREATMENT OF THE SLAG. 645 Methods of granulating the slag. The success of the granulation depends on bringing the slag into contact with the water as soon as possible after it has left the furnace. The effects of the process will be found to vary with, (a) the temperature of the slag at the point of contact; (b) the temperature of the water; (c) the amount of water used, and (d) its method of application. Taking up the last point first it may be noted that two general methods of application of the water have been used. In the first method the stream of slag, as it issued from the furnaces, was struck by a jet of steam under pressure. This method, which was used at one time in slag-cement plants in the Middlesboro district, England, had the effect of blowing the slag into fine threads with attached globules. It is, in fact, much the same as the process still used in. the manufacture of mineral wool. From an economic point of view it had the distinct advantage of putting the slag in a condition in which it was easily pul- verized by the grinding-machinery ; but it had certain inconveniences, and has been almost or entirely superseded by the method now to be mentioned. The second way in which the water may be applied is to allow the stream of slag as it issues from the furnace to fall into a trough contain- ing a rapidly flowing stream of cold water. Care must be taken that the fall into the trough is not too great, and that the stream of water is deep enough and fast enough, for otherwise the slag will acquire sufficient momentum in its fall to solidify in a mass on the bottom of the trough . This method is in use at all slag-cement plants of the present day, being occasionally modified by the use (either in addition to or in place of the flowing stream of water in the trough) of a jet of water playing on the slag before it strikes the trough. The following two examples, taken from present-day practice at American slag-cement plants, will serve to indicate two methods, dif- fering in minor details only, of slag granulation. At the first plant the furnaces are located on an embankment , about 8 feet above and 20 feet away from a standard gauge-switch track. A rectangular trench about 1 foot in width is dug from the furnace to near the edge of the embankment. Here a section of semi- circular sheet-iron troughing, 12 to 15 inches in diameter and about 10 feet in length, meets the trench. The inner end of the trough is fixed, and is at such a level that the bottom of the trough is about 6 inches below the bottom of the trench. The outer end of the trough is free and supported by wire ropes so that it can be readily swung into position over a box car on the switch track below. 646 CEMENTS, LIMES, AND PLASTERS. As already noted, the bottom of the earthen trench is about 6 inches above the bottom of the iron trough. This is done to tillow the inser- tion at this end of the trough of a 3-inch water-pipe. Slag from the furnace flows through the trench and into the trough, which is set at an inclination of about 1 inch in 10. Water is injected through the 3-inch pipe, under 10 or -15 feet hpad, into the trough. If enough water is used, the slag will be granulated as soon as it enters the trough, and will be readily carried down it into the car below, rarely flowing with a greater depth than 6 inches in the trough. If insuffi- cient water is used the slag puffs up and fills the trough, so that the slag-mass has to be broken into with an iron rod and pushed along. The car into which the slag flows is provided with four 3-inch holes in its sides, to allow the surplus water to escape. At another slag-cement plant recently visited by the writer, the granulated slag is caught in cylindrical masonry tanks, 15 feet in diameter and 10 feet in depth. The stream of molten slag flows from the furnace to and over the edge of the tank and through a semi-circular trough about 10 inches in diameter, which enters the tank at its top rim and projects G inches over the edge. About 6 inches below the bottom of this trough a pipe, carrying cold water under slight pressure, enters the tank, projecting into it for 4 inches. This pipe is 3 inches in diameter for most of its length, but the portion projecting into the tank is flattened so as to give an orifice 4 or 5 inches wide and about half an inch high. The stream of slag, flowing slowly along the trough and over the edge of the tank, is struck by the jet of cold water from the pipe, and is granulated. The granulated slag is taken from the tank by bucket elevators running continuously. Effects of granulating the slag The physical effect of causing hot slag to come in contact with cold water is to break the slag up into small porous particles. As this materially aids in pulverizing the slag, it is probable that granulation would be practiced on this account alone. But as a matter of fact, granulation has in addition to its purely physical result two important chemical effects. One is to make th slag, if it be of suitable chemical composition, energetically hydraulic; the other is to remove a portion of the sulphides contained in the slag in the form of hydrogen disulphide. Le Chatelier states that the hydraulic properties of granulated slag are due to the presence of a silico-alumino ferrite of calcium correspond- ing in composition to the formula 3CaO, A1 2 O 3 , 2SiO 2 . This com- pound appears also in Portland cements, but in them it is entirely inert, owing to the slow cooling it has undergone. When, however, as in SLAG CEMENT. REQUISITES AND TREATMENT OF THE SLAG. 647 the case of granulated slags, it is cooled with great suddenness, it becomes an important hydraulic agent. When so cooled "it is attackable by weak acids and also by alkalies. It combines particularly with hydrated lime in setting, and gives rise to silicates and aluminates of lime iden- tical with those which are formed by entirely different reactions during the setting ,of Portland cement. It is upon this property that the manu- 6000 4000 2000 , ^v l^-^-< X L $2&- to^- ^ e a OL < D / /> ^_StA TO_2_U< ' o 3 L | WE ^ ME ~L ) co d ffi <* /\ > Q. Qj CO D z GRANULA NOT GRA NULA a. _-o-- _ .- - < iE-fS=S ===^1 ~~ = =r=^= ) w 70 140 210 280 350 AGE IN DAYS COMPRESSION FIG. 159.* Effect of granulating slag. (Tetmajer.) 70 140 210 AGE IN DAYS TENSION FIG. 160.* Effect of granulating slag. 280 350 (Tetmajer.) facture of slag cements, which assumes daily greater importance, is based ". Increased hydraulicity due to granulation. The striking increase in the hydraulic properties of the slag when it is granulated was well brought out by Frost's investigations. The following table (244) ; giving the results of tests of tensile and compressive strength of briquettes of * From Johnson's " Materials of Construction", p. 190. 648 CEMENTS, LIMES, AND PLASTERS. both granulated and ungranulated slag, as determined by Prost, is of interest in this connection. The results of other tests, by Tetmajer, are shown diagrammatically in Figs. 159 and 160. TABLE 244. STRENGTH OF GRANULATED AND UNGRANULATED SL\O. (PROST.) Proportions of Mixture by Weight. V* Resistance in Kilograms per Square Centimeter. 28 Days. 84 Days. 210 Days. 360 Days. a Compression. g S 1 Compression. Tension. Comp revision . Tension.^ Compression. 33 parts lime, 100 parts slag: Granulated . . 33.7 32.1 27.6 259.9 233.7 205.2 43.5 5.4 38.1 5.4 34.3 377.5 308.2 248.9 46.4 10.7 40.5 10.5 38.9 7.6 440.5 50.5 326.7 54.1 267.8 47.6 44.4 13.8 35.3 13.3 38.1 10.8 438.7 59.9 350.9 62.4 253.1 63.8 Not granulated 66 parts lime, 100 parts slag: * Granulated Not granulated 100 parts lime, 100 parts slag: Granulated Not granulated Desulphurization due to granulation. When molten slag is poured into water, a very large proportion of the sulphur contained in the slag is carried off by the water. The extent to which the desulphuriz- ing of the slag is secured by the simple method of granulating is shown by the following result from actual practice at the slag brick works at Kralovedvoor, Bohemia. Here the slag is granulated, just as in slag- cement works, by running it into flowing cold water. Examination * -of the water used showed that it had increased in temperature from 14i C. to about 56 C., and that it carried in 10,000 parts the follow- ing parts of mineral matter in solution: SiO 2 0.426 CaSO 4 0.749 FeSO 4 . 108 MgSO 4 0.448 NaaSO 4 . 178 NaCl .038 N a2 SiO 2 0.693 CaS .271 H^ . 047 2.958 * Engineering and Mining Journal, April 16, 1898. SLAG CEMENT. REQUISITES AND TREATMENT OF THE SLAG. 649 Drying the Slag. The slag as it is brought to the cement mill from the granulating tanks carries from 15 to over 40 per cent of water absorbed during granulation. As will be noted later attempts have been made to utilize this contained water in the slaking of the lime, but these attempts have hitherto proved unsuccessful. As the manufacture is at present conducted, therefore, the large percentage of water carried by the slag is of no service, and in order to get good results from the grinding machinery the water must be removed as completely as possible before pulverization is attempted. Before describing the various types of driers in use, a few words on the general problem may be serviceable. The slag may carry, as above noted, from 15 to over 40 per cent of water, varying with the method of granulation, the fineness of grain, etc. In test runs slag can be thoroughly granulated without the use of more than 10 to 15 per cent of water, but in actual practice it will usually be found that the granulated slag carries from 30 to 45 per cent. As the slag must be reduced to extreme fineness it is necessary that this moisture be reduced as much as possible. With a well-conducted rotary drier it is possible to economically reduce the percentage of moisture in the dried product to about one-fourth of one per cent. The temperature to which the product is carried in drying is not a matter of serious moment so long as it does not pass the point at which the slag begins to re-fuse. Theoretically, of course, it is neces- sary only to carry the temperature above 212 F., but in practice it is economically impossible to keep it as low as this. It may be carried as high as a dull-red heat without injury to the slag. Indeed, it is probably the case that drying at relatively high temperatures improves the hydraulic properties of the slag, rather than otherwise, as it is well known that the natural puzzolanic materials are improved by roasting. It would not, therefore, be a matter of surprise if drying the slag at a higher temperature than is actually necessary should result in mate- rially accelerating the set of the resulting cement and also in increasing the strength of briquettes made from it. The Ruggles-Coles drier (see Fig. 161) consists of two concentric hollow cylinders bolted together and revolving on an axis slightly inclined from the horizontal. The outside cylinder is made of steel plates, the longitudinal seams having butt joints with inside lapping straps. The inner cylinder, which is also made of steel, is connected with the outer cylinder at its middle by heavy cast-iron arms A solidly riveted to 650 CEMENTS, LIMES AND PLASTERS. both cylinders, while the cylinders are further connected at each end by two sets of adjustable or swinging arms B, which prevent the joints being affected by the expansion or contraction of the cylinders. At the head or upper end the inner cylinder projects beyond the outer cylinder, passing into a stationary head or air chamber E to the hot air flue D of the furnace '"< with which it is connected. At the lower or discharge end is another stationary head E forming an air chamber, through an opening in the bottom of which the dried material is dis- charged. This head is supplied with a damper to regulate the tem- perature, which gives perfect control. The cylinders are set at an inclination of about f inch to the foot. The outer cylinder is secured to two heavy rolled steel-bearing rings, FIG. 161. Ruggles-Coles drier. which rest and revolve upon eight bearing wheels supported by oscil- lating arms or rockers. The lateral motion of the cylinder is taken up by four thrust wheels. The drier is revolved by a cast gear H secured to the outer cylinder, and this is driven by a shaft and pinion K extended beyond the end of the machine and supported in two babbitted journal boxes fitted to the frame. The entire machine is fitted and secured to a heavy frame of 8-inch I beams braced and framed together and usually set on a concrete foundation. The exhaust fan is placed where most convenient to drive, and is connected with the outer cylinder by a suitable flue L. The furnace G is built inde- pendent of the rest of the drier, and is connected with the head end of the inner cylinder by an iron flue D built with fire-brick. A specially designed burner is substituted for the furnace when oil, gas, or powdered coal are to be used. The heated air passes through the inner cylinder (which is shown by the dotted lines in the illustration) and returns, between the inner and outer cylinders, to the fan. The direction of the hot-air current is shown by arrows. The wet raw material is fed into the space between the inner and outer cylinders through a spout C in the stationary SLAG CEMENT. REQUISITES AND TREATMENT OF THE SLAG. 651 head at the upper end of the drier. The material is picked up by buckets or carriers fastened to the inner surface of the outer cylinder, and is carried partly around during the rotation of the drier. On dropping from these buckets it is caught by nights fastened to the outer surface of the inner cylinder. These flights carry the material partly around and then drop it on the outer cylinder, when the cycle of opera- tions commences again. While the movements of the material are occurring, it is being dried both by the heated-air current which flows through the space between the two cylinders, and by contact, with the warm outer surface of the inner cylinder; and it is also being carried slowly toward the lower or discharge end of the machine. The following table shows working results obtained in the use of the Ruggles-Coles drier on blast-furnace slag at various slag-cement plants : TABLE 245. WORKING RESULTS OF RUGGLES-COLES DRIER. User. Number of Driers. Original Percentage Moisture. Final Percentage Moisture. Water Evaporated per Hour. Knickerbocker Cement Co. Maryland Cement Co. .... Birmingham Cement Co. . . Southern Cement Co. . 1 2 2 2^ 41.82 20.32 45 40 0.29 0.25 4401 Lbs. 4114 " 4181 " 4707 " Stewart Cement Co 3 12 85 2272 " User. Dry Material Delivered pefSHour. Coal Used per Hour. Water Evaporated per Pound of Coal. Knickerbocker Cement Co. 6,399 Lbs. 560 Lbs. 7 87 Lbs. Maryland Cement Co ... 16,173 " 542 " 7 59 " Birmingham Cement Co. . 4,987 " 537 " 7 60 " Southern Cement Co 7,061 " 550 " 8 56 " Stewart Cement Co 15 408 " 334 " 6 80 " The Hoist drier is used at Donjeux and Mallstadt, and consists essentially of a sheet-iron cylinder 9 meters long and 0.8 meter in diam- eter, into which the slag is fed automatically by a screw feed. In the cylinder a helical screw revolves on a hollow central shaft, causing the slag to advance slowly through the cylinder. The fireplace is below and near one end of the cylinder and the heat is caused to pass under the cylinder to the other end, thence through the hollow shaft to the stack in a direction contrary to that in which the slag is moving. The cylinder is protected from the direct flame by brickwork. This appa- 652 CEMENTS, LIMES, AND PLASTERS. ratus dries from 7 to 8 metric tons of slag per day with a coal consump- tion of about 5 per cent of the weight of slag dried. At Vitry, France, a simple and effective non-rotary drier, operated by gravity, is employed, plan and section of which is given in Fig. 162. It consists of drying compartments (each of which is lettered, a, 6, c, d r in the plan), arranged about a central flue (c, d, c, d in plan), through which passes the heated gases from' a furnace. The central flue is 1 m. FIG. 162. Vitry slag-drier. square; the drying compartments 0.5X1 m. in area, and both are 7 m. in height. Each of the drying compartments contains 10 sheet- iron plates (of which only six, E, are shown in each compartment of the section). These plates are inclined and so arranged that the wet slag, shoveled in at the top of each compartment, descends by gravity and finally issues from the lowest plate in the heaps F, from which it is shoveled and sent to the grinding mills. According to Prost, a drier of this type and size will dry from 12 to 15 metric tons of slag per working day. From 6 to 6.5 Ibs. of coke are necessary to dry each 100 Ibs. of slag. Tower driers, resembling those used at one or two American Portland cement plants, could of course be used in drying slag. At present, however, every slag-cement plant in the United States uses rotary driers. CHAPTER XLIII. SLAG CEMENT: LIME, MIXING AND GRINDING. AFTER the slag has been granulated and dried, as described in the preceding chapter, it must be mixed with a carefully slaked lime, in proper proportions, and the mixture must be finely ground. These points will be taken up first in the present chapter, after which data on the general processes and costs of slag-cement manufacture will be presented. Composition and selection of the lime. The lime used for admix- ture with the slag may be either a quicklime (common lime) or a hydraulic- lime. In usual American practice, and also at most European plants, a common or quicklime is used. At a few American, French, and German plants, however, limes which have more or less hydraulic properties are employed. Prost has carried on experiments touching this point and decided that the use of a hydraulic lime did not notice- ably increase the tensile strength of the resulting cement, but that it did increase the value of the product in another way. This incidental advantage is that slag cements made by using hydraulic lime are less liable to fissure and disintegrate when used in air or in dry situations than cement in which common quicklime is used. As above noted, this method of improving the product has been tried, to the writer's knowledge, at only a few of the American plants. At Konigshof, Ger- many, the general practice at which plant is described on page 662, a somewhat hydraulic lime is used, whose analysis will serve as fairly representative of materials of this type, though most hydraulic limes would run considerably higher in silica and alumina. ANALYSIS OF HYDRAULIC LIME, KONIGSHOF, GERMANY. Per Cent. Silica (SiO 2 ) . . . 12.421 Alumina (A1 2 O 3 ) ........ 2 . 620 Iron oxide (Fe 2 O 3 ) .-., .883 Manganese oxide (MnO 2 ) . . . tr. Lime (CaO) . . 81 .646 Magnesia (MgO) . . 1 . 751 Soda (Na,O) 0.211 Carbon dioxide (CO 2 ) 0. 194 Moisture (H 2 O). ....... .. 0.425 653 654 CEMENTS, LIMES, AND PLASTERS. The following analyses are of limes used at different slag-cement plants in the United States: TABLE 246. ANALYSES OF LIMES USED IN AMERICAN SLAG-CEMENT PLANTS. ' 1. 7*2. 3. 4. 5. 6. Silica (SiO 2 ) 3 24 3 50 1 62 10 20 78 1 38 Alumina (A1 2 O 3 ) 1 , n Iron oxide (Fe 2 O 3 ) J4.26 3.92 2.62 3.60 0.52 0.62 Lime (CaO) 81 92 83 20 82 40 81 33 98 40 97 80 Magnesia (MgO) n d n d n d 1 17 10 18 Of the analyses above tabulated, it will be seen that Nos. 1 to 4 inclusive are of the semi-hydraulic type whose value has been noted. Analyses 5 and 6, on the other hand, are representative of the very pure limes used at most slag-cement plants. Burning the lime. As a matter of convenience, and also to reduce freight charges, the limestone is burned near the quarry. The subject of burning the lime requires only brief mention here, as it involves no points of particular interest or novelty. Only two minor details demand notice, as affecting the value of the cement. The first is, that the lime should be burned as thoroughly as possible, for unburned lumps of lime- stone are absolutely valueless to the cement-manufacturer, and must be removed before mixing with the slag. The second point to be noticed is, that the lime should be shipped t$ the cement-plant as soon as pos- sible after it is burned, in order to prevent any considerable proportion of it from air-slaking. Air-slaked particles, while not absolutely inert, are still of little value to the cement. Slaking the lime. The granulated slag as it comes to the mill from the tanks to which it is carried in granulating it carries a very large percentage of water. The amount of water carried will vary in prac- tice at different plants between 25 and 50 per cent as limits. Early in the history of slag-cement manufacture attempts were made to utilize this surplus water. To this end the wet slag was mixed with dry unslaked lime, the expectation being that the water in the slag would serve to slake the lime. In practice, however, it was soon found that this plan was not successful. The lime was only partially and very irregularly slaked, and the mixture was not left in such a con- dition as to be economically handled by the pulverizing machinery. In present-day practice, therefore, the lime is slaked previous to being mixed with the slag. SLAG CEMENT: LIME, MIXING AND GRINDING. 655 Sieving and grinding the lime. If lime has been thoroughly burned and carefully slaked it will all be in the form of a very fine powder, much finer than can be obtained by any economically practicable grinding- machinery. In practice, however, it will be found that after slaking the lime has not all fallen to powder, but still contains a certain pro- portion of hard lumps. The degree of carefulness with which the burn- ing and slaking have been conducted may be roughly judged by observ- ing the relative proportions of lumps and powder. The material remaining as lumps is of three different kinds. First, and in greatest proportion, are fragments of limestone which have not been thoroughly burned in the kiln. Such unburned pieces would be inert if used in the cement. Second, part of the lumps represent fragments of limestone which have been overburned in the kiln and have, therefore, partly clinkered. This is particularly likely to happen if the limestone contained any large proportion of silica or alumina. These partly clinkered lumps, being really poor-grade natural cements, can if pulverized do no particular harm to the slag cement, but on the other hand they cannot do as much good as an equal amount of lime. The third kind of material that may be present in lump form consists of fragments of well-burned lime, which, through accident or carelessness, have not been well slaked. These lumps of quicklime would, if incor- porated in the cement, be actively injurious. The preceding description and discussion of the three classes of mate- rial which are likely to remain as lumps in the slaked lime have been intentionally made detailed in order to point out an error in practice committed occasionally at slag-cement plants. It has been seen that the materials composing these lumps are of such a character as to be either useless or actively injurious if used in a slag cement. It should be obvious, therefore, that the only rational method of treatment is to sieve the slaked lime and to reject entirely all the material failing to pass through the sieve. This is the best practice and the method usually followed. Occasionally, however, urged by a false idea of economy or by inaccurate reasoning, the manufacturer saves the mate- rial failing to pass the sieve, crushes it, and adds it to the cement at a later stage in the manufacture. Proportions of lime and slag. Prost, in consequence of his experi- ments with various proportions of lime, advocated the proportion, to secure the best results, of from 35 to 40 parts of lime to 100 parts of slag. He also stated that the amounts of lime used in actual practice, for each 100 Ibs. of slag were: at Choindez, 40 to 45 Ibs.; at Donjeux, 40 Ibs.; at Brunswick, 33 Ibs., and at Cleveland, 33 Ibs. Mahon, in report- 656 CEMENTS, LIMES, AND PLASTERS. ing his experiments for the Maryland Steel Company, states that the best results were secured by the use of 25 parts of lime to 100 parts of slag, by weight. At another American plant the proportions used are 20 Ibs. lime to 100 Ibs. slag. In the manufacture of slag brick, which is in reality merely a branch of the slag-cement industry, the amount of lime added may fall as law as 10 Ibs. to 100 Ibs. of slag. These rules are, of course, purely empirical; and it is time that some better method of calculating the mixture should be presented. ThL: of course, can be accomplished by the use of the same device which has been previously discussed in connection with" hydraulic limes, natural cements, and Portland cements. Calculating the mixture. If we determine the Cementation Index * of a series of representative American slag cements, such as is given on page 667, we will find that the value obtained ranges from about 1.6 to 1.9 Accepting these values as fairly typical the information thus gained can be employed in devising a method for determining accurately the proportions in which any given slag should be mixed with any given lime in order to secure a good slag cement. Operation 1. Slag. Multiply the percentage of silica in the slag by 2.8, the alumina by 1.1, and the iron oxide by 0.7; add all the products together. From the sum subtract the percentage of lime in the slag plus 1.4 times the magnesia. Call the result "m ". Operation 2. Limestone. Multiply the percentage of silica in the unslaked quicklime by 2.8, the alumina by 1.1, and the iron* oxide by 0.7, and add the products together. Subtract this sum from the total percentage of lime (CaO) plus 1.4 times the magnesia. Call the result "n". Operation 3. Divide lOOXm.by 1.7 Xn. The quotient, -= , will equal the number of parts of quicklime to be used for each 100 parts of slag. The factor by which n is to be multiplied is here taken as 1.7, a very satisfactory value*. Values as low as 1.6 and as high as 1.9 would, however, give the proportions used in practice at various plants. * As previously explained in detail (pp. 170-171), the Cementation Index is the value obtained from the formula (2.8 X percentage silica) + (1. 1 X percentage alumina) Cementation Index = + (0. 7 X percentage iron oxide) ; _^^ (Percentage lime) + (1.4 X percentage magnesia) SLAG CEMENT: LIME, MIXING AND GRINDING. 657 Example. Assume that the two raw materials have the following composition : Slag. Limestone Silica (SiO 2 ) 32.2 1.8 Alumina (A1 2 O 3 ) 12.0 1.2 iron oxide (FeO, Fe 2 O 3 ) 0.6 0.4 Lime (CaO) 48.1 94.0 Magnesia (MgO) 2.3 1.2 Operation 1. Slag. Silica X2. 8 = 32. 2X2. 8= 90.16 Alumina XL 1=12. 0X1.1= 13.20 Iron oxide X0.7 = 0.6X0.7= 0.42 103.78 Lime Xl.O =48.1X1.0= 48.1 Magnesia Xl.4= 2.3X1.4= 3.22 51.32 103. 78-51. 32 =ra = 52. 46 Operation 2. Lime. Silica X2.8= 1.8X2.8= 5.04 Alumina Xl.l= 1.2X1.1= 1.32 Iron oxide X0.7= 0.4X0.7= 0.28 6.64 Lime Xl.O =94.0X1.0 = 94.00 Magnesia Xl.4= 1.2X1.4= 1.68 95.68 95.68-6.64 = n = 89.04 Operation 3. 100 w 100X52.46 5246 [34.6 = parts unslaked quicklime for each I. In 1.7X89.04 151.4 \ 100 parts dry slag. Pulverizing and mixing. The greatest differences in practice exist in the processes for grinding and mixing the slag and lime. The state- ment has been made in several publications that the differences in hardness between dry granulated slag and slaked lime is so great that it is impracticable to pulverize them together in a continuously operated mill. A number of plants, therefore, have installed small discontinuous mills, each of which is charged, locked, operated for a sufficient time to pulverize both constituents of the mixture, and discharged. The dis- advantages of this intermittent system are obvious and it seems especially unfitted for American conditions. The statement that no continu- ously operated mill was able to handle the mixture seemed inherently 658 CEMENTS, LIMES, AND PLASTERS. improbable, in view of the great variety of material successfully handled by the modern ball and tube mills when operated continuously in Port- land-cement practice. Several years ago I referred the question to a leading firm of manufacturers and was informed that nothing in their experience justified the unfavorable conclusion, and that their con- tinuously operated tube mills hacU successfully pulverized mixtures of slag and lime. It seems probable that the most economical prac- tice would be to send the dried slag through a small crusher, Griffin mill, or ball mill, mixing the crushed slag with }ime and completing the mixture and reduction in continuously operated tube mills. What- ever system of reduction is employed, it is necessary that the slag be dried as completely as possible, and, with modern dryers, the amount of moisture in the dried slag can be economically kept well below 1 per cent. In this connection it may be of service to note the results attained in the grinding of basic Bessemer slag (for use as a fertilizer) by the Pottstown Iron Company. A 2000-mm. Jensch ball mill was there employed. This mill consumed about 13 H.P. Its normal output was 20,000 Ibs. in ten hours, though a maximum of 29,000 Ibs. in ten hours had been reached on perfectly dry slag. The fineness of the product was such that 95 to 98 per cent would pass a 100-mesh sieve and 70 to 75 per cent a 150-mesh sieve. A West tube mill in use at an American slag-cement plant grinds 8| barrels per hour of mix to a fineness of 95 per cent through 200-mesh, or 10 barrels per hour to a fineness of 90 per cent. In doing this it uses 67 H.P., equivalent to power consumption of 8 H.P. hours or 6.7 H.P. hours, respectively. Regulation of set. Slag cements will normally set very slowly compared to Portland cements. As this interferes with their use for certain purposes, many attempts have been made by various treat- ments to reduce their set ting- time. There is, unfortunately, another reason why the manufacturer should desire to hasten the set of his product. Most of the slag cements sold in this country masquerade as Portland, and it is desirable to the manufacturer, therefore, to make such of their properties as are brought out in ordinary tests or analyses approximate to those of true Portland cement. The set of slag cements can be hastened by the addition of puzzolanic materials. Of these, burned clay, certain active forms of silica, and slags high in alumina are the cheapest and most generally obtainable. The most important method of regulation is, in this country at least, the Whiting process, which is followed at two large American plants. United States Patent No. 544,706, issued in 1895 to Jasper Whiting, SLAG CEMENT: LIME, MIXING AND GRINDING. 659 covers the use of "caustic soda, potash, sodium chloride, or equiva- lents or any substance of which the latter are ingredients", added either as aqueous solutions or in a dry state at any stage of the process of slag-cement manufacture. In the specifications accompanying the application for this patent, the patentee states that, in the case of dry caustic soda the amount added will vary from 0.125 to 3 per cent, " depending chiefly upon the use for which the cement is intended". The patent was subsequently conveyed to the Illinois Steel Company, and the process covered by it is used by that company in the manu- facture of its "Steel Portland " cement. A license has been issued to the Brier Hill Iron and Coal Company, of Youngstown, Ohio, under which license this company manufactures its "Brier Hill Portland " cement. The process, as practised in the slag-cement plant of the Illinois Steel Company, Chicago, 111., is described as follows: The quicklime used is obtained from the calcination of Marblehead or Bedford lime- stone and carries less than 1 per cent MgO. On its arrival at the mill it is unloaded into bins, beneath which are placed two screens of differ- ent mesh, the coarser at the top. A quantity of lime is drawn upon the upper screen, where it is slaked by means of the addition of water containing a small percentage of caustic soda. As the lime is slaked it falls through the coarse screen onto the finest screen, through which it falls into a conveyor which carries it to a rotary drier. After heat- ing, the resulting slaked and dried lime is carried by elevators to hoppers above the tube mills, where it is mixed in proper proportions with the granulated slag, which has been dried and powdered. General Practice. The general practice followed at a number of American and European slag-cement plants will now be described. A very recent and typical installation is shown in Fig. 163, which gives the plan and elevation of the slag-cement plant of the Stewart Iron Co., at Sharon, Pa. It will be seen that the granulated slag is passed through Ruggles-Coles driers, three of which are in use, and is then elevated to a dry-slag bin on the second floor of the mill. The lime is slaked in an adjoining room, and is also elevated to the second floor. Here the two materials are fed in proper proportions to a screw conveyor, which carries them to a Broughton mixer. The mix is then conveyed to three West tube mills, which deliver the finished product. The Maryland Cement Company,* at Sparrows Point, Md., obtains the slag from the furnaces of the Maryland Steel Company. The slag is dried in Ruggles-Coles driers, and after mixing with the slaked lime * Lewis, F. H. Cement Industry, p. 184. CEMENTS. LIMES, AND PLASTERS. is ground in discontinuous West pebble mills. Mahon's experiments preliminary to the establishment of this plant are discussed on an earlier page. The slag-cement plant of the Illinois Steel Company, Chicago, III., obtains its slag from the blast-furnaces of that company. The speci- FIG. 163. Elevation and plan of Stewart slag-cement plant. (The Iron Age.) fications under which this slag is received, with analyses showing its actual range in composition, will be found on a previous page. After granulation and drying in a specially designed dryer the slag receives its preliminary reduction in Griffin mills. Meanwhile the lime has been slaked as described in detail on a previous page (p. 659), caustic soda being added to regulate the set of the product. The ground slag and SLAG CEMENT: LIME, MIXING AND GRINDING. 661 this prepared lime are then mixed, and the mixture receives its final reduction in Davidsen tube mills. At the plant of the Birmingham Cement Company,* at Ensley, Ala., slag is obtained from the furnaces of the Tennessee Coal and Iron Com- pany, located in near-by towns. The slag is granulated at the fur- naces. Oh arrival at the mills, carrying about 40 per cent of water, it is dried in Ruggles-Goles driers. Two of these, of the A2 style, are in operation. After drying, the slag and slaked lime are fed together to West ball mills, four of which are in use, and the mixture is finally reduced in West tube mills. The Southern Cement Company, at North Birmingham, Ala., dries its slag in a style A2 Ruggles-Coles drier.. The dried slag is crushed in a Kent mill. After mixing with the slaked lime, the final reduction takes place in West tube mills. Two brands of slag cement are mar- keted. One, a normal slag cement, is said to average about CaO 55 per cent, Fe 2 3 , A1 2 3 , 12 per cent, Si0 2 , 27 per cent. The other brand is quicker setting and is said to carry about 10 per cent less CaO and about 10 per cent more SiO 2 . At Skinningrove, England, slags were used of a composition varying between the following limits: Si0 2 , 30 to 32 per cent; CaO, 30 to 33 per cent; A1 2 O 3 , Fe 2 O 3 , 25 to 28 per cent. The slag on issuing from the furnace was run into w ter; ground, before drying, under edge runners, and dried on iron plates in a drying chamber. The dried material was ground under millstones; sieved, and mixed with lime (which had been slaked and screened) in the proportions usually of lime 33 Ibs., slag 100 Ibs. The resulting cement varied in composition between the follow- ing limits: Si0 2 , 24 to 26 per cent; CaO, 45 to 47 per cent; A1 2 3 , Fe 2 O 3 , 20 to 22 per cent. At Vitry, France, the slag is struck by a jet of water immediately upon issuing from the furnace and carried by it into a masonry storage tank. From this tank the granulated slag is elevated and carried to the mill. Five driers of the style shown in Fig. 162 are employed, the dimensions being slightly different from those used at Choindez. After drying the slag is sieved, to remove the coarser particles, passed through six mills of different types, and again sieved. After having been thus reduced to the proper fineness, it is mixed with the slaked lime in ball mills operated discontinuously ; the proportions being about 40 Ibs. of lime to 100 Ibs. of slag. * Eckel, E. C. Engineering News, Jan. 23, 1902. 662 CEMENTS, LIMES, AND * PLASTERS. The slag-cement plant at Konigshof,* Germany, utilizes slag from the Carl-Emil furnaces. A typical analysis of this slag shows : Per Cent. SiO 2 26 . 29 A1 2 3 18.71 FeO .v 1 . 80 CaO '.:' f. 49.16 MgO e 2.45 The more important constituents commonly + vary between the following limits: Per Cent. SiO 2 24 to 27 A1 2 O 3 17 " 19 CaO. 49 ' ' 54 Tht slag is granulated, dried, and ground to such fineness that all passes a sieve with 900 meshes per square centimeter, and 85 per cent passes a sieve of 5000 meshes per square centimeter. The limestone is obtained from quarries at Koneprus, and is burned in continuous shaft kilns. Analysis of the resulting lime shows: Per Cent. SiO 2 12 . 421 A1 2 O 3 2 . 620 Fe 2 O 3 0.883 CaO 81 .546 MgO 1 . 751 CO 2 . 194 Moisture . 425 From this analysis it would seem probable that the lime is itself somewhat hydraulic. It is carefully slaked, and stored until the slaking is complete, after which it is screened to remove the coarser particles. The slag and lime are then mixed and ground together in propor- tions giving a cement of the following typical composition: Per Cent. SiO 2 20 . 81 A1 2 O 3 10.50 Fe 2 3 1.90 CaO 55.90 MgO 1.41 S 0.58 S0 3 0.91 Loss on ignition 3 . 50 * Jour. Iron and Steel Inst., vol. 2, 1900, p. 508. SLAG CEMENT: LIME, MIXING AND GRINDING. 663 The specific gravity of this cement ranges between 2.80 arid 2.90. In all its properties it resembles other slag cements. Slag cement is made at the Cockerill plant * at Seraing, Belgium, from blast-furnace slags ranging within the following limits: Per Cent. SiO 2 27 to 32 A1 2 O 3 12 " 22 CaO 49 " 55 The slag is granulated and dried, the latter taking place at a tem- perature of about 500 C., and requiring a fuel (coke) consumption of about 9 per cent of the weight of slag dried. The slag is ground so as to all pass a sieve of 76 meshes to the inch, and leave a residue of only 8 to 12 per cent on a sieve of 180 meshes to the inch. Grinding to this fineness requires 25 to 30 H.P. for the production of 450 to 800 kilo- grams per hour of powdered slag. Lime is burned, slaked by immer- sion, and stored eight to ten days, at the end of which time it is screened to pass a 76-mesh sieve. It is then mixed with the slag in the propor- tion of 15 to 20 parts of lime to 100 parts slag. Costs of manufacture. Data regarding the cost of manufacture of slag cement have been recently published.! The figures quoted are said to have been the costs of actual manufacture some years ago at the plant of the Maryland Cement Company. They are as follows, being based on a production of 5000 barrels per month: Per Barrel. Mill force, labor and superintendence/ $0 . 160 125 tons of coal at $3.05 per ton .076 3000 bushels of lime at $0.16 per bushel . 100 900 tons of slag at $0.50 per ton .090 Repairs, $100 per month .020 Oil and grease, $40 per month .007 Contingencies 0.011 Cost of administration. . . $0 . 121 $0.585 These figures seems rather high in some respects. For American, plants I should say that the average cost of manufacture should not be over 35 cents per barrel. * Eng. and Min. Jour., vol. 64, pp. 515-516. f Boilleau and Lyon. Cost of making slag cement. Municipal Engineering, vol. 26, p. 321. May 1904. 64 CEMENTS, LIMES, AND PLASTERS. This would be itemized about as follows: TABLE 247. COSTS OF SLAG-CEMENT MANUFACTURE PER BARREL. Min. Max. Slag 04 .10 Lime >'..... .^ 07 .12 Coal 03 .08 Oil, grease, waste, etc 005 .01 Repairs 01 .03 Labor U)5 .08 Superintendence, testing, etc 03 .05 .23| .47 Several American plants have to my knowledge worked quite 2, 22 to 30 per cent; A1 2 O 3 + Fe 2 O 3 , 11 to 16 per c#nt; CaO, 49 to 52 per cent; MgO, less than 4 per cent; S, less than 1.5 per cent; ignition loss, 2.5 to 7.5 per cent. TABLE 250. ANALYSES OF EUROPEAN SLAG CEMENTS. Components. Bruns- wick, Germany Choindez, Switzer- land. Bilbao, Spain. Donjeux. Saulnes, France. SiCX, 25 56 19 5 30 56 23 85 24 85 24 55 22 45 ALOo 11 20 17 5 13 31 13 95 12 10 14 05 13 95 FeO 25 1 10 3 85 1 85 3 30 CaO 49 70 54 45 01 51 40 49 20 49 25 51 10 MgO . 2 96 1 95 1 75 1 60 1.35 S .. *4 63 1 30 SO 3 fl 41 45 1 35 0.60 0.35 Loss on ignition 7.05 5.65 7.75 7.50 *CaS. t CaSO 4 . Physical Properties. Specific gravity. The specific gravity of slag cements usually ranges from 2.7 to 2.9, as compared with the 3.15 which may be considered a fair average for the specific gravity of a good Portland cement. The slag cements are, therefore, appreciably lighter than Portlands, and more bulk is obtained for the same weight. The following determi- nations of the specific gravity of three American slag cements have been made at Philadelphia: Toltec 2.861 Climax. 2 . 888 Penn 2.831 Aside from its use as a method of distinguishing slag cements from Portlands, the determination of the specific gravity of the cement is of little engineering importance. A point which is of engineering impor- tance, however, appears to have been overlooked by experimenters. Sa far as the writer knows, the relative specific gravities of set briquettes, composed of neat-slag cement and neat-Portland cement respectively, have never been determined. A knowledge of the two values would be of service, at times, in selecting the type of cement to be used. For SLAG CEMENTS: COMPOSITION AND PROPERTIES. some purposes, as in dams, a heavy material is preferable; for others, as in floors, the lighter cement would be better. Color of slag cements. Slag cements can usually be distinguished from Portlands by being much lighter in color and slightly different' 140 210 AGE IN DAYS 860 FIG. 164.* Effect on tensile strength of slag cements of hardening in air or in* water. (Tetmajer.) 140 210 AGE IN DAYS 2SO 350 PIG. 165.* Effect on compressive strength of slag cement of hardening in air or in water. (Tetmajer.) in tint, while from most natural cements they differ markedly in tint. They are commonly bluish-white to lilac, the exact color of any speci- men depending partly on the respective colors of the lime and th& slag which have been used in its manufacture, but more largely on. * From Johnson's " Materials of Construction ", p. 576. 670 CEMENTS, LIMES, AXD PLASTERS. the relative proportions in which these ingredients have been mixed. Slag cements do not stain masonry; and an imported cement of closely related origin (Meier's Pozzuolan) has long been in favor in this country for architectural uses, because of this non-staining property. Rapidity of set. Normally slag cements are slower setting than Portlands. Whether this, property is a disadvantage or not will depend on the use to which the cement is -to be applied. As before mentioned the rapidity of set increases naturally with the amount of alumina in the slag. Set can be artificially hastened by the addition of puzzo- lanic material to the cement; burned clay, active forms of silica, slags high in alumina, etc., are additions which are both effective and cheap. The treatment of the cement during manufacture with alkalies to accel- erate the set has already been discussed. Strength. While slag cements fall below high-grade Portlands in tensile strength, good American slag cements develop sufficient strength to pass the usual specifications for Portlands. Tested neat they do not approximate so closely to the Portlands as they do if tested in 2 :1 or 3:1 mortars. Part of this property may be due to the fact that they are in general ground finer than Portlands, especially than foreign Portlands. Prof. W. K. Hatt recently made a large series of tests on American slag cements, and reported that there was no noticeable deficiency in strength of briquettes kept in air as compared with those kept in water. Other investigators have arrived at opposite conclu- sions; and it is probable that these conflicting results arise from differ- ences in the chemical composition of the various brands tested. Resistance to mechanical wear. Slag cements are notably deficient in this property, and are therefore not available for use for the surface of pavement, floors, etc., where this quality must be highly developed; they seem to be well fitted, however, for pavement foundations, or .indeed for any w r ork which will not be exposed to dry air, and in which a high strength is not necessary. Ratio of tensile to compressive strength. This ratio, which is of importance (as noted in the discussion of Portland cements) seems to be much lower for slag cements than for Portlands. In the case tabu- lated below, the results of tests show the ratio for slag cement to aver- age 5.3:1, in place of the 10:1 ratio, which is a fair average for Port- land cements. , ,. compressive strength , The average value for ~ -, for the whole series tensile strength is 5.3. SLAG CEMENTS: COMPOS ITIOX AND PROPERTIES. 671 TABLE 251. TENSILE vs. COMPRESSIVE STRENGTH OF SLAG CEMENTS. Mixture. Test. 7 Days. 28 Days. 7 Days. 28 Days. 7 Days. 1 Month. Neat cement Tension 441 528 480 503 Compression tf-i-T 2054.5 4.66 2470 5.15 2830 5.63 1 cement, 3 sand. . Tension 170 219 145 200 171 243 <. a Compression 486.5 933 938 1138 1529 0.+ T 2.86 6.44 4.69 6.65 6.29 List of references on properties and testing of slag cements. In addition to the list given below, many of the papers cited on pages 664-665 will be found to contain data on the properties and testing of slag cements. Bonnami, H. Fabrication et controle des chaux hydrauliques et des ciments. 8vo, 276 pp. Paris, 1888. Candlot, C. Ciments et chaux hydrauliques. 8vo. Paris, 1889. Datienne, H. Manufacture and properties of slag cement. Revue univer- selle des mines, Sept., 1897. Elbers, A. D. Notes on the manufacture and properties of blast-furnace slag cement. Eng. and Mining Journal, vol. 64, pp. 515-516. 1897. Hatt, W. K.' American slag cements. 21st Ann. Rep. Proceedings Indiana Engineering Soc., pp. 45-65. 1901. Also in Engineering News, March 7, 1901. Le Chatelier, H. Tests of hydraulic materials. Trans. Am. Inst. Min. Eng., vol. 22, pp. 3-52. 1894. Mahon, R. W. Slag-cement experiments. Journal Franklin Institute, vol. 137, pp. 184-190. 1894. May, E. Slag cement : its production and properties. Stahl und Eisen, March and April, 1898. Abstract in Iron Age, Sept. 1, 1898. Prost, M. A. Note sur la fabrication et les proprietes des ciments de laitier. Annales des .Mines, 8th series, vol. 16, pp. 158-208. 1889. Redgrave, G. R. v Manufacture and properties of slag cement. Proc. Institu- tion Civil Engineers, vol. 105, pp. 215-230. 1891. Rohland, P. Influence of catalysers on velocity of hydration of cements. plasters, and limes. Zeitschrift anorg. Chemie, vol. 31, pp. 437-444. Abstract in Journal Soc. Chem. Industry, vol. 21, p. 1233. 1901. U. S. Army Board of Engineers. Report on Steel Portland Cement. Svo, 112 pp. Washington, 1900. Whiting, J. The definition of Portland cement. Engineering Record, July 30, 1898. Anon. The distinction between slag and Portland cements. Engineering Record, July 9, 1898. 672 CEMENTS, LIMES, AND PLASTERS. Specifications for slag cement. So far as known, the only American specifications for slag cement are those prepared* and published in 1902 by the Engineer Corps, U. S. Army. These are reprinted below. SPECIFICATIONS FOR P.UZZOLAN CEMENT. 1 . v* Engineer Corps, U. S. A., 1902. (1) The cement shall be a Puzzolan of uniform^ quality, finely and freshly ground, dry, and free from lumps, made by grinding together without subsequent calcination granulated blast-furnace slag with slaked lime. (2) The cement shall be put up in strong sound barrels well lined with paper, so as to be reasonably protected against moisture, or in stout cloth or canvas sacks. Each package shall be plainly labeled with the name of the brand and of the manufacturer. Any package broken or containing damaged cement may be rejected, or accepted as a fractional package, at the option of the United States agent in local charge. (3) Bidders will state the brand of cement which they propose to furnish. The right is reserved to reject a tender for any brand which has not given satisfaction in use under climatic or other conditions of exposure of at least equal severity to those of the work proposed, and for any brand from cement works that do not make and test the slag used in the cement. (4) Tenders will be received only from manufacturers or their authorized agents. (The following paragraph will be substituted for paragraphs 3 and 4 above when cement is to be furnished and placed by the contractor: No cement will be allowed to be used except established brands of high-grade Puzzolan cement which have been in successful use under similar climatic conditions to those of the proposed work and which come from cement works that make the slag used in the cement.) (5) The average weight per barrel shall not be less than 330 Ibs. not. Four sacks shall contain 1 barrel of cement. If the weight as determined by test weighings is found to be below 330 Ibs. per barrel, the cement may be rejected, or, at the option of the engineer officer in charge, the contractor may be required to supply, free of cost to the United States, an additional amount of cement equal to the shortage. (6) Tests may be made of the fineness, specific gravity, soundness, time of setting, and tensile strength of the cement. DEPARTMENT OF CIVIL ENGINEERING --., SLAG CEMENTS: COMPOSITION AND PROPERTIES. 673 (7) Fineness. Ninety-seven per cent of the cement must pass through a sieve made of No. 40 wire, Stubb's gauge, having 10,000 openings per square inch. (8) Specific gravity. The specific gravity of the cement, as deter- mined from a sample which has been carefully dried, shall be between 2.7 and 2.8. (9) Soundness. To test the soundness of cement, pats of neat cement mixed for five minutes with 18 per cent of water by weight shall be made on glass, each pat about 3 inches in diameter and \ inch thick at the center, tapering thence to a thin edge. The pats are to be kept under wet cloths until finally set, when they are to be placed in fresh water. They should not show distortion or cracks at the end of twenty-eight days. (10) Time of setting. The cement shall not acquire its initial set in less than forty-five minutes and shall acquire its final set in ten hours. The pats made to test the soundness may be used in determining the time of setting. The cement is considered to have acquired its initial set when the pat will bear, without being appreciably indented, a wire -^ inch in diameter loaded to \ Ib. weight. The final set has been acquired when the pat will bear, without being appreciably indented, a wire ^ inch in diameter loaded to 1 Ib. weight. (11) Tensile strength. Briquettes made of neat cement, after being kept in air under a wet cloth for twenty-four hours and the balance of the time in water, shall develop tensile strengths per square inch as follows : After seven days, 350 Ibs.; after twenty-eight days, 500 Ibs. Briquettes made of one part cement and three parts standard sand by weight shall develop tensile strength per square inch as follows: After seven days, 140 Ibs.; after twenty-eight days, 220 Ibs. (12) The highest result from each set of briquettes made at any one time is to be considered the governing test. Any cement not show- ing an increase of strength in the twenty-eight-day tests over the seven- day tests will be rejected. (13) When making briquettes neat cement will be mixed with 18 per cent of water by weight, and sand and cement with 10 per cent of water by weight. After being thoroughly mixed and worked for five minutes the cement or mortar will be placed in the briquette mould in four equal layers and each layer rammed and compressed by thirty blows of a soft brass or copper rammer, f of an inch in diameter or T \ of an inch square, with rounded corners, weighing 1 Ib. It is to be allowed to drop on the mixture from a height of about half an inch. 674 CEMENTS, LIMES, AND PLASTERS. When the ramniing has been completed the surplus cement shall be struck off and the final layer smoothed with a trowel held almost hori- zontal and drawn back with sufficient pressure to make its edge follow the surface of the mould. (14) The above are to be considered the minimum requirements. Unless a cement has been recently .used on a work under this office, bidders will deliver a sample barrertor test before the opening of bids. If this sample shows higher tests than those given above, the average of tests made on subsequent shipments must come up to those found with the sample. (15) A cement may be rejected in case it fails to meet a,ny of the above requirements. An agent of the contractor may be present at the making of the tests, or, in case of the failure of any of them, they may be repeated in his presence. If the contractor so desires the engi- neer officer in charge may, if he deems it to the interest of the United States, have any or all of the tests made or repeated at some recog- nized testing laboratory in the manner herein specified, all expenses of such tests to be paid by the contractor. All such tests shall be made on samples furnished by the engineer officer from cement actually delivered to him. CHAPTER XLV. SLAG BRICKS AND SLAG BLOCKS.* UNDER the names of "slag brick ", "slag tile ", "slag block ", "scoria brick ", etc., two very different products have been included by various writers. Both products are made from blast-furnace slag, but the two classes differ so greatly in their methods of manufacture and properties that it seems necessary to describe them separately. This has accord- ingly been done, the names "slag bricks " and "slag blocks " being supplied to the respective classes. As here used, the term "slag brick " will be confined to those bricks, tiles, etc., which are made by mixing slaked lime with ground slag, molding the mixture by hand or in a brick-machine, and drying or steaming the product. The term "slag blocks ", on the other hand, will be applied to the products made by pouring molten slag into brick -shaped molds. Slag Bricks. The structural products included in this chapter under the head of "slag bricks " include those which are made by mixing granulated slag with slaked lime or with slag cement, molding the mixture in a brick-press or by hand, and drying it in the, air, with or without the use of steam. It will be noted that all the raw materials used in this industry are the same as those utilized in the manufacture of slag cement; and indeed the manufacture of slag bricks may be considered as being merely a specialized phase of the slag-cement industry. Though the slag-cement industry of the United States is in a fairly satisfactory condition no serious attempt seems to have been made to prepare slag bricks, tile, pipes, etc., on a commercial scale. Small amounts of slag bricks have been made for use about the mills and furnaces * Over half of the material contained in this chapter is reprinted, by courtesy of Engineering News, from an article by the present writer published in its issue of April 30, 1903. 675 676 CEMENTS, LIMES, AND PLASTERS. and for the local market, but apparently no attempt has been made to extend the manufacture. Methods of manufacture. The slags used are basic blast-furnace slags, but a somewhat greater range in composition is allowable for slag bricks than when the slags are to be used in cement-manu- facture. The analyses quoted in the* present chapter may be regarded as fairly representative of the class of slags used in slag-brick manufac- ture. It will be seen that the silica ranges from 22.5 per cent to 35 per cent; the alumina and iron oxide together, from 1^6.1 per cent to 21 per cent; the lime, from 40 per cent to 51.5 per cent. As in slag cements, sulphur is an objectionable constituent. Much of it, fortunately, is removed during the process of granulating the slag. The general steps in slag-brick manufacture may be stated as follows; Slags of proper composition are granulated by being run into a stream of cold water immediately upon issuing from the furnace. This causes the slag to break up into little porous particles, thereby greatly reducing the expense of subsequent grinding: Granulation also confers hydraulie properties on the slag, and removes part or all of its contained sulphur. The granulated slag is dried and pulverized. Powdered slaked lime is added in sufficient uantity to bring the total calcium oxide content of the mixture up to about 55 per cent. This mixing, as well as the previous burning and slaking of the lime, must be carefully and thor- oughly done in order to prevent subsequent disintegration of the bricks. Usually, during or after the mixing, a small amount of water is added. The mixture is then molded into shape, either by hand or in a brick- machine. After shaping, the bricks are dried in the open air, this usu- ally taking six to ten days in dry weather. In the best practice, the bricks are retained for several months, after drying, in order that they may be well hardened before marketing. Though over 90 per cent of the total production of slag brick is at plants following the above methods, three other methods may be briefly noted. At a few plants the granulated slag is mixed, without drying, with the unslaked lime; the slag furnishing sufficient water to slake the lime. Slaking in this way is very imperfectly done, however, and the practice should never be followed if high-grade bricks are expected. At a few other plants, notably at the Bilbao plant described below, slag is mixed with slag cement instead of with lime. At certain English plants, also noted below, the slag bricks are hardened in steam cylinders like the cylinders used in lime-sand brick manufacture. Slag bricks vary in color from a grayish white to dark gray. They weigh less than clay bricks of equal size, are said to require less mortar SLAG BRICKS AND SLAG BLOCKS. 677 in laying up, and are at least equal to clay bricks in crushing strength. The product usually seems to find a ready market, though, of course, the low value of the material, relative to its bulk and weight, precludes long railroad transportation. Methods at special plants. Slag bricks were manufactured at the Cleveland Slag Works, Middlesborough, England; but the manufac- ture has been discontinued for some years At this plant the wet granu- lated slag was mixed with "selenitic lime" (see Ch. XV) instead of with common lime. The selenitic lime was composed of 80 per cent quicklime, 10 per cent gypsum, and 10 per cent iron oxide. About 670 Ibs. of this selenitic lime was sufficient for 1000 bricks. The mix- ture of slag and lime was pressed to shape in a brick-press; and the bricks were stacked in sheds for a week, to harden enough to handle well. After this they were stacked in the open air for five or six weeks more, when they were ready for use. The bricks were dull-gray in color, and very hard and tough. Buildings onstructed of them over twenty years ago are still in a good state of preservation. The manu- facture of slag bricks at these works was given up for reasons not con- nected with the technical value of the product, which seems to have acquired an excellent local reputation At Vitry, France, the manufacture * of slag bricks and pipes is carried on in connection with the manufacture of slag cement. The bricks are made by mixing 60 parts of slaked lime with from 250 to 300 parts of granulated slag. Sufficient water is added to this mix- ture to make a firm paste, from which the bricks are molded in hand- or steam-presses. These^ slags are found to be especially useful for foundations or basement work, pavements, etc. "Facing brick " are made from a similar mixture, with the addition of some fine sand. Sewer pipes are made from a mixture consisting of 500 kgs. of slag cement and 1 cu. m. of sand. This mixture is made into a stiff mortar, and forced into steel molds by iron rammers. The molds are removed as soon as the ramming has finished. The pipes are then dried for three days, after which they are immersed in water for twenty- four hours. They are then stacked in the factory ground for several months, after which they are ready for market. Slag bricks are manufactured f on a large scale at Kralovedvoor, near Prague, Bohemia. The slags normally used at this plant vary in composition within the following limits; * Engineering News, Jan. 1, 1897. f Engineering and Mining Journal, April 16, 1898. 678 CEMENTS, LIMES, AND PLASTERS. Per Cent. Silica (SiO 2 ) 25.8 to 27.0 Alumina (A1,O 3 ) 17.3 " 19.3 Iron oxide ( FeO) 1.5" 1.7 Manganese oxide (MnO) 0.0" 0.1 Lime (CaO) 51 . 4 " 51 . 5 Magnesia (MgO) y ^ 0.4" 25 Sulphur (S) ; 1.3" 1.8 As the slag issues from the furnaces it is run into an inclined iron trough in which cold water is flowing. In addition*. to granulating the slag, a considerable portion of the sulphur is removed in this way. The granulated slag is run into tanks, from which it is carried to the mixing floor, as required, by conveyors. Here the slag is dumped into mixers, together with thoroughly slaked lime in a pasty condition. The lime is obtained by the calcination of a limestone of the following range in composition : Per Cent. Silica (SiO 2 ) 0.2 to 0.6 Alumina ( A1 2 O 3 ) ^02" 08 Iron oxide (Fe 2 O 3 ) / Lime carbonate (CaCO 3 ) 97.0 " 98.4 Magnesium carbonate (MgCO 3 ) 0.9" 1.9 The mixture is then molded into shape under pressure in a brick- machine with a capacity of 1000 "bricks per hour. These bricks are taken to the drying house, where they remain about eight days, at the end of which time they are sufficiently hard to stand transportation. In the size usually made, the dry bricks weigh about 4.75 kgs. each, and will stand a pressure of 18 kgs. per square centimeter. In color they vary from nearly white to grayish. Cement and mortar adhere to them as well as to clay bricks. Occasionally bricks are made at this plant from slags of the follow- ing average composition, derived from the smelting of a manganiferous ore different from that commonly used at these furnaces: Per Cent Silica (SiO 2 ) 33 .00 Alumina (A1 2 O 3 ) 18.67 Iron oxide (FeO) 1 .00 Manganese oxide (MnO) 4 . 25 Lime (CaO). , 40.00 Magnesia (MgO) 2.33 Sulphur (S) 1 .33 Bricks made from this slag are dark colored, owing to the compara- tively large percentage of manganese present. More lime must be used, SLAG BRICKS AND SLAG BLOCKS. 679 in proportion to the slag, and the bricks made from this slag require a longer time to dry and harden than is needed by those made from the ordinary slag. Slag bricks are made at Ekaterinoslav,* Russia, from blast-furnace slags showing the following range in composition: Per Cent Silica (SiO 2 ) 22.5 to 35.0 Alumina (A1 2 O 3 ) 14.0 " 15.0 Iron oxide (Fe 2 O 3 ) 1.1" 3.3 Manganese oxide (MnO) 0.0" 0.3 Lime (CaO) 45.0 " 51.0 Magnesia (MgO) tr. " 1.4 Sulphur (S) 0.3" 0.4 Loss on heating 2.3" 7.5 The slag is granulated, sieved on a revolving screen, dried, and ground in a ball mill. Lime is slaked and sieved. Enough of this slaked-lime powder is added to the slag to bring the lime (CaO) con- tent of the mixture up to about 55 per cent. With slags of the range in composition above indicated, this would require the mixture to con- sist of from 5 to 12 parts of lime to 100 parts of slag. The mixing is carried on in a screw mixer, and the powdered mix is then pressed into brick in a dry press. On issuing from this press the bricks are set aside to harden, and at the end of six days are usually hard enough for use. Their tensile strength is about 312 Ibs. per square inch; and the crush- ing strength varies from 1250 to 5600 Ibs. per square inch; both, of course, increasing with age. The bricks are gray in color, well shaped, weigh less than stone, and require little mortar in laying up. They withstand temperature changes well, and are particularly well adapted for use in damp situations or under water. Toldt has described, the manufacture of slag bricks at Bilbao, Spain, where the blast-furnace slag from the Vizcaya furnaces is used. Slag cement is made by mixing three parts, of granulated and dried slag with one part powdered slaked lime, and grinding this mixture in a ball mill. The bricks are then made by mixing one part by volume of this cement with four parts of wet granulated slag, and pressing this mixture into shape in a brick-press. A Belgian form of press with twelve molds is used. This turns out twenty bricks per minute, with thirteen workmen. It will be noted that in a slag brick made in this fashion the strength of the brick must be almost entirely derived from the slag cement used in the mixture, for the uncrushed slag will be almost inert. * Engineering and Mining Journal, 1896. 680 CEMENTS, LIMES, AND PLASTERS. Hardening in steam-cyLnders. A new method of slag-brick manu- facture has recently been introduced * in Eng and. In this process the use of lime is dispensed with (except when slags carrying less than 35 per cent CaO are used), while a hardening cylinder is employed exactly as in the manufacture of lime-sand brick (see pp. 136-140). The slag is allowed to cool normally; it is then broken up and fed to an edge-runner mill, where it is crushed and ground, and falling thence into a deep pit under the mill, it is collected by an elevator and thrown on a 10-mesh screen. All capable of passing this goes to the mixer, the coarser particles being rejected and returned to the mill for further grinding. "The ground slag is moistened in the mixer with from 5 to 10 per cent of water, and is then delivered by the mixer into the brick- making machine, where it is molded into bricks under great pressure, the pressure employed being from 100 to 150 tons on each brick. As the bricks are made they are stacked onto steel platform-wagons made to carry from 700 to 800 bricks. The loaded wagons are allowed to stand for twelve hours, to allow the bricks to take a slight initial set, after which they are run into the steel chamber, and the bricks are here subjected to the action of steam at a pressure of from 105 to 120 Ibs. per square inch. Ten hours under this treatment is sufficient to harden the bricks and render them on withdrawal ready for building purposes. " It is necessary that the machinery employed should be of a very strong and durable character. For effecting the grinding an edge-run- ner mill is most suitable, as it is not easily put out of order by the iron which is often found in the slag in large pieces. The roller rims and false bottom should be of steel, preferably manganese steel; and the perforated grate should also be of steel . the rollers should be made of a suitable weight, depending upon the hardness of the slag generally from three to five tons each. Their width should not exceed 12 inches. " A specially designed brick-making machine is employed. This con- sists of a rotating table containing the molds, a feeding-pan, and power- ful toggle-press. As the table revolves, the molds pass alternately under the feeding-pan where they are fed with the charge of material, then under the press, and a further rotation brings the mold over the ejecting plunger and the brick is discharged ready for removal. The machine is capable of exerting a pressure on each brick of 150 tons, and is fitted with a simple contrivance to insure the corners of the bricks being well pressed up. Its operation is first to give the material in the mold at * Sutcliffe, E. R. Utilization of blast-surface slag. Amer. Mfr. and Iron World, vol. 74, pp. 555-563, May 5, 1904. SLAG BRICKS AND SLAG BLOCKS. 681 a top pressure by means of a wedge-shaped plunger, forcing the mate- rial well into the sides and corners of the mold, and a final pressure from below, which completely presses out the indentation made by the wedge-shaped plunger and gives a good finish to the sides and cor- ners of the brick. The necessity of this arrangement will be apparent when it is understood that ground slag does not become plastic under pressure as does ordinary clay, and that a material when filled into a mold by gravity naturally piles in the center, and if directly pressed would produce bricks of greater density in the center than at the sides. "The hardening-chamber s like a boiler without flues, 45 ft. long by 6 ft. in diameter. In contains 6000 bricks, and must be capable of withstanding the pressure of steam, which is used for their indurating. One end of the chamber is removable and held in place by hinged bolts threaded on to a back ring, the joint being made by a projection on the cover fitting into a recess in the shell, the bottom of the recess being filled with ordinary red rope packing. The chamber will per- mit of two steamings per day one during the daytime and the other at night. Hence each chamber with high-pressure steam serves for 12,000 bricks per day. " The brick wagons must be strongly constructed, as any deflection of the platform might tend to crack and spoil the bricks, which in the green state require some care in handling. It is necessary that roller or ball bearings be used for the axles, as under the action of the steam any oil or grease would be burnt out of ordinary bearings. " It will be noted that no binding material- whatever is mixed with the slag. The process is really the production of a concrete. In grind- ing the slag fine enough to pass a 10 per inch mesh a very large pro- portion of it is reduced to a fine dust, which acts as a hydraulic cement, the coarser particles forming the aggregate. Where the slag is very hard, and consequently only a small proportion of dust is produced, it is necessary to reduce a portion in a ball mill or other suitable fine-grind- ing machine. The precise action which takes place during the harden- ing is difficult to determine; but evidently the result is due to a com- bination being effected between the free lime found in all limy slags and the silica and alumina. " It may be assumed that the silicious compounds in the slag become soluble in the presence of heat and moisture, in which condition it is readily attacked by the free lime present in the slag. " With some slags high-pressure steam gives better results than low- pressure, besides requiring less time to effect the hardening. In speak- 6S2 CEMENTS, LIMES, AND PLASTERS. ing of high-pressure steaming, it is to be understood that this refers to any pressure above the atmosphere and low pressure to at or under this. With other slags low pressure is quite as effectual as, and in some instances is better than, high pressure. To determine which is the most suitable process is a question for experiment with the particular slag. Where low-pressure steaming' is adopted the chambers may either be made of thin sheet steel or tunnels may be constructed of brickwork. In the author's experiments a steel high-pressure chamber was used steaming up to 150 Ibs. pressure per square inch a*nd for low pressure a chamber constructed of brickwork. In general, for low-pressure steam- ing for about forty hours, and for high-pressure steaming twenty hours, will be found most suitable and convenient. The author has not formed a definite opinion as to what element in the slag causes the different effect in the action of high- and of low-pressure steam, but is inclined to think that it is principally due to the proportion of sulphur in the slag. During the steaming some sulphur is driven out of the bricks, and the final hardening does not seem to be completed until this vola- tile or unfixed sulphur is driven out or combined. It is probable that the sulphide of calcium present is slowly being split up, the hydrogen of the water combining with the sulphur forming sulphuretted hydro- gen, and the oxygen with the calcium* forming lime. By subjecting the slag to steam, thus keeping it moist and hot at the same time, this action is accelerated. " Generally slags high in sulphur^can be hardened best under pro- longed low-pressure steam, and in one or two instances no hardening effect was produced by high-pressure steam, whereas low-pressure steam produced the desired effect. From this it would seem that the chemical action is only accelerated up to a certain temperature, and that at a higher temperature a different effect is produced; or it may be that at a higher temperature the action is too violent, causing an expansion and separation of the particles without actually producing cracks or disintegration of the bricks, but sufficient to prevent the final com- bination. Seemingly the presence of this unfixed sulphur retards the action of the lime on the silicates and aluminates, and only when it is finally driven off can the full combination be effected. " In the case of a slag which falls to powder on exposure to the atmos- phere a grinding-mill is unnecessary; and with some slags of this character it is only necessary to moisten and then press it into bricks and harden as before. Again, with others it would be necessary to grind a portion to dust in a ball mill before the setting could be obtained. In the former case the slag powder would consist of a fine dust mixed SLAG BRICKS AND SLAG BLOCKS. 683 with coarser particles, and in the latter case it would be like a fine even- grained sand without any really fine dust. " The slag used for brickmaking should preferably be new; but it has been found that a slag which had been exposed for twenty years still possessed setting properties when acted upon by steam. One of the bricks exhibited was made in the summer of 1901 from slag which had been exposed to the atmosphere for over twenty years. " In the case of a slag which disintegrates on exposure to the atmos- phere, it would not be wise to use it directly after it has cooled unless the ground-moistened slag is permitted to stand until the free lime is thoroughly hydrated. This could be effected in silos erected directly over the brickmaking machine; and twenty-four hours in this condi- tion would be sufficient. In general the better plan would be to allow the slag to stand for about ten days before being used, as in such cases the grinding would be facilitated by the disintegrating. " The slag for brickmaking should preferably be cast in thin layers capable of being easily broken up in sizes suitable for being passed into the grinding-mill, rendering a stone-breaker unnecessary. " The bricks are almost perfect in form, there being no twisting or distortion produced by the induration, and in strength and other qualities they will compare with the best qualities of clay bricks. TABLE 252. CRUSHING STRENGTH OF INDURATED SLAG BRICKS. No. of Specimens. Dimensions in Inches. Cracked at Tons per Sq.Ft Crushed at Tons per Sq.Ft Remarks. 1 9X4fX2 227 340 1 9 X 4f X 2} 303 375 Not completely crushed 1 9X4f X2 227 375 Not completely crushed 1 9 X 4f X 2 1 246 370 " Objections have been raised against granulated slag bricks on account of their porosity, which ranges in some cases as high as 15 per cent. No objections of this kind can be raised against these indurated slag bricks, the absorption being remarkably low, as shown in Table 253. " The bricks before testing were thoroughly dried at 212 and then immersed for twenty-four hours. TABLE 253. POROSITY OF SLAG BRICKS. No. Dimensions in Inches. Weight Before Immersion. Weight After Immersion. Gain in Weight. Pef Cent. 1 2 9 X 4f X 2^ 9X4fX2| 8 Ibs. 10 oz. 8 Ibs. 12 oz. 9 Ibs. 1 oz. 9 Ibs. 4 oz. 7 OZ. 8 oz. 5.07 5.71 684 CEMENTS, LIMES, AND PLASTERS. " These bricks were tested by burning them in an ordinary continu- ous brick kiln, and a brick treated in this way withstood the fire suc- cessfully and is still a good hard brick, the only change being in the color, which is now a light buff. During the burning the loss in weight averaged 6 ounces, which equals 4.47 per cent; and the absorp- tion after burning was 16- ounces, or 11.9 per cent, after ^twenty-f our hours' immersion. " The following is the estimated cost of production, based on a pro- duction of 10,000 bricks per day of ten hours: Labor. s. d. 1 man at grinding-mill at 6d. per hour 5 2 men at brickmaking machine, taking off, at 6d. per hour 10 1 youth attending to moistening of material 3 6 4 wheelers and stackers at 6d. per hour 1 1 foreman. . 060 246 Cost in labor per 1000 bricks, say 4s. 6d. " To this must be added the cost of getting the slag to the machinery, wear and tear, depreciation, and such charges as may be added for power and steam. 11 As regards the slag, this should be run from the furnaces on to a level floor and then broken up and taken to the machinery. This will mean a little extra cost over that of running the slag in wagons and depositing it in balls on the slag heap; but 6d. per ton should cover the whole cost of casting the slag in this way and running it to the machinery. Wear and tear on machinery will necessarily be high, considering the wearing action of the slag. This will be well provided for at Is. 6d. per 1000 bricks. " As to power and steam, this would be generated from the furnace gases, and if not used for this purpose might be considered as wasted; but assuming this at the value of coal, if such were used we should require 2| tons of coal per day, which, at 10s. per ton, works out at 2s. 6d. per 1000 bricks. If we allow 6d. per 1000 for generation, we get 3s. per 1000 bricks. " The cost of a complete plant such as described would be about 3000, including buildings and all requirements. Taking depreciation at 7J per cent on the whole, and reckoning on 300 working days, we get 15s. per day, or Is. 6d. per 1000 bricks. SLAG BRICKS AND SLAG BLOCKS. 685 SUMMARY OF COST OF PRODUCTION PER 1000 BRICKS. s. d. Cost of labor 4 6 Slag at 6d. per ton (4 tons) 2 Wear and tear 1 6 Power and steam 3 Depreciation 1 6 Oils, sundries, etc 6 Total cost of production 13 " The above calculation is based on only producing 10,000 bricks per day. The plant would be capable of making up to 12,000 per day; so that by only calculating on this reduced output sufficient allow- ance is made for unforeseen losses. If a larger plant were installed the cost could be very considerably reduced; on a plant producing, say, 20,000 bricks per day, the cost per 1000 should not be more than 10s. to 11s. "This refers more particularly to limy slags; but in the case of slags not so rich in lime, hydrated lime can be added to the ground slag and the hardening effected in the same way, but in such cases the cost of production is increased by the cost of this added lime. " As before pointed out, most limy slags have setting properties with- out being steamed; and with slags containing from 40 to 48 per cent of lime, bricks may be made by merely grinding and pressing the mate- rial and permitting the bricks to stand out in the open air, the same conditions being observed as in making granulated slag bricks, but this method is not so satisfactory as the hardening by steam. In many slags there is a proportion of soluble salts which tend to spoil the bricks when allowed to harden naturally by appearing as efflorescence. This in some cases is so violent that the outer crust will be forced away from the brick; but the same effect does not happen when they are steamed, the steaming either turning the salts into a stable compound or driving them off. " These bricks will withstand the weather equally with a high-class clay brick. Bricks have been exposed to the weather the whole winter and no effect whatever is noticeable upon them. They have been soaked, then frozen, and afterwards put into hot water without deterioration." Slag Blocks. Under this heading will be considered all these products ("slag blocks", "slag tiles", etc.) made by running molten slag, direct from the furnaces, into molds of proper shape. The term slag block will be employed as a general but distinctive name for this class of products 686 CEMENTS, LIMES, AND PLASTERS. in order to distinguish them from the slag bricks made by mixing granu- lated slag with slaked lime, which have been discussed in previous sec- tions of this chapter. Slag blocks, if properly made, are stronger than slag bricks. They are, however, impervious to air and moisture; and on that account are not good building materials, for dwellings constructed of them are apt to be damp and unhealthful. : Their chief uses are for foundations or for paving blocks, for the latter of which they are particularly well adapted. Many smelters and furnaces have made small amounts of slag blocks for local use. For the past thirty years or so a considerable quantity have been made in the Lehigh iron district of Pennsylvania, their earliest recorded use being in the slag-block pavements laid in Philadelphia about 1876. The properties required in a slag block to be used for paving work are density, resistance to abrasion, toughness, and roughness of surface. These properties are found to vary with the chemical composition of the slag, the rapidity with which the slag is allowed to cool, and the character of the moulds used. By properly varying the last two factors, slags of almost any composition can be utilized in this industry. The three requisite properties first mentioned i.e., density, resist- ance to abrasion, and toughness vary directly with the rate of cooling, the slowly cooled blocks being the best. Blocks cast in sand molds and heavily covered with loose sand, cool very slowly, and give very much better results than those cast in iron molds. Slowness in cooling, however, requires much greater storage space than if rapid cooling is practiced; and casting in sand molds demands a higher grade of work- manship than casting "n iron molds. The roughness of surface or non-slipperiness of blocks intended for paving use is highly important, especially as slipperiness has been the chief defect charged against slag blocks, which defect is prevalent in blocks cast in iron moulds. In English practice it has been overcome by casting the block in a double size mold, having a projection inside which results in a notch on the slag block. The block is, after coating, split apart at this notch, and the rough fracture-surface of each half is laid uppermost in paving. This method of avoiding slipperiness adds considerable to the labor cost of the blocks, and is therefore not well adapted to American practice. Slag blocks cast on a sand bed are free from the defect noted (slipperiness); or at least it can be avoided if sufficiently coarse sand be used. Slag blocks manufactured by the Tees Scoria Brick Company, of SLAG BRICKS AND SLAG BLOCKS. 687 Middle.sborough, England, have been somewhat extensively employed * as street pavement in Rotterdam. Holland clay bricks, limestone blocks, and porphyry bricks are employed in the same city, and will be useful for comparison with the slag blocks. The foundation, in all cases, is simply a bed of sand, carefully packed. The thickness or depth of pavement laid on this varies, according to the paving mate- rial, as follows :" Clay bricks, 4J inches; slag blocks, 5 inches; limestone or porphyry blocks, 6 inches. The cost of material and laying per square yard is: Clay bricks, 62^ cents; limestone blocks, 62^ cents to $1.25; slag blocks, $1.25; porphyry blocks, $1.56. No data as to proportions of each pavement in use, or durability of the different types, are available. The adjunct director of public works of Rotterdam stated that for light traffic the clay-brick pavements were regarded as the best; for medium traffic, slag blocks or limestone; for heavy traffic, porphyry blocks. The manufacture of slag blocks from copper slags at Mansfeldt, Saxony, has been described f in detail by Egleston. The industry, as carried on in this locality, presents certain features of interest which warrant a somewhat lengthy abstract of the paper cited. The slags used are high in silica, ranging from 40 per cent to 60 per cent. When cooled rapidly, they form a dark colored brittle glass, but if cooled with great slowness the product becomes gray and crystal- line. These slowly cooled slags arc both hard and tough, and there- fore serviceable in the manufacture of structural hiaterial. The process employed at Krug Hutte is as follows: The slag, as it comes from the furnaces, is carried in slag wagons to the molding ground, where the bricks are cast. The bed of the molds is sand, which has been sieved to remove coarse particles. The bed is then carefully gone over with a shovel, which is pressed into it an inch or so to make the sand soft. It is then smoothed over with the shovel, and into the corners a piece of iron QOL.8*m..to 0.20 m. long, and 0.15 m. wide is laid, inclined so as to facilitate the passage of the slag in the slag-runners which go round the whole space. The molding-bed io then so divided by iron partitions pushed down into the sand as to give the size of blocks required. These partitions have several round holes, about 0.05 m. in diameter, near their tops, to permit the entrance of slag. Previous to use, the partitions are washed with clay and sprinkled with sand to prevent the slag from * Streets and Highways in Foreign Countries. Special Consular Report, vol. 3, p. 190, Washington, 1897. f School of Mines Quarterly, vol. 12, pp. 112-117. 688 CEMENTS, LIMES, AND PLASTERS. sticking to them. Around each of the molds thus formed is a space 0.20 m. wide, through which the slag flows. When all the partition's are in place, the bottom of each mold is made flat by pressing down into it a piece of sheet iron (of the same size as the compartment), attached to a handle. When the molds are ready slag is* brought from the furnace in slag wagons, and allowed to flow through the interspaces and into the molds. When the slag has about half filled a mold, a little sand is thrown on it to prevent too rapid cooling. When the molds are entirely full, they are covered with about a foot of sand and allowed to stand for forty-eight hours. At the end of this time the slag is cool, the sand is shoveled off, and the iron partitions removed. During rainy weather the molding ground is kept covered with boards until the slag is ready for pouring, and as soon as this operation is finished the molds are again covered with boards. The blocks are usually cubes, 0.15 m. on the edge, though larger sizes and different shapes are occasionally cast. The material which has solidified in the spaces between the molds is broken up for use as road metal. At Koch Hutte similar processes are employed. Large blocks, however, are cast in cast-iron molds, with a cover that is shut down in order to compress the slag. Similar work is carried on at Kupfer- kammer Hutte. Analyses of typical slags from the Mansfeldt district are given in Table 251. A very interesting example of the manufacture of slag blocks or tiles from a copper blast-furnace slag has been described * by Braden as having been seen in operation at a furnace located near Santiago de Chile. His description is as follows: The slag and matte are tapped from the blast-furnace into a slag- pot. After settling for a few moments the slag is poured from ladles into molds which are 6 inches square and 1 inch deep. The molds after being filled with slag are placed on a hearth which has a movable cover, and covers are placed on the molds as well as on the hearth. A very light heat is kept up, so that the slag cools very slowly. When it appears black the molds are lifted from the hearth and the slag tiles are dumped into cold water. The tiles thus made are very light and portable. When laid they have proven to be tough and durable. For this manufacture a slag carrying a considerable excess of iron * Trans. Am. Inst. Min. Engrs., vol. 26, pp. 52-53. 1896. SLAG BRICKS AND SLAG BLOCKS. TABLE 254. ANALYSES OF SLAG, MANSFELDT. Krug Hutte. Koch Hutte. Eckhardt Hutte. Kupferkammer Hutte. Year 1888 Per Cent. 18.35 6.732 14.825 4.725 0.697 0.063 1.165 0.232 0.289 47.63 1888 Per Cent. 23.187 2.22 17.001 4.643 0.328 tr. 0.692 0.118 0.277 48.465 1888 Per Cent. 21.51 0.847 16.525 2.768 0.744 tr. 0.934 tr. 0.3 46.39 1881 Per Cent. 19.29 3.23 16.35 10.75 1.26 0.75 48.22 1881 Per Cent. 20.29 4.37 15.67 8.73 l.ll 0.67 50.0 Magnesia Alumina ...... Iron oxide Zinc oxide Lead oxide Copper oxide Silica Fluor Total 97.708 96.931 90.018 99.85 100.84 Kupferkammer Hutte. Sangerhausen Hutte. 1881 Per Cent. 19.50 8.02 18.17 5.89 1888 Per Cent. 19.15 3.677 17.636 7.213 0.827 0.038 2.056 0.065 0.333 46.81 1881 Per Cent. 33.10 1.67 4.43 4.37 0.25 53.83 2.09 1881 Per Cent. 23.40 0.87 7.83 7.47 .30 57.43 1.97 Lime Alumina Iron oxide Manganese oxide Nickel and cobalt oxides Zinc oxide 3.57 Lead oxide Copper oxide 0.23 48.38 0.99 Silica Fluor Total 99.75 97.802 99.74 99.27 has been preferred. The tiles are sold for from $30 to $60 (pesos Chilenos) per thousand. Slag blocks have been manufactured at a Montana copper smelter by a process which contrasts strongly with the practice at Mansfeldt and Santiago. The copper slag was poured into iron molds built up by putting together iron plates of proper form. The process was carried on in the open air and no covering of any kind was placed on the blocks. The slag in consequence cooled very rapidly. Though the product was, therefore, not as dense or tough as that secured at Mansfeldt, the Montana practice effected a great saving of time and space. INDEX. Aalborg kiln, recommended for natural cement, 225 used for lime-burning, 101-103 Portland cement, 474 Abrasion, resistance to, by slag-cements, 670 Absorption-, of clay bricks, 145 lime-sand bricks, 142, 143, 144, 145 sandstones, 142, 146 Accelerators for plasters, 50, 64, 66 slag-cement, 659 Acidity index, definition of, 299 , see also Silica-alumina ratio Adhesive strength of plasters, 62 Alabaster, 15 Algfr, aid in marl deposition, 338 Alit, in Portland-cement clinker, 568, 571, 573 Alkalies, effect in Portland cement, 310, 388, 391 in flue-dust of cement-kilns, 510 limestones, 310 used as accelerator for slag cement, 659 flux in Portland cement, 391 Alkali waste, analyses of, 349 , used as Portland-cement material, 348-350 Alum, used in manufacturing Keene's cement, 76 Alumina bricks, for kiln-linings, 490 Alumina, clays high in, 491 , effect of, in Portland-cement mixtures, 298, 299, 387, 573 in sea-water, 602 silica-alumina ratio in clays, 354 limestones, 313 Portland cement, 299 Ammonia process, see Alkali waste. Analyses of alkali waste, 349 alumina brick for kiln-linings, 491 anthracite ash, 396 arenes, 637 ash of coke and coal, 396 691 692 INDEX. Analyses of ash, volcanic, 633-636 brick for kiln -linings, 491, 492 calcined magnesite, 155 cement, grappier, 185 , natural, 253-262, 286 , Portland, 397, 575, 577-579 , puzzolan, 662, 667, 66 , slag, 662, 667, 668 "cement" plaster, 57 chalks, 321 clays for kiln-brick, 491. 492 clays for Portland cement, 325, 350, 355-357, 359, 365, 390, 397 coal ash, 396 coal ash for rotary kilns, 506, 508, 512, 513 coke ash, 396 fire-brick for kiln-linings, 491, 492 flint pebbles for tube mills, 465 flue-dust, 510 fuel-ash, 393 gas, natural, 523 , producer, 525 , waste from rotary kilns, 507 gas-coke ash, 396 grappiers, 182, 185 grappier cements, 185 gypsite, 54 gypsum earth, 54 gypsum used in making Keene's cement, 77 plasters, 53, 54 Portland cement, 545 hard-finish plasters, 77 high-alumina clays for kiln-brick, 491 high-calcium limes, 116 hydraulic lime, 175, 179, 182, 188 hydraulic lime used in slag cement, 653, 654, 662 hydraulic lime-rock, 175, 176, 177, 187 Keene's cement, 77 kiln-brick, 491, 492 kiln-coals, 506, 508, 512, 513 kiln -gases, 507 Lafarge cement, 185 lean limes, 118 Lehigh cement rock, 326, 329 lime, 95, 116, 118, 119 hydraulic, 175, 179, 182, 188 used in slag cement, 653, 654, 662 .lime, used in slag cement, 653, 654, 662, 678 lime-sand brick, 142 limestones, magnesian, 122, 157, 321 INDEX. 693 Analyses of limestones, used for making hydraulic lime, 175-177, 187 lime, 122 magnesia, 157 natural cement, 204-206, 208-211, 213, 215, 217 Portland cement, 314, 321, 326, 329, 333, 342 , see also Chalk, Marl, magnesia, 155, 157, 158 magnesia brick, 161 magnesian limes, 119 magnesian limestones, 157, 327 magnesite, 152, 153 calcined, 155 marls, 342, 390, 397 natural cement, American, 253, 254, 255, 256, 257, 258, 259, 260, 286 , Austrian, 262 , Belgian, 261 , English, 261 , French, 261 , German, 262 natural-cement rock, American, 204, 205, 206, 208, 209, 210, 211, 213 , Belgian, 215 , English, 217 natural gas, 523 "natural Portland" cement, Belgium, 261 rock, Belgian, 215 old Portland cements, 575 oyster-shell lime, 95 oyster-shells, 95 Parker's cement, see Natural cements, pebbles, flint, 465 plaster, used in Portland cement, 545 plasters, 44, 57 Portland cement, American, 397, 577, 578, 579 , old, 575 Portland-cement mixtures, 394, 397, 506 pozzuolana, 633, 634, 635 producer-gas, 525 puzzolan cements, 662, 667, 668 puzzolanic materials, 633 to 638, 643 Roman cements, see Natural cements, santorin, 636 shales, see Clays, shell-lime, 95 shells, 95 slag cement, 662, 667, 668 slags, average blast-furnace, 351 , for Portland cement, 352 694 INDEX. Analyses of slags, for slag blocks, 689 s!a~ bricks, 678, 679 slag cement, 643, 662, 663 slate, roofing, 364 , used for Portland cement, 365 stack-gases, 507 stucco, 57 '-' . ^ tetin, 635 tosca, 635 trass, 636 tube-mill pebbles, 465 volcanic ash, 633-636 waste gases from kiln-stack, 507 water from slag granulation, 648 Analytical methods, 576-582 Anhydrite, 15 Anthracite ash, analysis of, 396 Ash of fuel, analysis of, 396 Atlas cooling system for clinker, 529 Atlas plant, costs at, 559 Atomic weights of elements, table of, 11 Austria, natural cements of, analyses, 262 Ball mills, 425, 452-457 , Bonnot, 452 , Gates, 452 , Jensch, 658 , Krupp, 455 , Smidth, 516 Basic slags, 641 Belit, in Portland-cement clinker, 569 Berthelet separator, 245 Blake crushers, 24, 238, 430 see also Crushers. Blatchley, W. S., on intermittent lime-kilns, 99, 100 origin of marl deposits, 337 Bleininger, A. V., on fineness of natural cements, 247 Portland cement, 585 Portland-cement mix, 425 Blocks, slag, 685-689 Blood as retarder for plaster, 64 Boilleau and Lyon, on costs of Portland cement, 558 slag cement, 663 Bonnot ball mill, 452 Bonnot tube mill, 457 Borax used in hard-finish plasters, 32 Breakers, rock, see Crushers. Brick, alumina, for kiln-linings, 490-492 magnesia, 160-162 INDEX. 695 Brick, sand-lime, 130-147 slag, 675-685 Brick-press, for sand-lime brick, 138 slag brick, 678, 679, 680 Brigham, S, T., on strength of hydrated lime, 128 Brines, as, sources of magnesia, 158, 159 British thermal unit defined, 12 Broughton mixers, 50 Buhrstones, 35, 239-243, 437 Burners, for coal, 486-489 natural gas, 489, 490 Burning, see Kilns, Fuels. Cactus, used as plaster-retarder, 65 Calcination, see Calciners, Fuels, Kilns. Calcined plaster, 32 Calciners, kettle, for plasters, 37-46 , oven, for plasters, 37 , rotary, for plasters, 46-50 Calcium carbonate, see Limestones. Calcium chloride as retarder for cement, 546, 54? Calcium oxide, see Lime. Calcium hydrate, see Lime, hydrated. Calcium hydroxide, see Lime, hydrated. Calorie denned, 12 Campbell, E. D,, on fineness of cement grinding, 424 Campbell kiln for natural cement, 225-228 Campbell lime-hydrater, 126 Canada, gypsum deposits, 25 magnesite deposits, 152 Candlot, E., on dome kilns, 471 Aalborg kilns, 475 composition of fuel-ash, 396 Carbon dioxide, in limestones, 97 by-product from lime-kilns, 108 from magnesite, 155 Carbonate of calcium, see Limestone. Carbonate cementing materials, 5-7 Carbonate of lime, see Limestone. Carbonate of magnesium, see Magnesite. Carbonic acid, see Carbon dioxide. Carpenter, R. C,, on tests of rotary kilns, 508-510 effects of plaster, 541 lime chloride, 546 Caustic-soda waste as Portland-cement material, 348-350 Caustic soda, used in slag cement, 659 Celit, in Portland -cement clinker, 568, 571, 573 Cementation index; calculation of, 170 explanation of, 170, 171 696 INDEX. Cementation index; hydraulic limes, 170-173, 176, 180, 182, 187 natural cements, 196-201, 214 Portland cements, 391-393, 397 slag cements, 656-667 Cementing materials, classification of, 2-10 , production in U. S., 2 " Cement" plasters, analyses of, -'57 ^ definition of, 32 manufacture of, 31-55 properties of, 56-67 Chamber-kiln for lime and cement, 106-108, 472-474 . Charcoal, as rotary kiln-fuel, 526 Chemical analyses, see Analyses. Chemical analysis, methods of, 576-582 Chemical compounds, table of, 12 Chemical elements, table of, 11 Chert, see Flint. Clark pulverizer, 243 Classification of cementing materials, 2-10 Classification of crushing machinery, 429-430 Clays; analyses of, 355, 356, 357, 359, 491, 492 origin of, 353 used for kiln-brick, 490-492 used in Portland cement manufacture, 353-363 Clinker cooling, 527-530 Clinker grinding, 530-534 Closson process, 158 Coal for kilns, analyses of, 513 , cost of preparation, 520 , crushing and pulverizing, 514-520 , distribution of, 513 , drying of, 515 , explosion and fire risks, 520 , treatment of, 514-522 Coke ash, analysis of, 396 Complex cementing materials, 6 Composition, chemical, see Analyses. Compressive strength; of clay bricks, 145 hydraulic limes, 183, 189, 190 lime-sand bricks, 142-146 natural cements, 274-275 plasters, 61-62 Portland cement, 588-592 puzzolan cements, 670-671 sand cement, 594, 596, 597 sandstones, 146 selenitic limes, 192-193 slag bricks, 678, 679, 683 silica cement, 594, 596, 597 INDEX. 697 Compress! ve strength; of slag cements, 670-671 Constitution of hydraulic limes, 172-176 natural cements, 195-199, 223 plasters, 31-32 Portland cement, 566-575 slag cement, 667 Cost of burning lime, 106, 108, 109-112 natural cement, 248-250 plasters, 52 Portland cement, 558-561 Cost of dredging marl, 379-381 Cost of drying cement materials, 378 coal, 520 slag, 649, 663, 664 Cost of erecting Portland-cement plant, 555-557 Cost of excavating cement rock, 219-221, 378-379 clay, 373 gypsum, 29, 30, 52 limestone, 379 , marl, 379-381 natural-cement rock, 219-221 Portland-cement materials, 378-381 shale with steam-shovel, 373 Cost of fuel, see Cost of burning. Cost of labor in lime plants, 106, 108, 110-112 natural-cement plants, 248-250 plaster plants, 52 Portland-cement plants, 558-559 Cost of land and quarries, Portland cement, 555 Costs, of manufacturing cement plaster, 52 hydrated lime, 128 lime, 106, 108, 110-112 lime-sand brick, 140-142 natural cement, 248-250 oxychloride cements, 167 plaster of Paris, 52 Portland cement, 554-561 slag bricks, 684-685 slag cement, 663-664 Sorel stone, 167 Cost of mining, see Cost of excavating. Cost of plant for hydrated lime, 128 lime-sand brick, 140 Portland cement, 555-557 slag brick, 684 . Cost of preparing coal for kiln use, 520 Cost of quarrying, see Cost of excavating. Crackers, 35, 239, 430 Creighton, Prof., on strength of natural cements, 276 698 INDEX. Crushers, Blake, 34, 238, 430 , cone-grinders, 34, 238, 430-433 , crackers, 35, 239, 430 , definition of group, 429-430 , Gates, 34, 238, 431-433 , gyratory, 34, 238, 431-433 , jaw, 34, 238, 430 ? , McEntee, 239, 430 , Mosser, 430 , nippers, 34 , rotary, 238, 430-433 , Sturtevant, 125, 238 , used for coal, 514 gypsum, 34 lime, 125 natural cement, 221, 238 plaster, 34 Portland cement, 429-433, 530 Crushing machinery, classification of, 429-430 for coal, 514-520 gypsum, 34-37 lime, 124-125 natural cement, 236-247 plaster, 34-37 Portland-cement clinker, 422-468, 530-535 Portland-cement materials, 404-410,414,417-418,422-468 slag cement, 657-658 slaked lime, 124-125 see also Crushing practice, Crushers, Ball mills, Tube mills, Kominuters, Rolls, Mills, Mill-stones, Edge-runners, Crackers. Crushing practice, general discussion, 422-428 types of machinery used, 238, 429-430 use of separators, 245, 426-428 see also Crushing machinery,. Cummer calciner for plaster, 46-48 Cummings mill, 241-243 Cyclone pulverizer, 238 Cylinders, hardening, for lime-sand brick, 139-140 slag brick, 680-682 Davidsen tube mill, 457-462 see also Tube mills. Davis, C. A., on origin of marl deposits, 337, 338, 340 Deval, L., on effect of alumina, 602 Dietzsch kiln, 474 Dodge process for lime-hydrating, 125 Dolomite, composition of, 6, 90-91 see also Limestones, Magnesian, INDEX. 699 Dome kiln, for lime, 99 Portland cement, 470-471 Dredging, cost of, 379-381 marl, 357-376, 379-381 Dryers, Edison, 402-403 , for slag, 649-652 limestone, shale, etc., 400-404 , Hoist, 651-652 , rotary, 400-401, 649-652 , Ruggles-Coles, 401, 649-651 , tower, 402-403 , Vitry, 652 , Cummer, see Calciners, rotary. Drying Portland-cement materials, 378, 399-404 Dr^y-pans, 437-438 Dyckerhoff, Prof., on effect of plaster, 539, 541, 543 Edge-runner mills, 243, 437-438 Edison drier, 401 kiln, 484 rolls, 435-437 Eldred process of lime-hydrating, 125 Elements, chemical, table of, 11 Emery mill, 35, 239-241 England, natural cements of, 217, 261 Estrichgips, 68-76 Examination of chalk deposits, 321 clay deposits, 360 limestone deposits, 315-318, 321 marl deposits, 343-346 shale deposits, 360 Excavation of raw materials, see Costs, Dredging, Mining, Quarrying. Felit, in Portland-cement clinker, 569 Ferric oxide, formula, 12 , see qlso Iron oxide. Ferrous oxide, formula, 12 , see also Iron oxide. Fiber-machine, for wall-plaster, 51 Fineness, of coal for kiln fuel, 515 marl, 340 natural cement, 236-237, 246-247, 282 Portland cement, 531, 584-585, 606 Portland-cement mixture, 423-425 plasters, 58-59 raw materials for Portland cement, 423-425 sand cement, 592-593 sand used in lime-sand brick, 134-135 700 INDEX. Fineness, of slag cement, 657-658, 673 Fire-brick, for kiln-linings, 490-492 Flint, in limestones, 91, 309 pebbles for tube mills, 461-462, 463-465 Flooring-plaster, 32, 68-76 France, flint pebbles from, 463-465 natural cements of, analyses, 261 ^ Fuel consumption; in hydraulic-lime kilns, 177 lime-kilns, 98-99, 101, 103, 106 magnesia-kiln, 154-155 natural-cement manufacture, 233-235 *- plaster manufacture, 52 Portland-cement manufacture, 492-496, 497-511 rotary dryers, 403, 651 Fuels, preparation of coal for rotary kiln, 514-520 suggested use of charcoal in rotary kiln, 526 used in Portland-cement manufacture, 512-526 use of coal in rotary kiln, 486-489, 512-522 use of producer gas in rotary kiln, 489-490, 524-526 use of natural gas in rotary kiln, 489-490, 522-524 use of oil in rotary kiln, 522 see also Fuel consumption. Gas, natural, in rotary kiln, 489-490, 522-524 , producer, in rotary kiln, 489-490, 524-526 Gates ball mill, 452-454 crusher, 34, 238, 425 tube mill, 457 Germany, natural cements of, analyses, 262 plaster manufacture in, 49-50, 68-76 Gillmore, Q. A., on Hoffmann kilns, 106-107 Sorel stone, 163-167 crackers, 239 Glue, used as plaster-retarder, 64 Grant, on tests of selenitic limes, 191-193 Granulating slag, 644-648 Grappier cements, analyses of, 185 definition of, 185 manufacture of, 185 properties of, 185-186 Grappiers, analyses of, 182, 185 definition of, 180, 185 Greenland, flint pebbles from, 463-465 Greensand marls, 335 Griffin mill, 243, 404, 425, 440-443, 516 Grimsley, G. P., on setting of plasters, 64 costs of plaster manufacture, 52 Gypsite, see Gypsum earth. Gypsum, analyses of, 53, 54 INDEX. 701 Gypsum, composition of, 14 , distribution of, 16-28 earth, analyses, 54 , character, 15 , methods of excavating, 30 , excavation of, 28-30 , origin of, 15 , physical properties, 15 , specific gravity, 15 , used in natural cement, 264 plaster manufacture, 14-30 Portland cement, 534-545 , varieties, 15 Hair, used in wall-plaster, 51 Hair-picker, 51 Hale, D. J., on prospecting marl deposits, 344-346 Hardening gypsum, methods for, 67 Hardening-cylinders for lime-sand brick, 138-140 slag brick, 680-685 Hard-finish "cements," see Hard-finish plaster. Hard-finish plasters, analyses, 77 , definition, 32 , manufacture, 76-78 , properties, 78 Harris system of pumping marl, 376-377 Hauenschild kiln, 475-477 Heat consumption, see Fuel consumption. Heat losses in rotary kiln, 502-511 Heat requirements in burning lime, 98-99 Portland cement, 497-501 Heat units, definition of, 12 Helbig, A. B., on heat used in rotary kiln, 505, 509 Hoffmann kiln for burning lime, 106-108 Portland cement, 472-474 Huennekes system, lime-sand brick, 143 Huntingdon mills, 438-440 Hurry and Seaman " blast-furnace " methods, 415-416 Hydrate cementing materials, definition, 4 Hydrated lime, cost of installation for, 128 , methods of making, 124-129 , packing weights, 128 , with Portland cement, 129 Hydraulic index, defects of, 169 , explanation of, 1 C9 of various cementing materials, 169 , use in classification, 169 Hydraulic limes, analyses, 179, 188 , classification of, 173 702 INDEX. Hydraulic limes, definition, 172 , manufacture, 174-182 , properties, 182-184, 188-190 , specific gravity, 182 , used in slag cements, 653-654' "Improved" natural cement, 259, 263, 268 j> Index, acidity, 299 , cementation, see Cementation index. , hydraulic, sec Hydraulic index. Iron disulphide, see Pyrite. Iron oxide, in limestones, 91, 310 Portland cement, 383, 388, 569-570, 573-574 Jaw crushers, see Crushers. Jensch ball mill, used in grinding basic slag, 658 Johnson kiln for Portland celnent, 471-472 Keene's cement, analyses, 77 , manufacture, 76-77 , properties, 78 Kent mill, 443-447 Kettles, calcining, for plasters, 38-46 Keystone lime-kiln, 104-106 Kilns, Aalborg, 101, 225, 474-475 , Campbell, 225-228 , chamber, 106-108, 472-474 , Dietzsch, 474 ., dome, 99-101, 225, 470-471 , Edison, 484 for hydraulic limes, 177-178 for lime-burning, 99-108 for natural cements, 225-233 for Portland cements, 469-496 , Hauenschild, 475-477 , Hoffmann, 106-108, 472-474 , intermittent, 99-101, 470-471 , Johnson, 471-472 , Keystone, 104-106 , O'Connell, 106-107 , Ransome (rotary), 480-496 , ring, 106-108, 472-474 , rotary, 480-498 , Schofer, 101,225,474-475 , Schwarz, 477-478 Kirkwood gas-burner for cement-kilns, 489-490 Kominuter, description, 448-452 , used for natural cement, 243, 246 , used for Portland cement, 448-452 INDEX 703 Krupp ball mills, 455-457 Krupp tube mills, 462-463 Labor, cost of, see Costs of labor. Lafarge cement, 185-186 Leblanc process, see Alkali waste. Le Chatelier, H., on constitution of hydraulic limes, 174-175 Portland cement, 569-570, 573, 574 effect of sea-water on cement, 601 expansion of magnesia brick, 161-162 setting properties of Portland cement, 572 Lewis, F. H., on effects of plaster, 537, 538, 544 Lignite used for producer-gas, 524-526 Lime, effects in Portland cement, 384-386, 496, 566-575 Lime, hydrated, composition of, 119 , manufacture of, 124-127 , methods of lime-slaking, 120-122, 124-127 , physical properties of, 128 Lime of Teil, see Hydraulic limes. Lime, slaked, see Limes, hydrated. L*ime carbonate, see Limestones. Lime chloride, used as cement-retarder, 546-547 Limes, analyses, 116, 118, 119 , classification of, 97-98 , composition, 115-119 , costs of manufacture, 106, 108, 109-112 , fuel consumption in burning, 98-99, 101, 103, 106 , groups of, 97-98 , hydraulic, see Hydraulic limes. . kilns used for, 99-108 , magnesian vs. high-calcium, 115 , methods of manufacture, 98-109 , physical properties, 120-123 , raw materials, 88-95 , statistics of production, 112-114 Limestones, analyses of, 157, 175-177, 187, 203-213, 215, 217 , composition of, 6, 89 , cost of excavation, 109, 219-221, 378-379 , distribution of, 91-94 , excavating, 109, 219-221, 367-375 , impurities of, 91, 309-310 , magnesian, 6, 90-91, 157 , mining, 219-221, 375 , modules of, see Sepatria. , origin of, 88-89 , properties of, 310-311 , quarrying, 109, 219-221, 367-375 , used for hydraulic lime, 175-177, 187 , used for lime, 88-95 704 IXDEX. Limestones, used for natural cement, 200-222 , used for Portland cement, 307-347 , varieties of, 89 , water contained in, 378 see also Chalk, Marl. Lindhard kominuter. 243, 246, 448-452 Linings for rotary kilns, 490-492- , ^ tube mills, 459 Mack's cement, 78 Magnesia, analyses of, 155, 158 , carbonate of, see Magnesite. , chemical formula of, 97, 148 , preparation of, 148-159 , in limestones, 90, 157 , in limes, 115, 118-119 , in natural cements, 197, 200 , in Portland cements, 298, 386, 631 Magnesia bricks, analyses, 161 , manufacture, 160-161 , properties, 161-162 Magnesian limestones, see Limestone, Magnesiao, Magnesite, analyses of, 152-153 , burning of, 154-155 , composition of, 148-149 , distribution of, 149-152 , imports of, 153 , origin of, 149 , production of, 153 Magnesium chloride, as source of magnesia, 158-159 use in Sorel stone, etc., 162-167 Mahon, R. W., on slags suitable for slag cement, 642 Mannheim system of calcining plaster, 49-50 Marble, 89 Marl, analyses of, 342 , composition of, 341-343 , definition of, 334 , distribution of, 339-340 , dredging methods and costs, 375-377, 379-381 , drying, 378 , examination of deposits, 343-347 , greensand, 335 , origin of, 335-339 , physical properties, 340-341 , pumping, 376-377 , water contained in, 343, 378 , weight, 340-341 Marston, Prof., on tests of plasters, 60-64 McEntee cracker, 239 INDEX. 705 McKenna, C,, on properties of Lafarge cement, 185-186 Mill, Cummings, 241-243 , emery, 35, 125, 239-241 , Griffin, 243, 404, 425, 440-443, 516 , Huntingdon, 404, 425, 438-440 , Kent, 443-447 , Sturtevant, 35, 125, 239-241 , Williams, 244-245, 466-468, 516 see also Crushing machinery. Millstones, 125, 239-243, 437 Mills, G. S., on tests of lime, 122 Mining gypsum, 29 limestone, 219-221, 375, 379 natural-cement rock, 219-221 see also Costs of excavation. Mixer, Broughton, 50 Natural cements, analyses, 253-262 , compressive strength, 274 , compressive-tensile ratio, 275-276 , cost of manufacture, 248-251 , definition of, 195 , effect of gypsum or plaster, 264 , effect of heat on strength, 275 , effect of salt, 267 , fineness, 236-237, 246-247 , history, 285-289 , methods of manufacture, 223-247 , modulus of elasticity, 277 , packing weights, 248 , physical properties, 262-277 , rapidity of set, 263 , raw materials for, 200-222 , statistics, 289-292 , tensile strength, 267, 283 Natural gas in rotary kilns, 489-490, 522-524 "Natural Portland" cements, 214-217 Newberry, S. B., on change in composition during burning, 396-397 constitution of Portland cement, 567, 569, 570, 574 formula for cement mixtures, 391-393 heat used in kiln, 505, 509 Kent mill, 445 Nihoul and Dufossez, on effects, pf plaster, 537, 539, 540 Nippers, used in grinding gypsum, 34-35 O'Connell lime-kiln, 106-107 Oil used in rotary kilns, 522 Organic matter, as retarder for plasters, 65 , in marls, 341, 343 706 INDEX. Ovens, used in plaster manufacture, 37 Oxychloride cements, 9, 162-167 Packing weights of hydrated lime, 128 natural cements, 248 plasters, 52 . Portland cements, 548-553 slag cements, 672 Parian "cement", 33 Parker's cement, 217 Pebbles for tube mills, 463-465 Peppel, S. V., on lime-sand brick, 134-136, 140, 142 Petroleum in rotary kilns, 522 Phosphorus, effects of, in Portland-cement mixtures, 389-390, 673 Plaster, accelerators for, 50, 64-68 , adhesive strength, 62-63 t analyses, 57 , classification, 32 , compressive strength, 61-62 , cost of manufacture, 52 , fineness, 58-59 , groups of, 32, 56 ., imports of, 79, 86 , manufacturing methods, 33-55 , packing weights, 52 , physical properties, 57-67 , production of, 79-87 , raw materials for, 14-30 , retarders for, 50, 64-68 , specific gravity, 57 , statistics of production, 79-87 , tensile strength, 58-61 , used in natural cement, 264 Portland cement, 534-545 , weight per cubic foot, 57 Plaster of Paris, definition, 32 , see also Plaster. Porosity, see Absorption. Portland cement, analyses of, 575, 577-579 , analytical methods, 604 , cementation index of, 398-399 , compressive strength, 588-589 , compressive tensile ratio, 589-591 , constitution of, 382-391, 566-582 , costs of manufacture, 554-561 , definition of, 297 , effect of freezing, 598-601 gypsum, 534-545 heat, 597-598 INDEX. 707 Portland cement, effect of plaster, 534-545 salt, 598-601 sea-water, 602-603 , fineness, 584-585, 606 , methods of analysis, 604 , methods of manufacture, 398-565 , modulus of elasticity, 591-592 , origin of names, 294-295 ' , packing weights, 548-553 , physical properties, 582-613 , production of, 561-565 , raw materials for, 300-381 , specific gravity, 385-386, 605-606 , specifications for, 614-631 , tensile strength, 586-588, 612 , use of gypsum or plaster, 535-545 Potash, see Alkalies. Prospecting, see Examination. Prost, S., on effects of granulating slag, 648 slags suitable for slag cement, 641-642 Pulverized coal as kiln-fuel, 486-489, 512-522 Pulverizer, Clark, 243 , Cyclone, 238 , Raymond, 466 Pulverizing machinery, see Crushing machinery. Purington, C. W., on use of steam-shovels, 373 Puzzolan cements, definition, 9, 632 , raw materials for, 632-640 , see also Slag cement. Pyrite in limestone, 91, 310, 388 see also Sulphur, Sulphides. Quarrying clays, 370-375 gypsum, 28, 30 limestone, 109, 219-221, 367-375, 378-379 natural-cement rock, 219-221 Portland-cement material, 367-375, 378-379 shales, 370-375 Ransome kiln, see Kiln, rotary. Ratio between compressive and tensile strength: natural cement, 275-276 Portland cement, 589-591 slag cement, 670-671 silica and alumina, see Alumina, Index. Raymond pulverizer, 466 Retarders for plasters, 50, 64-66 Portland cement, 534-547 Richards, J. W., on tests of rotary kiln, 506, 507, 509 Richards, R. H., on costs of crushing, 433 70S INDEX. Richardson, C., on constitution of Portland cement, 567-575 phosphorus in cement mixture, 389 specific gravity of natural cement, 263 Ring kiln, 106-108, 472-474 Rock-crushers, see Crushers. Rock-emery mill, 35, 125, 239-241 Rock excavation, see Excavation. Rohland, on set of plasters, 64 Roman cements, 214, 217 Roofing slate, see Slate. Rotary calciner, see Calciners. drier, see Driers. kiln, see Kiln. Sabin, L. C., on properties of lime, 123 natural cement, 263-265, 273, 276 Portland cement, 539, 540. 600 Salt, brines as sources of magnesia, 158-159 , effect on natural cement, 267 Portland cement 598-601 Sampling, marl deposits, 343-346 see also Examination. Sand cement, 592-597 Sawdust, used as plaster-retarder, 64 Schiebler process, 157 Schofer kiln for lime, 101 Portland cement, 473 Schwarz kiln, 477-479 Schwarz process for lime-sand brick, 136-137 Scott's cement, 190-193 Seasoning clinker, 235-236 Sea-water, as source of magnesia, 158-159 , effect on Portland cement, 602-603 Selenite, 15 Selenitic lime, manufacture, 190-191 , properties, 191-193 Separators, Berthelet system, 244-246 , in coal-pulverizer, 520 natural-cement plants, 244-246 Portland-cement plants, 426-428 Septaria, 217 Shales, analyses of, 357-379 , excavating, 370-374 , origin, 353 Shell-lime, analyses of, 95 Shells, analyses of, 95 , in marl deposits, 339 , used for lime-burning, 94-95 Shovel, steam, use of, 370-374 INDEX. 709 f5ilex linings for tube mills, 459 Silica cement, 592-597 Silica, in limestones, 309-310 Portland cement, 383, 386 Simple cementing materials, 3-7 Slag-blocks, analyses of slags used for, 689 , definition, 685 , methods of manufacture, C85-689 , properties of, 686 Slag-bricks, analyses of slags used for, 678, 679 , definition, 674 t methods of manufacture, 674-685 , properties, 676-679, 683 Slag cement, color of, 669 , composition of slags used, 641-644 , costs of manufacture, 663-664 , methods of manufacture, 641, 645-665 production of, 664 , properties and tests, 666-671 , specific gravity, 668-669 , specifications for, 672-674 , tensile strength, 670 , see also Puzzolan cements. Slags, used in Portland-cement manufacture, analyses, 352 , methods, 411-415 slag-cement manufacture, analyses, 643 , composition, 641-644 , drying, 649-652 , granulation, 644-648 , methods of use, 645-665 Slaked lime, see Lime hydrate. Slaking lime, 120-122, 126 natural-cement clinker, 235-236 Slate, analyses of, 364-365 , distribution of, 363-364 , origin of, 363 , used as Portland-cement material, 363-366 Slosson and Moudy, on accelerators and retarders for plasters, 65-67 temperature of plaster burning, 43-44 tests of plasters, 61-62 Smidth ball mill, 516 Soda, used as slag-cement accelerators, 659 , see Alkalies. Sorel stone, 9, 162-167 Specifications for natural cement, 276-284 Portland cement, 614-631 puzzolan cement, 672-674 slag cement, 672-674 Specific gravity, method of determining, 605-606 710 INDEX. Specific gravity, of anhydrite, 15 "cement plaster". 57 grappier cements, 185 gypsum, 15 Keene's "cement", 77 Lafarge cement, 185 lime, 115 limestone, 310 f* magnesia, 154 natural cement, 263-282 plasters, 57 Portland cement, 585 puzzolan cement, 668-669 slag cement, 668-669 Stack-dust from rotary kilns, composition and use, 510 Stack-gases from cement-kilns, composition of, 507 , use of, 509 lime-kilns, use of, 108-109 Statistics, see Imports, Production. Strength, see Adhesive strength, Compressive strength, Tensile strength. Structural materials, production of, in U. S., 1 Stucco, analyses, 57 , definition, 32 Sturtevant crusher for slaked lime, 125 rock-emery mill, 35, 125, 239-241 system of feeding powdered coal, 486-489 Stedman disintegrator, 36, 244 Sulphate of lime, see Gypsum, Plasters, Sulphur, Sulphide of iron, see Pyrite. Sulphides, in limestones ,'91, 310, 388 , presence in Portland cement, 388 slag cements, 666-667 Sulphur, effect in slag cements, 666 . , in alkali waste, 348, 349 , in kiln-coal, 512-513 , in limestones, 91, 310, 388 , in Portland cements, 388 , presence in slags, 648-666 , removal by granulating slag, 648 Sulphuric acid, see Sulphur trioxide. Sulphur trioxide, in gypsum, 14 limestones, 91, 310, 388 Tankage, used as plaster-retarder, 65 Teil, lime of, see Hydraulic-limes. Temperature of burning lime, 96-97 magnesite, 154 natural cement, 223-225 plasters, 31, 41, 44 INDEX. 711 Temperature of Portland cement, 499-500 Tensile strength, of grappier cements, 186 hard-finish plasters, 78 hydraulic limes, 183, 188 Keene's cement, 78 Lafarge cement, 186 limes, 122-123 natural cements, 267, 283 plasters, 58-61 Portland cements, 586-588, 612 puzzolan cements, 670, 673 selenitic limes, 191 slag cements, 670, 673 Testing methods, standard, 603-613 Tests of cementing materials, see Adhesive strength, Compressive strength, Tensile strength, Tests of efficiency of rotary kiln, 505-509 Tetmajer, Prof., effects of plaster, 543 tests of natural cement, 266-267 Portland cement, 505 Thermal efficiency of rotary kiln, 497-511 units defined and compared, 12 Thompson, S. ; on weight of cement barrels, 551-553 Travertine, 89 Tube mills, 404, 425, 430, 457, 516, 657-658 , Bonnot, 457 , Davidsen, 244, 457-462 , description of class, 430, 457 , discontinuous, 657-658, 660 , Gates, 457 , Jensch, 658 , Krupp, 457, 462-463 , pebbles for, 463-465 , used in natural-cement plants, 237, 243-244, 246 , West, 660 Tufa, 89 Tuff, 633, 638 United States, imports of gypsum, 85-86 magnesia and magnesite, 152-153 plasters, 85-86 Portland cement, 564-565 lime, 112-114 magnesite, 152-153 natural cement, 289-292 plasters, 79-84 Portland cement, 561-565 production of gypsum, 79-84 slag cement, 664 712 INDEX. United States, structural materials, 1 total cementing materials, 2 Van't Hoff, Prof., on constitution of flooring plasters, 69-75 Vegetable matter, as retarder for plaster, 65 , in marls, effects of, 341-343 , used in wall-plaster, 51-52* Wall-plasters, 33, 51-52 Water, amount present in chalks, 318 clays, 378 granulated slag, 649 gypsum, 14, 31 limestones, 378 marls, 340, 341, 343 plaster, 31, 44 shales, 378 required in slaking lime, 120 , sea-, effect on Portland cement, 602-603 , used in granulating slag, 644-648 Wear, resistance of slag cement to mechanical, 670 Weight per cubic foot of "cement plaster," 57 clay, 305 lime, 115 lime-sand bricks, 142 limestone, 305, 310 magnesia bricks, 161 marl, 340-341 plaster, 57 Portland cement, 551-553 shale, 305 slag bricks, 678, 683 , see also Specific gravity, Packing weights. Whiting process, regulation of set in slag cements, 658-659 Williams mill, 244-245, 466-468, 516 Wilder, F. A., on flooring plaster, 75 rotary calciners for plaster, 49-50 Wilkinson, P., on kettle process for plasters, 38-43 Wood-fiber, used in wall-plaster, 51 Woolson, I., on strength of clay bricks, 145 lime-sand bricks, 144 SHORT-TITLE CATALOGUE OF THE PUBLICATIONS OF JOHN WILEY & SONS NEW YORK LONDON: CHAPMAN & HALL, LIMITED ARRANGED UNDER SUBJECTS Descriptive circulars sent on application. Books marked with an asterisk (*) are sold at net prices only. All books are bound in cloth unless otherwise stated. AGRICULTURE HORTICULTURE FORESTRY. Armsby 's Principles of Animal Nutrition 8vo, $4 00 Budd and Hansen's American Horticultural Manual: Part I. Propagation, Culture, and Improvement 12mo, 1 50 Part II. Systematic Pomology 12mo, 1 50 Elliott's Engineering for Land Drainage 12mo, 1 50 Practical Farm Drainage. (Second Edition, Rewritten.) 12mo, 1 50 Graves's Forest Mensuration 8vo, 4 00 Principles of Handling Woodlands. (In Press.) Green's Principles of American Forestry 12mo, 1 50 Grotenfelt's Principles of Modern Dairy Practice. (Woll.) . 12mo, 2 00 * Herrick's Denatured or Industrial Alcohol 8vo, 4 00 Kemp and Waugh's Landscape Gardening. (New Edition, Rewritten. In Press.) * McKay and Larsen's Principles and Practice of Butter-making 8vo, 1 50 Maynard's Landscape Gardening as Applied to Home Decoration 12mo, 1 50 Sanderson's Insects Injurious to Staple Crops 12mo, 1 50 Sanderson and Headlee's Insects Injurious to Garden Crops. (In Prepa- ration.) * Schwarz's Longleaf Pine in Virgin Forest, 12mo, 1 25 Solotaroff's Shade Trees in Towns and Cities. (In Press.) Stockbridge's Rocks and Soils 8vo, 2 50 Win ton's Microscopy of Vegetable Foods 8vo, 7 50 Woll's Handbook for Farmers and Dairymen 16mo, 1 50 ARCHITECTURE. Baldwin's Steam Heating for Buildings 12mo, 2 50 Berg's Buildings and Structures of American Railroads 4to, 5 00 Birkmire's Architectural Iron and Steel 8vo, 3 50 Compound Riveted Girders as Applied in Buildings 8vo, 2 00 Planning and Construction of American Theatres 8vo, 3 00 1 Birkmire's Planning and Construction of High Office Buildings 8vo. $3 50 Skeleton Construction in Buildings 8vc, 3 00 Briggs's Modern American School Buildings 8vo, 4 00 Byrne's Inspection of Materials and Wormanship Employed in Construction. 16mo, 3 00 Carpenter's Heating and Ventilating of Buildings 8vo, 4 00 * Corthell's Allowable Pressure on Deep Foundations 12mo, 1 25 Fieilag's Architectural Engineering 8vo, 3 50 Fireproofing of Steel Buildings 8vo, 2 50 Gerhard's Guide to Sanitary- Inspections. (Fourth Edition, Entirely Re- vised and Enlarged.) .' J? 12mo, 1 50 * Modern Baths and Bath Houses '. 8vo, 3 00 Sanitation of Public Buildings 12mo, 1 50 Theatre Fires and Panics 12mo, 1 50 * The Water Supply, Sewerage and Plumbing of Modern City Buildings. 8vo, 4 00 Johnson's Statics by Algebraic and Graphic Methods 8vo, 2 00 Kellaway's How to Lay Out Suburban Home Grounds 8vo, 2 00 Kidder's Architects' and Builders' Pocket-book 16mo, mor., 5 00 Merrill's Stones for Building and Decoration 8vo, 5 00 Monckton's Stair-building 4to, 4 00 Patton's Practical Treatise on Foundations 8vo, 5 00 Peabody's Naval Architecture 8vc, 7 50 Rice's Concrete-block Manufacture 8vo, 2 00 Richey's Handbook for Superintendents of Construction 16mo, mor. 4 00 Building Foreman's Pocket Book and Ready Reference. . 16mo, mor. 5 00 * Building Mechanics' Ready Reference Series: * Carpenters' and Woodworkers' Edition 16mo, mor. 1 50 * Cement Workers' and Plasterers' Edition 16mo, mor. 1 50 * Plumbers', Steam-Fitters', and Tinners' Edition. . .16mo, mor. 1 50 * Stone- and Brick-masons' Edition 16mo, mor. 1 50 Sabin's House Painting 12mo, 1 00 Siebert and Biggin's Modern Stone-cutting and Masonry 8vo, 1 50 Snow's Principal Species of Wood 8vo, 3 50 Wait's Engineering and Architectural Jurisprudence 8vo, 6 00 Sheep, 6 50 Law of Contracts 8vo, 3 00 Law of Operations Preliminary to Construction in Engineering and Architecture 8vo, 5 00 Sheep, 5 50 Wilson's Air Conditioning 12mo, 1 50 Worcester and Atkinson's Small Hospitals, Establishment and Maintenance, Suggestions for Hospital Architecture, with Plans for a Small Hospital 12mo, 1 25 ARMY AND NAVY. Bernadou's Smokeless Powder, Nitro-cellulose, and the Theory of the Cellu- lose Molecule l2mo, 2 50 Chase's Art of Pattern Making 12mo, 2 50 Screw Propellers and Marine Propulsion 8vo, 3 00 * Cloke's Enlisted Specialists' Examiner 8vo, 2 00 * Gunner's Examiner 8vo, 1 50 Craig's Azimuth 4tu, 3 50 Crehore and Squier's Polarizing Photo-chronograph 8vo. 3 00 * Davis's Elements of Law 8vo, 2 50 * Treatise on the Military Law of United States 8vo, 7 00 * Dudley's Military Law and the Procedure of Courts-martial. ..Large 12mo, 2 50 Durand's Resistance and Propulsion of Ships 8vo, 5 00 * Dyer's Handbook of Light Artillery 12mo. 3 00 Eissler's Modern High Explosives 8vo, 4 00 * Fiebeger's Text-book on Field Fortification Large 12mo, 2 00 Hamilton and Bond's The Gunner's Catechism 18mo, 1 00 * Hoff's Elementary Naval Tactics 8vo, $1 50 Ingalls's Handbook of Problems hi Direct Fire . . . 8vo, 4 00 * Lissak's Ordnance and Gunnery 8vo, 6 00 * Ludlow's Logarithmic and Trigonometric Tables 8vo, 1 00 * Lyons's Treatise on Electromagnetic Phenomena. Vols. I. and II..8vo,each, 6 00 * Mahan's Permanent Fortifications. (Mercur.) 8vo, half mor. 7 50 Manual for Courts-martial 16mo,mor. 1 50 * Mercur's Attack of Fortified Places 12mo, 2 00 * Elements of the Art of War 8vo, 4 00 Nixon's Adjutants' Manual 24mo, 1 00 Peabody's Naval Architecture 8vo, 7 50 * Phelps's Practical Marine Surveying 8vo, 2 50 Putnam's Nautical Charts 8vo, 2 00 Rust's Ex-meridian Altitude, Azimuth and Star-Finding Tables Svo, 5 00 * Selkirk's Catechism of Manual of Guard Duty 24mo, 50 Sharpe's Art of Subsisting Armies in War 18mo, mor. 1 50 *Taylor's Speed and Power of Ships. 2 vols, Text 8vo, plates oblong 4to, 7 50 * Tupes and Poole's Manual of Bayonet Exercises and Musketry Fencing. 24mo, leather, 50 * Weaver's Military Explosives 8vo, 3 00 * Woodhull's Military Hygiene for Officers of the Line Large 12mo, 1 50 ASSAYING. Betts's Lead Refining by Electrolysis 8vo, 4 00 *Butler's Handbook of Blowpipe Analysis 16mo, 75 Fletcher's Practical Instructions in Quantitative Assaying with the Blowpipe. 16mo, mor. 1 50 Furman and Pardoe's Manual of Practical Assaying 8vo, 3 00 Lodge's Notes on Assaying and Metallurgical Laboratory Experiments.. 8vo, 3 00 Low's Technical Methods of Ore Analysis 8vo, 3 00 Miller's Cyanide Process 12mo, 1 00 Manual of Assaying 12mo, 1 00 Minet's Production of Aluminum and its Industrial Use. (Waldo.). ..12mo, 2 50 Ricketts arid Miller's Notes on Assaying Svo. 3 00 Robine and Lenglen's Cyanide Industry. (Le Clerc.) 8vo, 4 00 * Seamen's Manual for Assayers and Chemists Large 12mo, 2 50 Ulke's Modern Electrolytic Copper Refining Svo, 3 00 Wilson's Chlorination Process 12mo, 1 50 Cyanide Processes , , , 12mo, 1 50 ASTRONOMY. Comstock's Field Astronomy for Engineers Svo, 2 50 Craig's Azimuth 4to, 3 50 Crandall's Text-book on Geodesy and Least Squares Svo, 8 00 Doolittle's Treatise on Practical Astronomy Svo, 4 00 Hayford's Text-book of Geodetic Astronomy Svo, 3 00: Hosmer's Azimuth 16mo, mor. 1 00 * Text-book on Practical Astronomy Svo, 2 00 Merriman's Elements of Precise Surveying and Geodesy Svo, 2 50 * Michie and Harlow's Practical Astronomy Svo, 3 00 Rust's Ex-meridian Altitude, Azimuth and Star-Finding Tables Svo, 5 00 * White's Elements of Theoretical and Descriptive Astronomy 12mo, 2 00 CHEMISTRY. * Abderhalden's Physiological Chemistry in Thirty Lectures. (Hall and Defren.) Svo, 5 00 * Abegg's Theory of Electrolytic Dissociation, (von Ende.) 12mo, 1 25 Alexeyeff's General Principles of Organic Syntheses. (Matthews.) Svo, 3 00 Allen's Tables for Iron Analysis Svo, 3 00 Armsby's Principles of Animal Nutrition Svo, 4 00 3 Arnold's Compendium of Chemistry. (Mandel.) Large 12mo, $3 50 Association of State and National Food and Dairy Departments, Hartford Meeting, 1906 8vo, 3 00 Jamestown Meeting, 1907 8vo, 3 00 Austen's Notes for Chemical Students , 12mo, 1 50 Baskerville's Chemical Elements. (In Preparation.) Bernadou's Smokeless Powder. Nitro-cellulose, and Theory of the Cellulose Molecule 12mo, 2 50 * Biltz's Introduction to Inorganic Chemistry. (Hall and Phelan.). . . 12mo, 1 25 Laboratory Methods of Inorganic Chemistry. (Hall and Blanchard.) ^ 8vo, 3 00 * Blanchard's Synthetic Inorganic Chemistry. 4 12mo, 1 00 * Browning's Introduction to the Rarer Elements 8vo, 1 50 * Butler's Handbook of Blowpipe Analysis 16mo, 75 * Claassen's Beet-sugar Manufacture. (Hall and Rolfe.) 8vo, 3 00 Classen's Quantitative Chemical Analysis by Electrolysis. (Boltwood.).8vo, 3 00 Cohn's Indicators and Test-papers k . . 12mo, 2 00 Tests and Reagents 8vo, 3 00 * Danneel's Electrochemistry. (Merriam.) 12mo, 1 25 Dannerth's Methods of Textile Chemistry 12mo, 2 00 Duhem's Thermodynamics and Chemistry. (Burgess.) 8vo, 4 00 Effront's Enzymes and their Applications. (Prescott.) 8vo, 3 00 Eissler's Modern High Explosives 8vo. 4 00 * Fischer's Oedema 8vo, 2 00 * Physiology of Alimentation Large 12mo, 2 00 Fletcher's Practical Instructions in Quantitative Assaying with the Blowpipe. 16mo, mor. 1 50 Fowler's Sewage Works Analyses 12mo, 2 00 Fresenius's Manual of Qualitative Chemical Analysis. (Wells.) 8vo, 5 00 Manual of Qualitative Chemical Analysis. Part I. Descriptive. ( Wells. )8vo, 3 00 Quantitative Chemical Analysis. (Cohn.) 2 vols 8vc, 12 50 When Sold Separately, Vol. I, $6. Vol. II, $8. Fuertes's Water and Public Health 12mo, 1 50 Furman and Pardoe's Manual of Practical Assaying 8vo, 3 00 * Getman's Exercises in Physical Chemistry 12mo, 2 00 Gill's Gas and Fuel Analysis for Engineers 12mo, 1 25 * Gooch and Browning's Outlines of Qualitative Chemical Analysis. Large 12mo, 1 25 Grotenfelt's Principles of Modern Dairy Practice. (Woll.) 12mo, 2 00 Groth's Introduction to Chemical Crystallography (Marshall) 12mo, 1 25 Hammarsten's Text-book of Physiological Chemistry. (Mandel.) 8vo, 4 00 Hanausek's Microscopy of Technical Products. (Win ton.) 8vo, 5 00 * Haskins and Macleod's Organic Chemistry 12mo, 2 00 * Herrick's Denatured or Industrial Alcohol 8vo, 4 00 Hinds's Inorganic Chemistry 8vo, 3 00 * Laboratory Manual for Students 12mo, 1 00 * Holleman's Laboratory Manual of Organic Chemistry for Beginners. (Walker.) 12mo, 1 00 Text-book of Inorganic Chemistry. (Cooper.) 8vo, 2 50 Text-book of Organic Chemistry. (Walker and Mott.) 8vo, 2 50 * Holley's Lead and Zinc Pigments Large 12mo, 3 00 Holley and Ladd's Analysis of Mixed Paints, Color Pigments, and Varnishes. Large 12mo, 2 50 Hopkins's Oil-chemists' Handbook 8vo, 3 00 Jackson's Directions for Laboratory Work in Physiological Chemistry. .8vo, 1 25 Johnson's Rapid Methods for the Chemical Analysis of Special Steels, Steel- making Alloys and Graphite Large 12mo, 3 00 Landauer's Spectrum Analysis. (Tingle.) 8vo, 3 00 Lassar-Cohn's Application of Some General Reactions to Investigations in Organic Chemistry. (Tingle.) 12mo, 1 00 Leach's Inspection and Analysis of Food with Special Reference to State Control 8vo, 7 50 Lob's Electrochemistry of Organic Compounds. (Lorenz.) 8vo, 3 00 Lodge's Notes on Assaying and Metallurgical Laboratory Experiments.. 8vo, 3 00 Low's Technical Method of Ore Analysis ; 8vo, 3 00 Lowe's Paint for Steel Structures 12mo, 1 00 Lunge's Techno-chemical Analysis. (Cohn.) 12mo. 1 00 4 * McKay and Larsen's Principles and Practice of Butter-making 8vo, $1 50 Maire's Modern Pigments and their Vehicles 12mo, 2 00 Mandel's Handbook for Bio-chemical Laboratory 12mo, 1 50 * Martin's Laboratory Guide to Qualitative Analysis with the Blowpipe 12mo, 60 Mason's Examination of Water. (Chemical and Bacteriological.) 12mo, 1 25 Water-supply. (Considered Principally from a Sanitary Standpoint.) 8vo, 4 00 * Mathewson's First Principles of Chemical Theory 8vo, 1 00 Matthews's Laboratory Manual of Dyeing and Textile Chemistry 8vo, 3 50 Textile Fibres. 2d Edition, Rewritten 8vo, 4 00 * Meyer's Determination of Radicles in Carbon Compounds. (Tingle.) Third Edition 12mo, 1 25 Miller's Cyanide Process 12mo, 1 00 Manual of Assaying 12mo, 1 00 Minet's Production of Aluminum and its Industrial Use. (Waldo.). ..12mo, 2 50 * Mittelstaedt's Technical Calculations for Sugar Works. (Bourbakis.) 12mo, 1 50 Mixter's Elementary Text-book of Chemistry 12mo, 1 50 Morgan's Elements of Physical Chemistry 12mo, 3 00 Outline of the Theory of Solutions and its Results 12mo, 1 00 * Physical Chemistry for Electrical Engineers 12mo, 1 50 * Moore's Outlines of Organic Chemistry 12mo, 1 50 Morse's Calculations used in Cane-sugar Factories 16mo, mor. 1 50 * Muir's History of Chemical Theories and Laws 8vo, 4 00 Mulliken's General Method for the Identification of Pure Organic Compounds. Vol. I. Compounds of Carbon with Hydrogen and Oxygen. Large 8vo, 5 00 Vol. II. Nitrogenous Compounds. (In Preparation.) Vol. III. The Commercial Dyestuffs : Large 8vo, 5 00 * Nelson's Analysis of Drugs and Medicines 12mo, 3 00 Ostwald's Conversations on Chemistry. Part One. (Ramsey.) 12mo, 1 50 Part Two. (Turnbull.) 12mo, 200 Introduction to Chemistry. (Hall and Williams.) (In Press.) Owen and Standage's Dyeing and Cleaning of Textile Fabrics 12mo, 2 00 * Palmer's Practical Test Book of Chemistry 12mo, 1 00 * Pauli's Physical Chemistry in the Service of Medicine. (Fischer.) . . 12mo, 1 25 Penfield's Tables of Minerals, Including the Use of Minerals and Statistics of Domestic Production 8vo, 1 00 Pictet's Alkaloids and their Chemical Constitution. (Biddle.) 8vo, 5 00 Poole's Calorific Power of Fuels 8vo, 3 00 Prescott and Winslow's Elements of Water Bacteriology, with Special Refer- ence to Sanitary Water Analysis 12mo, 1 50 * Reisig's Guide to Piece-Dyeing 8vo, 25 00 Richards and Woodman's Air, Water, and Food from a Sanitary Stand- point 8vo. 2 00 Ricketts and Miller's Notes on Assaying 8vo, 3 00 Rideal's Disinfection and the Preservation of Food 8vo, 4 00 Sewage and the Bacterial Purification of Sewage 8vo, 4 00 Riggs's Elementary Manual for the Chemical Laboratory 8vo, 1 25 Robine and Lenglen's Cyanide Industry. (Le Clerc.) 8vo, 4 00 Ruddiman's Incompatibilities in Prescriptions 8vo, 2 00 Whys in Pharmacy 12mo, 1 00 * Ruer's Elements of Metallography. (Mathewson.) 8vo, 3 00 Sabin's Industrial and Artistic Technology of Paint and Varnish 8vo, 3 00 Salkowski's Physiological and Pathological Chemistry. (Orndorff.) 8vo, 2 50 Schimpf 's Essentials of Volumetric Analysis 12mo, 1. 25 Manual of Volumetric Analysis. (Fifth Edition, Rewritten) 8vo, 5 00 * Qualitative Chemical Analysis 8vo, 1 25 * Seamon's Manual for Assayers and Chemists Large 12mo. 2 50 Smith's Lecture Notes on Chemistry for Dental Students 8vo, 2 50 Spencer's Handbook for Cane Sugar Manufacturers 16mo, mor. 3 00 Handbook for Chemists of Beet-sugar Houses 16mo, mor. 3 00 Stockbridge's Rocks and Soils 8vo, 2 50 Stone's Practical Testing of Gas and Gas Meters 8vo, 3 50 * Tillman's Descriptive General Chemistry 8vo, 3 00 * Elementary Lessons in Heat 8vo, 1 50 Treadwell's Qualitative Analysis. (Hall.) 8vo, 3 00 Quantitative Analysis. (Hall.) 8vo. 4 00 5 Turneaure and Russell's Public Water-supplies .8vo, $5 00 Van Deventer's Physical Chemistry for Beginners. (Boltwood.) 12mo, 1 50 Venable's Methods and Devices for Bacterial Treatment of Sewage 8vo, 3 00 Ward and Whipple's Freshwater Biology. (In Press.) Ware's Beet-sugar Manufacture and Refining. Vol. I. 8vo, 4 00 Vol. II ...8vo, 5 00 Washington's Manual of the Chemical Analysis of Rocks. . . .^. 8vo, 2 00 * Weaver's Military Explosives , 8vo, 3, 00 Wells's Laboratory Guide in Qualitative Chemical Analysis 8vo, 1 50 Short Course in Inorganic Qualitative Chemical Analysis for Engineering Students .-.'., .^ 12mo, 1 50 Text-book of Chemical Arithmetic .; ^ . 12mo, 1 25 Whipple's Microscopy of Drinking-water 8vo, 3 50 Wilson's Chlorination Process 12mo, 1 50 Cyanide Processes 12mo, 1 50 Wmton's Microscopy of Vegetable Poods 8vo, 7 50 Zsigmondy's Colloids and the Ultramicroscope. (Alexander.).. Large 12mo, 3 00 CIVIL ENGINEERING. BRIDGES AND ROOFS. HYDRAULICS. MATERIALS OF ENGINEER- ING. RAILWAY ENGINEERING. Baker's Engineers' Surveying Instruments 12mo, 3 00 Bixby's Graphical Computing Table Paper 19 X 24$ inches. 25 Breed and Hosmer's Principles and Practice of Surveying. Vol. I. Elemen- tary Surveying .' . . 8vo, 3 00 Vol. II. Higher Surveying 8vo, 2 50 * Burr's Ancient and Modern Engineering and the Isthmian Canal 8vo, 3 50 Comstock's Field Astronomy for Engineers 8vo, 2 50 *-Corthell's Allowable Pressure on Deep Foundations 12mo, 1 25 Crandall's Text-book on Geodesy and Least Squares 8vo, 3 00 Davis's Elevation and Stadia Tables 8vo, 1 00 Elliott's Engineering for Land Drainage 12mo, 1 50 * Fiebeger's Treatise on Civil Engineering 8vo, 5 00 Flemer's Photographic Methods and Instruments 8vo, 5 00 Folwell's Sewerage. (Designing and Maintenance.) 8vo, 3 00 Frei tag's Architectural Engineering 8vo, 3 50 French and Ives's Stereotomy 8vo, 2 50 Goodhue's Municipal Improvements 12mo, 1 50 * Hauch arid Rice's Tables of Quantities for Preliminary Estimates. . . 12mo, 1 25 Hayford's Text-book of Geodetic Astronomy 8vo, 3 00 Hering's Ready Reference Tables (Conversion Factors.) 16mo, mor. 2 50 Hosmer's Azimuth 16mo, mor. 1 00 * Text-book on Practical Astronomy 8vo, 2 00 Howe's Retaining Walls for Earth 12mo, 1 25 * Ives's Adjustments of the Engineer's Transit and Level 16mo, bds. 25 Ives and Hilts's Problems in Surveying, Railroad Surveying and Geod- esy 16mo, mor. 1 50 * Johnson (J.B.) and Smith's Theory and Practice of Surveying . Large 12mo, 3 50 Johnson's (L. J.) Statics by Algebraic and Graphic Methods 8vo, 2 00 * Kinnicutt, Winslow and Pratt's Sewage Disposal 8vo, 3 00 * Mahan's Descriptive Geometry 8vo, 1 50 Merriman's Elements of Precise Surveying and Geodesy 8vo, 2 50 Merriman and Brooks's Handbook for Surveyors 16mo, mor. 2 00 Nugent's Plane Surveying 8vo, 3 50 Ogden's Sewer Construction 8vo. 3 00 Sewer Design 12mo, 2 00 Parsons's Disposal of Municipal Refuse 8vo, 2 00 Patton's Treatise on Civil Engineering .8vo, half leather, 7 50 Reed's Topographical Drawing and Sketching 4to, 5 00 Rideal's Sewage and the Bacterial Purification of Sewage 8vo, 4 00 Riemer's Shaft-sinking under Difficult Conditions. (Corning and Peele.).8vo, 3 00 Siebert and. Biggin's Modern Stone-cutting and Masonry 8vo, 1 50 6 Smith's Manual of Topographical Drawing. (McMillan.) 8vo, $2 50 Soper's Air and Ventilation of Subways 12mo, 2 50 * Tracy's Exercises in Surveying '. 12mo, mor. 1 00 Tracy's Plane Surveying 16mo, mor. 3 00 * Trautwine's Civil Engineer's Pocket-book 16mo, mor. 5 00 Venable's Garbage Crematories in America 8vo, 2 00 Methods and Devices for Bacterial Treatment of Sewage 8vo, 3 00 Wait's Engineering and Architectural Jurisprudence 8vo, 6 00 Sheep, 6 50 Law of Contracts 8vo, 3 00 Law of Operations Preliminary to Construction in Engineering and Architecture. . , 8vo, 5 00 Sheep, 5 50 Warren's Stereotomy Problems in Stone-cutting 8vo, 2 50 * Waterbury's Vest-Pocket Hand-book of Mathematics for Engineers. 2fX5l inches, mor. 1 00 * Enlarged Edition, Including Tables mor. 1 50 Webb's Problems in the Use and Adjustment of Engineering Instruments. 16mo, mor. 1 25 Wilson's Topographic Surveying 8vo, 3 50 BRIDGES AND ROOFS. Boiler's Practical Treatise on the Construction of Iron Highway Bridges.. 8vo, 2 00 * Thames River Bridge Oblong paper, 5 00 Burr and Falk's Design and Construction of Metallic Bridges 8vo, 5 00 Influence Lines for Bridge and Roof Computations 8vo, 3 00 Du Bois's Mechanics of Engineering. Vol. II Small 4to, 10 00 Foster's Treatise on Wooden Trestle Bridges 4 to, 5 00 Fowler's Ordinary Foundations 8vo, 3 50 Greene's Arches in Wood, Iron, and Stone 8vo, 2 50 Bridge Trusses 8vo, 2 50 Roof Trusses 8vo, 1 25 Grimm's Secondary Stresses in Bridge Trusses 8vo, 2 50 Heller's Stresses in Structures and the Accompanying Deformations.. . .8vo, 3 00 Howe's Design of Simple Roof-trusses in Wood and Steel 8vo. 2 00 Symmetrical Masonry Arches 8vo, 2 50 Treatise on Arches 8vo, 4 00 * Jacoby's Structural Details, or Elements of Design in Heavy Framing, 8vo, 2 25 Johnson, Bryan and Turneaure's Theory and Practice in the Designing 1 of Modern Framed Structures Small 4to, 10 00 * Johnson, Bryan and Turneaure's Theory and Practice in the Designing of Modern Framed Structures. New Edition. Part 1 8vo, 3 00 Part II. Rewritten. (In Press.) Merriman and Jacoby's Text-book on Roofs and Bridges: Part I. Stresses in Simple Trusses 8vo, 2 50 Part II. Graphic Statics 8vo', 2 50 Part III. Bridge Design 8vo, 2 50 Part IV. Higher Structures 8vo, 2 50 Sondericker's Graphic Statics, with Applications to Trusses, Beams, and Arches 8vo, 2 00 Waddell's De Pontibus, Pocket-book for Bridge Engineers 16mo, mor. 2 00 * Specifications for Steel Bridges 12mo, 50 Waddell and Harrington's Bridge Engineering. (In Preparation.) HYDRAULICS. Barnes's Ice Formation 8vo, 3 00 Bazin's Experiments upon the Contraction of the Liquid Vein Issuing from an Orifice. (Trautwine.) 8vo, 2 00 Bovey 's Treatise on Hydraulics 8vo, 5 00 Church's Diagrams of Mean Velocity of Water in Open Channels. Oblong 4to, paper, 1 50 Hydraulic Motors 8vo, 2 00 7 Coffin's Graphical Solution of Hydraulic Problems IGmo, mor. $2 50 Flather's Dynamometers, and the Measurement of Power 12mo, 3 00 Folwell's Water-supply Engineering 8vo, 4 00 Frizell's Water-power 8vo, 5 00 Fuertes's Water and Public Health 12mo, 1 50 Water-filtration Works 12mo, 2 50 Ganguillet and Kutter's General Formula for the Uniform Flow of Water in Rivers and Other Channels. (Hering and Trautwine.) 8vo, 4 00 Hazen's Clean Water and How to Get It Large 12mo, 1 50 Filtration of Public Water-supplies 8vo, 3 00 Hazelhurst's Towers and Tanks for Water-worifs 8vo, 2 50 Herschel's 115 Experiments on the Carrying Capacity of Large, Riveted, Metal Conduits 8vo, 2 00 Hoyt and Grover's River Discharge 8vo, 2 00 Hubbard and Kiersted's Water-works Management and Maintenance. 1. 8vo, 4 00 * Lyndon's Development and Electrical Distribution of Water Power. 8vo, 3 00 Mason's Water-supply. (Considered Principally from a Sanitary Stand- point.) 8vo, 4 00 Merriman's Treatise on Hydraulics 8vo, 5 00 * Molitor's Hydraulics of Rivers, Weirs and Sluices 8vo, 2 00 * Morrison and Brodie's High Masonry Dam Design 8vo, 1 50 * Richards's Laboratory Notes on Industrial Water Analysis 8vo, 50 Schuyler's Reservoirs for Irrigation, Water-power, and Domestic Water- supply. Second Edition, Revised and Enlarged Large 8vo, 6 00 * Thomas and Watt's Improvement of Rivers 4to, 6 00 Turneaure and Russell's Public Water-supplies 8vo, 5 00 Wegmann's Design and Construction of Dams. 5th Ed., enlarged 4to, 6 CO Water-Supply of the City of New York from 1658 to 1895 4io, 10 00 Whipple's Value of Pure Water Large 12mo, 1 00 Williams and Hazen's Hydraulic Tables 8vo, 1 50 Wilson's Irrigation Engineering 8vo, 4 00 Wood's Turbines 8vo, 2 50 MATERIALS OF ENGINEERING. Baker's Roads and Pavements 8vo, 5 00 Treatise on Masonry Construction 8vo, 5 00 Black's United States Public Works Oblong 4to, 5 00 Blanchard's Bituminous Roads. (In Preparation.) Bleininger's Manufacture of Hydraulic Cement. (In Preparation.) * Bovey 's Strength of Materials and Theory of Structures 8vo, 7 50 Burr's Elasticity and Resistance of the Materials of Engineering 8vo, 7 50 Byrne's Highway Construction 8vo, 5 00 Inspection of the Materials and Workmanship Employed in Construction. 16mo, 3 00 Church's Mechanics of Engineering 8vo, 6 00 Du Bois's Mechanics of Engineering. Vol. I. Kinematics, Statics, Kinetics Small 4to, 7 50 Vol. II. The Stresses in Framed Structures, Strength of Materials and Theory of Flexures Small 4to, 10 00 * Eckel's Cements, Limes, and Plasters 8vo, 6 00 Stone and Clay Products used in Engineering. (In Preparation.) Fowler's Ordinary Foundations. . .' 8vo, 3 50 * Greene's Structural Mechanics 8vo, 2 50 * Holley's Lead and Zinc Pigments Large 12mo, 3 00 Holley and Ladd's Analysis of Mixed Paints, Color Pigments and Varnishes. Large 12mo, 2 50 * Hubbard's Dust Preventives and Road Binders 8vo, 3 00 Johnson's (C. M.) Rapid Methods for the Chemical Analysis of Special Steels, Steel-making Alloys and Graphite Large 12mo, 3 00 Johnson's (J. B.) Materials of Construction Large 8vo, 6 00 Keep's Cast Iron 8vo, 2 50 Lanza's Applied Mechanics 8vo, 7 50 Lowe's Paints for Steel Structures 12mo, 1 00 8 Maire's Modern Pigments and their Vehicles 12mo, $2 00 Maurer's Technical Mechanics 8vo, 4 00 Merrill's Stones for Building and Decoration .' . ,8vo, 5 00 Merriman's Mechanics of Materials .* 8vo, 5 00 * Strength of Materials 12mo, 1 00 Metcalf 's Steel. A Manual for Steel-users 12mo, 2 00 Morrison's Highway Engineering . . . . 8vo, 2 50 Patton's Practical Treatise on Foundations 8vo, 5 00 Rice's Concrete Block Manufacture 8vo, 2 00 Richardson's Modern Asphalt Pavement 8vo, 3 00 Richey's Building Foreman's Pocket Book and Ready Reference. 16mo,mor. 5 00 * Cement Workers' and Plasterers' Edition (Building Mechanics' Ready Reference Series) 16mo, mor. 1 50 Handbook for Superintendents of Construction 16mo, mor. 4 00 * Stone and Brick Masons' Edition (Building Mechanics' Ready Reference Series) 16mo, mor. 1 50 * Ries's Clays : Their Occurrence, Properties, and Uses 8vo, 5 00 * Ries and Leighton's History of the Clay-working Industry of the United States 8vo. 2 50 Sabin's Industrial and Artistic Technology of Paint and Varnish 8vo, 3 00 * Smith's Strength of Material 12mo 1 25 Snow's Principal Species of Wood 8vo, 3 50 Spalding's Hydraulic Cement 12mo, 2 00 Text-book on Roads and Pavements 12mo, 2 00 * Taylor and Thompson's Extracts on Reinforced Concrete Design 8vo, 2 00 Treatise on Concrete, Plain and Reinforced 8vo, 5 00 Thurston's Materials of Engineering. In Three Parts 8vo, 8 00 Part I. Non-metallic Materials of Engineering and Metallurgy. . . . 8vo, 2 00 Part II. Iron and Steel 8vo, 3 50 Part III. A Treatise on Brasses, Bronzes, and Other Alloys and their Constituents 8vo, 2 50 Tillson's Street Pavements and Paving Materials 8vo, 4 00 * Trautwine's Concrete, Plain and Reinforced 16mo, 2 00 Turneaure and Maurer's Principles of Reinforced Concrete Construction. Second Edition, Revised and Enlarged 8vo, 3 50 Waterbury's Cement Laboratory Manual 12mo, 1 00 Wood's (De V.) Treatise on the Resistance of Materials, and an Appendix on the Preservation of Timber 8vo, 2 00 Wood's (M. P.) Rustless Coatings: Corrosion and Electrolysis of Iron and Steel 8vo, 4 00 RAILWAY ENGINEERING. Andrews's Handbook for Street Railway Engineers 3X5 inches, mor. 1 25 Berg's Buildings and Structures of American Railroads 4to, 5 00 Brooks's Handbook of Street Railroad Location 16mo, mor. 1 50 Butts's Civil Engineer's Field-book 16mo, mor. 2 50 Crandall's Railway and Other Earthwork Tables 8vo, 1 50 Crandall and Barnes's Railroad Surveying 16mo, mor. 2 00 * Crockett's Methods for Earthwork Computations 8vo, 1 50 Dredge's History of the Pennsylvania Railroad. (1879) Paper, 5 00 Fisher's Table of Cubic Yards. Cardboard, 25 Godwin's Railroad Engineers' Field-book and Explorers' Guide. . 16mo, mor. 2 50 Hudson's Tables for Calculating the Cubic Contents of Excavations and Em- bankments 8vo, 1 00 Ives and Hilts's Problems in Surveying, Railroad Surveying aiiJ Geodesy 16mo, mor. 1 50 Molitor and Beard's Manual for Resident Engineers 16mo, 1 00 Nagle's Field Manual for Railroad Engineers .... 16mo, mor. 3 00 * Orrock's Railroad Structures and Estimates 8vo, 3 00 Philbrick's Field Manual for Engineers 16mo, mor. 3 00 Raymond's Railroad Field Geometry 16mo, mor. 2 00 Elements of Railroad Engineering 8vo, 3 50 Railroad Engineer's Field Book. (In Preparation.) Roberts' Track Formulae and Tables 16mo, mor. 3 00 9 Searles's Field Engineering IGmo, mor. $3 00 Railroad Spiral. 16mo, mor. 1 50 Taylor's Prismoidal Formulae and Earthwork 8vo, 1 50 * Trautwine's Field Practice of Laying Out Circular Curves for Railroads. 12mo, mor. 2 50 * Method of Calculating the Cubic Contents of Excavations and Em- bankments by the Aid of Diagrams 8vo, 2 00 Webb's Economics of Railroad Construction Large 12mo, 2 50 Railroad Construction 16mo, mor. 5 00 Wellington's Economic Theory of the Location of Railways Large 12mo, 5 00 Wilson's Elements of Railroad-Track and Construction 12mo, 2 00 DRAWING. K Barr's Kinematics of Machinery 8vo, 2 50 * Bartlett's Mechanical Drawing 8vo, 3 00 * " Abridged Ed 8vo, 150 * Bartlett and Johnson's Engineering Descriptive Geometry 8vo, 1 50 Coolidge's Manual of Drawing 8vo, paper, 1 00 Coolidge and Freeman's Elements of General Drafting for Mechanical Engi- neers Oblong 4to. 2 50 Durley's Kinematics of Machines 8vo, 4 00 Emch's Introduction to Projective Geometry and its Application 8vo, 2 50 Hill's Text-book on Shades and Shadows, and Perspective 8vo, 2 00 Jamison's Advanced Mechanical Drawing 8vo, 2 00 Elements of Mechanical Drawing 8vo, 2 50 Jones's Machine Design: Part I. Kinematics of Machinery 8vo, 1 50 Part II. Form, Strength, and Proportions of Parts 8vo, 3 00 Kaup's Text-book on Machine Shop Practice. (In Press.) * Kimball and Barr's Machine Design 8vo, 3 00 MacCord's Elements of Descriptive Geometry 8vo, 3 00 Kinematics; or, Practical Mechanism 8vo, 5 00 Mechanical Drawing 4to, 4 00 Velocity Diagrams 8vo, 1 50 McLeod's Descriptive Geometry Large 12mo, 1 50 * Mahan's Descriptive Geometry and Stone-cutting 8vo, 1 50 Industrial Drawing. (Thompson.) 8vo, 3 50 Moyer's Descriptive Geometry 8vo, 2 00 Reed's Topographical Drawing and Sketching 4to, 5 00 * Reid's Mechanical Drawing. (Elementary and Advanced.) 8vo, 2 00 Text-book of Mechanical Drawing and Elementary Machine Design.. 8vo, 3 00 Robinson's Principles of Mechanism 8vo, 3 00 Schwamb and Merrill's Elements of Mechanism 8vo, 3 00 Smith (A. W.) and Marx's Machine Design Svo, 3 00 Smith's (R. S.) Manual of Topographical Drawing. (McMillan.) Svo, 2 50 * Titsworth's Elements of Mechanical Drawing Oblong Svo, 1 25 Warren's Drafting Instruments and Operations 12mo, 1 25 Elements of Descriptive Geometry, Shadows, and Perspective 8vo, 3 50 Elements of Machine Construction and Drawing Svo, 7 50 Elements of Plane and Solid Free-hand Geometrical Drawing. . . . 12mo, 1 00 General Problems of Shades and Shadows Svo, 3 00 Manual of Elementary Problems in the Linear Perspective of Forms and Shadow 12mo, 1 00 Manual of Elementary Projection Drawing 12mo, 1 50 Plane Problems in Elementary Geometry 12mo, 1 25 Weisbach's Kinematics and Power of Transmission. (Hermann and Klein.) Svo, 5 00 Wilson's (H. M.) Topographic Surveying Svo, 3 50 * Wilson's (V. T.) Descriptive Geometry Svo, 1 50 Free-hand Lettering Svo, 1 00 Free-hand Perspective Svo, 2 50 Woolf's Elementary Course in Descriptive Geometry Large Svo, 3 00 10 ELECTRICITY AND PHYSICS. * Abegg's Theory of Electrolytic Dissociation, (von Ende.!) 12mo, $1 25 Andrews's Hand-book for Street Railway Engineering 3X5 inches, mor. 1 25 Anthony and Ball's Lecture-notes on the Theory of Electrical Measure- ments 12mo, 1 00 Anthony and Brackett's Text-book of Physics. (Magie.) ... .Large 12mo, 3 00 Benjamin's History of Electricity 8vo, 3 00 Betts's Lead Refining and Electrolysis 8vo, 4 00 Classen's Quantitative Chemical Analysis by Electrolysis. (Boltwood.).Svo, ' 3 00 * Collins's Manual of Wireless Telegraphy and Telephony 12mo, 1 50 Crehore and Squier's Polarizing Photo-chronograph 8vo, 3 00 * Danneel's Electrochemistry. (Merriam.) 12mo, 1 25 Dawson's "Engineering" and Electric Traction Pocket-book. . . . 16mo, mor. 5 00 Dolezalek's Theory of the Lead Accumulator (Storage Battery), (von Ende.) 12mo, 2 50 Duhem's Thermodynamics and Chemistry. (Burgess.) 8vo, 4 00 Flather's Dynamometers, and the Measurement of Power 12mo, 3 00 * Getman's Introduction to Physical Science 12mo, 1 50 Gilbert's De Magnete. (Mottelay ) 8vo, 2 50 * Hanchett's Alternating Currents 12mo, 1 00 Hering's Ready Reference Tables (Conversion Factors) 16mo, mor. 2 50 * Hobart and Ellis's High-speed Dynamo Electric Machinery 8vo, 6 00 Holman's Precision of Measurements 8vo, 2 00 Telescopic Mirror-scale Method, Adjustments, and Tests.. . .Large 8vo, 75 Hutchinson's High Efficiency Electrical Illuminants and Illumination. (In Press.) Karapetoff' s Experimental Electrical Engineering: * Vol. 1 8vo, 3 50 * Vol. II 8vo, 2 50 Kinzbrunner's Testing of Continuous-current Machines .8vo, 2 OO 1 Landauer's Spectrum Analysis. (Tingle.) 8vo, 3 00 Le Chatelier's High-temperature Measurements. (Boudouard Burgess.) 12mo, 3 00 Lob's Electrochemistry of Organic Compounds. (Lorenz.) 8vo, 3 00 * Lyndon's Development and Electrical Distribution of Water Power. .8vo, 3 00 * Lyons's Treatise on Electromagnetic Phenomena. Vols, I .and II. 8vo, each, 6 00 * Michie's Elements of Wave Motion Relating to Sound and Light 8vo, 4 OO Morgan's Outline of the Theory of Solution and its Results 12mo, 1 00 * Physical Chemistry for Electrical Engineers 12mo, 1, 50 * Norris's Introduction to the Study of Electrical Engineering 8vo, 2 50 Norris and Dennison's Course of Problems on the Electrical Characteristics of Circuits and Machines. (In Press.) * Parshall and Hobart's Electric Machine Design 4to, half mor, 12 50 Reagan's Locomotives: Simple, Compound, and Electric. New Edition. Large 12mo, 3 50 * Rosenberg's Electrical Engineering. (Haldane Gee Kinzbrunner.) . .8vo, 2 00 Ryan, Norris, and Hoxie's Electrical Machinery. Vol. 1 8vo, 2 50 Schapper's Laboratory Guide for Students in Physical Chemistry 12mo, 1 00 * Tillman's Elementary Lessons in Heat 8vo, 1 50 * Timbie's Elements of Electricity Large 12mo, 2 00 Tory and Pitcher's Manual of Laboratory Physics Large 12mo, 2 00 Ulke's Modern Electrolytic Copper Refining 8vo, 3 00 LAW. * Brennan's Hand-book of Useful Legal Information for Business Men. 16mo, mor. 5 00 * Davis's Elements of Law 8vo, 2 50 * Treatise on the Military Law of United States 8vo, 7 00 * Dudley's Military Law and the Procedure of Courts-martial. . Large 12mo, 2 50 Manual for Courts-martial 16mo, mor. 1 50 Wait's Engineering and Architectural Jurisprudence 8vo, 6 00 Sheep, 6 50 11 Wait's Law of Contracts 8vo, $3 00 Law of Operations Preliminary to Construction in Engineering and Architecture 8vo, 5 00 Sheep, 5 50 Baker's Elliptic Functions 8vo, 1 50 MATHEMATICS. Briggs s Elements of Plane Analytic Geometry. (B6cher.) 12mo, 1 00 * Buchanan's Plane and Spherical Trigonometry 8vo, 1 00 Byerly's Harmonic Functions 8vo, 1 00 Chandler's Elements of the Infinitesimal Calculus 12mo, 2 00 * Coffin's Vector Analysis 12mo, 2 50 Compton's Manual of Logarithmic Computations 12mo, 1 50 * Dickson's College Algebra Large 12mo, 1 50 * Introduction to the Theory of Algebraic Equations Large 12mo, 1 25 Emch's Introduction to Projective Geometry and its Application 8vo, 2 50 Fiske's Functions of a Complex Variable 8vo, 1 00 Ha,lsted's Elementary Synthetic Geometry 8vo, 1 50 Elements of Geometry 8vo, 1 75 * Rational Geometry 12mo, 1 50 Synthetic Projective Geometry 8vo, 1 00 * Hancock's Lectures on the Theory of Elliptic Functions 8vo, 5 00 Hyde's Grassmann's Space Analysis 8vo, 1 00 * Johnson's (J. B.) Three-place Logarithmic Tables: Vest-pocket size, paper, 15 * 100 copies, 5 00 * Mounted on heavy cardboard, 8 X 10 inches, 25 * 10 copies, 2 00 Johnson's (W. W.) Abridged Editions of Differential and Integral Calculus. Large 12mo, 1 vol. 2 50 Curve Tracing in Cartesian Co-ordinates 12mo, 1 00 Differential Equations 8vo, 1 00 Elementary Treatise on Differential Calculus Large 12mo, 1 60 Elementary Treatise on the Integral Calculus Large 12mo, 1 50 * Theoretical Mechanics 12mo, 3 00 Theory of Errors and the Method of Least Squares 12mo, 1 50 Treatise on Differential Calculus Large 12mo, 3 00 Treatise on the Integral Calculus Large 12mo, 3 00 Treatise on Ordinary and Partial Differential Equations. . .Large 12mo, 3 50 Karapetoff's Engineering Applications of Higher Mathematics. (In Preparation.) Laplace's Philosophical Essay on Probabilities. (Truscott and Emory.) . 12mo, 2 00 * Ludlow's Logarithmic and Trigonometric Tables 8vo, 1 00 * Ludlow and Bass's Elements of Trigonometry and Logarithmic and Other Tables 8vo. 3 00 * Trigonometry and Tables published separately Each, 2 00 Macfarlane's Vector Analysis and Quaternions 8vo, 1 00 McMahon's Hyperbolic Functions 8vo, 1 00 Manning's Irrational Numbers and their Representation by Sequences and Series 12mo, 1 25 Mathematical Monographs. Edited by Mansfield Merriman and Robert S. Woodward Octavo, each 1 00 No. 1. History of Modern Mathematics, by David Eugene Smith. No. 2. Synthetic Projective Geometry, by George Bruce Halsted. No. 3. Determinants, by Laenas Gifford Weld. No. 4. Hyper- bolic Functions, by James McMahon. No. 5. Harmonic Func- tions, by William E. Byerly. No. 6. Grassmann's Space Analysis, by Edward W. Hyde. No. 7. Probability and Theory of Errors, by Robert S. Woodward. No. 8. Vector Analysis and Quaternions, by Alexander Macfarlane. No. 9. Differential Equations, by William Woolsey Johnson. No. 10. The Solution of Equations, by Mansfield Merriman. No. 11. Functions of a Complex Variable, by Thomas S. Fiske. Maurer's Technical Mechanics 8vo, 4 00 Merriman's Method of Least Squares 8vo, 2 00 Solution of Equations 8vo. 1 00 12 Moritz's Elements of Plane Trigonometry. (In Press.) Rice and Johnson's Differential and Integral Calculus. 2 vols. in one. Large 12mo, $1 50 Elementary Treatise on the Differential Calculus Large 12mo, 3 00 Smith's History of Modern Mathematics 8vo, 1 00 * Veblen and Lennes's Introduction to the Real Infinitesimal Analysis of One Variable 8vo, 2 00 * Waterbury's Vest Pocket Hand-book of Mathematics for Engineers. 2$X5f inches, rnor. 1 00 * Enlarged Edition, Including Tables mor. 1 50 Weld's Determinants 8vo, 1 00 Wood's Elements of Co-ordinate Geometry 8vo, 2 00 Woodward's Probability and Theory of Errors 8vo, 1 00 MECHANICAL ENGINEERING. MATERIALS OP ENGINEERING. STEAM-ENGINES AND BOILERS. Bacon's Forge Practice 12mo, 1 50 Baldwin's Steam Heating for Buildings 12mo, 2 50 Barr's Kinematics of Machinery 8vo, 2 50 * Bartlett's Mechanical Drawing 8vo, 3 00 * " Abridged Ed 8vo, 1 50 * Bartlett and Johnson's Engineering Descriptive Geometry 8vo, 1 50 * Burr's Ancient and Modern Engineering and the Isthmian Canal 8vo, 3 50 Carpenter's Experimental Engineering 8vo, 6 00 Heating and Ventilating Buildings 8vo, 4 00 * Clerk's The Gas, Petrol and Oil Engine 8vo, 4 00 Compton's First Lessons in Metal Working 12mo, 1 50 Compton and De Groodt's Speed Lathe 12mo, 1 50 Coolidge's Manual of Drawing 8vo, paper, 1 00 Coolidge and Freeman's Elements of General Drafting for Mechanical En- gineers Oblong 4to, 2 50 Cromwell's Treatise on Belts and Pulleys 12mo, 1 50 Treatise on Toothed Gearing 12mo, 1 50 Dingey's Machinery Pattern Making 12mo, 2 00 Durley 's Kinematics of Machines 8vo, 4 00 Flanders's Gear-cutting Machinery Large 12mo, 3 00 Flather's Dynamometers and the Measurement of Power 12mo, 3 00 Rope Driving 12mo, 2 00 Gill's Gas and Fuel Analysis for Engineers 12mo, 1 25 Goss's Locomotive Sparks '. 8vo, 2 00 Greene's Pumping Machinery. (In Preparation.) Hering's Ready Reference Tables (Conversion Factors) 16mo, mor. 2 50 * Hobart and Ellis's High Speed Dynamo Electric Machinery. 8vo, 6 00 Hutton's Gas Engine 8vo, 5 00 Jamison's Advanced Mechanical Drawing 8vo, 2 00 Elements of Mechanical Drawing 8vo, 2 50 Jones's Gas Engine 8vo, 4 00 Machine Design: Part I. Kinematics of Machinery '. 8vo, 1 50 Part II. Form, Strength, and Proportions of Parts 8vo, 3 00 Kaup's Text-book on Machine Shop Practice. (In Press.) * Kent's Mechanical Engineer's Pocket-Book 16mo, mor. 5 00 Kerr's Power and Power Transmission 8vo, 2 00 * Kimball and Barr's Machine Design 8vo, 3 00 Leonard's Machine Shop Tools and Methods 8vo, 4 00 * Levin's Gas Engine 8vo, 4 00 * Lorenz's Modern Refrigerating Machinery. (Pope, Haven, and Dean). .8vo, 4 00 MacCord's Kinematics; or, Practical Mechanism 8vo, 5 00 Mechanical Drawing 4to, 4 00 Velocity Diagrams 8vo, 1 50 MacFarland's Standard Reduction Factors for Gases 8vo, 1 50 .Mahan's Industrial Drawing. (Thompson.) 8vo, 3 50 13 Mehrtens's Gas Engine Theory and Design Large 12mo, $2 50 Oberg's Handbook of Small Tools Large 12mo, 2 50 * Parshall and Hobart's Electric Machine Design. Small 4to, half leather, 12 50 * Peele's Compressed Air Plant for Mines. Second Edition, Revised and En- larged 8vo, 3 50 Poole's Calorific Power of Fuels 8vo, 3 00 * Porter's Engineering Reminiscences, 1855 to 1882 8vo, 3 00 * Reid's Mechanical Drawing. (Elementary and Advanced.) 8vo, 2 00 Text-book of Mechanical Drawing and Elementary Machine Design. 8vo, 3 00 Richards's Compressed Air 12mo, 1 50 Robinson's Principles of Mechanism v j 8vo, 3 00 Schwamb and Merrill's Elements of Mechanism 8vo, 3 00 Smith (A. W.) and Marx's Machine Design 8vo, 3 00 Smith's (O.) Press-working of Metals 8vo, 3 00 Sorel's Carbureting and Combustion in Alcohol Engines. (Woodward and Preston.) Large 12mo, 3 00 Stone's Practical Testing of Gas and Gas Meters *.- 8vo, 3 50 Thurston's Animal as a Machine and Prime Motor, and the Laws of Energetics. 12mo, 1 00 Treatise on Friction and Lost Work in Machinery and Mill Work. . .8vo, 3 00 * Tillson's Complete Automobile Instructor 16mo, 1 50 * Titsworth's Elements of Mechanical Drawing Oblong 8vo, 1 25 Warren's Elements of Machine Construction and Drawing 8vo, 7 50 * Waterbury's Vest Pocket Hand-book of Mathematics for Engineers. 2JX5| inches, mor. 100 * Enlarged Edition, Including Tables mor. 1 50 Weisbach's Kinematics and the Power of Transmission. (Herrmann Klein.) 8vo, 5 00 Machinery of Transmission and Governors. (Hermann Klein.). .8vo, 500 Wood's Turbines 8vo, 2 50 MATERIALS OF ENGINEERING. * Bovey's Strength of Materials and Theory of Structures 8vo, 7 50 Burr's Elasticity and Resistance of the Materials of Engineering 8vo, 7 50 Church's Mechanics of Engineering 8vo, 6 00 * Greene's Structural Mechanics 8vo, 2 50 * Holley's Lead and Zinc Pigments Large 12mo 3 00 Holley and Ladd's Analysis of Mixed Paints, Color Pigments, and Varnishes. Large 12mo, 2 50 Johnson's (C. M.) Rapid Methods for the Chemical Analysis of Special Steels, Steel-Making Alloys and Graphite Large 12mo, 3 00 Johnson's (J. B.) Materials of Construction 8vo, 6 00 Keep's Cast Iron 8vo, 2 50 Lanza's Applied Mechanics 8vo, 7 50 Lowe's Paints for Steel Structures 12mo, 1 00 Maire's Modern Pigments and their Vehicles 12mo, 2 00 Maurer's Technical Mechanics - 8vo, 4 00 Merriman's Mechanics of Materials 8vo, 5 00' * Strength of Materials 12mo, 1 00 Metcalf 's Steel. A Manual for Steel-users 12mo, 2 00 Sabin's Industrial and Artistic Technology of Paint and Varnish 8vo, 3 00 Smith's ((A. W.) Materials of Machines 12mo, 1 00 * Smith's (H. E.) Strength of Material 12mo, 1 25 Thurston's Materials of Engineering 3 vols., 8vo, 8 00 Part I. Non-metallic Materials of Engineering 8vo, 2 00 Part II. Iron and Steel 8vo, 3 50 Part III. A Treatise on Brasses, Bronzes, and Other Alloys and their Constituents 8vo, 2 50 Wood's (De V.) Elements of Analytical Mechanics 8vo. 3 00 Treatise on the Resistance of Materials and an Appendix on the Preservation of Timber 8vo, 2 00- Wood's (M. P.) Rustless Coatings: Corrosion and Electrolysis of Iron and Steel 8vo, 4 00 14 STEAM-ENGINES AND BOILERS. Berry's Temperature-entropy Diagram 12mo, $2 00 Carnot's Reflections on the Motive Power of Heat. (Thurston.) 12mo, 1 50 Chase's Art of Pattern Making 12mo, 2 50 Creighton's Steam-engine and other Heat Motors 8vo, 5 00 Dawson's "Engineering" and Electric Traction Pocket-book. .. . Idmo, mor. 5 00 * Gebhardt's Steam Power Plant Engineering 8vo, 6 00 Goss's Locomotive Performance 8vo, 5 00 Hemenway's Indicator Practice and Steam-engine Economy 12mo, 2 00 Hutton's Heat and Heat-engines 8vo, 5 00 Mechanical Engineering of Power Plants 8vo, 5 00 Kent's Steam Boiler Economy 8vo, 4 00 Kneass's Practice and Theory of the Injector 8vo, 1 50 MacCord's Slide-valves .8vo, 2 00 Meyer's Modern Locomotive Construction 4to, 10 00 Moyer's Steam Turbine 8vo, 4 00 Peabody's Manual of the Steam-engine Indicator 12mo, 1 50 Tables of the Properties of Steam and Other Vapors and Temperature- Entropy Table .8vo, 1 00 Thermodynamics of the Steam-engine and Other Heat-engines. . . .8vo, 5 00 Valve-gears for Steam-engines 8vo, 2 50 Peabody and Miller's Steam-boilers 8vo, 4 00 Pupin's Thermodynamics of Reversible Cycles in Gases and Saturated Vapors. (Osterberg.) 12mo, 1 25 Reagan's Locomotives: Simple, Compound, and Electric. New Edition. Large 12mo, 3 50 Sinclair's Locomotive Engine Running and Management 12mo, 2 00 Smart's Handbook of Engineering Laboratory Practice .- 12mo, 2 50 Snow's Steam-boiler Practice 8vo, 3 00 Spangler's Notes on Thermodynamics 12mo, 1 00 Valve-gears 8vo, 2 50 Spangler, Greene, and Marshall's Elements of Steam-engineering 8vo, 3 00 Thomas's Steam-turbines 8vo, 4 00 Thurston's Handbook of Engine and Boiler Trials, and the Use of the Indi- cator and the Prony Brake 8vo, 5 00. Handy Tables : . .8vo, 1 50 Manual of Steam-boilers, their Designs, Construction, and Operation 8vo, 5 00 Manual of the Steam-engine 2 vols., 8vo, 10 00 Part I. History, Structure, and Theory 8vo, 6 00 Part II. Design, Construction, and Operation 8vo, 6 00 Wehrenfennig's Analysis and Softening of Boiler Feed-water. (Patterson.) 8vo, 4 00 Weisbach's Heat, Steam, and Steam-engines. (Du Bois.) 8vo, 5 00 Whitham's Steam-engine Design 8vo, 5 00 Wood's Thermodynamics, Heat Motors, and Refrigerating Machines. . .8vo, 4 00 MECHANICS PURE AND APPLIED. Church's Mechanics of Engineering 8vo, 6 00 * Mechanics of Internal Works 8vo, 1 50 Notes and Examples in Mechanics 8vo, 2 00 Dana's Text-book of Elementary Mechanics for Colleges and Schools .12mo, 1 50 Du Bois's Elementary Principles of Mechanics: Vol. I. Kinematics 8vo, 3 50 Vol. II. Statics 8vo. 4 00 Mechanics of Engineering. Vol. I Small 4to, 7 50 Vol. II Small 4to, 10 00 * Greene's Structural Mechanics 8vo, 2 50 * Hartmann's Elementary Mechanics for Engineering Students 12mo, 1 25 James's Kinematics of a Point and the Rational Mechanics of a Particle. Large 12mo. 2 00 * Johnson's (W. W.) Theoretical Mechanics 12mo, 3 00 Lanza's Applied Mechanics 8vo, 7 50 15 * Martin's Text Book on Mechanics, Vol. I, Statics 12mo, $1 25 * Vol. II, Kinematics and Kinetics. 12mo, 1 50 Maurer's Technical Mechanics 8vo. 4 00 * Merriman's Elements of Mechanics 12mo, 1 00 Mechanics of Materials 8vo, 5 00 * Michie's Elements of Analytical Mechanics 8vo, 4 00 Robinson's Principles of Mechanism 8vo, 3 00 Sanborn's Mechanics Problems Large 12mo, 1 50 Schwamb and Merrill's Elements of Mechanism 8vo, 3 00 Wood's Elements of Analytical Mechanics. 8vo, 3 00 Principles of Elementary Mechanics 12mo, 1 25 ' . ** MEDICAL. * Abderhalden's Physiological Chemistry in Thirty Lectures. (Hall and Defren.) *. 8vo, 5 00 von Behring's Suppression of Tuberculosis. (Bolduan.) 12mo, 1 00 Bolduan's Immune Sera ." 12mo, 1 50 Bordet's Studies in Immunity. (Gay.) 8vo, 6 00 * Chapin's The Sources and Modes of Infection Large 12mo, 3 00 Davenport's Statistical Methods with Special Reference to Biological Varia- tions 16mo, mor. 1 50 Ehrlich's Collected Studies on Immunity. (Bolduan.) 8vo, 6 00 * Fischer's Oedema 8vo, 2 00 * Physiology of Alimentation . Large 12mo, 2 00 de Fursac's Manual of Psychiatry. (Rosanoff and Collins.) Large 12mo, 2 50 Hammarsten's Text-book on Physiological Chemistry. (Mandel.) 8vo, 4 00 Jackson's Directions for Laboratory Work in Physiological Chemistry. .8vo, 1 25 Lassar-Cohn's Praxis of Urinary Analysis. (Lorenz.) 12mo, 1 00 Mandel's Hand-book for the Bio-Chemical Laboratory 12mo, 1 50 * Nelson's Analysis of Drugs and Medicines 12mo, 3 00 * Pauli's Physical Chemistry in the Service of Medicine. (Fischer.) ..12mo, 1 25 * Pozzi-Escot's Toxins and Venoms and their Antibodies. (Cohn.). . 12mo, 1 00 Rostoski's Serum Diagnosis. (Bolduan.) 12mo, 1 00 Ruddiman's Incompatibilities in Prescriptions 8vo, 2 00 Whys in Pharmacy 12mo, 1 00 Salkowski's Physiological and Pathological Chemistry. (Orndorff.) .. ..8vo, 2 50 * Satterlee's Outlines of Human Embryology 12mo, 1 25 Smith's Lecture Notes on Chemistry for Dental Students 8vo, 2 50 * Whipple's Tyhpoid Fever Large 12mo, 3 00 * Woodhull's Military Hygiene for Officers of the Line Large 12mo, 1 50 * Personal Hygiene 12mo, 1 00 Worcester and Atkinson's Small Hospitals Establishment and Maintenance, and Suggestions for Hospital Architecture, with Plans for a Small Hospital 12mo, 1 25 METALLURGY. Betts's Lead Refining by Electrolysis 8vo, 4 00 Bolland's Encyclopedia of Founding and Dictionary of Foundry Terms used in the Practice of Moulding 12mo, 3 00 Iron Founder 12mo, 2 50 Supplement 12mo, 2 50 Borchers's Metallurgy. (Hall and Hayward.) (In Press.) Douglas's Untechnical Addresses on Technical Subjects 12mo, 1 00 Goesel's Minerals and Metals: A Reference Book 16mo, mor. 3 00 * Iles's Lead-smelting 12mo, 2 50 Johnson's Rapid Methods for the Chemical Analysis of Special Steels, Steel-making Alloys and Graphite Large 12mo, 3 00 Keep's Cast Iron 8vo, 2 50 Le Chatelier's High- temperature Measurements. (Boudouard Burgess.) 12mo, 3 00 Metcalf's Steel. A Manual for Steel-users 12mo, 2 00 Minet's Production of Aluminum and its Industrial Use. (Waldo.). . 12mo, 2 50 * Ruer's Elements of Metallography. (Mathewson.) 8vo, 3 00 16 Smith's Materials of Machines 12mo, $1 00 Tate and Stone's Foundry Practice 12mo, 2 00 Thurston's Materials of Engineering. In Three Parts 8vo, 8 00 Part I. Non-metallic Materials of Engineering, see Civil Engineering, page 9. Part II. Iron and Steel 8vo, 3 50 Part III. A Treatise on Brasses, Bronzes, and Other Alloys and their Constituents 8vo, 2 50 Ulke's Modern Electrolytic Copper Refining 8vo, 3 00 West's American Foundry Practice 12mo, 2 50 Moulders' Text Book 12mo. 2 50 MINERALOGY. Baskerville's Chemical Elements. (In Preparation.) * Browning's Introduction to the Rarer Elements 8vo, 1 50 Brush's Manual of Determinative Mineralogy. (Penfield.) 8vo, 4 00 Butler's Pocket Hand-book of Minerals 16mo, mor. 3 00 Chester's Catalogue of Minerals 8vo, paper, 1 00 Cloth, 1 25 * Crane's Gold and Silver 8vo, 5 00 Dana's First Appendix to Dana's New "System of Mineralogy". .Large 8vo, 1 00 Dana's Second Appendix to Dana's New " System of Mineralogy." Large 8vo, 1 50 Manual of Mineralogy and Petrography 12mo, 2 00 Minerals and How to Study Them 12mo, 1 50 System of Mineralogy Large 8vo, half leathor, 12 50 Text-book of Mineralogy 8vo, 4 00 Douglas's Untechnical Addresses on Technical Subjects 12mo, 1 00 Eakle's Mineral Tables 8vo, 1 25 Eckel's Stone and Clay Products Used in Engineering. (In Preparation.) Goesel's Minerals and Metals: A Reference Book 16mo, mor. 3 00 Groth's The Optical Properties of Crystals. (Jackson.) (In Press.) Groth's Introduction to Chemical Crystallography (Marshall) 12mo, 1 25 * Hayes's Handbook for Field Geologists 16mo, mor. 1 50 Iddings's Igneous Rocks 8vo, 5 00 Rock Minerals 8vo, 5 00 Johannsen's Determination of Rock-forming Minerals in Thin Sections. 8vo, With Thumb Index 5 00 * Martin's Laboratory Guide to Qualitative Analysis with the Blow- Pipe 12mo, 60 Merrill's Non-metallic Minerals: Their Occurrence and Uses 8vo, 4 00 Stones for Building and Decoration 8vo, 5 00 * Penfield's Notes on Determinative Mineralogy and Record of Mineral Tests. 8vo, paper, 50 Tables of Minerals, Including the Use of Minerals and Statistics of Domestic Production. . 8vo, 1 00 * Pirsson's Rocks and Rock Minerals 12mo, 2 50 * Richards's Synopsis of Mineral Characters 12mo, mor. 1 25 * Ries's Clays: Their Occurrence, Properties and Uses 8vo, 5 00 * Ries and Leighton's History of the Clay-working Industry of the United States 8vo, 2 50 Rowe's Practical Mineralogy Simplified. (In Press.) * Tillman's Text-book of Important Minerals and Rocks 8vo, 2 00 Washington's Manual of the Chemical Analysis of Rocks 8vo, 2 00 MINING. * Beard's Mine Gases and Explosions Large 12mo, 3 00 * Crane's Gold and Silver 8vo, 5 00 * Index of Mining Engineering Literature 8vo, 4 00 * 8vo, mor. 5 00 * Ore Mining Methods 8vo. 3 00 Douglas's Untechnical Addresses on Technical Subjects 12mo, 1 00 Eissler's Modern High Explosives 8vo, 4 00 17 Goesel's Minerals arid Metals: A Reference Book 16mo, mor. $3 00 Ihlseng's Manual of Mining 8vo, 5 00 * Iles's Lead Smelting 12mo, 2 00 Peele's Compressed Air Plant for Mines 8vo, 3 00 Riemer's Shaft Sinking Under Difficult Conditions. (Corning and Peele.)8vo, 3 00 * Weaver's Military Explosives 8vo, 3 00 Wilson's Hydraulic and Placer Mining. 2d edition, rewritten 12mo, 2 50 Treatise on Practical and Theoretical Mine Ventilation 12mo, 1 25 SANITARY SCIENCE. Association of State and National Food and Dairy Departments, Hartford Meeting, 1906 8vo, 3 00 Jamestown Meeting, 1907 8vo, 3 00 * Bashore's Outlines of Practical Sanitation .*.... 12mo, 1 25 Sanitation of a Country House 12mo, 1 00 Sanitation of Recreation Camps and Parks 12mo, 1 00 * Chapin's The Sources and Modes of Infection Large 12mo, 3 00 Folwell's Sewerage. (Designing, Construction, and Maintenance.) 8vo, 3 00 Water-supply Engineering 8vo, 4 00 Fowler's Sewage Works Analyses 12mo, 2 00 Fuertes's Water-filtration Works 12mo, 2 50 Water and Public Health 12mo, 1 50 Gerhard's Guide to Sanitary Inspections 12mo, 1 50 * Modern Baths and Bath Houses 8vo, 3 00 Sanitation of Public Buildings 12mo, 1 50 * The Water Supply, Sewerage, and Plumbing of Modern City Buildings. 8vo, 4 00. Hazen's Clean Water and How to Get I ' Large 12mo, 1 50 Filtration of Public Water-supplies 8vo, 3 00 * Kinnicutt, Winslow and Pratt's Sewage Disposal 8vo, 3 00 Leach's Inspection and Analysis of Food with Special Reference to State Control 8vo, 7 50 Mason's Examination of Water. (Chemical and Bacteriological) 12mo, 1 25 Water-supply. (Considered principally from a Sanitary Standpoint). 8vo, 4 00 Mast's Light and the Behavior of Organisms. (In Press.) * Merriman's Elements of Sanitary Engineering 8vo, 2 00 Ogden's Sewer Construction 8vo, 3 00 Sewer Design 12mo, 2 00 Parsons's Disposal of Municipal Refuse 8vo, 2 00 Prescott and Winslow's Elements of Water Bacteriology, with Special Refer- ence to Sanitary Water Analysis 12md, 1 50 * Price's Handbook on Sanitation 12mo, 1 50 Richards's Conservation by Sanitation. (In Press.) Cost of Cleanness 12mo, 1 00 Cost of Food. A Study in Dietaries 12mo, 1 00 Cost of Living as Modified by Sanitary Science 12mo, 1 00 Cost of Shelter 12mo, . 1 00 * Richards and Williams's Dietary Computer 8vo, 1 50 Richards and Woodman's Air, Water, and Food from a Sanitary Stand- point 8vo, 2 00 * Richey's Plumbers', Steam-fitters', and Tinners' Edition (Building Mechanics' Ready Reference Series) IGmo, mor. 1 50 Rideal's Disinfection and the Preservation of Food 8vo, 4 00 Sewage and Bacterial Purification of Sewage 8vo, 4 00 Soper's Air and Ventilation of Subways 12mo, 2 50 Turneaure and Russell's Public Water-supplies 8vo, 5 00 Venable's Garbage Crematories in America 8vo, 2 00 Method and Devices for Bacterial Treatment of Sewage 8vo, 3 00 Ward and Whipple's Freshwater Biology. (In Press.) Whipple's Microscopy of Drinking-water 8vo, 3 50 * Typhoid Fever Large 12mo, 3 00 Value of Pure Water Large 12mo, 1 00 Winslow's Systematic Relationship of the Coccaceae Large 12mo, 2 50 18 MISCELLANEOUS. Emmons's Geological Guide-book of the Rocky Mountain Excursion of the International Congress of Geologists Large 8vo. $1 5O Ferrel's Pooular Treatise on the Winds 8vo, 4 OO Fitzgerald's Boston Machinist 18mo, 1 OQ Gannett's Statistical Abstract of the World 24mo, 75 Haines's American Railway Management 12mo, 2 5O Hanausek's The Microscopy of Technical Products. (Winton) 8vo, 5 00 Jacobs's Betterment Briefs. A Collection of Published Papers on Or- ganized Industrial Efficiency 8vo, 3 50- Metcalfe's Cost of Manufactures, and the Administration of Workshops.. 8vo, 5 OQ Putnam's Nautical Charts 8vo, 2 00- Ricketts's History of Rensselaer Polytechnic Institute 1824-1894. Large 12mo, 3 00 Rotherham's Emphasised New Testament Large 8vo, 2 00 Rust's Ex-Meridian Altitude, Azimuth and Star-finding Tables 8vo 5 00 Standage's Decoration of Wood, Glass, Metal, etc , . < 12mo 2 00 Thome's Structural and Physiological Botany. (Bennett) 16mo, 2 25 Westermaier's Compendium of General Botany. (Schneider) 8vo, 2 00' Winslow's Elements of Applied Microscopy 12mo, 1 50"' HEBREW AND CHALDEE TEXT-BOOKS. Gesenius's Hebrew and Chaldee Lexicon to the Old Testament Scriptures. (Tregelles.) Small 4to, half mor, 5 00 Green's Elementary Hebrew Grammar 12mo, 1 25 19 UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. " i h 1951 SEP 28 1996 LD 21-100m-9,'48(B399sl6)476 \ V_/ \_^T I W U. C. BERKELEY LIBRARIES 813219 Library UNIVERSITY OF CALIFORNIA LIBRARY