m 
 
 
 »wM i i&iMSs 
 
' * . s * <.vy 
 
 
 
 
 
 ■ ^ 
 
 ':/ ; 
 
 
 : \^- ^ '•" 
 
 ,^ "%. 
 
 '<£> ^ 
 
 
 A ■ < " ' C- V' % S 
 
 
 - %; 
 
 %*** :i 
 
 - ^>.^ v "-V 
 
 
 
 
 .0 O 
 
 
 \ \^«-y / 
 
 
 
 
 
 
 
 -^ 
 
 *•» t'\ 
 
 
 " 
 
 ^ 
 
 

 
 
 
 
 
 
 
 ".**' 
 
 
 
 
 
 
 N q. 
 
 '->" 
 
 
 
 
 
 
 ^ ^ 
 
 /W 
 
 
 v- V 
 
 .0- -o. '" •» , 
 
 \^' 
 
 ■ k -v\ 
 
 
STEEL RAILS 
 
 THEIR HISTORY, PROPERTIES, STRENGTH 
 AND MANUFACTURE 
 
 WITH NOTES ON THE 
 
 PRINCIPLES OF ROLLING STOCK AND TRACK DESIGN 
 
 WILLIAM H. SELLEW 
 
 PRINCIPAL ASSISTANT ENGINEER, MICHIGAN CENTRAL RAILROAD 
 
 361 ILLUSTRATIONS -83 FOLDING PLATES 
 
 NEW YORK 
 D. VAN NOSTRAND COMPANY 
 
 Twenty-five Park Place 
 LONDON 
 
 CONSTABLE & COMPANY, Ltd. 
 1913 
 
.S4 
 
 Copyright, 1913 
 D. VAN NOSTRAND COMPANY 
 
 \ 
 
 
 ( 
 
 Stanbope ipress 
 
 . H.GILSON COMPANY 
 
 SCi.A:M6058 
 
PREFACE 
 
 In this work the author has endeavored to systematize the knowledge in 
 existence upon the subject, and to present in a concise yet clear form the most 
 important features of the problem. 
 
 The first chapter treats of the development of the present design of section 
 with a comparison of the American rails with those in use on English Railways 
 and on the Continent. 
 
 In chapters two to five, inclusive, the external forces acting on the rail 
 and the corresponding stresses they produce in the rail are discussed. The 
 necessity and desire for information on this subject are widespread. While a 
 considerable amount of general information is to be found scattered through 
 the technical press and in the proceedings of the various Railway Associations 
 and Engineering Societies,, yet very little has been published dealing broadly 
 with the principles of design of the jail in reference to the rolling stock and 
 track structure. 
 
 In recent years much thought has been given to the manufacture of rail 
 steel, and investigators, it would seem, have turned their attention more to an 
 examination of the various defects found in the process of manufacture than 
 to the study of the duty of the rail. 
 
 Quite early the question of the intensity of pressure existing between the 
 wheel and the rail began to receive attention, but it was not until later that the 
 bending stresses in the rail were investigated. Purely theoretical contributions 
 to the latter subject were made by Zimmerman in 1888. The first practical 
 investigations of the bending stress in the rail were apparently those made 
 by the United States Government in 1894 by measuring the strains in the rail 
 under the static load of the locomotive wheels. These were followed by 
 Dr. P. H. Dudley's stremmatograph experiments for measuring the effect 
 of dynamic loads. In the time elapsed since the publication of these investi- 
 gations hardly anything has been done to further elucidate this problem. 
 
 The sixth chapter deals with the detail of manufacture of the rail. The 
 different stages in the process are described and the influence of each upon the 
 finished product is pointed out. It would be outside the limits of the present 
 
iv PREFACE 
 
 work to attempt a complete treatise on the manufacture of steel; the discussion 
 concerns itself, therefore, chiefly with the practical results obtained rather than 
 with theoretical considerations. 
 
 In the last chapter are given Rail Specifications representing the best 
 modern practice in this country and abroad. The forms recommended by the 
 American Railway Engineering Association for reports and record of the rail 
 are added for the sake of completeness in an appendix. 
 
 The greater part of the bibliography of rail specifications given in article 
 39 was prepared for the present work by the Secretary of the American Soci- 
 ety of Civil Engineers. The other shorter bibliographies appended to the 
 discussions of several of the subjects were compiled by Mr. McClelland, 
 Technology Librarian of the Carnegie Library of Pittsburgh. These, which 
 are intended to supplement certain parts of the text, are not exhaustive but 
 are thought to contain most of the important articles since 1906 which come 
 within their scope. A bibliography for the years 1870-1906 with chronologi- 
 cal arrangement appears in the Transactions of the American Institute of 
 Mining Engineers, Vol. 37, pp. 617-627. 
 
 A comprehensive bibliography of steel manufacture would be so extensive 
 as to be unwieldy. Exhaustive bibliographies on this subject appear in the 
 various volumes of the Journal of the Iron and Steel Institute, and a good 
 selective bibliography of iron and steel manufacture appears in Bradley 
 Stoughton's "Metallurgy of Iron and Steel." For the average reader who 
 desires a more detailed discussion of the processes of manufacture of steel, 
 Harbord's and Hall's excellent volume on the Metallurgy of Steel will gen- 
 erally be found sufficient. It is believed that these references, together with 
 the information contained in the footnotes throughout the book, will permit 
 a thorough examination of any of the subjects to be made. 
 
 The work is essentially a compilation. The author has, however, in every 
 case endeavored to give credit where anything has been drawn from an outside 
 source, and if he has been remiss in this respect it has been unintentional. In 
 the discussion of the granular structure of steel he has been much indebted to 
 the work of Mr. J. W. Mellor, from whose writings a considerable part of 
 article 25 has been taken. 
 
 The publications of the " Railway Age Gazette " have been freely quoted 
 from and the author wishes to express his appreciation of the courteous permis- 
 sion given for the use of this material and for the many quotations taken from 
 other sources, especially those from articles which have appeared in the Proceed- 
 ings of the American Railway Engineering Association. 
 
PREFACE v 
 
 He has much pleasure in expressing his indebtedness to the following gentle- 
 men who have given assistance in revising manuscript or proofs of the parts named : 
 Professor Gaetano Lanza, mathematical discussions of Chapters II to V inclusive; 
 Dr. A. B. Pierce, checking the author's calculations in these chapters; Professor 
 W. F. M. Goss, wheel pressures; Dr. Hermann von Schrenk, forestry; Professor 
 W. K. Hatt, strength of tie timber; Dr. P. H. Dudley, stremmatograph tests; 
 Mr. Harry D. Tiemann, impact; Mr. James E. Howard, repeated stress; Mr. A. 
 L. Colby, manufacture and specifications; Mr. Robert W. Hunt, influence of 
 detail of manufacture; Mr. Bradley Stoughton, effect of temperature during rolling, 
 and Mr. E. T. Howson for examination of the proofs before finally going to press. 
 
 WILLIAM H. SELLEW. 
 
 Detroit, Michigan, July, 1912. 
 
CONTENTS 
 
 Chapter I. — Development op the Present Section p age 
 
 1. Early sections 1 
 
 2. Present sections 14 
 
 Chapter II. — Pressure of the Wheel on the Rail 
 
 3. Speeds of modern locomotives 21 
 
 4 Weights of modern locomotives 29 
 
 5. Effect of excess balance and angularity of main rod 35 
 
 6. Effect of irregularities in the track 45 
 
 7. Effect of rocking of the engine 49 
 
 8. Effect of flat spots in the wheels 54 
 
 9. Impact tests 62 
 
 10. The dynamic augment of the wheel load 69 
 
 11. Electric locomotives 74 
 
 12. Cars 81 
 
 Chapter III. — Supports of the Rail 
 
 13. The tie 90 
 
 14. Bearing of the rail on the tie 122 
 
 15. Fastening of the rail to the tie 138 
 
 16. Strength of the tie 153 
 
 17. Bearing on the ballast 179 
 
 18. Bearing on the subgrade 180 
 
 19. Supporting power of the tie 188 
 
 Chapter IV. — Stresses in the Rail 
 
 20. Stress at point of contact of the wheel with the rail 193 
 
 21. Proposed solutions of the bending stress in the rail 210 
 
 22. Tests to determine the bending stress in the rail 218 
 
 23. Calculation of the bending and shearing stress in the rail 239 
 
 24. Effect of the joint 259 
 
 Chapter V. — -Strength of the Rail 
 
 25. Influence of stress and strain on the strength of the rail 270 
 
 26. Effect of low temperatures on the strength of the rail 284 
 
 27. Physical tests of the strength of the rail '. . . 288 
 
 28. The strength of the rail and proper weight for various conditions of loading . 310 
 
viii CONTENTS 
 
 Chapter VI. — Influence of Detail of Manufacture p AGB 
 
 29. Chemical composition 326 
 
 30. Extraction of the iron from its ore 344 
 
 31 . Conversion of the steel 366 
 
 32. The Ingot 395 
 
 33. Influence of mechanical work 420 
 
 Chapter VII. — Rail Specifications 
 
 34. Comparison of American specifications 463 
 
 35. Specifications (New York Central Lines) for basic open-hearth rails 478 
 
 36. British standard specifications of bull head railway rails 484 
 
 37. British standard specifications of flat bottom railway rails 488 
 
 38. Specifications for street railway rails 491 
 
 39. Bibliography of rail specifications 494 
 
 Appendix 
 
 Reports and records 501 
 
 Index 525 
 
LIST OF ILLUSTRATIONS 
 
 Fig. Page 
 
 1. Comparison of Rail Failures between Different Sections of Bessemer Steel Rails, from Oct. 31, 
 
 1908, to April 30, 1909 9 
 
 2. Increase in Axle Loads, 1885-1907 15 
 
 3. Classification of Locomotives 29 
 
 4. Progress in Locomotive Building 30 
 
 5. Decapod Locomotive of 1903 and American Type of 1857 31 
 
 6. Rail Pressures. Eight-wheel Engines 35 
 
 7. " " Ten-wheel Engines. (Light Weights.) 36 
 
 8. " " 442 (Atlantic) Type Engines 40 
 
 9. " " 462 (Pacific) Type Engines . . . , 41 
 
 10. " " 460 (Ten-wheel) Type Engines. (Heavy Weights.) 42 
 
 11. " " 260 (Mogul) Type Engines 43 
 
 12. " " 280 (Consolidation) Type Engines 44 
 
 13. Damaging Effect of Badly Balanced Locomotive , 39 
 
 14. Profile of Rail from Cuenot's Track Experiments 46 
 
 15. Rail Profile taken with a Railroad Automatic Track Inspector Machine 47 
 
 16. "Valley" or Local Depression in Track Profile 47 
 
 17. Summit between Two Depressions of Track Profile 47 
 
 18. Locomotive Driving Wheel Springs 48 
 
 19. Deflection of Locomotive Springs 49 
 
 20. Recording Device and Cab Controlling Mechanism for Testing Driving Wheel Springs 50 
 
 21. Recording Device in Place on Driving Wheel Spring 50 
 
 22. General Arrangement of Apparatus for Testing Driving Wheel Springs 51 
 
 23. Main Stylus used in Driving Wheel Spring Tests 51 
 
 24. Stress-strain Diagram. Locomotive Driving Wheel Springs 53 
 
 25. Flat Spot in Wheel 55 
 
 26. Irregularity in the Roundness of Present-day Chilled Car Wheels 58 
 
 27. Apparatus for Measuring the Effect of a Flat Spot 60 
 
 28. Diagram of Tests on Freight Car with Flat Wheels 61 
 
 29. Wire Tests 63 
 
 30. Deformation of Bridge Members under Passing Trains 64 
 
 31. Dynamic Wheel Loads of Typical Passenger Steam Locomotives 72 
 
 32. " " " " " Freight Steam Locomotives 73 
 
 33. Detroit River Tunnel Company's Locomotive 75 
 
 34. Pennsylvania" Electric Locomotive in Use in the New York Tunnels 76 
 
 35. Details Pennsylvania Electric Locomotive 77 
 
 36. Typical Load Diagrams for Electric Locomotives 78 
 
 37. Box Car 81 
 
 38. Flat Car 82 
 
 39. Gondola Car 82 
 
 40. Coke Car 83 
 
 41. Stock Car 83 
 
 42. Vestibuled Coach 83 
 
 43. Twelve-section Sleeping Car 84 
 
 44. Steel Combination Passenger and Baggage Car 84 
 
 45. Vestibuled Dining Car 84 
 
LIST OF ILLUSTRATIONS 
 
 e Car 84 
 
 47. Typical Load Diagrams for Cars 85 
 
 48. Typical Dynamic Load Diagrams for Motor Cars 85 
 
 49. 70-foot McKeen Motor Car 86 
 
 50. 70-foot General Electric Gas Electric Motor Car 86 
 
 51. Electric Railway Cars 87 
 
 52. Electric Railway Cars 88 
 
 53. Typical Load Diagrams for Electric Railway Cars 89 
 
 54. Carnegie Steel Tie : 90 
 
 55. Carnegie Steel Ties on the Bessemer and Lake Erie Railroad 91 
 
 56. Effect of Three Derailments on Steel Ties 91 
 
 57. Steel Tie after Four Years Service 92 
 
 58. Carnegie Steel Tie with Wedge Fastener 93 
 
 59. Hill Fastening on Carnegie Steel Tie 93 
 
 60. Hansen Steel Tie 94 
 
 61. Universal Metallic Tie on Pennsylvania Lines 95 
 
 62. Snyder Steel Tie 95 
 
 63. Buhrer Combined Steel and Wood Tie on L. S. and M. S. Ry 96 
 
 64. Mexican Railway Steel Tie 97 
 
 65. Buhrer Concrete Tie 98 
 
 66. Bottom Surface of Buhrer Concrete Tie 99 
 
 67. Section of Track on Chicago and Alton R.R., showing Kimball Tie 99 
 
 68. Kimball Tie put in Track on N. Y. C. &. St. L. R.R., July, 1904 100 
 
 69. Kimball Tie, showing Spiking Plugs 100 
 
 70. Percival Concrete Tie 101 
 
 71. Sarada Tie 102 
 
 72. Adriatic Railway Tie 102 
 
 73. Riegler Concrete Tie 103 
 
 74. " " " Appearance in the Track 103 
 
 75. Forest Regions of the United States 107 
 
 76. Hunnewell Plantation 109 
 
 77. Farlington Forest 110 
 
 78. Standard Prussian Ties of Baltic Pine 117 
 
 79. Standard Oak and Beech Ties on the French Eastern Railway 117 
 
 80. Distribution of Pressure from Tie Plate 118 
 
 81. Half Round Tie Proposed by the Forest Service 118 
 
 82. Spacing of Half-round Ties 118 
 
 83. Pole Tie . . 119 
 
 84. Extreme Form of Half-round Tie 120 
 
 85. Test on McKee Tie Plate 122 
 
 86. Wear of Tie under Tie Plate 123 
 
 87. Loblolly Pine Tie. Section of Tie under Rail Bearing 124 
 
 88. " " " Section from Middle 124 
 
 89. Belgian State Railways, 105-pound Rail and Tie Plate 125 
 
 90. " " " 115-pound Rail and Tie Plate 126 
 
 91. Kingdom of Wurttemberg State Railways, Tie Plate 127 
 
 92. Bavarian State Railways, Joint Hook Plate 128 
 
 93. Kingdom of Saxony State Railroad, Joint Hook Plate 129 
 
 94. Elsass-Lothringen State Railways, Tie Plate 130 
 
 95. Prussian State Railways, Tie Plate 131 
 
 96. Bavarian State Railways, Intermediate Wedge Plate 132 
 
 97. Wooden Tie Plate on French Eastern 132 
 
 98. Plain Bearing Plates, German Experiments on Tie Plates 134 
 
 99. Hook Plates, German Experiments on Tie Plates 135 
 
 100. Hook Plates with Clips, German ExDeriments on Tie Plates 135 
 
 101. Group 1, German Experiments on Tie Plates 136 
 
LIST OF ILLUSTRATIONS xi 
 
 Fig. Page 
 
 102. Group 2, German Experiments on Tie Plates 136 
 
 103. Group 3, German Experiments on Tie Plates 137 
 
 104. Short Leaf Pine Tie, after 2 Years' Service, cut through Spike Holes 138 
 
 105. Cross Section through the Spike Holes of Short Leaf Pine Tie 139 
 
 106. Common Spike 140 
 
 107. Common Screw Spike 140 
 
 108. Screw Spike used by Grand Duchy of Baden State Railways 140 
 
 109. Early French Screw Spikes 141 
 
 110. Machine Preparing Ties for Screw Spikes 142 
 
 111. Showing Application of Screw Spikes on A. T. & S. Fe R.R 142 
 
 112. French Railways — Rail Fastenings 143 
 
 113. German Railways — Rail Fastenings 145 
 
 114. English and Scotch Railways — Rail Fastenings 146 
 
 115. Screw Spike deduced from European Practice 147 
 
 116. French Screw Spike 149 
 
 117. Wooden Tie Plug used on French Railways 149 
 
 118. Collet Trenail 150 
 
 119. Cross Section of Pine Tie through Dowel 151 
 
 120. Three Ties of Baltic Pine on the Prussian State Railways 151 
 
 121. Comparative Resistance to Vertical Pressure of Screw Spikes in Pine Ties 152 
 
 122. " " " " " " " " " Beech Ties 152 
 
 123. Control Plan — Creosote Tie Tests 161 
 
 124. Tie Plate Forms used in Tests at Purdue University 170 
 
 125. Elastic Curve of Tie, 7 feet 10.4 inches long 172 
 
 126. " " " " 8 feet 10.3 inches long 172 
 
 127. Wood and Composite Ties used in Cuenot's Experiments 173 
 
 128. Measuring Apparatus for Ties under Static Load 174 
 
 129. " " " " " Dynamic Load 175 
 
 130. Results of M. Cuenot's Tests on Ties 176 
 
 131. Strain Diagram of Entire Tie 177 
 
 132. " " " Tie between Rails 178 
 
 133. " " " Tie outside of Rails 178 
 
 134. Ballast Experiments — Schubert. Six inches of sand and 6 inches of gravel 181 
 
 135. " " " Six inches of sand and 6 inches of stone 181 
 
 136. " " " Stone with thin layer of sand 182 
 
 137. " " " Stone resting on clay subgrade 182 
 
 138. Effect of Overloading the Subgrade 183 
 
 139. Pennsylvania Track Testing Apparatus 1 183 
 
 140. Distribution of Pressure to Subgrade 186 
 
 141. Bell's Apparatus for Measuring Depression of the Track 191 
 
 142. Reaction of Tie 191 
 
 143. Compression Modulus — Condition of Free Flow 193 
 
 144. " - " Partially Restricted Flow 193 
 
 145. " " Restricted Flow 193 
 
 146. Area of Contact between Wheel and Rail 195 
 
 147. Relation between Areas of Contact and Load on Wheel 196 
 
 148. Tire Wear, Ten-wheel Engines 198 
 
 149. " " Eight-wheel Engines 199 
 
 150. Two Pieces of a Worn 100-pound Rail after Testing 204 
 
 151. Reciprocating Machine for Testing Flow of Metal in Head of Rail 206 
 
 152. Section of 70-lb. Bessemer Rail Tested for Flow of Head 207 
 
 153. D : stribution of Tie Pressure under Rail 213 
 
 154. "Class I" Engine with 75 per cent Impact 214 
 
 155. Track Depression under "Class I" Loading 214 
 
 156. " Class K" Engine with 75 per cent Impact 215 
 
 157. Track Depression under " Class K" Loading 216 
 
xii LIST OF ILLUSTRATIONS 
 
 Fig. Page 
 
 158. Bending Moment of Rail placed on Ties 218 
 
 159. Railroad Track Experiments, Boston and Albany R.R 219 
 
 160. " " " Photograph of Leveling Instrument for Measuring the Depres- 
 
 sion of the Track 220 
 
 161. " " " Photograph of Micrometer for Determining the Fibre Stress in 
 
 the Base of the Rail 220 
 
 162. " " " C.B.&Q.R.R 222 
 
 163. Advance Wave Determinations 223 
 
 164. Movement of Rails Laid Alongside of Track 224 
 
 165. Railroad Track Experiments, View showing Micrometer for Measuring Strains in Rails, in 
 
 Position on Base of Rail under Driving Wheel 226 
 
 166. " " Pennsylvania R.R. Depression in Ballast 234 
 
 167. " " " " " Stress in Rail 235 
 
 168. Stremmatograph Tests at 19 and 40 m.p.h 236 
 
 169. " " " Slow Speeds 237 
 
 170. Wheel Loads for Different Spacing of Drivers 240 
 
 171. Rail Diagram for Wheel Spacing of 60 Inches 242 
 
 172. " " " " " " 70, 80 and 90 Inches 245 
 
 173. Distribution of Llorizontal Stress in Rail 248 
 
 174. Shearing Stress of Point Distant y' from Neutral Axis 249 
 
 175. Shearing Stress in 100-pound A. S. C. E. Rail 250 
 
 176. Lines of Principal Stress in Beam 251 
 
 177. Diagram of Pieces tested for Sag of Rail Head and Bending of Web 254 
 
 178. Method of Stationary Tests for Sag of Rail Head and Bending of Web 254 
 
 179. Sag of Rail Head in Stationary Tests 255 
 
 180. " " " " " Rolling Tests 257 
 
 181. Rails after Rolling Test with Load of 90,000 Pounds 258 
 
 182. Shearing Stress in 100-pound A. S. C. E. Rail and Splice Ear 262 
 
 183. 100 per cent Joint 263 
 
 184. Joint showing Uneconomical Distribution of Metal 263 
 
 185. " " Economical Distribution of Metcl 263 
 
 186. Diagram of Watertown Arsenal Tests on 100-pound Joints 265 
 
 187. Pure Swedish Iron 270 
 
 188. Pure Copper 270 
 
 189. Copper-bismuth Alloy 271 
 
 190. Iron with 1.8 per cent Carbon 271 
 
 191. Cleavage Planes with Crystals arranged Symmetrically 273 
 
 192. " " " " " in an Irregular Manner 273 
 
 193. Iron Strained beyond the Elastic Limit 273 
 
 194. Lead Strained beyond the Elastic Limit 273 
 
 195. Cross Section of Unstrained Metal 274 
 
 196. Cross Section of Metal after being Stressed 274 
 
 197. Slip Bands 275 
 
 198. Polished Surface with Small Cracks 275 
 
 199. " " " Large Cracks 276 
 
 200. Behavior of 0.55 Carbon Steel under Repeated Alternate Stresses 279 
 
 201. Behavior of 0.82 Carbon Steel under Repeated Alternate Stresses 279 
 
 202. Comparison of the Behavior of Different Grades of Steel under Repeated Alternate Stresses 281 
 
 203. Number of Repetitions before Rupture in Endurance Tests of Materials 283 
 
 204. Standard Drop Testing Machine 290 
 
 205. Diagram of Tests with Drop Testing Machines of Old and New Design 291 
 
 206. Relation of Work Done in Bending Rail in Drop and Static Tests 294 
 
 207. Time-deflection Curve, Massless Beam, within the Elastic Limit 296 
 
 208. " " Beam Stressed beyond the Elastic Limit 298 
 
 209. Scleroscope 298 
 
 210. Scleroscope Tests on Open-hearth Rail 299 
 
LIST OF ILLUSTRATIONS xiii 
 
 Fig. Page 
 
 211. Scleroscope Tests on Bessemer Rail 299 
 
 212. " " " New Titanium Rail 300 
 
 213. Amsler-Laffon Instrument for Measuring Hardness 301 
 
 214. Machine for Testing Rail Wear at Pennsylvania Steel Company 304 
 
 215. Diagram of Round Test Pieces; Tensile Tests on Rail Steel 304 
 
 216. Diagram of Flat Test Pieces; Tensile Tests on Rail Steel 308 
 
 217. Location and Numbers of Test Pieces used in Waterhouse's Tests 309 
 
 218. Effect of Repeated Loads on Beams 311 
 
 219. Resistance of Sub-grade to Pressure of the Track 314 
 
 220. Prices of Iron and Bessemer Steel Rails, 1855-1910 325 
 
 221. Comparative Wear of Rails of Similar Chemical Composition 327 
 
 222. Tenacity of Iron-Carbon Alloys 329 
 
 223. Influence of the Proportion of Nickel and Varying Heat-Treatment upon the Tensile Strength 
 
 of Nickel Steel 336 
 
 224. Influence of the Proportion of Nickel and Varying Heat-Treatment upon the Ductility of 
 
 Nickel Steel 337 
 
 225. Influence of the Proportion of Manganese on the Tensile Strength of Manganese Steel .... 337 
 
 226. Influence of the Proportion of Manganese on the Ductility of Manganese Steel 338 
 
 227. Tensile Strength and Ductility of Carbon Steel and of Manganese Steel 339 
 
 228. Elasticity and Ductility of Carbon Steel and of Manganese Steel 340 
 
 229. Ore Roasters, Norway Furnace, 1883 345 
 
 230. Open Pit Mine on Mesaba Range, Mountain Iron Mine, near Hibbing, Minnesota 346 
 
 231. View of the West Cut, looking North, Biwabik Mine 347 
 
 232. Steel Ore Dock at Two Harbors, Minn 348 
 
 233. Steamer " Augustus B. Wolvin " 349 
 
 234. Great Northern Railway Ore Dock at Allouez Bay, Superior, Wis 350 
 
 235. The "Wolvin"; a Typical Lake Steamer for the Transportation of Ore 351 
 
 236. Ten-ton Bucket of Unloader in Hold of the "Wolvin" 352 
 
 237. General View of Ore Unloader with Steamer at the Dock 352 
 
 238. Brown Hoist Unloader Unloading Cargo of Ore 353 
 
 239. Blast Furnace with Stoves and Buildings 354 
 
 240. Ground Plan, Showing the General Arrangement of Blast Furnace No. 4, Built at the Hazelton 
 
 Plant of the Republic Iron and Steel Co 355 
 
 241. Sectional View of Hazelton Blast Furnace No. 4 356 
 
 242. Top Rigging of Blast Furnace 357 
 
 243. The 450-ton Furnaces, Hot Stoves, and Gas-cleaning Plant in Course of Erection at Gary. . 358 
 
 244. The Whitwell Hot-blast Stove 359 
 
 245. Julian Kennedy Stove 360 
 
 246. Isabella Furnace, Carnegie Steel Company 361 
 
 247. Operation of Isabella Furnace on Dry Blast 362 
 
 248. 300-ton Mixer 364 
 
 249. Ten-foot Iron Cupola, Maryland Steel Company 365 
 
 250. Early Experiments of Blowing Air through Bath 367 
 
 251. Bessemer SteeL Works, Johnstown, Pa 368 
 
 252. American 5-ton Bessemer Plant. Plan 369 
 
 253. " " " " Section 369 
 
 254. Arrangement of Converters at Maryland Steel Company 369 
 
 255. 18-ton Converter, Maryland Steel Company 370 
 
 256. Typical 16-ton Bessemer Converter 371 
 
 257. Charging Bessemer Converter 372 
 
 258. Bessemer Converter in Full Blast 373 
 
 259. Modern Open-hearth Furnace 376 
 
 260. Open-hearth Plant 377 
 
 261. Wellman Tilting Open-hearth Furnace 378 
 
 262. Pouring Steel into Ladle at Open-hearth Furnace 379 
 
 263. Charging Platform of the Open-hearth Furnaces at Gary 381 
 
xiv LIST OF ILLUSTRATIONS 
 
 FlG - Page 
 
 264. Heroult Electric Furnace 384 
 
 265. Stassano Electric Furnace 385 
 
 266. Roechling-Rodenhauser Furnace 386 
 
 267. Details of Casting Ladle 389 
 
 268. Crushed Head 39-2 
 
 269. Crushed Head 394 
 
 270. Teeming Ingots at Bessemer Converter 395 
 
 271. " " " Open-hearth Furnace 396 
 
 272. Stripping the Mold from Ingots 397 
 
 273. Soaking Pits — Gary 397 
 
 274. Soaking Pits 398 
 
 275. Formation of Pipe in Ingot 399 
 
 276. Section of Ingot, containing Cavity of 128 cubic inches 400 
 
 277. Bloom from an Ingot of Same Heat and of Same Size as Fig. 276, showing Reduction of Cavity 400 
 
 278. Structure A. — Brinell's Tests 403 
 
 279. " B. — Brinell's Tests 403 
 
 280. " C. — Brinell's Tests 403 
 
 281. " D. — Brinell's Tests 403 
 
 282. " E. — Brinell's Tests 403 
 
 283. " O. — Brinell's Tests 403 
 
 284. " H. — Brinell's Tests 403 
 
 285. Ordinary Steel Ingot and Titanium Steel Ingot 406 
 
 286. Sulphur in Ordinary Steel 407 
 
 287. " " Titanium Steel 407 
 
 288. Phosphorus in Ordinary Steel 408 
 
 289. " " Titanium Steel 408 
 
 290. Carbon in Ordinary Steel 409 
 
 291. " " Titanium Steel 409 
 
 292. Influence of Conditions of Casting as shown by Wax Ingots. (Figs. 1-6.) 412 
 
 293. " " " " " " " " " " (Figs. 7-13.) 413 
 
 294. Illingworth's Press for Compressing Steel Ingots 414 
 
 295. Williams' Abdominal Liquid Compression of Solidifying Steel Ingots 414 
 
 296. Whitworth's Hydraulic Press for the Compression of Steel Ingots 415 
 
 297. Harmet's Liquid Compression by Wire Drawing 416 
 
 298. Steel Entering the Rolls 421 
 
 299. Cross Section of 8 by 8-inch Rail Bloom 422 
 
 300. Rail from Early Pass in Roughing Rolls .♦ 422 
 
 301. Same Rail as shown in Fig. 300 after Further Reduction 423 
 
 302. Finished Raii from Same Ingot as Bloom and Pieces from Roughing Rolls 423 
 
 303. Cooling Curve of Solid Copper 425 
 
 304. Cooling Curve of Water 425 
 
 305. Recalescence 425 
 
 306. Cooling Curve of Iron 426 
 
 307. Cooling and Heating Curves of Steel 426 
 
 308. Cooling of "Solid Steel" 427 
 
 309. The Influence of the Finishing Temperature on the Size of Grain 428 
 
 310. Influence of Finishing Temperature on the Size of Grain of Steel of 0.50 per cent Carbon . . . 429 
 
 311. Diagram of Results of Experiments on Rolling at Different Temperatures 430 
 
 312. Rail "B" near Surface 431 
 
 313. " "A" near Surface 431 
 
 314. " "B" Center of Head 431 
 
 315. " "A" Center of Head 431 
 
 316. Top View at Top of Head, 70-lb. Rail 432 
 
 317. " " " Center of Head, 70-lb. Rail ". 432 
 
 318. Side View at Top of Head, 70-lb. Rail 432 
 
 319. " " " Center of Head, 70-lb. Rail 432 
 
LIST OF ILLUSTRATIONS xv 
 
 Fig, Page 
 
 320. Transverse View at Top of Head, 70-lb. Rail 433 
 
 321. " " " Center of Head, 70-lb. Rail 433 
 
 322. Pieces for Microscopic Views shown in Figs. 316-321 433 
 
 323. Rail Mill, Algoma Steel Company 439 
 
 324. Housing for 28-inch Three-high Mill 440 
 
 325. Rolls used in Three-high Rail Mills 440 
 
 326. Three-high Rolls in the Rail Mill at Gary 441 
 
 327. Pass Diagram, Rail Mill, Illinois Steel Company, South Works 442 
 
 328. Rail Mill, Illinois Steel Company, South Works 443 
 
 329. Saw Runs of American Rail Mills 445 
 
 330. Head Sweep * 445 
 
 331. Cold Straightening Press, Maryland Steel Company 446 
 
 332. Value of V/E for Tables XCVI, XCVII and XCVIII 447 
 
 333. Diagram of Cogging Rolls, Tables XCVI, XCVII and XCVIII 448 
 
 334. Bar Sections of Passes 14-23 (Tables XCVI, XCVII and XCVIII) 449 
 
 335. Sections in which only "Direct Pressure" occurs in the Process of Rolling 456 
 
 336. Illustration of "Indirect Pressure" 456 
 
 337. Effect of Inclination of Inner Surface of the Rail Flange on Energy required in Rolling. . . . 457 
 
 338. Work done in Accelerating the Rotating Masses in Reversing Mill 459 
 
 339. Recent Rail Sections 460 
 
 340. Shrinkage Allowed in American Specifications in 1909 475 
 
 341. Test Pieces C and D, British Standard Specifications of Rails 487 
 
 342. M. W. 401. — Report of Chemical and Physical Examination 503 
 
 343. M. W. 402. — Certificate of Inspection 504 
 
 344. M. W. 403. — Report of Shipment 505 
 
 345. M. W. 404. — Report of Rail Failures in Main Tracks 506 
 
 346. M. W. 405. — Superintendent's Monthly Report of Rail Failures in Main Tracks 508 
 
 347. M. W. 406. — Annual Statement of Steel Rails Existing in Main Tracks 510 
 
 348. M. W. 407. — Laboratory Examination of Special Rails 511 
 
 349. Standard Locations of Borings for Chemical Analyses and Standard Tensile Test Pieces. . . . 512 
 
 350. M. W. 408. — Summary of Steel-rail Failures for One Year Compared with the Same Period 
 
 of Previous Year 513 
 
 351. M. W. 409. — Summary of Steel-rail Failures for a Period of Years 514 
 
 352. M. W. 410. — Comparative Number of Failures of Steel Rails of Different Section or Pattern, 
 
 Rolled by Different Steel Companies 515 
 
 353. M. W. 411. — Position in Ingot of Steel Rails which Failed 516 
 
 354. M. W. 412. — Cover Page for Forms M. W. 408, 409, 410 and 411 517 
 
 355. M. W. 413. — Location Diagram One inch equals one mile 518 
 
 356. M. W. 414. — " " Two inches equal one mile 519 
 
 357. M. W. 415. — Diagram showing Lines of Wear 520 
 
 358. M. W. 416. — Record'of Comparative Wear of Special Rail 521 
 
 359. M. W. 417. — Cover Page for Forms M. W. 413, 414, 415 and 416 522 
 
 360. Defective Rail Sheet 523 
 
 361. Diagram of Rail Failures, Harriman Lines 524 
 
LIST OF PLATES 
 
 (All plates except V and VI are in the back of the book.) 
 (Plates V and VI are between pages 11 and 12.) 
 
 - I. Standard Rail Sections of the American Society of Civil Engineers. 
 
 II. Rail Sections used before the Adoption of the A. S. C. E. Standard Sections in 1893. 
 
 III. Standard Wheel Sections showing Coning of Wheel. 
 
 IV. Rail Sections used during the Period between the Adoption of the A. S. C. E. Standard 
 
 Sections in 1893 and the Recommendation of the New Standard Sections by the 
 American Railway Association in 1907. 
 V. Examples of Defective Rails; Broken Rails, Flow of Metal and Crushed Head. 
 VI. " " " " Split Head, Split Web and Broken Base. 
 
 ~VII. Proposed Standard Rail Section of the American Railway Association, Series "A." 
 VIII. " " " " " " " " " Series "B." 
 
 - IX. Standard "P.S." Rail Section of the Pennsylvania Railroad System. 
 " X. Rail Sections of the Vignole Type. 
 " XL Rail Sections used on German Railways. 
 ~XII. Midland Railway, Permanent Way. 
 "XIII. L. & N. W. R. Details of Permanent Way. 
 ~XIV. British Standard Bull Head Railway Rails. 
 
 "XV. " " Flat Bottom Railway Rails. 
 
 -XVI. Rail Sections for Street Railways, Tram Girder Rails and High Tee Rails. 
 XVII. " " " " " Standard Girder Sections of the American Electric 
 
 Railway Engineering Association. 
 1 XVIII. British Standard Tramway Rails. 
 
 XIX. Deflection of Driving Wheel Spring, Consolidation Engine No. 1064, Boston & Maine 
 Railroad. 
 XX. Passenger Locomotive Diagrams. 
 " XXI. Freight Locomotive Diagrams. 
 XXII. Examples of American Tie Plates. 
 
 XXIII. Rail Diagram of Love. 
 
 XXIV. Examples of American Rail Joints. 
 XXV. Joints Tested at the Watertown Arsenal. 
 
 XXVI. Dynamic Wheel Loads for Various Rails and Axle Spacing. 
 
 XXVII. Bending Moments in Different Weights of Rail Corresponding to Loading on Plate 
 XXVI. 
 XXVIII. Weight of Rail for Various Conditions of Loading and Classes of Track. 
 XXIX. Plan of Gary Steel Plant. 
 XXX. Reversing Cogging Mill. 
 
 XXXI. American Three-high Mill, with 36-inch Rolls. 
 XXXII. Power required to Roll Rails about 35.5 kg. per meter. 
 XXXIII. Form M. W. 418, Am. Ry. Eng. Assn. 
 
STEEL RAILS 
 
 CHAPTER I 
 DEVELOPMENT OF THE PRESENT SECTION 
 
 1. Early Sections 
 
 Apparently the use of steel rails was first resorted to on account of the 
 poor quality of the iron rails of later manufacture. The wear of these iron rails 
 took the form of crushing or lamination, which destroyed the running surface 
 of the rail and rendered it unfit for use. An iron rail when manufactured, 
 even in the best way, was little more than a bundle of rods; and the top slab 
 under the heavy pounding of the locomotive had a tendency to spread side- 
 ways and become laminated. A steel rail, on the contrary, was rolled from a 
 solid ingot and for that reason was much more durable.* Iron, in the matter 
 of wear, exhibited very great irregularity, some rails showing signs of distress 
 within a year or two of being laid down, while others afforded very satis- 
 factory results. 
 
 f As an illustration of the latter assertion we can instance the experience 
 on the main line of the North-Eastern Railway on certain sections of its system 
 which may be taken as fair samples of the others. On that extending between 
 Newcastle and Berwick, 66.8 miles of double way, the iron rails laid down in 
 1847 weighed 65 pounds per yard. Renewals commenced in 1855 and terminated 
 in 1867. In these the weight was increased to 82 pounds per yard. The maxi- 
 mum duration of the 65-pound rails was 21 years and the minimum 8 years, 
 the average being 12.8 years. 
 
 Mr. T. E. Harrison stated in 1867 that on 700 miles of permanent way of 
 the North-Eastern Railway the average duration of the last complete set of 
 rails was found to be about 15.5 years; and some which were laid down in 1834 
 were still in use. 
 
 * See The Manufacture and Wear of Rails by C. P. Sandberg, Minutes of Proceedings of the 
 Institution of Civil Engineers, Vol. XXVII, Session 1867-8, and R. Price Williams's Paper "on the 
 Maintenance of Permanent Way," ibid., Vol. XXV, p. 353. 
 
 t Principles of the Manufacture of Iron and Steel. I. Lowthian Bell. London, 1884. 
 
2 STEEL RAILS 
 
 The statements just submitted do not afford any proper criterion of the 
 resisting powers of iron rails; for this can only be determined by the comparative 
 weights of the engines, the amount of traffic, and the speed of the trains which 
 have passed over them. According to Mr. R. Price Williams the average life of 
 an iron rail, on the most heavily worked portions of the railways in the United 
 Kingdom in the year 1878, may roundly be taken at about 17| millions of tons. 
 
 There is nothing speculative in the assertion that iron rails made before 
 the complete discontinuance of refining were, generally speaking, longer lived 
 than those of later manufacture. No doubt in the later days of iron rails the 
 permanent way was much more severely taxed than was formerly the case. 
 The engines were more ponderous, the traffic was heavier, and the speed greater; 
 but the experience of the North-Eastern Railway at all events indicates that 
 rails of iron have occasionally been made to give very satisfactory results. 
 Whether this be due to their having been made from refined metal, or whether 
 indeed they were so made, we have unfortunately little means of proving. It 
 is significant to note that during the twenty years preceding 1868 the price of 
 iron rails had been gradually reduced to one-third of their original cost, and that 
 this reduction was accompanied by the production of an inferior quality of rail. 
 
 * In America several of the railway companies began to use steel rails 
 as far back as 1864. In that year the Chicago and Northwestern, the Phila- 
 delphia, Wilmington and Baltimore, and the Old Colony and Newport each laid 
 portions of track with this metal. In the following year the Boston and Albany, 
 the Boston and Providence, the Connecticut River Railroad, the Chicago, Rock 
 Island and Pacific, and the Chicago and Alton each began the use of steel. 
 
 In September, 1869, a commission appointed to ascertain the extent to 
 which steel rails had been tried in the United States ascertained that of fifty- 
 seven railways then in operation, from which reports had been obtained, twenty- 
 six had made use of steel in weights varying from 100 to 15,000 tons, the whole 
 bulk reported as in use being 49,800 tons, equal to about 518 miles of tracks. 
 This, however, did not by any means represent the total weight of steel rails 
 laid throughout the States. The Commission already referred to was, indeed, 
 particular in calling attention to the fact that on the first of January, 1870, 
 there were at least 100,000 tons of steel rails laid down in America, and 10,000 
 tons of steel-headed rails besides. Of this quantity the largest bulk had been 
 supplied by England, and almost entirely by the Atlas, Barrow and Dowlais 
 Works, although several thousand tons had been contributed by three estab- 
 lishments in Germany. 
 
 * Steel — Its History, Manufacture, Properties, and Uses. J. S. Jeans. London, 1880. 
 
DEVELOPMENT OF THE PRESENT SECTION 3 
 
 In France the Paris, Lyons, and Mediterranean Railway Company decided, 
 so early as 1867, to use only steel rails in relaying its permanent way on 860 
 kilometers of the Paris and Marseilles line, where more than 10,000 trains ran 
 over each line of way yearly, at speeds which might reach 90 kilometers per hour. 
 
 In Austria steel rails were used as early as 1859 on the Northern Railway, 
 connecting Vienna with Cracow. They were, however, of puddled steel, manu- 
 factured at the works of the Archduke Albrecht, at Carlshutte. 
 
 In 1861 one German mile (about 4.8 miles) of the main line was laid with 
 steel rails by way of experiment. So much satisfaction was afforded by this 
 trial that in 1865 it was resolved to reconstruct in steel the whole of the main- 
 line permanent way, and by the close of that year thirty-five English miles had 
 been laid. Previous to this, however, steel had been largely used in Austria 
 for railway crossings — so much so, indeed, that at the close of 1864 there were, 
 on the Northern line, 468 steel crossings as compared with 977 of iron. 
 
 In Russia several lots of steel rails were laid down previous to 1872, but 
 the use of that metal cannot be said to have received a thorough impulse until, 
 in the year named, the Russian Imperial Administration approved the con- 
 struction with steel rails of the railways from Wjasma to Tula, Rjask, and 
 Jetetz, and from Morschunsk to Siezran, a total length of 1200 kilometers. 
 Previous to this time steel rails had been laid experimentally on the Nicolas 
 Railway, where they were found to answer so well that in 1872 about 70,000 
 tons of steel rails were ordered for Russia, chiefly from Creusot. 
 
 In England, before the end of 1861, steel rails had been laid down on the 
 Caledonian, Lancashire and Yorkshire, London, Brighton and South Coast, 
 and Rhymney railways, as well as on the London and North-Western. 
 
 The first Bessemer steel rails made in America were rolled at the North 
 Chicago Rolling Mill on the 24th of May, 1865, from hammered blooms made 
 at the Wyandotte Rolling Mill from ingots of steel made at experimental Steel 
 Works at Wyandotte, Mich. The experimental Steel Works at Wyandotte 
 were erected in 1864, and were the first works started in the country for conduct- 
 ing the pneumatic or Bessemer process. The rolls upon which the blooms were 
 rolled at the North Chicago Rolling Mill were those which had been in use for 
 rolling iron rails, and, though the reduction was much too rapid for steel, the 
 rails came out sound and well shaped. The first steel rails rolled in the United 
 States upon order, in the way of regular business, were rolled by the Cambria 
 Iron Company at Johnstown, Pa., in August, 1867,* from ingots made at the 
 
 * See paper on the Development of the American Rail and Track by J. Elfreth Watkins, Trans. 
 Am. Soc. of Civil Engrs., April, 1890, Vol. XXII, p. 228. 
 
4 STEEL RAILS 
 
 works of the Pennsylvania Steel Company, at Harrisburg, Pa., rails were rolled by 
 the Spuyten Duyvil Rolling Mill Company, at Spuyten Duyvil, N. Y., early in 
 September of that year, from ingots made at the Bessemer Steel Works, at Troy, 
 N. Y., then owned by Winslow and Griswold, but these were on experimental 
 orders, and not regular ones from any railway company.* 
 
 Before, however, the American steel works had produced any Bessemer 
 rails, or, indeed, before any such works had been executed in this country, the 
 Pennsylvania Railway Company had imported from England a lot of about 
 150 tons. This was towards the close of the year 1863. Some little delay took 
 place in slotting the rails to receive the track fastenings ; they were not laid 
 down until the early part of 1866, when they were placed on sidings in the yards 
 at Altoona and Pittsburg, where they would be subjected to considerable use. 
 As the rails appeared very brittle, it was not deemed expedient to place them 
 in the main track where they would be passed over by trains at high rates of 
 speed. None of them, however, were broken in the track, and as they exhibited 
 little or no appearance of wear, other steel rails were ordered in 1867, of a quality 
 combining more toughness with a sufficient degree of hardness, and experiments 
 were continued to test the relativ3 merits of the several descriptions of rail. 
 
 About 1864 the Erie Railway Company ordered from John Brown and 
 Company, of Sheffield, England, 1000 tons of Bessemer steel rails at 25£ per ton. 
 
 The American Commission of 1869 concluded, as the result of comparing 
 reports obtained from twenty-six railways then using steel rails: (1) That ex- 
 tremes of temperature do not injuriously affect steel rails. The Grand Trunk 
 Railway reports them as not injured by a temperature of 30 degrees below zero 
 of Fahrenheit, and no other road appears to find them unable to stand a cold 
 winter. (2) That the durability of steel rails far exceeds that of the best iron 
 rails. The Erie Railway reports their steel rails as having outworn thirteen 
 sets of iron rails, and as showing scarcely any sign of wear. The Philadelphia, 
 Wilmington and Baltimore reports them as having outworn seventeen iron rails, 
 and as showing little wear. The Chicago and Northwestern say that steel rails 
 have outworn fifteen iron rails, and show no perceptible wear. 
 
 In 1874 a small committee of American experts f conducted a very careful 
 and elaborate inquiry into the form, endurance, and manufacture of rails. In 
 
 * Private communication from Mr. Robt. W. Hunt. 
 
 t Appointed January 8, 1873, by the American Society of Civil Engineers, to determine " the 
 best form of standard rail sections of the United States; the proportion which the weight of rails should 
 bear to the maximum loads carried on a single pair of wheels of locomotives or cars; the best methods 
 of manufacturing and testing rails; the endurance, or, as it is called, the ' life ' of rails; the causes of 
 the breaking of rails and the most effective way of preventing it, and the experience of railways in America 
 
DEVELOPMENT OF THE PRESENT SECTION 5 
 
 speaking of the comparative value of steel and iron rails this committee stated, 
 that "while steel rails as we get them are tolerably uniform in quality, iron 
 varies so much that no comparison can be made except of particular qualities 
 or of averages of qualities widely different. We can as yet do little more than 
 give the results of our own experience. In so doing we shall not only compare 
 steel and iron, but also the effects of some different circumstances on the duration 
 of both. It seems probable that the best iron, if homogeneous and the head 
 of uniform hardness, so as to wear off evenly like steel, would, with machinery 
 of moderate weight, wear a third or even half as long as steel. The chairman 
 has found that his 62-pound iron rail, after carrying about 14,000,000 tons 
 gross load, has worn off only about 25 per cent more than the steel rails on the 
 same track and under the same circumstances. Probably it will not wear so well 
 when the top crust is worn through. But owing to want of homogeneousness 
 and uniformity the iron scales, splinters, laminates, or somehow disintegrates 
 or mashes in spots before it wears out." 
 
 Ashbel Welsh, the chairman of this committee, subsequently presented a 
 final report, giving particulars of the behavior of the steel rails, 53 pounds per 
 yard. Believing that a very thin stem and a very thin base would possess 
 sufficient strength, he designed a pattern in which as much metal as possible 
 should be placed in the head, and as little as possible anywhere else. The height 
 was 4 inches, the width of base 4 inches, the head fully 2f inches wide and 1} 
 inches deep; radius of sectional curvature of the head 12 inches, stem T 7 ? inch 
 thick, base f\ inch thick at the edges; angle of base and of under sides of head 
 14 degrees, length of rails 30 feet, weight 53 pounds per yard. 
 
 The rails were rolled by John Brown and Company, and were laid in 1867 
 and 1868 at places exposed to very heavy traffic, on the railway between Phila- 
 delphia and New York, where iron rails had lasted only four months. In straight 
 portions of the line, after having carried a gross weight of about 50,000,000 
 tons, mostly at high speeds, the heads had been worn down \ inch, having lost 
 in weight about 6 pounds per yard. In some sharp curves the sides of the heads 
 were so much worn that the rails were taken up in June, 1876. 
 
 The early steel rails were naturally made to the existing iron pattern. 
 These were generally pear-headed in order to prevent the side of the head from 
 breaking down, and were therefore not adapted to fishing. In 1866, as we 
 have seen, Mr. Ashbel Welsh designed a section differing but slightly from the 
 
 in the use of steel rails." See paper on the Form, Weight, Manufacture and Life of Rails. A 
 Report by Ashbel Welsh, C. E.; M. N. Forney, M. E.; O. Chanute, C. E.; and I. M. St. John, C. E. 
 Trans. Am. Soc. of Civil Engrs., Vol. Ill, p. 87. 
 
6 STEEL RAILS 
 
 modern rail,* and in 1874 Mr. Chanute, chief engineer of the Erie Railway, 
 investigated to determine the proper contour of the head by observing 
 the contour of the rails worn down by the action of the wheels. The width 
 and shape of the head having been provided for, the rail was considered 
 as a beam, and as much metal as possible was taken from the web and flange 
 to deepen it. 
 
 With the older sections the connections at the joints were very unsatis- 
 factory, the design preventing the fishplate from supporting the head. If the 
 plate could bear against horizontal surfaces, it would not be forced out laterally 
 by the loads, but the rail could not be properly filled by rolling and the play 
 would rapidly increase and could not be taken up. Mr. Chanute experimented 
 to determine the correct angle of the under side of the head to hold the fishplate 
 and found that with an angle above 15 degrees the plate was loosened by 
 stretching of the bolts. This relieved the pressure and friction of the plate 
 against the nuts and allowed them to turn. He therefore adopted the angle of 
 15 degrees under the head, and to avoid unnecessary metal in the flange he 
 made its angle 12 degrees. 
 
 The adoption of an improved section was very slow, and as late as 1881 
 119 patterns of steel rails of 27 different weights per yard were regularly manu- 
 factured, and 180 older patterns were still in, use, making a total of nearly 300 
 different patterns. This great variety of sections in use required the mills 
 to keep a large number of different rolls in stock, and finally to standardize 
 the design of the rail the present A. S. C. E. section, shown in Plate I, was 
 presented to the society on August 2, 1893. These sections met with favor, 
 and were adopted by many railroads, so that in a few years about two-thirds 
 of the output of the rail mills conformed to this design. 
 
 The gradual evolution of the present design of rail is shown in Plate II. 
 The earlier rails show the pear shape of the old iron rails, followed by the rails 
 where the section was more adapted to fishing and having a better distribution 
 of the metal to afford a stiffer rail. 
 
 The question of cylindrical tires and flat top rails was one on which there 
 existed for a long time a great deal of difference of opinion among railroad engi- 
 neers. In the early days of railroading the wheels were generally coned to a ratio 
 of 1 in 20, and after the organization of the Master Car Builders' Association this 
 ratio was adopted as the standard. This particular ratio apparently grew out of 
 
 * Robert L. Stevens in 1830 designed a " T " section of iron rail for the Camden and Amboy 
 Railroad, and is generally considered to have been the inventor of the flat-footed rail. See Trans. 
 Am. Soc. of Civil Engrs., Vol. IV, p. 236, and ibid., Vol. XXII, pp. 209, 216. 
 
DEVELOPMENT OF THE PRESENT SECTION 7 
 
 the prevailing practice at the car wheel foundries, and not from any theoretical 
 consideration of the relation of the wheel to a curve. 
 
 The agitation for the cylindrical wheel grew out of efforts to measure the area 
 of contact between the wheel and the rail, to determine the intensity of pressure 
 on the metal, and led the Master Car Builders' Association, in 1886, to change 
 their standard wheel section and reduce the coning ratio from 1 in 20 to 1 in 38, 
 which was about the last draft that would allow free withdrawal from the mold. 
 
 This section is shown on Plate III which also shows the cylindrical wheels 
 considered by the Association at this period. The section thus recommended and 
 adopted by the Association passed into general use on the railways of the country. 
 It was noticed, however, that the change was followed by a large increase in the 
 number of broken and sharp flanges, and after using the section for over twenty 
 years it was restored to the former ratio of 1 in 20 as shown by the 1910 wheel 
 given on the plate. 
 
 The rails of heavier section manufactured within the last few years are not 
 giving the service that should be expected of them. The fault may lie in im- 
 proper methods of manufacture or in the design of the rail itself, which, while 
 suitable for the conditions existing nineteen years ago, may be unfitted for 
 the heavy wheel loads of to-day. 
 
 It has been claimed that the old committee of the American Society of 
 Civil Engineers did not properly appreciate the importance of low finishing 
 temperature in designing their rails, and that the sections recommended in its 
 report in 1893 do not permit of a low enough finishing temperature in rolling 
 owing to the wide, thin flanges. 
 
 As a matter of fact this was one of the points which received most careful 
 consideration, not only by discussion between the members of the committee, 
 but also in consultation with rail manufacturers. But a peculiarity of the 
 situation comes from the fact, that at that time, what we now consider sections 
 of necessary weight were then not in general use. The committee was in- 
 structed to devise sections from 100 pounds per yard down, decreasing by 5 
 pounds, but 80-pound sections were then regarded as the heaviest likely to 
 be extensively used. Only one railroad at that time had heavier sections, and 
 that was the Philadelphia and Reading, which had a few 90-pound rails in use. 
 The New York Central had put in 80-pound rails, and perhaps they had a few 
 heavier ones, but their standard was 80-pound. The Delaware and Hudson had 
 adopted the 80-pound rail, also the Michigan Central. 
 
 The question was to devise a section which the committee considered a 
 good one and which could be easily rolled. Unfortunately the sections be- 
 
5 9 
 
 1 ° °" 
 
 {"0Z 
 
 \S3umivj 
 
 
 
 
 
 
 
 2 ^^ 3sva 
 
 & 
 
 
 
 
 
 
 
 
 §1 
 
 S3A1 
 
 
 
 
 
 
 
 
 OVSH 
 
 
 
 
 
 
 
 
 ; 111 
 
 usmaa 
 
 ° 
 
 —~6Z 
 
 S31M1IVJ 
 
 
 
 
 
 
 
 %^m 3svs 
 
 1 
 
 
 
 
 
 
 
 § SJ/M 
 
 
 
 
 
 
 
 3| QV3H 
 
 
 
 
 
 
 
 2 II 
 
 i 1 i nil illllll Nsxoua 
 
 Jgl 
 
 -~zr~~ 
 
 S3umivj 
 
 ■ON TV 101 
 
 
 
 
 
 
 
 ~i ssva 
 
 s 
 
 ^ 
 
 
 
 
 
 
 
 !il a3M 
 
 
 
 
 
 
 
 2 J 0V3H 
 
 
 
 
 
 
 
 m N sxoaa 
 
 S3umivj 
 
 
 
 
 
 
 
 "i\ 3SVS 
 
 % 
 
 
 
 
 
 
 
 flf 33M 
 
 
 
 
 
 
 
 fS OVSH 
 
 
 
 
 
 
 
 
 rillljjlljjjjj N3X0H8 
 
 ^3 
 1"§ 
 
 — 9 
 
 S3vmivd 
 
 ■ON 1V101 
 
 
 
 
 
 
 
 = | 3SVB 
 
 s 
 
 5i 
 
 
 
 
 
 
 
 §| asM 
 
 
 
 
 
 -1 
 
 
 | OVSH 
 
 
 
 
 
 
 
 
 Nsxoaa 
 
 jgp 
 
 S3ntmv3 
 
 ■ON 1V101 
 
 
 
 
 
 
 
 % 
 
 ssva 
 
 
 
 
 
 
 
 
 ; 
 
 S3M 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 nsxoaa 
 
 "PS 
 
 S3ymivj 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 1 
 
 3SV8 
 
 s 
 
 Q 
 
 
 
 
 
 
 
 = 
 
 asM 
 
 
 
 
 
 
 
 1 
 
 QV3H 
 
 
 
 
 
 
 5 III 
 
 
 N3X0UB 
 
 i g 
 
 "'1 
 
 
 
 
 
 
 
 3 
 
 ssva 
 
 =s 
 
 
 
 
 
 
 
 % 
 
 asM 
 
 
 
 
 
 
 
 i 
 
 OVSH 
 
 
 
 
 
 
 
 
 ;:||||l 
 
 mxoua 
 
 
 S3ymivj 
 
 ■ON li '101 
 
 
 
 
 
 
 
 31 ssn 
 
 1 
 
 
 
 
 
 
 
 
 ol S3M 
 
 
 
 
 
 
 
 2H OVSH 
 
 
 
 
 
 
 
 Sill 
 
 Nsxoaa 
 
 "~818~~~ 
 S3HmiVJ 
 
 
 
 
 
 
 
 3 
 
 3sva 
 
 § 
 
 
 
 
 
 
 
 §111 
 
 a3M 
 
 
 
 
 
 3 jp == 
 
 
 | OVSH 
 
 
 
 
 
 
 
 
 Nsxoua 
 
 
 "zr~' 
 
 S3iimivj 
 
 
 
 
 
 
 
 :■ 
 
 ssva 
 
 1 
 
 § 
 
 
 
 
 
 
 
 | 
 
 as/A 
 
 
 
 
 
 
 
 
 QV3H 
 
 
 
 
 
 
 
 311 
 
 
 Nsnoaa 
 
 
 S3umivj 
 
 ON 1V101 
 
 
 
 
 
 
 
 i 
 
 ssva 
 
 % 
 J 
 
 
 
 
 
 
 
 § 
 
 a3M 
 
 
 
 
 
 
 
 ill ll ||| | JJIUJI NSMOUa 
 
 
 
 
 
 
 
 
 "6ZL"' 
 
 S3umivj 
 
 ■ON 1V101 
 
 
 
 
 
 
 
 "M ssva 
 
 £ 
 
 Q 
 
 
 
 
 
 
 
 gj asM 
 
 
 
 
 
 
 5 
 
 0V3H 
 
 
 
 
 
 
 
 ~|| Nsvoua 
 
 ~'zsr" 
 
 S3 y mi v 3 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 sH ssva 
 
 * 
 
 
 
 
 
 
 
 Si 93M 
 
 
 
 
 
 
 
 -\ OVSH 
 
 
 
 
 
 
 
 ' iff! Nixoua 
 
 
 '~'9Z~~' 
 
 S3M11IVJ 
 ■ON IV 101 
 
 
 
 
 
 
 
 °jm ssva 
 
 £ 
 
 si 
 
 
 
 
 
 
 
 g] 83/H 
 
 
 
 
 
 
 
 5 mm J UTom 
 
 jl* 
 
 ~"ZP~~" 
 
 S3umivj, 
 
 ■ON 1V101 
 
 
 
 
 
 
 
 2jj§ 3Sra 
 
 
 
 
 
 
 
 
 g]j| g3/M 
 
 
 
 
 
 
 1' jjiliijjl ||i| , IDUIIJ Ww:ja 
 

 -— J— - 
 
 SBumivj 
 
 'OH IV 101 
 
 
 
 
 
 
 
 §1 3SV9 
 
 ! 
 
 q; 
 
 
 
 
 
 
 
 j 93M 
 
 
 
 
 
 
 
 3rt av3H 
 
 
 
 
 
 
 
 m 
 
 N3X0U9 
 
 ""p 
 
 sin mi vi 
 
 ■ON TV 101 
 
 
 
 
 
 
 
 
 3SV3 
 
 1- 
 
 
 
 
 
 
 
 ~-ii mini 
 
 93M 
 
 
 
 
 
 
 
 
 2^ 0V3H 
 
 
 
 
 
 
 
 M ™* 
 
 <2 9 to 
 
 * ^ cs 
 
 "89ZS- 
 
 S3 n mi vi 
 
 •ON IV 101 
 
 
 
 
 
 
 
 _ ^ 3SV9 
 
 1 
 
 \ 
 
 
 
 
 
 
 
 51 
 
 93M 
 
 
 
 
 
 
 - 
 
 
 0V3H 
 
 
 
 
 
 
 
 jJlil 
 
 N3X099 
 
 "IPl'S" 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 3§i ssv8 
 
 5 
 
 
 
 
 
 
 
 31 S3M 
 
 
 
 
 
 
 
 0V3H 
 
 
 
 
 
 
 
 llllllf N3X0U9 
 
 1 £*"~ 
 
 —Q£— 
 
 sjamivj 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 Si 3SS/8 
 
 £ 
 
 CO 
 
 a- 
 
 
 
 
 
 
 
 Sj 93M 
 
 
 
 
 
 
 
 3JI 0V3H 
 
 
 
 
 
 
 
 Sljjj n3»oua 
 
 "Z91— 
 
 sjvm/VJ 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 Si 3SV3 
 
 * 
 
 
 
 
 
 
 
 S III S3/M 
 
 
 
 
 
 "SI 
 
 
 -l"=-^=| OPJtf 
 
 
 
 
 
 
 
 21 
 
 N3X0U9 
 
 jl 5 
 
 S3UD1IVJ 
 ■ON IV 101 
 
 
 
 
 
 
 
 § 
 
 3SVS 
 
 S 
 
 
 
 
 
 
 
 
 
 - 
 
 93M 
 
 
 
 
 
 
 
 1 
 
 0V3H 
 
 
 
 
 
 
 
 § 
 
 N3X09S 
 
 -— g~— 
 
 S3 u mi v j 
 
 ■ON 1V101 
 
 
 
 
 
 
 
 = 
 
 3SV9 
 
 1 
 
 
 
 
 
 
 
 3g 
 
 93M 
 
 
 
 
 
 
 
 S^^ o«w 
 
 
 
 
 
 
 
 || N3H0V8 
 
 1S9 
 
 tfc "=c ^ 
 
 sir" 
 
 S3umivj 
 
 
 
 
 
 
 
 Si 3SfS 
 
 1 
 
 
 
 
 
 
 
 
 
 
 ., |lllll]llli| ff3/M 
 
 
 
 
 
 
 
 "SIlllllllllllllJ 13X0X9 
 
 
 
 
 
 
 
 '"l£Z"~ 
 
 
 
 
 
 
 
 2 1 jsps 
 
 ^ 
 
 
 
 
 
 
 
 III S3A1 
 
 
 
 
 
 = i = H avsH 
 
 
 
 
 
 
 3 till 
 
 
 N3H0U9 
 
 ^9 
 
 ' — 9~~~ 
 SBumivj 
 
 •ON IV 101 
 
 
 
 
 
 
 
 1 
 
 3SV9 
 
 £ 
 
 1 
 
 CO 
 
 
 
 
 
 
 
 -. 
 
 93M 
 
 
 
 
 
 
 : 
 
 
 0V3H 
 
 
 
 
 
 
 
 ; 
 
 N3H0U9 
 
 / 
 
 S3unnvd 
 
 ■on IV 101 
 
 
 
 
 
 
 
 r 
 
 3SV9 
 
 5 
 
 
 
 
 
 
 
 ; 
 
 93M 
 
 
 
 
 
 
 
 31 
 
 0V3H 
 
 
 
 
 
 
 
 1 
 
 N3X099 
 
 II 
 
 —£P£— 
 S3nmivj 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 SI 
 
 3SV9 
 
 I 
 
 § 
 
 
 
 
 
 
 
 
 
 51 
 
 83M 
 
 
 
 
 
 
 _»J 
 
 5|||d MUOW 
 
 "'819'" 
 
 S3umivd 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 
 * 
 % 
 
 
 
 
 
 
 
 si 1 ™ 
 
 
 
 
 
 
 
 ^^^J avaH 
 
 
 
 
 
 
 ■J 
 
 lllllll H3X0U3 
 
 ^32 
 
 —8£— 
 ssyniivj 
 
 ■OH IV 101 
 
 
 
 
 
 
 
 3II ssva 
 
 1 
 
 | 
 
 
 
 
 
 
 
 1 83/M 
 
 
 
 
 
 
 
 °Hl U g 
 
 
 
 
 
 
 
 11 lllllll 
 
 
 ~~TL 
 S3umiVd 
 
 ■ON IV 101 
 
 
 
 
 
 
 
 Si 3SfS 
 
 j 
 
 
 
 
 
 
 
 : 
 
 QV3H 
 N3X089 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 "'lllllll 
 
 tim 33 
 
 o - -c *, sc 
 
 
10 STEEL RAILS 
 
 yond 80 pounds were matters of compromise, and as they progressed arith- 
 metically less satisfactory results were obtained as the weight increased. 
 
 It has been the invariable experience in changing from a light to a heavy 
 section, in any class of rolled steel, that difficulties have been met and modifi- 
 cations have been made in the methods of rolling, in order to get as good structure 
 in the heavier sections as was formerly obtained in the lighter sections. In 
 ordinary sections other than rails it was a comparatively easy matter to over- 
 come the trouble and get a good structure; but the thin flange of the rail, and 
 the higher carbons called for in the heavier sections, further complicated 
 matters. 
 
 The greatest need at this time is for reliable statistical information taken 
 from properly kept records. The Committee on Rail of the American Railway 
 Engineering Association have been engaged for several years in collecting 
 statistics of defective rails on American roads. The classification adopted by 
 the committee is as follows: 
 
 1. Broken Rail. 
 
 2. Damaged. 
 
 3. Flow of Metal. 
 
 4. Crushed Head. 
 
 5. Split Head. 
 
 6. Split Web. 
 
 7. Broken Base. 
 
 It is the intention of the classification that all rails which broke in service, or 
 which have a straight crack working from top to bottom or from bottom 
 to top, which would very quickly result in a broken rail, should be classified as 
 "broken rails," regardless of internal defects. All of the other defective rails 
 which are removed, not being "broken" or damaged on account of wrecks, 
 broken wheels or similar causes, are to be classified under one of the other 
 heads, from 3 to 7, both inclusive. 
 
 Fig. 1 shows a comparison of rail failures between different sections. (See 
 Plates I, IV, VII and IX for description of sections.) The most striking char- 
 acteristic of the diagram is the comparatively large number of head failures of 
 85 N, 85, both on tangent, 36.1 failures per 10,000 tons, and on curve 27.7 
 failures per 10,000 tons laid. The legend on the diagram explains that this is a 
 Chicago, Burlington and Quincy section. It can hardly be said that the carbon 
 is excessively high, although pretty high, unless it is badly segregated, the chem- 
 ical constituents being: 
 
DEVELOPMENT OF THE PRESENT SECTION 11 
 
 Carbon 48 to .58 
 
 Phosphorus 10 
 
 Manganese 80 to 1.10 
 
 Silicon 20 
 
 Section 852, 85-pound, also a Chicago, Burlington and Quincy section, 
 has the same composition, but the failures are not so numerous. 
 
 Carbon 58 
 
 Phosphorus 10 
 
 Manganese 80 to 1.10 
 
 Silicon 20 
 
 The next most numerous head failures are in the A. S. C. E. 90-pound on 
 tangent, 15J failures per 10,000 tons, and on curve 12J failures per 10,000 tons 
 laid. The P. R. R. 100-pound head failures on curve are 12.8 per 10,000 tons 
 laid, while the head failures on tangent are small, and the head failures of the 
 New York Central 80-pound are about as large, both on tangent and curve, 
 11.4 per 10,000 tons laid. The A. S. C. E. 80-pound on tangent, the P. R. R. 
 85-pound on tangent, and the A. S. C. E. 85-pound on tangent and curve have 
 had the same number of head failures as the New York Central 80-pound. 
 The A. S. C. E. 100-pound on tangent and curve comes next, and then 852, 
 85-pound on curve, while the rest were all less than 5 failures per 10,000 tons laid. 
 
 The breakages are most numerous in 85 N, 85-pound on tangent, 9.6 per 
 10,000 tons laid, and next of 852, 85-pound on tangent, with the A. S. C. E. 90- 
 pound on tangent and the Dudley 80-pound on curve, both the same, following 
 closely. Next comes the New York Central 80-pound and 852 85-pound on 
 curve and the Boston and Maine 75-pound on tangent, all the same, and then 
 A. S. C. E. 100 and 85-pound on tangent. The breakages of the others are less 
 than 4 per 10,000 tons laid. 
 
 It will be observed that the breakages of so-called stiff sections are more 
 numerous than those of the lower sections with the heavier head. The carbon 
 is generally higher in the C. B. & Q. sections than in the A. S. C. E. and P. R. R. 
 sections. The web and base failures are less than 4 per 10,000 tons laid. 
 
 Plates V and VI present photographs of typical rail failures collected by 
 the committee. 
 
 Nearly nine million of tons of Bessemer steel rails, from seven different 
 mills and varying in weight from 100 pounds to 75 pounds, were reported in 
 the tracks of the American railroads on October 31, 1910. This corresponded 
 to 21,503,803 rails, and for the twelve-month period from October 31, 1909, 
 
* 
 
 "n. 
 

12 STEEL RAILS 
 
 to October 31, 1910, there were 30,086 failures or one defective rail for every 
 714 rails laid in the track. 
 
 It will be interesting to turn to the conditions of twenty years ago. The 
 following table * shows the rail failures on one of the American railways during 
 the years 1884 to 1888 inclusive. 
 
 In track, June 1, 1884 121,685 tons 
 
 In track, January 1, 1889 162,526 tons 
 
 Removed from track, 1884 to 1888 inclusive, on account of: 
 
 Broken 1,293| tons 
 
 Bruised l,435f tons 
 
 Split l,353f tons 
 
 Worn out 28| tons 
 
 No fault 35f tons 
 
 Total 4,147 tons 
 
 The record is deceptive in some respects without an explanation. Many 
 of the breaks in the older rails were caused by punching bolt holes. The record 
 of the bruised or battered rails, which constituted the largest item, would have 
 been still greater, except for the fact that long pieces of track laid with soft or 
 so-called " pewter" rails showed up in such bad shape that they were taken up 
 and the rails sent to branches or used in sidings after having been in use only a 
 few months on the main line. 
 
 The road received the last of these soft rails in 1884, and the record given 
 below of the rails received in the following four years is very good. All of the 
 rails purchased in this period weighed 65 pounds per yard, but after 1888 an 
 80-pound section was adopted as standard. It will be observed that there 
 were no failures from bruising in the harder rails received after 1884. 
 
 Year Rolled. 
 
 1885 
 
 18 86 
 
 1887 
 
 1888 
 
 Total. 
 
 Total number of rails received during year 
 
 Rails removed to April 1st, 1889. Account: 
 
 23,208 
 
 39 
 
 
 13 
 
 
 52 
 
 23,156 
 
 0.22% 
 
 0.06% 
 
 21 
 
 29,171 
 
 5 
 
 12 
 
 
 17 
 
 29,154 
 
 0.06% 
 0.02% 
 
 7 
 
 41,678 
 
 43 
 
 
 16 
 2 
 
 61 
 
 41,617 
 
 0.14% 
 
 0.07% 
 
 24 
 
 30,366 
 
 4 
 
 1 
 2 
 
 30,359 
 
 0.02% 
 0.02% 
 
 7 
 
 124,423 
 91 
 
 Bruised 
 
 
 
 Split 
 
 42 
 
 No fault 
 
 4 
 
 Total 
 
 137 
 
 Number of rails in the track, April 1st, 1889 
 
 Defective rails: 
 
 124,286 
 
 
 
 
 
 
 
 * Cylindrical Wheels and Flat Topped Rails for Railways, D. J. Whittemore, Trans. Am. Soc. 
 of Civil Engrs., Vol. XXI, 1889, pp. 185, 186. 
 
DEVELOPMENT OF THE PRESENT SECTION 13 
 
 There appears to have been a critical period occurring about every twenty- 
 years in the history of the rail. When the iron rails replaced the old strap 
 irons and other early forms of track construction, they proved very satisfactory 
 for the light wheel loads of the day. The wheel loads, however, were rapidly 
 increased and soon demanded heavier sections and a metal better able to resist 
 wear at the running surface of the rail. It was claimed that the metal in the 
 larger sections was poorer than that found in the early iron rails. Various ex- 
 periments were tried with rails having steel heads, but it was not until the in- 
 vention of the Bessemer process for making steel, which enabled a stronger and 
 more uniform rail to be made, that the difficulty was successfully met. 
 
 The use of steel in place of iron for rails commenced about 1865 and enabled 
 heavier wheel loads to be used with safety. The early steel rails were generally 
 made of mild steel which, while suitable for the loads of the early seventies, was 
 found to be too soft for the heavier equipment of the next decade. The situ- 
 ation was unfortunately complicated by the experiments on the Pennsylvania 
 Railroad which showed, or seemed to show, that low carbon steel rails were to 
 be preferred to those made from steel of greater hardness, and for several years 
 following 1881 the rails were made too soft, and, while there was not a return to 
 the serious difficulties of the time of the iron rails, the condition of affairs was 
 far from satisfactory. 
 
 Relief was found by increasing the hardening constituents in the steel, 
 but with the constant increase in the weight of engines and cars, as well as the 
 greater density of traffic incident upon the growth of the industrial resources of 
 the country, the situation again reached an acute stage about 1905 when the 
 failures of rails became so numerous as to cause the gravest concern on the 
 part of those in charge of the operation of the roads. 
 
 The failures as before were principally a question of wear rather than of break- 
 age. It appeared that each increase in section produced a rail that wore out more 
 rapidly than the lighter section which preceded it. This condition was further 
 accented by the form of the American Society of Civil Engineers' sections with 
 their thin bases, which turned black in the rolls while the heads were still hot, and 
 the fact that the larger sections took more time to cool and so underwent a partial 
 annealing, making the metal more readily abraded. 
 
 Three principal reasons were advanced as to the probable cause of the 
 poor service of these latter rails. It was claimed that the wheel loads in use in 
 this country were exceeding the limits of strength of the steel in the rail and, 
 without resorting to extraordinary methods of manufacture and consequently 
 greatly increased cost, the rails could not be made to carry the loads imposed 
 
14 STEEL RAILS 
 
 upon them with a proper degree of safety. The standard sections then in use 
 were those of the American Society of Civil Engineers and this design of rail, in 
 the heavier sections then demanded, was stated to be an impracticable one to roll. 
 
 The manufacturers of rails proposed these explanations as the real reason 
 which accounted for the failures of the rails in service. The railways, on 'the 
 other hand, while admitting that the metal of the rails would not stand the 
 heavy wheel loads, claimed that this was due to the fact that the steel was of 
 poorer quality than that obtainable in rails of earlier make, and that sufficient 
 care was not being given to the details of manufacture in the various processes 
 at the mills. The increase in the number of rail failures of the type designated 
 as "crushed heads" and "split heads" the manufacturers claimed was caused by 
 the metal breaking down under the excessive pressure of the heavy wheel loads, 
 and the railways contended that they were due to some defect in the structure of 
 the individual rails. 
 
 No one at all conversant with the situation will attempt to maintain that 
 the subject is not a pressing one. The making of steel rails for use under 
 high-speed passenger trains is something more than a mere commercial propo- 
 sition. Both the producer and the consumer have great responsibilities in 
 the matter, and neither can lay them aside nor shift them upon the other. 
 
 2. Present Sections 
 
 Realizing the importance of the question, the American Railway Asso- 
 ciation appointed a special committee on Standard Rail and Wheel Sections. 
 This committee, through a subcommittee on which the manufacturers were 
 represented, devoted a large amount of time and attention to the matter of 
 sections and specifications for steel rails and presented a preliminary report to 
 the association, October 1, 1907. 
 
 While the A. S. C. E. section was apparently well adapted for the light- 
 weight rails of 65 pounds and 75 pounds in use when it was designed, the 
 increase in weight on railway wheels (see Fig. 2) necessitated a heavier rail, 
 and the manufacturers of rails claimed that it was difficult to make such rails 
 of the A. S. C. E. section, due to the thin edge of the base. 
 
 Accompanying the report of the committee were two series of proposed 
 standard rail sections: Series "A " designed to meet the requirements of those 
 who advocate a rail with thin head and a high moment of inertia, and series "B" 
 to meet the requirements of those who think that there should be a narrow, 
 deep head, with the moment of inertia a secondary matter. These sections are 
 shown in Plates VII and VIII. 
 
DEVELOPMENT OF THE PRESENT SECTION 
 
 15 
 
 The one known as Series "A" is characterized by a shallow head, wide base, 
 thin flanges, and a greater height of section than Series "B." It is appar- 
 ently advocated by those who think that more of the duty of the track should 
 be borne by the rail and less by the other elements. It is obvious that the 
 stronger the rail, as a beam or girder, the more the strains are distributed, and 
 the less need, therefore, for exacting attention to the other features of track 
 maintenance. Its advocates think that the distribution of metal between head, 
 web, and foot, is such that the rolling difficulties, and especially the question 
 of finishing temperatures, can be met with better success. 
 
 6aooo 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •$ 
 
 
 
 
 
 
 "§ 3Q00Q 
 
 8 
 
 % 
 
 £2Q000 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 QOOO 
 
 
 
 
 
 
 
 
 Curve Showing Increased 
 "Greatest Load'on each Me. 
 
 
 
 
 
 1 
 
 
 Dates was 
 
 Fig. 2.- 
 
 6QOOO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 £ 4a0 °° 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 § 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 2Q000 
 
 
 
 
 
 
 
 
 
 
 
 
 1Q0OO 
 
 
 
 Curve Showing the Increased 
 Average Weight on each Mb . 
 
 
 
 
 1 1 
 
 >s i9oo im m Dates ms 
 
 i Axle Loads, 1885 to 1907. (Railway and Engineering Review - 
 
 It is entirely problematic whether this section will prove the best of the 
 two under consideration, and especially whether the transference of more of the 
 duty to the rail will result in ultimate track economy. Those who oppose this 
 section fear that the shallow head is an element of weakness. According to 
 their view, with such steel as it is at present possible to get in rails, the pounding 
 of the heavy traffic will lead to such crushing and splitting of the heads, owing 
 to internal physical defects in the metal, that the section will prove a failure, 
 especially on roads with heavy wheel loads and dense traffic. 
 
 Series "B" is modified to meet this latter view. The distribution of metal 
 is believed, as in the "A" section, successfully to meet the manufacturers' criti- 
 cism, the head and foot in the 100-pound rail having slightly over forty per 
 cent each of the metal, and the web the balance. This section is weaker as a 
 girder than the "A" section, and it would appear that for lighter rails the 
 
16 STEEL RAILS 
 
 section "A" should preferably be used to obtain the greatest stiffness. Where 
 the wheel loads are sufficiently large to require the heavy head of the "B" 
 section, the design of the rail should, however, approach more nearly the latter 
 section, even at the expense of lack of stiffness which may be compensated for 
 by strengthening the track structure or increasing the weight of rail used. 
 
 There is no good reason why the same characteristics of design should be 
 carried out for an entire series. With the excessive loads borne by the heavier 
 rails more attention must undoubtedly be given to the effect of the concen- 
 trated pressure at the point of contact of the wheel and the distribution of 
 this force to the base of the rail. The bending stress while of equal impor- 
 tance can be reduced by strengthening the track structure as a whole. Hence 
 under the most severe conditions a section should be used in which ample 
 provision has been made for the former stresses and the question of the bend- 
 ing stress, while not lost sight of, becomes of secondary importance. As the 
 section decreases in weight the importance of the stiffness of the rail increases, 
 until in the lighter sections, supporting small wheel loads, a very much higher 
 relative moment of inertia is to be desired than in the heavier rails of the same 
 series. 
 
 The sections "A" and "B" have been proposed as "recommended prac- 
 tice" by the American Railway Association, and have been referred to the 
 American Railway Engineering Association to study and accumulate data and 
 make a report after the sections have been sufficiently tried in service to enable 
 an opinion to be formed as to their respective merits. 
 
 The American Railway Association Committee, in its report of October 1, 
 1907, submitted a statement of cardinal principles which should govern the 
 design of a series of rail sections, as follows: 
 
 (a) There should be such a distribution of metal between the head and 
 the base as to insure the best control of temperature in the manufacture 
 of the rail. 
 
 (b) The percentage of metal in the base of the rail should preferably be 
 equal to or slightly greater than that in the head, and the extremities of the 
 flanges should be sufficiently thick to permit the entire section to be rolled at 
 low temperatures. The internal stresses and the extent of cold straightening 
 will be reduced by this means to a minimum, and at the same time the texture 
 of the section will be made approximately homogeneous. 
 
 (c) The sections should be so proportioned as to possess as great an amount 
 of stiffness and strength as may be consistent with securing the best conditions 
 of manufacture and the best service. 
 
DEVELOPMENT OF THE PRESENT SECTION 17 
 
 (d) The following limitations as to dimension details of the sections are 
 considered advisable for the various weights per yard: 
 I. The width of base to be \ inch less than the height. 
 II. The fishing angles to be not less than 13 degrees and not greater than 15 
 degrees. 
 
 III. The thickness of the base to be greater than in the existing sections of 
 
 corresponding weight. 
 
 IV. The thickness of the web to be no less than in the existing A. S. C. E. 
 
 sections of corresponding weight. 
 V. A fixed percentage of distribution of metal in head, web, and base for 
 the entire series of sections need not be adhered to, but each section 
 in a series can be considered by itself. 
 VI. The radii of the under corner of head and top and bottom corners of base 
 
 to be as small as practicable with the colder conditions of rolling. 
 VII. The radii of the fillets connecting the web with head and base to be 
 as great as possible, for reinforcing purposes, consistent with securing 
 the necessary area for bearing surface under the head for the top of 
 the splice bar. 
 VIII. The sides of the head should be vertical, or nearly so. 
 IX. The radii of the top corners of the head should not be less than | inch 
 so long as the wheels continue under the present standard of the 
 Master Car Builders' Association. 
 
 The principles (a), (&), and (c), above enumerated, appear to cover the 
 proper design of T-rail sections. The (d) limitations as to dimension details 
 should be approached tentatively rather than regarded as a cardinal principle. 
 
 Since October, 1907, a large tonnage has been rolled of rails substantially in 
 accordance with the new sections, both series "A" and "B." It has been dem- 
 onstrated that these sections can be finished in the mill at a lower temperature 
 than the A. S. C. E. sections,* and therefore a finer grained and better wearing 
 rail should be secured with the new section. However, great care must be ex- 
 ercised at the mills to see that rails are actually rolled at lower temperatures. 
 The 90-pound series " A " is now used on a majority of the Western prairie 
 roads, and the " B " section is used on the group of coal roads in Maryland 
 and Virginia. On account of the heavier head found in the " B " section, it 
 seems to be preferred by the crooked roads of the East, especially those in the 
 
 * This refers to the temperature of the head; no part of the new sections is finished as cold as 
 the thin bases of the A. S. C. E. rails. 
 
18 STEEL RAILS 
 
 mountains of Pennsylvania, Virginia and Maryland ; while on the prairie roads, 
 where little curvature is found, the series " A " rail with the lighter head finds 
 more general use. 
 
 On June 5, 1907, a joint committee of the Pennsylvania Railroad system — 
 Mechanical and Civil Engineers east and west of Pittsburg — was appointed 
 to study the rail question, and on September 20, 1907, their labors resulted in 
 the designs for 85-pound and 100-pound rail sections shown in Plate IX. 
 
 It will be noted that the sides of the head are vertical in the 85-pound 
 section and sloping in the 100-pound section. This difference is not intentional, 
 but arises from the method of constructing the sections. The bearing surface 
 underneath the head for supporting the shoulder of the splice bar was con- 
 sidered of great importance, and the committee was not warranted, in the light 
 of past experience, in reducing this surface. This was therefore fixed at not 
 less than the existing bearing surface. 
 
 As the same equipment is run over the 85-pound and 100-pound sections, 
 there was no good reason why the width of the head on top should not be the 
 same in each case, and, after studying the contour of the wheel tread, this width 
 was fixed at 2| inches. The result was sloping sides in one case and vertical 
 sides in the other. 
 
 This section, known as the "P. S." section, is a step farther away from the 
 "A" section. It has a still heavier head, a narrower base, and thicker flanges 
 than the "B" section. The radius of the web is smaller, thus producing 
 more of a buttress where the head and web join. The experience of the 
 Pennsylvania system seems to be that with their heavy wheel loads and 
 dense traffic, and with the grade of steel that it is now possible to get in 
 rails, more rails fail from crushing and disintegration of the head, apparently 
 due to the pounding of the traffic, than from any other one cause, and 
 accordingly in this section the maximum effort has been made to strengthen 
 the rail in its weakest point. The distribution of the metal is satisfactory, 
 and the strength of the rail as a girder or beam is practically the same as the 
 "B" section. 
 
 In Europe a T-rail section is used. The Vignole rail used extensively 
 abroad was invented in England in 1836 by Mr. Charles Vignoles. Plate X 
 shows the type of Vignole rail used on the French Eastern Railway. It is their 
 intention eventually to modify the section by decreasing the height of the head 
 a little and increasing its width. The weight of the section is 45 kilos per 
 meter or about 91 pounds per yard. The maximum axle load is about 40,000 
 pounds. 
 
DEVELOPMENT OF THE PRESENT SECTION 19 
 
 Plate X shows the rail used on the Paris, Lyon and Mediterranean, weigh- 
 ing 48 kilos per meter, or 96.7 pounds per yard. The maximum axle load on 
 this road is, for passenger service, 38,000 pounds, and, for freight service, 35,000 
 pounds. 
 
 Plate XI illustrates types of German rails. The figures show that the 
 corner radius of the head in all the rails is y-g- inch, as prescribed by the " Tech- 
 nical Conventions of the Union," which also recommends a width of head of at 
 least 2\ inches with a minimum radius of the top of the head of 7f inches. On 
 the majority of the German rails the latter radius is from 7| to 8| inches. An 
 inclination of io is given to the rails, which is obtained by the use of wedge 
 tie plates. 
 
 Plate X shows the Vignole rail recently used on the Egyptian State Rail- 
 ways, of 46 kilos, or 92.7 pounds.* 
 
 In England the idea seems to prevail that a T-rail track is undesirable, 
 and a double-headed, or bull-headed, rail is generally used on the English rail- 
 ways. Plate XII shows the construction of the permanent way of the Midland 
 Railway of England, and Plate XIII the permanent way of the London and 
 North Western Railway of England. On the latter road the British standard 
 bull-headed rail is used, as shown on Plate XIV. Plate XV shows the British 
 Standard flat bottom rail. 
 
 For street railway work either a T-rail or grooved-rail section is employed, 
 as illustrated on Plates XVI and XVII, which show the sections recommended 
 by the American Electric Railway Engineering Association. This association 
 has only taken up the study of girder and high T-rail sections, and the sections of 
 tram rails given on Plate XVI, while representative of good practice, are not the 
 standard sections of the association. Plate XVIII shows the British Standard 
 tramway rails. 
 
 While the T-rail is generally recognized in this country as having superior 
 merits for street railway purposes, it is necessary, however, on heavily traveled 
 streets to use the grooved rail. The use of a number of different rail sections 
 in street railway work is open to the same objections as are found in steam 
 railway practice, and the tendency is toward the adoption of standard sections. 
 
 In determining the proper form for a rail, the subject should be considered 
 from two points: First, and most important, the duty required. Second, and 
 about equally important, the influence of detail of manufacture upon the char- 
 acter of the finished product. It is perfectly proper that all the stresses to 
 
 * This rail has been replaced by another weighing 95 lbs. per yard, according to British Standard 
 sections, which was laid from the year 1911. 
 
20 STEEL RAILS 
 
 which it will be subject should be considered and calculated, but its ability to 
 resist them will depend quite as much upon the character of the metal as upon 
 the form of section. 
 
 In considering the duty, we have first to examine the external forces acting 
 upon the rail, which consist of the pressure exerted by the wheel on the rail and 
 the supporting forces represented by the ties. When these are known, the 
 stress induced in the rail can be calculated for different sections. 
 
CHAPTER II 
 pressure of the wheel on the rail 
 
 3. Speeds of Modern Locomotives 
 
 The eight-wheel or American engine was formerly the favorite type for 
 fast passenger service. The arrangement of this engine provides a four-wheel 
 leading truck and four-coupled driving wheels and afforded ample starting 
 capacity for the trains of moderate weight used at that time. 
 
 The Atlantic type is the result of the demand for large heating surface and 
 grate area in combination with large driving wheels in an effort to meet condi- 
 tions which could not be met successfully by the preceding American-type 
 engine. The Atlantic-type engine combines a four-wheel leading truck and 
 four-coupled driving wheels with trailing wheels. 
 
 The increase in the weight of the train due to heavier equipment and longer 
 trains has resulted in the use of the Pacific locomotive with six-coupled wheels 
 in place of the Atlantic type with four-coupled wheels. The latter engine is 
 better suited to high-speed service than the former, but it cannot accelerate 
 heavy trains to running speed nor maintain speed on grades as well as the 
 Pacific. The internal friction of the Pacific engine is much greater than that 
 of the Atlantic and it reaches its speed limit sooner, and in fact these powerful 
 engines have not been able to show any material increase in the speed of our 
 fast trains. 
 
 A train * was recently made up for test purposes which was intended to 
 represent modern express equipment which could be hauled at high speed on 
 level track. The six cars weighed 350 tons, and the Pacific locomotive 194 tons, 
 total 544 tons. The Pacific locomotive, which was selected for its good record 
 on that line, was not able to accelerate the train to more than a fraction above 
 60 miles per hour on a straight level track where atmospheric conditions were 
 normal. 
 
 On several railways in the West it was for a time thought that it would be 
 necessary to electrify the mountain divisions in order to attain speeds which 
 would carry the large volume of traffic over the grades and avoid congestion 
 and blockade. The Mallet locomotives have overcome this difficulty, and their 
 
 * Railway Age Gazette, January 28, 1910. 
 
22 
 
 STEEL RAILS 
 
 remarkable performance has for the time rendered the electric locomotive on 
 mountain lines, where there are no long tunnels, unnecessary. 
 
 On account of the good results obtained by the use of the Mallet compound 
 locomotives it will prove interesting to consider the question of adopting these 
 machines for fast service. The principal advantage to be derived from the use 
 of the Mallet type appears to lie in its ability to develop enormous force at the 
 draw-bar, but it will be observed that these forces are only possible at compar- 
 atively low speeds. 
 
 At speed,* whatever the type may be, it is the boiler and not the adhesion 
 that limits the output of power. The moment the speed is increased by any 
 considerable amount, high draw-bar forces become impossible and the wheel 
 arrangement peculiar to the Mallet type unnecessary. The assumption even 
 of a moderate speed will permit wheel arrangements, now common, to absorb 
 the full power of the largest boilers now considered practicable. For example, 
 assume that a locomotive is used which is to have sufficient boiler capacity to 
 permit 2000 h.p. to be developed in compound cylinders at all practicable 
 speeds. Such a locomotive would require a boiler having in the neighborhood 
 of 5000 feet of heating surface which, if fired with coal, would need to be supplied 
 with 6000 or 7000 pounds per hour. The draw-bar force equivalent to 2000 h.p. 
 for several different speeds is as follows: 
 
 At 1 mile an hour, the tractive force will be 750,000 lbs. 
 
 5 miles 
 : 7* 
 10 
 20 
 30 
 50 
 
 150,000 
 100,000 " 
 75,000 " 
 37,500 " 
 25,000 " 
 15,000 " 
 
 Assuming the driving axle of the proposed locomotive to carry a load of 
 50,000 pounds, and assuming the adhesion to be 25 per cent, each driving axle 
 will serve to develop 12,500 pounds tractive force. A Mallet compound having 
 eight axles would be capable of developing a maximum tractive force of 100,000 
 pounds, which force is equivalent to the development of 2000 h.p. in the cylinder 
 at a speed of 1\ miles per hour. At speeds lower than this the adhesion derived 
 from eight axles will not permit the cylinders to develop this rated power, and 
 for speeds higher than this the full adhesion of eight axles will not be necessary 
 to the development of 2000 h.p. 
 
 * Railway Age Gazette, April 22, 1910. 
 
PRESSURE OF THE WHEEL ON THE RAIL 23 
 
 Table I shows fast and unusual runs in the last three decades.* The 
 foregoing table is severely condensed. f The time in every case is from the 
 beginning to the end of the run, including stops. In speeds alone, for moderate 
 distances, there has been little change since 1895. For example, the engines of 
 the Atlantic City Railway make substantially the same time as was made over 
 the same line ten years ago; but with the larger boilers and fireboxes now used, 
 heavier trains are hauled without loss of speed. The Empire State Express of 
 the New York Central, which for years was limited to four cars, now usually 
 has five cars, and still makes its trip of 440 miles at the scheduled speed of 
 
 * Locomotive Dictionary, 1909 Edition, Chicago. 
 
 f March 1, 1901. — The record of 107.9 miles an hour is given by an officer of the road. 
 The grade was descending, mostly at 30 feet per mile. 
 
 March 24, 1902. — This run was made on a descending grade, which for some of the way 
 was as much as 32 feet per mile. 
 
 June 21, 1902. — This run is notable by reason of the rising grade. Altoona is 861 § feet 
 higher than Harrisburg. 
 
 June 19, 1903. — This run was made without a stop, but there were two engines. The 
 weight of the train was 1,008,000 pounds. There are a number of long ascending grades in the line. 
 
 August 8, 1903. — On this and the later run between the same places there were, of course, 
 many changes of engines. The record gives no data concerning the sizes of the engines, but most 
 or all of them were of the most powerful types made in the United States at that time. 
 
 June 9, 1904. — On this run engines were changed at Bristol. The dimensions given are 
 those of the engine used on the second stage of the journey. A car was left at Bristol and the 
 weight given is the average weight for the whole journey. The first engine had drivers 6 feet 
 8 inches in diameter, four-coupled; cylinders, 18 X 26 inches. The train making this run was the 
 regular mail train scheduled to run regularly, without a stop, from Plymouth to London in 4 hours 
 25 minutes. 
 
 July 20, 1904. — This is the best record which has been made over this line. The run of 
 June 19, 1906, was made with one more car. 
 
 1905. — Eighteen hours between New York and Chicago is the regular schedule time of one 
 daily train each way over the New York Central lines, and one over the Pennsylvania, the latter 
 being about 60 miles shorter. There is no published record of less time through, though on many 
 occasions the trains of both roads have made up much lost time. The run of November 3, 1905, 
 is an example of what has been done in such cases. In this run the number of cars was three, 
 except over portions of the road where a dining car was added, making four. 
 
 October 23, 1905. — This run and that of May 5, 1906, were not undertaken with a view 
 of making the highest possible speed, and each of the divisions over which these trains traveled 
 has been traversed no doubt in shorter time; but these transcontinental records are notable for 
 the long distances covered, even though the time be not the very highest of which the engines are 
 capable. Both of these runs were made by special trains throughout, except that in the run of 
 May, 1906, the run east of Buffalo was that of the regular Empire State Express. 
 
 June 19, 1906. — On this run a distance of 12 miles was traversed in 8 minutes (90 m.p.h.). 
 
STEEL RAILS 
 
 rocpocp ■ -cp 
 
 6 S S 
 
 oo s ogg gogg gggo oooo oggo o ooos S g£ go O J 
 
 
 8§g,M Ph s g'-S g8\£» 
 ^odm gzom <Sp2h& 
 
 
 ■s.a .1 K§ 
 ho o z'ShsS 
 
 ill 1 1||| I Hi iiii III! fii! J fill ill ill HI 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — 
 
 
 
 
 
 
 d 
 
 
 :tf : 
 
 
 
 
 
 
 
 
 
 :« 
 
 # " : 
 
 : p4 
 
 
 
 c4 cJcJci H 
 
 
 
 
 i^ 
 
 i* 
 
 
 :Wo 
 
 
 
 - 
 
 6 
 
 
 
 :W§W I : 
 
 :W 
 
 
 
 W KffiM J« 
 
 
 
 
 IZ 
 
 PC 
 
 £ 
 
 c 
 > 
 
 
 
 1 
 
 - 
 
 1? 
 
 * 
 
 7 
 
 
 i 
 
 * 
 
 Pj 
 
 :=8P- 
 
 
 .jd* 
 
 OS 
 
 
 d ddd^^g. 
 
 
 - 
 
 |>^=3 ^H^<a fe|=8oi e>l>< « oi^^-g 4^ >^ I »d<e !Z jb 
 a;o&on r dlj&Q &ZJJ J£Od pflifcfc fcZoZ a p^msio fch4!s fcfcfcjpIo,j<j 
 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 25 
 
 cicici 
 
 oaaffi os ; 
 
 & = 
 
 :3U 
 
 .'- 
 
 Ill it* i 
 
 SI! J|| : 1 : i IS f 
 
 
 
 883 888838888 8 8 3 888 8S38 
 
 MB- « 
 
 1 -3 $ J ||J fill || || 
 
 C3<<j<! IccffiH SSfc^^z^ Q<«o<6cco o m ^S 6«H £< 
 
 <o oooo 
 
 illl gj. 2 j H H .llli Ji§-tlN1l 
 
 5 g j : d c i g >, : J £| 
 
 ^ll ^^5 '-"loGc 
 
 "111 ^™ tf-s-sll**) «sd£l*,8JJ I 
 
 ooSh izzjj o;ss5<oec 6o<.3^d:3p 5 
 
 ■* -^ 00 00 fflffiOO OUJUJQONH <M€o4<'«*r!-t^0000'-l <N 
 
 S|5 2|22 |S2i 
 
26 
 
 STEEL RAILS 
 
 g 3 gg-g gggg «ggg ggg g g >; 
 
 
 III 
 
 
 111 ^ IH&ii ll'il 1JM Hi 
 
 PhOoS OSmW PhHSH 3o2s 6<H 5 £ h 
 
 115 1 
 
 0.2.23 ^feO, 
 
 
 
 
 raglz; d,pqpL,,4 t-i^SO «P-, i-iEj-aj coSOJH pi^fn otPOJ^ 
 
 S3S SSS S S g 
 
 22 2 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 27 
 
 53.3 miles an hour and with remarkable punctuality. Some of the records above 
 80 miles an hour lack the elements necessary to make them entirely credible. 
 
 It will be observed that the notable performance of October 24, 1895, 
 which made a number of records that stood unsurpassed for years, was set aside 
 by another performance equally remarkable just ten years later — October 24, 
 1905. 
 
 Table II shows the best records for given distances. They are classified by 
 speeds alone, no account being taken of the modifying effects of load or grade, 
 or size of engine. It will be borne in mind that these records are all those of 
 steam locomotives. For distances under 15 miles, the electric locomotive 
 which was tried on the Berlin-Zossen line in Germany in 1903 made speeds over 
 130 miles an hour which have not yet been equaled by any steam locomotive. 
 
 TABLE II. — BEST RECORDED SPEEDS OF STEAM LOCOMOTIVES 
 
 (Locomotive Dictionary) 
 
 
 Distance, Miles. 
 
 Rate, Miles per Hour. 
 
 Date. 
 
 Road. 
 
 1 
 
 3,255 
 
 45.60 
 
 May 5, 1906 
 
 Various. 
 
 2 
 
 2,246 
 
 50.00 
 
 July 9, 1905 
 
 A. T. & S. F. 
 
 3 
 
 1,025 
 
 54.27 
 
 Feb. 15, 1897 
 
 C. B. & Q. 
 
 4 
 
 717 
 
 56.00 
 
 Nov. 3, 1905 
 
 Perm. 
 
 5 
 
 525 
 
 69.53 
 
 June 13, 1905 
 
 L. S. & M. S. 
 
 6 
 
 257 
 
 74.55 
 
 Oct. 24, 1905 
 
 Penn. 
 
 7 
 
 131 
 
 77.81 
 
 Oct. 24,1905 
 
 Perm. 
 
 8 
 
 55.5 
 
 78.26 
 
 May 14, 1905 
 
 Atlantic City. 
 
 9 
 
 50 
 
 79.00 
 
 June 8, 1905 
 
 Penn. 
 
 10 
 
 15 
 
 98.66 
 
 Mar. 24, 1902 
 
 C. B. &Q. 
 
 11 
 
 4.8 
 
 107.90 
 
 Mar. 1, 1901 
 
 S. F. & W. 
 
 Considering passenger service alone, the crowning achievement of the 
 locomotive designers of the past twenty years has been not speed alone, nor 
 speed and power combined, but speed, power, and reliability. The Penn- 
 sylvania trains running daily between Jersey City and Chicago, 905 miles, at 
 50.9 miles an hour, were on time at destination during the year ending June 11, 
 1906, 328 times out of 365, or 89.8 per cent of the trips of the year, west bound, 
 and 85.2 per cent of the trips east bound. Of the 37 late arrivals at Chicago 
 14 were not over 10 minutes late. The New York Central reported for its 
 similar trains a somewhat less favorable record; but the Central fast trains 
 travel at a higher speed, the distance being greater, and the trains were often 
 made up of five, six, or seven cars for a part of the distance. 
 
 From a table published in the Railroad Gazette of January 12, 1906, page 34, 
 giving the speeds of a large number of regular scheduled trains between London 
 and other English cities, the examples shown in Table III are selected: 
 
STEEL RAILS 
 TABLE III. — REGULAR ENGLISH EXPRESS TRAINS, 1905 
 
 
 Railway. 
 
 Miles. 
 
 Speed, Miles per 
 
 London to — 
 
 
 118 
 194 
 246 
 201 
 125 
 126 
 165 
 162 
 
 59.2 
 
 
 
 56.7 
 
 
 
 55.7 
 
 
 L. &N. W 
 
 56.1 
 
 
 56.8 
 
 
 
 57.5 
 
 Sheffield 
 
 
 58 1 
 
 Sheffield 
 
 
 57.2 
 
 
 
 
 On the two important long lines of Great Britain, the West Coast and the 
 East Coast routes to Scotland, the best schedules in effect in 1909 were as 
 follows: London & North Western and Caledonian, London to Glasgow (mid- 
 night train), 401.5 miles; 8 hours; rate, 50.2 miles an hour. Great Northern, 
 North Eastern and North British, London to Aberdeen (day train), 523.5 miles; 
 11 hours, 7 minutes; rate, 47.1 miles an hour. 
 
 Table IV below shows the best performances of American railroads. The 
 fast trains between New York and Philadelphia, which for years were notable 
 as the fastest trains in America, are now outclassed by the New York-Chicago 
 trains. The Pennsylvania's Chicago train is regularly scheduled from Jersey 
 City to North Philadelphia, 84 miles, in 83 minutes. 
 
 TABLE IV. — SCHEDULED SPEEDS OF FAST REGULAR TRAINS ON 
 
 AMERICAN RAILROADS, AUGUST, 1906 
 
 (Locomotive Dictionary) 
 
 Jersey City 
 New York. 
 New York. 
 Washington 
 Jersey City, 
 
 Chicago. . 
 Chicago. . 
 
 Chicago. . 
 Buffalo.. . 
 Boston. . . 
 Jersey City 
 Washington 
 
 Camden. 
 Atlantic 
 
 City 
 
 Southern Pacific, Union Pacific, Chi- 
 cago & North Western 
 
 Atchison, Topeka & Santa F6 
 
 New York Central & Hudson River 
 and Lake Shore & Mich. Southern. 
 
 Pennsylvania 
 
 New York Central 
 
 New York, New Haven & Hartford . . . 
 
 Pennsylvania 
 
 Central of New Jersey, Philadelphia & 
 Reading, Baltimore & Ohio 
 
 West Jersey & Seashore (Pennsylvania) 
 
 Atlantic City (Philadelphia & Reading) 
 
 964 
 905 
 440 
 232 
 224 
 
 18:00 
 17:46 
 
 53.5 
 50.9 
 53.3 
 46.4 
 47.0 
 
 At the International Railway Congress held at Berne, Switzerland, July 
 4 to July 16, 1910,* Mr. Blum expressed the opinion that the primary reason for 
 
 * Track Strengthening for Increased Weight of Locomotives and Speed of Trains. Bulletin of 
 the International Railway Congress. London and Brussels, 1910, Vol. XXIV, p. 2497. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 29 
 
 the strengthening of tracks was not the increase in the speed of the wheel loads 
 but the greater increase in the traffic, particularly freight traffic. The weakest 
 part of the track, the rail joint, was stated to be most fatigued not by the fast 
 trains but by the slower running trains. Speeds of 130.5 miles an hour had 
 actually been attained on a line having a weaker superstructure than that now 
 used for the express lines of the Prussian State Railways without any danger 
 to the track, and but little anxiety need be felt if higher speeds were used on 
 the existing track. That was the opinion of the majority of the railways from 
 which he had elicited opinions. 
 
 In calculating the effect of the wheel pressure on the track a speed of 60 
 miles per hour will be taken for passenger service and 40 miles per hour for 
 freight service. As is seen from the preceding tables these speeds are exceeded 
 in special runs, but from the evidence of tests we may conclude that, at the 
 higher speeds, while the dynamic force of the wheel increases the track possesses 
 at the same time a greater resisting power. This subject will receive further 
 consideration in the following chapters. 
 
 4. Weights of Modern Locomotives 
 
 IOO 
 
 060 ^QQQ 
 
 0440 ^OO OO 
 
 aqoq OOP 
 
 0880 ^OOnn Q QOO ARTICULATED 
 
 2440 An OO OO ARTICULATED 
 
 ■2662 AnOOO OOOn articulated 
 
 2882 AOOnn OQQOnARTICULATED 
 
 080 A OOOO 8 WHEEL 
 
 240 inOO 
 
 260 ioOOO 
 
 Anooo 
 
 2100 inOOOOO 
 
 oOO 
 
 460 A n oOOO 
 
 480 iononnn 
 
 A OOn 
 
 062 A OOOr 
 
 082 A OOOOo 
 
 044 iOOnn 
 
 FORNEY 4 COUPLED 
 
 n*A JOnOnn 
 
 FORNEY 6 COUPLED 
 
 rufi A O O n n n 
 
 „ 4 .. 
 
 flfifi A OOOn on 
 
 ,, 6 .. 
 
 94? And On COLUMBIA 
 
 K? An OOOn 
 
 PRAIRIE 
 
 ?S? inOnOOn 
 
 MIKADO 
 
 ?«? inOOOOn 8 COUPLED 
 
 .OOOQOn 
 
 ,OOr 
 
 ^loOOOr 
 
 284 ^nOOOOnn 
 
 246 An OO r 
 
 ^nOOOr 
 
 loo OOo 
 
 AnOnOr 
 
 An nO O n n 
 
 DOUBLE ENDER 
 
 AnOOOr 
 
 446 AnOOoPr 
 
 Fig. 3. — Classification of Locomotives (Whyte's System). 
 
 The classification given in Fig. 3 will be adopted for the different types of 
 engines under discussion. This locomotive classification is based on the repre- 
 sentation by numerals of the number and arrangement of the wheels, commenc- 
 
30 STEEL RAILS 
 
 ing at the front. Thus, 260 means a Mogul, and 460 a ten- wheel engine; the 
 cipher denoting no trailing truck is used. 
 
 The total weight is expressed in 1000 of pounds. Thus, an Atlantic loco- 
 motive weighing 176,000 pounds would be classified as a 442-176 type. If the 
 engine is compound, the letter C should be substituted for the dash ; thus, 442C 
 176 type. If tanks are used in place of a separate tender, the letter T should 
 be used in place of the dash. Thus, a double-end suburban locomotive with 
 two-wheeled leading truck, six drivers, and six-wheeled rear truck, weighing 
 214,000 pounds, would be a 266T 214 type. 
 
 The locomotives shown in Fig. 4 give an idea of the progress in engine 
 building from 1832 to 1911. The three engines shown in the illustration were 
 
 - Progress in Locomotive Building. 
 
 built by the Baldwin Locomotive Works. The engine in the upper figure of the 
 illustration was the Old Ironsides, built in 1832, and weighed in working order 
 about four and a half tons. The engine built in 1876 weighed 103,000 pounds 
 exclusive of the tender and had 22,250 pounds on each driving-wheel axle, while 
 the modern locomotive appearing in the lower figure weighs 462,450 pounds 
 exclusive of the tender and has axle loads exceeding 50,000 pounds. Fig. 5 
 shows a further comparison of early and modern locomotives. 
 
 The following tables illustrate the extent of development that has taken 
 place in the Pennsylvania locomotives from 1850 to 1910.* 
 
 * Paper on Scientific Management of American Railways by S. M. Felton, at the Congress of 
 Technology held at Boston, Mass., on April 10, 1911, under the auspices of the Massachusetts Insti- 
 tute of Technology. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 31 
 
 Passenger Locomotives. 
 
 Weight on each driving axle (approx.) 
 
 Weight on all drivers 
 
 Weight on trucks 
 
 Weight, total 
 
 Freight Locomotives. 
 
 Weight on each driving axle (approx.) 
 
 Weight on all drivers 
 
 Weight on trucks 
 
 Weight, total 
 
 7,500 
 
 I .-,.000 
 :;ii.()(io 
 45,000 
 
 59,500 
 178,500 
 
 93,500 
 272,000 
 
 13,000 
 26,000 
 19,000 
 45,000 
 
 54,100 
 216,450 
 
 24,495 
 240,945 
 
 Since 1910 the weight of the Pacific type 
 passenger engine shown in the table has been 
 increased from 272,000 pounds to 292,000 
 pounds by the introduction of superheaters and 
 mechanical stokers, in the case of some engines. 
 The Pennsylvania Railroad also have running 
 a Pacific type engine, built in 1912, weighing 
 317,000 pounds which has axle loads of about 
 66,000 or 67,000 pounds, and an experimental 
 Atlantic type engine with 68,800 pounds on an 
 axle. 
 
 In Tables V to XI are given weights and 
 axle loads of modern engines. The passenger 
 engines are noticeable for their trailing trucks 
 and four-wheel leading trucks, while the freight 
 engines commonly use a two-wheel leading truck 
 and no trailer, although the Mikado type (282) 
 is designed with the trailing truck. This engine 
 is a comparatively new development, and is not 
 shown in the tables. It is rapidly coming into 
 favor for heavy freight service in which the 
 Consolidation locomotive was formerly em- 
 ployed. 
 
 The weight of the engine can, of course, 
 be determined accurately. The maximum rail pressure of a driving 
 when the locomotive is running is, however, not at all indicated by the 
 load of the wheel on the rail. 
 
 wheel 
 static 
 
32 
 
 STEEL RAILS 
 
 The dynamic augment* is due to several causes: First, the effect of the 
 "excess balance" necessary to counteract the reciprocating parts is to cause 
 an impressed load. Second, the angularity of the main rod causes an increase 
 of pressure on the main wheel. Third, consideration of impact or imperfections 
 existing in the rolling stock or roadway. Fourth, the rocking of the engine on 
 its springs. 
 
 -AXLE LOADS AND TOTAL WEIGHTS OF ATLANTIC TYPE (442) 
 LOCOMOTIVES 
 
 (From data furnished by the American Locomotive Company) 
 
 Name of railroad 
 
 Works 
 
 Weight on leading truck, pounds. . 
 Weight on driving wheels, pounds 
 Weight on trailing truck, pounds. , 
 Weight, total of engine, pounds. . . 
 
 Weight of tender, pounds 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 Wabash 
 Brooks 
 
 42,000 
 111,000 
 
 38,000 
 191,000 
 130,000 
 7' 6" 
 30' Hi" 
 
 Oregon Short Line 
 Brooks 
 
 47,500 
 
 99,500 
 
 48,000 
 195,000 
 142,500 
 
 7' 0" 
 27' 7" 
 
 Southern 
 Richmond 
 42,000 
 114,500 
 39,500 
 196,000 
 144,500 
 7' 6" 
 
 Rock Island 
 Schenectady 
 49,000 
 116,000 
 37,000 
 202,000 
 150,000 
 7' 0" 
 30' 10" 
 
 TABLE VI. — AXLE LOADS AND TOTAL WEIGHTS OF PACIFIC TYPE (462) 
 LOCOMOTIVES 
 
 (From data furnished by the American Locomotive Company) 
 
 Name of railroad 
 
 Works 
 
 Weight on leading truck, pounds. . 
 Weight on driving wheels, pounds 
 Weight on trailing truck, pounds. . 
 Weight, total of engine, pounds i . . 
 
 Weight of tender, pounds 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 C, M. & St. P. 
 
 Vandal ia 
 
 N. Y. C. 
 
 Pa. Lines 
 
 Brooks 
 
 Schenectady 
 
 Schenectady 
 
 Pittsburg 
 
 43,200 
 
 49,500 
 
 48,000 
 
 48,000 
 
 167,300 
 
 162,000 
 
 170,500 
 
 176,500 
 
 41,300 
 
 44,500 
 
 48,000 
 
 45,500 
 
 251,800 
 
 256,000 
 
 266,500 
 
 270,000 
 
 152,600 
 
 145,900 
 
 164,500 
 
 144,000 
 
 14' 0" 
 
 13' 10" 
 
 14' 0" 
 
 13' 10" 
 
 35' 7" 
 
 34' Si" 
 
 36' 6" 
 
 35' 2f" 
 
 * See Rail Pressures of Locomotive Driving Wheels. Barnes. Trans. Am. Soc. Mech. Engra., 
 Vol. XVI, 1895, pp. 249-289. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 33 
 
 TABLE VII. — AXLE LOADS AND TOTAL WEIGHTS OF PRAIRIE TYPE (262) 
 LOCOMOTIVES 
 
 (From data furnished by the American Locomotive Company) 
 
 Name of railroad 
 
 Works 
 
 Weight on driving wheels, pounds 
 Weight on leading truck, pounds. 
 Weight on trailing truck, pounds. 
 Weight, total of engine, pounds . . 
 
 Weight of tender, pounds 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 C.,B. &Q. 
 Brooks 
 158,000 
 25,500 
 33,500 
 217,000 
 148,500 
 13' 4|" 
 
 P. L. W. of P. 
 
 Schenectady 
 163,000 
 
 26,500 
 
 40,500 
 230,000 
 140,000 
 14' 0" 
 34' 3" 
 
 P. R. R. 
 
 Schenectady 
 167,000 
 27,000 
 40,500 
 234,500 
 139,500 
 14' 0" 
 
 L. S. & M. i 
 Brooks 
 170,000 
 28,000 
 47,000 
 245,000 
 159,000 
 14' 0" 
 
 TABLE VIII. — AXLE LOADS AND TOTAL WEIGHTS OF TEN-WHEEL TYPE (460) 
 LOCOMOTIVES 
 
 (From data furnished by the American Locomotive Company) 
 FREIGHT LOCOMOTIVES 
 
 Name of railroad 
 
 Works '. . 
 
 Weight on driving wheels, pound. 1 
 Weight on leading truck, pounds. 
 Weight, total of engine, pounds. . 
 
 Weight of tender, pounds 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 C. & N. W. 
 
 Schenectady 
 135,000 
 46,000 
 181,000 
 143,500 
 14' 10" 
 25' 10" 
 
 D. N.W. &P. 
 
 Schenectady 
 143,000 
 47,000 
 190,000 
 149,000 
 14' 10" 
 25' 9" 
 
 St.L.&S.F. 
 Brooks 
 142,000 
 50,000 
 192,000 
 127,000 
 14' 10" 
 25' 9" 
 
 C. P. R. 
 
 Montreal 
 143,000 
 52,000 
 195,000 
 131,000 
 14' 10" 
 
 PASSENGER LOCOMOTIVES 
 
 Name of railroad 
 
 Works 
 
 Weight on driving wheels, pounds 
 Weight on leading truck, pounds. 
 Weight, total of engine, pounds. . 
 
 Weight of tender, pounds 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 Ore. R. R. & Nav. Co, 
 Brooks 
 161,000 
 44,000 
 205,000 
 164,000 
 13' 10" 
 25' 10" 
 
 N. Y. C. 
 Schenectady 
 158,000 
 51,000 
 209,000 
 148,000 
 15' 10" 
 26' 10|" 
 
 D. N.W. &P. 
 
 Schenectady 
 161,500 
 49,500 
 211,000 
 142,000 
 14' 10" 
 25' 9" 
 
 D., L. &W. 
 
 Schenectady 
 170,000 
 48,000 
 218,000 
 135,500 
 14' 4" 
 25' 6" 
 
34 
 
 STEEL RAILS 
 
 -AXLE LOADS AND TOTAL WEIGHTS OF MOGUL TYPE (260) 
 LOCOMOTIVES 
 
 (From data furnished by the American Locomotive Company) 
 
 Name of railroad. . . . 
 
 Works 
 
 Weight on driving wheels, pound: 
 Weight on leading truck, pound; 
 Weight, total of engine, pounds 
 Weight of tender, pounds 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 TABLE X. — AXLE LOADS AND TOTAL WEIGHTS OF CONSOLIDATION TYPE (280) 
 LOCOMOTIVES 
 
 (From data furnished by the American Locomotive Company) 
 
 Name of railroad 
 
 Works 
 
 Weight on driving wheels, pounds. . 
 Weight on leading truck, pounds. . . 
 Weight, total of engine, pounds. . . . 
 
 Mich. Cen. 
 Schenectady 
 
 215,500 
 25,500 
 
 241,000 
 
 154,000 
 
 17' 6" 
 
 26' 5" 
 
 C, I. &s. 
 
 Brooks 
 214,000 
 27,500 
 241,500 
 148,000 
 17' 3" 
 26' 5" 
 
 Union R. R. 
 Pittsburg 
 226,000 
 25,000 
 251,000 
 136,000 
 15' 7" 
 24' 4" 
 
 D. &H. 
 
 Schenectady 
 223,000 
 29,000 
 252,000 
 151,000 
 17' 0" 
 25' 11" 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 -AXLE LOADS AND TOTAL WEIGHTS OF FOUR-CYLINDER ARTIC- 
 ULATED COMPOUND LOCOMOTIVES 
 
 (From data furnished by the American Locomotive Company) 
 
 Type 
 
 Name of railroad 
 
 Works 
 
 Weight on driving wheels, pounds. 
 Weight on leading truck, pounds. . 
 Weight on trailing truck, pounds. . 
 
 Weight, total of engine 
 
 Weight of tender 
 
 Wheel base, driving 
 
 Wheel base, total of engine 
 
 0660 
 
 B. &0. 
 
 Schenectady 
 
 334,500 
 
 334,500 
 
 139,000 
 
 10' 0" & 10' 0' 
 
 C. &0. 
 
 Schenectady 
 324,000 
 22,000 
 46,000 
 392,000 
 163,000 
 10' 0" & 10' 0' 
 48' 10" 
 
 St. L. & S. F. 
 Schenectady 
 360,000 
 25,500 
 32,500 
 418,000 
 150,000 
 15' 6" & 15' 6' 
 56' 10" 
 
 D. &H. 
 
 Schenectady 
 445,000 
 
 445,000 
 167,000 
 ' 9" & 14' 9" 
 40' 2" 
 
pressure of the wheel on the rail 35 
 
 5. Effect of Excess Balance and Angularity of Main Rod 
 
 ISOOO 
 
 
 
 (/ 
 
 B.D^ 
 
 
 
 1/ 
 
 
 IOOOO 
 5000 
 
 ^N^X. 
 
 --</ 
 
 
 
 
 
 
 
 25000 
 20000 
 ISOOO 
 
 
 
 
 "Xs^- 
 
 
 
 
 
 
 
 
 
 IOOOO 
 5000 
 
 V 
 
 / 
 
 
 
 V 
 
 J 
 
 
 
 ALL. DRIVERS 
 
 RIGHT DRIVE! 
 
 LRS-^ C LEFT DRIVERS 
 
 ©£)©©© 
 
 Fig. 6. — Rail Pressures. Eight-wheel Engines. (Am. Ry. M . M . Assn.) 
 
 The effect of the excess balance and the angularity of the main rod can 
 
 be accurately calculated, and is shown on the diagrams given in Figs. 6 to 12.* 
 
 The calculation for the diagrams given on Figs. 6 and 7 were all made from 
 
 * Figs. 6 and 7 are taken from Proceedings Am. Ry. M. Mech. Assn., 1895. Figs. 8 to 12 are 
 from data kindly furnished by the American Locomotive Co. 
 
STEEL RAILS 
 
 20000 
 ISOOO 
 IOOOO 
 
 D 9P R,GHT 18 0° C 2 
 
 O 30 
 
 
 
 
 >W 
 
 
 
 
 
 
 
 
 
 
 
 
 N^ 
 
 
 
 
 
 
 
 \ 
 
 ■FlB.o/V 
 
 9Q !80_ 
 
 
 /" 
 
 \ 
 
 
 M.D. 
 
 / B.D. 
 
 \ , 
 
 /^' 
 
 ^> 
 
 ***4= 
 
 =^r 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 \ 
 
 
 
 A 
 
 \ 
 
 
 
 
 / \ 
 
 
 \ 
 
 J \ 
 
 /M 
 
 ^■"^ 
 
 
 \ 
 
 
 
 
 
 \> 
 
 ^J 
 
 
 
 
 \ 
 
 \ J 
 
 
 
 \ 
 
 
 
 
 \ 
 
 \ 
 
 
 
 
 \J 
 
 
 
 ALL DRIVERS 
 
 OQOOO 
 
 Fig. 7. — Rail Pressures. Ten-wheel Engines (Light Weights). (Am. Ry. M. M. Assn.) 
 
 data obtained from eight- and ten-wheel engines on the C. M. & St. P. Ry. 
 The following are the principal dimensions and weights of each * 
 
 * Report of Committee on the Wear of Driving-Wheel Tires, Proceedings Am. Ry. M. Mech. 
 Assn., 1895. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 37 
 
 
 Eight-wheel Engine. 
 
 Ten-wheel Engine. 
 
 
 16 by 24 inches 
 
 160 pounds 
 
 56 inches 
 
 8 feet, 6 inches 
 
 7 feet, 2\ inches 
 
 2i inches 
 
 480 pounds 
 
 54,000 pounds 
 
 19 by 26 inches 
 
 
 
 
 
 
 
 
 10 feet 
 
 
 3J inches 
 
 
 
 
 
 
 
 The piston, piston rod, crosshead, and front end of the main rod are taken 
 as reciprocating parts, the back end of main rod as a revolving weight, in all 
 calculations which follow. 
 
 The weights of the ends of the rods were found by supporting each end 
 at the center of the box or bearing, and resting them alternately on scales. 
 
 The eight-wheel engines had the entire weight of the reciprocating parts 
 balanced, by adding one-half this weight in each driving wheel to the weight 
 necessary to balance the revolving parts' when weighed at the crank pin. The 
 ten-wheel engines were not counterbalanced alike, but all agreed in having 
 the forward and back wheels overbalanced; that is, with a heavier counter- 
 balance than that required to balance the revolving parts only; while the main 
 wheels of thirty-five of the fifty-three engines from which measurements were 
 taken were underbalanced for the revolving parts alone, and all of them under- 
 balanced according to the rule of adding to the weight necessary to balance 
 the revolving parts two-thirds of the weight of the reciprocating parts, divided 
 equally between the driving wheels. 
 
 The counterbalance in the wheels of each of these engines was carefully 
 weighed by resting the journals of each pair of drivers on level straight edges, 
 placing the crank horizontally, and hanging on the crank pin a sufficient weight 
 to just balance the counterbalance opposite. From this weight the weight of 
 the revolving parts attached to that pin was subtracted, the remainder being 
 the amount of overbalance weighed at the crank pin. If the weight of the 
 revolving parts exceeded the weight so found, of course the wheel was under- 
 balanced by the amount of such excess. 
 
 The actual average condition of the counterbalance in the wheels of the 
 fifty-three ten-wheelers was as follows: 
 
 Average overbalance weighed at the crank pin above that required to 
 balance revolving parts only: 
 
 Front wheel 271 pounds overbalance 
 
 Main wheel 80 pounds underbalance 
 
 Back wheel 237 pounds overbalance. 
 
38 STEEL RAILS 
 
 These weights* are used in the calculations for the ten-wheel engines 
 plotted on Fig. 7. 
 
 The following formulae have been used in calculating the forces in action: 
 NOTATION 
 P = Pressure of one driving wheel on rail. 
 W = Weight of each wheel on rail, engine at rest. 
 C = Centrifugal force of the excess weight in the counterbalance over that 
 
 required to balance the revolving parts. 
 A = Horizontal accelerating (or when negative retarding) force of the recip- 
 rocating parts. 
 Pi = Pressure against crosshead pin from steam in cylinder. 
 
 a = Angle of the crank with the horizontal. 
 N = Ratio of length of main rod to length of crank. 
 
 Hence, P = W - C sin a + ( Pl ~ A) • (1) 
 
 'A/2 
 
 Vsi 
 
 snra 
 But, 
 
 w = Weight of the excess in the counterbalance over that required to balance 
 the revolving parts. 
 
 v = Velocity of the center of gravity of the overbalance. 
 
 r = Radius of the center of gravity of the overbalance. 
 w' = Weight of the reciprocating parts. 
 v' = Velocity of the crank pin. 
 
 I = Length of the crank. 
 
 g = The acceleration of gravity, 32.16. 
 
 Hence, C =—, (2) 
 
 gr 
 
 A W ' V ' 2 /n\ 
 
 A = — =- cos a. (3) 
 
 gi 
 
 Or, by substituting in (1) the values of C and A given in (2) and (3), 
 
 Pi T~ cos a 
 
 wv 2 , V gl / /,x 
 
 P = W sin cH y (4) 
 
 gr ijp_ _ 1 
 
 V sin 2 a 
 
 The above formulae include the centrifugal force of the overbalance in the 
 
 drivers, the effect of the acceleration and retardation of the reciprocating parts, 
 
 and the angularity of the main rod. Formula (3), for the acceleration of the 
 
 reciprocating parts, assumes that they move as they would were the main rod 
 
 * These weights are the equivalent weights at a distance from the center equal to the crank length, 
 and not the actual counterbalance weights used. 
 
PRESSURE OF THE WHEEL ON THE RAIL 39 
 
 infinitely long, but the error this produces is too small to affect the accuracy of 
 the results, while the formulae are much simplified. 
 
 The left-hand ends of the diagrams correspond to the position of the engine 
 when the right crank is on the forward center, positive rotation being that 
 produced by running the engine forward. 
 
 The pressures upon the piston used in the calculation for Figs. 6 and 7 
 were obtained from actual indicator cards taken at these speeds, and with a 
 point of cut-off found by the examination of a large number of cards to be the 
 usual point at which an engine is worked at the speed taken. 
 
 The points of cut-off used are: 
 
 Eight-wheel engine, just starting, 22 inches; 40 miles per hour, 6 inches; 
 60 miles per hour, 6 inches. 
 
 Ten-wheel engine, just starting, 22 inches; 10 miles per hour, 13 inches; 
 20 miles per hour, 11 inches; 30 miles per hour, 8 inches; 40 miles per hour, 
 6 inches; 60 miles per hour, 5f inches. 
 
 Curves for just starting, ten and twenty miles per hour, show that the total 
 pressure of the main driver on the rail is always greater at these speeds and 
 cut-offs than the actual weight of driver on the rail when the engine is at rest. 
 
 Fig. 13. — -Damaging Effect of Badly Balanced Locomotive. 
 
 This is due to the angularity of the main rod always causing an increase of 
 pressure on the main wheel. There is, of course, a corresponding upward 
 pressure on the guide, reducing the weight on the truck. 
 
 Figs. 8 to 12 are for heavier engines and are calculated from some of the 
 largest engines that have been built of each type. 
 
 Fig. 13 shows, the damaging effect upon the track of a badly balanced 
 locomotive. 
 
40 
 
 STEEL RAILS 
 
 A 
 
 
 
 
 
 
 
 / 
 
 \ 
 
 
 
 
 -=-- 
 
 
 *r=r=: 
 
 =.*&- 
 
 
 
 
 
 
 
 /~ 
 
 s ~^ v 
 
 
 
 
 
 ----- 
 
 ^--m 
 
 :=. F ^, 
 
 7000O 
 600OO 
 
 
 
 
 
 
 
 / 
 
 „\ 
 
 5OO0O 
 
 "^-Cr" 
 
 -.--J 
 
 
 — i-^ 
 
 70000 
 
 
 
 
 MD 
 
 
 
 
 f _ 
 
 
 50O0O 
 
 v --» 
 
 ~v^ 
 
 ^L 
 
 _^>* 
 
 
 
 
 A 
 
 I4O0OO 
 
 
 
 
 W 
 
 120000 
 
 
 
 
 
 rev 
 
 
 
 
 
 
 
 ftso»* 
 
 r 
 
 
 
 ^ J 
 
 
 h 
 
 
 
 \_y 
 
 
 
 aoooo 
 
 70000 
 
 80 Ml 
 
 PER MR. 
 
 
 Vd. 
 
 
 
 
 //"" 
 
 N\ 
 
 
 ^ — 
 
 =^Z 
 
 / x 
 
 p.bT\ 
 
 
 
 
 I6§Q0 ©GO©© 
 
 Fig. 8. — Rail Pressures. 442 (Atlantic) Type Engines: Cylinders 21|" X 26", Wheels 79", Workin 
 Pressure 180 lbs. (Am. Locomotive Co.) 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 
 
 
 \M.D, 
 
 
 
 
 
 
 
 
 r.&B.Df 
 
 9Q I8Q 27Q 36Q 
 
 
 
 
 
 V^ M.D. 
 
 60O0O 
 
 
 
 
 
 -~ « 
 
 .^.-^ 
 
 _= = - 
 
 E8B.D 
 
 50000 
 
 
 
 
 
 
 IO Ml. 
 
 =ER HR. 
 
 
 
 
 ' / 
 
 -^M.D. 
 
 ""----___ 
 
 ■ >/ 
 
 X 
 
 \ 
 
 / 
 / 
 / 
 
 
 
 
 
 
 
 
 
 >v M.D. 
 
 -==.-=.-= 
 
 .==^- 
 
 ^"= — 
 
 f^rdV-' 
 
 20 Ml. PER HR. 
 
 
 
 
 M.D. 
 
 *=■- 
 
 — ^" 
 
 ----- 
 
 — -- 
 
 
 
 
 •vM.D. 
 
 f.sb.d\ 
 
 ^.-~ 
 
 — ^>- 
 
 ■*= r - 
 
 £^~ 
 
 40 Ml. PER HR 
 
 
 
 
 A 
 
 
 
 
 H 
 
 
 IOM.P.H 
 STARTING 
 
 
 
 
 
 
 20M.RH. 
 
 30M.PH. 
 
 m 
 
 
 
 
 
 
 
 
 
 40M.PH. 
 
 
 ^/ 
 
 /50MPH 
 
 
 
 VA- 
 
 60M.RH. 
 
 
 ALL DRIVERS 
 
 3HT DRIVERS^ £ LEFT DRIVERS 
 
 Fig. 9. — Rail Pressures. 462 (Pacific) Type Engines: Cylinders 22" X 28", Wheels 79", Working 
 Pressure 200 lbs. (Am. Locomotive Co.) 
 
42 
 
 STEEL RAILS 
 
 RIGHT SIDE 
 
 9Q iao 27o 3e o 
 
 
 
 
 /- 
 
 -V- 
 
 
 / 
 
 
 / 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 ER HOUR 
 
 
 S 
 
 1 
 
 
 2 
 
 ?0 3 
 
 50 oooo 
 
 
 
 
 \» 
 
 
 
 
 / / 
 
 VT 
 
 B *^\ 
 
 
 ^c- 
 
 
 
 — \ 
 
 SOOOO 
 
 \ 
 \ 
 \ 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 \ 
 
 / 
 
 \ 
 
 50000 
 
 K-" 
 
 
 
 ^i 4 !^ 
 
 ^ 
 
 
 
 
 
 
 
 
 
 //<:: 
 
 ^v 
 
 50000 
 
 v 
 
 
 / 
 
 \ 
 
 
 
 
 A 
 
 
 
 
 
 v\ 
 
 
 
 sjsr« 
 
 
 a 
 
 
 
 
 
 ^■^ 
 
 
 
 
 
 
 
 K 
 
 \ y 
 
 
 3 ° MPM 
 
 ISOOOO 
 140000 
 I3QOOO 
 
 
 
 
 
 
 \y 
 
 
 
 ALL DRIVERS 
 
 
 
 
 ~N"> 
 
 
 
 / / 
 
 
 ^T 
 
 
 / 
 
 N^ 
 
 Fig. 10. — Rail Pressures. 460 (Ten-wheel) Type Engines (Heavy Weights) : Cylinders 22" X 26" 
 Wheels 69", Working Pressure 200 lbs. (Am. Locomotive Co.) 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 
 
 RIGHT 
 
 SIDE 
 
 
 60000 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 70000 
 60000 
 
 
 
 
 
 
 
 
 
 
 -=-— — 
 
 "-jizj^ 
 
 f^ 
 
 ^=- 
 
 
 
 30MI. 
 
 
 
 
 RIGHT 
 
 SIDE 
 
 O 56 
 
 
 
 
 
 
 
 y 
 
 \\ 
 
 
 
 / 
 
 
 
 
 X 
 
 / 
 
 
 
 
 
 f - 
 
 "N \ 
 
 
 
 // 
 
 1/ 
 
 \\ 
 
 \JV 
 
 ^ ', 
 
 3 
 
 \ 
 
 / 
 / 
 
 
 
 RIGHT DRIVERS. 
 
 
 Fig. 11. — Rail Pressures. 260 (Mogul) Type Engines: Cylinders 21" X 28", Wheels 63", Working 
 Pressure 200 lbs. (Am. Locomotive Co.) 
 
STEEL RAILS 
 
 
 
 
 
 
 
 
 
 "^ 
 
 
 
 
 
 
 
 
 
 b;d.> 
 
 
 700O0 
 
 eoooo 
 
 
 
 
 ^ 
 
 
 
 
 i- l i20v. 
 
 soooo 
 
 
 
 
 
 
 
 /~ 
 
 V M.D. 
 
 
 
 M 
 
 ^ 
 
 *"' 
 
 
 
 
 y^7J"~^v 
 
 
 
 
 
 ~N_ 
 
 
 
 
 y^ 
 
 ~-<\ 
 
 
 ^v 
 
 ^/J 
 
 ^Cpd 
 
 \ 
 
 
 -v 
 
 
 
 
 
 "\\ 
 
 
 
 
 \ 
 
 ^ 
 
 
 
 
 
 
 
 A 
 
 
 
 
 A 
 
 
 
 
 ' \ 
 
 
 """""n 
 
 
 A 
 
 
 J) 
 
 
 ^N 
 
 
 
 ^-fr— 
 
 20M.P.H. 
 
 
 
 
 
 
 
 i 
 
 
 30MRM. 
 
 
 
 
 
 /UoMPH 
 
 
 
 
 /sOM.P.M. 
 
 
 
 \\ 
 
 
 
 
 u 
 
 
 
 RIGHT DRIVERS — 
 
 eoooo oooo© 
 
 Fig. 12. — Rail Pressures. 280 (Consolidation) Type Engines: Cylinders 23" X 32", Wheels 63", 
 Working Pressure 200 lbs. (Am. Locomotive Co.) 
 
pressure of the wheel on the rail 45 
 
 6. Effect of Irregularities in the Track 
 
 Fig. 14 shows the exaggerated profile of the rail observed by M. Cuenot 
 in his track experiments. 
 
 Figs. 15 and 16 show the rail profile taken with a Railroad Automatic 
 Track Inspector machine. These diagrams show the unloaded profile of the 
 rail, or the permanent set left in it by the passage of the trains. Evidently 
 the loaded profile will be below the unloaded line, and both profiles will probably 
 show the same general features, as indicated by the approximate loaded position 
 of the rail shown by the dotted line in Fig. 16. 
 
 The wheel as it passes over the curved surface of the rail shown in the 
 
 figure is constrained to move in a curved path whose radius is about 5000 feet, 
 
 Mv 2 
 and the pressure of the wheel on the rail is the centrifugal force, C = -5-' 
 
 directed away from the center of curvature. For 30,000-pound wheel loads, 
 
 M = — = ' 1fi , where the units are in pounds and feet. 
 
 For 60 miles per hour, 
 
 _ 5280 X 60 
 
 V 60 x 60 
 
 and R = 5000. 
 
 mi. e n Mv 2 30,000 x 88 x 88 1 , , c , 
 
 Therefore C - -^ = ^ x mQ - 1445 pounds, 
 
 which is the excess wheel pressure caused by the irregularity in the track shown 
 by the figure. To be on the safe side, it would seem desirable to increase this 
 amount. If, however, 4000 pounds be taken to represent the excess wheel 
 pressure, due to this cause, an ample factor of safety will apparently have been 
 provided. 
 
 It will be seen from the above that the increase of wheel pressure, due to 
 any change in the grade line, will be so small as to be negligible. It is good 
 practice to change from one rate of grade to another with a vertical curve, 
 changing the grade at each 100-foot station by 0.1 feet; this would give a radius 
 for the vertical curve of 50,000 feet, and a corresponding value for C of about 
 150 pounds. 
 
 Let us now consider the path of the wheel when passing over the summit 
 between two of the depressions shown in the track profile, Fig. 15. When the 
 wheel is in the act of leaving the valley, or depression, its path lies in a direction 
 away from the surface of the rail before it. It is, however, under the influence 
 of two forces, — neglecting for the moment the action of the springs. First, its 
 
 88 feet per second, 
 
STEEL RAILS 
 
 i+V. y ire?-l\M 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 47 
 
 momentum, acting along a line of direction tangent to the vertical curve of the 
 rail; and, second, the force of gravity. The trajectory of the wheel acting under 
 
 - Rail Profile taken with a Railroad Automatic Track Inspector Machine. 
 
 
 
 
 
 
 
 •"\ 
 
 
 
 
 
 
 \ 
 
 
 
 SURFACE O 
 
 F TRACK^ 
 
 J 
 
 
 
 
 
 
 
 
 --^ 
 
 ^-^^ 
 
 
 -~^ ,*.-- 
 
 ' 
 
 
 
 ~~- ~~_r" 
 
 •^O^so- - 
 
 '"PATH O 
 
 F WHEEL 
 
 Fig. 16. — • " Valley " or Local Depression in Track Profile. 
 
 these forces will be a parabola with its axis vertical. The greatest height of 
 ascent, y, and the horizontal range, x, are given by the following equations: 
 
 y = h sin 2 a, 
 
 x = 2 h sin 2 a. 
 
 h being the ideal height due to the velocity, we therefore have for a speed 
 of 60 miles per hour, 
 
 v 2 = 2 gh, 
 
 ~2g 
 
 , = 121 feet. 
 
 2 x 32.16 " 
 Fig. 17 is derived from the 
 same record as that from which 
 the diagram of Fig. 15 has been 
 taken and shows a summit be- 
 tween two depressions in the 
 profile of the track. We see from 
 the figure that the value of a is ' 
 
 0° 14' . Substituting these Values Fig. 17. — Summit between Two Depressions of Track Profile. 
 
 in the expressions for x and y, there results for the greatest height of ascent 0.002 
 feet, or 0.024 inches, and for the horizontal range, 1.97 feet. 
 
48 
 
 STEEL RAILS 
 
 SET 
 
 LOAD 
 
 iW 
 
 FREE 
 
 1 V4" 
 
 21870 
 
 5/8" 
 
 27300 
 
 SET 
 
 LOAD 
 
 4V 
 
 FREE 
 
 2" 
 
 20650 
 
 1*6" 
 
 25812 
 
 Fig. 18. — Locomotive Driving Wheel Springs. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 49 
 
 Fig. 17 shows that this curve practically coincides with the profile of the 
 rail. It is hardly conceivable, therefore, that the wheel can leave the rail when 
 passing from one depression to another, as the action of the springs, as well 
 as the resilience of the rail, which would tend to prevent this, are neglected in 
 the preceding discussion. 
 
 7. Effect of Rocking of the Engine 
 
 The pressure caused by the rocking of the engine on its springs can best 
 be determined by observing the amount the springs deflect under their load. 
 
 By referring to Plates XX and XXI, it will be seen that the wear of the 
 guides of the driving boxes will give a means of telling how much the springs 
 deflect. The maximum amount of wear 
 is probably about one inch. Turning to 
 Figs. 18 and 19, which show the springs 
 used for the locomotive drivers, we see 
 that the depression of the spring one inch 
 corresponds to a range of pressure of 
 about 8000 pounds. However, as the 
 rocking of the engine causes at times a 
 less pressure as well as a greater on the 
 springs, one-half of this amount, or 4000 
 pounds, should be taken as the pressure which will cause the spring to deflect 
 an amount equal to that obtained under service conditions. 
 
 A careful series of experiments have been made by Messrs. Coes and 
 Howard * to determine the live load on locomotive driving springs under actual 
 running conditions. 
 
 The apparatus consists of three distinct parts: (1) a recording device, 
 which fits on the spring band or saddle; (2) a spanner bar or beam, which is 
 fastened to each end of the spring link hangers and is connected to the record- 
 ing apparatus; (3) a battery box, which is in the cab with rheostat, switches, keys, 
 clock, and all the necessary controlling mechanism. See Figs. 20, 21, and 22. 
 
 The recording apparatus (Fig. 20) is in a box, which is bolted to a steel 
 plate (1) by four bolts; this plate in turn is bolted to a U-shaped band (2) which 
 is fastened to the spring band by four hardened steel set-screws. The record 
 is made on metallic-faced paper, 4 inches wide and about 750 feet long. This 
 paper is wound up upon a detachable drum (3) and travels across a curved brass 
 guide plate (4) under two guide rolls (5) on to the main drum (6). 
 
 * Thesis 1906 at the Massachusetts Institute of Technology, under the supervision of Prof. Lanza. 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 f 
 
 Jr 
 
 
 
 s 
 
 
 
 
 
 o 
 
 
 
 
 
 i a. 
 
 DEFLECTION INCHES 
 
 - Deflection of Locomotive Springs. 
 
50 
 
 STEEL RAILS 
 
 
 Bm 
 
 -QBpW 
 
 
 
 dp 
 
 i ^^§Slrr 
 
 I jb 
 
 
 S^ heostat 
 
 
 t&- 
 
 ^Br 
 
 __3 
 
 DRY BATTERIES 
 
 Fig. 20. — Recording Device and Cab Controlling Mechanism for Testing Driving 
 Wheel Springs. (Coes and Howard.) 
 
 
 
 
 £JP 
 
 " ^BP* 
 
 ■ // » 
 
 2. 
 
 ik^Ka 
 
 
 -J&b^jM 
 
 
 
 * ^■P^* s 
 
 
 
 Mgg 
 
 
 
 P^ ^ 
 
 
 Bp^ •'v — 
 
 Fig. 21. — Recording Device in Place on Driving Wheel Spring. (Coes and Howard.) 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 51 
 
 The main drum is driven by a motor (m) behind the curved plate, through 
 a worm and wheel drive (w). The main stylus is on a steel bar (6) machined 
 to fit two steel boxes (7) and is free to slide up and down. Considerable trouble 
 
 PI 
 
 zzr 
 
 f 
 
 f^J5 
 
 
 - General Arrangement of Apparatus for Testing Driving Wheel Springs. 
 (Coes and Howard.) 
 
 was encountered with the stylus on account of the excessive vibration and 
 jarring, and finally the type shown in Fig. 23 was designed, which gave entire 
 satisfaction, and with which the whole apparatus was equipped. 
 
 This is so constructed as to make the stylus spring always work in tension, 
 which is better than using the spring in compression. The spring is suffi- 
 ciently long to be sensitive and still not be thrown from the plate when the 
 engine strikes a curve, a trouble characteristic of all former instruments. Be- 
 sides the main stylus there are three others 
 of identical construction. First, the zero 
 stylus (8), which draws a straight line across 
 the roll and to which all deflections are 
 referred; second a stylus (9) which is on a 
 magnet that is operated by a Morse key in f'ig.23.— Main styiusused in Driving wheel 
 
 .a. u it.- j i. 1 /im -u- t. • Spring Tests. (Coes and Howard.) 
 
 the cab; third, a stylus (10) which is on a 
 
 magnet and is operated automatically by a clock in the cab. 
 
 The spanner bar (11) is shown in Fig. 21 and needs no description except 
 its method of fastening to the spring link hangers and its mode of operation. 
 It is fastened to the hangers by means of two blocks, which are slotted and fit 
 over the ends of the hangers, these blocks being held on to the hangers by four 
 
 ~*TO» 
 
52 STEEL RAILS 
 
 hardened steel set-screws. The spanner bar (11) is connected to the stylus 
 bar (6) by means of a short link (Z). Thus, whatever relative movement is 
 given the recording apparatus by the spring is transmitted as a vertical line on 
 the paper by means of the stylus bar (6) and the spanner bar (11). Hence, 
 since the paper is being driven horizontally by the motor we have a wavy line 
 giving a complete record of every movement made by the spring, and by means 
 of the records made by the key and the times recorded by the clock we can 
 account for most of the deflections due to frogs, switches, curves, crossings, 
 brake applications, and bridges. 
 
 The cab apparatus (Fig. 20) consists of a suitable box containing a portable 
 storage battery and six dry batteries. The storage battery gives 5 amperes 
 for 8 hours at a pressure of 6.6 bolts. This runs the motor. The six dry 
 batteries operate the clock and the key. On top of the box is a key (K), which 
 is connected by means of flexible lamp cord, fastened to the running board, to 
 the magnetic stylus (9). By means of the key the operator can record by code 
 any observation that may be necessary in working up the records. On the side 
 of the box is fastened a clock, which automatically records 15-second intervals 
 on the paper by a magnetic stylus (10). The motor is kept at the proper speed 
 by a rheostat fastened to the top of the box. 
 
 The first successful run was made on engine 1064, consolidated type 2-8-0, 
 with 36-inch springs, 17 leaves, 4 full-length leaves 4 inches by f inch. This run 
 was made on the Fitchburg Division of the Boston and Maine Railroad from 
 Boston to Ayer Junction on February 17, 1906. The spring tested was the 
 second (counting from the cylinder) on the left side. From this run was 
 obtained a maximum deflection of 0.34 inch. (See Plate XIX.) 
 
 A second run was made March 3, 1906, over the same route and on the 
 same engine to see if the same deflections were obtained. The curves obtained 
 by this latter run were practically identical with the test of February 17, 1906. 
 (See Figs. A and B, Plate XIX.) Figs. C D, E, and F present curves taken 
 at other points of the track. 
 
 The spring from engine 1064 was taken out and sent to the Engineering 
 Laboratory of the Institute ancl tested on the 100,000-pound Olsen Machine. 
 The results of this test are plotted and shown on Fig. 24. Two tests 
 were made, one with rollers under the knife-edges and one without. The set 
 had been measured on the engine and the ends of the leaves had also been 
 marked. The spring was then placed in the testing machine and the loads 
 applied, corresponding micrometer readings of the deflections being taken until 
 the spring had been loaded down to the set as measured on the engine. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 53 
 
 DEFLECTION AS OBTAINED 
 FROM RECORDING MACHINE 
 MINIMUM MAXIMUM 
 \+ _ Q.48^ Q.34" ^,1 
 
 O 
 DEFLECTION IN INCHES 
 
 Fig. 24. — Stress-strain Diagram-Locomotive Driving Wheel Springs. (Coes and Howard.) 
 
54 STEEL RAILS 
 
 This load was 14,200 pounds, or the static load on the engine, the dead 
 load for which the spring was designed being 14,144 pounds. Next the maxi- 
 mum deflection, as recorded by the spring apparatus, which was 0.34 inch, 
 was put in the spring, the load corresponding being 3500 pounds gradually 
 applied, making a total load on the spring of 17,700 pounds. 
 
 The excess load applied to the engine when running should be classed as 
 a suddenly applied load. If we consider the relation between the load slowly 
 applied to the spring in the testing machine and that suddenly applied when 
 the engine is in service, we see that in the first case the load gradually increases 
 from 14,200 pounds up to 17,700. The average load acting through 0.34 inch 
 is only 15,950 pounds and the total work done on the spring amounts to: 
 15,950 x 0.34 = 5423 inch-pounds. 
 
 In the case of the suddenly applied load under service conditions it will 
 be observed that owing to the load reaching nearly its full intensity before the 
 spring deflects, the load producing the deflection in this case would be obtained 
 by dividing 5423 inch-pounds by the deflection,* 
 
 or y-^-7 = 15,950 pounds. 
 
 This amounts to 12.3 per cent more than the static load of the engine 
 on the spring (14,200 pounds) and represents the dynamic augment of the 
 spring-borne weight of the locomotive. 
 
 It will be noticed that the dynamic augment as determined by these experi- 
 ments is considerable in excess of the figure arrived at by observing the wear of 
 the guides of the driving boxes. In the latter case the average deflection corre- 
 sponded to a load in the testing machine of 4000 pounds. Taking half of this or 
 2000 pounds as the dynamic load producing the same deflection of the spring, 
 we find a dynamic augment to the 25,000-pound static wheel load used of 
 08 per cent. This may very probably be accounted for by the fact that the 
 maximum deflection obtained is so infrequent as to cause no perceptible wear. 
 
 8. Effect of Flat Spots in the Wheels 
 We have now taken account of all the forces exerted by the wheel on the rail 
 
 except the impact caused by the existence of flat spots in the tread of the wheel. 
 The violence of the blow upon the track, delivered at every rotation of 
 
 the flat wheel, is a matter of common observation; but the amount of its force 
 
 and the damage done thereby are very hard to determine. 
 
 * See Applied Mechanics, Gaetano Lanza, 1895, page 246. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 55 
 
 A theoretical discussion of the force of the impact is not likely to lead to 
 any practical results because of the indefiniteness of the shape of the spot, due 
 to the rounding of the corners by wear, also to the lack of knowledge of the 
 effect of the springs and the resilience of the track. 
 
 The kinetic energy of the impact is represented by the expression \ Mv 2 , 
 where M represents the mass and v the velocity. In order to show a loss of 
 energy there must be a change of velocity, but any perceptible change in the 
 horizontal velocity of a moving car, due to the impact of the flat spot, is quite 
 inconceivable. There may, however, be a change in the vertical velocity of 
 the load as the flat spot comes over the rail. 
 
 Professor Hancock, of Purdue University, has made a very careful study 
 of the mathematical relations existing between the speed, impact, and length 
 of spot.* 
 
 Following Professor Hancock's analysis, let 
 A, in Fig. 25, be the center of a car wheel D 
 inches in diameter, revolving as shown by the 
 arrow, and CP be a flat spot L inches long just 
 beginning its contact with the rail. The whole i 
 wheel is turning about the point C, and will so 
 turn until P reaches R and the blow is struck on 
 the rail. At this latter instant A will have reached 
 A' and will be moving downward with a velocity 
 represented by the line be. If the velocity of A', 
 which is practically the same as that of the train, 
 is assumed as v feet per second, then 
 
 -Flat Spot in Wheel. 
 (Hancock.) 
 
 be = 
 
 CP 
 V CB = 
 
 If we regard the mass of the wheel and its load as concentrated at A and call 
 the total weight W pounds, the kinetic energy of the mass just before the rail 
 is struck will be: 
 
 This formula will give for the energy of impact of a flat spot 2.5 inches long 
 in a wheel 33 inches in diameter, carrying a load of 20,000 pounds when the 
 
 * Paper read before the Indiana Engineering Society, January, 1908. See also discussion by 
 L. S. Spilsbury, presented by H. H. Vaughan in the American Engineer and Railroad Journal, December, 
 1908. 
 
56 STEEL RAILS 
 
 train is traveling 60 miles per hour, 13,800 foot-pounds. At this speed it would 
 seem, however, that the results obtained by the formula would be open to 
 question. In the derivation of the formula it is assumed that the wheel turns 
 about C until P reaches R. This assumption only holds true for speeds from 
 zero up to about five miles per hour; * at speeds greater than five miles per hour 
 the point C will tend to leave the rail, and the whole wheel will revolve for an 
 instant entirely clear of the rail. 
 
 The above discussion neglects the effect of the springs, which will be to 
 increase the acceleration caused by gravity, and the resilience of the rail, which 
 will cause it to rise to meet the flat spot. 
 
 It is very questionable whether, on account of the very small time interval 
 required for the wheel to pass the length of the flat spot,f there is an appreciable 
 increase in the stress in the rail, except at the point of contact of the wheel with 
 the rail. 
 
 To increase the load on the rail a change in the vertical velocity of the 
 load must be made; but at high speeds, when the effect of the flat spot is most 
 detrimental, the time required to go the length of the flat spot is so small that 
 the acceleration of the wheel and its load, even when augmented by the action 
 of the springs, is so small as to be negligible. The real danger seems to lie 
 in the metal of the running surface of the head of the rail; the metal here is 
 under a high state of compression (see Figs. 146 and 147), which is momentarily 
 relieved by the passage of the flat spot and then applied suddenly. 
 
 When the flat spot is long enough so that the surface of the flat spot is 
 brought in contact with the rail, a sensible change in the vertical movement 
 of the load results and the load on the rail is increased. This is well shown by 
 the following example given by Mr. L. R. Clausen,! of the Chicago, Milwaukee 
 & St. Paul Railway: 
 
 "Some time in the year 1900 we had an engine with a flat spot on rear 
 right-hand driver 32 inches long and /? inch deep, which broke about 27 rails 
 during one week's time (85-pound rails, not to exceed one or two years old) ". 
 
 This flat spot was not apparent to the eye and was only detected by cen- 
 tering the wheel and then measuring around it with a gauge. 
 
 * E. E. Stetson, Railroad Age Gazette, December 4, 1908. 
 
 f The present allowable length of flat spots in car wheels is 2| in. This rule was adopted by the 
 Master Car Builders' Association in 1878. In 1909 the question of reducing the limit for freight wheels 
 to less than 2| inches was considered by committees of the Master Car Builders' Association and the 
 American Railway Engineering Association, but it was not then considered advisable to make any 
 change in the rule. 
 
 t Proceedings Am. Ry. Eng. & M. of W. Assn., 1909, Vol. 10, Part 2, p. 1158. Report on Flat 
 Spots on Car Wheels. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 57 
 
 In the Railway Age Gazette of March 16, 1910, was reported 200 85-pound 
 rails broken in 14 miles by a flat spot which had grown to a length of 6 inches, and 
 a maximum depth of f inch. In the extreme cold weather experienced in the months 
 of January and February, 1912, many tires failed by shelling out, and the following 
 examples, taken from the same authority, are representative of the conditions ex- 
 isting on lines in the Northern parts of the country during this period. 
 
 
 Broken Rails 
 
 January, 1912. 
 January 7, 1912. 
 January 14, 1912. 
 
 January 20, 1912. 
 January 24, 1912. 
 February, 1912. 
 
 February, 1912. 
 
 Minnesota. 
 Savanna, 111. 
 South Dakota. 
 
 New York State. 
 New York State. 
 Ohio. 
 
 Flat spot 4 ins. long on a rolled- 
 steel tire in passenger service. 
 
 Flat spot 5j ins. long on steel-tired 
 wheel in passenger service. 
 
 Two steel wheels with flat spots, 
 on different trucks of a dining 
 
 Flat wheel on a fast train. 
 
 Flat wheel on an observation car. 
 
 Shelled-out steel-tired wheel; at 
 end of run the flat spot was 9 ins. 
 long. 
 
 Flat spot on steel-tired wheel un- 
 der a baggage car. 
 
 9 80-lb. rails in 3 
 150 rails. 
 500 rails. 
 
 Nearly 100 rails. 
 
 15 rails. 
 
 960 rails in 200 miles. 
 
 50 rails in 70 miles. 
 
 It is generally known by those familiar with the manufacture and use of 
 chilled car wheels that only a very small percentage of them are evenly chilled. 
 This, apart from weakening the wheel, also produces a lack of roundness tend- 
 ing to cause pounding on the rail. The following information upon tests on 
 the roundness of tread of chilled car wheels has been furnished by Mr. S. K. 
 Dickerson, Assistant Superintendent of Motive Power, and Mr. H. E. Smith, 
 Engineer of Tests, of the Lake Shore and Michigan Southern Railway Company.* 
 
 To make these tests six pairs of wheels cast by different founders were 
 selected. An axle with a wheel pressed on each end was placed in a lathe and 
 the centers were firmly pressed. The wheels were then hand-turned. This 
 done, the tread was divided into eight sections, each the same distance from 
 the flat edge, and a specially constructed micrometer used to discover any 
 variations in the roundness. All the testing was done with great care and 
 precision. 
 
 The tests are illustrated in Fig. 26. The dotted line in each diagram is a 
 circle through that point on the tread having the smallest radius, and is assumed 
 as the datum line. In plotting the diagram the variations from this datum 
 line have been multiplied by five in order to emphasize the irregularity of the 
 
 * Proceedings Am. Soc. for Test. Materials, 1910, Vol. X, p. 307. Unevenly Chilled and Untrue 
 Car Wheels by Thomas D. West. 
 
58 
 
 STEEL RAILS 
 
 Fig. 26. —Irregularity in the Roundness of Present-day Chilled Car Wheels. (The Iron Age.) 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 59 
 
 Fig. 26. — Continued. 
 
60 
 
 STEEL RAILS 
 
 tread. It is to be understood, however, that the figures given are the actual 
 variations in the radii of the wheels from the datum circle. 
 
 Owing to the present imperfect state of our knowledge on this subject it 
 would seem desirable to determine experimentally the exact effect of the blow 
 delivered by a flat wheel on the rail. 
 
 Professor Benjamin* has designed an apparatus for such tests, which is shown 
 in Fig. 27. The apparatus shown in the figure will permit of the continuous 
 
 Fig. 27. — Apparatus for Measuring the Effect of a Flat Spot. (Benj; 
 
 operation of one wheel upon one section of rail indefinitely and permit at the 
 same time measurements of the effects of the blow. The truck is so supported 
 that one wheel turns freely upon an idle pulley, while the other wheel on the 
 same axle rests on a section of steel rail and in turning drives the latter by 
 friction. The section of rail is bent to a circle, lying in a horizontal plane, and 
 is firmly riveted and bolted to a supporting web, which is then fastened to a 
 central hub of cast iron or steel. This hub turns freely on a vertical mandrel 
 and is supported by a thrust bearing underneath. The rail and its attachments 
 thus turn in a horizontal plane under the rotating car wheel. The portion of 
 the rail immediately under the wheel is supported by friction rollers, which turn 
 
 * Paper presented at Meeting of Western Railway Club, November 17, 1908. See also dis- 
 cussion of Professor Benjamin's paper by H. H. Vaughan, American Engineer and Railroad Journal, 
 December, 1908, and a further article by Professor Benjamin in the Railway Age Gazette, June 28, 
 1912, p. 1613. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 61 
 
 2 !£ ' flat spot 
 
 RECORDING DEVICE 
 
 *?T 
 
 freely in a steel box or yoke. This latter forms a portion of the main casting 
 supporting the hub of the rail, and this casting is bolted to a wooden pier so 
 as to have a certain amount of elasticity. On the lower side of this casting and 
 directly beneath the point of contact between the wheel and the rail is a hard- 
 ened steel hammer, or ball, resting on a strip of soft metal. The soft metal is 
 supported on a heavy anvil of cast iron and is fed slowly beneath the hammer 
 by friction rollers. 
 
 The truck being loaded with the desired amount of pig iron or other material, 
 the wheels and their axles are rotated by means of a variable speed motor, and 
 the energy of a blow delivered by a flat spot on the wheel is measured by the 
 indentations of the strip of soft metal underneath the hammer. The amount 
 of energy due to any given indentation can be readily measured by producing a 
 similar indentation under a drop press. The curving of the rail in a horizontal 
 direction is not sufficient to interfere with the action of the wheel and the energy 
 of the blow is transmitted directly to the soft metal. 
 
 A subcommittee of the American Railway Engineering Association have 
 made an attempt to measure 
 the force of the blow caused 
 by flat wheels under working 
 conditions. 
 
 For this purpose an 
 80,000-pound capacity car 
 was equipped with regis- 
 tering devices to measure 
 the compression of the car 
 springs and a pair of wheels 
 with flat spots was placed 
 in one of the trucks. The 
 position of the flat wheels, 
 springs, and the recording 
 device is shown in Fig. 28. 
 The recording device con- 
 sisted of an apparatus for 
 measuring the maximum 
 deflection of the springs. 
 The springs were calibrated, 
 and it was found that a load of 32,500 pounds applied to a nest of four springs 
 produced a compression of one inch. 
 
 R|EAR truck 
 
 r|ear trujck ~> TRUE 
 
 ■Qfi|WARD TRluCK 
 
 2 %" FLAT SPOT 
 
 POSITIONS OF 
 
 RECORDING DEVICES 
 
 I 
 z 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 z 
 a. 
 
 0. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 IL p „ 
 
 
 
 
 
 
 
 
 
 
 z 
 o 
 
 
 
 
 
 
 
 
 
 <!) 
 
 
 
 
 
 
 
 
 
 
 a. 
 
 T JO 
 
 
 
 
 
 
 
 
 
 
 
 20 
 
 3 
 
 
 
 
 
 
 
 5 
 
 O 
 
 6C 
 
 - Diagram of Tests on Freight Car with Flat 
 Wheels. (Am. Ry. Eng. Assn.) 
 
62 STEEL RAILS 
 
 The car was loaded with splice bars to its full capacity, the load being 
 uniformly distributed. The train was then run for a short distance, brought 
 to a stop, and the maximum deflection of the springs noted. Several different 
 tests were made in this manner at different rates of speed, the results of which 
 are shown by the diagram in Fig. 28. 
 
 The diagram shows quite uniformly for all of the tests a greater deflection 
 of about a sixteenth of an inch for the trucks with flat wheels, corresponding 
 to an increase in pressure on these trucks of from 1000 to 2000 pounds. The 
 results would appear to indicate, then, that the flat wheels, either by increasing 
 the oscillation of the car or for other reasons, cause an increase of pressure on 
 the track. 
 
 On account of the small number of tests made and the fact that they were 
 confined to one end of a car the results should not be regarded as conclusive. 
 However, assuming that the effect of the flat spot in the wheel is to cause addi- 
 tional rocking of the vehicle, as the experiments would appear to indicate, it 
 will be noted from article 7 that this force is already taken account of in the 
 consideration of the excess pressure caused by the rocking of the locomotive 
 on its springs. 
 
 9. Impact Tests 
 
 Professor Goss * experimented to determine the effect of the counter- 
 balance pressure with the Purdue Test Locomotive. This engine is mounted 
 with its drivers resting upon wheels of approximately the same diameter with 
 the drivers, and when the drivers are turned by the engine the supporting wheels 
 roll in contact with them, and the engine as a whole remains stationary. The 
 engine was in complete horizontal balance and was counterbalanced heavier 
 than it would be in ordinary road service, the main wheel being about 0.4 per 
 cent and the rear wheel about 54.0 per cent more heavily balanced than is 
 the usual practice. 
 
 A common annealed iron wire 0.037 inch in diameter was used and run 
 under the drivers. Fig. 29 shows the effect of the drivers on the wire. 
 
 Wire I shows slight variation. 
 
 Wire II shows a jump of the wheel just after the counterbalance left the 
 highest point, the lifting being retarded, probably due to inertia of the mass 
 to be lifted. 
 
 * An Experimental Study of the Effect of the Counterbalance in Locomotive Drive-wheels 
 upon the Pressure between Wheel and Rail. — Goss. Trans. Am. Soc. of Mech. Engrs., Vol. XVI, 
 1895, p. 305. 
 
PRESSURE OF THE WHEEL ON THE RAIL 63 
 
 Wires III, IV, and V show the more marked lifting effects, due to increased 
 
 A light nick from a sharp chisel was made across the face of the wheel to 
 serve as a reference mark, which left a clean-cut projection upon the wire. It 
 was found at high speeds that the single nick across the face of the wheel leaves 
 two projections upon the wire, usually about one-eighth of an inch apart. The 
 contact between the wheel and the track would evidently appear, therefore, 
 not to be continuous, but a succession of exceedingly rapid impacts. Professor 
 Goss derived the following conclusions from his experiments: 
 
 Speed -58. 3 miles per Iiour-312 revolutions per 
 c„7„ Length One divislon=S' 
 
 Scale Thic hneas Onedivision = .Ol' 
 
 %Ly ffd.) mieel (7) Position Q% (?-4 
 
 Professor Goss. 
 
 (a) When a wheel is lifted through the action of its counterbalance, its 
 rise is comparatively slow and its descent rapid. The maximum lift occurs 
 after the counterbalance has passed its highest point. 
 
 (b) The rocking of the engine on its springs may assist or oppose the action 
 of the counterbalance in lifting the wheel. It therefore constitutes a serious 
 obstacle in the way of any study of the precise movement of the wheel. 
 
 (c) The contact of the moving wheel with the track is not continuous, even 
 for those portions of the revolution when the pressure is greatest, but a rapid 
 succession of impacts. 
 
64 
 
 STEEL RAILS 
 
 The question of impact has received a great deal of attention from bridge 
 engineers. Recent work in this direction consists of a large number of experi- 
 ments that were made on the Baltimore and Ohio Railroad by a subcommittee 
 of the American Railway Engineering Association to determine the amount of 
 impact in bridge members. The records of these tests were, unfortunately, 
 burned in the Baltimore fire before they were put in permanent form. The 
 experiments were continued by the Association under the direction of Professor 
 Turneaure. 
 
 The importance of this subject was widely realized, and most of the large 
 
 railroads in the country con- 
 tributed towards a fund which 
 enabled the work to be carried 
 on with a great degree of 
 thoroughness. 
 
 Professor Turneaure's ap- 
 paratus consisted of an auto- 
 graphic extensometer for 
 recording the deformation of 
 bridge members and multi- 
 plying that deformation by 
 some factor like eighty or 
 ninety and recording it on a 
 moving strip of paper. On 
 Fig. 30 are shown some of the 
 records taken by one of these 
 machines in 1906. It will be 
 seen from an examination of 
 the records that the effect of the different wheels can be readily traced in 
 the diagrams. 
 
 During the season of 1907 * further work was done by the committee. 
 The instruments used consisted of a deflectometer and eight extensometers. 
 The deflectometer is the instrument described in Transactions A. S. C. E., 
 Vol. XLI, June, 1899, p. 411. (Some Experiments on Bridges under moving 
 Train-loads.) The instrument is itself attached to the bridge, while the con- 
 nection with the ground below is made by means of a wire attached to a 
 heavy weight resting upon the ground. The deflection is multiplied by two and 
 
 * Proceedings Am. Ry. Eng. and M. of W. Assn., 1908. Report of Committee on Iron and 
 Steel Structures. Impact Tests. 
 
 NOTE:- ALL RECORDS READ FROM LEFT TO RI9HT 
 
 Fig. 30. — Deformation of Bridge Members under Passing 
 Trains. (Am. Ry. Eng. Assn.) 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 65 
 
 recorded on a moving strip of paper. The extensometers are clamped to various 
 members of a structure and record the extensions or compressions of the mem- 
 bers over a length of about four feet. The ratio of multiplication of the exten- 
 someters is about 50. 
 
 Test trains were made up of a selected type of locomotive, followed generally 
 by a sufficient number of loaded cars to cover the span. Longer trains were 
 not used, for the reason that it was desired to secure speeds as high as possible, 
 and also because many observations under the freight trains of the regular 
 traffic showed that at the speeds practically attainable the impact effect was 
 much less than with the shorter test train. 
 
 In carrying out the tests the train was headed in the more favorable direc- 
 tion for speed, and was moved back and forth over the structure at various 
 rates of speed. Such speeds were selected as fairly to cover the range from 
 about 20 miles per hour to the maximum attainable. A few movements were 
 made at from 10 to 20 miles per hour. Little difference was noted in the results 
 at various speeds below 15 miles per hour, and in general the results at 10 miles 
 per hour may be considered as practically equal to static stresses. 
 
 The speed of the train was determined by the use of stop watches and 
 signals by observers stationed at the ends of a 500-foot base line. The loco- 
 motive was generally working when crossing the span, but in some cases was 
 not. Differences in this respect caused no noticeable differences in results, 
 so far as the field observers were able to judge, although this point was con- 
 sidered mainly with respect to the higher speeds. Each test was given a serial 
 number, and all records obtained for that test were given the same number. 
 
 TABLE XII. — CALIBRATION TABLE, IMPACT TESTS ON BRIDGES. 
 
 Instrument 
 Numbed 
 
 Multiplication. 
 
 Value of 1 inch Ordinate, 
 £=30,000,000, 
 L=48 inches. 
 
 Speed of Paper, 
 Inches per Second. 
 
 Deflection 
 1 
 2 
 3 
 
 4 
 5 
 6 
 
 7 
 8 
 
 2 
 55.4 
 
 
 
 11,300 
 
 0.642 
 
 48.4 
 53.7 
 51.7 
 54.2 
 54.5 
 54.3 
 
 12,900 
 11,600 
 12,100 
 11,500 
 11,500 
 11,500 
 
 0.547 
 0.613 
 0.523 
 0.703 
 0.583 
 0.602 
 
 Table XII contains calibration data of the instruments and the approximate 
 speed of movement of the record paper. Table XIII contains detailed data of 
 one of the tests. This sheet represents the work of one day on one structure, 
 
STEEL RAILS 
 
 -oooiip 
 "oocve - ( 
 
 » 
 
 1 £ 
 
 •may 
 
 
 
 
 
 
 
 
 
 ■3[TO d 
 
 
 
 
 
 
 
 
 •daiy 
 
 
 2 
 
 s 
 
 
 
 
 •XB K 
 
 gsasasssssssssgssssggRgssKgggsgggSKSS 
 
 - 
 
 Lis 
 
 •maa 
 
 m : 
 
 ■3[TOa 
 
 •:■::::::::::: : £ : : : ::::::::::::::::: : 
 
 •Cuv 
 
 
 
 
 
 
 •XB W 
 
 SSSSgS8SSSSgSSSSggSS!2fSSgSSg!sggSSS8SSggS 
 
 . 
 
 jig 
 
 ■UI8H 
 
 
 
 
 •3[B9a 
 
 
 
 
 •dray 
 
 
 
 ssgsg ggs?g2ggjsg2 ; ; : 
 
 - 2 § 
 
 
 
 •XTJJ\[ 
 
 g :sgsgg§s 
 
 g83SSg£Sg£gSSggS!£SK£KgSS?gSSg 
 
 . 
 
 ■a -S 
 
 lis • 
 
 o£ra2 
 o 
 
 •may 
 
 
 
 
 
 
 
 ■3(153,! 
 
 
 
 
 
 
 
 •daiy 
 
 
 :-22SSSS2 
 
 22§gSS22 
 
 
 SS 
 
 •XE M 
 
 sssss 
 
 SSgSSSgggSSgggSSSSSggSS 
 
 :S33Sg : 
 
 - 
 
 III* 
 
 •may 
 
 
 
 
 
 
 
 
 
 •VM 
 
 
 
 
 
 
 
 
 
 •duiy 
 
 
 
 
 
 
 
 
 
 •xbm 
 
 •UI9H 
 
 sfesissaesssassssssKsaagasssgsssgsBSssss 
 
 - 
 
 ^1 S3 
 So, 
 
 •J[B 3d 
 
 ::::::: :o : ;°°£g° 
 
 
 -***-* :*""■"*■* ::::::: :■*<*■ : 
 
 •XB M 
 
 SgSSKggSSSSSSSgKggggSgSSSSSgSSSSfSSgSg 
 
 - 
 
 fells 
 
 mm 
 
 ■mag 
 
 
 
 
 
 
 
 
 
 •^ad 
 
 
 
 
 
 
 
 
 
 •duiy 
 
 
 IS : 
 
 
 
 gg gssgggs 
 
 2 
 
 :g 
 
 ■xbh 
 
 g 
 
 SgHKSSSgggSSSSSBSSSSggg : igSSSSSSBSgg 
 
 - 
 
 
 ■man 
 
 
 
 
 
 
 
 
 
 ■jpraa 
 
 
 
 
 
 
 
 
 
 •duiy 
 
 
 
 
 
 
 
 
 :222 
 
 ■xbh 
 
 
 Q 
 
 Jill 
 
 fiSDc 
 
 •UI8H 
 
 
 •duxy 
 
 »- : iessssssssssssssssssssss : - : .S»2»2gggS 
 
 •xbh | §22S22222SS222222SSS2SS^2222S23^222SS 
 
 
 oil 
 
 
 1 
 
 is 
 
 m^maJaJtzJaJaJaiZZZZZZZZZZt/jcGtG \mwmmmmmmmwwmmm 
 
 ooaost-DOMQf-om-fMMO-HOOOH : «as««an«««<o«N« 
 
 ikS 1 
 
 111 
 
 S" 
 
 """s 
 
 -««*■ 
 
 go^oo 
 
 35 1 
 
 -«— 8—— r«— 
 
 l°" 
 
 ""IS 
 
PRESSURE OF THE WHEEL ON THE RAIL 67 
 
 and with a particular locomotive and train. At the top of the sheet are given 
 the weights of test train, and a diagram of the bridge, showing the general 
 location of the instruments by number. The table then gives, first, the number 
 of the test, then the speed in miles per hour, then the position of the counter- 
 balance, as will be described more fully later. Then follow the data relating 
 to the deflection and measurements of the various extensometers from Nos. 1 
 to 8. The location of each instrument is given immediately below the number. 
 
 In the column headed "Maximum" is given generally the value of the 
 maximum ordinate of the diagram resulting from the test, expressed in hun- 
 dredths of inches. An exception to this is where there was obviously considerable 
 instrumental vibration, in which case an effort was made in scaling the diagrams 
 to eliminate this instrumental effect. This could be done in many cases quite 
 satisfactorily, but not in other cases, and all such records are open to more or 
 less doubt. 
 
 In the column marked "Amplitude" is given the amplitude of the vibra- 
 tions in the diagrams where such vibration is apparently due to the structure 
 and is not instrumental vibration. Vibrations may be due in general to three 
 causes: (a) vibrations of the structure as a whole, as shown most clearly in the 
 deflection diagrams and the chord or flange stresses; (6) vibrations of individual 
 members, especially eyebars, and (c) instrumental vibrations. 
 
 Other marked variations in the diagrams occur in such members as stringers, 
 floor beams, and hip verticals. It can only be stated at present that it is generally 
 possible to distinguish instrumental vibrations from others because of their 
 much greater rapidity. It may be said, however, that in most of the diagrams 
 obtained the instrumental vibrations are not serious. 
 
 In the column marked "Peak" are given the measured ordinates to the 
 highest points of the curves, including instrumental vibration. The excess 
 of this value over the "Maximum" shows the extent of the instrumental vibra- 
 tion as estimated. In many cases this is small, but in many cases also it is 
 large. In some cases it is so excessive that no attempt has been made to measure 
 the records. The deflectometer gave no trouble in this respect. 
 
 In the column headed "Remarks" are noted various remarks by the use of 
 letters: "I" signifies instrumental vibration. 
 
 Returning to column three, headed "Counterbalance": In the later tests 
 the position of the counterbalance was determined with reference to some panel 
 point of the bridge. This was done by inserting an index made of J-inch steel 
 into the rim of one of the drivers, exactly opposite the counterbalance. Then, 
 alongside the rail was placed a 2- by 4-inch strip, on which was placed a ridge 
 
68 STEEL RAILS 
 
 of clay or putty at such a height as to be indented by the index as it passed 
 along. The position of the indentation is noted in feet north or south of the point 
 of reference, which point of reference is shown on the truss diagram by the 
 letter "C." 
 
 During the seasons of 1908 and 1909 the experiments were continued, 
 and very complete data gathered of a number of bridges under different con- 
 ditions of loading.* 
 
 The experiments obtained in this series of tests indicate that with track 
 and rolling stock in good condition the main cause of impact is the unbalanced 
 condition of the drivers of the ordinary locomotives. The great importance of 
 unbalanced drivers is well brought out by a comparison of the results of com- 
 parative tests with the ordinary locomotive and the balanced compound and 
 electric locomotive. The impact caused by balanced compound and electric 
 locomotives was very small. 
 
 While it is interesting to study the effect of the impact in the different 
 members of a bridge in a consideration of the stresses in the rail, it is doubtful 
 whether the two cases are sufficiently alike to afford anything more than a 
 general comparison. 
 
 In considering the impact stresses in the rail the effect of the inertia of the 
 track must be borne in mind. 
 
 If we examine what takes place in the track under the impact of the wheel 
 it is seen that there is a force F acting between the wheel and the rail. When 
 impact occurs this force is increased by a force F' which produces a change in 
 the relative velocities of the wheel and the track, but on account of the nature 
 of the track the change in its velocity is almost impossible to determine. 
 
 The average value of this force F' is exactly equal to the change in momen- 
 tum produced by its action divided by the time required to produce this change 
 
 t-t' 
 
 where m = mass, 
 
 v = velocity, 
 t = time. 
 
 It will be observed that the interval of time t - t' during which the impact 
 acts is very small, and is not sufficient to allow the force F' to depress the 
 track at high speeds, with the result that the force F' is overcome mainly by 
 the resistance of the rail to compression. The heavier and harder the rail, there- 
 
 * Proceedings Am. Ry. Eng. and M. of W. Assn., 1911, Vol. 12, Part 3, p. 12. Report of Sub- 
 committee on Impact. 
 
PRESSURE OF THE WHEEL ON THE RAIL 69 
 
 fore, the greater is this force F' , other things being equal. Indeed it is quite 
 possible that it may at times exceed the crushing strength of the rail on its upper 
 surface. 
 
 A division of the forces acting on the wheel into those which produce local 
 strains in the running surface of the head and the forces which tend to set up bending 
 stresses in the rail is difficult to determine; probably the effect in producing bend- 
 ing stresses of all the forces except the static load on the wheel and the centrifugal 
 force described in Article 6 are considerably lessened owing to the inertia of the 
 track. 
 
 This especially applies to the pressure exerted by the springs of the loco- 
 motive when the engine is rocking. If the full effect of the compression of the 
 springs was realized the pressure might well be twice what has been taken. 
 
 This was illustrated in the impact tests on bridges when the excess pressure, 
 caused by the counterbalance, tended to produce well-defined strains, and at or 
 near the critical speeds would set up vibrations in the bridge itself, while the 
 impact from the rocking of the locomotive on its springs apparently caused a 
 much less serious effect. 
 
 10. The Dynamic Augment of the Wheel Load 
 While we have examined the various causes of the increase in wheel pressure 
 when the locomotive is running, it is still necessary to consider the effect of the 
 velocity of the wheel load. 
 
 Assuming the track perfectly smooth, the wheel without imperfections and 
 all of the rotating parts perfectly balanced, the effect of a load moving over the 
 track at a high rate of speed depends wholly upon the vertical curvature of 
 the track and the effect which this curvature has upon the path over which 
 the center of gravity of the load travels. 
 
 In the case of a bridge, if we assume the track originally straight and 
 absolutely rigid, the amount of impact or centrifugal force resulting from the 
 deflection of the structure can be approximately determined on theoretical 
 grounds. Such an analysis has been made by Dr. H. Zimmermann for the 
 case of a single rolling load, and a formula which is very closely approximate 
 to his exact formula is as follows: 
 
 F = P 1 , 
 
 16 vH 
 in which F = centrifugal force, P = weight of rolling load, v = velocity in feet 
 per second, d = deflection of structure, and I = span length. If, for example, 
 
70 
 
 STEEL RAILS 
 
 d = 2100 of span length and v = 90 feet per second (about 60 miles per hour),. 
 
 we have 
 
 F = P , 
 
 0.595 Z -3' 
 
 a formula which is practically exact for spans greater than 15 feet. For a 
 
 25-foot span this gives 8.7 per cent impact, and for a 50-foot span 3.7 per cent. 
 
 For a 100-foot span the value would be 1.7 per cent. 
 
 With the yielding supports under the rail the center of gravity of the load 
 tends to travel along a straight line (assuming the track to be perfectly uniform). 
 It seems therefore probable that the dynamic stress in the rail is not increased by 
 any appreciable amount by the velocity of the wheel load, and the maximum 
 pressure to be used in calculating the stress in the rail can be taken to be made 
 up of the static wheel load, the excess pressure due to the counterbalance and 
 angularity of the main rod, the pressure due to the wheel passing over irregu- 
 larities in the surface of the track, and the pressure caused by the rocking of 
 the engine on its springs. 
 
 The imposed pressure due to the angularity of the main rod and the excess 
 balance is given in Table XIV. 
 
 TABLE. XIV. — MAXIMUM DYNAMIC PRESSURE, IN POUNDS, DUE TO 
 
 ANGULARITY OF MAIN ROD AND EXCESS BALANCE, 
 
 FOR SPEEDS UP TO 60 MILES PER HOUR 
 
 Type. 
 
 Weight on Axle. 
 
 Excess Pressure Due to Angularity of 
 Main Rod an. 1 Evv-s < 'nun' re- 
 balance. (One Side Only.) 
 
 Main 
 Driver. 
 
 Front or Back 
 Drivers. 
 
 - Driver. 
 
 Front or Back 
 Drivers. 
 
 4-4-2 
 4-6-2 
 4-6-0 
 2-6-0 
 2-8-0 
 
 55,000 
 60,000 
 55,000 
 54,000 
 54,000 
 
 52,000 
 
 56,000 
 52,000 
 53,000 
 50,000 
 
 12,000 
 
 10,000 
 11,000 
 15,000 
 14,000 
 
 9,000 
 
 7,000 
 10,000 
 11,000 
 
 9,000" 
 
 This pressure is not a direct function of the wheel load and should be ex- 
 pressed in pounds for each class of locomotive. 
 
 The excess pressure for the freight locomotive, types 2-6-0 and 2-8-0, is 
 noticeably greater than is that of the passenger locomotive for the high speeds 
 given in Table XIV. Inasmuch as the engines in freight service are not called 
 upon for as high speed as in the passenger service, the excess pressure at 40 miles 
 per hour may be taken as the maximum for the engines of this class. The 
 greatest pressure occurs with the Mogul locomotive, type 2-6-0, and is 10,000 
 pounds for the main driver and 5000 pounds for the front and rear drivers. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 71 
 
 The extra pressure due to irregularities in the track is dependent on the 
 weight on the wheel and may be expressed in per cent of the wheel load, and is 
 
 o 00 a = 13 per cent of the weight on the driver. 
 
 The pressure due to the rocking of the engine on its springs is likewise a 
 function of the wheel load, and is 13 per cent of the load on the driver. 
 
 Dr, P. H. Dudley has observed that the drawbar pull and the tender on 
 stiff rails can become important factors in distributing the effect of the expended 
 tractive power to a longer portion of the track than that occupied by the driving 
 wheel base. 
 
 The distribution of the expended tractive effort through the drawbar 
 pull may be extended in 6-inch, 100-pound rails to all of the wheels under the 
 locomotive. In the 80-pound rail this reduces to about two-thirds of the length 
 of the total wheel base of the locomotive. 
 
 Two stremmatograph tests were made for the purpose of tracing the dis- 
 tribution of the stresses due to the expended tractive power of the locomotive 
 when drawing its train. The first test was made with the engine light, backing 
 over the stremmatograph, and then in a few moments coming forward. The 
 second test was made with the same engine drawing its train. It indicated 
 that the unit fiber atresses are increased under all of the wheels of the loco- 
 motive, and also the bending moments. 
 
 The effect of the expended tractive effort through the drawbar pull and 
 the stiff rails was distributed for the entire length of the wheel base of the loco- 
 motive, instead of being restricted to that of the driving wheels. The total 
 wheel effects under the light locomotive for one rail were 1,368,706 inch-pounds, 
 and 1,864,363 inch-pounds when drawing the train, an increase of nearly 500,000 
 inch-pounds, or 1,000,000 inch-pounds, practically, for both rails. 
 
 It is evident that' this increase in pressure on the rail is allowed for by the 
 dynamic augment to the spring-borne load of the locomotive, and we may now 
 prepare Table XV, giving the dynamic augment to be added to the static 
 wheel load. 
 
 TABLE XVa. — DYNAMIC AUGMENT TO BE ADDED TO THE STATIC 
 WHEEL LOAD WHEN THE LATTER IS 25,000 POUNDS OR OVER 
 
 Note. — The dynamic augment 
 as given in pounds in columns 1 or 
 2 should be added to the per cent 
 of the wheel load given in column 3, 
 to give the total dynamic augment 
 for the drivers. Column 3 only 
 should be used for the truck wheels . 
 
 Class. 
 
 Dynamic Augment. 
 
 Pounds. 
 
 »— ■ 
 
 Main 
 Drivers. 
 
 Front and Back 
 Drivers. 
 
 1 
 
 2 
 
 3 
 
 Passenger 
 
 Freight 
 
 12,000 
 10,000 
 
 9,000 
 5,000 
 
 26 
 26 
 
72 
 
 STEEL RAILS 
 
 TABLE XVb. — DYNAMIC AUGMENT TO BE ADDED TO 
 THE STATIC WHEEL LOAD WHEN THE LATTER 
 IS LESS THAN 25,000 POUNDS 
 
 Class. 
 
 Dynamic Augment. 
 
 Main Drivers. 
 
 Front and Back 
 Drivers. 
 
 Truck Wheels. 
 
 Passenger 
 
 Freight 
 
 er^cen 
 66 
 
 erren 
 
 45 
 
 Per cent 
 
 26 
 26 
 
 Prairie Type, 262-234. 
 
 Fig. 31. — Dynamic Wheel Loads of Typical Passenger Steam 
 Locomotives. 
 
 See Plate XX for static loads and Table XV for relation between 
 static and dynamic loading. 
 
 Note. — Total weight on drivers assumed to be equally divided 
 between the driving axles; if the main wheel is more heavily loaded the 
 dynamic pressure will be increased accordingly. 
 
 By referring to 
 Figs. 6 and 7 it is seen 
 that the excess pres- 
 sure, caused by the 
 angularity of the main 
 rod and counterbal- 
 ance, is less in amount 
 for these lighter en- 
 gines. Table XVb 
 shows the dynamic 
 augment of engines 
 having wheel loads less 
 than 25,000 pounds. 
 
 We may now pro- 
 ceed to construct typi- 
 cal load diagrams for 
 the different classes of 
 locomotives. Plates 
 XX and XXI give the 
 principal dimensions of 
 each type of engine 
 under discussion, from 
 which, with the aid of 
 Table XV, the load 
 diagrams given in Fig. 
 31 and Fig. 32 have 
 been prepared. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 73 
 
 % 
 
 e 
 e 
 
 -e 
 
 OQO/2 T 
 
 4*.- h 
 
 e: 
 
 vd 
 
 M 
 
 OOOQI 
 
 i 
 
 OOOOS | 
 
 
 e-i 
 e 
 
 §- 3 
 
 T 
 
 a „5 a c >>i2 .> * 
 
 -2 a § jg s * „ § 
 
 oooo* I J | I ^ g ~ g 
 
 ^ 1 S |i t? s J 
 
 I- I oooo* cc _g t) a 3 § - 
 
 J 
 
 ooog? £ 
 
 "®TT 
 
 -"J OC09£ i- 
 
 ® 
 
 > Ja ^ 1 .g 
 
 H ^ .2 g g I J 
 
 I 3 « I '? -i 8 
 
 O C3 S g g « S 
 
 
 03 H B m t, " 
 
 "fslT- °°° s " 3 ! "^ 
 
 &S 
 
 LL 
 
74 
 
 STEEL RAILS 
 
 11. Electric Locomotives 
 
 * In dealing with the electric locomotive the question of the excess balance 
 necessary to counteract the reciprocating parts can be entirely neglected, and these 
 machines, when properly constructed, would appear to have a more favorable 
 action on the track than is the case with a steam engine of the same capacity. 
 
 The low center of gravity possessed by the earlier locomotives of this type 
 imposed, under certain conditions, a very severe duty on the rail, f In order 
 to bring out the facts experimentally, the Pennsylvania Railroad Company, 
 who were about to design locomotives for their tunnel entrance into New York 
 City, constructed a special test track with apparatus for measuring side pres- 
 sures upon the rail; they built sample locomotives of different designs and 
 instituted a series of tests of electric and steam locomotives to determine their 
 relative riding qualities at speed. 
 
 It was found that all types of locomotives were practically steady at speeds 
 under 40 miles per hour, but that above this speed marked differences appeared; 
 that the steadiest riding machines were those with high center of gravity and 
 with long and unsymmetrical wheel base. In other words, that the nearer 
 steam-locomotive design is approached in wheel arrangement, distribution of 
 weight, height of center of gravity, and ratio of spring-borne to under-spring 
 weight, the less the side pressures registered on the rail head. 
 
 table xvi. - 
 
 - ELECTRIC and steam locomotives, comparison of weights 
 
 AND CENTERS OF GRAVITY. (Gibbs.) 
 
 
 Electric. 
 
 Steam. 
 
 Type 
 
 0-4-4-0 
 Pennsylvania. 
 Experimental. 
 
 195,140 
 
 42.5 
 
 50 
 
 28 
 
 4-4-4-4 
 Pennsylvania. 
 Experimental. 
 
 304,000 
 
 55 
 
 46.3 
 
 33.5 
 
 4-4-4-4 
 
 Pennsylvania. 
 N. Y. Tunnel. 
 
 332,000 
 
 63.75 
 
 16.7* 
 
 30.2* 
 
 2-4-4-2 
 N.Y..N.H.&H. 
 
 . Main Line. 
 
 202,000 
 37.4 
 
 2-4-4-2 
 Pennsylvania. 
 Main Line. 
 
 176,600 
 
 73 
 
 22.7 
 
 33 
 
 4-4-0 
 
 Service 
 
 Total weight of locomotive, 
 
 running order in pounds 
 
 Height, center gravity, complete 
 
 Main Line. 
 138,000 
 
 Per cent of weight of running gear 
 below springs to total weight 
 
 Heights of center of gravity of 
 running gear from rail, indies. 
 
 22.7 
 29 
 
 * Does not include motors as they are mounted in cab. 
 
 Table XVI presents a comparison of weights and centers of gravity of 
 modern electric locomotives and steam locomotives. Fig. 33 shows the Detroit 
 River Tunnel Company's locomotive. 
 
 * Seethe Railroad Age Gazette, Vol. XLVII, 1909, pp. 271, 319, 537, 881, and the Railway Age 
 Gazette, Vol. XLVII, 1910, p. 829, for descriptions of electric locomotives given in this article. 
 
 t Electric Traction by George Gibbs, report presented before the International Railway Con- 
 gress, July, 1910. See also a very complete article " The Electrification of Railways " by George 
 Westinghouse. Appendix No. 2. Data on Electric Locomotives of American Design, pp. 970-979. 
 Trans. Am. Soc. of Mech. Engrs., 1910. Vol. 32. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 75 
 
76 
 
 STEEL RAILS 
 
 Fig. 34 illustrates the type of the Pennsylvania Electric locomotives which 
 are used for handling the Pennsylvania Railroad trains into the New York 
 
 station. 
 
 This locomotive incorpo- 
 rates many novel features in 
 electric-locomotive design, and 
 is the result of several years' 
 cooperative development be- 
 tween the Pennsylvania Rail- 
 road Company and the 
 Westinghouse Electric and 
 Manufacturing Company. It 
 is distinctively a high-powered 
 machine, built for high speed 
 operation. 
 
 In wheel arrangement, 
 weight distribution, trucks and 
 general character of the run- 
 ning gear, it is the practical 
 equivalent of two American 
 type locomotives coupled per- 
 manently back to back. 
 
 The connecting rods are 
 all rotating links between rota- 
 ting elements, and are thus 
 perfectly counterbalanced for 
 all speeds. 
 
 The employment of this 
 transmission permits the 
 mounting of the motors upon 
 the frame, secures their spring 
 support, and, in common with 
 the rest of the locomotive, the 
 center of gravity at approxi- 
 mately the same height above 
 the rails, found desirable in 
 the best high-speed steam 
 experience. The same freedom of motion in the wheels and axles that is 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 77 
 
 characteristic of the present steam locomotive is also obviously secured. It 
 will be seen from Fig. 35 that the locomotive is an articulated machine and 
 that each half carries its own motor and has four driving wheels, 68 inches 
 
 
 
 
 hSoI 
 
 r JGI 
 
 
 H 
 
 
 1 
 
 DHf-f 
 Mm 1 v 
 
 • it''W^ 
 
 si 1 
 
 
 i 
 
 mm ! " 
 
 
 H ! 
 
 H^p^rafe 
 
 
 mm 
 
 
 V—-J6L- 
 
78 
 
 STEEL RAILS 
 
 in diameter, and one four-wheel swing bolster swivel truck with 36-inch wheels. 
 In these locomotives the variable pressure of the unbalanced piston of the steam 
 locomotive is replaced by the constant torque and constant rotating effort of 
 the drive wheels, and the pull upon the drawbar is thereby constant and 
 uniform. It might to the casual observer appear that by this arrangement of 
 driving a return has been made to steam locomotive practice as regards \ 
 counterbalancing difficulties, but it will, upon examination, be seen that nothing 
 of the kind is true. There are no questions of unbalanced reciprocating weights 
 involved, and all weights are revolving ones and directly counterbalanced. 
 
 In Table XVII is given the general characteristics of the electric locomo- 
 tives of this country. 
 
 In determining the dynamic augment of the wheel load in the case of the 
 electric locomotive, the effect of the counterbalance, which plays such an im- 
 portant part in the pressure of the driver of the steam locomotive, can be entirely 
 neglected. The other causes remain approximately the same, and by referring 
 to Table XV (column 3), it is seen that the dynamic augment amounts to 
 26 per cent of the wheel load. Fig. 36 gives the load diagrams of electric loco- 
 motives based upon this assumption. 
 
 B.&O.R.R. 18 95 
 
 D — Dynamic Load per Wheel, pounds. 
 /S = Static Load per Wheel, pounds. 
 D =1.26 5, 
 
 o tn o m 
 
 L6^r 
 
 PENNSYLVANIA R.R. 1909 
 
 9'- 4" i 7'- 2." >{ 9'- 9 " .1. 7 - 2? 
 
 & & CO 
 
 oo 
 oo 
 
 in O 
 
 OO 
 OO 
 
 mo 
 
 O if) 
 Fig. 
 
 OO 
 O O 
 OO 
 
 O O 
 OO 
 O O 
 
 OO 
 OO 
 
 o o 
 
 oo oo 
 
 O Q O O 
 
 o o mo 
 
 if) CQ 10 tO 
 
 oo 
 oo 
 
 UT.Q 
 vOK> 
 
 i. — Typical Load Diagrams for Electric Locomotives. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 79 
 
 
 
 ft 
 
 
 
 
 _ 
 
 
 
 s. . 
 
 
 
 
 
 
 
 «0 
 
 pJ^n 
 
 8 
 
 
 
 1 
 
 ||gJ«,«,S 
 
 |8||| 
 
 
 
 
 !*.E ~ ^ 
 
 j: = S : : : coco 
 
 ™ s 
 
 
 
 
 |s 
 
 assays 
 
 
 
 
 
 
 
 
 1 
 
 s 
 
 ".6 i 
 
 1 .1.1 8 J 
 
 Hill 
 
 
 
 
 
 
 
 
 1 
 
 I 
 
 || * § 
 
 |s s s s s : SS 
 
 s 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 K 6 
 
 
 
 
 
 
 
 «o : 
 
 
 
 
 
 
 1 
 
 iy oo § 
 
 »«;LJJ!| 
 
 1 
 
 |lj 
 
 9 
 
 
 
 
 
 
 
 
 
 H 
 
 £j : 
 
 S22g aoo § 
 
 
 
 
 1 
 
 1 
 
 13 
 
 .11 
 
 |: : : : : : S£ 
 
 lillj 
 
 S 
 
 
 
 B § 
 
 
 
 
 =8 
 
 
 <o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ° d 
 
 
 
 
 
 
 
 
 | 
 
 
 
 O 
 
 oo^-o : 00 || 
 
 liil.s 
 
 § 
 
 z 
 
 
 <HO 
 
 
 
 
 
 
 65 
 
 s 
 
 
 
 
 
 
 dO 
 
 
 
 1 
 
 
 ■g 
 
 
 ^IsSoc £ 
 
 
 | 
 
 111 
 
 
 1 
 
 1 
 
 
 |= = = :::S 
 
 ° 
 
 ~ 3 i 
 
 
 
 
 S§ : 
 
 
 
 
 
 2 
 
 
 <;o 
 
 
 
 
 
 
 
 
 
 
 
 
 •a 
 
 § 
 
 
 is 1 
 
 J. . . . . . ■ 
 
 
 s 
 
 
 | 
 
 1 
 1 
 
 y^oo £ 
 
 1 | 
 
 « 
 
 III 
 
 
 
 m 
 
 £ 
 
 •<cj : 
 
 1™^ 
 
 
 ™ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 j 
 
 
 
 
 
 
 1 
 
 I- 
 
 
 f-s 
 
 
 fj 
 
 1! 
 
 
 
 
 111 
 
 
 illill 
 
 ■SfcCue-if^O 
 
 
 
 
 lffi3QH222 
 
 ii- 
 
 
 
 O 
 
 
 
 
 Q 
 
 
 
 
 £ 
 
 
 
 GQ 
 
80 
 
 STEEL RAILS 
 
 li 
 
 
 ^6 
 
 
 
 
 
 |. 
 
 
 
 
 
 
 200,000 
 
 7,280 
 
 42.5 inches 
 
 35 
 
 
 ! 
 
 Is ° * 
 
 gg 
 
 |= = = = = = ." 
 
 1 
 
 1.1 
 
 s 
 
 "° 1 
 
 1 | 
 
 1 
 
 ooj 
 
 i« 
 
 ^ 
 
 1-3° 1 
 
 |= = = = = = § 
 
 
 
 
 
 <!0 -i 
 
 323SSS3 
 
 
 
 
 ii 
 
 1L 
 
 1^ 
 
 ^6 
 
 Ifcs- 
 
 ■11 
 
 d 
 
 < 
 
 1= : = 
 
 r : : 
 
 °°l 
 
 1 
 
 111 
 
 a 
 
 
 
 .6 
 
 1 
 
 1= = = = = = 
 
 
 
 
 i 
 
 1 
 
 111?- 
 
 1 
 d 
 
 M! ° M 2 0<0 2| 
 
 I 
 
 p 
 
 2 
 
 
 
 m& 
 
 3 
 
 r3r " 
 
 
 
 
 =8 
 
 
 
 1 
 
 jj jal js 
 
 
 
 
 I 
 
 £ 
 
 II 
 
 d 
 
 1= = = = = = 8 
 
 1 
 
 §p 
 
 a 
 
 
 1 
 
 
 J= = = = = = 
 
 Ills 
 
 
 w 
 
 « 
 
 •Is "? 
 
 J= = = = = = 3S? 
 
 <N 
 
 
 s 
 
 
 m& •* 
 
 ™ *"" N M 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 6 
 
 *6 o 
 
 2 
 
 d 
 
 1 | 
 
 1= = = = = = s 
 
 l 
 
 - - 
 
 § 
 
 2 
 
 
 ms= 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 :-8 8 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .1 
 
 1 
 
 
 i 
 
 h 
 
 Is. 
 
 1 
 
 | 
 I 
 
 | 
 1 
 
 | 
 
 1 
 
 1J 
 
 1 
 
 
 1 
 
 | 
 
 f * 
 i. 
 
 1 
 
 ■ ■- 
 
 1 ! 
 ; : 
 
 II 
 - 
 
 55 
 
 |1 
 
 .5 = 
 
 Mi 
 
 ill 
 •s<£t£ 
 
 ill 
 ill 
 
 1 i 
 
 •f §. 
 
 11 
 
 „|| 
 
 
 
 a 
 
 
 
 
 
 
 Q 
 
 
 
 
 
 
 is 
 
 
 CO 
 
PRESSURE OF THE WHEEL ON THE RAIL 81 
 
 12. Cars 
 
 In dealing with the pressure of the wheels of cars it will be noted that the 
 dynamic load is much less than is the case with the heavy steam locomotive. 
 Here the only dynamic effect caused by the rapidly moving wheel is that of 
 the rocking of the car on its springs, the irregularities of the track and lack of ' 
 roundness of the wheel. 
 
 Fig. 28 gives a convenient means for estimating the excess pressure caused 
 by the rocking of the car. It will be noted from the figure that the maximum 
 dynamic pressure corresponds to a depression of the nest of springs of about jq 
 inch at a speed of 50 miles per hour, or, as the springs are capable of resisting a 
 slowly applied load of 32,500 pounds for a depression of 1 inch, the dynamic 
 load at 50 miles per hour for the two wheels supported by the nest of springs 
 is probably about 5000 pounds, or 2500 pounds for each wheel greater than the 
 static load. 
 
 Fig. 37. — Box Car. Capacity 100,000 lbs. Weight 48,000 lbs. Length 40 ft. 
 
 An allowance of 10,000 pounds per wheel would appear to be ample to 
 cover higher speeds obtained and heavier cars than that used in the test to which 
 this figure refers, as well as any increase in pressure caused by irregularities in 
 equipment. With lighter loads the dynamic augment may be decreased, and for 
 wheel loads less than 15,000 pounds it has been taken as 0.7 of the static load. 
 
 Figs. 37 to 41 * illustrate types of freight cars and Figs. 42 to 46 * passen- 
 ger cars. It will be seen that the freight cars use a four-wheel truck, each 
 wheel of which sustains about 20,000 pounds for 100,000-pound capacity cars 
 when loaded to their full capacity. The minimum spacing of wheels used under 
 100,000-pound capacity cars is 5 feet 6 inches. Cars in passenger service as 
 seen from the figures use four-wheel trucks and also six-wheel trucks. The wheel 
 
 * Figs. 37 to 46 have been taken from the Car Builders Dictionary. 
 
82 
 
 STEEL RAILS 
 
 spacing of the four-wheel trucks is 8 feet and for those having six wheels, 
 5 feet 3 inches. 
 
 The increase in weight in American passenger equipment in recent years 
 has been very great. Steel coaches with wood inside finish used by one of 
 the Chicago lines weigh 139,000 pounds. New buffet library cars for another 
 
 Fig. 38. — Flat Car, Pressed Steel Underframe. Capacity 100,000 lbs. Weight 38,000 lbs. 
 
 western line weigh 153,000 pounds. In the design for new sleepers for the 
 Pennsylvania it is difficult to limit the weight to 160,000 pounds, and 75 
 to 80 tons may be taken as the weight of modern sleeping cars. The 70-foot 
 steel coaches, which are being built in large numbers, are so heavy that it is 
 necessary to use six-wheel trucks under them, and these alone weigh over 
 40,000 pounds for two trucks.* 
 
 Fig. 39. — Gondola Car. Wooden Body. Pressed Steel Underframe. Capacity 100,000 lbs. 
 Weight 44,500 lbs. 
 
 The proposed use of 80-ton freight cars with four-wheel trucks for 
 special service f suggests some comparison of these loads with those imposed 
 by the engine drivers. The proposed 80-ton car will weigh about 50,000 
 pounds and the total weight of the loaded car will be 210,000 pounds, which, 
 when carried by eight wheels, produces a static load on the rail of 26,250 pounds. 
 
 Weights on drivers have always greatly exceeded the load on smaller car 
 wheels, and the reason for this seems to have been the greater strength of 
 the larger wheels. Owing to the work locomotive tires receive in rolling they 
 
 * Railway Age Gazette, January 28, 1910. t Ibid., December 22, 1911. 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 83 
 
 
 
 1 Eil 
 
 iJJ.1 1 ! 1 
 
 ^ ^m -C.LS.4E. ■ 
 
 Ir4i,i 
 
 ^^^"Si 
 
 
 el 
 
 Fig. 40. — Coke Car, All Steel. Capacity 100,000 lbs. Weight 47,600 lbs. 
 
 
 Fig. 41. — Stock Car, Steel Underframe. Capacity 100,000 lbs. Weight 47,400 lbs. 
 
 
 m* 
 
 
 
 m ■ \_ j~ 
 
 liii k 
 
 111! i'l ii ti iiintiiH gil 1 
 
 . 
 
 H 1 
 
 
 fj jjMfn ■ i 1.33— — ^^BttT~~T^fo;».^//* 3 
 
 
 
 Fig. 42. — Vestibuled Coach, All Steel. 
 
84 
 
 STEEL RAILS 
 
 Fig. 46. — Baggage Car, All Steel. 
 
PRESSURE OF THE WHEEL ON THE RAIL 85 
 
 would appear better able to carry heavy loads than the smaller car wheels. 
 While the loads on car wheels in themselves are not necessarily detrimental 
 to the rail, it has seemed desirable to provide for the possible affect of a de- 
 fective wheel by taking a relatively high dynamic augment of the wheel load. 
 
 L 5- 6"j Freight Cars. 
 
 (T) (|) Weight of Car Empty 50,000 
 
 ^-^ — - W - Weight of Load 100,000 
 
 OO OO 10%Overload 10,000 
 
 o o o o 
 
 OO OO Total 160,000 
 
 O O O O 
 
 fl Q y, Passenger Cars. 
 
 4-wheel trucks. 
 
 I, 8 L Q"j Lbs. 
 
 JtT - ~?\ Weight of Car Empty 120,000 
 
 (!) (|) Weight of Load 20,000 
 
 Q Total 140,000 
 
 O O O O 
 
 O O O O Passenger Cars. 
 
 <0 <0 CO' CO a . . , n 
 
 (vj — oj — 6-wheel trucks, 
 
 n in Q lO Lb3 - 
 
 Weight of Car Empty 160,000 
 
 l 5-3"^ S'-3"a Weight of Load 20,000 
 
 Total 180,000 
 
 o O O O O o 
 
 ^^oonn ^ = Dynamic load per wheel, pounds 
 
 to m in m m in *S = Static load per wheel, pounds 
 
 cm - oj - cm - D = S + 10,000 
 
 a a) a (/) a in 
 
 Fig. 47. — Typical Load Diagrams for Cars. 
 
 Fig. 47 shows typical load diagrams for passenger equipment and freight cars. 
 Fig. 48 shows the dynamic wheel diagrams of motor cars used on steam 
 roads, and Figs. 49 and 50 illustrate these cars. 
 
 |< 7 "°"> | < 3 - 6 ^ 1_P^_J 
 
 <b S5 TD *<b * 2 "(D ^ 
 
 oooo OO OO 
 
 on on oo OO 
 
 8S |° SE 5? 
 
 oi- 
 
 ow 
 
 7-0 
 
 on a n a w 
 
 ^ 7-6\L &A-*" _. 
 
 <i><D . ^Ct)<bt 
 
 on § « C = D y namic load P er wheel > pounds %% °0 
 
 ^j oj — oi s = Static load per wheel, pounds 55 5 5 55 
 
 - - -- D = S + 10,000 where S is more than 15,000 ^ ^ „ 
 
 qu) QW £> = 1.7 £ where S is less than 15,000 aV) ow 
 
 Fig. 48. — Typical Dynamic Load Diagrams for Motor Cars. 
 
86 
 
 STEEL RAILS 
 
PRESSURE OF THE WHEEL ON THE RAIL 
 
 87 
 
 In Figs. 51 and 52 are given examples of the cars in use on electric rail- 
 ways, both for city service and on the longer runs of the interurban lines. 
 
 " Single-end " Smoking and Express, Mail and Baggage Car. 
 Seat ing capacity, 38. Weight of car body, about 34,000 lbs. 
 
 Wheel base of trucks, 6'6". Weight of trucks, 21,460 lbs. 
 
 Weight complete, about 38j tons. 
 
 62-ft. Buffet Observation Parlor Car. 
 Scat ing capacity, 35. Weight of car body, about 44,000 11 
 
 Wheel base of trucks, 6'6". Weight of trucks, 21,460 lbs. 
 
 Weight complete, about 44 tons. 
 
 51-ft. " Double-end " Special Parlor Car with Smoking Room. 
 Seating capacity, chairs 26. Weight of car body, about 29,000 lbs. 
 
 Seating capacity, seats 52. Weight of trucks, 20,322 lbs. 
 
 Wheel base of trucks, 6'6". Weight complete, about 39 tons. 
 
 Fig. 51. — Electric Railway Cars (Niles Car and Mfg. Co.). 
 
STEEL RAILS 
 
 51-ft. "Single-end," Two-compartment, Fast Interurban Car. 
 Seating capacity, 52. Weight of car body, about 26,500 11 
 
 Wheel base of trucks, 6'6". Weight of trucks, 18,600 lbs. 
 
 Weight complete, about 32 tons. 
 
 
 
 
 
 
 — __,. ___ ^ 
 
 
 
 ,ao - HI' In 
 
 
 
 
 ~m 
 
 i&. 
 
 
 
 
 
 
 
 
 
 48-ft. Center-vestibule, Arch-roof, Steel Prepayment Car. 
 Seating capacity, 54. Weight of car body, 22,000 lbs. 
 
 Wheel base of trucks, 6'4". Weight of trucks, 16,132 lbs. 
 
 Weight complete, about 26 tons. 
 
 42-ft. Double-truck, "Single-end," Pay-As-You-Enter Car. 
 Seating capacity, about 45. Weight of car body, about 16,000 lb 
 
 Wheel base of trucks, 5'0". Weight of trucks, about 12,000 lbs. 
 
 Fig. 52. — Electric Railway Cars (Niles Car and Mfg. Co.). 
 
PRESSURE OF THE WHEEL ON THE RAIL 89 
 
 Fig. 53 presents the typical dynamic load diagrams of this class of equipment. 
 The dynamic augment has been taken the same for these cars and the motor 
 cars as was used for the freight and passenger cars on steam roads. 
 
 Weight of car and equipment 43,000 lbs. 
 
 Weight of passenger load 15,000 lbs. 
 
 (100 (a) 150 lbs.) 
 
 WHEEL 33 DIA. 
 
 'ML 
 
 Equipped with two motors weighing 34,000 lbs. 
 each, both motors on axle of the rear truck. 
 
 City Cars. 
 
 j (j) WHEELS 36-DIA. (j) | (J) 
 
 Interurban Passenger Cars. 
 
 Weight of car body 35,000 lbs. 
 
 Weight of trucks 24,000 lbs. 
 
 Weight of motor equipment 19,000 lbs. 
 
 Weight of passenger load 11,250 lbs. 
 
 (75 <a) 150 lbs.) 
 
 Total 89,250 lbs. 
 
 Equipped with four motors hung on four axles, so that the load is evenly distributed. 
 
 CD i CD wheels 56 " dia - CD i CD 
 
 oo 
 oo 
 oo 
 
 Interurban Freight Cars. 
 
 Weight of car and equipment 62,320 lbs. 
 
 Weight of load (rated capacity) 50,000 lbs. 
 
 plus 10 per cent overload 5,000 lbs. 
 
 Total 117,320 lbs. 
 
 Equipped with four motors, one motor to each axle. 
 
 — Typical Dynamic Load Diagrams for Electric Railway C 
 D = Dynamic load per wheel, pounds 
 S = Static load per wheel, pounds 
 D = S + 10,000 where S is more than 15,000 
 D = 1.7 S where S is less than 15,000 
 
CHAPTER III 
 SUPPORTS OF THE RAIL 
 
 13. The Tie 
 
 The most common material used for the tie is wood. Some have sug- 
 gested (and this suggestion is made with increasing frequency) that ties should 
 
 be made out of materials other 
 than wood. Granite ties were 
 among the earliest substitutes 
 offered ; they were used for some 
 time in Dublin, Ireland, and 
 on the old Boston and Lowell 
 Railroad in Massachusetts. For 
 some fifty years various forms 
 of metal ties have been sug- 
 gested, and a large number of 
 steel ties have been tried in 
 various countries. In recent 
 years concrete ties have been 
 made. 
 
 The following examples 
 present in a general way what 
 has been attempted as a substi- 
 tute for the wooden tie. The list is far from complete, and must necessarily 
 remain so, as new forms of metal and concrete ties are constantly being developed. 
 The subject has been very fully reported upon by the committee on ties of the 
 American Railway Engineering Association.* 
 
 Steel ties have been used quite extensively on the Union Railroad and on the 
 Bessemer and Lake Erie Railroad. The total number of steel ties on these two 
 roads is over one million or enough to lay 300 miles of track. There are a large 
 number of steel ties of the Carnegie type (Fig. 54) in use throughout the country. 
 
 * See Report of Committee on Ties, Proceedings Am. Ry. Eng. Assn., 1909, 1910, 1911, and 
 1912, pp. 343-370. 
 
SITPOKTS OF THE RAIL 
 
 91 
 
 Fig. 55 shows the insulated tie in use on the Bessemer and Lake Erie Rail- 
 road. The insulated tie is made by placing a piece of fiber on the steel tie and 
 
 Fig. 55. — Carnegie Steel Ties on the Bessemer and Lake Erie Railroad. 
 (Am. Ry. Eng. Assn.) 
 
 — Effect of Three Derailments on Steel Ties. (Am. Ry. Eng. Assn.) 
 
92 
 
 STEEL RAILS 
 
 then firmly riveting a steel plate over the top of the fiber for the rail to rest on. 
 This is intended to stop all wear on the fiber. 
 
 Fig. 56 shows the effect of three derailments on the steel ties, which, in 
 this case, merely bent down the upper flange of the ties, and in no way injured 
 their usefulness as a tie. 
 
 
 i&at* > 
 
 pp|f>; 
 
 
 r * *** 
 
 
 
 Fig. 57. — Steel Tie after Four Years Service. (Am. Ry. Eng. Assn.) 
 
 Fig. 57 shows a steel tie taken from the track that had been in service four 
 years. Very little rust was found on the web of the tie and the bottom flange 
 of this tie showed very little corrosion. There were no signs whatever of the 
 tie failing in any respect. The cutting of the slots or holes in the web of the tie, 
 as shown in the figure, has been abandoned, as it was found that with slag or 
 stone ballast the holes with the web turned out were not necessary in order to 
 keep the track from sliding sideways. 
 
SUPPORTS OF THE RAIL 
 
 93 
 
 The tie was smooth on the upper face where the base of the rail rests and 
 showed very little, if any, wear. Providing the wear in years to come is no 
 greater in proportion than it has been during the past four years, the tie would 
 be good for 25 or 30 years. 
 
 Fig. 59. — Hill Fastening on Carnegie Steel Tie. (Am. Ry. Eng. Assn.) 
 
94 
 
 STEEL RAILS 
 
 Fig. 58 shows the Carnegie steel tie with the wedge fastener. 
 
 Carnegie Steel Tie with Hill Fastening. — (Fig. 59.) Approximately 
 100,000 of these ties have been installed in yard tracks at the Duquesne Plant 
 of the Carnegie Steel Company. 
 
 Hansen Steel Tie. (Am. Ry. Eng. Assn.) 
 
 Hansen Tie. — (Fig. 60.) Five hundred of these ties were placed in the 
 track, July, 1905, near Emsworth, Pa., on the Pennsylvania Lines West of Pitts- 
 burg. A great deal of trouble was experienced with the insulation, also from 
 the ties sliding transversely and longitudinally through the stone ballast, and 
 the ties were in consequence removed from the main track in November, 1905, 
 and placed in a passing siding. 
 
 Universal Metallic Tie. — (Fig. 61.) The figure shows these ties in the 
 Pennsylvania Lines tracks near Emsworth, Pa. The design is the trough type, 
 being a 6- by 8-inch by 8-foot steel channel. Holes are cut in the web of the 
 channel on each side of the rail, and this metal is bent up vertically on each 
 
SUPPORTS OF THE RAIL 
 
 95 
 
 side of a wooden block which fits in the channel under the rail. Clamps, fitting 
 over the base of the rail and extending down vertically outside these bent-up 
 portions of the channel, 
 bind the block, rail and 
 tie together. The clamp 
 on the gauge side of the 
 rail extends through the 
 hole in the 'base of the 
 channels about 4 inches 
 into the ballast, giving 
 an additional bond with 
 the roadbed. A bolt, 
 with a tapering head at 
 one end and with a taper- 
 ing washer at the other 
 end, holds the connection 
 tight. An insulating 
 fiber is inserted between 
 the rail and the clamp. 
 The weight of this tie is 
 175 pounds. 
 
 Fig. 61. — Universal Metallic Tie on Pennsylvania Lines. 
 (Railway Age Gazette.) 
 
 Fig. 62. — Snyder Steel Tie. (Am. Ry. Eng. Assn.) 
 
96 
 
 STEEL RAILS 
 
 Snyder Steel Tie. — (Fig. 62.) The illustration shows these ties in the 
 Conemaugh yards of the Pennsylvania Railroad. There is also about one 
 mile of these ties in use at Derry, Pa., on the same road, none of them being in 
 the main tracks. The standard type of the Snyder tie consists of a steel shell 
 T 3 e inch thick, 8 feet long, 7 inches wide, 7 inches deep, and with the bottom 
 open. The interior of the shell is filled with a mixture of asphalt and crushed 
 stone. In 20 of the ties the mastic had disintegrated and fallen out of the ends of 
 the ties after four years service. With this exception the Snyder tie has given very 
 satisfactory service in the tracks of the Pennsylvania Railroad at Conemaugh and 
 Derry. 
 
 Fig. 63. — Buhrer Combined Steel and Wood Tie on L. S. & M. S. Ry. 
 (Am. Ry. Eng. Assn.) 
 
 Buhrer Steel and Wood Tie. — (Fig. 63.) The figure shows the fourth or 
 freight track of the Lake Shore and Michigan Southern Railroad, east of Toledo, 
 tied with the Buhrer combined steel and wood tie. Early in 1907 the Carnegie 
 steel ties on the Lake Shore and Michigan Southern Railroad were removed 
 from the high-speed track. To care for the insulation the top flange of the tie 
 was cut off and two wooden blocks bolted to the web of the tie for spiking strips 
 and for the rail to rest on. These strips also rest on the bottom flange of the 
 steel tie. 
 
 Mexican Railway Tie. — (Fig. 64.) Practically the whole of the Mexican 
 Railway system of 361 miles is laid with these ties. These ties weigh about 125 
 
SUPPORTS OF THE RAIL 
 
 97 
 
 pounds, and cost $2.25. The ties are apparently giving excellent service. The 
 axle load on this road, however, is not heavy on the light grades, and on the 
 mountain grades, where axle loads as high as 50,000 pounds are employed, the 
 3 slow. 
 
 4 - -O HA LF LENGTH O F TI E. _ 
 
 PLAN 
 
 Fig. 64. — Mexican Railway Steel Tie. (Am. Ry. Eng. Assn.) 
 
 Buhrer Concrete Tie. — (Fig. 65.) About 600 of these ties were used on 
 the Pennsylvania Lines west of Pittsburg during 1903 and 1904 in stone ballast. 
 Nearly 500 were subjected to heavy and high-speed traffic and the balance to 
 medium traffic. The ties failed under traffic, the concrete breaking and crumb- 
 ling off from the reenforcement. The ties were removed from time to time and 
 by December, 1906, all had been removed on account of breaking. 
 
STEEL RAILS 
 
 ■a 
 
 - 
 
 D 
 
 \ 
 
 1 
 
 
SUPPORTS OF THE RAIL 
 
 99 
 
 Fig. 66 shows the bottom or bearing surface of this tie, which illustrates 
 how the concrete is left out at the center to provide against side motion. 
 
 
 
 ■ s.. 
 
 £ *^ m '* 
 
 <? 
 
 'n^^^ 
 
 ***. ^^^ 
 
 Fig. 66. — Bottom Surface of Buhrer Concrete Tie. (Am. Ry. Eng. Assn.) 
 
 Fig. 67. — Section of Track on Chicago and Alton R. R. showing Kimball Tie. 
 
 Kimball Concrete Tie. — (Fig. 67.) This figure shows the Kimball tie in 
 the track of the Chicago and Alton Railroad, near Lockport, 111., during October, 
 
100 
 
 STEEL RAILS 
 
 - Kimball Tie put in Track on N. Y. C. & St. L. R. R., July, 1904. 
 Photograph taken January, 1909. 
 
 Fig. 69. — Kimball Tie Showing Spiking Plugs. (Am. Ry. Eng. Assn.) 
 
SUPPORTS OF THE RAIL 101 
 
 1905. The track is kept in good condition, but several of the ties have developed 
 cracks which may be due to improper application. Fig. 68 presents further ex- 
 amples of the Kimball tie, and Fig. 69 shows a tie in good condition, taken 
 from the track. In this tie the spikes entered the spiking plug and the con- 
 crete was not damaged, as was found to be the case in most of the ties which 
 
 Fig. 70. — Percival Concrete Tie. (Am. Ry. Eng. Assn.) 
 
 developed cracks. Ties were not rusted to any extent in the center of the track 
 between concrete ends. In 1912 there were about sixty Kimball ties in successful 
 use in the track of the Chicago and Alton Railroad, the ties having been installed 
 in 1905. 
 
 Percival Concrete Tie. — (Fig. 70.) The figure shows ties which were 
 used on the Pittsburg and Lake Erie Railroad for about two years. They were 
 
102 
 
 STEEL RAILS 
 
 removed from the track in 1908 for the reason that the ties failed. The figure 
 illustrates very plainly how and where these ties failed. 
 
 The Sarada and Adriatic Railway ties, given in Figs. 71 and 72, illustrate 
 concrete ties used on the continent, and Fig. 127 shows the combined wood and 
 metal ties used in France. 
 
 Fig. 71. — Sarada Tie. (Concrete Review.) 
 
 Sarada Tie. — (Fig. 71.) 3.9 inches in center and 5.9 inches under rails 
 by 9| inches by 8 feet long. Reinforcement, 4 sheets of expanded metal, set 
 vertically and connected transversely by iron wires. The rail fastening bolts 
 enter from below and are held in tubular castings embedded in the ties. Weight 
 about 310 pounds. 
 
 V 
 
 • 
 • 
 
 \l 
 
 \A 
 
 • 
 
 ^L. 
 
 
 
 K 
 
 A 
 
 N 
 
 A 
 
 
 1/!! ;;\l 
 
 KTHiN | 
 
 m 
 
 Fig. 72. — Adriatic Railway Tie. (Concrete Review.) 
 
 Adriatic Railway Tie. — (Fig. 72.) Reinforcement, 29 rods having a total 
 area of 3 square inches. The rail is fastened by bolts passing through the tie 
 and inserted from below. The beveled rail seat is in accordance with European 
 practice. Weight about 286 pounds. 
 
 Riegler Concrete Tie. — (Fig. 73.) Some of these ties have been in ser- 
 vice on the high-speed tracks of the Pennsylvania Lines for several years 
 without showing signs of deterioration. The ties have a large bearing surface 
 and fifteen are used for a 33-foot rail, instead of eighteen standard wooden ties. 
 
SUPPORTS OF THE RAIL 
 
 103 
 
 The ties have rounded sides, which assist in distributing the downward thrust 
 over a short distance on each side of the tie, and the reaction assists in holding 
 the tie from slewing, all the ties remaining as first placed at right angles to 
 
 Fig. 74. — Riegler Concrete Tie, Appearance in the Track. (Railroad Age Gazette.) 
 
 the track. The weight of the complete tie is about 850 pounds. It is proposed 
 to cast a ring in the end to which a short rope can be attached for hauling into 
 the track. 
 
 Table XVIII presents a summary of the service tests on concrete ties. 
 
STEEL RAILS 
 
 ! 
 I 
 
 2 cracked in 2 weeks but held together, 10 cracked in 
 6 months, all out in 1J- years, bolts loosened and 
 
 iliMniciruiefl concrete dry mixture). 
 
 3 broken by a derailment, rest in good condition in 
 1904. 
 
 Some cracks within first year. 
 
 1 
 
 I 
 
 PC 
 
 1 
 
 1 
 
 1 
 1 
 
 Trolley line on paved street. 
 
 Cracked in 2 months — did not last through the 
 
 1 
 
 | 
 
 i 
 
 I f 
 
 ill 
 
 1 
 - 
 
 s 
 
 1 
 I 
 
 1 
 1 
 
 6 
 < 
 
 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 H 
 
 ■5 | ||s | US 
 
 
 i* 
 
 
 
 M 
 
 ■s .S 
 
 1 1 
 
 a 
 
 1 1 III 1 III 
 
 1 II II 1 
 
 1 
 
 I 1 11 1 
 
 £*& 
 
 g O -££ D. =0 
 
 
 
 S 2- S3 S» - : 
 
 "*§ 
 
 2 
 
 ■•3 ja i^ji ^ J " : 3 Sg g £12 &Z 
 p K Qow « slol 6 Id II HH tftf 
 
 ~ ^ JJ 1 
 p in mm < 
 
 1 
 1 
 
 *6 « 1 
 
 O «8 ai-c.2 3 ^g | || 
 
 
 1; II 
 
 Sri 11 
 
 « sa m . 
 
 d 1 III 1 
 
 I 
 
 1 
 
 
 1 
 
 < 
 
 
 
 1 
 
 o 
 
 
 
 
 1 
 
 1 
 < 
 
 
 
 
 ^ 
 
 
 1 
 
 Q 
 
 1 1 
 
 p s 
 
 « M < 
 
SUPPORTS OF THE RAIL 
 
 105 
 
 1 
 
 1 'IN 
 
 i il 
 1 jh 
 
 Ills 
 
 S |S'l 
 
 llll 
 OKOD 
 
 
 1 
 
 1 j 
 1 J 
 
 S j 
 
 Rails spread J — U baits weak. 
 
 100 replaced within 6 months (said to have been 
 frozen before setting), and all had been removed 
 by December, 1906. 
 
 
 1 f 
 1 ^ 
 
 
 1 
 < 
 
 
 
 
 
 
 
 6 
 
 8 
 
 3"= F-"hh 
 
 1 
 
 
 111 
 
 ^J^^,^^^^ c c 
 
 s ££%'■■ 
 
 1 
 
 >H >. >^ >- >-. >H >H g £ 
 
 1 
 
 1 1 II 
 1 
 
 Ills IIIIIII III III 1 
 £<2 o s ^^o QS^ 
 
 llco 
 
 Z °H 
 
 g j — „o ||g?SS^ S« i | 8 28SK 3 : 
 
 
 1 
 
 
 6 
 
 1 
 g 1 
 
 I 1 
 
 
 Q 
 
 |l 
 
 6| 
 
 1 
 
 1 
 
 SSSSSSh x -g 
 
 I.U11.I.I °. s ^ 
 6606066 1 J 1 
 
 lllllll ShS fr £ 
 
 1 
 1 
 
 1 i ■? ■? Ill 
 
 1 1 1 f III 
 
 
 £°? HHHHHKH = = « .> J g ';:.£.£ 1 K £ C J? 
 |.2 ^^h,.-;^ -g-g-g g gw^J^c -'g~_: 
 
 1 
 1 
 
 I § 
 
 ^ - '-a = 
 
 ■i 1 V 
 
 > D K fa 
 
 ! 
 
 
 
 
 
 I 
 6 
 
 5 PQ 
 
 
 ■3 
 1 
 
 
106 STEEL RAILS 
 
 Probably no form of reinforced concrete tie has been made which is suitable 
 for heavy and high-speed traffic. The real field of usefulness for the concrete 
 tie appears to lie in its application in places where speed is slow and where 
 conditions are especially adverse to the life of wood or metal. 
 
 The steel tie seems much more promising, but the fact remains that most 
 of the railroads in this country to-day are using wood, and, so far as the author is 
 able to judge from present tendencies, are likely to continue to do so for some time. 
 
 The question of a future timber supply for wood ties is a very important 
 one. The railroads are rapidly exhausting the available timber near their lines 
 and not only is the tie becoming dearer, but in many instances it is found im- 
 possible to obtain a sufficient supply to meet the annual requirements of the road. 
 
 The experience as set forth in a paper read at the American Forest Congress 
 by Mr. L. E. Johnson, President of the Norfolk and Western Railway, is typical 
 of most roads.* 
 
 "Originally the country passed through by the railroad was well timbered. 
 The first extensive depletion of timber land was on the first hundred miles 
 adjacent to the seaboard, where the original timber was cypress and Virginia 
 or loblolly pine. 
 
 "Up to the year 1888 the road used a great many cypress ties, but such 
 timber is no longer procurable. The second growth of Virginia loblolly pine 
 in this district is very knotty, and, further, it is not suitable for crossties until 
 it is treated to improve its lasting qualities. 
 
 "All the balance of the road is in territory where both white oak and 
 chestnut oak is indigenous, and up to quite recently all the crossties that have 
 been needed have been obtained within moderate hauling distance from the 
 railroad line. 
 
 "The average requirements in oak ties per year for renewals are 310 per 
 mile, aggregating, in round numbers, 800,000 ties per year for the entire road. 
 At prevailing prices 800,000 ties cost per annum about $315,000, which is shown 
 to be about 15 per cent over the cost of a like number ten years ago. This 
 total is far below what some railroads less fortunately situated must pay for 
 a like number." 
 
 The general distribution and character of the original forests f of the 
 United States are shown by Fig. 75. A glance at this discloses that five groups 
 
 * Proceedings of The American Forest Congress, Washington, 1905, p. 265. 
 
 t The Timber Supply of the United States, Kellogg. Forest Service, Circular 97. Original 
 Forests, R. S. Kellogg, Vol. 2, pp. 179, 180. Report of the National Conservation Commission, Feb- 
 ruary, 1909. 
 
SUPPORTS OF THE RAIL 
 
 107 
 
 of states embrace the natural timbered areas of the country, — the Northeastern 
 states, the Southern states, the Lake states, the Rocky Mountain states and 
 the Pacific states. Of these, the two groups last mentioned are occupied by 
 forests in which practically all the timber-producing trees are coniferous, the 
 
108 
 
 STEEL RAILS 
 
 first three of both conifers and hardwoods. The earliest attack was upon 
 the white pine of the Northeast, the original stand of which is almost 
 entirely cut out. 
 
 The Northeastern states reached their relative maximum in 1870 and the 
 Lake states in 1890. The Southern states are undoubtedly near their maximum 
 to-day, and the time of ascendency of the Pacific states is rapidly approaching. 
 There will be no more shifting after the Pacific states take first place, since 
 there is no new region of virgin timber to turn to. 
 
 The percentage of the total lumber cut, furnished by the principal regions 
 since 1850, according to census figures, is given in Table XIX. 
 
 TABLE XIX. 
 
 -GEOGRAPHIC DISTRIBUTION OF TOTAL 
 LUMBER PRODUCT 
 
 Year. 
 
 New England 
 
 &. 
 
 Southern 
 States. 
 
 K! 
 
 1850 
 
 Per cent. 
 54.5 
 
 36.2 
 36.8 
 24.8 
 18.4 
 16.0 
 
 Per cent. 
 6.4 
 13.6 
 24.4 
 33.4 
 
 27^4 
 
 Per cent. 
 13.8 
 16.5 
 9.4 
 11.9 
 15.9 
 25.2 
 
 Per cent. 
 
 3.9 
 6.2 
 3.8 
 3.5 
 7.3 
 9.6 
 
 1860 
 
 1870 . . . 
 
 1880. . . . 
 
 1890 
 
 1900. .. . 
 
 
 It is evident that at the present rate of consumption the available supply 
 of the present timber used for ties, especially white oak and yellow pine, will 
 be exhausted to a serious degree before many years, and that the railroads must 
 consider the question of what course they are to pursue in the future. 
 
 Under these conditions there are obviously two courses: First, the reduction 
 of the amount consumed, which can be done by the substitution of other 
 materials for wood and by the use of preservative methods for prolonging the 
 life of the tie, which, by increasing its durability, will diminish the annual 
 requirements for renewals; second, by the adoption of forestry methods, 
 having for their purpose the proper care and management of the forests still 
 remaining and the cultivation of new tree plantations. 
 
 The question of forest preservation and perpetuation is beginning to receive 
 attention in this country through the several State Bureaus of Forestry which 
 have been established, and attention is given to forest preservation by these 
 as well as by the National Government. 
 
 It has been found that the most important need for most of the railroads 
 at this time is definite technical information. It is not sufficient to know that 
 
SUPPORTS OF THE RAIL 
 
 109 
 
 timber supplies are being exhausted, but one should also know exactly what 
 these supplies are, and what the rate of exhaustion is, and what the probable 
 rate of regrowth is in any particular region upon which that particular road is 
 depending. 
 
 The need of such investigation is being universally felt, and has manifested 
 itself in very striking form, as shown by the two meetings of the governors of 
 
 Fig. 76. — Hunnewell Plantation (Catalpa). (Bureau of Forestry, Bulletin 37.) 
 Average diameter 3.85 ins., 21st year. 
 
 the various states, called by the President in May and December, 1908, in 
 Washington, D. C. 
 
 Many tree species * in the United States are adapted to a certain degree 
 at least for the production of crossties. Notwithstanding this, in making the 
 majority of railroad plantations only two species have been used. These two 
 species are catalpa and black locust. 
 
 Catalpa f has been planted for a great many years on a great variety of soils 
 
 * Proceedings Am. Ry. Eng. and M. of W. i 
 t Practical Arboriculture, J. P. Brown. 
 
 , 1908, Vol. 9, p. 715. 
 
110 STEEL RAILS 
 
 and throughout a wide range of territory, and although many plantations have 
 reached the age of twenty-five years or more,* so far as known, the trees in none 
 of the plantations have reached a size suitable for crossties (Figs. 76 and 77). 
 The black locust, although it is a rapid grower and thrives on a variety of soils, 
 is so subject to the attacks of insects that trees seldom reach a sufficient size 
 
 Fig. 77. — Farlington Forest (Catalpa). (Bureau of Forestry, Bulletin 37.) 
 Average diameter 4.39 ins., 21st year. 
 
 to make a crosstie. Trees which do reach this size are usually so weakened by 
 numerous cavities made by the boring of the insects that the wood cannot be 
 used with safety. 
 
 Table XX shows that of the total number of trees planted, the locusts 
 predominate, with the catalpa second; the results to date favor the former, 
 although it is perhaps too early fairly to estimate the ultimate value of any of 
 the plantations now under cultivation. 
 
 * The Hardy Catalpa, Bureau of Forestry, Bulletin No. 37. The Farlington Forest, p. 15. The 
 Hunnewell Plantation, p. 26. 
 
SUPPORTS OF THE RAIL 
 
 111 
 
 I 
 
 M 
 it 
 i 
 
 o definite knowledge obtained 
 from results to date, but condi- 
 tions are favorable for obtaining, 
 within about 20 years, f ie- and 
 posts amounting to double the 
 
 erty up to present time. Expect 
 to plant 10,000 trees in 1910. 
 
 s'f.sf 
 
 | 
 
 I 
 
 ost about 3,000 trees which were 
 planted on low ground and did 
 not stand the winter weal her. 
 Balance are thrifty and look well. 
 Cannot now estimate ultimate 
 
 rom results thus far obtained no 
 reliable estimate of ultimate 
 value can be made. Present re- 
 sults do not warrant additional 
 planting. Catalpa is decided 
 failure in arid and semi-arid 
 districts of Texas. In the humid 
 districts of Louisiana and Texas 
 conditions do not seem favorable 
 to its growth. 
 
 
 2 
 
 Eh 
 
 13 
 
 i-! 
 
 fe 
 
 l« 
 
 1 
 
 |s 
 
 
 
 °. 
 
 8 
 
 Is 
 
 1 
 I* 
 
 fj-2 § 
 
 tfsl 
 
 
 
 
 I 
 
 1 
 
 Is 
 
 8 
 
 Is 
 
 I! 5 
 
 « 
 
 8 
 
 6 
 
 
 < 
 
 S3 2,3 
 
 
 
 s 
 
 8 
 
 {{iff 
 
 
 s 
 
 <M 
 
 o 
 
 as 
 
 § 
 
 o|^|§ 
 
 
 si 
 
 a 
 
 2 
 
 s 
 
 
 iJ s a| 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 3*1 
 
 
 
 
 
 
 
 i-Jij 
 
 
 M 
 
 I 
 
 
 IS* a 
 
 
 i 
 
 
 6 
 
 m«S 
 
 6^3tS 
 
 ££ 
 
 6 
 
 ri 
 
 § 
 
 §|| 
 
 | 
 
 1111 
 
 1 §§ 
 
 | 
 
 o 
 
 
 
 
 
 
 
 
 
 S 
 
 1 
 
 l§S 
 
 | 
 
 III"-" 
 
 i 51 
 
 s" 
 
 s 
 
 £ 
 
 
 ^~ 
 
 M 
 
 
 N 
 
 
 
 
 
 ■a« : 
 
 
 
 
 
 1 
 
 
 
 1' 
 
 §|| | 
 
 
 
 1 86 1 
 
 1 
 
 
 
 I a 
 
 
 1 
 
 ^i! 
 
 . 
 
 § 
 
 ||| 
 
 § 
 
 _• 
 
 jjgS 
 
 1 
 5 
 
 1 
 o 
 
 l^S 
 
 1 
 1 
 
 IP 
 
 
 
 3 
 
 
 
 
 •§i 
 
 
 1 
 
 
 
 
 
 1 
 
 
 1 
 
 
 t£ 
 
 
 I 
 
 JN 
 
 I 
 
 •8 
 
 <3 
 
 i4 
 
 ||| 
 
 
 O 
 
 fc 
 
 
 >- 
 
 
 Q 
 
 
 o 
 
STEEL RAILS 
 
 s 
 s 
 
 fi 
 
 I'l 
 
 1 
 
 Too early for any estimate as to 
 ultimate value. Trees have se- 
 cured Vr root growth and it is 
 expected to cut all crooked and 
 branched trees to ground line 
 this winter. About 5,000 seed- 
 lings failed to take root at time 
 of planting. 
 
 Total of 123 acres planted. Lati- 
 tude of extreme coldness was 
 
 unsuitable for catalpa, and prac- 
 tically none of the trees grew. 
 
 and experiment discontinued. 
 Company has large holdings of 
 mountain timber, also tracts in 
 Eastern Virginia, containing Iml li 
 soft and hard woods, original and 
 second growths, which are being 
 held, only mature timbers being 
 
 Can give no estimate of ultimate 
 
 No remarks. 
 
 lift 
 
 Is 
 
 ll 
 
 •J 
 
 S 
 
 
 § i 
 
 
 1 
 Is 
 
 1 
 
 £ 
 
 
 g s 
 
 
 
 § 
 
 
 8 S 
 
 
 1 
 
 1 If If 
 
 o 6 6 do 
 
 
 1 
 
 1 1 1 1 1 
 
 
 1 
 
 Q 
 
 1 11 1 111 
 
 1 
 
 s 
 
 « '11 I ^ ! 
 
 H --3 ►". .2 S a 
 
 1 
 
 1 1 
 
 1 £ 
 
 1 
 
 5 
 
 1 ^ r 
 
 * «■ * 
 
 
 C3|| 5 
 
 i i - - 
 
 1 1 s g 
 -S'd 1 £ 
 
SUPPORTS OF THE RAIL 113 
 
 Tree planting as such by railway companies has not been a very successful 
 matter, and it is generally felt that the planting should be regarded as supple- 
 mentary to other methods for securing a tie supply, particularly to the manage- 
 ment of forest lands. 
 
 There are, without question, large areas of timber in the South which can 
 be obtained at a reasonable cost at the present time, and it seems to be very 
 much more advisable to buy forest regions, or where cut-over lands are pur- 
 chased, to encourage the growth of natural forest trees, rather than to go into 
 extensive experiments for the planting of new trees. 
 
 Forest planting in some cases may be desirable when a railroad has waste 
 land for which it has no particular use. It is a good object lesson to the farmers, 
 and if the plantations are successful they will net a fair return on the invest- 
 ment and furnish a limited supply of tie and timber for the future. 
 
 It should be observed, however, that it would not be practicable for the 
 individual roads to plant enough trees to supply their timber requirements, 
 and further the critical period of scarcity and high prices would come before 
 any of the trees so planted would reach maturity. 
 
 The information assembled by the Committee on Ties of the American 
 Railway Engineering Association, in 1910 (Table XX), shows what has been 
 done by the railroads in the way of tree planting; the situation is very little 
 changed at the present time, and, in the opinion of those best able to judge, 
 relief from this source is very uncertain. 
 
 If the railroads wish to provide against future scarcity and excessive 
 prices with any degree of certainty it will be necessary for them to actively 
 engage in forestry operations, having for their purpose the management of 
 mature timberlands and the cultivation and reforestation of the cut-over lands 
 within the forest area. This is an individual problem with every road, but, 
 generally speaking, it is the only sound policy which will provide for the future 
 requirements fifteen or twenty years hence. 
 
 Some of the railroads have now undertaken to preserve the timberlands 
 which they acquired through land grants or otherwise. The Southern Pacific 
 in northern California and southern Oregon still have quite large areas of good 
 timber from which they can cut mature trees. The Northern Pacific has been 
 cooperating with the government for some years with a view to finding how 
 best to handle their western holdings, and provide a source of tie supply at the 
 eastern end of their lines. In the East, the Delaware and Hudson have put 
 about one hundred thousand acres in the Adirondacks under management.* 
 
 * Timber Supply in Relation to Wood Preservation, E. A. Sterling. Proceedings, American 
 Wood Preserver's Association, 1911, pp. 140-144. 
 
114 STEEL RAILS 
 
 While the great desideratum is the obtaining of a permanent source of 
 supply of tie timber, the economic side of the problem must as well be 
 considered. 
 
 The application of actuarial methods to forestry is, despite the obvious 
 difficulties about the assessments of the different factors used in making calcula- 
 tions, the only correct way of estimating the financial position of timber crops 
 as a commercial investment. 
 
 The most profitable rotation is what should, both in theory and in practice, 
 receive most consideration in the management of a forest. It is found by mak- 
 ing various calculations, each as if for a single crop, in accordance with Faust- 
 mann's formula, and ascertaining that particular rotation which shows the 
 greatest profit by indicating the maximum productivity or largest capital value 
 of land and growing stock. 
 
 Faustmann's formula is as follows:* 
 
 , _ F n +(T a xl.0y n - a ) + (T b xl.0p«- b )+ ■ ■ ■ (T o xl.0p»-«)-(Cxl.0y n ) g 
 
 i.or-i 0.0/ 
 
 where A = The productivity of the woodlands (as estimated by the net 
 
 value of the timber crop, etc.); 
 F n = The net income, free from cost of harvesting, yielded by the 
 mature fall at the year (n); 
 T a , T b , . . . T g = The net income, free from cost of harvesting, yielded by 
 the thinnings at the years a, b, . . . q; 
 p = The percentage or rate of interest which the woodlands are 
 supposed to yield annually on the investment represented 
 by their capital value; 
 C = The cost of forming the crop originally, or of regenerating 
 
 or replanting the area on the fall of the mature crop; 
 g = The annual outlay for general charges (supervision, protec- 
 tion, taxes, etc.). 
 
 After determining the most profitable period of rotation, the amount of 
 land required to produce a given amount of ties annually can be found. 
 
 The cost per acre that can be paid for the land is determined as follows: 
 The average annual charge, at present prices, for different kinds of ties may be 
 taken as about 12.8 cents. 
 
 * The Forester, Nisbet, Vol. II, p. 239, London, MCMV; and Economics of Forestry, Fernow, 
 New York, 1902. 
 
SUPPORTS OF THE RAIL 115 
 
 This may be arrived at by the following relation. The discounted present 
 value of an annual rental or return r obtainable for n years in all, the rate of 
 interest being p, is expressed by the formula: 
 
 r _ r(1.0p» -1) nr C(1.0p»x0.0p) 
 
 1.0p»x0.0p 1.0p»-l 
 
 In the case of a white oak tie, 
 
 C = $ .90, cost of tie in the track, and r = $ .14 annual charge. 
 
 n = 8 years, life of tie; 
 
 p = rate of interest, 5 per cent. 
 
 The table given below shows the annual charge for different kinds of 
 wood. 
 
 White oak 14.0 cents annual charge 
 
 Heart pine 12.5 cents annual charge 
 
 Red oak, untreated 12.7 cents annual charge 
 
 Miscellaneous 12.0 cents annual charge 
 
 Assuming the life of a treated tie produced by the forest to be 12 years, the 
 value of such a tie can then be expressed by the formula 
 c r(1.0p»-l) 
 
 l.Op-xo.op* 
 
 where C = value of the tie; 
 
 r = return or annual charge, 12.8 cents, obtainable 12 years in all; 
 p = rate of interest, 5 per cent. 
 
 Substituting these values in the formula, there results for the value of the 
 tie $1.13. From this there must be deducted: 
 
 Cutting • $0.10 
 
 Handling 0.05 
 
 Treatment 0.30 
 
 Transportation 0.20 
 
 Putting in track , . 0.15 
 
 Total $0.80 
 
 which leaves for the stump value of the tie $0.33. 
 
 The amount of ties produced by the forest will depend upon the kind of 
 trees grown and the location of the tract. An annual yield of three ties per 
 acre should be expected under careful management in most cases of moderately 
 rapid growing trees. This will bring in a return per acre of: 
 
116 STEEL RAILS 
 
 R = 3 ties at $0.33, less management and taxes $0.30 = $0.69, and the 
 investment per acre which will give a five per cent return will be: 
 
 The wasteful methods employed in cutting ties in the past have called forth 
 many protests and suggestions as to how this waste might be checked. The 
 Forest Service states in this connection: 
 
 " The suggestions made for economy in the cutting of ties have been largely 
 in the direction of preventing wasteful cutting. The manner in which they 
 have been cut from trees has been largely determined by the ease and rapidity 
 with which ties could be made, and by the knowledge that certain portions of a 
 log were more serviceable for tie purposes than others. 
 
 " Ties were usually made out of heart wood, using the best and only the 
 straight, live trees. No attention was paid to the waste incurred by cutting 
 off all the sapwood top section, by leaving dead trees, etc. But with the intro- 
 duction of treated ties certain new developments in tie making have taken 
 place. Treated ties allow the use of sapwood, of sawed dead timber, and of 
 sawed ties, consequently tie forms which were altogether impracticable under 
 the old methods are now within the field of possibility, and must be considered 
 on their merits." 
 
 In view of this Dr. von Schrenk has proposed a form of half-round tie which 
 has been used extensively abroad (Figs. 78 and 79). The following description of 
 the proposed form is taken from his excellent paper on Cross-Tie Forms and Rail 
 Fastenings, f 
 
 This form of tie is probably a more economical tie than the present rectangular 
 tie used in this country, and, on account of its proved merits, should properly be 
 considered as a possible substitute for the present form. 
 
 If we consider the manner in which the load is distributed from the base 
 of a rail resting on a 5-inch plate, which in turn rests on a tie 8 inches broad, 
 we shall find that the lines of force acting from such a tie plate are distributed 
 on the ballast as indicated in Fig. 80. 
 
 * The cost per acre that can be paid for the forest land is based upon the annual charge of un- 
 treated ties as representing the average outlay by the railroads for this material at the present time. 
 The use of treated ties would probably reduce the annual charge per tie, but at the same time it must 
 be borne in mind that owing to the rapidly increasing cost of the timber from which the tie is made, 
 the annual charge for a treated tie will probably rise as high as the present figure for a natural tie, 
 before sufficient time has elapsed for the treated ties to affect the general average. 
 
 t Cross-Tie Forms and Rail Fastenings, Von Schrenk; Bureau of Forestry, Bulletin No. 50. 
 
SUPPORTS OF THE RAIL 
 
 117 
 
 Fig. 78. — Standard Prussian Ties of Baltic Pine. (Bureau of Forestry, Bulletin No. 50.) 
 
 Fig. 79. — Standard Oak and Beech Ties on the French Eastern Railway. 
 (Bureau of Forestry, Bulletin No. 50.) 
 
118 
 
 STEEL RAILS 
 
 Keeping in mind the desirability of an increased bearing surface on the 
 ballast, it is suggested that a type of tie with a top-bearing surface of about 
 
 Distribution of Pressure from Tie Plate Distribution of Pressure from Tie Plate in Half- 
 
 in Ordinary Tie. round Tie. 
 
 Fig. 80. — • Distribution of Pressure from Tie Plate. 
 
 6 *< 
 
 .81. — Half-round Tie Proposed by the 
 Forest Service. 
 
 6 inches and a base-bearing surface of 
 anywhere from 8 to 12 inches will not 
 only give a sufficient bearing surface 
 for the rail, but will also give a much 
 more stable track. Such a tie is shown 
 in Fig. 81. 
 
 Fig. 82 shows the 7 by 8-inch tie 
 and tie with 6-inch top and 12-inch 
 base, spaced as closely as is consistent 
 with the proper use of the shovel or 
 other tool employed to tamp the tie. 
 
 Fig. 82. — Spacing of Half-round Ties. 
 
 The comparative showing of rectangular 7 by 8-inch and 7 by 9-inch ties 
 and of ties with 6-inch top and 12-inch base, spaced respectively at 11 and 10J 
 inches, as shown in Fig. 82, is given in Table XXI. 
 
SUPPORTS OF THE RAIL 
 
 119 
 
 TABLE XXI. — COMPARISON OF RECTANGULAR AND HALF-ROUND TIES 
 
 (Bureau of Forestry, Bulletin No. 50) 
 
 New Tie, 
 6-inch Top, 
 12-inch Base. 
 
 Distance between bearing centers, on both top and base 
 
 tie, inches 
 
 Increase in distance between bearing centers by use of ties 
 
 of the new form, inches 
 
 Total number of ties per mile 
 
 Number of ties per mile saved by use of new form 
 
 Total linear bearing on ballast per mile, feet 
 
 Bearing surface on ballast per mile, with 8-feet length, 
 
 square feet 
 
 Gain in bearing surface by use of tie of the new form, square 
 
 feet 
 
 3,242 
 
 426 
 
 2,161 
 
 352 
 2,376 
 
 19,008 
 
 3,520 
 
 2,816 
 
 " 2,816' 
 22,528 
 
 According to this table the number of ties of the new form required per 
 mile is 352 less than with the 7 by 9-inch tie, and 426 less with the 7 by 8-inch 
 tie, while the amount of bearing surface obtained is greater by 3,520 square feet 
 than that obtained by the 7 by 9-inch tie, — an increase in bearing surface of 
 over one-sixth. At the same time there would seem at first sight to be a con- 
 siderable saving from the smaller number of ties, but in reality there is little 
 difference in expense because of the larger number of feet, board measure, in 
 the new tie. 
 
 It now becomes necessary to consider 
 the changed tie form from a lumber 
 standpoint. Ties are now being cut from 
 trees of all diameters from 9 inches 
 upward. If cut but one from a cross 
 section, they are usually termed pole 
 ties. Most of these are rounded at the 
 edge and squared on two sides (Fig. 83), 
 with a required bearing surface of 6 to 
 8 inches. Pole ties are now cut from 
 trees as large as 17 inches in diameter. Most of them are hewn, and in the 
 hewing much of the outer portion of the tree is wasted. In larger trees also a 
 great deal of timber is wasted, even when ties are split in the most economical 
 fashion. In the majority of instances no wane is admitted for a first-class tie, 
 so that logs less than 10 inches in diameter will not make ties of this class. This 
 means that a great many tops are now left in the woods because they are too 
 small. By adopting the half-round tie suggested above (and here emphasis 
 
 Fig. 83. — Pole Tie. 
 
120 
 
 STEEL RAILS 
 
 ought to be laid upon the fact that ties cut according to this shape will all be 
 treated) it will be possible to utilize a great many logs which now do not make 
 ties, and also to cut a good many more ties out of the same amount of timber 
 than under the present specifications. 
 
 The cutting of ties of this new form will be essentially a sawmill proposi- 
 tion. Where now there is a great deal of waste in hewing, if the log were sawed, 
 
 it would mean the obtaining of several 
 boards on the side. The number of 
 boards to be sawed from a tree 16 inches 
 in diameter, making two ties, will depend 
 largely upon the value of the timber from 
 which the ties are made. For instance, 
 it will pay to make as many boards as 
 possible out of a 16-inch, two-tie log of 
 red oak or gum, while with timber like 
 loblolly pine, the lumber of which has a 
 low value, it will at present not pay to 
 cut off many boards. In the case of such 
 timber an extreme form of the half-round 
 tie will be applicable (Fig. 84). 
 The influence which the new tie form will have upon the size of trees cut 
 for tie purposes ought to be a marked one. It certainly would discourage the 
 cutting of pole ties to a very considerable extent. It would not pay to make a 
 tie out of a small tree when by leaving it for a few years two ties could be made 
 from the same tree. In other words, the present policy of cutting trees 11 or 
 12 inches in diameter would be found less profitable than cutting trees 16 or 
 17 inches in diameter. 
 
 There is probably no other branch of the lumber industry in which so 
 many small trees are annually destroyed and the possible regrowth of forests 
 retarded to such an extent as in the manufacture of ties. The practice of sawing 
 ties from logs is going to be more and more prevalent as the old feeling that a 
 sawed tie is not worth having disappears. This feeling is already rapidly dis- 
 appearing. It certainly will disappear entirely when railroad men realize that 
 with a chemically treated tie it makes no difference whether it be sawed or 
 hewn. With increasing permanency in the source of supply, it will pay more 
 and more to put up small sawmills, which will saw ties and such lumber as may 
 incidentally come to them. This will be particularly true in regions where 
 there are rapidly growing tree species, such, for instance, as loblolly pine. The 
 
 - Extreme Form of Half-round Tie. 
 
SUPPORTS OF THE RAIL 
 
 121 
 
 cutting of these trees will, moreover, make possible the use of large quantities 
 of timber which now is practically wasted and from which the lumberman has 
 no return. This is particularly true of tops. 
 
 As the rail should be designed to have sufficient stiffness to enable it to 
 distribute the load over a number of ties, allowing only such a proportion of 
 the wheel load to come on each tie as can be safely carried, it will be necessary 
 to determine the safe load that it will be proper to put on the tie. As a mean 
 representing the average general practice, we may take in the following discus- 
 sion a 7 by 8-inch by 8-foot 6-inch tie and a 7 by 9-inch by 8-foot 6-inch tie 
 (see Table XXII). It would seem desirable also to consider the strength of 
 the half-round tie. 
 
 TABLE XXII. — SIZE OF TIES AND SPACING 
 
 (Am. Ry. Eng. Assn.) 
 
 Southern 
 
 Penn. R.R 
 
 L. &N 
 
 B. &0 
 
 N. & W 
 
 P. &R 
 
 Penn. (S. W. Sys.). 
 
 Lehigh Valley 
 
 N.,C. &St. L 
 
 D. & H. Co 
 
 A., B. & A 
 
 Cent, of N.J 
 
 B.,R. & P 
 
 C, C. &0 
 
 A. C. L 
 
 Penn. (N. W. Sys.). 
 
 D., L. & W 
 
 Fla. East Coast... . 
 C.,C.,C. &St. L... 
 Hocking Valley.... 
 
 L. S. &M. S 
 
 Erie 
 
 Long Island 
 
 South. Pacific 
 
 Union Pacific 
 
 S. A. L 
 
 N. Y., N. H. &H.. 
 
 C. of Ga 
 
 G.,H. & S. A 
 
 Georgia 
 
 M. &0 
 
 Norfolk Southern . . 
 N. Y. C. &H. R... 
 Great Northern. . . . 
 S. P., L. A. & S. L. 
 Northern Pacific. . . 
 
 D. & R. G 
 
 C..B. &Q 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8^ 
 
 7X7 and 9X8| 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 ■7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 
 7X7 and 9X8! 
 7X9X8 
 7X9X8 
 7X9X8 
 7X9X8 
 7X9X8 
 7X9X8 
 7X9X8 
 7X9X8 
 7X9X8 
 7X9X8 
 7X8X8 
 7X8X8 
 7X8X8 
 7X8X8 
 6X8X8 
 
 2880 
 2880 
 2880 
 2880 
 
 2816 
 2816 
 2816 
 2816 
 3300 
 3050 
 3040 
 2720 
 2720 
 
 3164 
 2816 
 3200 
 
 2900 
 3200 
 3200 
 
 C, R. I. & P 
 
 St. L. &S. F 
 
 Grand Trunk 
 
 M.,K. & T 
 
 Col. & Sou 
 
 Maine Central 
 
 C. &E. I 
 
 C, I. & L 
 
 El. P. &S.-W 
 
 St. L., B. & M 
 
 Ft. W. & D. C 
 
 C. &N.-W 
 
 C.,M. & P. S 
 
 C.,M. & St. P 
 
 C. I. &s 
 
 St. L. S. W 
 
 M. &St. L 
 
 S. A. & A. P 
 
 Rutland 
 
 Mo. &N. Ark 
 
 S. Fe, P. &P 
 
 L. E. &W 
 
 G. R. &I 
 
 W. &L. E 
 
 N. W. Pac 
 
 Mo. Pac 
 
 B. &M 
 
 K. C.,M. & O 
 
 Tenn. Cent 
 
 C. G. W 
 
 C.,H. & D 
 
 M. C 
 
 Bangor & Aroostook . 
 
 N. Y.,0. & W 
 
 M., J. & K. C 
 
 a, st. p.,m.&o... 
 
 D., S. S. & A 
 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X8X8 
 6X6X8 
 6X9X8 
 7X9X9 
 7X7X8 
 7X7X8 
 
 3200 
 3200 
 3200 
 3200 
 3200 
 3200 
 3200 
 3200 
 
 3000 
 3000 
 2992 
 2992 
 2992 
 2992 
 2992 
 2900 
 2880 
 2880 
 
 2816 
 2816 
 2816 
 2816 
 
122 
 
 STEEL RAILS 
 
 14. Bearing of the Rail on the Tie 
 The general tendency at the present time is more and more towards the use of 
 tie plates. "With the introduction of the treated tie it is necessary to adopt some 
 means to protect the wood from wear at the rail bearing on account of the longer 
 life of the tie. 
 
 The objections which have been made to tie plates were, first of all, that 
 they buckled severely. This, however, has taken place only when the plates 
 
 were too thin, and the 
 following record of tests 
 made of a prominent 
 make of tie plate show 
 that the present plates 
 have ample strength to 
 resist buckling (Table 
 XXIII and Fig. 85). 
 
 Most plates have 
 been made with the idea 
 of being anchored to the 
 tie so as to prevent the 
 communication of the 
 motion of the rail to the plate. As a result, we have a large number of different 
 forms of plates, provided with prongs, spines, or flanges on the bottom, which 
 are pressed into the tie either by table xxiii.-test of McKEE tie plate 
 the weight of the passing load or 
 before the rail is laid (see Plate 
 XXII). 
 
 The chief objection which is 
 made to plates at this time, par- 
 ticularly in connection with the use 
 of softer woods, is that not only do 
 they not aid in preventing the wear 
 of fibers, but they actually assist 
 the rail to wear. This is well illus- 
 trated in Fig. 86, showing a tie plate 
 which has been in position on a loblolly pine tie for about four years. 
 
 The constant rocking motion of the rail, which had become very marked 
 as the spikes were pulled from the soft wood, had transmitted itself to the tie 
 plate, and when a load passed over the rail the tie plate moved back and forth 
 
 McKee Tie Plate. 
 
 Loads as Applied, Te 
 
 st No. 1, Tes 
 
 t No. 2, 
 
 Pounds. Deflec 
 
 ion Inches. DeEect 
 
 on Inches. 
 
 250 
 
 000 
 
 000 
 
 4,000 
 
 022 
 
 022 
 
 8,000 
 
 029 
 
 030 
 
 12,000 
 
 035 
 
 038 
 
 16,000 
 
 042 
 
 056 
 
 20,000 
 
 054 
 
 070 
 
 24,000 
 
 066 
 
 081 
 
 28,000 
 
 081 
 
 100 
 
 32,000 
 
 101 
 
 129 
 
 36,000 
 
 122 
 
 165 
 
 40,000 
 
 154 
 
 210 
 
 44,000 
 
 188 
 
 248 
 
 48,000 
 
 221 
 
 285 
 
SUPPORTS OF THE RAIL 
 
 123 
 
 in unison with the rail. It was not long before the soft fibers of the loblolly- 
 pine suffered under this treatment, and in the course of time so great did the 
 abrasion and crushing of the fibers by the plate become that a considerable hole 
 was made under the plate, in which water gathered. The plate gradually sank 
 
 . Fig. 86. — Wear of Tie under Tie Plate. 
 The upper illustration shows a Loblolly Pine Tie treated with Zinc Chloride, after four years' service 
 
 in Texas. 
 The lower illustration shows a longitudinal section through the spike hole of a Western Yellow Pine 
 Tie after several years' service in Texas. 
 (Bureau of Forestry, Bulletin No. 50.) 
 
 down into this hole, as shown in the illustration. When the tie was removed 
 it had disappeared in the wood, and the base of the rail was resting on the outer 
 edges of the tie beyond the plate. 
 
 This tie had been treated with zinc chloride. The water which gathered 
 under the tie plate leached out the salt, and as a result decay started on both 
 sides of the plate, as the illustration shows. The tie had to be removed, al- 
 though the rest of it was perfectly sound (Figs. 87 and 88). 
 
STEEL RAILS 
 
 
 % 
 
 
 
 /$ v5 , -*-,~ N . 
 
 i4 
 
 ft. 
 
 
 ■2 
 
 
 ^^^hUhhP 
 
 j— 
 
 Fig. 87. — Section of Tie under Rail Bearing showing Wear and Decay, 
 
 Fig. 88. — Section from Middle of Same Tie showing Entire Soundness. 
 
 LOBLOLLY PINE TIE TREATED WITH ZINC CHLORIDE, AFTER FOUR 
 YEARS' SERVICE IN TEXAS. 
 (Bureau of Forestry, Bulletin No. 50.) 
 
SUPPORTS OF THE RAIL 125 
 
 The types of plates used in Europe are without exception flat plates.* Figs. 
 I to 96 represent plates used by several European roads at the present time. 
 
 Plan of Tie Pla 
 
 Fig. 89. — Belgian State Railways, 105-lb. Rail and Tie Plate. (Am. Ry. Eng. Assn.) 
 
 * See "The Question of Screw Fastenings to Secure Rails to Ties." W. C. Cushing, Proceedings 
 Am. Ry. Eng. and M. of W. Assn., 1909, Vol. 10, Part 2. 
 
STEEL RAILS 
 
 . 
 
 
 
 3.42" 
 
 
 
 o 
 
 
 0.67" 
 
 5 
 
 , 138" 
 
 
 039 
 
 
 
 ,3|oflS" 
 
 °1 
 
 
 
 
 Fig. 90. — Belgian State Railways, 115-lb. Rail and Tie Plate. (Am. Ry. Eng. Assn.) 
 
SUPPORTS OF THE RAIL 
 
 127 
 
 Intermediate liooK Plate. 
 Fig. 91. — Kingdom of Wtirttemberg State Railways, Tie Plate. (Am. Ry. Eng. Assn.) 
 
128 
 
 STEEL RAILS 
 
 Fig. 92. — Bavarian State Railways, Joint Hook Plate. (Am. Ry. Eng. Assn.) 
 
SUPPORTS OF THE RAIL 
 
 129 
 
 Joint ' hook Plate.. 
 - Kingdom of Saxony State Railroad, Joint Hook Plate. (Am. Ry. Eng. Assn.) 
 
STEEL RAILS 
 
 Fig. 94. — Elsass-Lothringen State Railways, Tie Plate. (Am. Ry. Eng. Assn.) 
 
SUPPORTS OF THE RAIL 
 
 SECTION THROUGH JOINT TIE 
 
 1 (T) 
 
 
 
 \ 
 
 O 
 
 
 
 , l-3o' 
 
 
 ■ w , 
 
 l.46'j fe 
 
 
 
 
 1 1 
 
 4 
 
 
 , 1.89" 
 
 
 
 
 
 A] oj2i 
 
 
 
 
 
 
 
 6.30" 
 
 L° 
 
 3 
 
 
 . °i 95 ' 
 
 
 
 
 
 ; 
 
 
 
 11.42" 
 
 
 
 l 
 
 
 iKV 
 
 
 ,, 
 
 
 1 
 
 
 ,, 
 
 
 Fig. 95. — Prussian State Railways, Tie Plate. (Am. Ry. Eng. Assn.) 
 
132 
 
 STEEL RAILS 
 
 Intermediate Wedge Plate 
 Fig. 96. — Bavarian State Railways, Intermediate Wedge Plate. (Am. Ry. Eng. Assn.) 
 
 The general tendency on the Continent has been toward adopting more and 
 more rigidly flat plates, with firm fastenings. The almost universal adoption 
 j,,,^,^.^ of this principle is very striking at the 
 
 present day. 
 
 On the French Eastern* the rail 
 rests on the tie without metallic plates, 
 except on very sharp curves (of 984.25 
 feet radius and under) . Plates of poplar 
 or felt are placed under the rail, solely to 
 protect the wood against the mechanical 
 action of the base. These plates are 
 compressed before being used, so that 
 they will not be further compressed 
 under the pressure of the rail. The 
 plates are furnished 0.28 inch thick, and 
 
 Fig. 97. — Wooden Tie Plate on French Eastern, the Compression brings them to 0.16 inch. 
 * Bureau of Forestry, Bulletin No. 50, von Schrenk. 
 
SUPPORTS OF THE RAIL 133 
 
 The ties are adzed at the treating plant so that a place is left for this flat 
 wooden shim. When the track is laid, the shim is placed in position (Fig. 97) 
 and screw spikes are screwed into the tie. Their pressure holds the plate firmly 
 between the base of the rail and the tie. In Fig. 97 the wooden tie plate is 
 represented by the thin unshaded portion between the rail and the tie. It is 
 exactly the width of the rail. In the course of time the motion of the rail wears 
 out this shim, and a new one is substituted by giving the screw spike one or 
 two upward turns. A new plate is then shoved in endwise and the screw is 
 fastened. The length of life of one of the wooden shims on the main-line tracks, 
 such as that of the French Eastern from Paris to Strassburg, is about one and 
 one-half to two years. 
 
 Dr. von Schrenk gives the theory upon which this wooden plate is used as 
 follows : The principal function of the plate has been said to consist of prevent- 
 ing the wear of the fibers of the tie immediately under the rail base. This wear 
 consists in the actual breakage of the wood fibers under a grinding and tearing 
 action rather than in crushing them. 
 
 In considering the function of the tie plate we have three bodies to deal 
 with: the tie, the tie plate, and the rail. Motion might conceivably take place 
 either between the rail and the tie plate or between the tie plate and the tie. 
 When a metal tie plate is used on the hardwood tie, and is successfully anchored 
 in it, the tie plate and the tie act as one body, over which the rail moves back 
 and forth. As soon as the tie plate loses its holding power, however, the chances 
 are that when the rail moves across the tie the tie plate will oscillate back and 
 forth in unison with the rail. This results in breaking the wood fibers under- 
 neath the plate. Where a wooden plate is used, it adheres so closely to the 
 wood that when the rail moves across the tie the wooden plate and the 
 wooden tie are liable to act as one, even though the tie plate is not anchored 
 to the tie. 
 
 The Forest Service tests have not shown results favorable to wooden tie plates. 
 While the tests have not been very thorough, they have been thought to throw 
 much doubt on the efficiency of this form of plate. 
 
 * For some years the question of a satisfactory fastening between rails and 
 soft-wood ties has been a subject of continuous experiments on the Prussian 
 Government railroads. The first investigations followed the general use of 
 plain bearing plates, 7| by 6j inches (rolled steel) in size, shown in Fig. 98 a. 
 
 * Fastening of Rails to Soft Wood (Pine) Ties, Organ fur die Fortschritte des Eisenbahnwesens, 
 May 15, 1908, et seq. Translation appears in Vol. 10, Part 2, Proceedings Am. Ry. Eng. & M. of W. 
 Assn., p. 1533. 
 
134 
 
 STEEL RAILS 
 
 It was soon discovered that on soft-wood ties, the small adhesion between 
 the spike and the wood permitted the spikes to pull out to a more or less extent, 
 and the loose rail, under the sudden applications of load, would quickly batter 
 down the wood. Besides, the pressure on the tie not being uniform would 
 produce a kind of convex wear in the wood, as illustrated in Fig. 98 b. 
 
 Some improvement was obtained by the use of screw spikes, without, how- 
 ever, entirely overcoming the abnormal wear and the consequent looseness 
 
 Fig. 98. — Plain Bearing Plates, German Experiments on Tie Plates. 
 
 and inefficiency of the structure (Figs. 98 c and d). The hook plate, shown 
 in Fig. 99 a, was next tried. The hook, which was made to hold the outside 
 flange of the rail, necessitated a plate somewhat longer on the outside than on 
 the gage side, resulting in an uneven distribution of pressure on the wood and 
 a condition as shown in Fig. 99 6. The bending of the screw spikes observed 
 in this case was at first thought to be due to the lack of support of the head of 
 the spikes on the far side from the rail flange, and, to r-medy this supposed 
 defect, rail clips were introduced, giving the head of spikes a full support all 
 
SUPPORTS OF THE RAIL 
 
 135 
 
 around (Fig. 100 a). This arrangement proved to be much better than any- 
 previous one, but still did not produce a satisfactory fastening. Fig. 100 b indi- 
 cates clearly the manner of failure of these plates. 
 
 Hook Plates, German Experiments on Tie Plates. 
 
 It is evident that the plate hook, being rigid and incapable of producing 
 
 any actual pressure against the rail flange, would cause all the stresses to be 
 
 carried against the screw fastening, and as soon as this would wear in any of 
 
 its parts the rail would become loose under the hook, and the shocks would 
 
 CX 
 
 Fig. 100. — Hook Plates with Clips, German Experiments on Tie Plates. 
 
 begin their destructive work. Also, the direct pressure under the head of the 
 spikes would tend to pull these out of the tie, and constitute another element 
 of weakness to the general construction. 
 
 The impossibility of fastening the rail with the same amount of holding 
 power on both sides, besides the drawbacks enumerated above, led to the intro- 
 duction of an entirely different system of fastening. 
 
 The first set, or Group 1, of plates are shown in Fig. 101. These plates had 
 a bearing on the ties of 90 square inches as against 80 square inches in the 
 largest previous plate, and were fastened to the ties by means of four screw 
 spikes absolutely independent of the rail fastening. The rail was in its turn 
 
STEEL RAILS 
 
 fastened to the plate by means of two bolts and clips, these being independent 
 of the tie fastening. 
 
 The clips were made so as to be capable of adjusting the gage of the track 
 by being reversible, and also of such a shape as to take up and transmit hori- 
 zontal forces at the base of the rail to the shoulders provided for in the tie 
 plates. In this manner the upward forces would be resisted by all the screw 
 spikes, and similarly all the horizontal forces would be taken care of. Spring 
 
 . 
 
 r< 
 
 12% 
 
 ---320."?™- 
 
 > 
 
 
 IT 
 
 of 
 
 ( )( 
 
 1° 
 
 f 
 
 AJjLA * 
 
 
 
 IiiwmM ErLFN 
 
 Ljp. 
 
 
 ° 
 
 1 
 
 Fig. 101. — Group 1, German Experiments on Tie Plates. 
 
 washers were provided under the head of the screw spikes and the rail-fastening 
 nut. 
 
 Eighty-three of these plates were put in service in 1898 and removed from 
 the tracks, together with the ties, in 1907 for examination. The tie wear 
 was found to be very slight and very uniform under the base of the plate, 
 varying from a minimum of 0.14 millimeter (iJ-g- inch) to a maximum of 0.19 
 millimeter (-j-jj-g- inch), except in a few cases where spikes had become loose 
 and caused an increased as well as an irregular wear. 
 
 n 
 
 250 m 
 
 !S -, 
 
 
 LTo J 
 
 
 \ °T 
 
 
 nil 
 
 
 III, 
 
 
 „ 1 
 
 
 00 i 
 
 ill 
 
 ||lJ 
 
 : l 
 
 YL° \ 
 
 
 ( 6\ 
 
 
 Fig. 102. — Group 2, German Experiments on Tie Plates. 
 
 Sand, however, was found between the ties and plates, and this might 
 have caused even this slight wear. The rail seat in the tie plate had worn to 
 about the same extent (maximum iwo inch). No other sign of deterioration 
 was observed. 
 
 The second series of tests, Group 2 (Fig. 102), was carried on under con- 
 ditions similar to those for Group 1. The main difference between these plates 
 
SUPPORTS OF THE RAIL 
 
 137 
 
 and the plates of Group 1 lies in the size of the bearing surface over tie, being 
 about 70 square inches for Group 2 plates as against 90 square inches for Group 1 
 plates. 
 
 On removal of the plates, it was found that the wear of tie developed the 
 same uniform wear as in the previous group. The slight wear gave the impres- 
 sion of being purely compression, there being no indication whatever of side dis- 
 placement. As a matter of fact, after the screw spikes had been removed the 
 plate had to be knocked off with a hammer. Sand was found under the edge 
 of only a few plates. The screw spikes used on these plates were 4f inches by 
 f inch and had a deeper thread than those of Group 1. 
 
 In spite of the smaller bearing of this plate, as compared to Group 1, the 
 amount of tie wear was actually smaller, and the fastening generally more 
 satisfactory. 
 
 Center of Rail^l 
 
 10" 
 
 - 3 /8 ' 
 
 4 7 A 
 
 Fig. 103. — Group 3, German Experiments on Tie Plates. 
 
 Condition of test in Group 3 (Fig. 103) was similar to the previous tests in 
 Groups 1 and 2. The main difference in this case consisted in the sloped top 
 of the tie plate, which gave the rail a desired amount of inclination toward the 
 gage side. This arrangement brought the center of rail closer to the outer edge 
 of the tie plate by about 10 millimeters (f inch). Screw spikes used were similar 
 to those in Group 2, but with a somewhat better grip, the holes having been 
 drilled smaller. To increase the rigidity of the fastening, double spring washers 
 were employed on all the screw spikes. 
 
 The tie wear was smaller than in any previous instance. The gage of the 
 track was measured frequently and found to remain practically unchanged. 
 The spring washers, which had shown some failures when used singly, were 
 found in this test to have their original elasticity unimpaired. This design of 
 tie plate, however, failed in a few instances, as shown at "a," Fig. 103, which 
 would seem to indicate that a greater stress was carried against this point than 
 in the other arrangements. 
 
138 STEEL RAILS 
 
 It is clearly evident from the behavior of plates of Groups 2 and 3 that the 
 wear of ties is not at all directly proportional to the extent of the bearing surface 
 of the tie plate, but depends more upon the rigidity of the fastening. In the 
 case under consideration, the most important point developed is the necessity 
 of rigidly fastening the tie plates to the ties in order to preserve the life of the tie. 
 
 15. Fastening of the Rail to the Tie 
 In this country the ordinary nail spike is generally used for fastening a rail 
 to a wooden tie. The most important objections to the spike are: first, in 
 
 Fig. 104. — Short Leaf Pine Tie, after 2 Years' Service, cut through Spike Holes. 
 (Bureau of Forestry, Bulletin No. 50.) 
 
 the soft-wood tie the spike does not hold with sufficient firmness to keep the 
 rail securely to the tie; second, in driving the spike into the softer woods the 
 fibers are broken to an unusual extent (Fig. 104). As a result they do not 
 withstand lateral pressure of the rail, and consequently the spike hole is rapidly 
 increased to such an extent that the spike no longer holds. Water collects in 
 the enlarged hole and decay sets in (Fig. 105). 
 
 Table XXIV * compares the holding force of a common spike (Fig. 106), 
 weight 165 spikes to 100 pounds, with that of the common screw spike (Fig. 107), 
 similar to those used on the French and other continental railroads, weight 
 85 spikes to 100 pounds. 
 
 * Holding Force of Railroad Spikes in Wooden Ties, Forest Service, Circular 46. 
 
SUPPORTS OF THE RAIL 
 
 139 
 
 
 
 
 L ^ 
 
 
 
 1,1 
 
 - 
 
 -! - :t 
 
 
 ; -J 
 
 t>' • 
 
 till 
 
 
 : -"JS 
 1 
 
 
 
 VJ-> 
 
 Iv-"^ 
 
 vtfi 
 
 
 _iL*-*^ 
 
 a^i^jp^ 
 
 gii^'^-i^U%^^^*~ 
 
 - Cross Section through the Spike Holes of Short Leaf Pine Tie, treated with Zinc Chloride, 
 Texas. (Bureau of Forestry, Bulletin No. 50.) 
 
 TABLE XXIV.- 
 
 • HOLDING FORCE OF COMMON AND SCREW SPIKES 
 
 (Forest Service, Circular 46) 
 
 
 Number 
 of Tests. 
 
 Condition of Wood. 
 
 Force Required to Pull Spike. 
 
 of Spike. 
 
 Average. 
 
 Maximum. 
 
 Minimum. 
 
 White oak: 
 
 Common spike 
 
 5 
 5 
 
 Partially seasoned 
 
 Pounds. 
 
 6,950 
 
 13,026 
 1.88 
 
 Pounds. 
 
 7,870 
 14,940 
 
 6,160 
 11.050 
 
 
 
 
 
 5 
 
 8 
 
 
 
 
 Oak (probably red): 
 Common spike 
 
 4,342 
 
 11,240 
 
 2.61 
 
 5,300 
 13,530 
 
 3,490 
 
 ...do 
 
 8,900 
 
 
 
 
 
 28 
 26 
 
 
 
 
 Loblolly pine: 
 Common spike 
 
 3,670 
 7,748 
 2.11 
 
 6,000 
 14,680 
 
 2,320 
 
 ..do 
 
 4,170 
 
 
 
 
 12 
 14 
 
 Green 
 
 ....do 
 
 
 
 Hardy catalpa: 
 Common spike 
 
 3,224 
 8,261 
 2.56 
 
 4,000 
 9,440 
 
 2,190 
 6,280 
 
 
 
 
 11 
 11 
 
 Green 
 
 ....do 
 
 
 
 Common catalpa: 
 Common spike 
 
 2,887 
 6,939 
 2.42 
 
 4,500 
 8,340 
 
 2,240 
 5,890 
 
 
 
 4 
 5 
 
 
 
 
 Chestnut: 
 Common spike 
 
 2,980 
 9,418 
 3.15 
 
 3,220 
 11,150 
 
 
 ....do 
 
 7,470 
 
 
 
 
 
 
 
 
STEEL RAILS 
 
 t-%"- 
 
 Fig. 106. — Common S 
 
 Fig. 107. — Common Screw Spike. 
 
 Tables XXV and XXVI are taken from "Studies of the Stability of Rail- 
 way Tracks," by Jules Michel,* and give the holding power of hook and screw 
 spikes. 
 
 TABLE XXV. — PULLING FORCE NECESSARY TO PULL OUT FOR 0.20 INCH IN A 
 HOOK SPIKE AND A SCREW SPIKE BURIED 4.13 INCHES IN THE WOOD 
 
 (Jules Michel) 
 
 P.L.M. Hook S 
 
 P.L.M. Screw Spike. 
 
 Poplar 
 
 Larch 
 
 Baltic fir, creosoted 
 
 Beech, treated with sulphate of copper. . 
 
 Oak 
 
 American cypress 
 
 Pounds 
 
 992 
 1,598 
 
 Pounds 
 
 4,454 
 5,291 
 5,732 
 
 Figs. 108 and 109 present examples of early screw fastenings. 
 
 Fig. 108. — Screw Spike used by Grand Duchy of Baden State Railways (1860). 
 * Revue Generate des Chemins de Fer, July, 1884, and June, 1893. 
 
SUPPORTS OF THE RAIL 
 NORTHERN 
 
 Fig. 109. —Early French Screw Spikes (1863). 
 
 TABLE XXVI. — FORCES NECESSARY FOR EXTRACTING BY 0.20-INCH SCREW 
 SPIKES 0.79 INCH AND 0.91 INCH IN DIAMETER WITH THREADS OF 0.39-INCH 
 AND 0.59-INCH PITCH, SUNK 4.14 INCHES IN WOOD OF VARIOUS SPECIES AND 
 AGES 
 
 (Jules Michel) 
 
 Diameter 
 of Screw 
 Date of Spike. 
 Trial. 
 
 Pitch of 
 
 Northern 
 
 New Wood Currently Employed in Tracks 
 
 1875 
 
 0.79 
 
 0.79 
 0.79 
 
 0.79 
 0.79 
 0.79 
 0.91 
 
 0.39 
 
 0.39 
 0.39 
 
 0.49 
 0.59 
 0.49 
 0.49 
 
 5,733 
 
 
 9,481 
 10,143 
 
 9,923 
 
 10,584 
 11,576 
 
 12,844 
 12,458 
 
 0.59 
 
 0.59 
 0.55 
 
 0.55 
 0.55 
 
 
 1881 
 
 
 July, 1884. 
 
 1889 
 
 
 
 Rolled screw spikes, 
 24 trials. 
 
 1889 
 
 
 
 
 1889 
 
 
 
 
 
 1891 
 
 7,640 
 
 11,378 
 
 
 
 1889 
 
 13,010 
 
 12,348 
 
 0.67 
 
 
 
 
 
 
 Wood Having Been 9 Years in Track 
 
 0.79 
 0.79 
 
 0.39 
 0.49 
 
 10,143 
 11,576 
 
 Fig. 110 shows a machine used on the Atchison, Topeka and Santa Fe 
 for preparing ties for screw spikes. Wooden dowels as shown in Fig. Ill are 
 screwed into the ties.* Table XXVII gives the cost of equipping a mile of track 
 with screw spikes, the estimate being based on work actually done on a section 
 of track five miles in length on the Illinois division of the railway. 
 
 * Railroad Age Gazette, December 24, 1909. 
 
142 
 
 STEEL RAILS 
 
 Fig. 110. — Machine Preparing Ties for Screw Spikes. (Railroad Age Gazette.) 
 
 Hole Tapped Plug in Place Dressed for Tie Plate Completed Rail 
 
 Fastening 
 
 Fig. 111. — Showing Application of Screw Spike on A. T. & S. Fe R. R. (Railroad Age Gazette.) 
 
 TABLE XXVII. — ONE MILE OF TRACK WITH SCREW SPIKES AND DOWELS 
 
 12,000 spikes at 2.7 cents each : $ 324 
 
 6,000 tie plates at 21 cents each 1,260 
 
 Boring ties for, and driving, 24,000 dowels, at 1 cent each 240 
 
 24,000 wooden dowels at 1| cents each 360 
 
 Driving screw spikes (per mile) 150 
 
 Total $2,334 
 
 ONE MILE WITH CUT SPIKES 
 
 12,000 spikes $ 127 
 
 6,000 tie plates at 21 cents each 1,260 
 
 Driving cut spikes (per mile) 150 
 
 Total $1,537 
 
 * Apparently the French railways were about the first in Europe to begin 
 the use of the screw spike (tirefond) as a rail fastening, and it is to-day uni- 
 versally employed by the large systems (Fig. 112, Table XXVIII). 
 
 * For a very full discussion of the subject, see " The Question of Screw Fastenings to Secure Rails 
 to Ties," W. C. Cushing, Proceedings Am. Ry. Eng. & M. of W. Assn., 1909, Vol. 10, Part 2, p. 1456. 
 
SUPPORTS OF THE RAIL 
 
 14a 
 
 Fig. 112. — French Railways — Rail Fastenings. (Am. Ry. Eng. Assn.) 
 
144 
 
 STEEL RAILS 
 
 TABLE XXVIII. — FRENCH RAILWAYS — RAIL FASTENINGS 
 
 (Am. Ry. Eng. Assn.) 
 
 Railway. 
 
 Number and 
 Position of 
 Screw Spikes 
 at Each End 
 of Tie. 
 
 Type of 
 Rail Used. 
 
 Screw Spikes. 
 
 Number of 
 Ties Used 
 per Rail. 
 
 Joints. 
 
 Name. 
 
 - 
 
 J~ 
 
 hi 
 
 2oj 
 Q 3 
 
 
 g :: : : 
 
 Arrangement. 
 
 No. 
 Bolts. 
 
 tDe Paris a Lyon et 
 
 a la Meditei ranee 
 
 6194 
 
 4544 
 3631 
 
 3083 
 
 2445 
 
 2380 
 1812 
 
 2 inside and 2 
 outside 
 
 2 inside and 1 
 outside 
 
 1 inside and 1 
 outside 
 
 2 inside and 1 
 outside 
 
 2 inside and 2 
 outside 
 
 2 inside and 1 
 outside alter- 
 nating with 
 
 1 inside and 
 
 2 outside 
 
 2 inside and 1 
 
 97 lb. 
 T. 
 
 •B.H. 
 
 T. 
 ♦B.H. &T. 
 
 T. 
 T. 
 
 •B.H. 
 
 16^22 
 11.62 
 
 to 
 5.91 
 
 5.71 
 
 4.72 
 5^52 
 
 4.73 
 
 4.72 
 to 
 5.32 
 
 0.79 
 0.87 
 
 0.91 
 0.91 
 
 0.55 
 
 0.59 
 
 0.54 
 
 0.62 
 
 0.65 
 0.67 
 
 0.49 
 0.39 
 
 0.49 
 
 0.49 
 
 0.31 
 
 12 to 14 per 36' 
 
 17 per 39.37' 
 
 25 per 59' 
 
 14 to 16 per 36' 
 
 Square and 
 suspended 
 
 Square and 
 suspended 
 
 Square and 
 suspended 
 
 Square and 
 suspended 
 
 Square and 
 suspended 
 
 6 
 
 De L'Ouest 
 
 
 
 
 
 
 
 
 
 
 * B.H. means Bull Head. 
 
 t Mr. Cartault. As-istani Chief Engineer 
 often surpasses 15,432 to 17,637 pounds. 
 
 The above are all the important French n 
 
 i>ak or beech, the 
 
 i. extraction reaches ai 
 
 TABLE XXIX. — GERMAN RAILWAYS — RAIL FASTENINGS 
 
 (Am. Ry. Eng. Assn.) 
 
 
 Kind of Tie 
 Plate. 
 
 Number and 
 
 Screw Spikes 
 
 at Each End 
 
 of Tie. 
 
 Type 
 
 WciL'hl 
 of Rail 
 
 Screw Spikes. 
 
 Number o 
 
 Tie- !"sed 
 
 Joints. 
 
 
 Name of State 
 Railway. 
 
 
 
 ids 
 
 Q 3 
 
 III 
 
 '--- 
 
 Arrange- 
 
 No. 
 of 
 Bolts. 
 
 Wiirttemberg 
 
 Hook plate. 
 
 Hook outside 
 Hook plate. 
 
 Hook inside 
 
 At joints. 
 Hook plate. 
 Hook inside 
 At intermedi- 
 
 outside 
 2 inside and 1 
 outside 
 
 2 inside and 1 
 outside 
 
 Hook spikes 
 1 insifle and 
 2 outside 
 
 T. 
 
 92.74 
 T. 
 
 14.21 
 16.15 
 
 5.12 
 and 
 5.91 
 
 5.32 
 6.69 
 
 0.79 
 0.79 
 
 0.59 
 0.59 
 
 0.39 
 
 
 Suspended 
 Suspended 
 
 6 
 
 
 
 
 
 
 
 
 
 Hook plate. 
 Hook inside 
 
 2 inside and 1 
 outside 
 
 87.7 
 T. 
 
 16.58 
 
 6.5 
 
 0.79 
 
 0.59 
 
 0.39 
 
 
 Suspended 
 
 
 
 
 
 Baden (2) 
 
 
 Use hook an 
 
 73 T. 
 
 atcsancl 
 
 screw 
 
 
 fBava 
 
 •ianSt 
 
 te Rys. 
 
 Suspended 
 
 4 and 6 
 
 Elsass-Lothringen 
 
 Hook plate. 
 Hook outside 
 
 1 inside and 1 
 outside 
 
 91 T. 
 
 16.54 
 
 5.91 
 
 0.87 
 
 0.65 
 
 0.39 
 
 23 and 24 
 
 Square and 
 suspended 
 
 6 
 
 Prussian (3) 
 
 Hook plate. 
 Hook outside 
 
 2 outside and 
 1 inside 
 
 91 T. 
 
 16.54 
 
 5.91 
 
 0.87 
 
 0.65 
 
 0.39 
 
 19 and 24 
 per 49.2' 
 
 Suspended 
 
 6 
 
 (1) Use screw spikes on main tracks, and hook spikes on secondary tracks. 
 
 (2) Use steel ties almost exclusively. Use wooden ties on bridges with steel floor beams, in tunnels and for insulated joints in 
 electric signal districts. 
 
 (3) In tunnels use cast-iron chairs, wooden wedges, spikes, and trenails identical with Midland Ry., England. 
 " Hook spikes are now only used on lines of minor importance." 
 
SUPPORTS OF THE RAIL 
 
 145 
 
 The German railways did not adopt this style of fastening as early or as 
 generally as those of France, and the use of the hook spike is quite widespread. 
 In 1899, the general employment of the screw spike on all lines of the system 
 was prescribed for the Prussian Government Railways (Fig. 113, Table XXIX). 
 
 Fig. 113. — German Railways — Rail Fastenings. (Am. Ry. Eng. Assn.) 
 
 Same design of Screw Spike \ 
 
 I by the Wurttemberg, Bavarian and Baden State Railways as is 
 shown for the Saxon. 
 
146 
 
 STEEL RAILS 
 
 
 >L 
 
 The common hook spike used in the United States has been often severely- 
 condemned by writers in the technical press, and the readers have been usually 
 led to infer that it is employed everywhere in Europe, which is seen from the 
 above not to be the case. Indeed, the screw spike in Great Britain is almost 
 as rare as it is in the United States, at least on the large systems, the only one 
 
SUPPORTS OF THE RAIL 
 
 making use of it being the London and North Western, and that only partially 
 (Fig. 114, Table XXX and Plate XIII). 
 
 All of the large railway systems in Great Britain use the double-head rail, 
 held in position in large cast-iron chairs by wedges, and consequently the fasten- 
 ings are for securing the chairs to the ties. For the purpose of fastening the 
 chairs to the ties, the almost universal plan is to use two iron or steel spikes and 
 two wooden trenails. The spikes are not pointed, and are driven into previously 
 bored holes. Instead of the trenails, the London and North Western Railway 
 makes use of two screw spikes, which resemble those of the Belgian State 
 Railways. 
 
 TABLE XXX.— ENGLISH AND SCOTCH RAILWAYS — 
 
 (Am. Ry. Eng. Assn.) 
 
 RAIL FASTENINGS 
 
 
 Number and Kind of 
 Fastenings per Chair. 
 
 Spikes. 
 
 Trenails. 
 
 Railway. 
 
 Length 
 Head. 
 
 Head. 
 
 'of 
 Shank. 
 
 Total 
 Length. 
 
 Length 
 
 of 
 
 Top Cone 
 
 Top 
 Diam. 
 
 Inches. 
 
 Diam. 
 
 Point. 
 
 Lancashire & York- 
 Great Eastern 
 
 2 spikes and 2 trenails 
 2 spikes and 2 trenails 
 
 *2 spikes and 2 trenails 
 *2 spikes and 2 trenails 
 *2 spikes and 2 trenails 
 J|2 twisted spikes and 2 trenails 
 2i-inch bolts through ties 
 
 TT2 spikes and 2 screws 
 
 t6i 
 1 
 
 Inches. 
 1 
 
 Inches. 
 1 
 
 §61 
 
 Inches. 
 2 
 
 Inches. 
 
 Inches. 
 
 
 
 
 
 Great Northern 
 
 .ii • n-inn- in' hi ln"l 
 
 North British 
 
 J6 
 6 
 
 1A 
 
 1 
 
 ft 
 
 2 
 2 
 2ft 
 
 if 
 if 
 
 if 
 
 a 
 
 North Eastern 
 
 li 3 o 
 
 London & North 
 
 t6 
 
 11 
 
 \i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * Has short mileusre of chairs secured by two f-inch bolts thro 
 
 
 t Illustrated by Fig. 1 of Fig. 114. 
 
 
 t Illustrated by Fig. 2 of Fig. 114. 
 
 
 § Illustrated by Fig. 3 of Fig. 114. 
 
 
 11 Illustrated by Fig. 4 of Fig. 114. 
 
 
 On the Forth Bridge the North British Ry. uses flat-botton 
 
 
 II The twisted spikes are to be abandoned for plain ones. 
 
 
 ;ened to longitudinal beams by wood sc 
 
 Diameter of hole required for this ecrew fi" 
 
 - Screw Spike deduced from European Practice. (Cushing.) 
 
STEEL RAILS 
 
 Fig. 115 shows a design of screw spike deduced by Mr. Gushing from 
 European practice. The form of the thread seems to have little influence upon 
 the holding power of the screw spike. Table XXXI gives the resistance for 
 threads of right-angle form and those of isosceles triangular form. 
 
 TABLE XXXI 
 RESISTANCE OF SCREW SPIKES HAVING DIFFERENT THREADS 
 (Jules Michel — Revue Generate des Chemins de Fer, June, 1893) 
 
 Dimensions of Screw Spikes. 
 
 Isosceles Thread. Right-angled Thread. 
 
 Remarks. 
 
 Screw Spikes 0.79 inch in diameter. New tie. 
 
 Pitch of 0.39 inch 
 Pitch of 0.59 inch 
 Pitch of 0.49 inch 
 
 11,687 pounds 
 12,458 pounds 
 13,010 pounds 
 
 12,039 pounds 
 12,513 pounds 
 13,561 pounds 
 
 Average of 4 trials. 
 Average of 4 trials. 
 Average of 4 trials. 
 
 Screw Spikes 0.91 inch in diameter. 
 
 Pitch of 0.49 inch 
 
 13,424 pounds 13,424 pounds 
 
 New tie. 
 
 Screw Spike 0.79 inch in diameter. 
 
 Pitch of 0.39 inch 
 Pitch of 0.49 inch 
 
 10,253 pounds 
 
 9,923 pounds 
 11,576 pounds 
 
 Ties 9 years in service. 
 Ties 9 years'in service. 
 
 Screw Spikes 0.91 inch in diameter. 
 
 Pitch of 0.49 inch 
 
 11,246 pounds 11,025 pounds 
 
 Ties 9 years in service. 
 
 The proof that the screw spike is not a thoroughly efficient rail fastening 
 lies in the devices which have been invented to assist it in its work, — the square 
 plug, the Collet trenail, the Thiollier helical lining, and the Lakhovsky screw 
 and case.* 
 
 The main objection to the Collet trenail is its size; it is illustrated in Figs. 
 116, 117, and 118 with a screw spike and the wooden plug commonly used on 
 French railways for repairing old holes. The difference in size is large, the 
 Collet trenail being If inches in diameter outside the threads. This cuts away 
 a considerable portion of the critical part of a tie, and is considered by many 
 engineers to weaken the tie too much. The plug is only about an inch square. 
 Nevertheless this screw dowel is largely used in Germany. 
 
 The Collet trenail has been tested from its inception by the Chemins de 
 Fer de l'Est, but the square plug illustrated in Fig. 117 is preferred. The wooden 
 screw, often made of elm, cannot be put in place without removing the tie from 
 
 * Lakhovsky trenail, Revue Generate des Chemins de Fer. Paris, 1909, Vol. XXXIII, 
 pp. 324-327. 
 
SUPPORTS OF THE RAIL 
 
 149 
 
 %"ScrswSpiKe 
 
 Chermnde Fe.r 
 de UEsf 
 
 Fig. 116. — French Screw Spike. 
 (Am. Ry. Eng. Assn.) 
 
 Square Wooden Plug 
 used for repairing old holes. 
 
 Fig. 117. — Wooden Tie Plug used on French Railways. 
 (Am. Ry. Eng. Assn.) 
 
150 
 
 STEEL RAILS 
 
 Col le' 
 
 "■ I r&no i f 
 
 _ 
 
 somefirr 
 
 les used 
 
 as 
 
 substitute for square 
 
 01 ug 
 
 the track, and it frequently splits. The ties on the Chemins de Fer de 1'Est 
 are principally oak and beech. Figs. 119 and 120 illustrate pine ties with 
 
 dowels in place. 
 
 The diagrams of Figs. 121 and 
 122 give the comparative resistance 
 to vertical pressure of screw spikes 
 with and without dowels. 
 
 The Thiollier steel helical lining 
 is being experimented with as a sub- 
 stitute for the Collet trenail, and the 
 Lakhovsky screw and steel casing 
 (Bulletin of the International Railway 
 Congress, March, 1907) are considered 
 worth trying by the Chemins de Fer 
 de l'Est, de 1'Etat, and de Paris a 
 Orleans. 
 
 From its greater holding power, 
 the verdict of the engineers of the 
 French, Belgian, and German railways 
 is that the screw spike is superior to 
 the hook spike, because they consider 
 it very important to hold the rail fast 
 to the tie. 
 
 On the other hand, the British 
 railways do not seem to find the screw 
 spike necessary for their large and 
 heavy chairs, and they use creosoted 
 ties, as well as the Continental lines; 
 but the holes for their spikes are bored 
 in advance. 
 
 According to our present knowl- 
 edge, the amount of bearing surface 
 the tie plate has upon the tie is ap- 
 parently not the determining factor 
 in providing against wear. The 
 question of securing the plate firmly to the tie is fully as important as the size 
 of the plate used, and in selecting a proper unit stress for the bearing on the 
 tie it is evident, therefore, that the area of the bearing surface cannot be con- 
 
 Fig. 118. — Collet Trenail. (Am. Ry. Eng. Assn.) 
 
SUPPORTS OF THE RAIL 151 
 
 sidered without taking account of the kind of fastening employed to hold the 
 plate to the tie. 
 
 Fig. 119. — Cross Section of Pine Tie through Dowel. (Bureau of Forestry, Bulletin No. 50.) 
 
 Fig. 120. — Three Ties of Baltic Pine on the Prussian State Railways, Berlin, showing the manner in 
 which screw dowels appear in the tie when ready to be shipped. (Bureau of Forestry, Bulletin No. 50.) 
 
STEEL RAILS 
 
 16000 - r 
 15000- 
 
 
 
 
 
 -D 
 
 
 
 
 
 
 
 
 p 
 
 11^ 
 
 JE 
 
 : i 
 
 "IE 
 
 s 
 
 
 
 
 
 
 
 
 
 
 NEW 
 
 
 
 13000- 
 12000- - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 II000-- 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 r > 
 
 '/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 >/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 *' 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 V 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 I) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 <>7 
 
 
 
 
 
 
 
 
 
 
 
 4000 - - 
 
 "4 
 
 
 & 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 £i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 , / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 a 
 
 5 
 
 OS 
 
 10 
 
 .1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 06 
 
 .08 
 
 ■1 
 
 3 
 
 12 
 
 14 
 
 .16 
 
 VERTICAL DISPLACEMENT - INCHES 
 
 - Comparative Resistance to Vertical Pressure of Screw Spikes in Pine Ties, Old and New, 
 with and without Dowels. (Bureau of Forestry, Bulletin No. 50.) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 *y 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 $/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 6/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■y 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 W 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 & 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 >/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 <■■/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 rf 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 < 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 // 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 // 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 // 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .02 
 
 .C 
 
 4 
 
 
 6 
 
 .C 
 
 8 
 
 
 
 
 
 2 
 
 
 4 
 
 
 6 
 
 
 
 
 .0 
 
 z 
 
 D 
 
 4 
 
 
 
 
 
 8 
 
 
 
 
 
 
 
 16 
 
 Fig. 122. — Comparative Resistance to Vertical Pressure of Screw Spikes in Beech Ties, Old and New, 
 with and without Dowels. (Bureau of Forestry, Bulletin No. 50.) 
 
SUPPORTS OF THE RAIL 153 
 
 In the case of a white-oak tie, where the spike holds well and the life of the 
 tie is comparatively short, the ordinary working stress of the timber to resist 
 crushing at right angles to the grain may probably be safely taken in propor- 
 tioning the strength of the tie. With soft woods, however, which offer less 
 resistance to the spike pulling loose, and which, when treated, possess long life, 
 the ordinary working stress of the wood has little application to the bearing under 
 the tie plate unless some means are used to secure the plate firmly to the tie. 
 
 As will be seen in the discussion of the Supporting Power of the Tie (Article 
 19), one of the weakest points in the support of the rail lies at the bearing of 
 the tie plates on soft-wood ties, even when the normal crushing value of the 
 wood is taken as is done in the calculations. It is thus of considerable impor- 
 tance that with a soft-wood tie a more secure fastening than the ordinary spike 
 be used to hold the tie plate firmly to the tie. 
 
 With the increase in density of traffic there has developed a growing 
 tendency for the rail to creep or move in the direction in which the traffic 
 moves. On account of the joint ties being spiked through slotted holes in the 
 joint, these ties move with the rail, with the result that correct spacing of the 
 adjacent ties is not maintained. 
 
 To overcome this difficulty there have been devised numerous devices 
 for anchoring the rails to the ties. These are generally fastened to the base 
 of the rail and bear against the side of the tie; when employed in sufficient 
 numbers they are fairly efficient in preventing the movement of the rail* 
 
 16. Strength of the Tie 
 Assuming the tie to be in good condition and free from decay, we have 
 now to determine the strength of the wood of which it is composed. Let us 
 first examine the kinds of woods used in the United States. 
 
 * Some recent literature on this subject is as follows: 
 
 Kunze, W. — Das schienenwandern, ursache und abhilfe. 2,500 w. 111. 1909. (In Glasers 
 annalen fur gewerbe und bauwesen, Vol. 65, p. 122.) 
 
 Considers cause of creeping in rails and devices for its prevention. 
 
 Schldssel, L. — On the working loose of screws when used as rail fastenings, 21 p. 111. 
 1907. (In Bulletin of the International Railway Congress, Vol. 21, p. 3.) 
 
 Concludes that wedge fastenings should be substituted for screw fastenings. 
 
 Tex, K. den. — Creeping of rails in the direction of the trains. 800 w. 111. 1911. (In 
 Bulletin of the International Railway Congress, Vol. 25, p. 292.) 
 
 Use of rail anchors. 2,000 w. 111. 1911. (In Railway Age Gazette, Vol. 51, p. 125.) 
 
 Considers tendencies in the creeping of rails and forms of anchors most successful in over- 
 coming it. 
 
 Wirth, Alfred. — Die schienenwanderung und ihre verhiitung. 10,000 w. 1909.- (In Zeit- 
 schrift des Osterreichischen Ingenieur — und Architekten — • Vereines, Vol. 61, p. 317, 333.) 
 
 Discussion of rail creeping at some length, considering theory and prevention by rail-fastening 
 devices. 
 
154 
 
 STEEL RAILS 
 
 The following statements are based on the number of ties bought rather 
 than on the number actually used. For all practical purposes, however, the 
 two are identical, because the purchases in twelve months are an accurate index 
 of consumption for a corresponding period. 
 
 Table XXXII shows the number and value of the different kinds of ties 
 purchased by the steam and street railroad lines in the United States in 1906, 
 and contrasts the purchases of steam railroad companies in 1905 and 1906. 
 
 TABLE XXXII. — NUMBER AND VALUE OF TIES PURCHASED BY STEAM AND 
 
 STREET RAILROADS IN THE UNITED STATES IN 1905 AND 1906 
 
 (Forest Service, Circular 124) 
 
 
 Steam railroads, 1905. 
 
 Steam railroads, 1906. 
 
 Street railroads, 1906.* 
 
 
 Number. 
 
 «. 
 
 Aver- 
 YaTue 
 !fe. 
 
 Number. 
 
 Value. 
 
 Value 
 per 
 Tie. 
 
 Number. 
 
 Value. 
 
 Aver- 
 
 Oaks 
 
 34,677,304 
 18,351,037 
 6,962,827 
 3,633,276 
 4,717,604 
 3,483,746 
 
 (t) 
 
 3,060,082 
 
 1,713,090 
 
 590,852 
 
 (tt 
 
 (t) 
 
 791,409 
 
 $19,072,517 
 7,707,436 
 3,083,644 
 1,198,981 
 2,264,450 
 1,149,636 
 
 $0.55 
 .42 
 .44 
 .33 
 .48 
 .33 
 
 41,532,629 
 
 17,538,090 
 
 6,416,867 
 
 6,706,222 
 
 4,646,763 
 
 4,988,585 
 
 3,909,500 
 
 2,430,236 
 
 2,037,002 
 
 725,346 
 
 553,838 
 
 258,030 
 
 1,734,517 
 
 $21,256,518 
 
 8,905,009 
 
 3,044,446 
 
 2,782,967 
 
 2,132,984 
 
 1,813,500 
 
 1,673,359 
 
 837,217 
 
 576,896 
 
 248,844 
 
 210,458 
 
 76,833 
 
 661,501 
 
 $0.51 
 .51 
 .47 
 .41 
 .46 
 .36 
 .43 
 .34 
 .28 
 .34 
 .38 
 .30 
 .38 
 
 3,825,245 
 
 1,303,120 
 
 666,575 
 
 542,340 
 
 1,942,212 
 
 115,911 
 
 60,105 
 
 146,623 
 
 21,196 
 
 523,283 
 
 900 
 
 115,357 
 
 93,550 
 
 $2,021,534 
 662,736 
 265,670 
 227,425 
 862,958 
 48,635 
 24,668 
 52,344 
 6,072 
 287,328 
 360 
 74,219 
 64,643 
 
 $0.53 
 
 South'rnpinest 
 
 Cedar 
 
 Douglas fir. . 
 
 Chestnut 
 
 Cypress 
 
 Western pine. 
 Tamarack. . . . 
 
 Hemlock 
 
 Redwood 
 
 Lodgepole pine 
 White pine . 
 All others. . . . 
 
 .51 
 .40 
 .42 
 
 .44 
 .42 
 
 1,101,630 
 565,320 
 118,170 
 
 .36 
 .33 
 
 .20 
 
 .36 
 
 .29 
 .55 
 
 
 
 64 
 
 343,662 
 
 .43 
 
 69 
 
 Total. . 
 
 77,981,227 
 
 $36,585,446 
 
 $0.47 
 
 93,477,625 
 
 $44,220,532 
 
 $0.47 
 
 9,356,417 
 
 $4,598,592 
 
 $0 49 
 
 T I' mi I'MU includes \vl 
 t Included in southern 
 
 >aas in iwua. 
 
 ,ne, lodgepole pine, and western pine. 
 
 The purchases of ties reported by the steam railroad companies in 1906 
 exceeded those of 1905 by more than 15,000,000. Nearly one-half of this 
 excess was oak. The purchases of cedar ties showed a decrease of about one- 
 half million, due possibly to the sharp demand for cedar poles, which operated 
 against the production of ties. Douglas fir ties nearly doubled in quantity, 
 and both cypress and hemlock increased by a large percentage, but tamarack 
 purchases fell off more than one-fifth and chestnut about 1.5 per cent. 
 
 Oak, the chief wood used for ties, furnishes more than 44 per cent, nearly 
 one-half of the whole number, while the southern pines, which rank second, 
 contribute about one-sixth. Douglas fir and cedar, the next two, with approxi- 
 mately equal quantities, supply less than one-fifteenth apiece. Chestnut, 
 cypress, western pine, tamarack, hemlock, and redwood are all of importance, 
 but no one of them furnishes more than a small proportion. 
 
SUPPORTS OF THE RAIL 155 
 
 Table XXXIII shows, by kinds, the number and cost of the cross-ties pur- 
 chased by steam and electric railroads in the United States in 1907. 
 
 Table XXXIV gives a comparative statement showing the number of 
 cross-ties purchased by the steam and electric railroads during the years 1910, 
 
 1909, 1908, and 1907. 
 
 Of the total purchases of cross-ties during 1910, 139,596,000, or 94.2 per 
 cent, were made by steam railroads, while electric railroads purchased 8,635,000, 
 or 5.8 per cent. The steady increase in the number of cross-ties reported as 
 purchased for new track is noteworthy. The total for this purpose in 1910 was 
 22,255,000, as against 16,437,000 in 1909, 7,431,000 in 1908, and 23,557,000 in 
 1907; the total for 1910 exceeding that for 1909 by 35.4 per cent, for 1908 by 
 199.5 per cent, and nearly equaling that for 1907, the largest ever recorded. 
 Largely as a logical result of the greater demand for cross-ties during 1910, the 
 average cost per tie at point of purchase advanced to 51 cents, the same figure 
 reached in 1907, as compared with 49 cents in 1909 and 50 cents in 1908. 
 
 In 1910, as in preceding years, oak was the principal kind of wood used for 
 cross-ties. The number of oak cross-ties formed 46.1 per cent of the total for 
 
 1910, as compared with 46.2 per cent in 1909, 42.8 per cent in 1908, and 40.2 
 per cent in 1907. 
 
 A substantial increase in 1910 over 1909 is shown in the number of southern 
 pine cross-ties reported; the increase in the cut from this species over 1909 
 being 22.8 per cent, as against an increase of 20 per cent in the total number of 
 cross-ties reported from all woods. Douglas fir also showed for 1910 over the 
 preceding year a larger increase, namely, 28.2 per cent, than the increase in the 
 total purchase from all woods. On the other hand, chestnut, cedar, and cypress, 
 with increases over 1909 of 17.1 per cent, 7.8 per cent, and 17.6 per cent, 
 respectively, were bought in relatively smaller quantities. 
 
 While the bulk of the cross-ties were cut from the six woods mentioned dur- 
 ing each of the four years and while combined they contributed 85.5 per cent of 
 the total in 1910, 85.3 per cent in 1909, 86.5 per cent in 1908, and 87.2 per cent 
 in 1907, a remarkable and significant showing in connection with the figures for 
 1910 is noted with reference to certain woods which hitherto have been utilized 
 as cross-tie material to only a very limited extent. The increase in the number 
 of cross-ties over 1909, reported as cut from elm, was 451.7 per cent; gum, 328.8 
 percent; birch, 323.3 per cent; spruce, 121.5 per cent; and mesquite, 114.9 per 
 cent. A very large percentage of the cross-ties cut from these woods were given 
 some preservative treatment, thus increasing their life to or beyond that of 
 untreated cross-ties made from the more commonly used or standard cross-tie 
 
156 
 
 STEEL RAILS 
 
 TABLE XXXIII. — CROSS-TIES PURCHASED BY STEAM AND ELECTRIC ROADS OF 
 THE UNITED STATES IN 1907 
 
 (Bureau of the Census, Forest Products No. 8) 
 
 
 Total. 
 
 
 Steam Railroads. 
 
 
 
 Kind. 
 
 Hewed. 
 
 Sawed. 
 
 Number. 
 
 Total Cost. 
 
 cSl 
 
 per 
 
 Tie. 
 
 Number. 
 
 Total Cost. 
 
 Aver- 
 
 Cost 
 T P fe. 
 
 Number. 
 
 Total Cost. 
 
 Aver- 
 age 
 
 Cost 
 
 Tie. 
 
 Total... 153,699,620 
 
 $78,958,695 
 
 $0.51 
 
 112,309,246 
 
 $56,522,768 
 
 $0.50 
 
 31,776,434 
 
 $17,020,882 
 
 $0.54 
 
 Oaks 
 
 61,757,418 
 34,215,081 
 14,524,266 
 8,953,205 
 7,851,325 
 
 6,778,944 
 5,019,247 
 4,562,190 
 2,366,459 
 
 2,030,982 
 666,916 
 474,455 
 
 4,499,132 
 
 32,985,122 
 18,434,198 
 6,818,869 
 4,473,960 
 3,772,048 
 
 3,099,439 
 
 2,515,798 
 
 2,254,617 
 
 807,241 
 
 1,198,497 
 332,984 
 193,606 
 
 2,072,316 
 
 0.53 
 0.54 
 0.47 
 0.50 
 0.48 
 
 0.46 
 0.50 
 0.49 
 0.34 
 
 0.59 
 0.50 
 0.41 
 0.46 
 
 51,169,478 
 25,629,749 
 1,436,258 
 7,941,152 
 4,922,831 
 
 5,695,640 
 3,206,754 
 4,144,127 
 2,283,675 
 
 884,552 
 
 666,916 
 
 289,624 
 
 4,038,490 
 
 26,774,251 
 13,100,589 
 590,754 
 3,987,035 
 2,337,697 
 
 2,552,381 
 
 1.576,457 
 
 2,083,646 
 
 770,969 
 
 507,154 
 
 332,984 
 
 106,528 
 
 1,802,323 
 
 0.52 
 0.51 
 0.41 
 0.50 
 0.47 
 
 0.45 
 0.49 
 0.50 
 0.34 
 
 0.57 
 0.50 
 0.37 
 0.44 
 
 6,929,572 
 
 7,415,686 
 
 12,366,640 
 
 396,891 
 
 889,420 
 
 884,915 
 
 1,626,330 
 
 340,618 
 
 79,256 
 
 406,519 
 
 4,033,150 
 
 4,569,060 
 
 5,884,822 
 
 190,322 
 
 426,523 
 
 453,058 
 835,895 
 137,481 
 34,796 
 
 224,525 
 
 0.58 
 
 Southern pines 
 Douglas fir. . . 
 
 Cedar 
 
 Chestnut 
 
 Cypress 
 
 Western pine. . 
 Tamarack. . . . 
 Hemlock 
 
 Redwood 
 
 Lodgepole pine 
 
 White pine 
 
 All other 
 
 0.62 
 0.48 
 0.48 
 0.48 
 
 0.51 
 0.51 
 0.40 
 0.44 
 
 0.55 
 
 131,671 
 308,916 
 
 53,041 
 178,209 
 
 0.40 
 0.58 
 
 TABLE XXXIV.— CROSS-TIES PURCHASED BY STEAM AND ELECTRIC ROADS OF 
 THE UNITED STATES DURING THE YEARS 1910, 1909, 1908, AND 1907 
 
 (Bureau of the Census) 
 
 
 1910 
 
 1909 
 
 1908 
 
 19J7 
 
 Kind of Wood. 
 
 Number. 
 
 Cost at 
 
 Number. 
 
 Cost at 
 point of N 
 
 imber. 
 
 Cost at 
 
 purchase. 
 
 Number. 
 
 Cost at 
 purchase. 
 
 Total 
 
 148,231,000 
 
 875,889,000 
 
 123,751,090 
 
 860,321,000 112 
 
 467,000 
 
 S56,282,000 
 
 153,703,000 
 
 S78,959,000 
 
 Oak 
 
 68,382,000 
 
 2o.jh4.ooo 
 
 11. i>2:i, Olio 
 
 7,760.1)111) 
 
 7,:joj,oiiii 
 
 .3,3.10,000 
 
 3,103,000 
 
 4,iil2,ooo 
 
 3,40s, 000 
 
 2,10.3.01)0 
 
 1,621,000 
 
 798,000 
 
 773,000 
 
 548,000 
 
 499,000 
 
 429,000 
 
 393,000 
 
 238,000 
 
 178,000 
 
 134,000 
 
 476,000 
 
 37,731,000 
 13. :i30,ooo 
 
 .3,317, 
 
 3,716,000 
 
 3,430,000 
 
 2,390,000 
 
 2,U7«,ooi> 
 
 2,058,000 
 
 1,063,000 
 
 1,262,000 
 
 754,000 
 
 351,000 
 
 346,000 
 
 220,000 
 
 251,000 
 
 201,000 
 
 170,000 
 
 121,000 
 
 156,000 
 
 71,000 
 
 266,000 
 
 57,132,000 
 
 2I,3.\.3. I 
 
 9,067.000 
 0,020,001) 
 6,777,000 
 4,589,000 
 3,311,000 
 (i, 7 l .)7,ooii 
 2,012,000 
 
 2.OS.S.0OO 
 
 378,000 
 195,000 
 158,000 
 
 99,000 
 225,000 
 556,000 
 
 92,000 
 487,000 
 120,000 
 
 62,000 
 962,000 
 
 29,062,000 48 
 11,112,000 21 
 3,754,000 7 
 2,947.000 8 
 • 3,085,000 8 
 1,902,000 3 
 1,356,000 4 
 3,619,000 3 
 865,000 3 
 1,108.000 
 198,000 
 69,000 
 55,000 
 46,000 
 109,000 
 237,000 
 32,000 
 224,000 
 117,000 . . 
 39,000 
 385,000 1 
 
 110,000 
 
 ;,.,o, oiiii 
 
 9X.S.IIIIII 
 
 11,-4.000 
 172.000 
 457,000 
 
 023,0011 
 OH3, 
 
 120,0110 
 
 871,000 
 
 202,0110 
 
 102.01)11 
 
 131,000 
 05,ooo 
 
 111,000 
 
 707, 
 
 11 1,000 
 
 ol.s.ooo 
 
 ' 31.000 
 846,000 
 
 24,653,000 
 
 11,50!), 000 
 
 3,500,000 
 
 3,0.s2,ooo 
 
 4,028,000 
 
 1,320,000 
 
 2,010,000 
 
 1,573,000 
 
 1,179,000 
 
 444,000 
 
 117,000 
 
 86,000 
 
 68,000 
 
 24,000 
 
 66,000 
 
 335,000 
 
 38,000 
 
 247,000 
 
 61,757,000 
 
 31,21,3,000 
 11.53.3,01)0 
 7,S51,(HI0 
 ,s, 05 l.i ii in 
 6,780,000 
 4,502,000 
 5,1119, 000 
 2,307,000 
 2,032,000 
 15,000 
 52,000 
 
 32,9S4,000 
 
 
 
 
 
 Cedar 
 
 4,474,000 
 
 
 
 We-tern pine 
 
 2,516,000 
 
 
 
 
 
 
 
 
 
 
 
 
 
 104,000 
 475,000 
 
 
 
 
 
 
 Lod.iex.le pine 
 
 667,000 
 
 333,000 
 
 
 21,000 
 684,000 
 
 
 
 
 4,328,000 
 
 
 
 
SUPPORTS OF THE RAIL 157 
 
 woods. The growing scarcity of these last-mentioned woods, however, tends 
 to increase their cost and accounts largely for the introduction of substitutes 
 cut from cheaper species. The drift in this direction is clearly brought out by 
 a comparison of the figures relating to treated cross-ties during the past four 
 years. In 1907 the number of cross-ties reported as having been given some 
 preservative treatment was 19,856,000; in 1908, 23,776,000; in 1909, 22,033,00; 
 and in 1910, 30,544,000; the number for 1910 showing an increase over that for 
 the preceding year of 8,511,000, or nearly 39 per cent. 
 
 The question of tie preservation is becoming more and more important as 
 the demand for tie material increases and the traffic requirements become more 
 exacting. So long as plenty of white-oak ties could be secured, the necessity 
 for tie preservation was not felt; but with the constantly increasing use of pine 
 and other less decay-resistant woods, it has become a vital economic question. 
 The railroad companies have met the problem by establishing treating plants 
 in various parts of the United States and by laying experimental tracks with 
 treated ties to determine the efficiency of the several preservatives under vary- 
 ing conditions.* 
 
 Table XXXV, prepared by the Forest Service, f gives the results of an 
 elaborate series of tests upon the strength of treated and untreated pine ties. 
 
 In outlining the plan for these tests two divisions were made, dealing 
 respectively with the effect on the strength of timber of the preliminary proc- 
 esses of steaming, superheating, vacuum, etc., commonly employed in the 
 preservation of wood, and the effect of the preserving materials themselves. 
 The tests were confined to sapwood, and were made on small pieces taken from 
 the tie, and also on full-sized ties. 
 
 The effect of the preliminary processes was determined on both green and 
 seasoned timber. ' Both green and seasoned timber were also used in deter- 
 mining the effect of preservatives. The preservative fluids included only 
 creosote t and zinc chloride. 
 
 The material for the experiments was railroad ties 11 feet long. One 
 8-foot section of each tie was put through the particular treatment, and the 
 untreated section, 3 feet long, was used for control test pieces. 
 
 From each tie 12 pieces were taken, 4 from the control section and 8 from 
 the treated section. All of these pieces were 2 inches by 2 inches in cross section 
 and 36 inches long, with one side parallel to the direction of the annual rings 
 
 * Experiments with Railway Cross-ties, Forest Service, Circular 146. 
 
 t Experiments on the Strength of Treated Timber, Forest Service, Circular 39, by W. K. Hatt. 
 X The treatment was essentially the "Rueping" process, although this name is not used in tha 
 circular. 
 
158 
 
 STEEL RAILS 
 
 TABLE XXXV. — EXPERIMENTS ON THE STRENGTH OF TREATED TIMBER 
 
 (Forest Service, Circular 39) 
 
 EFFECT OF STEAMING AND PRESERVATIVE TREATMENTS ON THE STRENGTH OF GREEN 
 
 LOBLOLLY PINE 
 
 [Specimens, 2 by 2 inches; air-dried before tests! 
 
 
 Cylinder Conditions. 
 
 Strength. 
 
 per Inch. 
 
 Moisture. 
 
 Specific 
 Gravity 
 (dry). 
 
 
 Steaming. 
 
 h 
 
 Static. 
 
 Im- 
 
 8 
 
 
 
 
 
 
 
 
 
 
 
 § 
 
 
 «a 
 
 
 
 ■6 
 
 I 
 
 i 
 
 I 
 
 2 
 
 a 
 
 1" 
 < 
 
 S3 
 
 H 
 
 fl 
 6 
 
 --, =- 
 1 l fe 
 
 f 
 
 6 
 
 & 
 
 2 
 
 6 
 
 2 
 3 
 H 
 
 6 
 
 1 
 
 
 Hrs. 
 
 Lbs. 
 sq. in. 
 
 °F. 
 
 Lbs. 
 per 
 eu.it. 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 
 
 Per 
 
 Per 
 
 
 
 
 
 
 
 
 Untreated wood =100 percent. 
 
 
 
 
 
 
 
 
 f 4 
 
 20 
 
 25V 
 
 
 92.81 93.11 93.81 93.2 
 
 7.5 
 
 6.5 
 
 13.8 
 
 13.4 
 
 0.558 
 
 J. 646 
 
 Steam, at various pressures 
 
 4 
 
 HO 
 
 -m 
 
 
 99.8104.3 102.0 102.0 
 
 7.5 
 
 7.0 
 
 12.7 
 
 12.2 
 
 .553 
 
 .571 
 
 ^4 
 
 40 
 
 2i n 
 
 
 94.6 99.0107.7100.4 
 
 6.0 
 
 6.0 
 
 13.6 
 
 12.4 
 
 .525 
 
 .534 
 
 
 14 
 
 50 
 
 296 
 
 
 94.51 96.4|l03.5l 98.1 
 
 6.5 
 
 6.5 
 
 13.7 
 
 12.5 
 
 .514 
 
 • 508 
 
 Creosote, injectedat 150°F. 
 
 
 
 
 
 Steamed wood = 100 per cent. 
 
 
 
 
 
 
 
 under a pressure of 100 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 pounds per square inch. . . 
 
 4 
 
 20 
 
 258 
 
 25.4 
 
 81.6 
 
 79.9 
 
 102.4 
 
 88.0 
 
 6.5 
 
 6.0 
 
 13.2 
 
 
 .663 
 
 
 Zinc chloride: 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2.5 per cent solution. . . 
 
 4 
 
 20 
 
 249 
 
 
 87.4 
 
 92.8 
 
 113.8 
 
 98.0 
 
 6.5 
 
 6.0 
 
 12.8 
 
 13.7 
 
 .530 
 
 .534 
 
 3.5 per cent solution. . . 
 
 4 
 
 20 
 
 246 
 
 
 97.4 
 
 95.2 
 
 92.7 
 
 95.1 
 
 7.5 
 
 7.0 
 
 13.1 
 
 13.5 
 
 .538 
 
 .539 
 
 5.0 per cent solution. . . 
 
 4 
 
 20 
 
 24(i 
 
 
 99.8 
 
 96.7 
 
 78.9 
 
 91.8 
 
 5.5 
 
 6.0 
 
 12.3 
 
 13.4 
 
 .510 
 
 .506 
 
 10.0 per cent solution. . . 
 
 4 
 
 20 
 
 255 
 
 
 100.1 
 
 100.4 
 
 74.8 
 
 91.8 
 
 8.5 
 
 8.5 
 
 13.2 
 
 13.5 
 
 .582 
 
 .612 
 
 Physical Characteristics and Average Strengths of the Air-dried Untreated Wood 
 
 Moisture per cent 
 
 Weight per cubic foot (dry) pounds 
 
 Rings per inch 
 
 Modulus of elasticity pounds per square inch 1,1 
 
 Bending strength at elastic limit 
 
 Bending strength at rupture • 
 
 Compression strength parallel to grain 
 
 Compression strength at right angles to grain 
 
 Shearing strength radial to grain 
 
 13.4 
 33.6 
 
SUPPORTS OF THE RAIL 
 
 159 
 
 TABLE XXXV. — Continued 
 
 
 [Specimens 
 
 full-sized ties; 
 
 seasonec 
 
 treated 
 
 and reseasoned before tests] 
 
 
 
 
 
 
 Cylinder Conditions. 
 
 Strength (static). 
 
 Spike Pulling. 
 
 
 
 
 
 
 
 
 8 
 
 Force Re- 
 
 § 
 
 
 
 Steaming. 
 
 Bending 
 
 Compression. 
 
 
 
 g 
 
 
 Treatment. 
 
 
 
 
 
 H „; 
 
 Pull Spike. 
 
 8. 
 
 i 
 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 I 
 
 1 
 
 £ 
 
 1 
 
 1* 
 
 3 . 
 
 - 2 
 ■So 
 (33 
 
 < 
 
 < 
 
 1 
 
 a 
 a 
 6 
 
 s 
 
 •s 
 
 
 
 Hrs. 
 
 Lbs. 
 per 
 sq. in. 
 
 °F. 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 
 Lbs. 
 
 cu. ft. 
 
 
 
 
 
 
 Untreated wood = 100 per 
 
 cent. 
 
 
 
 
 
 f 4 
 
 10 
 
 237 
 
 99.2 
 
 79.3 
 
 91.1 
 
 89.9 
 
 118.5 
 
 110.7 
 
 4.9 38.0 
 
 
 
 1 4 
 
 20 
 
 258 
 
 93.7 
 
 78.4 
 
 99.1 
 
 90.4 
 
 103.6 
 
 109.4 
 
 5.2 37.3 
 
 Steam, at 
 
 various pressures — 
 
 i 4 
 
 30 
 
 274 
 
 87.8 
 
 83.4 
 
 92.7 
 
 88.0 
 
 100.1 
 
 96.5 
 
 5.3 37.9 
 
 
 
 It 
 
 40 
 
 286 
 
 88.4 
 
 78.1 
 
 74.6 
 
 80.4 
 
 93.0 
 
 77.9 
 
 5.2! 37.8 
 
 
 
 50 
 
 295 
 
 69.1 
 
 60.6 
 
 74.4 
 
 68-0 
 
 80.4 
 
 70.3 
 
 4.8 1 36.1 
 
 
 
 r 2 
 
 20 
 
 257 
 
 82.4 
 
 81-9 
 
 87.1 
 
 83.8 
 
 97.9 
 
 93.7 
 
 5.1 38.1 
 
 Steam, fo 
 
 various periods 
 
 1 4 
 
 20 
 
 258 
 
 93.7 
 
 78.4 
 
 99.1 
 
 90.4 
 
 103.6 
 
 109.4 
 
 5 ? 
 
 37.3 
 
 1 6 
 
 20 
 
 256 
 
 87.5 
 
 78-8 
 
 92.0 
 
 86.1 
 
 83.0 
 
 79.0 
 
 4.6 
 
 36.7 
 
 
 
 110 
 
 20 
 
 256 
 
 77.0 
 
 75.5 
 
 73.2 
 
 75.2 
 
 84.1 
 
 76.8 
 
 4.8 
 
 36.7 
 
 Zinc chloi 
 
 ide, 2.5 per cent solu- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 20 
 
 258 
 
 74.7 
 
 65.1 
 
 68.6 
 
 69.5 
 
 75.3 
 
 73.8 
 
 4.3 
 
 41.5 
 
 Creosote, 
 
 28 pounds per cubic 
 
 
 
 
 4 
 
 20 
 
 257 
 
 69.5 
 
 61.2 
 
 60.1 
 
 63.6 
 
 68.2 
 
 68.1 
 
 4.6 
 
 65.3 
 
 
 
 Physical Characteristics and Average Strengths c 
 
 : Untreated Wood 
 
 Moisture '. per cent (approximate) 
 
 Weight per cubic foot (air-seasoned) pounds 
 
 Rings per inch 
 
 Modulus of elasticity pounds per square inch 
 
 Bending strength at elastic limit 
 
 Bending strength at rupture 
 
 Compression strength parallel to grain 
 
 Compression strength at right angles to grain (rail-bearing) " 
 
 Spike pulling — common spike 
 
 Spike pulling — screw spike 
 
 1,568,000 
 3,429 
 6,458 
 4,452 
 503 
 3,598 
 7,748 
 
TABLE XXXV. — Continued 
 
 EFFECT OF STEAM AND CREOSOTE ON THE STRENGTH OF SEASONED LOBLOLLY PINE 
 
 [Specimens, 2 by 2 inches; tested immediately after treatment] 
 
 Cylinder Conditio: 
 
 Steam, at various pressures 
 
 Creosote, injected at 150° F. 
 under a pressure of 100 
 pounds per square inch. . 
 
 Soaking, wood previously 
 treated with creosote in- 
 jected at 150° F. under a 
 pressure of 100 pounds per 
 square inch 
 
 Per 
 
 Per 
 
 'Ood = 100perce 
 
 92.6|106.4|124.0|107.7 6 
 70.91 55.ll 72.9' 6" " 
 
 Steamed wood = 100 per ( 
 
 97.1|102.2|110.7|103.3 7 
 
 Seasoned wood = 100 per cent. 
 
 81.0| 78. 4| 82.0| 80.4 7 
 Soaked untreated wood = 
 100 per cent. 
 
 19.20 139.4 160.4 
 
 . 6.06 
 
 5 21.6 . 
 
 1X.< 
 
 5 78.1 . 
 
 Physical Characteristics and Average Strengths of the Untreated Seasoned Wood 
 
 Moisture per cent 
 
 Weight per cubic foot (dry) pounds 
 
 Rings per inch 
 
 Modulus of elasticity pounds per square inch 1,6 
 
 Bending strength at elastic limit " " " 
 
 Bending strength at rupture " " " 
 
 Compression strength parallel to grain " " " 
 
 Compression strength at right angles to grain " " " 
 
 Shearing strength radial to grain " " " 
 
 TABLE XXXV. — Concluded 
 
 9,444 
 4,819 
 
 Untreated wood = 
 100 per cent. 
 117.0 I 130.0 
 
 Weight (air-seasoned). 
 
 Control. Treated. 
 
 Physical Characteristics and Average Strengths of the Untreated Wood 
 
 Moisture per cent 
 
 Weight per cubic foot (dry) pounds 
 
 Rings per inch 
 
 Modulus of elasticity pounds per square inch 1 
 
 Bending strength at elastic limit " " " 
 
 Bending strength at rupture " " " 
 
 Compression strength parallel to grain " " " 
 
 Compression strength at right angles to grain " " " 
 
 Shearing strength radial to grain " " " 
 
 8,760 
 4,956 
 
SUPPORTS OF THE RAIL 
 
 161 
 
 and the other at right angles to it. After the bending tests had been made 
 on these pieces, smaller pieces, 2 inches by 2 inches in cross section and 4 inches 
 long, were cut from their ends and used for the compression and shearing tests. 
 In any tie the test pieces were taken out according to Fig. 123, variation being 
 allowed only to secure clear pieces. 
 
 The test pieces from each tie were marked consecutively from 1 to 12. 
 The untreated pieces, marked 1 and 2, were used for control-impact tests, and 
 those marked 3 and 4 for control-static tests. The treated pieces, marked 5 
 and 6, were used for impact tests; those marked 7 and 8 for static tests. The 
 
 4_ 2 
 
 3 
 
 CONTROL 
 
 TREATED \ 
 
 / 
 
 S 
 
 S 
 
 2 
 
 6 
 
 /O 
 
 3 
 
 7 
 
 " 
 
 
 9 
 
 
 IZ 
 
 
 10 
 
 
 II 
 
 
 Fig. 123. — Control Plan — Creosote Tie Tests. 
 
 treated pieces, marked from 9 to 12, were similarly tested, but were resoaked, 
 if necessary, to bring them back to the degree of moisture found in the control 
 pieces. Ordinarily the steaming process did not decrease the moisture content 
 of the wood, in which case tests on resoaked pieces were not required. 
 
 In addition to the tests on small pieces, the strength of full-sized ties in 
 bending and in compression, both parallel and at right angles to grain, was 
 obtained, as well as the capacity of the wood to hold a spike. The ties used 
 were 8 feet long. The entire tie was treated and afterwards tested in full size. 
 In the bending tests under a static load, the ties were supported on a span of 
 80 inches and loaded at the third points of the span. 
 
 Short sections of the ties were used for tests to determine the resistance 
 against compression parallel to grain, against compression at right angles to 
 grain (which is similar to that produced on a tie by the base of a rail), and 
 against the force withdrawing a spike. In the tests of compression at right 
 angles to grain, the width of the tool equaled that of the base of an 80-pound 
 A. S. C. E. rail. The force necessary to cause the yielding of the wood was 
 measured. Both screw spikes and common spikes were driven into the tie, 
 and the force necessary to pull them out directly along their length was meas- 
 ured. Any common spike was driven but once, since it was found that the resist- 
 ance against pulling diminished when the spike was redriven into new wood. 
 
 The weight of the tie before treating, after treating, and at the time of 
 test was determined. The physical characteristics of the wood, such as per 
 
162 STEEL RAILS 
 
 cent of sap, rate of growth, shakes, knots, and moisture content, were also 
 recorded. 
 
 Impact tests were made on certain of the full-sized ties. In general, it 
 was found that the influence of the various factors may be determined by both 
 static and impact tests. 
 
 The results of these tests form a body of evidence from which the fol- 
 lowing general conclusions may be drawn: 
 
 (1) A high degree of steaming is injurious to wood in strength and spike- 
 holding power. The degree of steaming at which pronounced harm results 
 will depend upon the quality of the wood and its degree of seasoning, and upon 
 the pressure (temperature) of steam and the duration of its application. For 
 loblolly pine the limit of safety is certainly 30 pounds for 4 hours, or 20 pounds 
 for 6 hours. 
 
 (2) The presence of zinc chloride will not weaken wood under static load- 
 ing, although the indications are that the wood becomes brittle under impact 
 when treated with solutions above 3.5 per cent concentration. 
 
 (3) A light treatment with creosote will not weaken wood of itself. Since, 
 apparently, it is present only in the openings of the cells, and does not 
 get into the cell walls, its action can only be to retard the seasoning of 
 the wood. 
 
 The Committee on Wood Preservation of the American Railway Engi- 
 neering Association in its report at the March, 1910, Convention of the Asso- 
 ciation presented the following conclusions based on the best data available 
 at the time on the strength of treated timber: 
 
 (a) High steaming will diminish the strength rapidly. 
 
 (6) Treating with strong solution of zinc chloride will render the timber 
 brittle, perhaps because of free acid in the solution. 
 
 (c) Creosote is inert. 
 
 (d) Seasoned timber treated with light doses of creosote is as strong as 
 the original timber. 
 
 Tables XXXVII and XXXVIII give the results of tests of the Forest Ser- 
 vice on a number of woods, and Table XXXIX shows the unit stresses recom- 
 mended by the Committee on Wooden Bridges and Trestles of the American 
 Railway Engineering Association. 
 
 The great variation in strength, which is noticeable in timber of the same 
 species, makes it necessary to accept with caution the result of a limited number 
 of tests representing the average of the species. One of the most troublesome 
 factors influencing the strength of wood is the amount of moisture in it. 
 
SUPPORTS OF THE RAIL 
 
 163 
 
 TABLE XXXVI. — ACCOUNT OF TEST MATERIAL USED IN TABLE XXXVII 
 
 SUMMARY OF MECHANICAL TESTS ON THIRTY-TWO SPECIES OF AMERICAN WOODS 
 (Division of Forestry, Circular 15) 
 
 of Me- 
 
 clianii al 
 
 Tests. 
 
 Localities and Number of Trees from Each. 
 
 Cuban pine 
 
 (Pinus heterophylla.) 
 Shortleaf pine 
 
 t l'niiis echinata.) 
 Loblolly pine 
 
 (Pinus taeda). 
 
 Whil 
 
 
 Spruce pine 
 
 i'ma-aialaa , 
 
 Bald cypress 
 
 (Taxodium distichur 
 
 White cedar 
 
 (Chamaecyparis thyoides.) 
 Douglas spruce 
 
 (Pseudotsuga taxifolia.) 
 White oak 
 
 (Quercus alba.) 
 O vercup oak 
 
 (Quercus lyrata.) 
 Post oak . . 
 
 (Que 
 
 r.) 
 
 Cow oak . . 
 
 (Quercus michauxii.) 
 Red oak 
 
 (Quercus rubra.) 
 Texan oak 
 
 (Quercus texana.) 
 Yellow oak 
 
 (Quercus velutina.) 
 Water oak 
 
 (Quercus nigra.) 
 Willow oak 
 
 (Quercus phellos.) 
 
 ' '" 
 
 lllio, II . 
 
 Waici fnrkon 
 
 ' II I I 
 
 Nutmeg hickory 
 
 (Hicoria myristicaeformis.) 
 Pecan hickory 
 
 (Hicoria pecan.) 
 Pignut hickory 
 
 (Hicoria glabra.) 
 White elm 
 
 (Ulmus amerieana.) 
 Cedar elm 
 
 (Ulmus crassifolia.) 
 White ash 
 
 (FraxiL 
 
 Green ash 
 
 (Fraxinus lanceolata.) 
 
 stpla 
 
 ■Iain (22); uplands (6); hill district (6); 
 
 .ting uplands (6), South Carolina, coast 
 
 sippi, low coast plain (2); Louisiana, low 
 
 a, gravelly soil (7); sandy loam (6); Texas, low 
 
 i (6); Georgia, uplands (1); South 
 
 it plain (6). 
 Alabama, coast p 
 
 Carolina, coast (.,. 
 Alabama, uplands (4); Missouri, low hilly uplands (6); 
 
 Arkansas, low lulls' uplands 'Hi, I e>.a -, uplands Mi;. 
 Alabama, mountainous plateau (8); low coast plain (6); 
 
 Arkansas, level flood plain (5); Georgia, level coast 
 
 plain (6); South Carolina, low coast plain (7). 
 Wisconsin, clay, uplands (51; sandy soils (4); sandy loam 
 
 a",): Michigan, level drill lands (3). 
 Wisconsin, drift (5); Michigan (3). 
 
 Alabama, low 
 South Carolina 
 
 -i plaii 
 
 le barren (6); river bottom (4); Louisi- 
 
 , border of lake (4); Mississippi, Yazoo 
 
 bottom (3); upland (3). 
 ississippi, low plain. 
 
 Alabama, ridges of Tennessee Valley (5); Mississippi, low 
 
 plain (7). 
 Mississippi, low plain (7); Arkansas, Mississippi bottoms 
 
 (3). 
 Alabama, Tennessee Valley (5); Arkai 
 
 bottom (3). 
 Alabama, Tennessee Valley (4); Arkai 
 
 bottoms (3); Mississippi, low plain (4). 
 Alabama, Tennessee Valley (5); Arkansas, Mississi 
 
 bottom (2).* 
 Arkansas, Mississippi bottom. 
 
 Alabama, Tennessee Valley (5). 
 
 Mississippi, low plain (4). 
 
 Alabama, Tennessee Valley (5); Arkansas, Mississippi 
 
 hotioiu 1 3 i; Mississippi low plain i ! i. 
 Alabama, Tennessee Valley (5); Arkansas, I 
 
 l.oiioin (3);] [ississippi, low plain (3). 
 Mississippi, alluvial plain (3); limestone (J 
 
 Miss 
 
 sippi, low plai 
 
 ., bottom. 
 Arkansas, bottom. 
 Mississippi, bottom. 
 
 Arkansas, bottom (3); Mississippi, low plain (4). 
 
 * These two should probably be classed as Southern red oak. They were collected before the distinction 
 
 Note. — The values for specific , 
 moisture below 15 per cent; the n_ 
 at its influence on specific gravity 
 
 ere given refer to "dry" wood of test material, i.e., wood containing variat 
 
 jffect has therefore not been taken into account, but more careful experime 
 
 such low per cent is so small that it may be neglected for practical purposes. 
 
STEEL RAILS 
 
 In Table XXXVII all values except those for the Southern pines have been 
 referred to 12 per cent moisture, which may be said to be the lightest average 
 moisture content of seasoned wood. 
 
 TABLE XXXVII. — RESULTS OF TESTS IN BENDING — AT RUPTURE 
 
 SUMMARY OF MECHANICAL TESTS ON THIRTY-TWO SPECIES OF AMERICAN WOODS 
 
 (Division of Forestry, Circular 15) 
 
 [Pounds per square inch] 
 
 No. 
 
 Species. 
 
 Number 
 
 of 
 Tests. 
 
 Single 
 Test. 
 
 Lowest 
 Single 
 Test. 
 
 10 Per 
 Tests. 
 
 I.nweM 
 
 10 Per 
 
 cent of 
 Tests. 
 
 Average 
 Tests. 
 
 o°Trats n 
 
 within 
 
 10 Per cent 
 
 Proportion 
 
 of Tests 
 
 within 
 
 25 Per cent 
 
 of Average. 
 
 1 
 2 
 
 Reduced to 15 Per cent Moisture 
 Longleafpine 
 
 1,160 
 390 
 330 
 650 
 
 95 
 
 170 
 
 87 
 41 
 218 
 216 
 
 256 
 57 
 117 
 40 
 31 
 153 
 257 
 187 
 75 
 14 
 25 
 72 
 37 
 30 
 18 
 
 87 
 10 
 118 
 
 17,800 
 
 17,(100 
 
 14^00 
 
 11,100 
 12,900 
 16,300 
 
 9! 100 
 
 20^300 
 19,600 
 
 23^000 
 16,500 
 19,500 
 15,000 
 16,000 
 16,000 
 17,300 
 23,300 
 20,700 
 18,000 
 19,500 
 16,600 
 18,300 
 25,000 
 14,000 
 19,200 
 15,000 
 16,000 
 14,400 
 
 2^900 
 5,000 
 3,900 
 
 4,600 
 3,100 
 3,100 
 2,300 
 3,500 
 3,800 
 5,700 
 4,900 
 5,100 
 3,300 
 5,700 
 8,200 
 5,100 
 5,800 
 3,200 
 5,000 
 5,700 
 5,300 
 5,300 
 7,000 
 6,700 
 5,600 
 11,100 
 7,300 
 6,600 
 5,000 
 5,100 
 5,100 
 
 14,200 
 14,600 
 12,400 
 13,100 
 
 10,100 
 12,300 
 13,600 
 11,700 
 8,400 
 12,000 
 18,500 
 14,900 
 15.300 
 12,500 
 15,400 
 16,900 
 14,600 
 15,700 
 13.800 
 15,600 
 20,300 
 19,700 
 17,300 
 19,300 
 15,600 
 18,100 
 24,300 
 13,600 
 17,300 
 14.200 
 16.000 
 12,700 
 
 8,800 
 8,800 
 7,000 
 8,100 
 
 5,000 
 4,900 
 
 5!000 
 4,000 
 4,100 
 
 6^00 
 7,400 
 
 9^100 
 10,000 
 5,700 
 7,200 
 5,400 
 6,900 
 9,400 
 7,900 
 5,400 
 8,700 
 8,100 
 10,300 
 11,500 
 7,300 
 8,500 
 6,300 
 5,100 
 6,000 
 
 10,900 
 11,900 
 9,200 
 10,100 
 
 7,900 
 9,100 
 10,000 
 7,900 
 6,300 
 7,900 
 13,100 
 11,300 
 12,300 
 11,500 
 11,400 
 13,100 
 10,800 
 12,400 
 10,400 
 12,000 
 16,000 
 15,200 
 12,500 
 15,000 
 12,500 
 15,300 
 18,700 
 10,300 
 13,500 
 
 1L600 
 9,500 
 
 Per cent. 
 
 44 
 
 43 
 28 
 43 
 25 
 32 
 
 39 
 47 
 47 
 32 
 
 64 
 28 
 
 33 
 40 
 46 
 
 21 
 
 28 
 
 38 
 43 
 44 
 50 
 37 
 20 
 39 
 
 Per cent. 
 
 84 
 83 
 
 
 
 
 fl 
 
 Reduced to 12 Per cent Moisture. 
 
 81 
 
 6 
 
 
 60 
 
 
 
 
 8 
 
 
 
 9 
 
 
 78 
 
 
 
 58 
 
 
 
 75 
 
 
 
 81 
 
 
 
 92 
 
 
 
 68 
 
 15 
 
 Red oak 
 
 84 
 86 
 
 
 
 
 
 
 
 
 
 70 
 
 
 
 
 
 
 
 
 
 
 
 
 
 24 
 
 I-Sll lernut liirknrv 
 
 60 
 
 
 Pecan hickory 
 
 
 
 
 72 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ♦Actual tests on " dry " material not reduced for moisture. 
 
SUPPORTS OF THE RAIL 
 
 165 
 
 TABLE XXXVII. 
 
 -RESULTS OF TESTS IN BENDING — AT RELATIVE 
 ELASTIC LIMIT 
 
 SUMMARY OF MECHANICAL TESTS ON THIRTY-TWO SPECIES OF AMERICAN WOODS 
 
 (Division of Forestry, Circular 15) 
 
 (Pounds per square inch] 
 
 No. 
 
 — 
 
 Number 
 
 of 
 
 Tests. 
 
 Highest 
 
 Test. 
 
 Test. 
 
 ,1 llmlii'M 
 
 10 Per 
 cent of 
 
 Tests. 
 
 Average 
 of I.owcM 
 10 Per 
 cent of 
 Tests. 
 
 Average 
 of all 
 Tests. 
 
 Proportion 
 of Tests 
 within 
 10 Per cent 
 of Average. 
 
 Proportion 
 of Tests 
 
 25 Per 'cent 
 of Average. 
 
 Modulus 
 of Elas- 
 
 ( Average of 
 all Tests). 
 
 
 Reduced to 15 Per 
 cent Moisture. 
 
 1,160 
 390 
 330 
 650 
 
 130 
 95 
 170 
 
 87 
 
 218 
 216 
 49 
 256 
 
 57 
 117 
 
 31 
 153 
 257 
 187 
 
 75 
 
 25 
 72 
 37 
 
 18 
 44 
 87 
 10 
 118 
 
 13.500 
 12.900 
 11,900 
 12,700 
 
 10.000 
 11,300 
 13,700 
 12,000 
 
 13J00 
 15,700 
 11,600 
 10,600 
 14,200 
 14,500 
 12,000 
 11,800 
 11,800 
 13,100 
 13,500 
 16,100 
 15,400 
 11,900 
 14,300 
 12,200 
 15,000 
 17,500 
 9,700 
 10,700 
 11,500 
 13,200 
 11,000 
 
 2,400 
 2,200 
 2,900 
 3,100 
 
 4,100 
 3,100 
 3,000 
 2,200 
 
 4!400 
 4,000 
 5,100 
 
 5400 
 5,900 
 4,900 
 4,500 
 2,700 
 5,100 
 5,400 
 4,300 
 4,100 
 7,500 
 4,200 
 5,800 
 7,400 
 5,300 
 
 3!600 
 3,200 
 3,500 
 
 11,100 
 11,500 
 
 10300 
 
 8.200 
 10,300 
 11,200 
 
 7J390 
 9,600 
 14,100 
 9,500 
 
 ll|600 
 
 13,600 
 11,400 
 11,100 
 11,400 
 10,000 
 11,600 
 14,200 
 
 11300 
 14,000 
 11,200 
 14,400 
 16,400 
 
 io!ioo 
 
 10,400 
 13,200 
 10,100 
 
 5,400 
 5,600 
 4,800 
 5,400 
 
 4.500 
 4,500 
 5,000 
 4,200 
 4,000 
 3,400 
 6,100 
 
 6^000 
 5,000 
 5,600 
 7,800 
 5,100 
 5,500 
 4,300 
 6,600 
 7,700 
 7,800 
 4,800 
 7,600 
 6,400 
 7,900 
 8,300 
 5,400 
 5,800 
 5,200 
 3,200 
 5.100 
 
 8,500 
 
 7^200 
 8,200 
 
 6,400 
 7,700 
 8,400 
 6,600 
 
 9^00 
 
 81400 
 7,600 
 9,200 
 9,400 
 8,100 
 8,800 
 7,400 
 8,600 
 11,200 
 11,700 
 9,800 
 11,100 
 9,300 
 11,500 
 12,600 
 7,300 
 8,000 
 7,900 
 8,900 
 
 Per cent. 
 
 43 
 42 
 48 
 46 
 
 38 
 
 25 
 44 
 32 
 37 
 
 34 
 50 
 15 
 62 
 
 40 
 42 
 
 50 
 39 
 21 
 44 
 46 
 
 33 
 57 
 43 
 40 
 46 
 
 81 
 83 
 81 
 85 
 
 85 
 73 
 
 66 
 86 
 56 
 73 
 
 76 
 95 
 49 
 94 
 75 
 84 
 81 
 80 
 89 
 
 86 
 84 
 93 
 89 
 83 
 71 
 91 
 83 
 
 82 
 
 
 
 
 2,300,000 
 
 
 
 
 
 
 
 
 Reduced to 12 Per 
 cent Moisture. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 10 
 
 1 ).iuu!:i- spruce * 
 
 1,680,000 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 17 
 
 Yellow oak 
 
 1,740.000 
 
 19 
 
 Willrm oak 
 
 1,750,000 
 
 21 
 22 
 23 
 
 25 
 26 
 27 
 
 Shagliark hickory 
 
 2,390,000 
 2,320,000 
 2,080,000 
 2,280,000 
 1,940,000 
 2,530,000 
 2,730,000 
 
 Water hickory 
 
 Bitternut hickory 
 
 Xut nit-Li hickory 
 
 Pecan hickory 
 
 Pignut hickory 
 
 
 
 
 
 
 
 
 
 
 
 
 1,700,000 
 
 
 
 * Actual tests on "dry" material not reduced for moisture. 
 
 TABLE XXXVII.— Concluded — RESULTS OF TESTS IN COMPRESSION, ACROSS 
 GRAIN,* AND SHEARING WITH GRAIN 
 
 SUMMARY OF MECHANICAL TESTS ON THIRTY-TWO SPECIES OF AMERICAN WOODS 
 
 (Division of Forestry, Circular 15) 
 
 [Pounds per square inch] 
 
 No. 
 
 Species. 
 
 of 
 
 Tests. 
 
 Compres- 
 G°rain S . 
 
 Shearing 
 with 
 Grain 
 not Re- 
 duced for 
 Moisture. 
 
 No. 
 
 Species. 
 
 Number 
 
 of 
 Tests. 
 
 Compres- 
 Grab. 
 
 Shearing 
 with 
 Grain 
 not Re- 
 duced for 
 Moisture. 
 
 1 
 
 Reduced to 15 Per cent 
 Moisture. 
 
 1,210 
 400 
 330 
 690 
 
 130 
 100 
 175 
 650 
 87 
 41 
 218 
 216 
 
 256 
 
 1,000 
 1,000 
 900 
 1,000 
 
 700 
 1,000 
 1,200 
 800 
 700 
 800 
 2,200 
 1,900 
 3,000 
 1,900 
 
 700 
 700 
 700 
 700 
 
 400 
 500 
 800 
 500 
 400 
 
 1,000 
 1,000 
 1,100 
 900 
 
 15 
 16 
 
 17 
 
 20 
 
 22 
 23 
 24 
 25 
 26 
 27 
 2* 
 29 
 30 
 31 
 
 Reduced to 12 Per cent 
 Moisture. — Concluded. 
 
 Red oak 
 
 Southern red oak 
 
 57 
 117 
 
 30 
 153 
 255 
 135 
 
 14 
 25 
 72 
 37 
 
 18 
 
 87 
 10 
 118 
 
 2,300 
 2,000 
 
 2|000 
 1,600 
 1,800 
 2,700 
 3,100 
 2,400 
 2,200 
 2,700 
 2,800 
 3,200 
 1,200 
 2,100 
 1,900 
 1,700 
 1,400 
 
 1,100 
 
 
 
 
 
 
 
 
 
 
 1,100 
 
 
 Reduced to 12 Per cent 
 Moisture. 
 
 Willow oak 
 
 Spanish oak 
 
 Shagbat 1; hickory 
 
 900 
 900 
 1,100 
 
 
 Water h.ckorv 
 
 Bitternut hickory 
 
 Nutmeg hickory 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Pignut hickory 
 
 White elm 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * To an indentation of 3 
 
 it of the height of the 
 
 in "dry" material not reduced for 
 
166 
 
 STEEL RAILS 
 
 TABLE XXXVIII. — STRENGTH VALUES FOR STRUCTURAL TIMBERS 
 
 (Forest Service, Circular 189) 
 BENDING TESTS ON GREEN MATERIAL 
 
 
 Sizes. 
 
 H 
 
 1 
 
 j3 
 
 F. S. a 
 
 E. L. 
 
 M. of R. 
 
 M. of E. 
 
 Calculated 
 Shear. 
 
 
 g 
 
 
 a 
 
 
 u4 
 
 
 j 
 
 
 ^ 
 
 
 Species. 
 
 1 
 % 
 
 s 
 
 o 
 
 
 
 
 
 II 
 
 
 
 
 II 
 
 
 ej 
 
 
 
 1 
 
 1 
 | 
 2 
 
 1 
 
 g, 
 
 II 
 
 |£ 
 
 
 1 -<' 
 
 2JJ M 
 
 5„. 
 
 
 5,. 
 
 
 Ins. 
 
 Ins. 
 
 
 
 
 Lbs. 
 
 
 Lbs. 
 
 
 1000 
 Lbs. 
 
 
 Lbs. 
 
 
 Longleaf pine 
 
 12X12 
 
 138 
 
 4 
 
 2S 6 
 
 9." 
 
 4099 
 
 83 
 
 6710 
 
 0.74 
 .71 
 
 1523 
 
 0.9! 
 
 261 
 306 
 
 0.86 
 
 
 10X16 
 
 168 
 
 4 
 
 26. S 
 
 16.' 
 
 4193 
 
 .85 
 
 6453 
 
 1626 
 
 1.05 
 
 1.01 
 
 
 8X16 
 
 156 
 
 7 
 
 28.4 
 
 14. ( 
 
 3147 
 
 .64 
 
 5439 
 
 .60 
 
 1368 
 
 
 390 
 
 1.29 
 
 
 6X16 
 
 132 
 
 1 
 
 10.8 
 
 21 ..- 
 
 4120 
 
 .83 
 
 6460 
 
 .71 
 
 1190 
 
 " 
 
 378 
 
 1.25 
 
 
 6X10 
 
 180 
 
 1 
 
 31.0 
 
 6.S 
 
 3580 
 
 .72 
 
 6500 
 
 .72 
 
 1412 
 
 .92 
 
 175 
 
 .58 
 
 
 6X 8 
 
 180 
 
 2 
 
 '27.0 
 
 8.5 
 
 3735 
 
 .75 
 
 5745 
 
 .63 
 
 1282 
 
 .83 
 
 121 
 
 .40 
 
 
 2X 2 
 
 30 
 
 15 
 
 33.0 
 
 14. 
 
 4950 
 
 1.00 
 
 9070 
 
 1.00 
 
 1540 
 
 1.0C 
 
 303 
 
 1.00 
 
 Douglas fir 
 
 8X16 
 
 180 
 
 191 
 
 31.5 
 
 11. ( 
 
 3968 
 
 76 
 
 5983 
 
 .72 
 
 1517 
 
 95 
 
 269 
 172 
 
 81 
 
 
 5X 8 
 
 180 
 
 84 
 
 30.1 
 
 10. * 
 
 3693 
 
 .71 
 
 5178 
 
 1533 
 
 .96 
 
 .52 
 
 
 2X12 
 
 180 
 
 27 
 
 35.7 
 
 20.( 
 
 3721 
 
 .71 
 
 5276 
 
 .64 
 
 1642 
 
 1.03 
 
 256 
 
 .77 
 
 
 2X10 
 
 180 
 
 26 
 
 32.9 
 
 21. ( 
 
 3160 
 
 .60 
 
 4699 
 
 .57 
 
 1593 
 
 1.00 
 
 189 
 
 .57 
 
 
 2X 8 
 
 180 
 
 29 
 
 33.0 
 
 17. ( 
 
 3593 
 
 .69 
 
 5352 
 
 .65 
 
 1607 
 
 1 01 
 
 171 
 
 .51 
 
 
 2X 2 
 
 24 
 
 508 
 
 30.4 
 
 11. ( 
 
 5227 
 
 1.00 
 
 8280 
 
 1.01 
 
 1597 
 
 L.0( 
 
 333 
 
 1.00 
 
 Douglas fir (fire-killed) . . 
 
 8x16 
 
 180 
 
 30 
 
 36. N 
 
 10.' 
 
 3503 
 
 .80 
 
 4994 
 
 .64 
 
 1531 
 
 .94 
 
 330 
 
 1.19 
 
 
 2x12 
 
 180 
 
 32 
 
 34.2 
 
 17.' 
 
 3489 
 
 .80 
 
 5085 
 
 .66 
 
 1624 
 
 .99 
 
 247 
 
 .89 
 
 
 2x10 
 
 180 
 
 32 
 
 38.9 
 
 18. 
 
 3851 
 
 .88 
 
 5359 
 
 .6! 
 
 1716 
 
 1.05 
 
 216 
 
 .78 
 
 
 2x 8 
 
 180 
 
 31 
 
 87.0 
 
 15." 
 
 3403 
 
 .78 
 
 5305 
 
 .68 
 
 1676 
 
 1.01 
 
 169 
 
 .61 
 
 
 2X 2 
 
 30 
 
 200 
 
 38.2 
 
 17.1 
 
 4360 
 
 1.00 
 
 7752 
 
 1.00 
 
 1636 
 
 1.00 
 
 277 
 
 1.00 
 
 Shortleaf pine 
 
 8X16 
 
 180 
 
 12 
 
 80.5 
 
 12. 
 
 3185 
 
 .73 
 
 5407 
 
 .70 
 
 1438 
 
 1.03 
 
 362 
 
 1.40 
 
 
 8X14 
 
 180 
 
 12 
 
 45.8 
 
 12." 
 
 3234 
 
 .74 
 
 5781 
 
 .75 
 
 1494 
 
 1.07 
 
 338 
 
 1.31 
 
 
 8X12 
 
 180 
 
 24 
 
 52 . 2 
 
 11. E 
 
 3265 
 
 .75 
 
 5503 
 
 .71 
 
 1480 
 
 1.06 
 
 277 
 
 1.07 
 
 
 5X 8 
 
 180 
 
 24 
 
 I7.s 
 
 11. 1 
 
 3519 
 
 .81 
 
 5732 
 
 .74 
 
 1485 
 
 1.06 
 
 185 
 
 .72 
 
 
 2X 2 
 
 30 
 
 25 1: 
 
 51.7 
 
 18. ( 
 
 4350 
 
 1.00 
 
 7710 
 
 1.00 
 
 1395 
 
 1.00 
 
 258 
 
 1.00 
 
 Western larch 
 
 8X16 
 
 180 
 
 82 
 
 51.0 
 
 25.: 
 
 3276 
 
 .77 
 
 4632 
 
 .64 
 
 1272 
 
 .97 
 
 298 
 
 1.11 
 
 
 8X12 
 
 180 
 
 30 
 
 50 . 3 
 
 23 . '. 
 
 3376 
 
 .79 
 
 5286 
 
 .73 
 
 1331 
 
 1.02 
 
 254 
 
 .94 
 
 
 5X 8 
 
 180 
 
 14 
 
 5(3.0 
 
 25 ( 
 
 3528 
 
 .83 
 
 5331 
 
 .74 
 
 1432 
 
 1 09 
 
 169 
 
 .63 
 
 
 2X 2 
 
 28 
 
 189 
 
 1(3.2 
 
 2(3 1 
 
 4274 
 
 1.00 
 
 7251 
 
 1.00 
 
 1310 
 
 i 00 
 
 269 
 
 1.00 
 
 Loblolly pine 
 
 8X16 
 
 180 
 
 17 
 
 45.8 
 
 6.1 
 
 3094 
 
 .75 
 
 5394 
 
 .69 
 
 1406 
 
 .98 
 
 383 
 
 1.44 
 
 
 5X12 
 
 180 
 
 94 
 
 00,9 
 
 5.S 
 
 3030 
 
 .74 
 
 5028 
 
 .64 
 
 1383 
 
 .96 
 
 221 
 
 .83 
 
 
 2X 2 
 
 30 
 
 44 
 
 70.9 
 
 5.4 
 
 4100 
 
 1.00 
 
 7870 
 
 1.00 
 
 1440 
 
 1.00 
 
 265 
 
 1.00 
 
 Tamarack 
 
 6X12 
 
 162 
 
 15 
 
 57.6 
 
 16.6 
 
 2914 
 
 .75 
 
 4500 
 
 .66 
 
 1202 
 
 1.05 
 
 255 
 
 1.11 
 
 
 4X10 
 
 162 
 
 15 
 
 43 . 5 
 
 11.4 
 
 2712 
 
 .70 
 
 4611 
 
 .68 
 
 1238 
 
 1.08 
 
 209 
 
 .91 
 
 
 2X 2 
 
 30 
 
 82 
 
 38.8 
 
 14.0 
 
 3875 
 
 1.00 
 
 6820 
 
 1.00 
 
 1141 
 
 1.00 
 
 229 
 
 1.00 
 
 Western hemlock 
 
 8X16 
 
 180 
 
 39 
 
 42 . 5 
 
 15.6 
 
 3516 
 
 .80 
 
 5296 
 
 .73 
 
 1445 
 
 1.01 
 
 261 
 
 .92 
 
 
 2X 2 
 
 28 
 
 52 
 
 51.8 
 
 12 J 
 
 4406 
 
 1.00 
 
 7294 
 
 1.00 
 
 1428 
 
 1.00 
 
 284 
 
 1.00 
 
 Redwood 
 
 8X16 
 
 180 
 
 14 
 
 X0.5 
 
 19. ! 
 
 3734 
 
 .79 
 
 4492 
 
 .64 
 
 1016 
 
 .96 
 
 300 
 
 1.21 
 
 
 6X12 
 
 180 
 
 14 
 
 S7 . 3 
 
 17. .s 
 
 3787 
 
 .80 
 
 4451 
 
 .64 
 
 1068 
 
 1.00 
 
 224 
 
 .90 
 
 
 7X 9 
 
 180 
 
 14 
 
 79.8 
 
 10.7 
 
 4412 
 
 .93 
 
 5279 
 
 .76 
 
 1324 
 
 1.25 
 
 199 
 
 .80 
 
 
 3X14 
 
 180 
 
 13 
 
 86.1 
 
 23.7 
 
 3506 
 
 .74 
 
 4364 
 
 .62 
 
 947 
 
 .89 
 
 255 
 
 1.03 
 
 
 2x12 
 
 180 
 
 12 
 
 70 9 
 
 18.6 
 
 3100 
 
 .65 
 
 3753 
 
 .54 
 
 1052 
 
 .99 
 
 187 
 
 .75 
 
 
 2x10 
 
 180 
 
 13 
 
 55.8 
 
 20 
 
 3285 
 
 .69 
 
 4079 
 
 .58 
 
 1107 
 
 1.04 
 
 169 
 
 .68 
 
 
 2x 8 
 
 180 
 
 13 
 
 S3 . 8 
 
 21.5 
 
 2989 
 
 .63 
 
 4063 
 
 .58 
 
 1141 
 
 1 .OS 
 
 .134 
 
 .54 
 
 
 2X 2 
 
 28 
 
 157 
 
 75.5 
 
 19.1 
 
 4750 
 
 1.00 
 
 6980 
 
 1.00 
 
 1061 
 
 1.00 
 
 248 
 
 1.00 
 
 Norway pine 
 
 6X12 
 
 162 
 
 15 
 
 50.3 
 
 12 5 
 
 2305 
 
 .82 
 
 3572 
 
 .69 
 
 987 
 
 1.08 
 
 201 
 
 1.17 
 
 
 4X12 
 
 162 
 
 18 
 
 47.9 
 
 14.7 
 
 2648 
 
 .94 
 
 4107 
 
 .79 
 
 1255 
 
 1.31 
 
 238 
 
 1.38 
 
 
 4X10 
 
 162 
 
 16 
 
 45 . 7 
 
 13.3 
 
 2674 
 
 .95 
 
 4205 
 
 .81 
 
 1306 
 
 1.30 
 
 198 
 
 1.15 
 
 
 2X 2 
 
 30 
 
 133 
 
 32 8 
 
 11.4 
 
 2808 
 
 1.00 
 
 5173 
 
 1.00 
 
 960 
 
 I 00 
 
 172 
 
 1.00 
 
 Red spruce 
 
 2X10 
 
 144 
 
 14 
 
 32.5 
 
 21.9 
 
 2394 
 
 .66 
 
 3566 
 
 .60 
 
 1180 
 
 1.02 
 
 181 
 
 .80 
 
 
 2x 2 
 
 26 
 
 60 
 
 37 8 
 
 21 3 
 
 3627 
 
 1.00 
 
 5900 
 
 1.00 
 
 1157 
 
 1 00 
 
 227 
 
 1.00 
 
 White spruce 
 
 2x10 
 
 144 
 
 16 
 
 40.7 
 
 9.3 
 
 2239 
 
 .72 
 
 3288 
 
 .63 
 
 1081 
 
 1.08 
 
 166 
 
 .83 
 
 
 2X 2 
 
 26 83 
 
 58.3 
 
 2 
 
 3090 
 
 1.00 5185 
 
 1.00 
 
 998 
 
 1. 00 
 
 199 
 
 1.00 
 
SUPPORTS OF THE RAIL 
 
 167 
 
 TABLE XXXVIII. — Continued 
 COMPRESSION AND SHEAR TESTS ON GREEN MATERIAL 
 
 
 Compression || to 
 
 grain. 
 
 
 Compression J_ 
 
 to grain. 
 
 Shear. 
 
 Species. 
 
 § 
 
 s 
 
 1 
 
 1 
 H 
 
 1 
 
 a 
 1 
 
 1 
 1 
 
 H-g 
 
 6 M 
 
 
 21 
 
 2& 
 
 i 
 < 
 
 1 
 
 X 
 
 a 
 
 H 
 
 | 
 
 1 
 1 
 
 ?5 
 
 h4 
 
 1 
 
 | 
 1 
 
 1 
 
 § 
 (2 
 
 l 
 §1 
 
 
 4X4 
 2X2 
 6X6 
 5X6 
 2X2 
 6X6 
 2x2 
 6X6 
 5X8 
 2X2 
 
 46 
 
 14 
 
 515 
 170 
 902 
 108 
 201 
 95 
 23 
 2S1 
 
 26.3 
 34 7 
 30 . 7 
 30.9 
 29.. s 
 34.8 
 37.9 
 -11.2 
 4:: . 5 
 51.4 
 
 Lbs. 
 3480 
 
 1000 
 Lbs. 
 
 Lbs. 
 
 4800 
 4400 
 3500 
 3490 
 4030 
 3290 
 3430 
 3436 
 3423 
 3570 
 
 Ins. 
 
 4X4 
 
 Ins. 
 4 
 
 22 
 
 25.3 
 
 Lbs. 
 568 
 
 44 
 
 21.8 
 
 Lbs. 
 
 973 
 
 
 
 
 2780 
 2720 
 3500 
 2620 
 
 1181 
 2123 
 1925 
 1801 
 
 4X8 
 
 16 
 
 259 
 
 30.3 
 
 570 
 
 531 
 
 29.7 
 
 765 
 
 
 
 
 
 
 
 
 
 
 
 
 Douglas fir (fire-killed) . . 
 
 6X8 
 
 16 
 
 24 
 
 33.7 
 
 368 
 
 77 
 
 35.8 
 
 631 
 
 
 2514 
 2241 
 
 1565 
 1529 
 
 5X8 
 5X8 
 5X8 
 5X5 
 2X2 
 6X8 
 6X8 
 4X6 
 4X4 
 8X4 
 4X4 
 
 16 
 14 
 12 
 8 
 2 
 16 
 12 
 6 
 4 
 8 
 8 
 
 12 
 12 
 
 24 
 24 
 277 
 22 
 20 
 53 
 30 
 16 
 38 
 
 37.7 
 
 42 . 8 
 53.0 
 47.0 
 4S.5 
 43.0 
 40.2 
 52. s 
 50.4 
 07 . 2 
 44.6 
 
 361 
 
 366 
 325 
 344 
 400 
 417 
 416 
 478 
 472 
 392 
 546 
 
 179 
 
 47.0 
 
 704 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6X6 
 2X2 
 
 107 
 
 491 
 
 49.1 
 50.6 
 
 2675 
 3026 
 
 1575 
 1545 
 
 3510 
 3696 
 
 179 
 
 40.7 
 
 700 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Loblolly pine 
 
 8X8 
 4X8 
 2X2 
 6X7 
 4X7 
 2X2 
 6X6 
 2X2 
 6X6 
 2X2 
 
 14 
 18 
 
 53 
 4 
 6 
 165 
 82 
 131 
 34 
 143 
 
 65 4 
 60.0 
 74.0 
 49.9 
 27.7 
 3(3.8 
 46 . 6 
 55.6 
 83 6 
 72.1 
 
 1560 
 2430 
 
 365 
 691 
 
 2140 
 3560 
 3240 
 3032 
 3360 
 3190 
 3355 
 3392 
 3882 
 3980 
 
 121 
 
 83.2 
 
 630 
 
 
 
 
 
 
 2332 
 2444 
 
 1432 
 1334 
 
 
 
 
 
 
 24 
 
 39 . 2 
 
 668 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Western hemlock 
 
 2905 
 2938 
 3194 
 3490 
 
 1617 
 1737 
 1240 
 1222 
 
 6X4 
 
 6 
 
 30 
 
 48.7 
 
 434 
 
 54 
 
 65.7 
 
 630 
 
 
 6X8 
 6X6 
 6X7 
 6X3 
 6X2 
 6X2 
 6X2 
 2X2 
 
 16 
 12 
 9 
 14 
 12 
 10 
 8 
 2 
 
 13 
 14 
 
 13 
 13 
 
 12 
 11 
 
 12 
 1S6 
 
 8G.7 
 53.0 
 
 71.7 
 75 6 
 .6 5 
 
 V, 
 56.7 
 
 75.5 
 
 473 
 424 
 477 
 411 
 430 
 423 
 396 
 569 
 
 148 
 
 84.2 
 
 742 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6X7 
 4x7 
 2X2 
 2x2 
 2X2 
 
 5 
 
 8 
 
 178 
 58 
 84 
 
 29 
 28 4 
 20 8 
 35 4 
 61.0 
 
 1928 
 2154 
 
 905 
 1063 
 
 2404 
 2652 
 2504 
 2750 
 2370 
 
 20 
 
 26.7 
 
 589 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2X2 
 2X2 
 
 2 
 2 
 
 43 
 46 
 
 31. S 
 50.4 
 
 310 
 270 
 
 30 
 40 
 
 32.0 
 3S.0 
 
 758 
 
 
 
 
 651 
 
 
 
 
 
168 
 
 STEEL RAILS 
 
 TABLE XXXIX. — UNIT STRESSES FOR STRUCTURAL TIMBER RECOMMENDED BY 
 THE COMMITTEE ON WOODEN BRIDGES AND TRESTLES 
 
 AM. RY. ENG. ASSN. 
 [Pounds per Square Inch] 
 
 
 Bending. 
 
 Shearing. 
 
 Compression. 
 
 o^ 
 
 Kind of 
 Timber. 
 
 Extreme 
 Fiber 
 Stress. 
 
 Modulus 
 of Elas- 
 ticity. 
 
 Parallel to 
 Grain. 
 
 Longitudi- 
 nal Shear in 
 Beams. 
 
 Perpendic- 
 Grain. 
 
 Parallel to 
 Grain. 
 
 For 
 Columns 
 
 15 Diams. 
 Safe 
 
 Formulas for 
 Safe Stress 
 
 Columns over 
 .15 Diams. 
 
 If 2 
 II 
 
 
 
 "§! 
 
 1 
 
 II 
 
 !| 
 
 E'i 
 
 "I 
 
 la 
 k5 
 
 fl 
 
 as 
 
 u 
 
 l! 
 
 Douglas fir .... 
 
 6100 
 
 1200 
 
 1,510,000 
 
 690 
 
 170 
 
 270 
 
 110 
 
 630 
 
 310 
 
 3600 
 
 1200 
 
 900 
 
 ""O-efe) 
 
 10 
 
 Longleaf pine. . . 
 
 6500 
 
 1300 
 
 1,610,000 
 
 720 
 
 180 
 
 300 
 
 120 
 
 520 
 
 260 
 
 3800 
 
 1300 
 
 980 
 
 M'-Wd) 
 
 10 
 
 Shortleaf pine... 
 
 5600 
 
 1100 
 
 1,480,000 
 
 710 
 
 170 
 
 330 
 
 130 
 
 340 
 
 170 
 
 3400 
 
 1100 
 
 830 
 
 M'-^d) 
 
 10 
 
 White pine 
 
 4400 
 
 900 
 
 1,130,000 
 
 400 
 
 100 
 
 180 
 
 70 
 
 290 
 
 150 
 
 3000 
 
 1000 
 
 750 
 
 ™( l -m) 
 
 10 
 
 S ruce 
 
 
 
 
 
 
 
 
 
 
 
 1100 
 
 830 
 
 1100(1 L ) 
 
 
 Norway pine. . . 
 
 4200 
 
 800 
 
 1,190,000 
 
 590 
 
 130 
 
 250 
 
 100 
 
 
 150 
 
 2600* 
 
 800 
 
 600 
 
 V 60 D ) 
 
 
 Tamarack 
 
 4600 
 
 900 
 
 1,220,000 
 
 670 
 
 170 
 
 260 
 
 100 
 
 
 220 
 
 3200* 
 
 1000 
 
 750 
 
 M'-do) 
 
 
 Western hem- 
 
 5800 
 
 1100 
 
 1,480,000 
 
 630 
 
 160 
 
 270* 
 
 100 
 
 440 
 
 220 
 
 3500 
 
 1200 
 
 900 
 
 1200 (l-^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 900 (l L ) 
 
 
 Bald cypress. . . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 White oak 
 
 5700 
 
 1100 
 
 1,150,000 
 
 840 
 
 210 
 
 270 
 
 110 
 
 920 
 
 
 3500 
 
 1300 
 
 980 
 
 M'-m) 
 
 12 
 
 Note. — T 
 
 ese un 
 
 it stre 
 
 ses are for 
 
 a gree 
 
 n cond 
 
 tion o 
 
 D = \ 
 
 
 
 >e usee 
 
 withe 
 
 th in inch 
 
 ig the live-load s 
 
 tressea 
 
 The difference between green and seasoned wood may amount to as much 
 as 50 per cent as shown by Table XL. The influence of seasoning consists in 
 (1) bringing by means of shrinkage about 10 per cent more fibers into the same 
 square inch of cross section than are contained in the wet wood; (2) shrinking 
 the cell-wall itself by about 50 per cent of its cross section and thus hardening 
 it, just as a cowskin becomes thinner and hardens by drying. 
 
 Table XL applies only to small, clear pieces of wood seasoned under special 
 conditions with great care. The Forest Service has found * that a comparison of 
 the results of tests on seasoned material with those from tests on green material 
 shows that, without exception, the strength of 2 by 2 inch specimens is increased 
 by lowering the moisture content, but that increase in strength of other sizes is 
 
 * Strength Values for Structural Timbers. McGarvey Cline. Forest Service, Circular 189, 
 Jan. 25, 1912. 
 
SUPPORTS OF THE RAIL 
 
 169 
 
 much more erratic. Some specimens, in fact, show an apparent loss in strength 
 due to seasoning. In the light of these facts it is not safe to base working stresses 
 on results secured from any but green material. 
 
 TABLE XL. — REDUCTION FACTORS FOR STRESS AT ELASTIC LIMIT IN BEND- 
 ING OF LONGLEAF PINE 
 
 (Forest Service, Bulletin 70) 
 
 
 From — 
 
 To — 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Moisture 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 *■ 
 
 6. 
 
 8. 
 
 10. 
 
 12. 
 
 14. 
 
 16. 
 
 18. 
 
 20. 
 
 22. 
 
 24. 
 
 26.* 
 
 2 
 
 1 
 
 1.13 
 
 1.31 
 
 1.53 
 
 1.75 
 
 1.99 
 
 2.20 
 
 2.39 
 
 2.54 
 
 2.70 
 
 2.85 
 
 2.99 
 
 3.14 
 
 4 
 
 .882 
 
 1 
 
 1.15 
 
 1.35 
 
 1.54 
 
 1.75 
 
 1.94 
 
 2.11 
 
 2.25 
 
 2.39 
 
 2.52 
 
 2.64 
 
 2.77 
 
 6 
 
 .767 
 
 .867 
 
 1 
 
 1.17 
 
 1.34 
 
 1.52 
 
 1.69 
 
 1.83 
 
 1.95 
 
 2.07 
 
 2.18 
 
 2.29 
 
 2.41 
 
 8 
 
 .656 
 
 .742 
 
 .856 
 
 1 
 
 1.15 
 
 1.30 
 
 1.44 
 
 1.57 
 
 1.67 
 
 1.77 
 
 1.87 
 
 1.96 
 
 2.06 
 
 10 
 
 .572 
 
 .648 
 
 .746 
 
 .882 
 
 1 
 
 1.13 
 
 1.26 
 
 1.37 
 
 1.45 
 
 1.54 
 
 1.63 
 
 1.71 
 
 1.80 
 
 12 
 
 .503 
 
 .570 
 
 .657 
 
 .768 
 
 .881 
 
 1 
 
 1.11 
 
 1.20 
 
 1.28 
 
 1.36 
 
 1.44 
 
 1.51 
 
 1.58 
 
 14 
 
 .455 
 
 .515 
 
 .594 
 
 .694 
 
 .795 
 
 .904 
 
 1 
 
 1.09 
 
 1.16 
 
 1.23 
 
 1.30 
 
 1.36 
 
 1.43 
 
 16 
 
 .419 
 
 .475 
 
 .547 
 
 .639 
 
 .733 
 
 .832 
 
 .921 
 
 1 
 
 1.07 
 
 1.13 
 
 1.19 
 
 1.25 
 
 1.32 
 
 18 
 
 .393 
 
 .445 
 
 .513 
 
 .600 
 
 .688 
 
 .781 
 
 .865 
 
 .938 
 
 1 
 
 1.06 
 
 1.12 
 
 1.17 
 
 1.24 
 
 20 
 
 .370 
 
 .420 
 
 .485 
 
 .565 
 
 .648 
 
 .735 
 
 .814 
 
 .884 
 
 .941 
 
 1 
 
 1.05 
 
 1.11 
 
 1.16 
 
 22 
 
 .351 
 
 .397 
 
 .458 
 
 .535 
 
 .614 
 
 .697 
 
 .771 
 
 .838 
 
 .593 
 
 .948 
 
 1 
 
 1.05 
 
 1.10 
 
 24 
 
 .335 
 
 .379 
 
 .437 
 
 .510 
 
 .585 
 
 .665 
 
 .735 
 
 .799 
 
 .851 
 
 .904 
 
 .954 
 
 1 
 
 1.05 
 
 *26 
 
 .318 
 
 .361 
 
 .415 
 
 .486 
 
 .556 
 
 .633 
 
 .700 
 
 .760 
 
 .810 
 
 .860 
 
 .908 
 
 .951 
 
 1 
 
 A recent instructive series of tests* have been conducted by W. K. Hatt in 
 the Laboratory for Testing Materials of Purdue University, in cooperation 
 with the Wood Preservation Committee of the American Railway Engineering 
 Association and with the following organizations: 
 
 Big Four Railroad Company. 
 
 Illinois Central Railroad Company. 
 
 American Creosoting Company. 
 
 Ayer and Lord Tie Company. 
 
 Atchison, Topeka and Santa Fe Railway Company. 
 
 Forest Service, U. S. Department of Agriculture. 
 The principal results of these tests are shown in Table XLI. One of the main 
 determinations of the tests was of the resistance of the ties to the direct pressure 
 of the rail. It was shown that the various treatments had not weakened the ties, 
 except in the case of ties newly treated with crude oil. 
 
 The tie was put through a planer, so that one surface was true. The 
 other surface was adzed at the place of bearing to provide a true bearing for 
 the plate representing the bottom of the rail or the tie plate. 
 
 * Fourth Progress Report of Tests on Treated Ties, Proceedings Am. Ry. Eng. & M. of W. 
 Assn., 1910, Vol. 11, Part 2. 
 
170 
 
 STEEL RAILS 
 
 TABLE XLI. — BEARING STRENGTH OF TIES UNDER THE RAIL 
 
 (See Proceedings Am. Ry. Eng. & M. of W. Assn., Vol. 11, Part 2, pp. 853-4) 
 
 Red oak 
 
 Loblolly pine 
 
 Shortleaf pine 
 
 Longleaf pine 
 
 Red gum 
 
 Total 
 
 1. Crude oil, 529. 
 1. Crude oil, 373. 
 
 1. Crude oil, 655. 
 
 Tests were made upon 581 half -ties to determine the relation of the crush- 
 ing strength of the ties with and without the tie plates. 
 
 The fiber stress per unit of area of wood under the tie plate at the elastic limit 
 in the case of the oak are less than those under the rail alone except for plate C 
 (see Fig. 124). Of course, the total load is greater. This is accounted for by 
 the perceptible springing of the tie plates, thus producing a non-uniform pres- 
 
 Ep3T* 
 
 □ D 
 
 y 
 
 ■ 7 /e 
 
 n 
 
 r;: 
 
 1^6 
 
 
 
 
 
 
 tl 
 
 
 
 
 
 7" 
 
 1 . 
 
 L 
 
 
 
 
 
 
 
 i 
 
 ^> 
 
 U 
 
 
 ft 
 
 K 
 
 K 
 
 TJ 
 
 m ^F 
 
 Fig. 124. — Tie Plate Forms used in 
 „ Tests at Purdue University. 
 
 Fiber Stress in Tie Plates at Elastic 
 ?<* Limit: A 37,500 lbs. persq. in.; B68,800 
 lbs. per sq. in.; C 59,800 lbs. per sq. in. 
 
SUPPORTS OF THE RAIL 
 
 171 
 
 sure on the wood under the tie plate. The loads, therefore, carried with the aid 
 of a tie plate, while larger, are not increased in the same ratio as the increase 
 of bearing surface. 
 
 In the loblolly-pine ties and in plate C on red oak, no perceptible spring- 
 ing of the tie plate was observed within the elastic limit of the timber, the load 
 being increased in practically the same ratio as the surface. 
 
 Tie plate A (see Fig. 124), i 7 e inch thick, was permanently bent at the 
 edge of the rail bearing when the test was carried to j-inch compression on oak 
 ties. The yielding was confined almost entirely to the edge of the rail bearing. 
 
 Tie plate B, | inch thick at edge of the rail, was not permanently bent 
 by the same test. It, however, springs as much, or more, than plate A, but 
 the springing was more uniform. Plate B is harder metal, and this would seem 
 to be an advantage in this test. 
 
 The three tie plates were tested under flexure to determine the quality 
 of the metal. The results are shown in Fig. 124. 
 
 In the calculations of the strength of the tie, if we take the strength of 
 the wood as shown by Table XLII, the result will be not far from correct. 
 The working stress at the rail bearing given in the table refers to the allowable 
 stress under the tie plate. 
 
 TABLE XLII 
 
 WORKING STRESSES FOR TIE TIMBER 
 
 Kind of Wood. 
 
 Working Stress (Pounds per Square 
 Inch). 
 
 Compression at 
 Rail Bearing. 
 
 Extreme Fiber 
 
 Stress in Cross 
 
 Bending. 
 
 Oak. . 
 
 500 
 
 325 
 200-250 
 
 1000 
 1200 
 750 
 
 Longleaf pine 
 
 
 Examining the safe load the tie in the track will carry, we have to con- 
 sider two sources of possible failure of the tie: 
 
 1. The compression of the fibers under the tie plate. 
 
 2. The rupture of the tie due to too great a bending moment in the tie. 
 A 6 by 9-inch tie plate gives an area of 54 square inches. Referring to 
 
 Table XLII, we find the permissible load on the tie plate to be 27,000 pounds 
 for oak ties, 17,500 pounds for longleaf yellow pine, and about 11,000 to 
 13,000 pounds for the inferior woods. 
 
 Let us now consider the bending moment in the tie. 
 
172 STEEL RAILS 
 
 If the tie were completely rigid, there would result a uniform distribution 
 of the pressure on the ballast. This is, however, never realized, and there is 
 an unequal distribution of the pressure. 
 
 The tie should be considered as a continuous beam, supporting a vertical 
 load at two points and resting on material which is, within certain limits, com- 
 pressible. Exactly what takes place in the ballast under the loaded tie is of 
 the greatest importance in determining the bending of the tie. 
 
 * M. Coiiard found that the vertical displacements of cross-ties hardly 
 reach three millimeters (f inch), and that they are not proportional to the 
 weights supported. He has concluded from his experiments that "the cross- 
 ties fixed to the rail remain, at certain points, suspended above the ballast, 
 and that right at the rail there is formed under even the best tamped cross- 
 ties some depressions of ballast on the edges of which the cross-tie is supported; 
 that under the passage of a wheel even lightly loaded the cross-ties come in 
 contact with the ballast and deflect to the depth of the depressions." 
 
 Shwedler, Hoffman, Schwald, Riese, and Zimmerman, from the theoretical 
 researches of Winckler, have derived the elastic curve of the tie represented 
 
 Fig. 125. — Elastic Curve of Tie, 7 feet 10.4 Fig. 126. — Elastic Curve of Tie, 8 feet 10.3 
 
 inches long. (After Winckler.) inches long. (After Winckler.) 
 
 by Fig. 125 or Fig. 126, according as the cross-tie was 2 m. 40 (7 feet 10.4 inches) 
 or 2 m. 70 (8 feet 10.3 inches) long. 
 
 Very careful experiments have been made by M. Cuenot on the relative 
 action of the tie and the ballast. f The following record of his tests is taken 
 from Mr. W. C. Cushing's translation of his work: 
 
 "The rails employed were of the type used on the Paris, Lyons and Medi- 
 terranean, either the P. M. type, of a weight of 39 kilograms per running meter 
 (78.6 pounds per yard), or the P. L. M.-A. type, of a weight of 34J kilograms 
 per running meter (69.5 pounds per yard). 
 
 "All my experiments, during nearly three years, have been made, first, on 
 a side track, then on track No. 2 of the line from Mouchard to Bourg, trav- 
 ersed by the express and fast trains, comparatively with oak cross-ties employed 
 on the P. L. M. system, and with composite cross-ties (wood and steel). (See 
 Fig. 127.) Finally, a special track for experiments was laid at the Bourg 
 
 * Revue des Chemins de Fer, July, 1897. 
 
 f Deformations of Railroad Tracks and the Means of Remedying Them. G. Cuenot, 1907, 
 New York. 
 
SUPPORTS OF THE RAIL 173 
 
 station, and there was tested, at the same time as the two types of cross-ties 
 mentioned, the metallic cross-tie in use on the State System. 
 
 "The wooden cross-ties were oak, creosoted, and of the following dimen- 
 sions: 
 
 Length 2 m. 60 (8 ft, 6.36 in.) 
 
 Width m. 22 to m. 25 (8.66 in. to 9.84 in.) 
 
 Depth m. 14 to m. 15 (5.51 in. to 5.91 in.) 
 
 "The composite cross-tie was composed of a metallic skeleton in the form 
 of an inverted trough, provided in the interior with two symmetrical blocks 
 of wood solidly fixed, and leaving between them an empty central space." 
 
 
 
 ; 
 
 - 
 
 
 - 
 
 
 -»»'--- 4 fcno^-JP<f 3 
 
 
 
 r 
 
 
 
 ; j i « 
 
 
 c^ i. 
 
 S 
 
 
 
 
 
 
 
 
 % 
 
 s , |j; 
 
 
 * (_ 
 
 
 
 ■■;; 
 
 
 
 
 
 
 B 
 
 
 
 " ' ! I 
 
 HALF UPPER PLAN 
 
 ££S 
 
 Fig. 127. — Wood and Composite Ties used in Cuenot's Experiments. 
 
 The measuring apparatus for experiments in the static state was as follows: 
 " There were first placed in the surface of the wood cross-ties, screws with 
 square heads distributed over their whole length and giving 15 or 16 fixed 
 points, which were to serve as bench marks for the determination of the defor- 
 mation. A rigid steel rule in the form of a T (Fig. 128) presented, right at the 
 points, whose spacing was the same for all cross-ties, vertical rods terminated 
 by a notch, in which was brought, while resting on the screw with square head, 
 a gage in the form of an inclined plane, whose divisions were calculated in a 
 manner to correspond with a tenth of a millimeter. The inclination of the 
 inclined plane had been so chosen that the interval between two divisions was 
 at least of 2 centimeters (fo 9 o inch), which allowed estimating the tenth of a 
 millimeter with exactness. 
 
174 
 
 STEEL RAILS 
 
 "The rule was fixed in an unchangeable manner to two stakes of strong 
 dimensions, buried in the embankment about 1 m. 10 (3.61 feet), in order to 
 eliminate the influence of the load on the supports of the rule. When the 
 rule was in place, an observer introduced the wedge-shaped gage in the notch, 
 while maintaining it horizontally on the head of the screw, and stopped it at 
 the moment when it commenced to become wedged; he then made a first read- 
 
 --^iiza a^f^-^tt Hw 
 
 a 
 
 Fig. 128. — Measuring Apparatus for Ties under Static Load. (Cuenot.) 
 
 ing on all the points of reference, proceeding, for example, from left to right, 
 then a second, proceeding in the reverse direction, from right to left. The 
 readings made were recorded and the mean taken, which thus gave the actual 
 position of the cross-tie. 
 
 "The vehicle, which served to load the cross-tie considered, was brought 
 up, always taking care to place the same wheels at the same spot with reference 
 to the piece submitted to the test; it was allowed to remain during about 10 
 minutes, and the readings were recommenced. Two successive readings were 
 made, and the mean of them was taken. The difference between the inscribed 
 means gave the deformation of a cross-tie under the load considered. 
 
 "The measuring apparatus (see Fig. 129) for the experiments in the 
 dynamic state was essentially composed of a stylus arranged in a stable manner 
 at the face of a plate of smoked glass and fixed on the points of the cross-tie 
 under observation. The black smoke deposited on the glass plate, which was 
 displaced at the same time and by the same amount as the points, was re- 
 moved by the point of the stylus ; the height of the part removed gave the value 
 of the deflection, or of the raising, of a cross-tie at the points considered. The 
 
SUPPORTS OF THE RAIL 
 
 175 
 
 reading of this height was made by means of a magnifying glass nearly to the 
 tenth of a millimeter. 
 
 "The stylus, with a flat point of tempered steel, was mounted on a very 
 flexible spring, which could be approached to or removed from the glass plate 
 at will, with the aid of a thumbscrew. The glass plate was fixed by screws 
 on one of the faces of a cross-tie, then smoked in a flame of a candle at the 
 moment when it was desired to put it in service. The thumbscrew passed 
 through an iron rod and simply 
 rested on the spring, which, left 
 free, moved back and forth on 
 the rod fixed by means of two 
 bolts on a stake deeply buried in 
 the soil. 
 
 "In order to make an ob- 
 servation, the screw is pressed 
 against the spring until the point 
 of the stylus comes in contact 
 with the blackened plate. In 
 this position a light blow is given 
 to it, which makes it oscillate 
 and defines a horizontal trace 
 of 2 or 3 millimeters' dfo to 
 iVo inch) length on the black 
 smoke, which forms the reference 
 mark. 
 
 "At this moment one can 
 either place the vehicle on the 
 cross-tie, or allow trains at speed 
 to pass over it. The height of the part of the glass plate rubbed off by 
 the point of the stylus gives, above the reference mark, the values of the 
 depression, and below, the uplift, of the cross-tie. The latter is always 
 inferior to the former; for the flexure is important in comparison with the move- 
 ment of uplift of this piece under the influence of loads at a distance. The 
 successive influence of each of the axles cannot be noted, but it is solely a maxi- 
 mum indication which is produced." 
 
 The results of the dynamic tests were always slightly less than those 
 obtained from the static tests. Fig. 130 shows the general results from the 
 large number of experiments made. 
 
 POSITION OF 5 REFERENCE MARKS R 
 THE LENGTH OF A CR055 TIE, IN PLAN. 
 
 Fig. 129. — Measuring Apparatus for Ties under 
 Dynamic Load. (Cuenot.) 
 
STEEL RAILS 
 
 WOOD TIE 
 
 to^l*****™**,. 
 
 •«_ 15.75". >L_ 15.75: 
 
 ._l5.75"--»U-_ 15.75". 
 PART TAMPED WITH TAMPING BAR PART TAMPED V 
 
 SPONTANEOUS 
 
 Fig. 130. — Results of M. Cuenot's Tests on Ties. 
 
 The following conclusions have been drawn by M. Cuenot from his ex- 
 periments : 
 
 " (a) The long ties, 8 feet 6.36 inches to 7 feet 6.6 inches, take, under the 
 load, the form of a basin with the bottom slightly raised in the center. 
 
 " (6) The short ties, 7 feet 0.6 inch to 6.5 feet, are deformed according to 
 a curve, convex or otherwise, and inclined toward the extremity. 
 
SUPPORTS OF THE RAIL 177 
 
 " (c) The ties between 7 feet 0.6 inch and 7 feet 2.64 inches are lowered 
 parallel with themselves without sensible curvature. 
 
 "(d) The unsymmetrical tamping raised the curve towards the center; a 
 very feeble lack of symmetry reacts very clearly in this direction. 
 
 "(e) It is possible, by increasing the rigidity of a cross-tie, notably by 
 concentrating the material about the supports, to reduce its sinking to the 
 quantity which is intended as a limit, and its flexure in such measure as one 
 would wish. 
 
 "(f) The permanent sinking of the ballast is variable according to the 
 case, but the elastic sinking, the only one there is reason to consider, is, so to 
 speak, constant, whatever be the length and type of the cross-tie adopted. The 
 deformation is slowly produced and augments with time." 
 
 It is here seen that a tie 8 feet 6 inches long, which is the usual length of 
 tie employed in this country, under proper conditions of tamping, will assume 
 
 Fig. 131. — Strain Diagram of Entire Tie. 
 
 the loaded position shown in Fig. 131. In the figure, the loads IF at A and 
 C represent the load at each rail. 
 
 Considerations of the tamping under the tie will not admit of any exact 
 mathematical formula for the distribution of the pressure of the ballast. If, 
 however, as a working hypothesis, we assume that the pressure of the ballast 
 is uniformly distributed between the rails and that the pressure is similarly 
 uniformly' distributed, but of greater intensity from the rail to the end of the 
 tie, and, further, that the tangent to the elastic curve of the loaded tie is hori- 
 zontal under the points of support of the rail, then approximate formulae can 
 be readily derived for the maximum bending moment in the tie and the greatest 
 intensity of pressure of the ballast in terms of the load on the tie at either rail 
 bearing. 
 
 This assumption, while not taking into account all the conditions of the 
 loading of the ballast nor giving the exact distribution of pressure under the 
 tie, will, nevertheless, when applied to an 8-foot 6-inch wood tie, afford a means 
 of determining the maximum bending moment and the greatest intensity of 
 pressure with sufficient accuracy for our purpose, and very possibly as exactly 
 as the present state of our knowledge of the subject warrants. 
 
178 STEEL RAILS 
 
 Where it is desired to investigate the action of ties in a more thorough 
 manner, the calculations may be proceeded with in the same manner as those for 
 the case where the rail acts as a continuous girder in Article 23. It should be 
 borne in mind, however, that the coefficient of the ballast or the ratio of pressure 
 to sinking is not the same for all parts of the tie, but with proper conditions of 
 tamping is considerably greater under the tie in the region adjacent to the rail 
 bearing. 
 
 Referring to Fig. 132, the moments at the supports A and C are equal 
 and each ^ W'L', where W is the total load uniformly distributed over 
 
 ■j— - 
 
 .iiw' 
 
 Fin. 132. — Strain Diagram of Tie between Rails. 
 
 the span U. The bending moment at the center of the span B is ii W'L'. 
 Therefore, the maximum bending moment occurs at the supports and is 
 M m = -h W'L'. 
 
 The free part of the tie outside the rail acts as a cantilever; the maximum 
 bending moment occurs at C and is M c = ^ W'L" , where W" is the total 
 load uniformly distributed over the span L" (Fig. 133). 
 
 Fig. 133. — Strain Diagram of Tie outside of Rails. 
 
 Considering the tie as a whole, Fig. 131, we have, from the principle of the 
 continuous girder, M c = j% W'L' = \ W"L", but for an effective length of the 
 tie of 100 inches, L' = 60 inches and L" = 20 inches; 
 therefore iV W 60 = \ W" 20 
 
 W' =2 W". 
 
 But the reaction at any support is equal to the algebraic sum of the shear 
 to the right and left of the support, and 
 
 W = J cl + J CT 
 W = \W' + W" 
 w = I W' + \ W' 
 W = W' = 2 W", 
 where 
 
 W = the load at either rail bearing, 
 
 J cl = the shear immediately to the left of C, 
 
 J„ = the shear immediately to the right of C. 
 
SUPPORTS OF THE RAIL 
 
 179 
 
 Substituting the value of W in the equation for the maximum bending 
 moment, 
 
 M m = M a = M e , 
 M m = iV W X 60, 
 M m = 5 W, 
 where 
 
 M m = maximum bending moment, 
 M a = bending moment at A, 
 M c = " " " C. 
 
 The extreme fiber stress, / = —f- = — —- , or W = ^— • 
 J I I 5y 
 
 For a rectangular tie, 
 and W = 
 
 y 12 " 2 " 
 30 " 
 
 Turning to Table XLII, we find the allowable extreme fiber stress in bend- 
 ing 1000 pounds per square inch for oak, and 750 pounds per square inch for 
 the inferior woods. We can, therefore, prepare Table XLIII, showing the safe 
 load that the tie can bear and not exceed a proper bending stress. 
 
 TABLE XLIII. — ALLOWABLE LOAD ON TIE AS DETERMINED 
 FROM EXTREME FIBER STRESS IN BENDING 
 
 Kinds of Wood 
 
 Size of Tie. 
 
 Allowable Load 
 Applied at Each 
 Rail Bearing. 
 
 Oak 
 
 Oak 
 
 . Inches. 
 
 7X8 
 
 7X9 
 *Half round 
 
 7X8 
 
 7X9 
 *Half round 
 
 Pounds. 
 
 13,100 
 14,700 
 15,000 
 9,800 
 11,000 
 11,300 
 
 Oak 
 
 
 
 
 17. Bearing on the Ballast 
 
 Considering now the bearing power of the ballast on which the tie rests, 
 the maximum loading on the ballast under the tie per linear inch of the tie, 
 from the preceding article, is 
 
180 
 
 STEEL RAILS 
 
 To express W in terms of the bearing power per square foot of the ballast 
 (p), and the width in inches of the base (6), we have the allowable load per 
 linear inch of the tie equal to 
 
 bp _ W 
 144 2 x 20' 
 
 For bearing on gravel or broken stone, not well confined, three tons per 
 square foot is as much as should be allowed. 
 
 We may now prepare Table XLIV, showing the safe load that can be 
 applied at each rail bearing as determined by the proper load on the ballast. 
 
 TABLE XLIV. — ALLOWABLE LOAD ON THE TIE AS DETERMINED 
 BY THE SAFE LOADING OF THE BALLAST 
 
 Width of Base of Tie. 
 
 Allowable Pressure 
 \pplieil at Each Rail 
 Bearing of the Tie. 
 
 Inches. 
 
 8 
 
 9 
 
 *12 
 
 Pounds. 
 
 13,500 
 15,000 
 20,000 
 
 18. Bearing on the Subgrade 
 
 Before assuming a proper bearing under the tie, an examination must be 
 made of the distribution of the load to the subgrade. The following experi- 
 ments have been made in Germany to determine the distribution of force upon 
 the subgrade.* 
 
 An experimental box, 37 inches long, 20 inches high, and 6 inches wide, was 
 filled with a layer of clay 8 inches high at the bottom, on top of which was placed 
 a layer of sand 6 inches high, and then a layer of gravel 6 inches high, upon 
 which a tie was laid. This tie was tamped with the ordinary tamping pick and 
 then subjected to a load of 57 pounds per square inch, or 8200 pounds per 
 square foot, by which the rail level was depressed. By the use of an eccentric 
 the loading was alternately lifted from the tie and again returned, thus imitating 
 the process of passing a loaded wheel over the track. As soon as the tie had 
 settled 1.2 inches, which was registered upon an attached sliding plate, the 
 tie was again raised and tamped. From time to time photog aphic views 
 and observations as to the stage or condition of the experiment were taken 
 
 * Glaser's Annalen fur Gewerbe und Bauwesen. May, 1899. (Director Schubert.) Transla- 
 tion appears in Proceedings Am. Ry. Eng. & M. of W. Assn., Vol. 7, p. 105. 
 
SUPPORTS OF THE RAIL 181 
 
 by removing the front wall of the experimental box. After the eleventh tamp- 
 ing the experiment was considered as completed, and the section shown in 
 Fig. 134 was taken. 
 
 From this view we can easily see how a short depression, measuring about 
 12 inches to 14 inches wide, has been formed in the clay, with an upward swelling 
 
 Fig. 134. — Ballast Experiments — Schubert. Six inches of sand and 6 inches of gravel. 
 
 on each side. The pressure transmitted from the tie has accordingly distributed 
 itself over this small width when the depth below the bottom of the tie was 
 12 inches. 
 
 In a subsequent experiment, broken stone was used in place of gravel; 
 otherwise the procedure was the same. From a photograph of the section 
 after the fifth tamping (see Fig. 135) a depression in the clay extending nearly 
 
 Fig. 135. — Ballast Experiments — Schubert. Six inches of sand and 6 inches of stone. 
 
 over the entire width of the experimental box (27| inches to 29| inches wide) 
 is noticeable. The distribution of the force is consequently double that of 
 the previous experiment. 
 
 Still more favorable appears this distribution when the height of the stone 
 ballast is increased. In doing this, it is judicious to retain a thin layer of sand 
 so as to prevent the larger pieces of broken stone from entering into the clay. 
 
182 
 
 STEEL RAILS 
 
 As will appear from the section shown in Fig. 136, a depression in the clay has 
 not taken place, and only a few of the broken stones have gone through the 
 sand to the clay. In emptying the box only a very unimportant depression 
 was noticeable. 
 
 - Ballast Experir 
 
 - Schubert. Stone with thin layer of sand. 
 
 Finally, the behavior of a foundation layer was investigated, and after 
 the fourth tamping the section shown in Fig. 137 was taken. The stones of 
 the foundation layer have penetrated the clay rather deep, and not only those 
 in the center, but also stones on the sides, from which we can conclude that 
 
 Fig. 137. — Ballast Experiments — Schubert. Stone resting on clay subgrade. 
 
 the force transmitted through the tie has distributed itself nearly over the 
 entire width of the box. 
 
 Hence, the most favorable distribution of forces is accomplished by the 
 use of ballast of broken stone, with or without a foundation layer. The latter 
 is, however, not suitable in a yielding subgrade, inasmuch as the stones pene- 
 trate into the grade, and the yielding soil will swell into the spaces, thus making 
 the drainage ineffective. 
 
 The effect of overloading the subgrade is very clearly shown in Fig. 138. 
 
SUPPORTS OF THE RAIL 
 
 183 
 
 * The road department of the Pennsylvania Railroad has installed an 
 interesting piece of apparatus on the grounds of the South Altoona foundry 
 to test the bearing qualities of different kinds of roadway and ballast. The 
 
 Fig. 138. — Effect of Overloading the Subgrade. (Am. Ry. Eng. Assn.) 
 
 particular ballast or subgrade to be tested is placed in three heavy boxes that 
 extend across the track and have sufficient depth to serve the purpose. The 
 track crosses this on a level and extends out on either side, terminating in a short 
 
 Fig. 139. — Pennsylvania Track Testing Apparatus. (Railway Age Gazette.) 
 
 and sharp incline. A four-wheel car on this track is loaded with pig metal to 
 obtain any desired weight on the wheels. This car is also equipped with electric 
 motors. A shed built across the track carries an overhead rail, from which a 
 motor current is obtained, and a contact shoe is on the car. Fig. 139 illus- 
 trates the apparatus. 
 
 * Railway Age Gazette, June 11, 1909, and July 21, 1911. 
 
184 STEEL RAILS 
 
 When current is turned on, the car moves out to the end of the conductor 
 rail, and here, as the contact is broken, the power of the motor is shut off. The 
 car runs on until stopped by the adverse grade, and meanwhile a trip reverses 
 the current connections to the motor. Stopped by the grade, the car runs back, 
 beneath the current rail, when its motor drives it to the other end, where the 
 movement is again reversed. In this way the car is made to travel back and 
 forth automatically over the track until the desired results are obtained, the 
 number of trips being automatically registered upon a counter. 
 
 These tests are the most extensive of the kind ever conducted in this 
 country. It was felt that while the data obtained by Mr. Schubert were very 
 instructive yet more valuable data could be obtained from a series of experi- 
 ments if made in a manner more nearly approaching actual track conditions.* 
 
 The track was 109 feet in length, built of new P. R. R. standard 
 85-pound rail with 7-inch by 9-inch by 8§-foot ties spaced 25£ inches center 
 to center. It being impracticable to run the car faster than about five 
 miles per hour, at which speed any effect upon the track, due to impact 
 alone, would be negligible, a weight of 75,000 pounds per axle was chosen 
 for the experimental truck. 
 
 A series of five tests has been completed; the first one beginning on Sept. 
 2, 1908, and the last one ending on Aug. 2, 1910. Table XLV gives general 
 data of the tests. Water was applied by sprinkling the boxes to observe the 
 effect of moisture on the ballast; the amount applied in each test is shown in 
 the table by inches of rainfall. 
 
 In test No. 1 the line of demarkation between the bottom of the ballast 
 and the roadbed material was not straight. The test showed conclusively that 
 a depth of 8 inches of trap-rock ballast, when laid on the usual roadbed ma- 
 terial, was not sufficient to distribute the weight carried by the ties uniformly 
 over the roadbed. 
 
 The results in the third box showed, however, that if 12 inches of per- 
 meable material, such as cinder, were used beneath the 8 inches of ballast, the 
 distribution of the weight over the roadbed material was much better. 
 
 The results of the first test led to the second test to determine how a depth 
 of 12 inches, 18 inches, and 24 inches of trap rock under the ties would behave. 
 In test No. 2 the dividing line between the ballast and the loam was quite 
 straight in box No. 3, but in boxes Nos. 1 and 2 there existed some depressions 
 in the line especially under the rail. 
 
 * Experiments to Determine the Necessary Depth of Stone Ballast. Report of the General 
 Manager's Committee Pennsylvania Railroad. Proceedings Am. Ry. Eng. Assn., 1912, Vol. 13. 
 
SUPPORTS OF THE RAIL 
 
 185 
 
 A study of the sections in test No. 3 showed that the loam was more evenly 
 depressed in box No. 3 than in the other boxes where stone had been substi- 
 tuted for part of the cinder during the test. 
 
 Test No. 4 showed that the gravel and slag distributed the pressure upon 
 the loam with about the same efficiency. 
 
 Test No. 5 was made to determine whether a combination of rock and 
 cinder would prove as satisfactory as the rock alone. It was found, however, 
 that the line between the ballast and the loam in box No. 3 was not as good 
 as in box No. 3 of Test No. 2. 
 
 TABLE XLV. — SUMMARY OF ROADBED TESTS AT ALTOONA. 
 
 Test No. 
 
 j 
 
 2 
 
 3 
 
 * 
 
 5 
 
 
 
 
 
 
 
 
 
 
 1st Part. 
 
 2nd Part. 
 
 1st Part. 
 
 2nd Part. 
 
 1st Part. 
 
 2nd Part. 
 
 Fr r 
 
 Sept. 2, 1908 
 Jan. 5, 1909. 
 
 Apr. 18, 1909 
 June 15, 1909. 
 
 June 28, 1909 
 July 20,1909. 
 
 July 21, 1909 
 Aug. 6, 1909. 
 
 Oct. 19, 1909 
 Nov. 17, 1909. 
 
 Nov. 17, 1909 
 Dec. 3, 1909. 
 
 May 19. 1910 
 June 24,1910. 
 
 June 24, 1910 
 Aug. 2, 1910. 
 
 Material in 
 Box No. 1 
 
 27" bad clay 
 
 12" trap rock 
 26" loam 
 
 24" cinder 
 12" loam 
 
 Removed 8 " 
 
 placed with 
 trap rock 
 
 24" slag 
 12" sandy 
 
 moved and 
 8" trap rock 
 added 
 
 24" cinder 
 13" sandy 
 
 6" cinder re- 
 moved and 
 6" trap rock 
 added 
 
 Box No. 2 
 
 8 " trap rock 
 27" sandy 
 
 18 " trap rock 
 19° loam 
 
 24" cinder 
 12" loam 
 
 Removed 12" 
 placed with 
 
 ^"oaf^ 
 
 12" slag re- 
 moved and 
 12" trap 
 rock added 
 
 13" sandy 
 
 8" cinder re- 
 moved and 
 
 added 5 r ° C 
 
 Box No. 3 
 
 12" cinder 
 
 15" bad clay 
 
 24 " trap rock 
 14" loam 
 
 24" cinder 
 12" loam 
 
 Unchanged 
 
 24" sandy 
 12" sandy 
 
 12" gravel re- 
 moved and 
 
 rock added 
 
 13" sandy 
 
 10" cinder re- 
 
 10" trap 
 rock added 
 
 No. of round 
 
 Settlement: 
 Box No. 1 
 Box No. 2 
 Box No. 3 
 
 Rainfall ■ 
 
 81,600 
 
 loft" 
 
 w 
 
 98*" 
 
 49,932 
 
 8|" 
 9|" 
 
 111" 
 
 45,500 
 
 % 
 8|" 
 
 40,100 
 2|" 
 
 fv 
 
 7" 
 
 45,561 
 
 124" 
 
 134" 
 
 8|" 
 
 40,060 
 
 21" 
 6" 
 
 19,210 
 
 1" 
 2i" 
 
 93,094 
 14 2 " 
 
 For computing the bearing on subgrade we are furnished a method * by 
 Mr. Thomas H. Johnson, who has made a study of Director Schubert's report 
 with a view to deriving a formula which would show the thickness of ballast 
 necessary to produce an equal distribution of the axle loads on the surface 
 of the roadbed underneath the ballast. 
 
 Referring to Fig. 140, the following formulae are suggested by Mr. Johnson: 
 For gravel, ab = x = b' + § d'. 
 For stone, ab = x = b' + d'. 
 
 The relatively small arcs will approximate to parabolas and may be con- 
 sidered as such. 
 
 * Proceedings Am. Ry. Eng. & M. of W. Assn., 1906, Vol. 7. 
 
186 
 
 STEEL RAILS 
 
 The intensity of pressures is proportional to the ordinates of the curve. 
 
 Areas of parabolic segment = %xy\ hence, mean ordinate = f y, or mean 
 pressure = § maximum pressure, or maximum pressure equal to f mean pressure. 
 
 Pressure at b = 0; hence, to obtain an approximately uniform distribution 
 over the surface of roadbed, the tie spacing S must be such that the curves 
 overlap and have a common ordinate, y' = \y. This will occur when db = \cb; 
 or eb = \ ab; or mo = f mn.* 
 
 We should obviously aim to space the tie so that the area of distribution 
 of adjacent ties will overlap and give approximately an equal distribution of 
 the axle loads on the surface of the roadbed underneath the ballast. 
 
 W -^-^- 
 
 Fig. 140. — Distribution of Pressure to Subgrade. (Johnson.) 
 
 With a tie spacing of 23 inches centre to centre of ties, by applying Mr. 
 Johnson's formulae, we find that it will be necessary to use 45 inches of gravel 
 ballast and 22 inches of stone ballast under the tie to obtain equal distribution 
 on the subgrade. 
 
 Tie spacing, 
 
 S = 23 inches = f x. 
 
 
 For gravel, 
 
 S = ! (6' + \ d'), 
 
 
 or 
 
 S = | V + | d', 
 
 \d' = S -lb', 
 
 
 and 
 
 d' = | (S - f b') = | (23 inches - f 8 inches) 
 
 = 45^ inches. 
 
 For stone, 
 
 S =!(&' + d'), 
 
 
 or 
 
 S = | V + f d', 
 
 
 and 
 
 d' = §(S-f&') = | (23 inches - f 8 inches) = 
 
 * Approximate; to be exact, db = 0.29 cb and mo = 0.71 mn. 
 
 = 22-f inches. 
 
SUPPORTS OF THE RAIL 187 
 
 It will be seen that unless an excessive depth of ballast is used, a uniform 
 distribution of pressure on the subgrade will not be obtained. However, if the 
 maximum pressure on the subgrade does not exceed its allowable bearing power, 
 the fact that it is not uniformly distributed will not necessarily prove detrimental. 
 
 From the above, we see that the maximum pressure = | mean pressure; 
 
 W 
 
 but from article 17 the mean pressure is tt— > and the maximum pressure is, 
 
 therefore, 
 
 3 W_ W 
 2 40 x 27 x 
 Substituting the value of x for stone and gravel ballast, we have the maxi- 
 mum pressures: 
 
 Gravel ballast, ^W+W) ' (1) 
 
 Stone ballast, ^jfyj-y (2) 
 
 We may take the depth d' for gravel ballast as 18 inches and for stone as 
 12 inches. Equations (1) and (2) will, therefore, reduce to: 
 
 Gravel ballast, ^L_, (3) 
 
 Stone ballast, W * +W) - (4) 
 
 Equations (3) and (4) are, then, the expressions for the maximum pressure 
 on the subgrade per square inch, in terms of the load on the tie and the width 
 of the base of the tie. 
 
 For bearing on clay foundation, subject to frost and usually made ground, 
 1 to 1| tons per square foot is good practice. Therefore, putting equations (3) 
 and (4) equal to the bearing power of the subgrade, we can obtain the value of W. 
 
 Gravel ballast, ^ = on ™ m , 
 
 144 27 (&' + 9) 
 
 c . , n i 3000 W 
 
 Stone ballast, — , —-_ ■ 
 
 From which 
 
 Gravel ballast, W = 563 (&' + 9), 
 Stone ballast, W = 563 (&' + 12), 
 
 Note. — Professor Talbot is now engaged on tests at the University of Illinois having for their 
 purpose the determination of the distribution of pressure in gravel. These tests are not complete, but 
 the evidence produced so far appears to indicate that the pressure under the center of the tie as shown 
 in Fig. 140 is less than that at the edges, due to a distinct arching effect of the material under the tie. 
 A very great difference in the distributing power of the sand was noted under different conditions of 
 dampness. These tests when finished will doubtless furnish information of value in reference to the 
 distribution of the rail pressure to the subgrade. 
 
188 
 
 STEEL RAILS 
 
 where W is the safe load in pounds applied to the tie at the rail bearing and 
 b' is the width of the tie at its base in inches. 
 
 Table XLVI shows the value of W for the different ties under consideration. 
 
 TABLE XLVI. — ALLOWABLE LOAD APPLIED TO TIE AT RAIL BEARING 
 AS DETERMINED FROM BEARING ON SUBGRADE 
 
 
 Ballast. 
 
 
 Width of Tie 
 
 
 
 Allowable Load 
 Applied to Tie at 
 
 
 
 
 Kind. 
 
 Depth below 
 Tie. 
 
 Rail Bearing. 
 
 Inches. 
 
 
 Inches. 
 
 Pounds. 
 
 8 
 
 Gravel 
 
 18 
 
 9,600 
 
 9 
 
 Gravel 
 
 18 
 
 10,100 
 
 12 
 
 Gravel 
 
 18 
 
 11,800 
 
 8 
 
 Stone 
 
 12 
 
 11,300 
 
 9 
 
 Stone 
 
 12 
 
 11,800 
 
 12 
 
 Stone 
 
 12 
 
 13,500 
 
 19. Supporting Power of the Tie 
 Table XLVI I assembles the information given in the previous tables. 
 
 TABLE XLVII. — BEARING POWER OF TIES IN THE TRACK 
 
 ' X (6, 12) Inches 
 S Feet 6 Inches. 
 (Half Round.) 
 
 Allowable load, in pounds, applied at bearing of 
 
 rail on tie, as determined by: 
 Bearing of tie plate, 
 
 Oak 
 
 Longleaf yellow pine 
 
 Inferior woods 
 
 Bending of tie, 
 
 Oak 
 
 Inferior woods 
 
 Bearing on ballast 
 
 Bearing on grade, 
 
 18-inch gravel ballast 
 
 12-inch stone ballast 
 
 27,000 
 
 17,500 
 
 j 11,000 
 
 I 13,000 
 
 13,500 
 9,600 
 
 27,000 
 17,500 
 11,000 
 13,000 
 
 27,000 
 17,500 
 11,000 
 13,000 
 
 15,000 
 11,300 
 
 11,800 
 13,500 
 
 It would apparently seem that the weakest part of the substructure of the 
 track lies in the bearing on the subgrade. In some cases of a very weak sub- 
 grade, as in the muskeg swamps of Canada, it has been found necessary to 
 resort to unusual methods of track construction in order to maintain the track 
 in proper condition. Mr. D. MacPherson reported at the January, 1912, 
 meeting of the Canadian Society of Civil Engineers the use of 12-foot ties in a 
 
SUPPORTS OF THE RAIL 189 
 
 stretch of track over muskeg, the resulting cheapening in cost of maintenance 
 apparently fully warranted the extra expense of the large ties. 
 
 If we consider the effect of the dynamic load, it will be noted from the 
 discussion in the previous articles that the sinking of the tie in the ballast under 
 the action of the dynamic load is little, if any, greater than under the static 
 load, although the dynamic load is from 50 to 75 per cent greater in amount 
 than the static load. 
 
 As the calculations of the strength of the track must be made for the greatest 
 load put upon it, which is the dynamic load, it would seem desirable to increase 
 somewhat the safe bearing values given in the table as determined by the bend- 
 ing of the tie and bearing on the ballast and subgrade. We are not warranted, 
 however, in assuming a like increase in strength at the bearing of the tie plate 
 under the action of the dynamic load, as the effect of the moving loads is, in 
 this case, to reduce the strength of the wood. 
 
 Examining Table XLVII as it applies to dynamic loading, it is seen that 
 a bearing value of 14,000 pounds, or 7 tons, on half the tie can probably be 
 taken with safety except in the case of the bearing of the plate on the soft-wood 
 tie. The use of a soft wood, as cedar or loblolly pine, for ties under heavy 
 traffic, with the customary form of plate and fastening in use in this country, 
 is to be discouraged, and the general tendency at the present time is to use a 
 wood better adapted to resist mechanical wear under these conditions. 
 
 The rail in the track acts as a continuous girder, resting upon yielding 
 supports. Evidently, therefore, not only must the allowable safe load on each 
 tie be determined, but the yielding of the tie under the pressure of the rail must 
 as well be considered. 
 
 The relation of the bearing power of the tie to the amount it is depressed 
 in the ballast is not thoroughly understood. 
 
 The German engineers, Weber, Winckler, and Zimmerman, have advanced 
 the theory that the pressure, P, of the ballast per unit of surface of the cross 
 tie which it supports is, at each point, in direct ratio with the sinking, Y, of 
 the latter; or P = CY when C is a coefficient depending upon the character 
 of the ballast. The researches of these engineers may be summed up as follows : 
 
 (a) The results of experiments are stated to agree quite closely with the 
 supposition that the pressure on the unit of surface is in direct proportion with 
 the measure of the sinking. 
 
 * (6) With a subsoil supposed to be good, the magnitude of the coefficient 
 of ballast has been found : for gravel ballast (without metalled bed) C = 3 ; for 
 
 * P in kilograms per square centimeter; Y in centimeters. 
 
190 
 
 STEEL RAILS 
 
 gravel ballast (with metalled bed) C= 8; for ballast of small stones and scoriae 
 C= 5. 
 
 (c) The sinking observed under a load in motion, at speeds varying from 
 40 to 60 kilometers (24.85 to 37.28 miles) per hour, was not much greater than 
 the sinking observed under the same load in a state of repose. 
 
 Here again the fact that we are dealing with a dynamic load must be borne 
 in mind, and at high speeds, when the dynamic augment of the wheel load is 
 greatest, the bearing value of the tie corresponding to a given depression in 
 the ballast is largely increased. 
 
 The amount the tie is depressed in the track may be judged from the follow- 
 ing evidence. 
 
 Dr. P. H. Dudley gives from 0.2 inch to 0.4 inch as the amount the 
 general running surface of the rail is below the trackman's surface. Director 
 Schubert states that a wooden tie is depressed from 0.3 inch to 0.4 inch 
 before it reaches a solid bearing. M. Couard observed that the maximum de- 
 pression of the tie was about 0.12 inch, and states that the amount of depression 
 is not proportional to the load. 
 
 Fig. 130 shows M. Cuenot's tests in which a depression is left under the 
 tie of about 0.04 inch and the loaded tie is depressed about 0.12 inch. 
 
 * Fig. 141 illustrates an apparatus used by Bell for measuring and recording 
 the deflection of the rails at various speeds. The following were the results 
 obtained by the passage of a train in which the weights were: 
 
 
 Tons. 
 
 Cwt. 
 
 Locomotive, running weight.. 
 
 Tender 
 
 Total weight of six carriages. 
 
 46 
 
 33 
 
 22 
 
 12 
 18 
 
 7f 
 
 Speed of Train. 
 
 Vertical Deflection. 
 
 Miles per Hour. 
 
 4.2 
 14.9 
 26.7 
 40.4 
 57.1 
 65.2 
 
 0.25 
 
 0.25 
 0.27 
 0.25 
 0.33 
 0.30 
 
 The depression of the tie in the ballast is very erratic. Table LIX shows 
 that in the tests made by the United States government on the depression of 
 
 * The Development of the Manufacture and Use of Rails i 
 Inst, of Civil Engrs., Vol. CXLII, April, 1900, p. 133. 
 
 Great Britain, Bell. Proceedings 
 
SUPPORTS OF THE RAIL 191 
 
 rails the mean depression, under the drivers of an engine having axle loads of 
 
 44,000 pounds, was as follows: 
 
 60-pound rail 073-inch deflection 
 
 70-pound rail 138-inch deflection 
 
 85-pound rail 233-inch deflection 
 
 All of these depressions were obtained in gravel ballast with static wheel loads 
 
 and give results the reverse of what might have been expected. 
 
 Fig. 141. — Bell's Apparatus for Measuring Depression of the Track. 
 
 Fig. 142 shows the relation of the depression to the pressure on the tie. 
 The dotted lines give Zimmerman's coefficients 3 and 8 and the dash line that 
 suggested by Freeman's discussion in Article 21. These curves are straight 
 
 LOAD PER LINEAR INCH UNDER ONE LI N E OF RAI LS 
 (TIES SPACED 20IN) TONS (2000 LBS) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ --rv 
 
 
 I '>-■>> 
 
 - v ->,' u , a 
 
 
 
 
 
 
 
 
 :--! 
 
 1 '*<;>' N 
 
 
 
 
 IH v-N 
 
 
 1 
 
 x %± ' 
 
 
 .^Zx - 
 
 
 j^,s^ _ 
 
 
 * ^ ■ - :: 
 
 
 
 
 - r r v^ 
 
 
 __~_.V_S 
 
 
 
 
 
 
 ^e^: "vi'v " 
 
 
 
 
 : : -^* .v!S " 
 
 
 - - s \ ; 
 
 
 __ -_ :*_:!. .. 
 
 LOAD ON ONE HALF OF TIE TONS (2000 LBS) 
 Fig. 142. — Reaction of Tie. 
 
 lines plotted through the origin, this appears to be Freeman's assumption, but 
 in the case of Zimmerman's analysis, owing to different parts of the tie depress- 
 ing unequally, some variation should probably be made from a straight line. 
 
192 STEEL RAILS 
 
 It is quite evident that under the tie at the rail there is formed a depression 
 of ballast, that even under a comparatively light pressure the tie deflects to 
 the depth of this depression, and that from this moment only is the relation of 
 the deflection to the. load of importance. 
 
 From Mr. Love's analysis of the Government rail experiments (Article 23) 
 we are furnished with a means of determining the relation between the pressure 
 and deflection after the tie comes to a bearing in the ballast. The points plotted 
 in Fig. 142 are obtained from Mr. Love's diagrams and represent the reaction 
 of the tie referred to the depression measured from the highest point in the 
 elastic curve of the rail between two drivers. The tie is assumed to come to a 
 solid bearing at 0.20-inch depression below the trackman's surface. 
 
 Bearing in mind that these points are obtained from a static load and 
 that as far as the stresses in the rail are concerned the depth the tie depresses 
 before it comes to a solid bearing is of comparatively small importance, we may 
 construct the curve of pressures shown by the full line in Fig. 142. 
 
 It is very probable that what really takes place is shown by the full line in 
 the figure. The rail deflects under light pressures in some cases to as much as 
 0.20 inch and the tie comes in contact with a compact bed of ballast and the 
 pressure from this point rises very rapidly in proportion to the deflection. In 
 general it was found from the government tests that the ties in the center of the 
 span between the drivers on light rail were supporting very light loads although 
 in some cases they were considerably depressed in the ballast (see Plate XXIII) 
 and for this reason it appeared that a better knowledge of the action of the 
 ballast would be gained by referring the depression to the highest point in the 
 rail between the wheels rather than to the trackman's surface. In the figure 
 the pressure on the tie under the highest point of the rail between two drivers 
 is plotted with a deflection of 0.20 inch below the trackman's surface and all the 
 other deflections in the span referred to this. 
 
 Above the limits of the experiments, the curve is flattened to provide for 
 more rapid sinking caused by the increased pressure. 
 
CHAPTER IV 
 
 STRESSES IN THE RAIL 
 
 20. Stress at Point of Contact of the Wheel with the Rail 
 
 Passing from an examination of the external forces acting upon the rail 
 to a consideration of the resulting stresses they produce in the material of the 
 rail, let us first examine the stress at the point of contact between the wheel 
 and the rail. 
 
 The essence of the wheel is that its theoretical bearing surface shall be a 
 mathematical line or point, affording no area of bearing surface whatever. 
 In practice this is not strictly the case, owing to the elastic compressibility of 
 the surface, but the bearing surface is always very small, nor can it be increased 
 to advantage by making either the wheel or bearing surface more compressible. 
 To such bearing surfaces the ordinary compression moduli of the textbooks have 
 no application, as they are derived from experiments upon prisms which have 
 the same bearing surface as the greatest section, or nearly so. 
 
 Fig. 143. 
 Condition of Free Flow. 
 
 Fig. 144. 
 Partially Restricted Flow. 
 
 :ion Moduli. (After Johnson.) 
 
 Fig. 145. 
 Restricted Flow. 
 
 * When a plain cylindrical column is subjected to a uniform compression 
 stress over its entire cross section, as in Fig. 143, it may be said to be in a con- 
 dition of "free flow," since it is free to spread in all directions throughout the 
 length of the column. In Fig. 144 the material is compressed uniformly over 
 a small area, as with a die. Here there is a flowing of the metal laterally, and 
 
 * Paper contributed by Professor Johnson to the Engineers' Club of St. Louis, December, 1892. 
 193 
 
194 STEEL RAILS 
 
 then vertically, finding escape around the edges of the die. This is a condition 
 of confined or restricted flow, and evidently the elastic limit here will be much 
 higher than with the simple column. 
 
 In Fig. 145 the surface is compressed by a cylinder, the greatest distortion 
 being at the middle of the area of contact. When this metal is forced to move, 
 or flow, it can find escape only out around the limits of the compressed area. 
 But at these limits the metal is very little compressed, and, hence, must be 
 moved from the center. The confined ring of metal inside the limits of external 
 flow is now much wider, and, hence, the resistance to flow much greater, so that 
 this condition will be found to have a higher elastic limit stress than that shown 
 in Fig. 144, and very much above the ordinary " elastic limit in compression " 
 which is found for the free-flow condition of Fig. 143. 
 
 A careful set of experiments was made by Professors Crandall and Marston* 
 to determine the elastic limits of steel rollers on steel plates. In these experi- 
 ments eleven rollers were employed, from one inch to 16 inches in diameter, 
 with pressures varying from 1000 to 14,000 pounds. Their results showed 
 that the elastic limit load with soft steel rollers on steel plates per linear inch is 
 
 P = 880 D, 
 where P = load in pounds per linear inch of roller, 
 
 D = diameter of roller in inches. 
 
 Professor Johnson f experimented to determine the area of contact between 
 locomotive and car wheels and rails. Sections of wheels were mounted in a 
 100,000-pound Riehle testing machine and short sections of rail were placed 
 in the machine so that the wheel treads rested upon them in a normal position. 
 They were then loaded with 5000-pound increments from 5000 to 60,000 
 pounds, the area of contact being measured after each loading. These actual 
 areas of contact are given two thirds actual size in Fig. 146, and in Fig. 147 the 
 areas are plotted with the area of contact as abscissa and the loads as ordinates. 
 
 Professor Johnson states that no permanent distortion was noted upon either 
 rails or wheels at the contact surface up to the 60,000-pound limit. It is seen 
 from Fig. 147 that these areas plot practically upon a straight line through the 
 origin, indicating that the area is directly proportional to the load. This being 
 true, it must follow that the load divided by the area of contact, or the average 
 stress per square inch over the area of contact, is a constant for all loads. This 
 constant is something over 80,000 pounds per square inch. 
 
 * Friction Rollers by C. L. Crandall and A. Marston. Trans. Am. Soc. of Civil Engrs., August, 
 1894, Vol. XXXII, pp. 99-129. 
 
 t Discussion of Crandall and Marston's paper on Friction Rollers. Trans. Am. Soc. of Civil 
 Engrs., September, 1894, Vol. XXXII, pp. 270-273. 
 
STRESSES IN THE RAIL 
 
 195 
 
 * Mr. Fowler, in his experiments on the relation of the load on the wheel 
 to the area of the spot, found : 
 
 Load on Wheel. 
 
 Area of Spot. 
 
 Pounds. 
 
 Square Inch. 
 
 6,000 
 
 0.11 
 
 10,000 
 
 0.12 
 
 11,500 
 
 0.13 
 
 14,500 
 
 0.15 
 
 16,500 
 
 0.17 
 
 17,500 
 
 0.18 
 
 19,000 
 
 0.19 
 
 25,000 
 
 0.20 
 
 The following conclusions are drawn by Mr. Fowler from tests on the 
 contact areas between wheels and rails: t That the average pressure on the 
 metal in wheel and rail is within the safe limits at low loads, but at a load of 
 20,000 pounds the elastic limit is reached and permanent set begins in the rail; 
 
 DIRECTION ALONG RAIL 
 
 2*3 r- 
 
 fiioO 
 
 20000 30000 40000 
 
 fooOOOOOO 
 
 5000 lOOOO 20000 30000 40000 50000 lbs. 
 
 Fig. 146. — Area of Contact between Wheel and Rail. (Johnson.) 
 
 that the accumulated pressure at the center of the contact area is excessive 
 with comparatively small loads, and is only prevented from doing injury by 
 the support of the surrounding metal; that the effect of difference in diameter 
 in wheels under the same load is insignificant and only appreciable when the 
 difference is great; that a hard and unyielding cast-iron wheel damages the 
 rail more than a steel wheel, and the wear of the rail will be greater with cast- 
 iron than with steel wheels. 
 
 * Proceedings Pittsburg Railway Club, November, 1907. 
 
 t Bulletin of the International Railway Congress, London and Brussels, 1908, pp. 651-663. 
 G. L. Fowler, Contact Areas between Wheels and Rails. See also G. L. Fowler, The Car Wheel 1907, 
 p. 161. Giving the results of a series of investigations made for the Shoen Steel Wheel Co. 
 
196 
 
 STEEL RAILS 
 
 * Honigsberg describes a method proposed to be applied to measure the 
 actual forces between the wheel and the rail. This is based on the fact that 
 polished surfaces of iron or steel show peculiar markings — sometimes known 
 as Luder's Lines — on the limit of elasticity being exceeded. As the limit of 
 elasticity can be artificially raised to any value between the primitive elastic 
 limit and the breaking strength, this gives a means of making standard test 
 LOAD ON WHEEL IN POUNDS 
 
 Fig. 147. — Relation between Areas of Contact and Load on Wheel. (After Johnson.) 
 
 pieces; since when lines appear it may be concluded that the artificially raised 
 limit has been exceeded. 
 
 If a wheel passing over two calibrated pieces of metal causes the lines to 
 appear on one and not on the other, it may be concluded that the actual stress 
 caused by the wheel lies between the elastic limits of the two standard pieces. 
 
 Tire wear would seem to indicate that the elastic limit of the metal was 
 exceeded or too closely approached. The investigation of a committee of the 
 Master Mechanics Association in 1895, on the wear of locomotive tires, has thrown 
 interesting light on this subject. 
 
 * Measurement of Forces between Rail and Wheel. 
 des Eisenbahnwesens, Wiesbaden, 1904, pp. 109-160. 
 
 O. Honigsberg, Organ fur die Fortschritte 
 
STRESSES IN THE RAIL 197 
 
 * Fig. 148 shows a diagram of the average wear of the tires of the 
 fifty-three ten-wheel engines for which the calculations plotted in Fig. 7 
 were made, and Fig. 149 shows the same data of the eight-wheel engines 
 shown in Fig. 6. 
 
 The lower diagrams in Figs. 148 and 149 show the ratio of the rotative 
 force to the weight on the rail, which we may call the " coefficient of slip." 
 Since the coefficient of slip is the rotative force at the rail divided by the total 
 weight of the drivers on the rail, it is evident that as this coefficient increases 
 the tendency of the drivers to slip increases, and when it just equals the coeffi- 
 cient of friction between the tire and rail the engine is on the point of slipping. 
 
 The committee of the Master Mechanics Association in its report says: 
 
 An examination of the tire wear shown in Figs. 148 and 149 shows no dis- 
 tinct relation between the worn spots and the curves of maximum pressure 
 of the wheels as given in Figs. 6 and 7. A very clear relation can, however, 
 be traced between the worn spots and the parts of the wheel where the greatest 
 coefficient of slip is combined with the heaviest wheel pressure. 
 
 Local peculiarities of the tire, such as soft spots in it, as well as flat spots 
 caused by slight sliding, affect the final contour of the worn tire, and it is only 
 by taking the average wear at the same point on a large number of tires that 
 the irregularities due to general conditions show themselves with the necessary 
 clearness. 
 
 Referring to the diagram of the average wear of the tire of the fifty-three 
 ten-wheel freight engines shown in Fig. 148, first we will consider the wear of 
 the front and back tires only, as these wheels were overbalanced, the main 
 wheel's being underbalanced, and, on account of the effect of the angularity of 
 the main rod, subject to quite different conditions from the others. 
 
 Directing our attention to the wheels on the right side of the engine, an 
 inspection of the figure shows quite uniformly, in both right forward and back 
 tires, two locations of maximum wear, one beginning at about 160° and attain- 
 ing its maximum at 220° or 230°, the other becoming pronounced at about 
 10° or 20° and attaining its maximum at about 50°. It will also be noticed that 
 both of these low spots are connected from 220° to 50° in the direction of rotation 
 by a portion of the tire much more worn than that portion from 50° to 220°. 
 
 To understand the cause of this irregular wear, it is necessary to bear in 
 mind that there are at least two ways in which driving wheels are slipped : first, 
 when the slipping is slightly but distinctly noticeable, extending through but 
 a small portion of the revolution; second, when the hold on the rail is entirely 
 
 * Proceedings Am. Ry. M. Mech. Assn., Vol. 28. 
 
198 
 
 STEEL RAILS 
 
 MAIN TIRE 
 FRONT TIRE 
 BACK TIRE 
 MAIN TIRE 
 FRONT TIRE 
 BACK TIRE 
 
 RIGHT DRIVERS^, 
 LEFT DRIVERS^ 
 
 WEAR ON TIRES 
 
 o © 
 
 
 
 
 
 J ^^^ V — V 
 
 
 
 
 / v 
 
 
 S \ S 
 
 
 
 
 
 
 
 
 
 COEFFICIENT 
 
 OF SLIP. E 
 
 NGINE JUST 
 
 STARTINO. 
 
 
 
 
 
 
 
 — ^^"v 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .12 
 
 
 
 
 
 
 
 
 
 OH 
 
 
 
 
 
 COEFFICIENT OF SLIP IO MILES PER HOUR. 
 
 COEFFICIENT OF SLIP 4-0 MILES PER HOUR 
 
 =?- 
 
 COEFFICIENT OF SLIP 60 MILES PER HOUR 
 Fig. 148. — Tire Wear, Ten-wheel Engines. (Am. Ry. M. Mech. Assn.) 
 
STRESSES IN THE RAIL 
 
 199 
 
 FRONT TIRE 'd 
 
 BACK TIRE £ £1 
 
 RONT TIRE £ 
 
 > RIGHT SIDE. 
 
 * tf,W" " ////A 
 
 j^wt mm 
 
 / /////////X///////// A/ ///////// 
 
 Vleft side. 
 
 WEAR OF TIRES 
 
 RIGHT DRIVER^. 
 
 'X F r A3 
 
 
 ^^ 
 
 
 y^ 
 
 ~^ 
 
 
 / N 
 
 / \ 
 
 s \ 
 
 
 
 / \ 
 
 / \ 
 
 / \ 
 
 / \ 
 
 
 / \ 
 
 / \ 
 
 / \ 
 
 s V 
 
 
 y w 
 
 V- 
 
 V- 
 
 
 '" 
 
 
 
 
 *^ 
 
 COEFFICIENT of slip engine JUST STARTING 
 
 
 
 
 
 
 
 
 / \ 
 
 
 
 
 y \ 
 
 / \ 
 
 
 
 
 ^ \ 
 
 / v 
 
 
 
 
 / - 
 
 
 \ 
 
 
 51 
 
 / 
 
 
 
 s^ \ 
 
 -0 
 
 
 
 
 
 COEFFICIENT OF SLIP 40 MILES PER HOUR 
 
 
 
 
 
 
 / \ 
 
 
 
 
 / ^ 
 
 
 
 
 / \ 
 
 ^\ 
 
 
 S — ^ ^""~ \. 
 
 / V 
 
 ' X 
 
 
 y 
 
 
 
 — -^^\ , 
 
 
 
 
 
 
 
 
 
 COEFFICIENT OF SLIP 60 MILES PER HOUR 
 
 Fig. 149. — Tire Wear, Eight-wheel Engines. (Am. Ry. M. Mech. Assn.) 
 
200 STEEL RAILS 
 
 broken, and the wheels slip through a number of revolutions, usually turning 
 with considerable velocity. 
 
 The first case, of slipping through but a small part of a revolution, occurs 
 almost without exception on heavy pulls at slow speed, often being seen when 
 an engine is pulling hard on a hill with just enough sand being used to avoid 
 serious slipping, but not enough to prevent a slight slip at points where the 
 rotative force is the greatest. The beginning of slip must occur under these 
 conditions at or near the maximum of the coefficient of slip. Referring to 
 Fig. 148, we find a maximum value of the coefficient of slip at 40° to 50°, and 
 130° to 140° with engine just starting. At 20 miles per hour, the maxima are 
 at 40° and 130°, and at this speed the tendency to slip at 100° is also almost as 
 great as at the other points. The figure shows a small spot following 100° on 
 the front tire, but none is seen on the back. The diagrams on Fig. 7 indicate 
 the cause, as the pressure of these wheels upon the rail at 100° is almost at a 
 minimum and is much less than at 140° to 160°. 
 
 It is also noticeable that the amount of wear following 160° is greater than 
 that following 40° or 50°, for the same reason. This variation in pressure upon 
 the rail increases rapidly with the speed, and Fig. 7 shows very clearly that 
 following 40° the pressure of the front and back wheels on the right side 
 decreases very rapidly, while the reverse is the case following 160°. 
 
 The same conditions as to pressure on the rail occur, for the left-hand front 
 and back wheels, just 90° back of those on the right side, and irregularities of 
 wear produced by the drivers slipping through a number of revolutions at consid- 
 erable velocity should occur on the left wheels at points 90° back of the corre- 
 sponding point on the right wheels: 90° back of 40° is 310°, and 90° back of 220° 
 to 230° is 130° and 140°. Fig. 148 shows the greatest depth of wear of tires of 
 the left front and back wheels to be almost exactly at these points. There is also 
 a small spot worn at 40°, due to the slipping at slow speeds when the influence 
 of the counterbalance is nil. 
 
 The irregularities of wear of the main wheels follow the same law as those 
 of the front and back wheels, but the conditions are considerably modified 
 by the difference in pressures caused by the influence of the angularity of the 
 main rod, and to a less degree from these wheels being under- instead of over- 
 balanced. 
 
 The spots caused by the slight slipping at slow speeds at about 40° and 
 130° should be found in these wheels as in the front and back wheels, unless 
 the accompanying condition of necessary pressure is absent. Fig. 7 shows from 
 16,500 to 17,000 pounds at 40° on the right main wheel, and from 12,700 to 
 
STRESSES IN THE RAIL 201 
 
 17,500 pounds on the left wheel at the same point, indicating greater wear on 
 the right than on the left tire at this point, which the diagram, Fig. 148, shows. 
 The wear at 130° is found in these wheels, but, owing principally to the influence 
 of the angularity of the main rod and partly to the wheels being underbalanced, 
 the conditions of pressure following 130° on the right main wheel are very dif- 
 ferent from those of the right front and back wheels. Fig. 7 shows that the 
 pressure on this wheel is always rapidly decreasing following 130°, instead of 
 increasing, and, consequently, the worn spot at this point extends but a short 
 distance in the direction of rotation. Not so, however, with the left main tire. 
 Here the pressure is always increasing following this point, and the figure shows 
 the great elongation of this spot in the direction of rotation, extending it as 
 far as 210°, while that on the right tire extends only to 165°. 
 
 There still remains to be explained why the heavy spot on the main tire 
 should slightly precede the point of the maximum coefficient of slip at 130°, 
 and why that on the left wheel still farther precedes this point and, in general, 
 is greater than on the right. An inspection of the diagram on Fig. 7 shows that 
 the pressure of the right main wheel on the rail is always greater preceding than 
 following the 130° point. Fig. 148 also shows that the coefficient of slip is high 
 as early as 110° after a speed of 10 miles per hour is attained, and increases 
 but slightly to its maximum at about 130°. Any slipping occurring between 
 110° and 130° will, on account of the pressure, cause a serious spot at this point 
 on the main wheels, which the diagram shows. 
 
 Fig. 148 shows the worn spot under consideration on the left tire, not 
 only elongated in the direction of rotation, which is explained by the difference 
 in pressure in this direction, but also in the opposite direction, extending beyond 
 the 80° point. This is doubtless due to the slight slip caused by the main rod 
 passing the forward center and suddenly thrusting this wheel back an amount 
 equal to the lost motion in the bearing shoe and wedge. The same thing occurs, 
 of course, on the right wheel, and the sharp, but slight, wear following the 350° 
 point shows it quite clearly. On the left wheel, however, this wear is imme- 
 diately followed by the more serious one due to the approach of the maximum 
 point of coefficient of slip from 110° to 130°, and becoming merged into it, both 
 are increased. 
 
 The upper diagram on Fig. 149 shows the wear of the tires on the engine 
 for which the calculations are plotted on Fig. 6 for the eight-wheel engine. This 
 shows in a general way the same characteristics of the average wear for the 
 fifty-three ten-wheelers, shown on Figs. 7 and 148, but is, undoubtedly, affected 
 to a considerable extent by unknowable local conditions. Here the front wheels, 
 
202 STEEL RAILS 
 
 of course, correspond most nearly to the main wheels on the ten-wheeler, and 
 here, as there, the left main tire shows the most serious irregularity of wear. 
 
 The committee presented the following conclusions as a result of its investi- 
 gation, which it should be borne in mind was in connection with much lighter wheel 
 loads than obtain at the present time. 
 
 " There is no doubt locomotive tires wear without slipping, and there should 
 be, and probably is, a portion of the irregular wear due to the pulverizing or 
 crushing action being greater under heavy than light loads. 
 
 " An experiment was made by removing all the overbalance in the counter- 
 balance of an engine, when the irregularities of wear in the main wheel were 
 almost exactly duplicated in location and to a remarkable degree in magnitude. 
 This, together with similar experiments attended by the same general results, 
 leads us to believe that the irregularities of wear of the tire are almost wholly 
 caused by abrasion from slipping, and that the pulverizing of the steel from 
 pressure alone is of secondary importance." 
 
 With the smaller wheels under the cars and locomotive tenders the con- 
 ditions are quite different.* The smaller steel wheel or tires do not render as 
 satisfactory service under heavy loads and high speed as the larger locomotive 
 tires, principally for the reason that the manufacturer is not able to put into 
 the small tire sufficient mechanical work to obtain uniform physical properties 
 for the full circumference of the tire, and in the service portions of the tread 
 thickness. 
 
 The experience with 36-inch steel-tired wheels under locomotive tenders 
 illustrates this point. The steel tender wheels have failed by shelling out, and 
 portions of the metal of inferior physical structure on one-third of the circum- 
 ference have worn so as to make an eccentric tire which has caused such severe 
 impacts on the rail as to require the removal of the wheels after a short service. 
 The average load on these wheels is 18,000 pounds, and the maximum static 
 load 20,400 pounds; many of them give only six months' service, or 30,000 
 miles, after a first or second turning. 
 
 In a paper read at the last International Railway Congress, published in 
 the Bulletin of October, 1911, Dr. P. H. Dudley has proposed a new method of 
 measuring the tonnage service of rails and wheels. He explains that the ton- 
 nage supported by a given portion of the bearing surface of a rail due to a pass- 
 ing wheel is the total load multiplied by the number of wheels passing over it. 
 The tonnage sustained by the metal in the treads of the wheel is the total wheel 
 load multiplied by the number of revolutions, and this tonnage accumulates 
 
 * Railway Age Gazette, December 22, 1911. 
 
STRESSES IN THE RAIL 203 
 
 more rapidly than that of the rail. It is greater also as the diameter is less 
 and the number of revolutions larger, so that the tonnage service of 36-inch 
 tender wheels is much greater than that of the 75- or 80-inch drivers, though 
 the loads on the latter may be much larger. 
 
 The pressure and movement of heavy loads on the rail causes a cold rolling 
 to take place on the head of the rail, which tends to expand the metal and, if 
 the rail were free to move, would cause it to assume a curved form with the 
 head on the convex surface.* As the rail cannot bend, this cold-rolled metal 
 is subjected to a compression stress. A tensile component would be expected 
 in the vicinity of and next below the part which was affected by compression 
 strains, and this has led to a theory of the cause of the oval silvery spots or 
 transverse fissures in the head of the rail observed by Mr. Howard, f These 
 rail fissures which resemble the smooth surfaces of a progressive fracture have 
 so far been largely confined to one steel company. The theory advanced as to 
 their probable cause should not be regarded as final until further substantiated, 
 and many engineers feel that it is not the true explanation. No other adequate 
 reason has been offered as yet to explain their formation, although a careful inves- 
 tigation is being made which will doubtless throw further light upon the subject. 
 
 Mr. Howard J states that: " The flow of the metal of the head, appar- 
 ent to the eye and witnessed very generally in portions of the track, may 
 be taken as evidence of exhausted ductility of the metal. The ability of 
 the steel to elongate, as found in the primitive state of the rail before going 
 into service, is lost by reason of its development, and the rail, at first tough 
 and capable of being bent, is now brittle and will bend only to a limited extent 
 before rupture. 
 
 " The brittleness is due to the flow of the metal at or immediately below 
 the running surface of the head. The structural continuity has not been de- 
 stroyed, as may be shown upon annealing the metal, which effects a restoration 
 in its ability to elongate. A rail from service will not bend well with the head 
 on the tension side, since the surface metal has been subjected to cold flow in 
 advance of its being worn away by abrasion." 
 
 * Mr. Howard found that in the case of a rail, exposed to this action, on cutting off the head from 
 the web, the former sprung into a curved shape with the running surface on the convex side. The 
 deflection at the middle of the length of the piece, 5 feet long, was 0.20 inch. It is probable, however, 
 that some of this curvature was caused by the strains set up in the rail when cooling. 
 
 t Appendix to Report by the Interstate Commerce Commission on Accident to a Lehigh Valley 
 Railroad Train at Manchester, N. J., on August 25, 1911. See also Broken Lehigh Valley Rail, Iron 
 Age, Vol. 88, Part 2, p. 800. 
 
 t Some Causes Which Tend toward the Fracture of Steel Rails. James E. Howard, Journal 
 Association of Engineering Societies, July, 1908. 
 
204 STEEL RAILS 
 
 Removing the surface metal, in the planer, restores the bending quali- 
 ties of the rail, but in this case it is necessary to plane away the metal from the 
 sides as well as from the top of the head, that is, as far down as the cold flow 
 has taken place. 
 
 The difference in the bending qualities of the same rail according to 
 the head being on the tension or compression side is shown by Fig. 150. The 
 
 Fig. 150. — Two Pieces of a Worn 100-lb. Rail after Testing. The 
 upper piece, with head on compression side, bent 21 degrees with- 
 out rupture. The lower piece, with head on tension side, bent 
 4| degrees and then ruptured. (Howard.) 
 
 upper piece of rail in the figure was bent with the head in compression, while 
 the lower one had the head on the tension side of the bend. 
 
 Rails of this series of tests have ruptured with a deflection of only 3 to 5 
 degrees when the head was in tension, but remained unruptured when bent 
 through an angle of 20 degrees or more with the base in tension. After an- 
 nealing these old rails, of exhausted toughness, the bending qualities were 
 restored, after which the rail could be bent in either direction through about 
 the same number of degrees without fracture. 
 
STRESSES IN THE RAIL 
 
 205 
 
 The effect of the exhausted metal in the head is well illustrated by 
 Table XLVIII, which presents some of the results of Kirkaldy's tests 
 on rails. * Kirkaldy states that " the rail appears to have been gradually 
 hardened under the action of the traffic, more especially on the immediate 
 skin or surface, until the steel thereon cracked under the upward flexion of the 
 rail in the regions just over the chairs, or the minute cracks may have been 
 induced by the severe action of the brakes on trains, so to speak, tearing up 
 or disintegrating the surface of the steel." 
 
 TABLE XLVIII. — 
 
 BENDI 
 
 VG TESTS ON 
 
 WORN RAILS. 
 
 (Kir 
 
 kaldy.) 
 
 (Dis 
 
 aneebetwc 
 
 en supports 5 feet, load applied a 
 
 t center) 
 
 
 
 
 Weight 
 
 Dime 
 
 nsion, 
 
 Stress. 
 
 Deflection. 
 
 
 
 Depth. 
 
 Web. 
 
 Elastic. 
 
 Ultimate. 
 
 At 40,000 rT ... 
 Pounds. Ultl 
 
 
 
 
 Pounds. 
 
 Inches. 
 
 Inch. 
 
 Pounds. 
 
 Pounds. 
 
 Inch. Inc 
 
 hes. 
 
 
 Worn rail; 80 lbs. main line, 23 
 Same, tested inverted. 
 
 | 77.62 
 
 5.00 
 
 .65 
 
 1 35,700 
 
 60,780 
 58,240 
 
 0.62 8 
 0.59 4 
 
 19 
 
 Removed uncracked. 
 Snapped. 
 
 New rail. 
 
 Same, tested inverted. 
 
 J 85.41 
 
 5.50 
 
 .68 
 
 ( 44,300 
 1 43,500 
 
 78,840 
 77,350 
 
 0.26 8 
 0.28 6 
 
 
 5 
 
 Removed unciackeJ. 
 Removed buckling. 
 
 Worn rail; 85 lbs., in road 5 
 
 vears with heavv wear. 
 Same, tested inverted. 
 
 ( 84.25 
 
 5.43 
 
 ... 
 
 ( 47,800 
 ( 47,800 
 
 88,130 
 
 0.24 8 
 0.24 6 
 
 
 5 
 
 Removed uncracked. 
 Removed buckling. 
 
 Worn rail; 82 lbs., 10 years, ser- 
 Same, tested inverted. 
 
 j, M 
 
 4.95 
 
 .66 
 
 ( 40,000 
 ( 40,000 
 
 73,130 
 55,260 
 
 0.31 8 
 0.31 1 
 
 
 44 
 
 Removed uncracked. 
 Snapped. 
 
 Worn rail; exposed to brake 
 
 | 74.70 
 
 4.86 
 
 .66 
 
 ( 35,400 
 ( 31,500 
 
 61,440 
 35,400 
 
 0.64 7 
 
 
 36 
 
 Removed uncracked. 
 Snapped, slight flaw. 
 
 
 
 The slipping of the driving wheel of the locomotive when starting a train 
 may cause roughness of the metal of the rail, accompanied by intense heating 
 of the immediate surface metal of the head. In addition to the loss in ductility 
 of the steel by reason of its flow under the wheel pressures, the metal at the 
 running surface is hardened through this action of the wheel. Showers of 
 sparks attend instances of this kind, from which the high temperature acquired 
 by the particles of the steel may be judged of. There follows also a sudden 
 reduction in temperature through conductivity of the cold metal below, which 
 has an effect similar to quenching steel from high temperatures in water or 
 other quenching liquids, and there results a surface hardening of the metal. 
 During this period of hardening the surface metal is placed in a state of intense 
 tension, relief from which is obtained by the development of cracks in the steel. 
 
 A very interesting experiment is reported by Wickhorst f on the flow of 
 metal in the rail head under the wheel load. The test was made to determine, 
 
 * Kirkaldy on Effects of Wear upon Steel Rails, Appendix II. Min. of Proceedings of the Inst, 
 of Civil Engrs., Vol. CXXXVI, January, 1899, p. 166. 
 
 t Flow of Rail Head under Wheel Loads. M. H. Wickhorst, Proceedings Am. Ry. Eng. & M. 
 of W. Assn., 1911, Vol. 12, Part 2, p. 535. 
 
206 
 
 STEEL RAILS 
 
 if possible, what change is made in the microstructure of the head of the rail 
 by the rolling over the rail of heavy wheel loads. At the same time, measure- 
 ments were made of the spread of the head and the width of the bearing produced. 
 The test was made on a new 70-pound Bessemer rail with a reciprocating 
 machine in which a piece of rail is moved back and forth under a wheel to which 
 a load can be applied by means of levers. A diagram of the machine is shown 
 in Fig. 151, from which it is seen that the rail is fastened to a steel bloom 
 
 Fig. 151. — Reciprocating Machine for Testing Flow of Metal in Head of Rail. 
 (Am. Ry. Eng. Assn.) 
 
 which runs on rollers running on another steel bloom that forms the bed of the 
 machine. The rail bed is connected by means of a connecting rod to the bed 
 plate of a planer, which furnishes the power to run the rail machine. The 
 weights applied to the weight hanger are multiplied 600 times, as applied to 
 the axle of the wheel. The piece of rail tested was 12 inches long, which was 
 set up between two other similar pieces, which acted as end pieces onto which 
 the wheel could roll when leaving the piece under test. The piece tested had the 
 sides of the head planed vertical to a width of head of 2 inches. This width was 
 used partly to accentuate the test and partly to do away with the rounded 
 corner, so as to allow of measuring the width closer to the top of the head, and 
 the sides were made vertical so the measurements could be made satisfactorily 
 with a micrometer along the whole depth of the head. The section of the 
 
STRESSES IN THE RAIL 
 
 207 
 
 original rail and as tested are shown in Fig. 152. Before testing, the width 
 of the head was determined at the top, at the bottom and halfway between, 
 both at the middle of the piece tested and at one end, by means of a micrometer. 
 To determine the sag of the head, two prick-punch marks were put on each side 
 of the rail at the middle of its length, one on the side of the head near its bottom 
 and the other on the top side of the base, about f inch from the web. In order 
 to have a vertical side on which to prick-punch the mark, the base was gouged 
 at the desired place. The distance between the marks was measured in .01 
 
 Fig. 152. — Section of 70-lb. Bessemer Rail Tested for Flow of Head. 
 (Am. Ry. Eng. Assn.) 
 
 inch by means of a fine-pointed toolmaker's dividers and a steel scale reading 
 to .02 inch. 
 
 The test was started with a load of 30,000 pounds applied to tie wre< 
 using 1000 double strokes or 2000 rollings of the wheel over the rail ur_c t 
 test. The bearing assumed a width of .64 inch. The only effect on the width 
 of head was to spread the top of the head .002 inch, and the load was, there- 
 fore, increased at once to 60,000 pounds and the test continued until the head 
 seemed to no longer spread as measured with the micrometer. The width of 
 the head of the rail and the width of the wheel bearing on the rail, st various 
 stages of the test under the load of 60,000 pounds, are shown in Table XI IX. 
 The spread of the top part of the head and the width of bearing in this table 
 
208 
 
 STEEL RAILS 
 
 includes also the effect of the preliminary rolling of 2000 wheel applications 
 with 30,000 pounds. The head did not show any sag throughout the test. 
 
 TABLE XLIX.— ROLLING TESTS ON RAIL HEAD WITH LOAD OF 60,000 POUNDS 
 
 
 Width of 
 
 Spread of Head. 
 
 Wheel 
 Rollings. 
 
 Middle of Rail. 
 
 End of Rail. 
 
 
 Top of Head. 
 
 Middle of 
 Head. 
 
 Bottom of 
 Head. 
 
 Top of Head. 
 
 Middle of 
 Head. 
 
 Bottom of 
 Head. 
 
 200 
 
 2,000 
 4,000 
 23,460 
 32,142 
 
 .92 
 .92 
 .94 
 1.02 
 1.04 
 
 .002 
 .009 
 .010 
 .013 
 .013 
 
 .000 
 
 .003 
 .004 
 .006 
 .006 
 
 .000 
 
 .000 
 .001 
 .001 
 .001 
 
 .004 
 .006 
 .006 
 .008 
 .008 
 
 .002 
 .002 
 .002 
 .003 
 .003 
 
 .000 
 
 .000 
 .000 
 .000 
 .000 
 
 It should be remarked that the wheel was beveled some, and the bearing 
 was, therefore, on one side of the top of the head, remaining throughout about 
 .2 inch from one side and increasing in width toward the other side. 
 
 A microscopic examination was made of the rail at the top of the head and 
 at the center of the head both before and after rolling. While the micro-photo- 
 graphs obtained indicated a slight stratification of the grains in a longitudinal 
 direction, there was little, if any, difference between the specimens before and 
 after rolling. The material tested was good ductile material of medium hard- 
 ness for rail steel, and as the maximum lateral stretch at the top of the head 
 was only .013 inch, much difference in the microstructure could hardly be 
 expected. 
 
 It is evident that the metal in the head of the rail must have a high elastic 
 limit to successfully meet the severe conditions of modern service. This fact 
 was clearly brought out several years ago by one of the writers in connection 
 with some service tests on annealed rails on the Philadelphia and Reading Rail- 
 way. An account of this investigation appears in the Proceedings of the New 
 York Railroad Club, December, 1906. Eleven 90-pound rails were sawed into 
 halves, and one half of each rail was annealed. The carbon content averaged 
 0.54 and the manganese 1.06 per cent. After 88 million tons traffic it was 
 found that the annealed rails averaged 31.9 per cent more wear and they also 
 showed a greater tendency to crush and splinter, but it was found on test that 
 the elastic limit had been reduced over 10 per cent. The annealed rails, in spite 
 of their finer structure and consequent greater toughness, did not wear so well 
 on account of the lower elastic limit. 
 
 In Article 9 attention was called to the effect of the inertia of the track on the 
 stresses produced by impact, and in Article 8 it was shown how the lack of round- 
 
STRESSES IN THE RAIL 209 
 
 ness of the wheel may cause excessive strains in the running surface of the head. 
 These factors make it much more difficult to control this stress than that produced 
 by bending; in the latter the forces acting on the rail can be determined within 
 closer limits and the remedy is easier to apply. 
 
 The rail is in fact called upon to perform two quite distinct functions, one of 
 which is to resist the strain produced at the area of contact between the wheel and 
 the rail and the other to resist the bending stress; the latter can be reduced by 
 increasing the moment of inertia of the section or strengthening the track structure, 
 but the former is in a measure independent of the form of the rail and requires a 
 change in the character of the material of which the rail is composed. 
 
 The defect known as "roaring rails" is caused by an imperfect surface or 
 corrugations in the head of the rail. These corrugations are confined almost 
 exclusively to the rails used on electric roads and few problems, confronting the 
 maintenance of way engineer of such roads, have attracted the attention and study 
 being given to this trouble. 
 
 There are many conflicting opinions as to the cause of the phenomena, 
 none of which appear adequate to properly explain it.* 
 
 The corrugations of rails in recent years have increased rapidly in number; 
 once they start they rapidly grow worse and it is important to remove them as 
 soon as the indentations appear. This is generally accomplished by means of a 
 rail grinding device which consists either of a carborundum block rubbing over 
 the rail or an emery wheel which grinds the rail to a true surface. 
 
 The cost of removing corrugations in rails varies from a few cents to 50 
 cents per foot of rail, depending on the depth of the waves; fortunately 
 after the corrugations have been removed there is little probability of their 
 ever returning. 
 
 * Some of the recent literature on this subject is as follows: 
 
 Andrews, J. H. M. — Some notes on rail corrugation. 1500 w. 1910. (In Electric Rail- 
 way Journal, Vol. 36, p. 370.) 
 
 Outlines prominent causes of corrugation and presents a few notes and conclusions. 
 
 Busse, A. — Rail corrugation. 1500 w. 1910. (In Electrician, Vol. 65, p. 930.) 
 
 Observations from many points indicate strongly that corrugation is due primarily to defects in 
 the rail metal, resulting from the rolling. 
 
 Panton, Joseph A. • — Rail corrugation. 26 p. 111. 1907. (In Journal of the Institution of 
 Electrical Engineers, Vol. 39, p. 3.) 
 
 Concludes, in summary, that corrugations are caused, directly or indirectly, by lateral play in 
 weak trucks, the weakness being intensified by unsymmetrically driven axles. 
 
 Wilson, C. A. Carns. — Rail corrugation. 3000 w. 111. 1908. (In Engineering, Vol. 86, 
 p. 90.) 
 
 Aims to show conditions under which corrugations are produced. 
 
210 STEEL RAILS 
 
 21. Proposed Solutions of the Bending Stress in the Rail 
 
 Following the path of the forces as they pass through the rail and are 
 distributed to the ties, we find very complex and unstable conditions. The 
 rail is supported on a series of yielding supports. These supports, through 
 their unequal yielding, bring about distributions of stress in the rail that are 
 difficult to calculate. 
 
 Before proceeding further with the discussion of this subject, let us turn 
 to some of the methods advanced for the proper solution of the problem. 
 
 * Mr. 0. E. Selby approaches it in the following manner: " Examining 
 first the bending stress in the rail, we have 50,000-pound axle loads on supports 
 20 inches apart. For these conditions the American Railway Engineering and 
 Maintenance of Way Association specifications for steel bridges, paragraph 5, 
 call for 100 per cent impact, making the stresses equivalent to those from a 
 100,000-pound axle load, or a 50,000-pound wheel load. 
 
 " For a simple beam the bending moment in one rail would be 250,000 
 inch-pounds. For a continuous beam with rigid supports, it would be two- 
 thirds that, and for a continuous beam with partially yielding supports, three- 
 fourths of the bending moment for a simple beam is reasonable, giving 187,500 
 inch-pounds. If we consider the wheel placed over a tie which yields enough 
 to carry one-fourth of the load to each adjacent tie, the resulting moment in 
 the rail is the same. The section modulus of an 80-pound A. S. C. E. rail is 
 10.0, making the extreme fiber stress 18,750 pounds per square inch. For a 
 100-pound rail the unit stress is reduced to 12,800. 
 
 " Passing to the load on the tie, we encounter an element which must 
 vary between rather wide limits with the stiffness of the rail and yielding of 
 the supports. With a simple beam and load placed midway between the sup- 
 ports, the reaction on each support would be one-half the load. The theory 
 of the continuous girder would make the reactions about 55 per cent to 67 per 
 cent, depending on whether the load is placed over a support or midway between. 
 
 " The yielding of supports would undoubtedly reduce these percentages. 
 Bridge specifications usually consider the load equally distributed among three 
 ties, but bridge ties are spaced usually 14 inches between centers, so that, if 
 the load going to one tie is proportioned to the tie spacing, the amount for 
 20-inch spacing would be 20 -=- 14 x 1 -f- 3 = 20 ■*- 42 = 47.6 per cent. There- 
 
 * O. E. Selby, Bridge Engineer, C, C, C. & St. L. Ry. A Study of the Stresses Existing in Track 
 Superstructures and Rational Design Based Thereon. Proceedings Am. Ry. Eng. & M. of W. Assn., 
 1907, Vol. 8. 
 
STRESSES IN THE RAIL 
 
 211 
 
 fore, the assumption that the maximum load on a tie is half the axle load seems 
 a proper one. 
 
 " Modern specifications call for an E-60 loading, which contains 60,000- 
 pound axle loads, spaced 5 feet between centers. A tie spacing equal to one-half 
 the wheel spacing would load the girder (rail) at the quarter points and produce 
 moments (in a simple beam) equal to those produced by a uniform load. 
 
 " A more practicable tie spacing, one-third of the wheel spacing, would 
 similarly produce moments j2 P ar t, or only 1.4 per cent greater than those 
 from a uniform load, so that if we design a rail for a uniform load equal to the 
 wheel load divided by the wheel spacing, the result would be very nearly correct. 
 The wheel load with impact is 60,000 pounds, or 1000 pounds per linear inch 
 of rail. For a continuous beam (indefinite number of spans and loads) the 
 maximum movement is 1 h- 12 x WL 2 = 1 -f- 12 X 1000 X 60 2 = 300,000 
 inch-pounds." 
 
 The following has been presented by Mr. Bland : * 
 
 " The rail acting as a beam under passage of wheel loads is in a condition 
 of ' restrained ' ends, and the maximum moment from a wheel load Q is given 
 by M = ± | Ql, ' Q ' being concentrated load and ' I ' being span center 
 to center of ties. The moment alternates from positive to negative, and alter- 
 nates equally. The dynamic augment to static wheel load is taken 60 per cent 
 for a speed of 75 miles per hour. 
 
 " Assume a static axle load of 60,000 pounds, giving static wheel load of 
 30,000 pounds. The dynamic augment for 75 miles per hour is 60 per cent, 
 making a dynamic wheel load of 48,000 pounds." 
 
 TABLE L. — RAIL STRESSES. (Bland.) 
 
 Rail Weight. 
 
 Tie Spacing, 
 c. toe. =2. 
 
 Minimum Rail 
 Modulus=Z. 
 
 Dynamic Moment. 
 
 Resulting Unit 
 Stresses. 
 
 Pounds per Yard. 
 60 
 
 70 
 85 
 100 
 
 24 
 
 24 
 22 
 22 
 
 6.70 
 8.30 
 11.30 
 15.00 
 
 Inch-pounds. 
 144,000 
 
 144,000 
 132,000 
 132,000 
 
 Lbs. per Sq. In. 
 
 21,500 
 17,350 
 11,680 
 8,800 
 
 The following investigation of the stresses in the rail was suggested by 
 Mr. F. B. Freeman, in a paper presented to the New York Central Lines Main- 
 tenance of Way Committee entitled " Investigation of Stresses in Track Super- 
 structure." This paper is of considerable interest, as the results obtained by 
 
 * J. C. Bland, Engineer of Bridges, Penn. Lines West of Pittsburg, on the Capacity of Modern 
 Heavy Rail for Existing Heavy Engines, 1907. 
 
212 STEEL RAILS 
 
 the methods proposed are compared with the stresses actually given by test 
 of the rails under the moving trains. An abstract of the paper follows: 
 
 By means of plotted curves recently published showing resultant wheel 
 loads in terms of the unit static loads at various speeds, Dr. P. H. Dudley shows 
 as a result of his experiments that with' smooth wheels rolling without accelera- 
 tion, the impact approaches 50 per cent as a limit at 100 miles per hour. A 
 curve of resultant wheel loads at various speeds with acceleration shows 100 per 
 cent impact as the limit at 100 miles per hour. At speeds of 50 to 70 miles per 
 hour the impact appears to be about 75 per cent when the train is under 
 acceleration. 
 
 By acceleration it is meant that the locomotive is exerting its maximum 
 tractive effort at that speed. 
 
 From his stremmatograph tests, Dr. Dudley finds that the maximum 
 extreme fiber stresses in the 100-pound rail under the present Class " I " loco- 
 motive (Atlantic type) sometimes run as high as 22,000 pounds, while with the 
 80-pound rail stresses as high as 28,000 pounds are not uncommon. One thing, 
 however, must be borne in mind : though the extreme fiber stress in steel rails 
 may seem high as shown by test, these maximum stresses are of very short dura- 
 tion, lasting but a small fraction of a second and often reversed immediately. 
 Within certain limits the stress seems to vary directly as a function of the speed ; 
 hence the greater the speed, the greater the stress, but the shorter its duration. 
 
 The present steel rails have an elastic limit between 50,000 pounds and 
 60,000 pounds and an ultimate strength of 110,000 pounds to 120,000 pounds 
 as shown by test. 
 
 If we investigate the stresses which are supposed to exist under the Class 
 "I" locomotives and then compare the resultant stresses with those of tests, 
 we may arrive at a conclusion regarding the trustworthiness of our assumption. 
 
 As the Atlantic type of locomotive (Class " I ") cannot draw an ordinary 
 train at a greater speed than 70 miles per hour with acceleration, we are justi- 
 fied in using 75 per cent impact in investigating the existing stresses under 
 Class " I." 
 
 The Pacific type (Class " K ") and the electric locomotive (Class " T ") 
 each may haul trains at speeds approaching 100 miles per hour,* and the former 
 will be investigated with 100 per cent impact. 
 
 As the wheels roll along the rail there is a general depression of the track, 
 local depressions being greater under the heaviest concentrations, and fairly 
 uniform where the wheel loads are equal and evenly spaced. According to 
 
 * This statement seems open to question; compare with Article 3. 
 
STRESSES IN THE RAIL 213 
 
 Dr. Dudley, this general depression varies from .10 inch to .20 inch, varying 
 with the stiffness of the rail, elasticity of subgrade, and tamping of ballast. 
 Due to the deflection of the rail, there will De a greater depression under the 
 wheels and the lesser depression about midway between them. The tie pressure 
 may be assumed proportional to the tie depression, and from the tie pressure 
 the stress in the rail may be approximated. 
 
 In considering the deflection of the rail, we may consider the rail as a con- 
 tinuous beam being supported by the wheels and having varying concentrated 
 loads applied by the ties. These concentrations decrease toward the center 
 of the span and give an effect practically similar to that of the uniform load, 
 equivalent to the sum of the tie pressure (Fig. 153). 
 
 Fig. 153. — Distribution of Tie Pressure under Rail. (Freeman.) 
 
 In order to approximate the rail deflection, there will be no great error in 
 considering the load of the drivers and second wheel of the first truck as uni- 
 formly distributed from a point between the two wheels of the forward truck 
 to a point midway between the rear driver and the trailer; likewise that the load 
 of the trailer be distributed from a point midway between the rear driver and 
 the trailer to a point midway between the trailer and the first wheel of the tender. 
 
 The load of the drivers of the present Class " I " locomotive, including 
 impact, is 96,000 pounds, which will be distributed over approximately 19 feet, 
 giving an equivalent uniform load of 6400 pounds per lineal foot. Under the 
 trailer the equivalent uniform load will be 4000 pounds per lineal foot, and 
 under the front wheel about 3000 pounds per lineal foot (Fig. 154). 
 
 Under the drivers we have a span of 7 feet and a uniform load of 6400 
 pounds per lineal foot. With the 100-pound rail this loading would cause a 
 deflection of .047 inch and with the 80-pound rail a deflection of .08 inch. 
 
 The average uniform load between the rear driver and the trailer being 
 about 5000 pounds per lineal foot, the deflection of the 100-pound rail becomes 
 .127 inch, while the 80-pound rail deflects .22 inch. 
 
 It will be seen by a glance at Fig. 155 that the tie reactions vary with the 
 stiffness of the rail, being much more uniform with the heavier rail. 
 
STEEL RAILS 
 
 6 
 o 
 
 in 
 
 r 
 
 .Q 7 
 
 O O 
 
 9. ° 
 
 O m 
 
 o o 
 o o 
 O m 
 
 ^i 9 '- 2 " c 
 
 i - 
 o 6 
 
 O 10 
 
 r^ to 
 cu - 
 
 =? 7 J 
 
 o 6 o 
 o o o 
 o o o 
 
 N - S 
 CM CM (0 
 
 "\\0'-7Vz [ . . »9'- 
 
 7 ^ + £ + ^ 
 O Q OO O O O 
 O O OQ OO O 
 mo m :r O - 
 N. oo mo "in w 
 
 CM -^ CM <t ~~ CM — 
 
 O 
 
 8 
 
 L 
 
 y \ly \i'J \L 
 
 
 J 
 
 3700 lbs. PER LIN. FOOT 
 
 4000 lbs. 
 
 P.L.F. 
 
 
 3000 lbs. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 154. — Class "I" Engine with 75 per cent Impact. (Freeman.) 
 
 Fig. 155. — Track Depression under Class "I" Loading. (Freeman.) 
 
 Plotting the most probable deflection of the 100-pound rail (shown in solid 
 line) and the 80-pound rail (shown in broken line), and then taking the tie re- 
 actions as proportional to the ordinates to the curves of deflection, we may 
 expect the following depressions and reactions: 
 
 TABLE LL — TIE DEPRESSIONS AND REACTIONS, CLASS "I" LOADING 
 (FREEMAN) 
 
 Tie. 
 
 100-pou 
 
 nd Rail. 
 
 80-pound Rail. 
 
 Depression. 
 
 Bearing. 
 
 Depression. 
 
 Bearing. 
 
 A 
 
 Inch. 
 0.10 
 
 0.15 
 0.19 
 0.15 
 0.10 
 0.10 
 0.17 
 0.22 
 0.22 
 0.18 
 0.18 
 0.22 
 0.22 
 
 5,300 
 8,100 
 10,200 
 8,100 
 5,300 
 5,300 
 9,400 
 11,800 
 11,800 
 9,700 
 9,700 
 11,800 
 11,800 
 
 0.04 
 0.18 
 0.23 
 0.18 
 04 
 0.04 
 0.17 
 0.2 
 0.24 
 17 
 0.17 
 24 
 0.24 
 
 2,200 
 10,000 
 12,600 
 10,000 
 2,200 
 2,200 
 9,500 
 13,400 
 13,400 
 9,500 
 9,500 
 13,400 
 13,400 
 
 B 
 
 c 
 
 D 
 
 E. . . 
 
 F 
 
 G. . . 
 
 h...: 
 
 I . 
 
 J 
 
 K 
 
 L 
 
 M 
 
STRESSES IN THE RAIL 
 
 215 
 
 If we use these tie reactions and assume with Winckler that the moments 
 in a continuous girder over yielding supports are three-fourths those of a 
 simple beam, we may compute the bending moments and the extreme fiber 
 stress in the rail. 
 
 Taking moments under the rear driver, we find an extreme fiber stress of 
 19,430 pounds with the 100-pound rail and 28,950 pounds with the 80-pound 
 rail. Under the trailer we find 17,750 pounds in the 100-pound rail and 19,800 
 pounds in the 80-pound rail. 
 
 These stresses given for Class "I" are practically those found from the 
 stremmatograph tests. 
 
 According to Dr. Dudley, when the Class "I" type engine strikes a piece 
 of soft or rough track, it begins pitching or rocking about its center of gravity, 
 located between the two drivers. This rocking increases the pressure on the 
 trailer sometimes as much as 50 per cent. This tends to increase the stresses 
 under the trailer by that amount. In the Class " K " type, this difficulty is 
 obviated to a great extent by the increased length of wheel base. 
 
 We may now compare Class " K " with Class "I," using 75 per cent impact 
 and considering the two engines working under similar conditions. 
 
 Following the same analysis as in the investigation of Class "I," we may 
 assume the weight of the drivers and the second wheel of the first truck as 
 uniformly distributed over a length of about 28 feet of rail, giving a uniform 
 load of 6500 pounds per lineal foot. Under the trailer we find a uniform load 
 of 3800 pounds per lineal foot, and under the tender a load of 3700 pounds per 
 lineal foot (Fig. 156). 
 
 Fig. 156. — Class "K" Engine with 75 per cent Impact. (Freeman.) 
 
 Computing the deflections of the rail, we find .049 inch for the 100-pound 
 rail and .083 inch for the 80-pound rail on the 7-foot span under the drivers; 
 on the 10-foot 11-inch span between the rear driver and the trailer a deflection 
 of .24 inch is shown by the 100-pound rail and .42 inch by the 80-pound rail. 
 
STEEL RAILS 
 
 Plotting the curve of probable deflections for the 100-pound and 80-pound 
 rail and using a general average depression under the drivers, proportional to 
 the uniform load under the drivers, we may find the tie depressions (Fig. 157). 
 
 Fig. 157. — Track Depression under Class "K" Loading. (Freeman.) 
 
 The depression under a uniform load of 6500 pounds is taken at .20 inch. 
 Upon this basis we get the following tie depressions and reactions : 
 
 TABLE LIT — TIE DEPRESSIONS AND REACTIONS, CLASS "K" LOADING 
 
 (FREEMAN) 
 
 
 100-pound Rail. 
 
 80-pound Rail. 
 
 Tie. 
 
 Depression. 
 
 Bea 
 
 
 Depression. 
 
 Bearing. 
 
 
 
 1=75 Per cent. 
 
 1=100 Percent. 
 
 
 1=75 Per cent. 
 
 1 = 100 Per cent. 
 
 A 
 
 0.05 
 
 0.05 
 0.14 
 0.20 
 0.20 
 0.14 
 0.05 
 0.05 
 0.13 
 0.23 
 0.23 
 0.18 
 0.18 
 0.23 
 0.23 
 0.18 
 0.18 
 
 3,000 
 3,000 
 
 7,000 
 10,000 
 10,000 
 7,000 
 3,000 
 3,100 
 7,900 
 14,000 
 14,000 
 11,000 
 11,000 
 14,000 
 14,000 
 11,000 
 11,000 
 
 3,400 
 3,400 
 8,000 
 11,400 
 11,400 
 8,000 
 3,400 
 3,500 
 9,000 
 16,000 
 16,000 
 12,600 
 12,600 
 16,000 
 16,000 
 12,600 
 12,600 
 
 0.03 
 0.03 
 0.11 
 
 0.25 
 0.25 
 0.13 
 0.03 
 0.04 
 0.14 
 0.24 
 0.24 
 0.16 
 0.16 
 0.24 
 0.24 
 0.16 
 0.16 
 
 1,560 
 
 1,560 
 5,700 
 13,000 
 13,000 
 6,800 
 1,100 
 1,300 
 9,100 
 15,600 
 15,600 
 10,400 
 10,400 
 15,600 
 15,600 
 10,400 
 10,400 
 
 1,780 
 
 B 
 
 1,780 
 
 C 
 
 6,500 
 
 D 
 
 14,850 
 
 E 
 
 14,850 
 
 F 
 
 7,800 
 
 G 
 
 H 
 
 1,260 
 1,500 
 
 I 
 
 10,400 
 
 J 
 
 17,800 
 
 K 
 
 L 
 
 M 
 
 N 
 
 
 
 P 
 
 Q 
 
 17,800 
 11,900 
 11,900 
 17,800 
 17,800 
 11,900 
 11,900 
 
 Taking moments under the rear driver, we find an extreme fiber stress of 
 22,000 pounds in the 100-pound rail and 32,150 pounds in the 80-pound rail. 
 These stresses may be expected up to speeds of 60 miles per hour. At speeds 
 of 90 to 100 miles per hour the impact will approach 100 per cent and the stress 
 in the 100-pound rail may run as high as 25,200 pounds and up to 36,600 pounds 
 in the 80-pound rail. These are stresses which may be expected in ordinary 
 
STRESSES IN THE RAIL 
 
 217 
 
 main-line track, at maximum speed. On a soft piece of track they may run 
 up 25 per cent higher. 
 
 Zimmerman's analysis of the stresses in the rail * is based upon the stiffness 
 of the rail and tie and on the compressibility of the ballast. For the bending 
 moment in the rail he derives the following equation: 
 
 8y + 7 
 
 in which 
 
 where 
 
 M 
 
 87 + 7 r a 
 "4 T + 10 4" 
 
 4 7 +10 
 
 Mo = Maximum bending moment in the rail, 
 m = Bending moment of a simple beam loaded in 
 
 the middle of the span a with a load G, 
 G = Wheel load, 
 
 a = Distance center to center of ties, 
 B 
 
 r> QEI 
 
 D = 
 
 CM 
 
 ■\f; 
 
 Cb 
 
 E = Modulus of elasticity of the rail, 
 / = Moment of inertia of the rail, 
 b = The width of the tie, 
 1= One-half the length of the tie, 
 C = Coefficient of the ballast, 
 z = An auxiliary value depending on the form of the tie. 
 
 The coefficient of ballast represents the pressure in kilograms per square 
 centimeter of the ballast which causes a depression of one centimeter. The 
 coefficient 3 corresponds with simple gravel and the coefficient 8 with gravel 
 on a bed of dry stone or on rocky soil. 
 
 Fig. 158 shows that the ratio — Q increases with y , that is to say, with 
 
 the stiffness of the rail and the flexibility of the tie. 
 
 Table LI 1 1 presents calculations for several German railroads by the 
 aid of Zimmerman's formulas. The table is taken from an article on " The 
 Track Superstructure of German Railways " by M. Blum in the Revue Generale 
 
 * Calculation of the Superstructure, Berlin, 1888. 
 
218 
 
 STEEL RAILS 
 
 des Chemins de Fer, No. 5, November, 1908, a translation of which is given 
 in Proceedings American Railway Engineering and Maintenance of Way 
 Association, Vol. 11, Part 2, 1910. 
 
 V 
 
 O 
 
 
 
 1 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 3 
 
 
 
 
 4 
 
 0.7 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 DQ 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 | | 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 "-, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1.5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •y=o i 2 
 
 Fig. 158. — Bending Moment of Rail placed o 
 
 Ties. (Zimmerman.) 
 
 In questions of such moment we cannot rely on mathematical analyses for 
 conclusions. We can deduce general results from specific experiments by their 
 aid, but it is somewhat unsafe to attempt any generalization on its evidence 
 alone. 
 
 22. Tests to Determine the Bending Stress in the Rail 
 
 * Experiments were made in the track of the Boston and Albany Railroad 
 in 1889. The experiments consist in measuring the depression of the rail at 
 different places along the length when loaded, and in measuring the extension 
 or compression of the metal at the upper surface of the base of the rail, near 
 the edge of the outside flange. 
 
 For measuring the depression a row of stakes was driven alongside the 
 track, three feet away, and the relative level of points on the base of the rail 
 and nails in the tops of the stakes was ascertained by means of a sensitive 
 spirit level. 
 
 For ascertaining the strains, gauged lengths of 5 inches each were estab- 
 lished and defined by center punch marks on the base of the rail at places 
 over the ties and midway between them, and the amount of extension or 
 compression, as the case might be, was measured on these gauged lengths. 
 
 The rails were 72 pounds per yard, 4£ inches high and 4£ inches width of 
 base. The results of the experiments are shown in Fig. 159. 
 
 * House Executive Documents, 1st Session, 51st Congress, 1889-90, Vol. 25, Tests on Metals. 
 

 
 
 
 
 TABLE LIII. — ] 
 
 DIMENSIONS 
 
 , WEIGHTS, 
 
 AND COST 
 
 OF DIFFERENT SYSTEMS OF 
 
 TRACK SUPERSTRUCTURE ON WOODEN TIES 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 S | 4 | 5 
 
 • 
 
 10 11 1. | 13 
 
 » 1 " 1 " 1 " 
 
 18 
 
 19 
 
 20 21 
 
 22 | 23 | 24 | 25 
 
 2S 27 
 
 - \*\m\*\m\u\u\m\ 
 
 
 Rails. 
 
 Ties. 
 
 Anglo Bars. 
 
 Weight of Track. 
 
 1 
 | 
 1 
 
 
 Rail, 
 
 Ties. 
 
 Angle Bar*. 
 
 Comparison. 
 
 o 
 
 
 is 
 
 1 
 
 
 & 
 
 1 
 1 
 
 Spacing. 
 
 I 
 
 Per Pair. 
 
 I 
 
 1 
 
 i 
 
 J 
 
 1 
 
 j 
 
 j 
 
 1 
 
 | 
 
 1 
 1 
 
 Deflection. 
 
 l 
 
 1 
 
 5 
 
 1 
 
 Rails. 
 
 Ties. | Angle Bars. 
 
 1- 
 
 Type of Rail, Etc. 
 
 a 
 
 j 
 
 1 
 
 j'! 
 
 Center, 
 
 Eitrera- 
 
 i 
 
 § 
 
 | 
 
 3 
 
 j 
 
 B H 
 
 r 
 
 
 
 tifgSta 
 
 Referred to Point 
 
 
 
 per'' 
 
 Ia.« 
 
 Ft. 
 
 Feet and laches. 
 
 . Ia. 
 
 .. 
 
 La. 
 
 Lbs. 
 
 In.<: 
 
 Lbs. 
 Yard. 
 
 % 
 
 % 
 
 % 
 
 perlcS. 
 
 K' 
 
 Lbs. 
 
 In. 
 
 A 
 
 la. 
 
 la. 
 
 L, 
 
 See Note. 
 
 With Prussian Track, Type 6a, No. 1 of Table. 
 
 
 Prussian Hessian Rys. ) 
 Type 6a, 1885-94 J 
 
 Prussian Hessian Rys., 1 
 Type 6e, 1885-99 ( 
 
 Prussian Hessian Rys., ) 
 Type 8a, 1890-94 J 
 
 Prussian Hessian Rys., 1 
 Type 8b, 1890-99 J 
 
 Prussian Hessian Rys., I 
 Type 15a, 1905 ) 
 
 Alsace-Lorraine Rys., 1 
 Type of 1893 J 
 
 Alsace-Lorraine Rys., ) 
 Type 1893-1903 ( 
 
 Bavarian State Rys., 1 
 Type of 1892 J 
 
 Bavarian State Rys., i 
 Type of 1898 J 
 
 Saxony State Rys., 1 
 Type of 1890 J 
 
 Saxony State Rys., Rail 1 
 with reinforced joint | 
 
 67.1 
 67.1 
 
 82.4 
 
 90.5 
 
 76.0 
 
 70.1 
 
 87.4 
 91.9 
 
 91.9 
 
 24.88 
 32.45 
 32.45 
 37.99 
 26.21 
 26.21 
 
 25.72 
 
 34.97 
 40.80 
 
 40.80 
 
 39.37 
 39.37 
 39.37 
 39.37 
 49.21 
 29.53 
 39.37 
 
 29.53 
 
 39.37 
 32.81 
 
 49.21 
 
 15 
 18 
 15 
 17 
 24 
 12 
 19 
 
 12 
 
 18 
 13 
 
 19 
 
 8' 10|"X10i''X6A" 
 8' KH"Xl0i"X6ft" 
 8' 10i"X10i"X6&" 
 8' 10r'Xl0J"X6A" 
 8' 10i"X10i"X6- r y' 
 8' 10J"X10i"X6A" 
 8' 10rxl0|"X6&" 
 
 8' 2J"Xl0i"X6A" 
 
 8' 10J"X6A"X6A" 
 8' 2i"Xl0i"X6&" 
 
 j*8' 2i"X10}"X6f " | 
 1|8' 10r'X10r'X6ft"j 
 
 33.45 
 27.56 
 33.45 
 30.32 
 
 24.80 
 31.50 
 
 32.13 
 
 30.04 
 32.48 
 
 32.48 
 
 20.87 
 19.69 
 22.05 
 20.87 
 21.20 
 23.62 
 13.39 
 
 9.69 
 
 16.54 
 21.20 
 
 21.20 
 
 27.17 
 31.10 
 28.35 
 32.28 
 35.04 
 34.65 
 34.65 
 [27.56] 
 I 2 ' 1 "'' 1 '' 
 33.47 
 35.43 
 
 33.47 
 
 60.6 
 67.5 
 81.9 
 91.5 
 90.0 
 
 49.1 
 
 75.5 
 101.0 
 
 20.16 
 20.16 
 28.96 
 29.01 
 29.01 
 23.52 
 14.40 
 
 10.42 
 
 14.88 
 9.84 
 
 27.60 
 
 1.23 
 1.23 
 1.12 
 
 1.12 
 1.31 
 
 1.82 
 
 2.47 
 
 2.35 
 4.14 
 
 1.48 
 
 413.0 
 473.4 
 448.7 
 491.7 
 548.8 
 458.9 
 504.1 
 
 405.4 
 
 537.2 
 473.2 
 
 469.2 
 
 61.1 
 53.8 
 55.6 
 57.0 
 56.1 
 60.6 
 
 55.5 
 
 53.9 
 48.2 
 
 48.4 
 
 9.1 
 
 10.7 
 9.5 
 
 10.2 
 10.8 
 9.3 
 
 9.9 
 
 13.6 
 
 13.0 
 
 12.4 
 
 32.5 
 
 28.2 
 36.7 
 33.3 
 32.9 
 33.1 
 30.1 
 
 34.6 
 
 32.5 
 
 1.40-5 
 1.55] 
 1.56 | 
 
 1.66] 
 1.84] 
 1.53 | 
 
 i*| 
 
 1.40 J 
 
 1.83 | 
 1.67J 
 
 1.67 j 
 
 108 
 108 
 
 289 
 
 289 
 
 108 
 289 
 
 108 
 289 
 
 108 
 108 
 
 289 
 108 
 
 108 
 289 
 
 17,570 
 14,000 
 
 16,490 
 13,310 
 
 14,970 
 11,920 
 
 14,420 
 11,620 
 
 13,550 
 9,480 
 
 16,360 
 13,020 
 
 15,400 
 12,460 
 
 17,630 
 14,020 
 
 14,000 
 
 13,200 
 10,760 
 
 13,200 
 10,760 
 
 0.143 
 0.074 
 
 0.122 
 0.059 
 
 0.132 
 0.067 
 
 0.122 
 0.059 
 
 0.100 
 0.051 
 
 0.135 
 0.069 
 
 0.115 
 0.057 
 
 0.146 
 
 0.073 
 
 0.119 
 0.061 
 
 0.127 
 0.066 
 
 0.127 
 0.066 
 
 620.0 
 746.6 
 
 617.2 
 630.0 
 
 613.0 
 
 607.2 
 635.6 
 
 600.0 
 601.5 
 
 620.0 
 701.0 
 
 508.6 
 612.8 
 
 504.8 
 
 600.0 
 
 605.7 
 624.2 
 
 550^3 
 
 493.4 
 550.3 
 
 0.146 
 0.070 
 
 0.142 
 0.059 
 
 0.144 
 0.065 
 
 0.142 
 0.061 
 
 0.140 
 0.057 
 
 0.145 
 0.066 
 
 0.141 
 0.059 
 
 0.156 
 
 0.071 
 
 0.142 
 0.060 
 
 0.152 
 0.065 
 
 0.152 
 0.065 
 
 0.016 
 0.016 
 
 0.015 
 0.013 
 
 0.016 
 0.015 
 
 0.015 
 0.014 
 
 0.014 
 0.013 
 
 0.016 
 0.015 
 
 0.015 
 0.013 
 
 0.026 
 
 0.022 
 
 0.015 
 0.014 
 
 0.025 
 0.020 
 
 0.025 
 0.020 
 
 -0 
 -0 
 
 -0 
 
 -0 
 -0 
 
 -0 
 -0 
 
 -0 
 -0 
 
 -0 
 -0 
 
 -0 
 
 -0 
 
 +0 
 
 -0 
 
 -0 
 -0 
 
 +0 
 -0 
 
 -0 
 -0 
 
 015 
 018 
 
 015 
 015 
 
 015 
 016 
 
 015 
 
 015 
 
 014 
 015 
 
 015 
 017 
 
 015 
 
 014 
 
 007 
 0004 
 
 015 
 
 oi.-. 
 
 007 
 0004 
 
 007 
 0004 
 
 1479 
 1215 
 
 1245 
 1036 
 
 1147 
 976 
 
 1116 
 959 
 
 1136 
 962 
 
 1164 
 924 
 
 1409 
 1176 
 
 1968 
 2024 
 1539 
 1581 
 
 1486 
 1398 
 
 2189 
 
 917 
 
 '.).-,(; 
 760 
 793 
 
 4836 
 3973 
 
 4072 
 3386 
 
 4913 
 4181 
 
 4200 
 
 3007 
 
 3935 
 3334 
 
 4047 
 3213 
 
 2400 
 
 4261 
 5578 
 3373 
 4358 
 
 2037 
 2763 
 
 3350 
 2425 
 
 4240 
 3510 
 
 fi.ooj 
 
 (l.llj 
 }l,ll 
 
 M 
 
 }l,9j 
 
 M 
 
 li.oo 
 
 (1.31J 
 
 M 
 I 1-19 ! 
 
 1.00 
 1.00 
 
 0.94 
 0.95 
 
 0.85 
 0.85 
 
 0.82 
 0.83 
 
 0.93 
 0.93 
 
 0^89 
 
 1.00 
 
 1.00 
 
 0.80 
 0.81 
 
 0.75 
 0.76 
 
 0.75 
 0.76 
 
 1.00 
 1.00 
 
 0.85 
 
 0.92 
 0.91 
 
 0^81 
 
 0.70 
 0.69 
 
 0.94 
 0.93 
 
 0>7 
 
 1.02 
 0.99 
 0.83 
 
 0.89 
 0.89 
 
 0.89 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 00 
 00 
 
 98 
 84 
 
 99 
 95 
 
 85 
 
 81 
 00 
 
 81 
 
 98 
 84 
 
 80 
 74 
 
 80 
 
 74 
 
 1 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 00 
 00 
 
 98 
 85 
 
 99 
 93 
 
 98 
 87 
 
 96 
 
 96 
 
 84 
 
 07 
 01. 
 
 98 
 
 85 
 
 04 
 92 
 
 04 
 92 
 
 1.00 
 1.00 
 
 0.84 
 0.85 
 
 0.78 
 
 0.70 
 0.79 
 
 0.77 
 0.79 
 
 0.79 
 0.76 
 
 0.95 
 0.97 
 
 1.33 
 1.37 
 1.27 
 1.30 
 
 1.00 
 1.15 
 
 1.18 
 1.30 
 
 0.62 
 0.65 
 0.63 
 0.65 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 00 
 
 00 
 
 84 
 85 
 
 02 
 05 
 
 87 
 91 
 
 81 
 84 
 
 84 
 81 
 
 61 
 
 61 
 
 88 
 15 
 
 10 
 
 61 
 70 
 
 61 
 
 289 
 
 289 
 108 
 
 108 
 
 108 
 
 108 
 289 
 
 108 
 
 2.S0 
 
 108 
 
 108 
 
 "a" = pounds per 
 square iucli produc- 
 ing 1 inch depres- 
 sion in ballast. 
 
 Inside angle bar. 
 < lulside angle bar. 
 Inside angle bar. 
 ( tulside angle bar. 
 
 Inside angle bar. 
 
 I lul side angle bar. 
 lu-ide angle bar. 
 Outside angle bar. 
 
 Sellew, " Steel Rails." 
 
 
 
 
 
 
 
 
 
 
 
 
 
 • Iatermediat 
 
 a ties. 
 
 
 t Joiat ties. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Note."— Columns 26 and .27 are, shown ir 
 
 nits, the original table not indicating what n 
 
STRESSES IN THE RAIL 219 
 
 * Experiments were made by the Ordnance Department, U. S. Army, during 
 the month of October, 1893, on the track of the Chicago, Burlington and Quincy 
 Railroad, at Hawthorne, 111. 
 
 The experiments consisted of measuring the depression of the rails under 
 the weights of different classes of locomotives, and the fiber stresses developed 
 in the base of the rail. 
 
 For the purpose of observing the depression of the rails, bench marks were 
 established on a row of stakes driven alongside the rail, 31 inches distant from 
 
 DEPRESSION 
 OF RAIL 
 
 STRAINS IN 
 BASE OF RAIL 
 5" LENGTH 
 
 COMPRESSION 
 
 Fig. 159. — Railroad Track Experiments, Boston and Albany R. R. 
 
 it. A beam carrying a micrometer and an astronomical level bubble were used 
 in observing the depression of the rail (see Fig. 160), 'first measuring the height, 
 using points on the outer flange, when the rail was unloaded, and repeating the 
 observations when the engine was standing on the track. 
 
 It was found that the roadbed in the vicinity of the locomotives was sensibly 
 depressed and that the bench marks were within the influence of that depression. 
 It was possible to detect a depression of the roadbed as far as 91 inches from the 
 locomotive at the side of the track. 
 
 A correction for the depression of the bench marks was obtained by means 
 of a cantilever supported 10 feet from the track, and the total depression of 
 
 * House Executive Documents, 3rd Session, 53rd Congress, 1894-95, Vol. 30, Tests of Metals, etc. 
 
220 
 
 STEEL RAILS 
 
 Fig. 160. — Railroad Track Experiments. 
 Photograph of Leveling Instrument for Measuring the Depression of the Track. 
 
 Fig. 161. — Railroad Track Experiments. 
 Photograph of Micrometer for Determining the Fibre Stress in the Base of the Rail. 
 
STRESSES IN THE RAIL 221 
 
 points on the rails was also determined with reference to the cantilevers in some 
 of the experiments instead of using stakes. 
 
 The fiber stresses were determined in the base of the rail by measuring 
 the elongation or compression of the metal on a gauged length of 5 inches, 
 established on the top surface of the outer flange, observing the strains when 
 the wheels were directly over or when spanning the gauged length (see Fig. 161). 
 
 The observed strains were then computed for the stresses per square inch, 
 assuming a modulus of elasticity of 30,000,000 pounds per square inch and 
 that the fibers in the base were strained proportionally to their distance from 
 the neutral axis of the rail ; the computed stresses referring to the outside fibers 
 most remote from the neutral axis. 
 
 It will be observed that the strains and the computed stresses refer to 
 a gauged length of 5 inches, and, consequently, the maximum stresses may be 
 somewhat greater than those shown, considering the maximum bending moment 
 to be directly under the point of application of the load. Some of the results 
 are graphically shown in Fig. 162. 
 
 The moment of inertia of the 66-pound rail tested was 19.127 and the 
 
 19.127 
 section modulus of the base ? '„, = 8.54. 
 
 Fig. 162A shows the depression of one rail its entire length and the ends 
 of contiguous rails, the locomotive occupying one position thereon as shown 
 with reference to the rail and ties. 
 
 Fig. 162B shows the curve of depression under another type of locomotive. 
 This engine had no leading truck nor tender, but had a two-wheeled trailing truck. 
 
 In the position it occupied during the test, the greatest depression of the 
 rail occurred under the forward drivers, the rail presenting a sharp acclivity 
 before the engine, and beyond the joint the contiguous rail rose slightly above 
 the normal level. 
 
 In the diagram, Fig. 162C, are shown the fiber stresses as measured on 
 the base of the rail at station 14|, midway between ties Nos. 14 and 15. 
 
 Advance wave determinations were made on the 66-pound rail on cinder 
 ballast (8 inches under the tie) with the same class engine as shown in Fig. 162A, 
 the engine weighing 125,000 pounds. With the locomotive slowly approaching, 
 an upward movement of the rail began when the leading truck wheel was about 
 15 feet away ; the wave increased while the locomotive continued to advance, 
 reaching a maximum of .0037 inch when the truck wheel was about 8| feet away. 
 Then followed a sudden depression, and the height of the rail was reduced to 
 the normal level when the truck wheel was about 1\ feet away. 
 
222 
 
 STEEL RAILS 
 
 The position of the locomotive when the upward motion of the wave first 
 reached the station could be identified with considerable precision, but, owing 
 to an appreciable interval of time being necessary for the level bubble of the 
 
 TENDER 66000lbs. 
 
 30200ibs. 32600lbs r 287001bs. I8500lbs. 
 
 NE AS SHOWN I N ' PC . . 
 
 Q C)OQ 
 
 ooo 
 
 _Q 
 
 66lb.RAIL 
 
 Same Track Stresses Lbs. 
 
 ,..,, PER SQ.IN. 
 
 AS A IN BASE 
 
 COMPRESSION 
 
 ASUREMENTS TAKEN ON A GAUGED LENGTH .OF 5 INCHES. 
 ION I4>i (SEE A.) 
 
 M 
 
 AAA, A 
 
 Fig. 162. — Railroad Track Experiments. C. B. & Q. R. R. 
 
 measuring instrument to stop and reverse the direction of its movement, the 
 position of the crest of the wave, as well as the time when the height of the rail 
 was returned to its normal level, could not be so well defined. 
 
 The wave length was probably somewhat less than the observations showed. 
 The abruptness with which the direction of the wave motion was changed and 
 
STRESSES IN THE RAIL 
 
 223 
 
 ■* 
 
 
 
 
 
 
 "" 
 
 ^^^^i' 
 
 i 
 
 
 rO N 
 
 
 —§— 
 
 
 
 
 ?\ 
 
 ^ i 
 
 
 
 
 
 -T— 
 
 
 CU 
 
 ^ 
 
 ^K 
 
 H h- 
 
 6.4" _ 
 
 H 
 
 '' 
 
 n 
 
 _|_ 
 
 __N__ 
 
 
 ^K. 
 
 s 
 
 F— -- 
 
 -A 
 
 
 sr^ 
 
 Fig. 163. — Advance Wave Determinations. (Cuenot.) 
 
 the rail returned to its normal level, after which, of course, it was depressed 
 below the normal, was a very striking feature of the observations.* 
 
 * Fig. 163 shows the advance wave observed by M. Cuenot. It was found that when the first 
 wheel of the engine is about 20 feet from a tie the upward movement commences and r 
 mum at about 10 feet. 
 
224 
 
 STEEL RAILS 
 
 The observations of the depression of the roadbed made in these experi- 
 ments are of importance. On cinder ballast that part of the roadbed in which 
 the stakes were driven (31 inches from the track) was depressed a maximum 
 of .049 inch and on gravel ballast the maximum was .036 inch. Wooden stakes 
 
 and iron bolts were driven 
 different depths into the 
 roadbed with similar re- 
 sults; in fact, the few 
 observations which were 
 made showed the longer 
 stakes to have been quite 
 as much depressed as the 
 shorter ones, which did 
 not penetrate the cinder 
 ballast. 
 
 Following out the 
 depression of the roadbed 
 in a lateral direction, 
 on cinder ballast, when 
 the middle driver of the 
 engine was abreast the 
 place of observation, 
 there was a measurable 
 depression at a distance 
 of 91 inches from the rail. 
 The recovery in the 
 depression of the roadbed 
 was not complete imme- 
 diately upon the removal 
 of the engine from that 
 
 Fig. 164. — Movement of Rails Laid Alongside of Track. vicinity. The principal 
 
 The right-hand rail lying by the near telegraph pole moved 40 feet, n^rt of the recovery at 
 The trail it left may be traced from a point near the angle bar in 
 
 the foreground. (Railroad Age Gazette, Dec. 17, 1909.) Once took place; the 
 
 remaining portion of the 
 depression, however, was very sluggish in returning. The length of time 
 required to effect complete resilience was not determined. One observation, 
 however, made nine minutes after the load was removed from the vicinity, 
 showed the resilience then incomplete. 
 
STRESSES IN THE RAIL 225 
 
 Fig. 164 shows an exaggerated case of the wave motion and depression of 
 the track. The road runs on an embankment about five feet above the level 
 of a wet meadow. The wave motion of the track and embankment is so great 
 that rails lying by the side of the track move along apparently of their own 
 accord at the rate of nearly a foot a day. This movement was undoubtedly 
 due to the undulatory movement of the track and entire fill and probably some 
 reaction of the fill itself against the track. 
 
 Further tests were made in 1894 and 1895 on the tracks of the Pennsylvania 
 Railroad and the Boston and Albany by the Government. * These experiments 
 comprise observations on the fiber stresses developed in rails in the track, the 
 depression of the rails, and the slope or inclination of the rails caused by the 
 weight of the different wheels of the locomotive. The results show some phe- 
 nomena displayed by rails in service under the static conditions of loading or 
 when a locomotive passes slowly over the track. 
 
 The series were made chiefly on the track of the Pennsylvania Railroad, 
 where exceptional opportunities existed for examining roadbed, embracing a 
 wide variety of conditions of weight of rails and different kinds of ballast and 
 its behavior under heavy types of freight and passenger locomotives. 
 
 The tests were made during the early part of the month of November, 1894, 
 on track in the condition it was found in service. 
 
 The experiments on the Boston and Albany Railroad were made with track 
 on frozen gravel ballast, in the month of February, 1895. 
 
 The fiber stress tests were made by means of a micrometer mounted on 
 the upper side of the flange of the base of the rail, at a place midway adjacent 
 ties. The instrument covered a gauged length of 5 inches. The micrometer 
 was adjusted in position, and then the several wheels of the locomotive were 
 successively brought over the gauged length, or until the same was midway 
 adjacent wheels. 
 
 The instrument was read when the locomotive was at each of these posi- 
 tions. It was found practicable to make the micrometer observations without 
 arresting the locomotive in all cases, taking the readings as the locomotive 
 passed slowly over the rail. In this manner the strains developed were measured, 
 an elongation of the metal showing tensile stress, and a contraction in the gauged 
 length showing compressive stress. 
 
 The measured strains were reduced to stresses per square inch, assuming 
 the modulus of elasticity of the steel to be 30,000,000 pounds per square inch, 
 and correcting the observed strains in order to obtain the maximum fiber stresses, 
 
 * House Documents, Vol. 46, 54th Congress, 1st Session, 1895-96. No. 54, Tests of Metals. 
 
226 STEEL RAILS 
 
 on the further assumption that the strains were proportional to their distance 
 from the neutral axis of the rail. 
 
 Fig. 165 shows the micrometer in position on the base of the rail, under 
 the driving wheel of a locomotive. 
 
 The depression of the rails was measured by means of a sensitive level 
 bubble, mounted on a rod, carrying at one end a screw micrometer, which 
 
 Fig. 165. — Railroad Track Experiments. 
 
 View showing Micrometer for Measuring Strains in Rails, in Position on Base of Rail under 
 
 Driving Wheel. 
 
 rested on a stake driven in the roadbed 30 inches from the rail; the other end 
 of the rod rested upon the base of the rail. The depression of the track was 
 thus measured with reference to the top of the stake used as a bench mark. 
 In this series it was necessary to arrest the movement of the locomotive at 
 each observation. 
 
 The slope tests, or inclination of the rails, were made by means of a sensitive 
 level bubble mounted on a frame 12 inches long. At one end of the frame there 
 
STRESSES IN THE RAIL 
 
 227 
 
 was a fixed supporting rod having a conical point; at the other end there was a 
 screw micrometer, the contact end of which was also made with a conical point. 
 
 In the use of this instrument, two center punch marks, 12 inches apart, 
 were made on the base of the rail. The conical points of the instrument entered 
 these center punch marks and furnished definite contact points with the rail. 
 The instrument was then leveled and the changes in slope, when the rail was 
 affected by the locomotive, were measured from this initial adjustment of the 
 level bubble. 
 
 Fig. 165 shows the slope instrument resting on the second tie to the right 
 of the fiber-stress micrometer. The rails examined ranged in weight from 60 to 
 100 pounds per yard, and were supported on oak ties resting on cinder, gravel, 
 and stone ballast, in the case of the Pennsylvania Railroad. 
 
 On the Boston and Albany Railroad, yellow pine ties, with shoulder tie 
 plates, were used, the roadbed being ballasted with gravel, which was in a 
 frozen condition at the time of the tests. 
 
 TABLE LIV. — RAILROAD TRACK EXPERIMENTS. GENERAL DIMENSIONS 
 
 OF RAILS 
 
 GOVERNMENT RAIL TESTS 
 
 (House Documents, Vol. 46, 54th Congress, 1st Session, 1895-96. No. 54, Tests of Metals) 
 
 Weight 
 
 Height, 
 
 Width of 
 
 Width of 
 
 Tlnckne-.- 
 
 Moment of 
 
 Moment of 
 
 Distance, Neutral Axis to 
 
 per Yard. 
 
 Base. 
 
 Head. 
 
 of Web. 
 
 Inertia. 
 
 Resistance. 
 
 Outside Fiber. 
 
 Pounds. 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 Inch. 
 
 / 
 
 R=L. 
 
 Headrc 
 Inches. 
 
 sk: 
 
 60 
 
 4i 
 
 41 
 
 2f 
 
 i 
 
 14.222 
 
 6.693 
 
 2.125 
 
 2.125 
 
 70 
 
 4i 
 
 4| 
 
 2A 
 
 
 18.055 
 
 8.282 
 
 2.32 
 
 2.18 
 
 85 
 
 5 
 
 5 
 
 2& 
 
 1| 
 
 26.374 
 
 10.853 
 
 2.57 
 
 2.43 
 
 100 
 
 51 
 
 5h 
 
 2f! 
 
 f 
 
 38.957 
 
 14.812 
 
 2.87 
 
 2.63 
 
 95 
 
 5* 
 
 51 
 
 3 
 
 I 
 
 32.28 
 
 13.563 
 
 2.65 
 
 2.38 
 
 The general dimensions of the rails are given in Table LIV. What was 
 then considered a heavy type of freight and passenger locomotive was em- 
 ployed, the weights of which are recorded in Table LV. 
 
 Referring to the tests on the Pennsylvania Railroad, the tensile fiber stresses 
 developed under the weight of the driving wheels ranged from 2810 to 19,540 
 pounds per square inch, and the compression stresses, when the gauged length 
 was between wheels, reached 7880 pounds per square inch. These values 
 belonged to the rails in their ordinary condition of service. A tie was removed 
 from the track, laid with 100-pound rail, which increased the distance between 
 centers of ties to 52 inches, and here the maximum tensile stress developed 
 was 18,970 pounds per square inch, against 9840 pounds per square inch for 
 another rail of the same section resting on ties 26 inches apart. 
 
228 
 
 STEEL RAILS 
 
 A splice bar on a 70-pound rail was strained 22,140 pounds per square 
 inch, tension, and 8300 pounds per square inch, compression stress, by the 
 driver of passenger engine No. 809. 
 
 TABLE LV. — WEIGHTS OF LOCOMOTIVES 
 
 GOVEE 
 (House Documents, Vol. 46, 54th C 
 
 NMENT RAIL TEST 
 
 3 
 
 6. No. 54, Tests of Metals) 
 
 
 Locomotive. 
 
 Total. 
 
 En g ine. 
 
 Tender. 
 
 Weight per Wheel. 
 
 Pilot. 
 
 Drivers. 
 
 Wheel. Pounds. 
 
 Tons. 
 
 Passenger No. 809, 
 Class PK. 
 
 Passenger No. 1515, 
 Class T. 
 
 Freight No. 557, 
 Class R. 
 
 Passenger No. 209, 
 B. &A. R.R. 
 
 Pounds. 
 
 197 050 
 222,500 
 188,600 
 
 199,700 
 
 39,750 
 50,300 
 11,000 
 
 40,700 
 
 Pounds. 
 
 87,300 
 
 95,200 
 
 800 
 
 75,000 
 
 Pounds. 
 
 70,000 
 77,000 
 63,800 
 
 84,000 
 
 Pilot 
 
 9,937 
 21,750 
 21,900 
 
 8,750 
 
 12,575 
 24,250 
 23,350 
 12,833 
 
 5,500 
 13,250 
 13,750 
 15,650 
 14,250 
 
 7,975 
 
 10,175 
 18,750 
 18,750 
 
 4.968 
 
 Driver, first. . . 
 Driver, second. 
 Tender 
 
 Pilot 
 
 Driver, first. . . 
 Driver, second. 
 Tender 
 
 Pilot 
 
 10.875 
 10.950 
 4.375 
 
 6.287 
 12.125 
 11.675 
 
 6.416 
 
 2.750 
 
 Driver, first. . . 
 Driver, second. 
 Driver, third. . 
 Driver, fourth.. 
 Tender 
 
 Pilot. . . 
 
 6.625 
 6.875 
 7.825 
 7.125 
 3.987 
 
 5 087 
 
 Driver, first. . . 
 Driver, second. 
 Tender 
 
 9.375 
 9.375 
 
 
 First truck. . . . 
 Second truck. . . 
 
 9,250 
 11,750 
 
 4.625 
 5.875 
 
 Table LVI shows the maximum tensile fiber stress caused by the wheels 
 of the pilot, engine, and tender on the different rails and kinds of ballast, also 
 the maximum compression stresses developed in each experiment. The place 
 of observation in these experiments was between ties and about one-quarter of 
 the length of the rail from the end. 
 
 From the irregular manner in which the stresses were developed in the 
 different weights of rail, it is evident that the peculiar condition of the track 
 at individual rails has an important influence on the magnitude of the fiber 
 stresses. 
 
 The lightest section of rail examined, 60 pounds per yard, resting on ties 
 on gravel ballast, gave exceptionally low fiber stresses, and it will be seen that 
 this rail was depressed a correspondingly small amount. 
 
 So much variation is found in the stresses as to practically obscure the 
 relative strength of the different weights of rails, and it seems necessary to 
 compare the extreme sections to show a well-defined difference in the maximum 
 stresses. 
 
STRESSES IN THE RAIL 
 
 229 
 
 On account of the peculiar conditions influencing the behavior of the 
 individual rails, the relative values of the different kinds of ballast are less 
 conspicuously shown in the fiber-stress experiments than in the series on the 
 depression of the rails. 
 
 TABLE L VI. — MAXIMUM FIBER STRESSES IN BASE OF RAIL 
 
 GOVERNMENT RAIL TESTS 
 
 (House Documents, Vol. 46, 54th Congress, 1st Session, 1895-96. No. 54, Tests of Metals) 
 
 Tensile Fiber Stress per Square 
 
 Inch. (Pounds.) 
 
 Pilot. 
 
 Drivers. 
 
 Tender. 
 
 6,180 
 
 11,670 
 
 2,750 
 
 3,430 
 
 7,550 
 
 3,430 
 
 11,860 
 
 19,540 
 
 9,770 
 
 11,160 
 
 16,050 
 
 9,770 
 
 10,730 
 
 17,170 
 
 10,020 
 
 8,970 
 
 18,620 
 
 8,280 
 
 7,590 
 
 13,790 
 
 6,210 
 
 10,070 
 
 14,390 
 
 7,910 
 
 6,470 
 
 11,510 
 
 6,470 
 
 9,450 
 
 18,180 
 
 10,910 
 
 13,840 
 
 22,140 
 
 9,230 
 
 7,160 
 
 10,030 
 
 5,020 
 
 5,730 
 
 12,180 
 
 7,880 
 
 3,580 
 
 10,030 
 
 5,020 
 
 10,750 
 
 12,180 
 
 6,450 
 
 9,310 
 
 17,120 
 
 9,310 
 
 7,160 
 
 10,030 
 
 2,870 
 
 7,160 
 
 10,750 
 
 4,300 
 
 4,300 
 
 10,030 
 
 5,020 
 
 6,320 
 
 9,840 
 
 5,620 
 
 10,540 
 
 18,970 
 
 8,430 
 
 3,510 
 
 8,430 
 
 4,220 
 
 6,870 
 
 9.920 
 
 6,870 
 
 7,630 
 
 11,450 
 
 6,870 
 
 Passenger No. 809 
 Freight No. 557... 
 Passenger No. 809 
 Freight No. 557... 
 Passenger No. 809 
 Passenger No. 809 
 Freight No. 557. . . 
 Passenger No. 809 
 Freight No. 557. . . 
 Passenger No. 809. 
 Passenger No. 809. 
 Passenger No. 809.. 
 Passenger No. 1515. 
 Freight No. 557.... 
 Passenger No. 809.. 
 Passenger No. 1515. 
 Freight No. 557. . 
 Passenger No. 809 
 Freight No. 557... 
 Passenger No. 809 
 Passenger No. 809 
 Freight No. 557.. . 
 Passenger No. 209 
 Passenger No. 209 
 
 1,370 
 
 1,400 
 4,290 
 5,520 
 
 8,300 
 3,580 
 4,300 
 4,300 
 4,300 
 5.020 
 7,880 
 4,300 
 3,580 
 4,220 
 2,110 
 2,810 
 3,050 
 
 * Taken at different points on the rail. 
 
 The relative effect of the several wheels of the locomotives are shown 
 with greater precision than some other features of the test, inasmuch as in this 
 comparison the action of all wheels are referred to the same point on the rail. 
 Table LVII shows the tensile stresses developed per ton weight on the dif- 
 ferent wheels of each locomotive on the several rails. From these results it 
 appears that the stresses are generally greatest under the outside wheels. 
 
 An examination of the results shows, as an extreme case, that the pilot 
 wheels of freight engine No. 557 on a 60-pound rail, with stone ballast, gave 
 a fiber stress of 4058 pounds per square inch per ton on the wheel, whereas 
 the first driver of the engine, per ton, strained the rail only 1685 pounds per 
 square inch. In this instance the total stress per square inch was the same 
 under these two wheels, namely, 11,160 pounds, although the weight on the 
 drivers was more than twice that on the pilot wheel. 
 
STEEL RAILS 
 
 TABLE LVII. 
 
 -TENSILE FIBER STRESSES IN BASES OF RAILS PER TON 
 WEIGHT ON THE DIFFERENT WHEELS 
 
 GOVERNMENT RAIL TESTS 
 
 (House Documents, Vol. 46, 54th Congress, 1st Session, 1895-96. No. 54, Tests of Metals) 
 
 PASSENGER LOCOMOTIVES 
 
 Tensile Fiber Stress (in Pounds) per Ton Weight on Wheels of 
 
 95 D 
 95 L 
 95 B 
 95 D 
 95 H 
 95 K 
 95 L 
 95 M 
 95 N 
 
 No. 809 
 No. 809 
 No. 1515 
 No. 809 
 No. 809 
 No. 809 
 No. 1515 
 No. 809 
 No. 809 
 No. 809 
 No. 809 
 No. 809 
 No. 809 
 No. 809 
 No. 209 
 No. 209 
 No. 209 
 No. 209 
 No. 209 
 No. 209 
 No. 209 
 No. 209 
 No. 209 
 
 Cinder 
 
 Cinder 
 
 Cinder 
 
 Gravel 
 
 Gravel 
 
 Gravel 
 
 Gravel 
 
 Stone 
 
 Stone 
 
 Stone 
 
 Stone 
 
 Stone, tie removed .... 
 
 Bridge 
 
 Splice bar 
 
 Frozen gravel, rail No. 1 
 Frozen gravel, rail No. 1 
 Frozen gravel, rail No. 2 
 Frozen gravel, rail No. 2 
 Frozen gravel, rail No. 2 
 Frozen gravel, rail No. 2 
 Frozen gravel, rail No. 2 
 Frozen gravel, rail No. 2 
 Frozen gravel, rail No. 2 
 
 1285 
 1927 
 2494 
 2110 
 
 1038 
 1038 
 519 
 
 FREIGHT LOCOMOTIVES 
 
 
 Locomotive. 
 
 Ballast. 
 
 Tensile Fiber Stress (in Pounds) per Ton Weight on Wheels of 
 
 Rail 
 Weight 
 per Yard. 
 
 Pilot. 
 
 Drivers. 
 
 Tender. 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 1 
 
 2 
 
 3 
 
 4 
 
 85 
 
 60 
 
 No. 557 
 No. 557 
 No. 557 
 No. 557 
 No. 557 
 No. 557 
 No. 557 
 No. 557 
 
 Cinder 
 
 1302 
 1247 
 2760 
 2604 
 4058 
 2353 
 1564 
 1276 
 
 540 
 518 
 
 1250 
 974 
 
 1685 
 761 
 865 
 424 
 
 1041 
 999 
 1805 
 1146 
 2335 
 1046 
 1041 
 511 
 
 366 
 878 
 1323 
 824 
 1872 
 1011 
 732 
 718 
 
 1408 
 1060 
 2512 
 1408 
 2056 
 1615 
 1408 
 1183 
 
 860 
 1558 
 
 720 
 2450 
 1264 
 1078 
 
 705 
 
 898 
 617 
 1384 
 720 
 1926 
 1084 
 898 
 880 
 
 720 
 860 
 
 1038 
 359 
 
 2275 
 903 
 898 
 351 
 
 1259 
 860 
 
 70 
 
 
 1384 
 
 85 
 60 
 
 Gravel 
 
 720 
 2450 
 
 70 
 85 
 100 
 
 Stone 
 
 Stone 
 
 Stone .. 
 
 1623 
 1259 
 
 1058 
 
STRESSES IN THE RAIL 
 
 231 
 
 Throughout this and earlier series of track experiments the same tendency- 
 has been found, the outside wheels exerting the most severe action on the rails 
 in proportion to the weight which they carry. 
 
 The maximum and minimum tensile stresses per ton on the different wheels 
 are shown in Table LVIII. 
 
 TABLE LVIII. 
 
 ■ MAXIMUM AND MINIMUM TENSILE STRESSES PER TON ON THE 
 DIFFERENT WHEELS 
 
 GOVERNMENT RAIL TESTS 
 ■use Documents, Vol. 46, 54th Congress, 1st Ses 
 
 -96. No. 54, Tests of Metals) 
 
 Maximum Stress. 
 
 
 Gravel . 
 Gravel . 
 Stone . . . 
 Stone . . . 
 Cinder.. 
 Gravel . 
 Gravel . 
 Stone . . . 
 
 Stone 
 
 Bridge 
 
 Splice bar. 
 Cinder..... 
 
 Cinder.. 
 Cinder. . 
 Gravel . 
 Gravel. . 
 Gravel . 
 Stone... 
 
 No. 8 
 Freight No. 557 
 Passenger No. 8 
 Freight No. 557 
 Passenger No. 809. 
 Passenger No. 8 
 Freight No. 557 
 
 No. 8 
 
 Stone ! 
 
 Stone | 
 
 Stone, tie removed 
 
 Stone 
 
 Gravel, frozen* . 
 Gravel, frozen* . 
 Gravel, frozenf . 
 
 Gravel, frozenf . 
 Gravel, frozenf . 
 Gravel, frozenf . 
 Gravel, frozenf . 
 
 Gravel, frozenf . 
 Gravel, frozenf . 
 
 Freight No. 557 
 Passenger No. 809. 
 Passenger No. 809, 
 Passenger No. 
 
 Passenger No. 1515. 
 Freight No. 557. 
 Passenger No. 8 
 Passenger No. 1515 
 Freight No. 557.. . 
 Passenger No. 809. 
 
 Freight No. 557. . . 
 Passenger No. 809. 
 Passenger No. 809. 
 Freight No. 557. . . 
 Passenger No. 209. 
 Passenger No. 209. 
 No. 209. 
 
 No. 209. 
 Passenger No. 209. 
 Passenger No. 209. 
 Passenger No. 209. 
 
 Passenger No. 209. 
 Passenger No. 209. 
 
 1st pilot. . 
 
 Pilot 
 
 Pilot 
 
 Pilot 
 
 4th tender . 
 4th tender. 
 
 Pilot 
 
 1st pilot. . 
 
 Pilot 
 
 4th tender . 
 1st pilot. . 
 1st pilot. . 
 
 1st tender. 
 4th driver. 
 1st pilot . . 
 1st pilot. . 
 
 Pilot 
 
 1st pilot. . 
 
 Pilot... 
 1st pilot 
 1st pilot 
 Pilot... 
 1st pilot 
 1st pilot 
 1st pilot , 
 
 2nd tender 
 1st pilot. . 
 1st pilot. . . 
 1st pilot . . . 
 
 1st pilot. . . 
 1st pilot. . . 
 
 Pounds. 
 
 1244 
 1247 
 2387 
 4058 
 2290 
 
 2353 
 2494 
 2789 
 1441 
 
 1228 
 1408 
 2164 
 1481 
 2604 
 1441 
 
 1564 
 1272 
 2122 
 1276 
 1199 
 1351 
 1500 
 
 1485 
 1500 
 1500 
 
 1st tender... 
 1st driver. .. 
 3rd tender. . 
 1st driver... 
 2nd driver. . 
 1st tender. . 
 3rd tender . . 
 1st, 2nd, 3rd 
 
 tender 
 1st driver. 
 2nd driver . 
 2nd tender. . 
 2nd, 3rd 
 
 tender 
 2nd pilot . . . 
 1st driver... 
 2nd pilot. . . . 
 1st tender. . 
 3rd tender . . 
 2nd, 3rd 
 
 tender 
 3rd driver 
 2nd pilot.. 
 3rd driver 
 3rd driver 
 1st driver. 
 1st tender. 
 3rd, 4th 
 
 tender 
 1st driver. 
 1st driver. 
 1st driver. 
 2nd, 3rd 
 
 tender 
 2nd tender 
 2nd tender 
 
 471 
 518 
 1595 
 
 1285 
 451 
 733 
 
 
 -164 
 
 Illustrative of the influence which the condition of the roadbed has on 
 the fiber stresses, the 60-pound rail on gravel ballast showed 1247 pounds 
 per square inch stress per ton under the pilot wheel of the engine, whereas, 
 
232 STEEL RAILS 
 
 with the same weight of rail on stone ballast, the same wheel gave 4058 pounds 
 per square inch. 
 
 The fiber stress experiments on the Boston and Albany Railroad were 
 made on rails 95 pounds per yard, on frozen gravel ballast, and observations 
 were taken at several points along the length of the rails. The observations 
 on rail No. 1 were made with the rail in the condition in which it was found 
 in the track. There was some looseness between the tie plates and the rail 
 and ties, which, in rail No. 2, was diminished as far as possible by redriving 
 the spikes and by the use of a number of additional ones. This is the only 
 instance in which spikes were redriven before testing. Rail No. 1 was examined 
 at two, and No. 2 was examined at seven, places along its length. 
 
 The tensile fiber stresses at the first end of rail No. 2 were higher than 
 those developed at the middle and near the second end of the rail. In this 
 rail, as the tensile stresses diminished at the second end, the compressive stresses 
 increased. At a space 33 inches from the end of the rail, the compressive stress 
 in the base reached 7630 pounds per square inch when this space was midway 
 the drivers. The same stress was also shown when the space was between the 
 tender trucks. 
 
 Concerning the relation between the fiber stress developed and the total 
 depression of the rail, the evidence generally favors the deduction that di- 
 minished depression will be accompanied by diminished fiber stress. 
 
 The depression of the rails examined on the Pennsylvania Railroad shows, 
 with the 60-pound rails, the least depression on the gravel ballast, the order 
 of rigidity being gravel, stone, and cinder ballast. With the 70-pound sections, 
 the order of rigidity is gravel, cinder, and stone ballast. Under the 85-pound 
 rails, the stone ballast gave greater rigidity than the gravel. No test for de- 
 pression was made with cinder ballast under the 85-pound rails, and only stone 
 ballast was used under the 100-pound rails. 
 
 Table LIX states the mean depression of the driving wheels, and also 
 the mean depression of all the other wheels of the locomotive in each experi- 
 ment. There is in the table a column of differences which states the excess 
 of depression of the drivers over that of the other wheels. The column of 
 differences is useful in showing the additional depression of the rails under 
 the weights of the driving wheels after they have been loaded by the other 
 wheels. 
 
 Under the 60- and 70-pound sections, the gravel ballast gave the greatest 
 rigidity under the drivers, as well as under the other wheels, and in the column 
 of differences the excess of depression was least for this kind of ballast. 
 
STRESSES IN THE RAIL 
 
 233 
 
 The total depression with 85-pound rails was less for the stone than for 
 the gravel ballast, although the excess of depression under the drivers was 
 practically the same in the two cases. 
 
 The depression of the rails on frozen gravel ballast, in which there was 
 no visible movement of the ties, would seem to represent about the attainable 
 limit of rigidity in track on wooden ties. 
 
 The fact that 60-pound rail on gravel compares favorably with the heavier 
 section on the frozen ballast indicates that this light section of rail was in a 
 condition approaching rigidity. 
 
 In the slope tests, the approach of the locomotive was felt for a distance 
 of 12 to 15 feet in front of the first wheel. The first observed movement was 
 
 TABLE LIX. — DEPRESSION OF RAILS — MEAN DEPRESSION UNDER DRIVING 
 WHEELS AND MEAN DEPRESSION UNDER PILOT AND TENDER WHEELS 
 
 GOVERNMENT RAIL TESTS 
 (House Documents, Vol. 46, 54th Congress, 1st Sess'on, 1805-96. No. 54, Tests of Metals) 
 
 Rail 
 Weight 
 per Yard. 
 
 Ballast. 
 
 Locomotive. 
 
 Drivers. 
 
 Pilot and 
 Tender. 
 
 Difference. 
 
 Pounds. 
 60 
 
 Cinder 
 
 
 .229 
 .073 
 .162 
 .230 
 .138 
 
 Inch. 
 
 .154 
 .042 
 .122 
 .157 
 .089 
 
 .075 
 
 60 
 
 
 .031 
 
 60 
 
 Stone 
 
 
 .040 
 
 70 
 
 
 .073 
 
 70 
 
 
 
 .049 
 
 70 
 
 Stone 
 
 
 .277 | .207 
 .233 .184 
 
 .070 
 
 85 
 
 
 .049 
 
 85 
 
 Stone 
 
 
 .144 
 .168 
 .139 
 
 .097 
 .116 
 .103 
 
 .047 
 
 100 
 
 
 .052 
 
 95 
 
 Gravel, frozen, rail No. 1 
 
 Passenger No. 209 
 
 .036 
 
 an upward one, the inclination of the rail sloping in a direction from the loco- 
 motive. This was followed by a reversal in the direction of the inclination, 
 which then sloped toward the locomotive. As the several wheels successively 
 passed over the place of observation, the inclination of the slope reached a 
 maximum and was reversed in direction, these motions being repeated under 
 each wheel with some modifications, according to the condition of the track. 
 After the locomotive had passed over the place of observation the inclination 
 gradually diminished, and eventually the rail practically resumed its original 
 level. A very critical examination led to the conclusion that each passage of 
 a locomotive left the rail in a slightly different state than it before occupied, 
 and that some sluggishness of recovery in the ballast had an influence on these 
 minute displacements. 
 
234 
 
 STEEL RAILS 
 
 Figs. 166 and 167 show graphically the results of the tests for depression 
 and stress in different kinds of ballast and weights of rails. 
 
 r 1 ' r! ~ v ' fen ■ 'Q'- 4 34" w 5-7 " ■w 4 -"- | A.< 5 '- 7 '" >. 
 
 @ i <h K°\Jk\J & (h (h <& 
 
 39750lbs 43500lbs ,438001b* TENDER 7O0O0lbs 
 
 LOCOMOTIVE N°809 CLASS PK. 
 
 TRACKMANS SURFACE O 
 
 ■trackman's SURFACE O 
 
 ^ ^ ^-TRACKMANS SURFACE 
 
 Gf?*l/eC^B4L-L /IS T 
 
 fT FiACKMANS SURFACE 
 
 trackman's surface 
 
 Fig. 166. — Railroad Track Experiments, Pennsylvania R. R. Depression in 
 
STRESSES IN THE RAIL 
 
 6 To 
 
 
 TENDER ?OOOQ 
 
 STRESSES IN BASE OF RAIL - GRAVEL BALLAST. 
 
 STRESSES IN BASE OF RAIL - STONE BALLAST. 
 
 Fig. 167. — Railroad Track Experiments, Pennsylvania R. R. Stress in Rail. 
 
236 
 
 STEEL RAILS 
 
 All of the experiments just described have been made with the static load 
 of the engine. In 1897, Dr. P. H. Dudley commenced a series of interesting 
 tests to determine the effect of the dynamic load of the engine. 
 
 * In Figs. 168 and 169 are shown the results of tests made by Dr. Dudley 
 with the stremmatograph. The principle of the stremmatograph is to record 
 on a moving strip the molecular compression or elongation of the metal in a 
 given length of the base of the rail, induced by the strains produced by each 
 
 
 5T£ 
 
 CAR 
 . -fiOC 
 
 4T2CAR 3S2CAR 2t!2CAR IST.CAR ' TENDER ENGINE, 
 noo ooo ooo ooo ooo onn ooo ooooo nn (. JUnnlk 
 
 5000lbs. 
 
 
 RECORD NO.l IOO lb. RAIL 6"HIGH « 
 
 19 Ml. PER HOUR ,| | 
 
 comp. o -jij-^ UjuUj^ljlJL^ 
 
 ~ WWWWWV MM 
 
 "MM Ml ,,M T T n 
 
 SOOOIbs 
 
 COM P. O A 1 
 
 TENSION "Ti 
 
 
 
 
 
 
 lOOOOlba 1 
 
 
 
 
 
 
 
 
 
 
 
 
 ' J 
 
 
 
 
 
 1 
 
 RECORD NO.2 80rb. RAIL 5"HIGH 
 
 
 40MI, PER HOUR 
 
 
 \ 1 
 
 SOOOOIbs 
 
 Fig. 168. — ■ Stremmatograph Tests at 19 and 40 m.p.h. 
 
 wheel of the moving train. These records can be measured by filar-micrometers 
 under a microscope, and then from the modulus of elasticity of the steel we may 
 compute the stresses which produce the given compression or elongation per 
 square inch of the extreme fiber in the base of the rail. 
 
 The object of the stremmatograph is to convert rails of any section and 
 weight, of any system of permanent way construction, into testing machines 
 in the track and show how much they are strained, due to the wheel loads and 
 
 * Dr. P. H. Dudley, The Railroad Gazette, May 20 and October 21, 1898, Stresses in Rails 
 under Moving Loads. Vol. Ill, Proceedings Am. Society for Testing Materials, 1903, Stremmato- 
 graph Tests, by P. H. Dudley. Vol. IV, ibid., 1904, Bending Moments in Rails, by P. H. Dudley. 
 
STRESSES IN THE RAIL 
 
 237 
 
 spacing of any type of locomotives and cars moving over the rails at the dif- 
 ferent speeds of service. In principle it is the same as the device for measuring 
 the strain in bridge members, described in Article 9. 
 
 
 O O 
 
 O O 
 
 OO nfes, 
 
 sooo 
 
 Record N?l. 
 IOO LB RAIL IOMlPerHR 
 
 comp, n aA aA A Aa/^ 
 
 — A/W\AW 
 
 5600 v y » 
 
 IQOOO 
 
 .nOOOn^ 
 
 c 
 
 XX 
 
 )L 
 
 Record N°5 
 ,nnn IOO-RAIL 
 
 COMP. A A N 
 
 TENSION 
 
 SOOO .. 
 
 V 
 
 IOOOO 
 
 V 
 
 5000 Record N?4 
 
 65 LB RAIL 
 COMP. Q | | K 
 
 TENSION 
 
 
 
 "9Q0Q 
 
 
 
 
 
 
 20000 
 
 1 
 
 
 
 ' 
 
 
 30000 
 
 
 
 
 *0000 
 
 
 
 
 SOOOO 
 
 
 Fig. 169. — Stremmatograph Tests at Slow Speed's. 
 
 Record No. 1, Fig. 168, is taken on the New York Central and Hudson 
 River Railroad tracks. The instrument was applied on the outside rail of a 
 3-degree curve at the Grand Central Terminal, over which nearly all of the 
 heavy trains from the terminal pass outward; the tonnage was from 20,000 to 
 
STEEL RAILS 
 
 25,000 per day, and there was more looseness in the track than generally found out 
 on the main line. The following is the data of the test. 
 
 Date, 
 
 
 June 28, 1898. 
 
 Weight of rail, 
 
 
 100 pounds per yard. 
 
 Height of rail, 
 
 
 6 inches. 
 
 Ballast, 
 
 
 Stone. 
 
 Ties, 
 
 
 Oak with tie plates spaced 24 inches, 
 center to center. 
 
 Speed, miles per 
 
 hour, 
 
 19. 
 
 Temperature, 
 
 
 90° F. 
 
 Locomotive and tender, 
 
 202,000 pounds. 
 
 First car, 
 
 
 95,000 pounds. 
 
 Second car, 
 
 
 86,200 pounds. 
 
 Third ear, 
 
 
 82,000 pounds. 
 
 Fourth car, 
 
 
 94,950 pounds. 
 
 Record No. 1, Fig. 169, shows the extreme fiber stresses in the same track 
 and under the same engine in a test of December 29, 1897, temperature 23° F., 
 when the engine was moving at a speed of 10 miles per hour. 
 
 When the December test was made, the rail was not only firm on the ties 
 but was under some tension, due to the low temperature. In this test the 
 small stresses under the front truck wheel show at once that the rail was very 
 firmly supported and was not loose on the ties. The rail being under tension 
 before loading, the actual stresses of tension in the rail were higher than indi- 
 cated, and those of compression lower. 
 
 When the June tests were made, the track had not been tamped or sur- 
 faced by the trackmen since the preceding October. Over 16,000,000 tons had 
 been carried over the rails since that date, or 12,000,000 tons since the previous 
 test. 
 
 Since the December test, air brakes had been applied to the engine trucks, 
 which adds more to her weight. While the engine truck carried more weight 
 in the June than in the December test, the increased stresses under the front 
 truck wheels and the much larger stresses of compression show that the rail 
 and ties were much looser in the June test. 
 
 Record No. 2, Fig. 168, was taken at West Albany (N. Y. C. & H. R. R. R.) 
 September 30, 1897. The engine was drawing five Wagner palace cars at a 
 speed of 40 miles per hour; 80-pound rail, 5 inches high; ties spaced 25 inches, 
 center to center. 
 
STRESSES IN THE RAIL 239 
 
 Records Nos. r 2 and 3, Fig. 169, show tests made at 2 miles per hour 
 and 10 miles per hour. The total weight of the locomotive was 96 tons; the 
 engine 60 tons, with 15,500 pounds on pony truck and 104,500 pounds on three 
 pairs of drivers. The tender weighed 72,000 pounds. The instrument was 
 placed on the outside rail on a 3-degree curve and down grade of 10 feet per 
 mile. The rail was 80-pound section, 5| inches high. The ties were yellow 
 pine, 7 by 9 inches, spaced 25 inches, center to center. Gravel ballast, the 
 track being in good condition. 30,000,000 pounds was taken as the modulus 
 of elasticity of the steel. 
 
 The two ties between which the stremmatograph was attached to the rail 
 were very firm in the ballast, and to the eye did not seem to depress as much 
 as those on either side; therefore the compression strains were probably 
 higher than on ties all practically depressing alike in the ballast. 
 
 Records Nos. 4 and 5, Fig. 169, were taken on 65-pound and 100-pound 
 rails, respectively. The 65-pound rails were of steel with an elastic limit of 
 60,000 pounds, while on the 100-pound rails it was 65,000 pounds. The loco- 
 motive was a switching engine, having 125,000 pounds upon drivers. The 
 instrument was placed between ties spaced 30 inches, center to center, having 
 tie plates. 
 
 Dr. Dudley states that tests with his stremmatograph show that the bend- . 
 ing moments in 80-pound rails under wheel loads used in 1905 may be as high 
 as 300,000 to 350,000 inch-pounds, indicating a unit fiber stress in the base of 
 the rail of as much as 30,000 or 35,000 pounds on worn 5-inch 80-pound sections. 
 With the 65-pound rail, stresses were frequently found as high as 40,000 to 45,000 
 pounds.* 
 
 23. Calculation of the Bending and Shearing Stress in the Rail 
 
 Examining, first, the moment and shear in the rail between two pairs 
 of driving wheels, we see that if the rail were completely rigid there would 
 result a uniform distribution of the wheel pressure to the ties, and if w equal 
 the pressure of one wheel divided by the wheel spacing, I, the resulting moment 
 and shear in the rail would then be 
 
 Maximum moment, M = zr~ wl 2 , 
 
 Maximum shear, J = — ■ 
 
 From the discussion on the support of the rail, we can take .35 ton as 
 a maximum value for w, and 150 inch-tons will be taken, for the present, as 
 
 * Private communication, June 7, 1912. 
 
240 
 
 STEEL RAILS 
 
 /' 
 
 
 
 / 
 / 
 / 
 / 
 / 
 
 \/ 
 
 / \ 
 
 \ \ 
 
 / / 
 
 
 \ 
 \ 
 \ 
 \ 
 
 \ 
 
 SPACING OF DRIVERS INCHES 
 
 Fig. 170. — -Wheel Loads for Different Spacing 
 of Drivers. 
 
 the maximum allowable moment, the following calculations being based on 
 100-pound A. S. C. E. rail. From the above we may construct the dotted 
 
 line shown in Fig. 170. 
 
 Considering, now, the effect of the 
 unequal distribution of the wheel pres- 
 sure to the ties: Fig. 142 shows the 
 relation of the reaction of the tie to the 
 depression in the ballast. We can, 
 therefore, by constructing the elastic 
 curve of the rail to give a shear curve 
 corresponding to the support of the 
 tie, given in Fig. 142, obtain the true 
 elastic curve of the rail, from which 
 the moment and shear diagrams can 
 be developed. 
 
 By this method we will assume the 
 probable curve of the rail and from this curve deduce the moment and shear 
 curves. The shear curve will give the reactions under the rail, and by comparing 
 these with the pressure exerted by the ties for the given depression, as shown 
 by the curve of the rail, the correctness of the latter can be checked. After a few 
 trials the actual curve taken by the rail for any span and wheel load can be drawn, 
 and the maximum bending moment and pressure on the ballast ascertained. 
 
 Mr. C. E. Love of the University of Michigan has made a very interesting 
 analysis of the results of the government tests of 1889, 1894, and 1895, described 
 in Article 22. As it is proposed to use the method adopted by Mr. Love a re- 
 view of one of the cases he has worked out will prove instructive and, on account 
 of the careful measurements made in the test selected, the accuracy of the 
 analysis can be shown before proceeding with its application to the general 
 case under consideration. 
 
 In Mr. Love's discussion the elastic curve of the rail, plotted from the careful 
 measurements given in these tests, is taken as the original curve, and the curve of 
 slopes, bending moment diagram, and shear diagram derived by the following 
 relations: Elastic curve, v=f(x), or ^ 
 
 Curve of slopes, C =f (x), or 
 
 dv 
 
 dx 
 
 Moment diagram, M = 
 Shear diagram, F = 
 
 EIf"(x), or EI ^, 
 Elf" (x), or EI ^- 
 
STRESSES IN THE RAIL 241 
 
 Plate XXIII shows one of the diagrams worked out by Mr. Love. This 
 plate is based on diagram No. 5 of the 1894 experiments.* The rail used was 
 
 30 feet long, with tie centers at points 1, 2, 3 18 (diagram A) and 
 
 loaded as shown, one horizontal space representing 2 inches. 
 
 In diagram B, points on the elastic curve were plotted from the observed 
 deflections and a smooth curve drawn through them, one vertical space 
 representing a deflection of .002 inch. This curve was then divided into 
 segments 6 inches (three spaces) in length along the horizontal axis and the 
 slope of each segment measured as if it had been straight. In determining the 
 tangent of each slope angle, the measurements were made in sixtieths of an 
 inch. The base, of course, was constant, three spaces equaling ^ inch of 
 actual distance on the paper, so that, calling the altitude dv, the actual slope was 
 tan = yi ' where dv was measured in sixtieths of an inch. In drawing the slope 
 curve, diagram C, however, one space corresponded to an altitude of gV inch. 
 For example, the extreme left-hand segment, diagram C, is three spaces below 
 the axis. This makes the value of one space in diagram C, 
 
 1 2 1 X 1 
 
 iSpaCe "l000'2'l4~i4^00* 
 
 The bending moment was constructed by plotting points directly from 
 diagram C, by taking the ordinates of the bending moments equal to the corre- 
 sponding breaks in the slope curve. The ordinates of the points in diagram D 
 are three times the actual slope of the curve in diagram C. The moment of 
 inertia (7) of the rail is 19.127, and a value of the modulus of elasticity (E) is 
 taken as 30,000,000. 
 
 d 2 v 
 Hence, since M = EI-r~ 2 , we have for the value of one space in diagram D, 
 
 1 space = — ^r- -|. |. 30,000,000 x 19.127 = 6800 inch-pounds. 
 
 The construction of diagram E is now very simple. The actual slope of a 
 segment of the moment curve was measured and the ordinate of the shear curve 
 taken as 10 spaces for each unit of the measured slope. Hence, for diagram E, 
 
 1 space = 6800- ^-^ = 340 pounds. 
 
 The bending moment at any point can now be scaled directly from dia- 
 gram D, and the change in shear under any tie in diagram E is the reaction at 
 the tie. 
 
 The total load on this rail is 56,500 pounds. After examining the re- 
 
 * House Executive Documents, 3rd Session, 53rd Congress, 1894-95, Vol. 30, Tests of Metals, etc. 
 
STEEL RAILS 
 
 actions at ties 1, 2, and 3, it is reasonable to suppose that 3000 pounds of the 
 load at No. 1 is supported by the rail to the left, thus leaving a net load of 53,500 
 
 pounds. The total up- 
 ward reaction is 52,300 
 pounds. 
 
 The error at No. 
 5| is 100 pounds, i.e., 
 the actual load is 16,- 
 000 pounds, while as 
 read from diagram E 
 it is 16,100 pounds; 
 at No. 10| the er- 
 ror is + 300 pounds; 
 at No. 13J, + 300 
 pounds. 
 
 Other experiments 
 were performed on this 
 rail, in which the fiber 
 stresses were meas- 
 ured. The maximum 
 bending moment here 
 is +98,500 inch- 
 pounds. In the other 
 experiments it was 
 117,000 inch-pounds, 
 caused by an engine 
 10 per cent heavier, 
 with the load at the 
 middle of a 24-inch 
 span. The minimum 
 here is - 24,000 inch- 
 pounds; in the other 
 experiments it was 
 - 38,500 inch-pounds. 
 Fig. 171 shows the 
 construction for 100- 
 pound A. S. C. E. rail 
 with spans of 60 inches, between centers of driving wheels. 
 
 
 
 
 
 
 1 
 
 O 1 0" 
 
 2 O" 3 O" 
 
 
 > 
 
 (T 
 
 
 
 
 0.26" 
 
 
 
 D 
 U 
 
 
 
 
 
 
 
 
 
 
 
 0.28" 
 
 
 
 CO 
 
 < 
 
 -J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Lu 
 
 
 
 
 
 
 
 
 
 
 
 
 O.JO'/' II 
 
 T 
 
 
 > 
 
 
 
 
 
 
 
 D 
 
 
 
 
 
 
 
 LU 
 
 Q. 
 
 
 
 
 
 
 
 
 
 
 
 O 
 
 _l 
 co 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 to 
 
 CD 
 
 
 
 
 
 
 LU 
 > 
 
 z* 
 
 
 
 
 
 
 
 
 
 .oooo 
 
 
 
 
 01 
 
 D 
 O 
 
 in 
 
 00 
 
 
 
 o vl 
 
 
 
 
 
 1- 
 
 Z 
 
 UJ 
 
 K) 
 
 UJ 
 
 
 
 
 
 
 
 
 
 
 Z 
 
 o 
 
 z 
 
 to 
 
 
 
 
 
 
 
 to 
 
 
 
 
 
 
 > 
 
 a. 
 
 to 
 <0 
 
 
 aoooc 
 
 
 
 
 
 >UB5.- s i:_J_L 
 
 
 
 
 
 o 
 
 Ul 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 CO 
 
 2 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
STRESSES IN THE RAIL 243 
 
 The construction of Fig. 171 is similar to that of Plate XXIII. The scale 
 to which the elastic curve is drawn is, 0.002 inch equals one space vertically, 
 and 2 inches equals one space horizontally. In deriving the slope curve, 
 the actual ordinates of the elastic curve are measured in fiftieths of an inch, 
 and one vertical space on the slope curve is taken to represent one-fiftieth of 
 an inch of the elastic curve ordinate. The base is taken in all the curves as 
 two spaces. 
 
 The value of one vertical space of the slope curve is, then, 
 2 1 1 1 
 
 1000*50' (&=*&) "2' 
 
 2 j. 50 1 
 1000*50' 10 *2* 
 
 2 _1 1 1 
 
 1000*10*2 10,000* 
 
 For example, the ordinate of the elastic curve at 2 inches to the right of the 
 
 ?el is about — inch; the actual slope is, therefore, tan = p. -r- ^ or — -=- — = .5, 
 5U 50 10 50 50 
 
 as shown by the slope curve between and 2 inches. The ordinate of the 
 
 elast 
 
 7.5 
 
 wheel is about — inch; the actual slope is, therefore, tan = p. -r- ^or— -=- — = .5, 
 50 50 10 50 50 
 
 itween 
 
 12 5 
 
 elastic curve at 4 inches is about -=_- or the rise from 2 inches to 4 inches is 
 50 
 
 rA - and the slope .75, as appears on the slope curve, between 2 inches and 4 
 50 
 
 inches. 
 
 The ordinates on the moment curve are taken the same number of spaces 
 
 as the corresponding breaks in the curve of slopes and consequently, on account 
 
 of the base being 2 spaces, represent twice the actual slope of the slope curve. 
 
 The value of one vertical space in the moment curve is, then, 
 
 1 1 1 FT 
 10,000 2 2 
 
 — [— ■ \>\ • 30,000,000 x 43.8 = 32,850 inch-pounds. 
 
 The ordinates of the shear curve are five times those of the moment curve, 
 and the value of one vertical space in the shear curve is 
 
 32,850-|-|-| = 1643 pounds. 
 
 The most simple way to construct the diagram is to approximate the 
 elastic curve of the rail and then follow out the operations, as shown by Table 
 LX, correcting the elastic curve from column 11 of the table and readjusting 
 the calculations, if necessary, to the corrected elastic curve values. 
 
244 
 
 STEEL RAILS 
 
 TABLE 
 
 LX. — CALCULATIONS OF RAIL DIAGRAMS FOR 60-INCH 
 
 WHEEL SPACING 
 
 Horizontal 
 
 Depression 
 track- 
 
 2 
 
 Pounds. 
 
 Shear. 
 
 Moment. 
 
 Slope. 
 
 Elastic curve 
 
 
 Pounds. 
 
 Spaces. 
 5 
 
 2Xtan. 
 6 
 
 Ordinate. 
 
 2Xtan. 
 
 Ordinate. 
 9 
 
 2Xtan. 
 
 10 
 
 Ordi- 
 11 
 
 
 
 0.30 
 
 "2800 " 
 
 20,080 
 
 12.2 
 
 2.44 
 
 6.1 
 4.9 
 
 
 
 
 
 
 
 
 2 
 
 4.9 
 
 2.5 
 
 4 
 
 
 17,280 
 
 10.5 
 
 2.10 
 
 4.9 
 
 4.9 
 
 
 6 
 
 
 2760 
 
 2.8 
 
 2.8 
 
 7.4 
 
 8 
 
 
 14,520 
 
 8.8 
 
 1.76 
 
 7.7 
 
 7.7 
 
 
 10 
 
 
 2720 
 
 1.0 
 
 1.0 
 
 15.1 
 
 12 
 
 
 11,800 
 
 7.2 
 
 1.44 
 
 8.7 
 
 8.7 
 
 
 14 
 
 
 2680 
 
 - .4 
 
 - .4 
 
 23.8 
 
 16 
 
 0.29 
 
 9,120 
 
 5.6 
 
 1.12 
 
 8.3 
 
 8.3 
 
 
 18 
 
 2640 
 
 -1.5 
 
 -1.5 
 
 32.1 
 
 20 
 
 
 6,480 
 
 3.9 
 
 .78 
 
 6.8 
 
 6.8 
 
 
 22 
 
 
 2610 
 
 -2.3 
 
 -2.3 
 
 38.9 
 
 24 
 
 
 3,870 
 
 2.4 
 
 .48 
 
 4.5 
 
 4.5 
 
 
 26 
 
 
 2590 
 
 -2.9 
 
 -2.9 
 
 43.4 
 
 28 
 
 0.28i 
 
 1,280 
 
 8 
 
 .16 
 
 1.6 
 
 
 
 1.6 
 
 
 
 30 
 
 1280 
 
 -3.1 
 
 -3.1 
 
 45.0 
 
 
 
 
 
 
 
 Note. — Col. 3 is found from col. 2 and Fig. 142. 
 
 Col. 4 is found from col. 3 and is I be total load rained from the center of the span. 
 Col. 5 is found from col. 4 by dividing the figures given in col. 4 by 1643. 
 Col. 6 is found from col. 5 by dividing the figures given in col. 5 by 5. 
 
 In Fig. 172 are given diagrams for spans of 70 inches, 80 inches, and 90 
 inches between centers of drivers. 
 
 The maximum bending moment can be read directly from the moment 
 curve and the wheel load is twice the load supported on half the span or twice 
 the total reaction shown by the shear curve of the diagrams. In comparing the 
 shear curve with the curve of pressures of the ties it would seem desirable to 
 assume the rail to be supported by a distributed load in place of a series of loads 
 concentrated at the ties, the effect will be practically the same and the calcula- 
 tions much simplified. 
 
 The results of static and dynamic tests of the stress in the rail both indi- 
 cate that the negative bending moment between the drivers is very much smaller 
 than the positive bending moment produced in the rail under the drivers. The 
 dynamic tests appear to show a ratio of about 1 to 4 or a compressive stress in 
 the base of the rail only one-fourth as great as the tensile stress, under normal 
 conditions. In poorly tamped track the compression stress seems to increase. 
 
 If the rail were uniformly supported between the wheels the compression 
 stress would be one-half the tension stress under the wheels of a set of drivers. 
 
 Such a condition of the pressure exerted by the tie would be represented by 
 a horizontal line in Fig. 142. The diagrams given in Figs. 171 and 172 show that, 
 as the spacing of the drivers increases, the negative bending moment or the 
 compression stress in the base of the rail decreases in relation to the tension 
 
STRESSES IN THE RAIL 
 
 245 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 WHEEL 
 
 TO" 
 
 CJOC 
 
 
 
 
 WHEEL 
 
 3 80" 
 
 CJOC 
 
 
 
 
 wheel: 
 
 90- 
 
 CioC 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 /■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 > 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 / 
 
 
 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 y 
 
 
 
 
 
 
 o 
 
 
 
 
 
 / 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 In 
 < 
 
 
 
 ^)^o 
 
 / 
 
 / 
 
 
 
 
 
 y 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r 
 
 \ 
 
 
 
 
 UJ 
 
 > 
 
 
 
 
 
 ^N 
 
 
 
 
 
 
 
 
 s 
 
 
 
 / 
 
 
 
 \ 
 
 
 
 a. 
 
 D 
 O 
 
 
 
 
 / 
 
 
 \ 
 
 
 
 
 
 
 
 \ 
 
 
 
 / 
 
 
 
 \ 
 
 
 
 uJ 
 
 
 
 
 / 
 
 
 
 \ 
 
 
 
 
 
 
 \ 
 
 
 
 / 
 
 
 
 
 \ 
 
 
 g 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 sooo 
 
 
 \ 
 
 
 
 
 oooo 
 
 
 \ 
 
 
 
 
 
 3IN.LBS 
 
 \ 
 
 
 
 
 
 
 r? 
 
 
 
 
 \ 
 
 \ 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 s 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i- 
 z 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2aocx 
 
 LBS 
 
 \ 
 
 
 
 
 240CX 
 
 )LBS 
 
 \ 
 
 
 
 
 220CX 
 
 
 
 
 
 
 
 
 > 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 to 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 172. — Rail Diagram for Wheel Spacing of 70, 80, and 90 inches, one-half size of original diagram. 
 
246 STEEL RAILS 
 
 stress; for drivers spaced 60 inches apart it is practically one-half the tensile 
 stress, but when the spacing is increased to 90 inches it is not much more than 
 one- third. 
 
 An examination of Fig. 142 makes this clear and shows that with the 
 greater deflection obtained in the 90-inch span the ties in the center of the 
 span support relatively much less of the load. With lighter rail the deflection 
 would be still further increased and a greater ratio of the tension to the 
 compression stress obtained. 
 
 The full line, shown in Fig. 170, shows the true allowable wheel loads given 
 by the diagrams of Figs. 171 and 172. Up to spans of 80 inches the wheel load 
 is limited by the safe bearing power of the tie and is obviously less than that 
 obtained from the assumption that there is a uniform distribution of the wheel 
 pressure to the ties upon which the dotted line of Fig. 170 is based. After the 
 spacing of the drivers exceeds 80 inches the wheel load is limited by the bending 
 moment in the rail; here the bending moment is greater for a uniformly dis- 
 tributed load and, consequently, the dotted line in the figure falls below the 
 full line. 
 
 Turning our attention to the allowable wheel load as determined by the 
 conditions at the front and rear drivers. Figs. 162 and 163 show that there is 
 a wave motion of the rail ahead of the engine and the rail rises slightly above 
 the trackman's surface. This lack of pressure on the rail at the outside wheels 
 causes these wheels to exert a more severe action on the rail. This is clearly 
 shown by Table LVIII and in Fig. 162, where the outside drivers, although 
 carrying less weight, gave practically the same stress as the middle driver. 
 Records Nos. 2 and 3 of Fig. 169 illustrate the same tendency. 
 
 The rail diagrams, just worked out, show that in increasing the wheel 
 spacing from 80 to 90 inches the permissible wheel load fell from 24 to 22 tons. 
 From the diagram for the 90-inch wheel spacing it is seen that the ties in the 
 middle of the span afford little support, and while it is somewhat problematic 
 what load will be carried by the rail ahead of the front driver, we will not be 
 very far wrong if we assume it to carry 8 tons with no leading truck, 9 tons 
 with a two-wheel leading truck, and 10 tons with a four-wheel leading truck. 
 
 On account of the load of the tender wheels and the effect of the draw- 
 bar pull, we may reasonably take 10 tons where a trailing truck is used and 
 9 tons where there is no trailer, as the load carried by the rail back of the rear 
 drivers. 
 
 Table LXI may now be prepared showing the allowable dynamic wheel 
 load under different conditions of wheel spacing. 
 
STRESSES IN THE RAIL 
 
 TABLE LXI.— ALLOWABLE DYNAMIC WHEEL LOAD (POUNDS) FOR 
 100-POUND A. S. C. E. RAIL 
 
 Middle wheel 
 
 Front wheel, 
 
 No leading truck 
 
 Two-wheel leading truck. . 
 Four-wheel leading truck. . 
 
 Back wheel, 
 
 No trailing truck 
 
 Trailing truck 
 
 36,000 
 38,000 
 40,000 
 
 38,000 
 40,000 
 42,000 
 
 40,000 
 42,000 
 
 Inches. 90 Inches 
 
 40,000 
 
 42,000 
 44,000 
 
 38,000 
 40,000 
 42,000 
 
 Referring to Figs. 31 and 32, of typical load diagrams of engines, it will be 
 seen that with the exception of the articulated engine there is had a very satis- 
 factory agreement between Table LXI and the diagrams. 
 
 We have now to consider the stresses in the rail caused by the bending 
 moment and shear derived in Figs. 171 and 172. The maximum bending 
 moment in these figures is 300,000 inch-pounds, and the maximum shear is 
 24,000 pounds. 
 
 It is beyond the scope of the present work to enter into the discussion 
 of mathematical investigations of continuous web strains, and in order to form 
 some conception of the nature of stresses in the continuous rail we shall view 
 the matter in the simplest manner possible.* 
 
 In the rail under the wheel it is evident that, by virtue of the bending 
 stress, that part of the rail above the neutral axis is subject to compression, 
 and that below to tension, both of which stresses attain maximum values at 
 the outermost fibers of the rail, and decrease to zero at the neutral axis. This 
 intensity of the stress at any point is at once obtained from the well-known 
 equation of flexure: 
 
 M 
 
 ~y = f, 
 
 (a) 
 
 where M is the bending moment, I the moment of inertia of the section of the 
 rail, y the distance of the point from the neutral axis, and / the intensity 
 of the stress at that point. 
 
 Table LXII gives the extreme fiber stress in the base due to bending in 
 different sections of 100-pound rail, caused by a bending moment of 300,000 
 inch-pounds. The high moment of inertia in the Series "A" of the American 
 Railway Association would giVe a slightly different elastic curve for this rail 
 than is shown for the A. S. C. E. section in Figs. 171 and 172, with the result 
 
 * See Plate Girder Construction, Isami Hiroi, New York. 
 
248 
 
 STEEL RAILS 
 
 that the bending moment would be increased and the unit of load supporting 
 the rail decreased. 
 
 TABLE LXIL— EXTREME FIBER STRESS DUE TO A 
 BENDING MOMENT OF 300,000 INCH-POUNDS 
 
 Section. 
 
 Weight. 
 
 Extreme Fiber 
 
 A. S. C. E 
 
 Am. Ry. Assn., Series "A" 
 
 Am. Ry. Assn., Series "B" 
 
 Pounds per 
 Yard. 
 100 
 100 
 
 100 
 
 Pounds per 
 
 18,600 
 16,900 
 19,100 
 
 Fig. 173. — Distribution of Horizontal Stress in Rail. 
 
 The bending moment M decreases as we proceed toward the center of the 
 space between the wheels, and with it evidently the intensity / of the hori- 
 zontal stress also; so that / varies not only in vertical directions on both sides 
 of the neutral axis, but also in the direction of the length of the rail. 
 
 x , Let xx and x'x', in Fig. 173, be 
 
 r\ "\" "P* two sections of a rail, very close to 
 
 each other, and NN the neutral axis. 
 The variation of the value of / 
 in both sections may be represented 
 by triangles with apices in the 
 neutral axis, and the variation in 
 the longitudinal direction between 
 these two sections by the difference of the areas of two triangles, as shown 
 shaded in the figure. This increase of horizontal stress from one section to 
 another produces at each longitudinal layer a force tending to slide it past the 
 layer next above it, and is transmitted undiminished toward the neutral axis, 
 where this shearing force, which has been increasing at every layer, attains its 
 maximum intensity. 
 
 This stress is called the longitudinal shear, and can be at once obtained 
 from equation (a). Thus, let /' be the corresponding value of / in section x'x'; 
 and let M and M' be the bending moment in the two sections xx and x'x' re- 
 spectively, and a an infinitely small cross area, distant y from the neutral axis. 
 The total horizontal stresses acting in that part of the section lying between 
 the extreme fiber distant h from the neutral axis and the layer y'y' distant y' 
 from the axis in xx and x'x' are respectively: 
 
 2>> 
 
 %f'a. 
 
STRESSES IN THE RAIL 249 
 
 The longitudinal shear in the layer y'y' between the two sections is, there- 
 fore, equal to 
 
 h h 
 
 %fa - %fa. 
 
 1/ 2/' 
 
 Substituting in the expression the values of / and /' given by equation 
 
 (a), we obtain, 
 
 ± f , 4w M' - M v 
 Zii a- Z,f a= f — Z, ya - 
 
 y' y' 1 2/ 
 
 Since the area on which this horizontal shear is acting is equal to b Ax, 
 when b is the breadth of the cross section at the layer y'y' and Ax the distance 
 between x and x', we obtain for the intensity of the shear, 
 
 -Jb^-^- (b) 
 
 Thus at every point in the rail there are two shearing actions taking place 
 at the same time, one the longitudinal shear and the other the vertical shear. 
 
 Imagine abed, Fig. 174, to be an infinitely small a ^-^ x - h - ^ 
 
 portion of the side of a rail at a point distant y' ^ ^ x 
 
 from the neutral axis. Suppose the side of this J* ^ ^y *•' 
 
 area element to be Ax and Ay, and the breadth of the J' J -j -J-' 
 
 beam at the point to be b. There are then found two FlG _ 174. —shearing stress of Point 
 shearing stresses on this element, one vertical and Distant y' from Neutral Axis, 
 the other horizontal. These two shears form two pairs of couples acting around 
 the body, as shown by the arrows. Let t x represent the intensity of the horizontal 
 shear at this point and t y that of the vertical. The amount of the horizontal 
 shear is equal to t x Ax b; that of the vertical shear is likewise equal to t y Ayb. 
 In order that the body be in equilibrium, the moment of these couples 
 must be equal, i.e., t x ax b Ay = t y Ay b ax. Consequently, t x = t y (which is 
 always the case), showing that at every point in the rail the intensities of the 
 vertical and horizontal shears are equal, and we will hereafter designate them 
 with the common letter t. The value of t x has already been deduced in equation 
 (b), namely: 
 
 , M ' - M ^ . . 
 
 - M 
 
 — = S z , where S x is the total vertical shear at the sec- 
 
 jaX 
 
 Substituting this value of — in equation (c) , there results: 
 
 t = fj%ay, (d) 
 
250 
 
 STEEL RAILS 
 
 or the intensity of the shearing stress at any point in the rail is equal to the 
 total shearing force on the entire cross section multiplied by the statical moment 
 of the area of the section outside the longitudinal plane of shear in question 
 about its axis in the neutral plane, divided by the product of the moment of 
 inertia of the entire section into the breadth of the section at that point. 
 
 Fig. 175 shows the intensity of the shearing stress in a 100-pound rail, 
 the total vertical shear at the section being 24,000 pounds. 
 
 Fig. 175. — Shearing Stress in 100-pound A. S. C. E. Rail. 
 
 There still remains to be considered the horizontal force /, whose value 
 is given in equation (a), tending either to compress together or pull asunder 
 the two faces ac and bd (Fig. 174), according as it is on the upper or lower side 
 of the neutral axis. 
 
 At the neutral axis where / = 0, t x and t y are then the only stresses, and 
 we know from mechanics that the resultant action of two equal shears at right 
 angles to each other, exactly as t x and t y are, is equivalent to that of two equal 
 and opposite stresses at right angles to each other, called the principal stresses 
 and making an angle of 45° with the shearing stresses. But at a distance each 
 
STRESSES IN THE RAIL 251 
 
 side of the neutral axis the third stress, /, now comes in, which evidently gives 
 a new direction to the line of resultant stress, turning the axis of principal stresses 
 toward itself more and more as its intensity increases. 
 
 Fig. 176 represents the appearance which the lines of principal stresses 
 thus obtained present in a beam loaded in the middle and supported at each 
 end. The lines of maximum tension are shown dotted and cut the lines of 
 compression always at right angles. Both lines cross the neutral axis at an 
 inclination of 45° and run almost parallel to it in the middle of the beam in the 
 neighborhood of extreme fibers. 
 
 Fig. 176. — Lines of Principal Sti 
 
 Now comes the question how the web should be proportioned to resist 
 such stresses : The greatest intensity of the vertical shearing stress on the verti- 
 cal section of the rail, shown in Fig. 175, is about* 10,000 pounds per square inch. 
 In modern bridge practice a shearing stress is allowed in web plates of 10,000 
 pounds per square inch, which gives a satisfactory thickness of the web for the 
 rail shown in the figure. But as has already been explained, the action of the 
 shearing stresses at the neutral axis is equivalent to compression and tension 
 at right angles to each other and of equal intensity, making an angle of 45° with 
 the axis, and the web is still in danger of failing by flexure under this com- 
 pression stress. 
 
 Consequently, the web with its thickness as already proportioned for 
 shearing must now be examined for its strength as a column. We will probably 
 be not far from correct if the length of the column is taken as h sec 45°, h being 
 the vertical distance between the top of the flange and bottom of the head of 
 the rail. Then, for the 100-pound A. S. C. E. rail, h sec 45° equals 4.4 inches, 
 and the load is 
 
 = p = 10,000 pounds per square inch. 
 
 This amount is correct for the bending stress caused by the load which is 
 central over the rail head. The wheel load is, however, rarely applied exactly 
 in line with the vertical axis of the rail, and the additional couple due to the 
 eccentricity causes a torsion in the rail. To make a correct analysis would 
 
252 STEEL RAILS 
 
 be very complicated and decidedly uncertain on account of the lack of experi- 
 mental evidence. 
 
 No column formula can be made to apply exactly to the web of the rail. 
 If we apply the formula for a column with an eccentric loading of .6 of an inch, 
 the resulting stress amounts to over 50,000 pounds per square inch. It is 
 doubtful, however, whether the stress introduced by this torsion can be com- 
 bined with those due to bending in this manner. 
 
 It will be observed that, even were this large stress correct, there is a 
 tensile stress acting at right angles to the compression stress and tending to 
 hold the strip in its true plane. Just what the restraining influence of this 
 tensile stress is cannot be determined theoretically, but the following experi- 
 ments show it to be of importance. 
 
 During 1910 tests were made at the Maryland Steel Company's plant at 
 Sparrows Point, Md., for the purpose of determining what effect the eccentric 
 loading of the wheel had on the head and web of the rail.* The tests Were made 
 with a 200,000-pound test machine by canting a piece of rail 18 inches long 
 and applying the load at the edge by means of a block with a radius of 16^ 
 inches, to represent a car wheel, where it came in contact with the rail. Other 
 tests were made with a reciprocating machine representing a loaded wheel 
 rolling back and forth on the edge of the canted rail. 
 
 For the tests a rail was taken from stock and six pieces each 18 inches 
 long were cut from it for test in the stationary test machine and six similar 
 pieces were used for test in the reciprocating machine. In order to have the 
 material as uniform as possible throughout the section and in the different 
 pieces, a " C " rail was selected, that is, the third rail from the top of the rail 
 bar. The rail was a 90-pound A. R. A. type B section and the pieces were planed 
 down to thicknesses of head at the side of | inch, \ inch, f inch, § inch, | inch, 
 and 1 inch, two pieces of each thickness, one for each kind of test. In each 
 case the brand side of the head of the rail, which was the bottom side as rolled, 
 was planed vertical to a width of li 3 g inches from the center line. Fig. 177 
 shows the dimensions of the section used and also gives diagrams of the pieces 
 tested. The essential dimensions of the head of the pieces tested, as indicated 
 by letters A, B, and C, on Fig. 177, measured as shown in Table LXIII. 
 
 Two samples were taken with a |-inch drill for analysis from a section 
 near the middle of the length of the rail, one close to the upper corner and the 
 other at the junction of the head and the web. The results of the analyses are 
 shown in Table LXIV. 
 
 * Strength of Rail Head, M. W. Wickhorst, Proceedings Am. Ry. Eng. & M. of W. Assn., 
 1911, Vol. 12, Part 2, p. 518. 
 
STRESSES IN THE RAIL 
 
 253 
 
 These results show the material to be very uniform. 
 
 TABLE LXIII. — DIMENSIONS OF HEADS AS TESTED FOR 
 STRENGTH OF RAIL HEAD 
 
 Test Numbers. 
 
 Thickness of Head. 
 
 Width. 
 
 Edge, 
 
 Center, 
 B. 
 
 Side to Center, 
 C. 
 
 
 1.02 
 91 
 .76 
 .62 
 .51 
 .38 
 
 1.23 
 1 15 
 1.02 
 .86 
 .75 
 .64 
 
 1.15 
 
 1.18 
 
 1.17 
 1.15 
 1.16 
 1.17 
 
 3 and 4 
 
 5 and 6 
 
 7 and 8 
 
 
 11 and 12 
 
 
 TABLE LXIV. — ANALYSES 
 FOR STRENGTH OF 
 
 OF RAILS TESTED 
 RAIL HEAD 
 
 
 Corner of 
 Head. 
 
 Junction of 
 Head and Web. 
 
 
 .538 
 
 .523 
 .070 
 .055 
 .81 
 .103 
 .18 
 None 
 
 Phosphorus 
 
 
 070 
 050 
 81 
 103 
 19 
 N"one 
 
 
 Silicon 
 
 Copper 
 
 Nickel 
 
 Chromium 
 
 Tensile tests were also made of pieces cut from near the middle of the rail, 
 two pieces |-inch diameter and 2-inch gauge length, for longitudinal test from 
 the center of the head, and two pieces |-inch diameter and 1-inch gauge length 
 for transverse test across the center of the head. The yield point in the 2-inch 
 pieces was determined by means of a Capp's multiplying dividers. The results 
 of the tests are shown in Table LXV. 
 
 TABLE LXV. — TENSILE TESTS OF RAILS TESTED FOR 
 STRENGTH OF RAIL HEAD 
 
 
 Yield Point 
 (Pounds per 
 
 Tensile 
 
 Strength 
 
 (Pounds per 
 
 Square Inch). 
 
 Elongation. 
 
 Eeduction of 
 
 Longitudinal a 
 
 2-inch gauge length b 
 
 Average 
 
 Transverse a 
 
 1-inch gauge length b 
 
 Average 
 
 51,000 
 52,700 
 51,850 
 
 111,600 
 111,500 
 111,550 
 
 110,200 
 111,200 
 110,700 
 
 16 
 
 16.5 
 
 16.3 
 
 6 
 7 
 6.5 
 
 29 
 29 
 
 29 
 
 7 
 9 
 8 
 
 
 
 These results show material of good ductility longitudinally and the stretch 
 crosswise of the head shows up well for a transverse test. 
 
STEEL RAILS 
 
 - Diagram of Pieces tested for Sag of Rail Head and I 
 (Am. Ry. Eng. Assn.) 
 
 «. 4':^ 
 
 
 
 4" 
 
 
 
 CT^sJ 
 
 : — — ^° 
 
 
 
 
 
 — 18" 
 
 
 - 
 
 
 — — — -— i -_-^__^ ,. -^ 
 
 
 -1 L 
 
 
 
 1 
 
 Fig. 178. — Method of Stationary Tests for Sag of Rail Head and Bending of Web. 
 (Am. Ry. Eng. Assn.) 
 
 The arrangement used for making the stationary tests is shown in Fig. 
 178, and is intended to represent a 33-inch car wheel resting on the edge of 
 the top of the rail. The head is thus tested as a cantilever, the load tending 
 to sag the head locally and to also bend the web. 
 
STRESSES IN THE RAIL 
 
 255 
 
 The load was applied in increments of 10,000 pounds up to 60,000 pounds 
 and then in increments of 20,000 pounds up to 200,000 pounds, the capacity 
 of the test machine. The sag of the head was determined by measuring the 
 distance by means of dividers, between prick-punch marks placed on the side 
 of the head near the bottom and on the base, as indicated in Fig. 178, the 
 load being on while taking the reading. The marks on the base were placed 
 about one inch from the web, by gouging some of the metal so as to have a 
 vertical surface on which to prick-punch the mark. The amount that the 
 opposite side of the head elevated, or the " lift," was determined in a similar 
 manner. The results of these tests are shown in Table LXVI. 
 
 TABLE LXVI. 
 
 -STATIONARY TESTS IN TEST MACHINE OF STRENGTH 
 OF RAIL HEAD 
 
 Sag and Lift in Inches 
 
 w 
 
 i -inch Head. 
 
 3-inch Head. 
 
 f-inch Head. 
 
 1-inch Head. 
 
 |-inch Head. 
 
 1-inch Head. 
 
 Sag. 
 
 Lift. 
 
 Sag. 
 
 Lift. 
 
 Sag. 
 
 Lift. 
 
 Sag. 
 
 Lift. 
 
 Sag. 
 
 Lift. 
 
 Sag. 
 
 Lift. 
 
 10,000 
 
 .00 
 .02 
 .05 
 .06 
 
 .08 
 .09 
 
 .00 
 .00 
 .00 
 .01 
 .01 
 .02 
 
 .00 
 .01 
 .02 
 .03 
 .04 
 .05 
 .06 
 .07 
 .08 
 .09 
 .10 
 .11 
 .13 
 
 .00 
 .00 
 .01 
 .01 
 .01 
 .02 
 .02 
 .02 
 .02 
 .02 
 .02 
 .02 
 .02 
 
 .00 
 
 .00 
 .00 
 .00 
 .00 
 .01 
 .01 
 .02 
 .03 
 .03 
 .03 
 .04 
 .04 
 .04 
 
 .00 
 
 .01 
 .01 
 .01 
 .02 
 .02 
 .02 
 .03 
 .03 
 .03 
 .03 
 .03 
 .03 
 .03 
 
 
 
 .00 
 .00 
 .01 
 .02 
 .02 
 .03 
 .04 
 .04 
 .05 
 .05 
 .06 
 .07 
 .08 
 
 00 
 
 20,000 
 
 
 01 
 02 
 04 
 06 
 07 
 10 
 11 
 12 
 13 
 14 
 15 
 16 
 
 
 00 
 01 
 01 
 02 
 03 
 04 
 05 
 07 
 OS 
 os 
 09 
 10 
 
 
 01 
 01 
 
 02 
 03 
 04 
 05 
 06 
 07 
 OS 
 08 
 09 
 
 in 
 
 .02 
 .02 
 .02 
 .03 
 .03 
 .04 
 .04 
 .04 
 .04 
 .04 
 .04 
 .03 
 
 
 30,000 
 
 40,000 .... 
 
 .00 
 
 50,000 
 
 01 
 
 60,000 
 
 02 
 
 80,000 
 
 .02 
 
 100,000 
 
 
 
 02 
 
 120,000 
 
 
 
 .03 
 
 140,000 
 
 .15 
 .17 
 .20 
 
 .03 
 .03 
 .05 
 
 03 
 
 160,000 
 
 .03 
 
 180,000 
 
 03 
 
 200,000 
 
 .02 
 
 
 
 
 
 
 
 
 
 
 
 These results 
 are plotted in Fig. 
 179, in which the 
 load is plotted 
 against the sag of 
 head for each thick- 
 ness of head tested. 
 The f-inch head, 
 according to these 
 curves, gave a 
 greater sag than 
 
 £h e --inch head. •^ IG - 1 ^- — ^ag °f *t au Head in Stationary Tests. (Am. Ry. Eng. Assn.) 
 
 Although this is according to the measurements obtained, it would seem to 
 be in error, due, perhaps, partly to errors of measurement, but probably also 
 
256 
 
 STEEL RAILS 
 
 due to some condition which cannot be accounted for, as, for instance, 
 application of the load. 
 
 The curves show that a load of 10,000 pounds does not sag the head with 
 the load applied to the edge of the top side, with any thickness down to f inch, 
 and probably neither does a load of 20,000 pounds, although, as the load was 
 on when the measurement was taken, we cannot say how much of the sag was 
 elastic and how much permanent. A load of 30,000 pounds seems to cause a 
 permanent sag with the f-inch head, but not much, if any, with the heads of 
 greater thickness. 
 
 It is interesting to note in this connection that the web seemed to stand 
 the load of 200,000 pounds successfully. 
 
 Tests were also made with the reciprocating machine, shown in Fig. 151, 
 in which a piece of rail is moved back and forth under a wheel to which a 
 load can be applied by means of a system of levers. The rail is fastened to 
 a steel bloom which runs on rollers running on another steel bloom that forms 
 the bed of the machine. The rail bed is connected by means of a connecting 
 rod to the bed plate of a planer, which furnished the power to run the rail machine. 
 The weights attached to the weight hanger are multiplied 600 times as applied 
 to the axle of the wheel, and in these tests weights of 50, 100, and 150 pounds 
 were used, so that the wheel loads were 30,000, 60,000, and 90,000 pounds re- 
 spectively. A piece of rail 18 inches long was tilted on an inclined plane of 
 1 in 10, as in the other tests, and the wheel, loaded with 30,000 pounds, was 
 run back and forth over a length of about 10 inches for 100 double strokes, 
 which made 200 movements of the loaded wheel over the rail. The sag of the 
 head and the width of the bearing taken by the wheel were then measured with 
 no load; the load was then increased to 60,000 pounds and the wheel again 
 run on the rail as before. The measurements were again taken and a final test 
 made with a load of 90,000 pounds. The results of these tests proved to be 
 interesting and fairly definite, and are shown in Table LXVII. 
 
 TABLE LXVII. — RESULTS OF ROLLING TESTS OF STRENGTH OF RAIL HEAD 
 
 Thickness of Head. 
 
 Sag of Head. 
 Inches. 
 
 Width of Bearing. 
 Inches. 
 
 30,000 
 Pounds. 
 
 60,000 
 Pounds. 
 
 90,000 
 Pounds. 
 
 30,000 
 Pounds. 
 
 60,000 
 Pounds. 
 
 Si 
 
 ^ inch 
 
 .05 
 .01 
 .00 
 .00 
 
 .00 
 
 .00 
 
 .13 
 
 .08 
 .05 
 .03 
 .01 
 .00 
 
 .17 
 
 .13 
 .13 
 .10 
 .08 
 .05 
 
 .56 
 .34 
 .31 
 .32 
 .32 
 .23 
 
 1.20 
 1.00 
 
 .76 
 .70 
 .58 
 .60 
 
 1.56 
 1.42 
 1.40 
 1.28 
 1.16 
 1.00 
 
 1 jj^h 
 
 1 - mc h 
 
 3 jjjgjj 
 
 | inch 
 
 
STRESSES IN THE RAIL 257 
 
 The results showing the relation of thickness of head and sag of head under 
 loads of 30,000, 60,000, and 90,000 pounds are plotted in Fig. 180. It will be 
 noted that 30,000 pounds 
 produces no sag when the 
 head is f inch or over in thick- 
 ness. With 60,000 pounds 
 the head must be 1 inch 
 or over in thickness. All 
 the samples tested sagged 
 under 90,000 pounds, but 
 by extending /the curve 
 it seems probable that a 
 head If inches thick would 
 hold up a rolling load of 
 90,000 pounds when con- 
 centrated at the edge. 
 
 < 
 
 Ul 
 
 U. 
 
 
 
 £.05 
 
 V 
 
 
 ROLLING TESTS 
 
 s 
 
 
 ^S^ 90,ooo 
 
 LBS. 
 
 
 
 
 S^ ^ -60,000 
 
 LBS. 
 
 
 
 
 30,ooo 
 
 LBS. 
 
 
 
 
 %" l/ 2 " 5/ 8 " %" 7/ e " ,•• 
 
 THICKNESS of HEAD 
 
 - Sag of Rail Head in Rolling Tests. 
 (Am. Ry. Eng. Assn.) 
 
 After the rails were tested they were cut in two and their sections are 
 shown in comparison with the original rail section in Fig. 181. 
 
 The rolling load of 90,000 pounds applied at the edge of the head pro- 
 duced very little or no bending of the web with the section used, which was a 
 90-pound A. R. A. type B, with a thickness of about -^ inch at the middle. 
 
 While there seems little liability of the web failing as a column, the height 
 of the rail in reference to the stability of the outer rail on curves must be 
 considered. 
 
 * Mr. E. E. Stetson found that in many cases the resultant of the hori- 
 zontal force and wheel pressure on curves falls entirely outside the base of the 
 rail in the 100-pound section. 
 
 In 1907 the Pennsylvania Railroad Company prepared a piece of track 
 on the West Jersey and Seashore Railroad, on a 1-degree curve, near Franklin- 
 ville, N. J., with special cast-steel ties and measuring apparatus, for the pur- 
 pose of comparing the effects of lateral horizontal forces on the outer rail of the 
 curve generated by different classes of electric locomotives and standard steam 
 locomotives. 
 
 The force exerted by the locomotive was communicated to steel plates by 
 means of hardened steel spheres of small diameter, and the effect of the force 
 
 * A Study of Rail Pressures and Stresses in Track Produced by Different Types of Steam Loco- 
 motives when Rounding Various Degree Curves at Different Speeds. E. E. Stetson, Proceedings Am. 
 Ry. Eng'. & M. of W. Assn., 1909, Vol. 10, Part 2, p. 1432. 
 
258 STEEL RAILS 
 
 was to cause the spheres to make a more or less deep impression in the steel 
 
 By means of laboratory tests, it was determined what forces in pounds 
 were required to produce various known depths of the impression of the steel 
 balls in the plates, and after this calibration had been made it was possible to 
 transform the depths of the impressions of the balls in the plates into pounds 
 
 Fig. 181. — Rails after Rolling Test with Load of 90,000 Pounds. (Am. Ry. Eng. Assn.) 
 
 of pressure. Table LXVIII-A gives the results of some of these tests for steam 
 locomotives, in order that they may be compared with the computations made 
 by Mr. Stetson, which are shown in Table LXVIII-B for the same degree of 
 curve and for one locomotive of the same class as used in the tests on the 
 Atlantic City line. Mr. Stetson's calculations are for speeds of 60 miles per 
 hour, which are from twenty to thirty miles per hour less than the actual 
 tests, but the weights of the locomotives are heavier. The lower speeds should 
 give smaller pressures, while, on the other hand, the heavier locomotives should 
 give higher pressures. 
 
STRESSES IN THE RAIL 
 
 259 
 
 TABLE LXVIII. - 
 
 HORIZONTAL PRESSURES EXERTED BY STEAM LOCO- 
 MOTIVES AGAINST RAIL ON CURVES 
 
 (Am. Ry. Eng. Assn.) 
 
 Run 
 
 Number. 
 
 Speed in Miles 
 
 Speed Corre- 
 Superelevation. 
 
 Type of Loco- 
 
 Condition of 
 Rafl. 
 
 Depth of Im- 
 
 Maximum 
 Pressure in 
 Pounds. 
 
 
 
 
 
 
 Inch 
 
 
 9 
 
 89.4 
 
 70 
 
 B 
 
 Dry 
 
 0.204 
 
 10,500 
 
 10 
 
 87.7 
 
 70 
 
 B 
 
 Dry 
 
 0.246 
 
 13,000 
 
 17 
 
 92.3 
 
 70 
 
 B 
 
 Dry 
 
 0.222 
 
 11,500 
 
 18 
 
 90.5 
 
 70 
 
 B 
 
 Dry 
 
 0.216 
 
 11,200 
 
 19 
 
 85.03 
 
 70 
 
 B 
 
 Dry 
 
 0.199 
 
 10,300 
 
 20 
 
 79.50 
 
 70 
 
 B 
 
 Dry 
 
 0.193 
 
 10,100 
 
 21 
 
 75.5 
 
 70 
 
 B 
 
 Wet 
 
 0.162 
 
 8,500 
 
 22 
 
 80.7 
 
 70 
 
 B 
 
 Wet 
 
 0.217 
 
 11,200 
 
 23 
 
 81.29 
 
 70 
 
 B 
 
 Wet 
 
 0.217 
 
 11,200 
 
 111 
 
 85.30 
 
 70 
 
 B 
 
 Dry 
 
 0.179 
 
 9,500 
 
 118 
 
 80.3 
 
 70 
 
 B 
 
 Wet 
 
 0.181 
 
 9,500 
 
 119 
 
 83.9 
 
 70 
 
 B 
 
 Wet 
 
 0.188 
 
 10,000 
 
 120 
 
 83.9 
 
 70 
 
 B 
 
 Wet 
 
 0.199 
 
 10,300 
 
 11 
 
 81.3 
 
 70 
 
 D 
 
 Dry 
 
 0.165 
 
 8,700 
 
 12 
 
 83.5 
 
 70 
 
 D 
 
 Dry 
 
 0.134 
 
 7,000 
 
 Class B is an Atlantic type locomotive, total weight = 176,600 pounds, height center of gravity 
 above base = 73 inches. 
 
 Class D is an American type locomotive, total weight = 138,000 pounds, height center of gravity 
 above base = 65 inches. 
 
 TABLE B. 
 
 — RESULTS OF COMPUTATIONS MADE 
 
 BY E. E. STETSON FOR A 1-DEGREE CURVE 
 
 Speed in Miles 
 per Hour. 
 
 Speed Corre- 
 sponding to 
 Superelevation. 
 
 Type of Loco- 
 
 Pounds. 
 
 Remarks. 
 
 60 
 70 
 
 60 
 70 
 
 60 
 70 
 
 60 
 60 
 
 60 
 60 
 
 60 
 60 
 
 ClaSS B 
 
 Class B 
 
 Class A 
 Class A 
 
 Class C 
 Class C 
 
 11,500 
 12,950 
 
 11,120 
 12,830 
 
 13,180 
 14,700 
 
 Class B is an Atlantic type, total weight = 
 
 183,150 pounds, height center of gravity 
 
 above rail taken as 70.5. 
 Class A is Pacific type, total weight 270,100 
 
 pounds, height center of gravity above rail 
 
 = 76.25. 
 Class C is consolidation type, total weight 
 
 238,200 pounds, height center of gravity above 
 
 rail = 62. 
 
 24. Effect of the Joint 
 
 The preceding discussion is based on the assumption that the joint affords 
 100 per cent efficiency. 
 
 If we examine the functions the joint performs in carrying the load 
 from one rail to the other, we see that the splice bars, by fitting tightly to the 
 inclined surfaces of the head and base of the rail, are able by their friction to 
 transmit large horizontal strains from one rail to the next. The proportion 
 of the bending moment of the rail transmitted to the splice bar by this means 
 is important in determining the correct proportions of the joint. 
 
260 
 
 STEEL RAILS 
 
 To determine the friction of the bar the following tests were made at the 
 Watertown Arsenal in 1904.* There was first made a series of track observa- 
 tions on the Boston and Albany Railroad at Faneuil station, near Boston, to 
 determine the resistance of nuts on bolts of splice bars as found in the track 
 against further tightening. 
 
 Tests were made with a wrench 33 inches long, the resistance against tighten- 
 ing being shown by the force required at the end of the wrench to turn the nuts 
 forward. The average of 60 observations was 52 pounds on a 33-inch wrench. 
 
 Tests were then made at the Arsenal on the frictional resistance of two 
 6-hole splice bars on two sections of 6-inch 100-pound rail. Spring nuts 
 were used under the nuts, f-inch bolts, 10 threads per inch, length of wrench 
 used 33 inches. The results of the tests are shown in Table LXIX. 
 
 TABLE LXIX. — FRICTIONAL RESISTANCE OF SPLICE BARS 
 
 (Watertown Arsenal) 
 
 Tightening Force Applied to Wrench 
 (Pounds). 
 
 Frictional Resistance of Joint. 
 
 Initial 
 
 (Pounds). 
 
 Continuous 
 Movement 
 (Pounds). 
 
 50 
 
 75 
 
 37,500 
 46,900 
 72,800 
 72,800 
 31,000 
 
 33,800 
 44,700 
 65,500 
 65,500 
 28,600 
 
 
 110 — 1 bolt 
 
 50 
 
 The maximum pull applied to five of the bolts in the third test, 85 pounds 
 on a 33-inch wrench, was the limit of strength of the bolts. This pull on the 
 wrench caused a permanent elongation of about .06 inch to .10 inch on each 
 of the five bolts. The sixth bolt resisted a pull of 110 pounds on the wrench 
 without material elongation. 
 
 After making observations on the frictional resistance in these tests, the 
 first test, with bolts tightened to 50 pounds' pull, was repeated. 
 
 The splice bars were now used on one piece of rail, using four bolts, the 
 nuts of which were tightened with a pull of 50 pounds on a 33-inch wrench. 
 The initial resistance was 50,900 pounds and movement continued under 
 31,200 pounds. 
 
 Tests with four bolts in one piece of rail, with 50 pounds' pull on the 
 wrench, were repeated with an initial resistance of 59,200 pounds. The 
 movement continued under 41,600 pounds. 
 
 * House Documents, Vol. 78, No. 291, 58th Congress, 3rd Session, 1904-05, Tests of Metals. 
 
STRESSES IN THE RAIL 261 
 
 The tension on the bolts was reduced during test, and after the last observa- 
 tions were made the nuts could be further tightened with a pull of 30 pounds. 
 Each nut could be turned up 90 degrees before again attaining a resistance of 
 50 pounds on the wrench. 
 
 One-half joint was again made up with four bolts and 50 pounds' pull on 
 the wrench. The initial resistance was 66,500 pounds. The slipping of the 
 angle bars occurred with a series of throbs, immediately followed in each 
 instance by a reduction in the load on the bars. The succeeding throbs took 
 place under gradually diminishing loads, following to 49,300 pounds at the 
 fourth throb. When removed from the testing machine, the nuts could be 
 turned on with an average pull of 35 pounds on the wrench. 
 
 * The experiments carried out by Messrs. Resal, Poutzen, and Menard on 
 the longitudinal slipping of rails connected by fishplates were of four descrip- 
 tions, viz.: 
 
 (1) On new rails with new fishplates and bolt holes; 
 
 (2) On the same rails lubricated with mineral oil; 
 
 (3) On old rails with worn bolt holes; and 
 
 (4) On the same with the addition of a thin layer of sand between the 
 
 surfaces of contact. 
 
 It was found that the old rails gave the best results and required a pressure 
 of 18 tons to effect any appreciable movement, whereas the new rails were 
 least satisfactory, particularly after oiling. As to the experiment with sand, 
 it was found that, owing to the reduction of the surfaces of contact caused by 
 the sand, the slipping was about the same as in the case of new rails. 
 
 Dr. P. H. Dudley found that a well-fitted splice bar for a 5-inch rail re- 
 quired over 4000 pounds per linear inch of one-half of the length of the bar to 
 overcome the friction in the rail ends, and for 90-pound and 100-pound 6-inch 
 rail, 4500 and 4800 pounds respectively 
 
 We are probably not warranted in taking the frictional resistance of the 
 joint at more than 40,000 pounds; nor can the friction between the rail and the 
 splice bar be well increased by the use of special joints, without at the same 
 time increasing to an undesirable extent the stresses in the rail, caused by 
 sudden changes in temperature. 
 
 It will be seen that this frictional resistance may cause an initial tensile 
 stress of about 4000 pounds per square inch in the 100-pound rail at times of 
 a sudden fall in temperature. 
 
 * Revue Generate des Chemins de Fer, Paris, 1908, Vol. 31, pp. 8-14. 
 
262 STEEL RAILS 
 
 The tension set up in rails of lighter section in falling temperatures, before 
 they render in the splice bars, is considered by Dr. P. H. Dudley to be 
 important and indirectly responsible for a large number of the cracked or broken 
 rails which occur during falling temperatures. Records and dates of broken 
 rails taken by Dr. Dudley for a number of years, when compared with the 
 dates of decided falling temperatures, were found to practically coincide, but 
 as soon as the temperature would rise, relieving the rails from tension or 
 putting them in compression, the breakages would cease, except in cases of a 
 development of a check which commenced in a falling temperature. 
 
 SHEAR 
 
 Fig. 182. — Shearing Stress in 100-pound A. S. C. E. Rail and Splice Bar. Total Shear, 24,000 Pounds. 
 
 If we consider the effect of the fractional resistance between the splice bar 
 and the rail, it is apparent that the bar shown in Fig. 182 will act as an 
 integral part of the rail until the longitudinal shear at the surfaces of contact 
 of the rail and the bar exceeds the resistance caused by friction on these 
 surfaces. This resistance for a 20-inch splice bar may be taken as 4000 
 pounds per linear inch for the entire joint, or 1400 pounds per square inch for 
 
STRESSES IN THE RAIL 
 
 263 
 
 the upper surface of contact, and 500 pounds per square inch for the lower 
 surface of contact. 
 
 It is seen from the figure that the surface friction is sufficient to carry a 
 total shear at the section of 24,000 pounds, and by referring to the rail diagrams 
 given in Figs. 171 and 172 it would appear that the maximum bending moment 
 in the rail would be transmitted to the splice bars without slipping. However, 
 between the two rails the splice bars must carry the entire moment, and unless 
 the section of the bar is increased at the middle of the joint there results an 
 excessive deflection at this point. 
 
 Fig. 183. — 100 per 
 cent Joint. 
 
 Fig. 184. — Joint showing 
 Uneconomical Distribu- 
 tion of Metal. 
 
 Fig. 185. — Joint showing 
 Economical Distribution 
 of Metal. 
 
 To overcome this source of weakness in the joint, the form shown in Fig. 183 
 has been found to embody most of the essential elements demanded by the 
 extra reinforcement needed at the center of the joint. This section is only 
 used at the middle of the bar and the section shown in Fig. 185 is used for the 
 rest of the bar. 
 
 It will be seen that the added metal is distributed in such a way as 
 to still keep the vertical axis within the vertical surface that is gripped by 
 the bolts. The sectional area and moment of inertia of the reinforcement 
 shown in Fig. 183 can readily be adjusted to match the stiffness of the rail 
 that is to be spliced, whereas, with the space limitations of Fig. 185, it is 
 not possible to get a higher relative percentage of strength than, say, 40, 
 as compared with the rail. 
 
 The splice bars, shown in Figs. 184 and 185, show the greater stiffness that 
 can be obtained by means of a proper distribution of the metal. Taking the 
 
264 STEEL RAILS 
 
 stiffness of the rail as 100 per cent, the relative stiffness of the bars, shown in 
 Figs. 184 and 185, is 29.1 per cent and 37.3 per cent respectively.* 
 
 As far back as 1876 quite full experiments were made with a modified 
 type of reinforcement shown in Fig. E, Plate XXIV, on the Swedish Government 
 Railroads, and in the German handbooks of somewhat later dates quite a variety 
 of sections are found of this same general shape. The reinforcing vertical 
 flange occupied a plane at some distance from the axis of the rail, which 
 causes the vertical axis of the splice bar to assume a position outside the ver- 
 tical surface that is gripped by the bolts, and in consequence the resisting 
 stresses in the flange itself must cause an outwardly rotating action, tending 
 to strip the threads of the bolts. 
 
 Plate XXIV shows types of joints used in this country, f Fig. A on this 
 plate illustrates the common type of angle bar. The variations from this 
 section, as applied to 80- and 85-pound rail, are in many directions. A com- 
 paratively frequent one is the thickening of the vertical web to f inch. Another 
 tendency is to put more metal into the upper part of the web near the under 
 side of the rail head. An extreme development of this latter practice is shown 
 in the Pennsylvania's angle bar for use with its new section of rail. Fig. B 
 shows this section. The horizontal extension of the lower flange of the bar 
 is another direction in which the angle-bar section is frequently modified. 
 
 There are six patented joints which are now in service in sufficient numbers 
 to merit consideration. They may be divided into two classes: those with deep 
 girder flanges, namely, the Hundred per cent, the Duquesne, and the Bonzano, 
 Figs. C, D, and E; and those which are base-supporting, as the Continuous, the 
 Weber, and the Wolhaupter, Figs. F, G, and H. 
 
 The Rail Committee of the American Railway Engineering Association 
 have recently made a series of interesting tests on rail joints at the Watertown 
 Arsenal, f 
 
 (1) Three joints of each kind were furnished, of which two were used for 
 testing and the third joint was reserved for future use if needed. 
 
 (2) All joints were full-bolted. Several of the joints first tested had various 
 sized openings between the rail ends. After the test of the first three joints, 
 all other joints were changed so that the opening between the ends of the rails 
 was as close to three-eighths of an inch as possible. The span between supports 
 in the testing machine was 30 inches. 
 
 * Railroad Age Gazette, April 9, 1909, p. 804. 
 
 t Railroad Age Gazette, March 19, 1909. 
 
 % Bulletin No. 123, May, 1910, Am. Ry. Eng. & M. of W. Assn. 
 
STRESSES IN THE RAIL 
 
 265 
 
 (3) One joint was tested with the load first applied to the base, in incre- 
 ments of 2000 pounds, until the limit of 32,000 pounds was reached, and then 
 the joint was reversed and the load applied on the head until the joint failed 
 or the limit of the machine was reached. 
 
 (4) The second joint was tested by first applying the load on the head and 
 then reversing it, applying the load on the base, until the limit was reached. 
 
 (5) With the exception of the joints furnished by the Cambria Steel Com- 
 pany and Mr. A. Morrison, the joints were selected from material which had 
 
 DEFLECTION. J N INCHFS 
 - Diagram of Watertown Arsenal Tests o 
 (Am. Ry. Eng. Assn.) 
 
 100-pound Joints. 
 
 been furnished by the manufacturers to the railroad companies in the regular 
 routine of business, and therefore fairly represent the material ordinarily fur- 
 nished by the manufacturers. 
 
 Figs. A to F (Plate XXV) show some of the joints tested; the results of 
 the tests on these joints are presented in Fig. 186. The material in the different 
 splice bars varies so widely that it is difficult to judge of the value of the differ- 
 ent designs. The excellent results obtained with the Dudley joint (Fig. D) is 
 probably due to the high strength of the metal as compared to the other joints 
 tested. 
 
266 STEEL RAILS 
 
 The rolling mills are reluctant to make splices of higher carbon, Bessemer 
 process, than .10 to .20 per cent. Some railroads have specified as high as 
 .63 per cent, and with good results both as to manufacture and experience. 
 The mill, however, suffers by such a high standard. One mill claims to have 
 broken one-third of the total quantity of bars in the straightening process, 
 and it also broke many of the punches. 
 
 Splices made from steel of .50 carbon appear to give much better results, 
 as might be expected, than the softer steel bars. It is necessary to hot-punch 
 the higher carbon steel, and when this is done there is no difficulty in properly 
 manufacturing them. The Cambria Steel Company are rolling bars of this 
 grade of steel which are hot finished and oil tempered. 
 
 * The economic advantage of high-carbon steel, hot finished, is that, with 
 the expenditure of about 10 per cent more than the cost of soft steel, a joint 
 is given a carrying capacity that can be equaled only by the addition of double 
 the quantity of metal of soft steel and at the additional cost of 100 per cent. 
 This latter joint will cost 100 per cent more for freight, while there is no addi- 
 tional cost for freight in the former. 
 
 The oil treatment of steel is a natural sequence of the use of high carbon, 
 and its advantages are about equal to those of high carbon over soft steel. 
 This, however, varies with the section of the bar and hardness of the steel. In 
 economy, oil-treated steel is as much in advance of high-carbon steel as is the 
 latter over soft steel. 
 
 Intimately connected with the rail- joint problem is the question of the 
 length of the rail. In a recent bulletin (August, 1909) of the International 
 Railway Congress, the practice in English-speaking countries is very fully dis- 
 cussed and abstracts from this report are given below. 
 
 In Great Britain and Ireland the railways have been gradually increasing 
 the length of rails, with a view to reducing the number of joints. Some rail- 
 ways still use rails only 30 feet long, and a few use 60-foot rails, but a large 
 number have 45-foot rails, and it appears that this may be taken as a standard 
 for the near future. The principal reasons given for limiting the length, as 
 given by the engineers of different railways, may be summarized as follows: 
 (1) Difficulty of straightening rails at the mills; (2) cost of manufacture; (3) 
 difficulties of transportation; (4) expansion and contraction; (5) unloading and 
 handling on the track. 
 
 So far the use of long rails does not appear to have called for the 
 adoption of any special arrangements other than proper proportioning 
 
 * Railway Age Gazette, March 16, 1910, daily edition. 
 
STRESSES IN THE RAIL 267 
 
 of the bolt holes and the play at the joints, and the strengthening of 
 the maintenance gangs (by consolidating neighboring gangs or otherwise) 
 while handling long rails. The temperature varies from 0° F. in winter to a 
 maximum of 130° F. in the sun in summer. To allow free play for expansion 
 during extreme heat, it is the practice of most engineers to ease the joints by 
 slackening the nuts. 
 
 In the United States the standard length on a number of railways is 33 feet, 
 and the reasons given for limiting the length are, in general, similar to those 
 noted above. Experiments have been made with rails of greater length, but 
 on the whole these have not been satisfactory, although the opinions expressed 
 by some of the railways give 40 feet, 45 feet, 50 feet, and 60 feet as admissible 
 lengths. The range of temperature is 100° F. in some parts of the country, 
 while in others it is 180° F. In some cases the nuts are slackened in the early 
 part of the summer. 
 
 In the roads of other countries investigated which include 17 railways 
 in South America, India, South Africa, Australia, and Canada, the limiting 
 length of rails varies from 30 to 40 feet. The range of temperature is from 
 100° to 155° in India, and 160° in South Africa and Australia. 
 
 Inquiry was made as to whether any railways contemplated welding the 
 rails at the joints. All the replies were in the negative, and the general opinion 
 was that continuous rails would be unsafe, on account of the temperature 
 changes. It is well known, of course, that welding the joint is common practice 
 in street railway work, but in such cases the rails are protected by the paving, 
 so that only a small portion is exposed. 
 
 The following very interesting report from the Pennsylvania Lines is quoted 
 in the bulletin : 
 
 " In 1897 a continuous rail, 1050 feet long, made up of 35 80-pound 30-foot 
 rails joined by angle bars with drilled holes and machine-turned bolts (no pro- 
 vision being made for expansion and contraction), was laid in the eastbound 
 main track, near New Brighton, Pa. The ends were held by bent rails bedded 
 in concrete, so placed as to bear against the ties. Special long and wide angle 
 bars were used at the ends, fastened to the anchor ties with lag screws. The 
 track was a tangent with stone ballast. 
 
 " The rail crept and kinked out of line badly. An examination made in 
 August, 1900, after three years' service, showed that the entire rail crept in 
 the direction of traffic (eastward). At the west anchorage, the vertical holding 
 rails had cut into the cross-ties forming the anchorage framework, while at the 
 east anchorage there was a space of If inches between the vertical rails and the 
 
268 STEEL RAILS 
 
 framework. All of the spikes were bent eastward, and both slots and spikes 
 were badly cut. The bolts were all slightly sprung. The alignment at the 
 joints was very bad." * 
 
 The conclusions presented in the bulletin are given below: In Great 
 Britain and Ireland, the lengthening of rails and the consequent reduction of 
 the number of joints has been steadily proceeding at an increasing rate during 
 the past 60 years. In 1840-50, the normal length of rolled iron rails was 
 from 15 to 18 feet. The length of these iron rails increased at the rate of about 
 3 feet in each decade until 1870-80, when steel rails from 24 to 30 feet long 
 were brought into general use. Since then the decennial increase has been about 
 4J feet, and at the present time rails 36 feet and 45 feet long are in general use, 
 while two railways have adopted 60 feet in length. 
 
 In 1904, when the Engineering Standards Committee issued the British 
 Standard sections, it recommended the adoption of the following as the normal 
 lengths of rails: 30, 36, 45, or 60 feet. In other countries embraced by the 
 report, the length of rails has been steadily and uninterruptedly increasing, 
 but within narrower limits than in Great Britain and Ireland. 
 
 The conclusions to be drawn from the numerous replies and remarks by 
 engineers throughout the English-speaking countries are that there is a maxi- 
 mum length of rail somewhere between 33 feet and 60 feet, which should not 
 be exceeded ; and continuous or welded joints over a long length of railway are 
 impracticable and dangerous. 
 
 BIBLIOGRAPHY 
 
 Blum. — Report ... on the question of rail joints. (All countries except France, Belgium, 
 Italy, Spain, Portugal, Austria-Hungary, Rumania, Bulgaria, Servia, Turkey, Egypt, and countries 
 using the English language.) 48 p. 111. 1910. (In Bulletin of the International Railway Con- 
 gress, Vol. 24, Part 1, p. 1701.) 
 
 Bouchard, H. — Note sur le joint asymetrique. 4000 w. 111. 1909. (In Revue generate des 
 chemins de fer, Vol. 32, p. 9.) 
 
 Describes theory, construction, and results with favorable rail joint. 
 
 Chateau. — ■ Report ... on the question of rail joints (France, Belgium, Italy, Spain, and 
 
 * The author several years ago had experience with a continuous rail designed by the late 
 Mr. Torrey. 
 
 The rail was about the length and of the same weight as in the Pennsylvania test. It was 
 laid on a branch line of the Michigan Central Railroad on a tangent and ballasted with a good 
 bed of gravel. Provision was made at intervals of about 500 feet throughout the test track for 
 the expansion of the continuous rail by means of special expansion joints, the rail being anchored 
 midway between these joints. 
 
 " When first installed the riding qualities of the track were exceptionally good, which may, 
 however, have been accounted for by the unusual care that had been taken in constructing it. 
 After several years the test track appeared to ride about as well and to require the same amount 
 of attention as the other track on the division. 
 
STRESSES IN THE RAIL 269 
 
 Portugal). 38 p. 111. 1910. (In Bulletin of the International Railway Congress, Vol. 24, Part 1, 
 p. 1427.) 
 
 Edelstein, Leon. — Prevention of play between rail and fishplate. 2000 w. 111. 1908. 
 (In Bulletin of the International Railway Congress, Vol. .22, Part 1, p. 436.) 
 
 Shows faults that develop with wear, and gives suggestions for prolonging life of both rails 
 and fishplates. 
 
 Godfernatjx, R. — Note about rail-joints. 24 p. 111. 1911. (In Bulletin of the Inter- 
 national Railway Congress, Vol. 25, p. 1480.) 
 
 Reviews development of rail joints and different forms used. 
 
 Haarmann, A. — Der schienenstoss. 5000 w. 111. 1911. (In Stahl und eisen, Vol. 31, Part 
 1, p. 49.) 
 
 Discusses types of rail joints and probable future practice in track construction. 
 
 Improvements and experiments in rail joints. 2500 w. 1910. (In Engineering News, Vol. 64, 
 p. 281.) 
 
 Abstracts information from reports to International Railway Congress on practice in the United 
 States, Great Britain, France, Belgium, and Austria-Hungary. 
 
 International Railway Congress. — [General discussion on rail joints.] 38 p. 1911. (In 
 Bulletin of the International Railway Congress, Vol. 25, p. 405.) 
 
 Kramer, Friedrich. — Report ... on the question of rail-joints (Austria, Hungary, Rumania, 
 Bulgaria, Servia, Turkey, and Egypt). 82 p. 111. 1910. (In Bulletin of the International Rail- 
 way Congress, Vol. 24, Part 1, p. 1967.) 
 
 Jones, Cyril Walter Lloyd. — Design of fishplate rail joints. 7000 w. 111. 1910. (In 
 Minutes of Proceedings of the Institution of Civil Engineers, Vol. 182, p. 282.) 
 
 Pellarin. — Etude des joints des rails. 27 p. 111. 1908. (In Annales des ponts et chaussees, 
 Memoires, series 8, Vol. 33, p. 9S.) 
 
 Report of commission for investigation of rail joints in Belgium, Italy, Switzerland, and Holland. 
 
 Ross, Alexander. — Report ... on the question of rail joints. (Countries using the English 
 language.) 44 p. 111. 1909. (In Bulletin of the International Railway Congress, Vol. 23, Part 2, 
 
CHAPTER V 
 
 strength of the rail 
 
 25. Influence of Stress and Strain on the Strength of the Rail 
 
 In determining the safety of any structure, not only the amount of stress 
 induced in the different members by the load must be found, but its character 
 and the effect it may produce on the material of which the structure is composed 
 must be considered. 
 
 A series of rapidly repeated stresses will, under certain conditions, affect 
 the breaking strength of the metal. The cause of this loss of strength under 
 the influence of repeated stress is a much-mooted question among engineers, 
 and it is of interest to examine the problem in detail.* 
 
 Fig. 187. — Pure Swedish Iron. (Mellor.) 
 
 Fig. 188. — Pure Copper. (Arnold.) 
 
 The junction of the crystalline grains of pure iron, shown in Fig. 187, and 
 of pure copper, shown in Fig. 188, are typical of pure metals; but when impuri- 
 ties are present the crystals of the pure metal, in the act of crystallizing, reject 
 the impurities which collect at the crystal boundaries. The particles of pure 
 metal slowly migrate and coalesce together, so as to form little islands sur- 
 rounded by the impurity; accordingly, in the solidified mass, we find the crystals 
 of pure metal enveloped by a film of the metal associated with the foreign sub- 
 
 * The author is indebted for the discussion on pp. 270-276 to Mr. J. W. Mellor's work, The Crys- 
 tallization of Iron and Steel, Longmans, Green & Co., London, 1905. 
 
 270 
 
STRENGTH OF THE RAIL 
 
 271 
 
 stance. This investing membrane separates the crystals of pure metal one 
 from the other. Obviously, the mechanical and physical properties of the alloy 
 will depend upon the character of the film. 
 
 The mass of pure metal, for example, may be quite ductile like gold, while 
 the mass of metal with the impurity may be quite brittle, as Arnold * found 
 to be the case with an alloy of gold with .2 per cent of bismuth; and copper 
 containing .5 per cent bismuth. A representation of the latter alloy is shown 
 in Fig. 189. When such a metal is fractured, the line of fracture follows the 
 junction of the grains. 
 
 Stead calls this ailment intergranular or intercrystalline weakness (inter, 
 between). We have had examples. Arnold's work on the influence of bismuth 
 
 Fig. 189. — Copper-bismuth Alloy. 
 (Arnold.) 
 
 - Iron with 1.8 per cent Carbon. 
 (F. Popplewell.) 
 
 on copper and on gold. One per cent of sulphur arranged as a mesh of iron 
 sulphide will entirely destroy the ductility of the iron, reducing the ultimate 
 stress from 20 to 2 tons per square inch. 
 
 The network of cementite which envelops the crystal grains of steel con- 
 taining over 1 per cent of carbon are the principal lines of weakness. The 
 metal when fractured generally breaks through the center of this brittle envelope. 
 The coefficient of contraction of the cementite cell walls is greater than of the 
 cell contents. Pearlite cells, for example, bound together by thick cementite 
 walls (Fig. 190), are liable to rupture, because the coefficient of contraction of 
 the cementite cell walls is greater than the cell contents. The mass is, in con- 
 sequence, very feebly held together, and a sudden blow will easily fracture the 
 metal.t 
 
 * J. O. Arnold and J. Jefferson, Engineering, 61, 177, 1896. 
 t J. O. Arnold, Metallographist, 5, 267, 1902. 
 
272 STEEL RAILS 
 
 Intergranular weakness resembles the weakness of a brick building with 
 faulty mortar. 
 
 There is another type of intergranular weakness which is due to imper- 
 fect union of the crystal grains. This is particularly marked in phosphorous 
 steels. The crystal grains, on cooling, contract unequally and tend either 
 to draw the grains away from each other or to leave the mass in a state of 
 unnatural tension. The fracture then follows the granular junctions. Thick 
 plates and bars are frequently brittle because comparatively little work has 
 been done on them. The crystals are not interlocked one with another, as in 
 steel which has been well worked. 
 
 Intergranular weakness may, therefore, be of two kinds: 
 
 (1) Brittle envelope surrounding the crystal grains; 
 
 (2) Imperfect union of the crystal grains. 
 
 Stead has pointed out another type of weakness in sheet steels which has 
 to do with the crystals themselves, without reference to the union of one crystal 
 with another. It is a kind of intracrystalline weakness (intra, within). 
 
 It is characteristic of some crystals to break more readily in some directions 
 more than in others. This property of crystals is called cleavage. The direc- 
 tions in which the crystal splits are called cleavage planes. If a bar of iron 
 could be cut from a single crystal, that bar would have three lines of weakness 
 in the direction of the three cleavage planes; while if the bar were built up of 
 a number of crystals whose cleavage planes were all in the same direction, that 
 bar would be more readily broken in the direction of its cleavage planes, neglect- 
 ing for the moment intergranular weakness. On the other hand, if the cleavage 
 planes of the adjacent crystals are inclined at considerable angles to one another 
 the bar would be less liable to break than one in which the crystals were arranged 
 symmetrically. Figs. 191 and 192 will make this clear. The dotted lines ab, 
 Fig. 191, represent the cleavage planes across a sheet of iron when the crystals 
 are arranged symmetrically, while in Fig. 192 the crystals are arranged in an 
 irregular manner. The cleavage planes of Fig. 191 run along parallel lines, 
 and the sheet would, therefore, be more liable to rupture than the sheet shown 
 in Fig. 192, where the lines of weakness are not in the same direction, and this 
 in spite of the fact that Fig. 191 has a finer grain. 
 
 Other things being equal, a fine-grained structure is stronger and tougher 
 than a coarse-grained piece. Figs. 191 and 192 show that this order of things 
 may be reversed. Fortunately, the crystals of one steel do not generally grow 
 symmetrically. 
 
STRENGTH OF THE RAIL 
 
 273 
 
 Fig. 191. — Cleavage Planes with Crystals ar- Fig. 192. — Cleavage Planes with Crystals arranged 
 [ Symmetrically. (J. W. Mellor.) in an Irregular Manner. (J. W. Mellor.) 
 
 Fig. 193. — Iron strained beyond the Elastic Limit. 
 (J. A. Ewing and W. Rosenhain.) 
 
 Fig. 194. — Lead strained beyond the Elastic 
 Limit. (J. A. Ewing and W. Rosenhain.) 
 
 Examining now what takes place in the metal under repeated stress, Ewing 
 and Rosenhain * found if a metal is strained past its " yielding point " — elastic 
 limit — the faces of the crystal grains (Fig. 193) show fine black lines, which 
 increase in number as the strain increases. Lines appear on certain crystals 
 nearly transverse to the pull, as the strain increases lines appear upon other 
 
 * J. A. Ewing and W. Rosenhain, Phil. Trans., 193, 353, 1899; 195, 279, 1900. J. A. Ewing 
 and J. C. W. Humfrey, Ibid., 200, 241, 1902. W. Rosenhain, Journal Iron and Steel Inst., 67, i, 335, 
 1904. F. Osmond and C. Fremont and G. Cartaud, Revue de Metallurgie, 1, i, 1904. 
 
274 STEEL RAILS 
 
 grains. Intersecting lines then make their appearance on some of the grains. 
 Such a strained surface is shown in Fig. 194. 
 
 The lines are apparently not actual cracks in the surface, but rather slips 
 along the cleavage planes of the crystal. They are called slip bands, or 
 slip lines. 
 
 Let AB (Fig. 195) represent a cross section through a polished surface 
 of metal. Let C be the junction between two contiguous grains, A and B. 
 When the metal is pulled in the direction of the arrows, a number of slips are 
 developed along the cleavage planes, a,b,c,d. . , and the surface now presents 
 
 Fig. 195. — Cross Section of Unstrained Metal. Fig. 196. — Cross Section of Metal after being 
 
 (J. W. Mellor.) Stressed. (J. W. Mellor.) 
 
 the appearance shown in Fig. 196. With still greater strains the slip bands 
 develop into actual cracks, and rupture takes place. Hence it follows that 
 under progressively augmented strain, rupture takes place, not at the crystal 
 boundaries, but through the crystals themselves. 
 
 Ewing and Humfrey have subjected Swedish, iron, with a breaking stress 
 of 23.6 tons per square inch, to a series of compression and tension stresses, 
 9 tons in magnitude, repeated 400 times per minute. On examination it was 
 found that fine slip bands appeared in a few crystals after a few, say, 5000 
 reversals of stress; with a greater number, say, 40,000, the slip bands increase 
 in number, and those which first appeared broaden and develop into small 
 cracks, as shown in Fig. 197. 
 
 If the specimen be repolished, so as to clear off the slip bands, the cracks 
 alone become visible, as at A (Fig. 198). The crack, or flaw, gradually creeps 
 across the specimen when the number of alternations is still further increased, 
 as shown in Fig. 199. Finally the specimen breaks. 
 
 Ewing and Humfrey state: " Whatever the selective action of the stress 
 is due to, the experiments demonstrate that in repeated reversals of stress 
 certain crystals are attacked, and yield by slipping, as in other cases of non- 
 elastic strain. Then, as the reversals proceed, the surfaces upon which the 
 slipping has occurred continue to be surfaces of weakness. The parts of the 
 crystal lying on the two sides of each such surface continue to slide back and 
 forth over one another. 
 
STRENGTH OF THE RAIL 
 
 275 
 
 Fig. 197. — Slip Bands. (J. W. Ewing and J. C. W. Humfrey.) 
 
 I \* 
 
 
 
 r. 
 
 
 h. 
 
 -A 
 
 Fig. 198. — Polished Surface with Small Cracks. (J. W. Ewing and J. C. W. Humfrey.) 
 
276 
 
 STEEL RAILS 
 
 " The effect of this repeated sliding or grinding is seen at the polished 
 surface of the specimen by the production of a burr or rough and jagged irregu- 
 lar edge, broadening the slip band, and suggesting the accumulation of debris. 
 Within the crystal this repeated grinding tends to destroy the cohesion of the 
 metal across the surface of the slip, and in certain cases this develops into a crack. 
 
 " Once the crack is formed, it quickly grows in a well-known manner, 
 by tearing at the edges, in consequence of the concentration of stress which 
 
 
 Fig. 199. — Polished Surface with Large Cracks. (J. W. Ewing and J. C. W. Humfrey.) 
 
 results from lack of continuity. The experiments throw light on the known 
 fact that fracture by repeated reversals or alternations of stress resembles 
 fracture resulting from ' creeping flaw ' in its abruptness, and in the absence 
 of local drawing out, or other deformation of shape." * 
 
 The rupture of steel is not caused by the gradual growth of the crystal- 
 line structure of the metal under the influence of shocks and vibrations. The 
 breaking down is due to fatigue. When fatigued, the metal breaks more readily. 
 Again, when subjected to sudden shock, the metal has no time to " flow." The 
 
 * P. Kreuzpointer, Journ. Franklin Inst., 
 P. Breuil, Suppl. Journ. I. & S. Inst., 1904. 
 
 153, 233, 1902. J. A. Ewing, Nature, 70, 187, 1904. 
 
STRENGTH OF THE RAIL 277 
 
 slipping of the crystal planes, or the plasticity of the metal, has no time to come 
 into play. The metal, in consequence, appears to be abnormally brittle. 
 
 * The experiments made by Wohler, from 1859 to 1870, were the first that 
 indicated the laws which govern the rupture of metals under repeated applica- 
 tion of stress. For instance, he found that the rupture of a bar of wrought 
 iron by tension was caused in the following ways : 
 
 By 800 applications of 52,800 pounds per square inch. 
 
 By 107,000 applications of 48,400 pounds per square inch. 
 
 By 450,000 applications of 39,000 pounds per square inch. 
 
 By 10,140,000 applications of 35,000 pounds per square inch. 
 Here it is seen that the breaking unit stress decreases as the number of 
 applications increases. It was further observed that a bar could be strained 
 from up to a stress near its elastic limit an enormous number of times without 
 rupture, and it was also found that a bar could be ruptured by a stress less than 
 its elastic limit under a large number of repetitions of stress which alternated 
 from tension to compression and back again. 
 
 f The apparatus used by Wohler and his successor, Spangenberg,! was 
 of four kinds: 
 
 (1) To produce rupture by repeated load; 
 
 (2) For repeated bending, in one direction, of prismatic rods; 
 
 (3) For experiments on loaded rods under constant bending stress; 
 
 (4) For torsion by repeated stress. 
 
 The amount of the imposed stress was determined by breaking several 
 rods of like material, ascertaining the breaking load, and taking some fraction 
 of this for the intermittent load. 
 
 From the results of these experiments of Wohler, extending over eleven 
 years, the following law was deduced: 
 
 " Wohler's Law: Rupture of material may be caused by repeated vibra- 
 tions, none of which attain the absolute breaking limit. The differences of 
 the limiting strains are sufficient for the rupture of the material." 
 
 The work of Wohler and Spangenberg has proved what was long before 
 supposed to be the fact: that the permanence and safety of any iron or steel 
 structure depends not simply on the greatest magnitude of the load to be sus- 
 tained, but on the frequency of its application and the range of variation of 
 its amount. 
 
 * A. Wohler, Engineering, 1871 ; Zeitschrift fur Bouwesen, 1870, p. 83, Berlin, 
 t Iron and Steel, Materials of Engineering, Part 2, Thurston, 1909, p. 618. 
 t Zeitschrift fur Bouwesen, 1874, p. 485, Berlin. 
 
278 STEEL RAILS 
 
 Prof. L. Spangenberg resumed the line of experiments at the point of its 
 discontinuance by Wohler, and his results tend to confirm the law of the latter. 
 Spangenberg directed his attention to other metals than iron and steel, and 
 also endeavored, by inspection of the surfaces of fracture, and by his hypoth- 
 esis as to the molecular constitution of metals, to explain the phenomena of 
 fracture. Among the several observations noted in his " Fatigue of Metals " 
 is the important fact that when subjected to often-repeated transverse stress 
 fracture of iron took place only on the tension side of the bar and extended 
 only to the neutral axis. From this he inferred that the working strength of 
 wrought iron is less than its elastic resistance. 
 
 Fowler states, in this connection, that a steel rail tested for transverse 
 strength in a machine will, as a rule, bend many inches, and fail by distortion of 
 the head under the compressive stress. In actual work hundreds of such rails 
 break, but it is the tensile and not the compressive stress which causes the 
 failure, and there is no distortion of the head, as in the testing machine. 
 
 Reynolds and Smith * extended Wohler's conclusions to steel bars tested 
 under direct tension and compression and at a rapid rate of alternation. The 
 work of Stanton and Bairstow,f published in 1906, while less concordant than 
 that of Reynolds and Smith, confirms their general conclusions; it extends the 
 conclusions of Ewing and Humfrey to notched specimens tested for endurance 
 under direct tension and compression, and it clearly points out the advisability 
 of testing a material for endurance in approximately the form in which it is 
 to be used in practice. 
 
 French engineers, commenting upon the work of Wohler, Spangenberg, 
 Weyrauch,^ and Launhardt, consider that the result is simply to base upon the 
 ultimate strength a deduced limit of working stress which corresponds closely 
 to the elastic limit, and generally urge the use of a reasonable factor of safety 
 related to the limit of elasticity.! 
 
 Figs. 200, 201, and 202 present three diagrams on the behavior of steels 
 under repeated alternate stresses, illustrating some of the tests which have 
 been made at the Watertown Arsenal laboratory.) | 
 
 On Fig. 200 the heavy vertical lines represent the number of loads which 
 were applied to a number of steel bars of .55 per cent carbon, and which 
 
 * Phil. Trans. Royal Society of London, A- Vol. 199, p. 265, 1902. 
 
 t Proceedings Inst, of Civil Engrs., Vol. 166, p. 78, 1906. 
 
 t Various Methods of Determining Dimensions. Dr. J. Weyrauch; translated by G. R. Bodmer. 
 Proc. Inst. C. E., 1882-83, Vol. LXXI. 
 
 § Resume de la Societe des Ingenieurs Civils, 1882. 
 
 II Notes on the Endurance of Steels under Repeated Alternate Stress. Howard, Proceedings 
 Am. Society for Testing Materials, 1907, Vol. VII. 
 
STRENGTH OF THE RAIL 
 
 279 
 
 caused rupture of the metal. Beginning with the highest fiber stress, 60,000 
 pounds per square inch, the progressive gain in endurance of the steel as the 
 loads were successively reduced will be noted, as indicated by the lengths of 
 the different lines. The lowest fiber stress experimented with did not end in 
 
 REPEATED ALTERNATE STRESSES 
 O .55 CARBON STEEL. 
 
 TENSILE TEST 
 ELASTIC LIMIT 59,000 LBS PER S0UNCH 
 TENSILE STRENGTH 111,200 
 ELONGATION 12 PER CENT 
 
 CONTRACTION 33.5 
 
 30,000 35,000 40,000 45,000 50,000 60,000 
 
 FIBRE STRESSES. LBS. PER SQUARE INCH. 
 
 Fig. 200. — Behavior of 0.55 Carbon Steel under Repeated 
 Alternate Stresses. (Am. Soc. for Testing Materials. — . 
 Howard.) 
 
 REPEATED ALTERNATE STRESSES 
 .82 CARBON STEEL., 
 
 TENSILE TEST 
 ELASTIC LIMIT 64,0 00 LBS. PER SDJNCH 
 
 TENSILE STRENGTH 142,800 - •• - •• 
 ELONGATION 7 PER CENT 
 
 CONTRACTION 11.8." 
 
 ii 
 
 40,000 45,000 50,000 55.000 60.000 
 FIBRE STRESSES, LBS. PER SQUARE INCH 
 •K: NOT RUPTURED 
 
 Fig. 201. — Behavior of 0.82 Carbon Steel under Repeated 
 Alternate Stresses. (Am. Soc. for Testing Materials. — 
 Howard.) 
 
 rupture of the shaft; after 76 million repetitions, in round numbers, under a 
 load of 30,000 pounds per square inch, the fiber stress was increased to 60,000 
 pounds per square inch, which higher load caused rupture after about 8000 
 rotations. The enormous gain in endurance of the steel, under 30,000 pounds 
 fiber stress, over its behavior with the higher loads, would be represented by 
 a vertical line about 28 feet in height, according to the scale of this diagram. 
 The results of the tensile tests of this grade of steel are entered on the diagram, 
 
280 STEEL RAILS 
 
 from which it may be seen that the several fiber stresses were, with one exception, 
 below the tensile elastic limit of the metal. 
 
 On Fig. 201 similar lines represent the behavior of specimens containing 
 .82 per cent carbon. The behavior of this grade resembles that of the previous 
 diagram, and similar to other steels belonging to this series of experiments. 
 The endurance under corresponding loads is seen to be greater than displayed 
 on the preceding diagram. After making 202 million rotations the test of the 
 shaft loaded with 40,000 pounds was temporarily discontinued, the steel being 
 unruptured. A line drawn to scale to represent the endurance under this load 
 would be about 73 feet in height. 
 
 On Fig. 202 appear curves representing the relative endurance of each 
 of the six grades of steel used in this series of experiments. Their endurance 
 under the higher fiber stresses only are shown, loads which caused compara- 
 tively early rupture of the steel in most of the tests. 
 
 It may be remarked that the fiber stresses experimented with were gener- 
 ally below the tensile elastic limits of the steels. The greater endurance of the 
 steels of .73 and .82 per cent carbon in comparison with either the higher or the 
 lower carbons is an interesting feature of the tests. 
 
 Elastic properties only are displayed by steels, prior to rupture when rup- 
 ture is caused by a large number of alternate stresses of tension and compression ; 
 no appreciable display of ductility, as shown by elongation and contraction of 
 area, need precede rupture, in any grade of steel, following the application of 
 stresses of this kind. If the fiber stresses somewhat exceed the tensile elastic 
 limit, a limited display of elongation, other than elastic, may occur, but rupture 
 caused by loads which are in the vicinity of or below the tensile elastic limit 
 is not attended with an appreciable display of ductility. 
 
 While tests by repeated alternate stresses are characterized by the absence 
 of ductility, as witnessed in tests by tension, there may be elastic movements 
 of the metal aggregating considerable distances. The aggregate extension of 
 the most strained fiber of the .82 per cent carbon steel which has successfully 
 endured 202 million repetitions amounts to nearly 5 miles per linear inch of 
 specimen, a distance quite incomparable to the permanent extension of the 
 metal in the tensile test. 
 
 * Little attention seems to have been given to the possibility of finding 
 a relation connecting the stress used in endurance tests with the number of 
 repetitions required for rupture. Spangenberg, Reynolds, and Smith, and 
 
 * The Exponential Law of Endurance Tests, O. H. Basquin, Proceedings of the American Society 
 for Testing Materials, 1910, Vol. X, p. 625. 
 
STRENGTH OF THE RAIL 
 
 281 
 
 Stanton and Bairstow, have shown stress-repetition curves drawn to ordinary 
 Cartesian coordinates. 
 
 Logarithmic coordinates present a distinct advantage' in the study of 
 simple exponential curves, because these curves become straight lines for these 
 
 REPEATED ALTERNATE STRESSES. 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 ■ 
 
 40.000 FS 
 
 i 
 
 
 1 m l 
 1 ^' 
 
 V I 1 
 
 en 
 hi 
 o 
 
 IS 
 
 o 
 
 r 
 
 
 
 
 ol 
 
 <1 
 
 
 
 1 coj 
 
 1 ^f 
 
 
 
 / 
 
 
 
 
 
 
 /. 
 
 60,000 LBS 
 
 
 ^Sffr 
 
 a£^ 
 
 GRADE .17 C. 
 E.L. 51000 
 
 .3.4C. .55 C. .73C. 
 
 54,000 59,000 B4.000 
 
 .82C. I09C. 
 
 64.000 77.000 
 
 Fig. 202. — Comparison of the Behavior of Different Grades of Steel under Repeated Alternate Stresses 
 (Am. Soc. for Testing Materials. — Howard.) 
 
 coordinates and their equations may be written at once. Fig. 203 shows 
 stress-repetition curves for nineteen sets of endurance tests, made by five dif- 
 ferent observers. The names of the experimenters, the kind of test, and the 
 material are given in Table LXX. 
 
282 
 
 STEEL RAILS 
 
 TABLE LXX. — EXPERIMENTS ON REPEATED STRESS 
 
 (Basquin) 
 
 Letter. 
 
 —■ ■ 
 
 Kind of 
 Test. 
 
 Material. 
 
 Coefficient 
 
 (Thousands) 
 
 C. 
 
 Exponent 
 
 A 
 B 
 
 Wohler 
 
 b 
 b 
 b 
 b 
 b 
 b 
 
 c 
 
 b 
 
 b 
 d 
 b 
 
 b 
 b 
 
 b 
 
 Wrought-iron axles, Phoenix Co . . . . 
 Wrought-iron axles, Phoenix Co. . . . 
 
 217 
 109 
 103 
 94 
 130 
 36 
 29 
 1000 
 920 
 320 
 310 
 90 
 97 
 115 
 250 
 102 
 66 
 110 
 150 
 135 
 
 -0.12 
 10' 
 
 C* 
 
 
 -0 09 
 
 D* 
 
 
 Cast steel, Borsig 
 
 Homogeneous iron, P. C. & Co 
 
 Bar copper, Heckmann 
 
 Cast iron, locomotive cylinder 
 
 Krupp's spring steel, hardened 
 
 Krupp's spring steel, unhardened. . . 
 Krupp's axle steel 
 
 11 
 
 E 
 
 << 
 
 -0 12 
 
 F 
 
 n 
 
 08 
 
 G 
 
 << 
 
 —0 09 
 
 H 
 
 << 
 
 
 J 
 
 
 -0 21 
 
 K] 
 
 << 
 
 
 L\ 
 
 
 18 
 
 M 
 
 " 
 
 
 07 
 
 N 
 
 
 
 
 
 P 
 
 Benjamin Baker 
 
 Reynolds & Smith . . 
 
 Bars from Forth Bridge 
 
 -0.10 
 
 15 
 
 QX 
 
 
 13 
 
 
 Annealed mild steel 
 
 -0.11 
 
 T 
 
 
 Steel A, grooved 
 
 Steel B, grooved 
 
 -0.15 
 
 u 
 
 " 
 
 
 
 
 * C has round shoulders near grip ; D has square shoulders, 
 t K has round shoulders near grip; L has square shoulders, 
 t Ordinates are " Half Range of Stress," instead of maximun 
 § Carbon, 0.32 per cent ; yield point, 39,000 pounds per squan 
 
 ich. Most points show the n: 
 
 Kinds of Tests : 
 
 (a) Bending in one direction only (+ and 0). 
 
 (b) Rotating under bending load (+ and — ). 
 
 (c) Tension only (+ and 0). 
 
 (d) Bending back and forth (+ and — ). 
 
 (e) Tension and smaller compression (+ and — ). 
 
 The curve A is represented by the equation 
 S = 217,000 R-° 12 , 
 which has the form 
 
 S = CR n , 
 
 in which S is the maximum stress used in each test and R is the number of repe- 
 titions of this stress required for rupture. The coefficient, 217,000, was found 
 by extending the line A to the left until it intersected the vertical line R = 1 
 (i.e., 10°), and the stress at this intersection was read off the logarithmic scale as 
 217,000 pounds per square inch. The coefficient is the stress given by the curve 
 for a single repetition. All the coefficients given in Table LXX in the column 
 marked C were found in the same way. The value of the exponent, —0.12, 
 was found by measuring the angle (130°) which this line makes with the hori- 
 zontal axis and then taking one-tenth of its natural tangent. The factor " one- 
 tenth " comes in because, in Fig. 203, the scale used along the vertical axis 
 
STRENGTH OF THE RAIL 
 
 283 
 
 in plotting the stresses is ten times the scale used along the horizontal axis in 
 plotting the repetitions. In the same way the exponent for each curve of 
 Fig. 203 has been found and is listed in the table under the column marked n. 
 
 In looking over the curves, Fig. 203, it is evident that in many cases the 
 straight line represents the results of endurance tests very accurately through- 
 out a considerable range of stress. One is also impressed with the approximate 
 
 Fig. 203. — Number of Repetitions before Rupture in Endurance Tests of Metals. 
 (Am. Soc. for Testing Materials. — Basquin.) 
 
 parallelism of many of these lines. Curves B, C, D, E, F, G, M, N, and S 
 represent tests made in much the same way, — by rotating a specimen under 
 bending load. The curve for hard steel, tested by Baker * in much the same 
 way, has a steeper slope; the same is also true of the grooved specimens tested 
 by Foppl.f Curves H, J, K, and L are approximately parallel and represent 
 tests in one direction only; i.e., the stresses are not reversed. They have about 
 
 * Some notes on the Working Stress on Iron and Steel, Trans. Am. Soc. of Mech. Engrs., 1887, 
 Vol. VIII, p. 157. 
 
 t Mitteilungen aus dem Mech. Engrs., Vol. 130, 1909. 
 
284 STEEL RAILS 
 
 double the slope of the other curves mentioned. Why curve A, on wrought 
 iron, does not fall into this class is not clear. 
 
 * Small changes in temperature occur when a bar of metal is stressed 
 within the elastic limit; it becoming cooler under tension and warmer under 
 compression. The measurements of these changes made by Turner, in 1902, 
 have shown that these changes in temperatures continue at a uniform rate up 
 to about three-fifths of the elastic limit, and that then a marked change occurs, 
 the bar under tension then beginning to grow warmer, while the temperature 
 of the bar under compression increases at a more rapid rate. 
 
 It thus appears for stresses higher than above three-fifths of the elastic 
 limit, at least, energy is converted into heat under repeated applications; prob- 
 ably this occurs also at lower stresses when repeated stresses range from ten- 
 sion into compression in a bar, or when a beam is subjected to alternating flexure. 
 
 In the case of the rail the bending stress alternates about in the proportion 
 of 4 to 1, and it is very probable that by taking a unit stress less than half the 
 elastic limit, we may safely ignore the effect of fatigue on the metal of the rail 
 produced by this stress. The disturbed metal at the running surface of the head 
 which is frequently conspicuous in old rails is evidence of the elastic limit of the 
 metal being exceeded rather than the effect of repeated stress below this limit. 
 Generally speaking the effect of repeated stress is not to produce distortion of the 
 metal, and prior to rupture elastic properties only are displayed. 
 
 26. Effect of Low Temperatures on the Strength of the Rail 
 
 Very complete investigations were made to determine the effect of 
 changes of temperature in modifying the physical properties of iron and steel by 
 Styffe and Sandberg. f The conclusion of Styffe is that the strength of iron 
 and steel is not diminished by cold. Arguing from these experiments, investi- 
 gators have assumed that the cause of the frequent breakage of rails in cold 
 weather, and of articles made of iron and steel, is unequal expansion and con- 
 traction and the rigidity of supports, where, as in the case with rails, frost may 
 very greatly affect them. 
 
 J Sandberg, while admitting the care and the accuracy which distinguished 
 this extensive series of experiments, still doubted whether the reasons just 
 
 * Mechanics of Materials, Merriman, p. 354, New York, 1905; and Trans. Am. Soc. of Civil 
 Engrs., Vol. XXVIII, 1902, p. 27, Thermo Electric Measurements of Stress, Turner. 
 
 t The Elasticity, Extensibility, and Tensile Strength of Iron and Steel, by Knut Styffe, trans- 
 lated from the Swedish, with an original appendix by Christer P. Sandberg; with a preface by John 
 Percy, M.D., F.R.S., London, 1869. (Sandberg's investigations appear in the appendix.) 
 % Iron and Steel; Materials of Engineering, Thurston, 1909 p. 556. 
 
STRENGTH OF THE RAIL 285 
 
 given were the sole reasons why metals should more readily break in cold than 
 in hot weather, and, having obtained the consent of the State Railway Adminis- 
 tration, he conducted a series of experiments in the summer and winter of 1867, 
 at Stockholm, to determine whether, with equal rigidity of supports, iron rails 
 would yield with equal readiness to blows at the two extremes of temperature. 
 
 The rails experimented upon were each cut in halves, and one piece was 
 tested in cold and the other in warm weather, at temperatures of 10 degrees and 
 84 degrees Fahr. respectively. The supports at the end of the rails were granite 
 blocks placed four feet apart, and resting on the smoothly leveled surface of the 
 granite rock. They were broken by a heavy drop weighing 9 cwt. 
 Sandberg's conclusions from twenty experiments are thus given: 
 
 " (1) That for such iron as is usually employed for rails in the three principal 
 rail-making countries (Wales, France, and Belgium), the breaking strain, as 
 tested by sudden blows or shocks, is considerably influenced by cold, such iron 
 exhibiting at 10 degrees Fahr. only from one-third to one-fourth of the strength 
 which it possesses at 84 degrees Fahr. 
 
 " (2) That the ductility and flexibility of such iron is also much affected 
 by cold; rails broken at 10 degrees Fahr. showing, on an average, a perma- 
 nent deflection of less than one inch, whilst the other halves of the same 
 rails, broken at 84 degrees Fahr., showed a set of more than 4 inches before 
 fracture. 
 
 " (3) That at summer heat the strength of Aberdare rails was 20 per cent 
 greater than that of the Creusot rails, but that in winter the latter were 20 per 
 cent stronger than the former." 
 
 Sandberg suggests that this considerable lack of toughness at low tem- 
 peratures may be due to the "cold-shortness" produced by the presence of 
 phosphorus. 
 
 Jouraffsky, of St. Petersburg, has reported* the results of tests of rails 
 made for the Russian government, which supplement the preceding in a very 
 valuable manner. It was found that by placing pieces of rail from 6 feet to 
 8 feet long in a mixture of ice and salt, the temperature of the rail could be 
 lowered in a very short space of time, during warm weather, 36 degrees Fahr. 
 below freezing point. 
 
 A special commission, Messrs. Erakoff, Beck, Guerhard, Nicolia, and Feo- 
 dossieff, was appointed to carry out a series of tests on this plan. Pieces of 
 rail 6 feet long were taken in pairs, one of which was tested at the 
 natural temperature, the others being placed in a box filled with a mixture of 
 
 * Communicated to the London Meeting of the Iron and Steel Institute, 1879. 
 
286 STEEL RAILS 
 
 two parts of broken ice and one part of salt, and, after being cooled to a 
 temperature of from +3 degrees to -6 degrees Fahr., which occurred in half 
 an hour, they were all submitted to the same tests. Altogether, 86 samples 
 were tested, and these were, for the sake of comparison, divided into two 
 groups, viz.: (1) Rails which broke under the test; and (2) rails which stood 
 the test. 
 
 The results indicated that the brittleness of the steel increases very much 
 at low temperature if it contains more than a moderate amount of phosphorus, 
 silicon, and carbon. The total of the three elements in the rails which broke 
 under the test averages 0.54 per cent, and in those which stood the same test 
 0.41 per cent, the first average (0.54 per cent) varying from 0.44 to 0.67 per 
 cent, and the second average (0.41 per cent) varying from 0.37 to 0.55 per cent. 
 
 The tests on steel at different temperatures made at the Watertown Arsenal 
 in 1888,* showed within the range of 0.37 to 0.71 per cent carbon a slight increase 
 in the elastic limit and tensile strength at degrees Fahr. above that at summer 
 heat, accompanied by very little change in the elongation or contraction of area. 
 
 Dr. P. H. Dudley recommends the use of basic open-hearth rails of 0.62 
 to 0.75 carbon with phosphorus under 0.04 to insure a more uniform range of 
 toughness and ductility of metal where exposed to low temperatures than has 
 been obtained in plain Bessemer with 0.50 carbon and 0.10 phosphorus. 
 
 Tests of Bessemer heats were made in which one-tenth of one per cent of 
 metallic titanium was added to the steel and the carbon increased from 0.50 to 
 a range of 0.60 or 0.70, the metal having higher physical properties and tough- 
 ness at the same time. The manganese and silicon are lowered slightly to 
 prevent shrinkage cavities in the ingots. 
 
 Tests were made under the drop comparing the ordinary Bessemer rails 
 and this grade of steel cooled to zero temperatures, and some specimens were 
 also cooled to 22 degrees Fahr. below zero. The tests of the Bessemer steel 
 with the one-tenth of one per cent of metallic titanium withstood a drop of 
 2000 pounds, falling sixteen feet without breaking, while the plain Bessemer 
 would fail at the low temperatures at about one-half that height. 
 
 Dr. Dudley f considers that basic open-hearth rails (C = 0.68; Mn = 
 0.86; Si = 0.10; P = 0.012) which show ductility ranging from 15 to 20 per cent 
 under the drop test are proper for use under high-speed trains where the tem- 
 perature in winter falls to 20 degrees Fahr. below zero. 
 
 * Tests of Metals, 1888, House Doc. No. 45, 50th Congress, 2nd Sess., p. 505. 
 t Dudley on Ductility Tests on Rails. Proceedings American Society for Testing Materials, 
 Vol. X, 1910, p. 229. 
 
STRENGTH OF THE RAIL 287 
 
 The question of obtaining a proper amount of ductility in rails, used in cold 
 climates, is a very important one. The experience of the railroads in the 
 Northern and Western States, during the very cold weather of the winter of 
 1911-1912, was more severe than ever before reported. The effects were so 
 great that rails which had heretofore been quite free from breakages were broken 
 in considerable quantities by the wheels of the passing trains. The record of 
 rails of large ductility or tenacity and toughness, on the other hand, showed a 
 much greater freedom from breakages. 
 
 With the recently adopted standard drop-testing machine (see Article 27) 
 the ductility of the rail can be measured with much more accuracy than was 
 possible in the machines which preceded it. With the older machines the re- 
 bound of a 2,000-pound hammer was as large as 12 to 14 inches, while in the 
 present machine it is confined to 3 or 4 inches. 
 
 The maximum elongation per inch can be obtained by stamping the base, 
 head, or edge of the base of the butt, as the case may be, before testing with a 
 spacing bar of six inches, directly under the point of impact for either a single 
 blow or for two or more required to exhaust the ductility of the metal. These 
 six inches include about two-thirds of the metal affected by the impact. 
 
 The elongation in the base of a 6-inch, 100-pound rail, with a moment of 
 inertia of 48.5, under a single blow for an 18-foot fall in the present drop-testing 
 machine, will be from 6 to 7 per cent for a steel of 0.50 carbon, 0.10 phosphorus, 
 and 1.00 manganese. The elongation of the metal under the drop-testing 
 machine compares favorably with that obtained by static loads in the tension 
 machine. The tendency is an increase of possibly one or more per cent, owing 
 to the fact that the base of the rail in stretching does not neck as in the case of 
 a tensile specimen. 
 
 To raise the mean ductility, it has been found necessary to reduce the 
 average percentage of carbon in heavy sections to 0.50, when the phosphorus 
 content is 0.10, for rails which are to be used where the temperatures fall below 
 zero. The Bessemer rails, owing to the greater content of phosphorus, oxides, 
 and nitrogen, show a greater tendency to irregular low ranges of ductility than 
 the basic open-hearth rails. The use of ferro-titanium in Bessemer steel, to 
 take up a large percentage of the oxides and also a part of the nitrogen, makes 
 it possible to increase the carbon content without any sacrifice to the ductility.* 
 
 Probably "some of the failures of rails in cold weather can be attributed to the 
 effect, already noted in article 24, of the contraction of the metal, which may set 
 
 * The subject of ductility in rail steel has been reported upon very fully by Dr. P. H. Dudley in 
 papers presented to the American Society for Testing Materials, see Proceedings, Vol. X, 1910, pp. 
 223-232, and Vol. XI, 1911, pp. 454-461. 
 
288 STEEL RAILS 
 
 up tensile stresses of some magnitude in the rails before the ends render in the 
 splice bars. 
 
 The effect of frost, in heaving the track where the ballast or subgrade con- 
 tains much moisture, is to cause an irregular surface and, on account of some 
 of the ties rising or "heaving" above the others, often produces larger bending 
 moments in the rail than would be expected if the ties were free to adjust them- 
 selves to the elastic curve of the rail. In the case of well-drained track, on stone 
 ballast, where heaving is absent, smaller bending moments will be found in the 
 frozen track than when in its natural condition. 
 
 27. Physical Tests of the Strength of the Rail 
 
 The impact hammer or drop test, introduced by Sandberg and Styffe, 
 in 1868, is most generally used in this country in testing the strength of the rail. 
 
 From the very prominent place given drop tests in rail specifications,* 
 it might be seen that the behavior of the rail under the drop test is generally 
 regarded as valuable information as to its character. As a matter of fact, 
 however, engineers differ widely as to the advisability of accepting this test as 
 an index to the reliability of the rail, on account of the great variation in the 
 results obtained. The test shows, it is true, whether the piece being tested is 
 brittle or not, and by observation of the permanent set, whether the steel 
 
 Note. For an account of the state of knowledge relating to impact tests and for a bibliography of 
 literature, reference is made to the following: 
 
 American Section, International Association for Testing Materials: Bulletin No. 5, October, 
 1899. Report of Committee on Present State of Knowledge Concerning Impact Tests. W. K. 
 Hatt and Edgar Marburg. 
 
 Bibliography on Impact Tests and Impact Testing Machines. Proceedings American Society 
 for Testing Materials, Volume II, page 283. W. K. Hatt and Edgar Marburg. 
 
 The Resistance of Metals under Impact. Mansfield Merriman. Proceedings American 
 Association for the Advancement of Science, Volume 43, 1894. 
 
 Theory of Impact and its Application to Testing Materials. H. D. Tiemann. Journal of 
 the Franklin Institute, October and November, 1909. 
 
 International Association for Testing Materials, Vth Congress, Copenhagen, 1909. Impact 
 tests papers, III,, III 2 , Ills, IIL, IIL, His, IIL, Ills. 
 
 Elongation and Ductility Tests of Rail Sections under the Manufacturers' Standard Drop- 
 Testing Machine. P. H. Dudley, Proceedings American Society for Testing Materials, Vol. X (1910), 
 p. 223. 
 
 The Same. Iron Trade Review, Vol. 47, p. 410. 
 
 Nouvelle Methode d'essai des Rails, Ch. Fremont Genie Civil, Vol. 59 (1911), p. 7, 26, 48, 72. 
 
 The same. Railway Age Gazette, Vol. 51, p. 1176. 
 
 New Types of Impact Testing Machines for Determining Fragility of Metals. T. Y. Olsen. 
 Proceedings American Society for Testing Materials, Vol. XI (1911), p. 815. 
 
 * Some Results Showing the Behavior of Rails under the Drop Test, and Proposed New Form 
 of Standard Drop Testing-Machine. S. S. Martin. Proceedings Am. Soc. for Test. Materials, 1908, 
 Vol. VIII. 
 
STRENGTH OF THE RAIL 289 
 
 is soft or hard. Recent work in Germany and France points to the conclusion 
 that some form of impact test is found necessary to detect faults of structure 
 that are not evidenced by the static test. 
 
 Average specifications, which a majority of the railroads have in recent 
 years used as a standard, contain the following clause as to drop test: 
 
 One test shall be made on a piece of rail, not less than 4 feet, nor more than 6 feet, selected 
 from each blow of steel. The test piece shall be taken from the top of the ingot. The rails shall 
 be placed head upwards on the supports, and the various sections shall be subjected to the follow- 
 ing impact tests under free falling weight: 
 
 70 to 79 pound rail, 18-foot drop. 
 
 80 to 89 pound rail, 20-foot drop. 
 
 90 to 100 pound rail, 22-foot drop. 
 
 If any rail breaks, when subjected to the drop test, two additional tests may be made of 
 other rails from the same blow of steel, also taken from the top of the ingot, and if either of these 
 latter rails fail, all the rails of that blow which they represent will be rejected; but if both these 
 additional test pieces meet the requirements, all the rails of the blow which they represent will 
 be accepted. 
 
 The drop-testing machine shall have a tup of 2000 pounds weight, the striking face of which 
 shall have a radius of not more than 5 inches, and the test rail shall be placed head upward on 
 solid supports 3 feet apart. The anvil block shall weigh at least 20,000 pounds, and the supports 
 shall be part of or firmly secured to the anvil. The report of the drop test shall state the at- 
 mospheric temperature at the time the test was made. 
 
 These specifications, while used by the railroads, had to be modified ac- 
 cording to the character of the drop-testing machines at the different mills. 
 Thus, we find machines answering closely the following descriptions : 
 
 1. A drop-test machine consisting of some concrete and loose stone, supporting a number 
 of 12 by 12 inch oak ties, 12 feet long, on which is placed an oak block 18 inches by 18 inches by 
 11 feet. On the oak block are two steel plates 1 by 18 inches by 7 feet, which become the bearings 
 for the rail supports. These supports weigh 1300 pounds. 
 
 2. A drop-test machine consisting of a wooden foundation 4 feet deep and 10 by 10 feet, 
 on which were placed two blooms, probably 8 by 10 inches by 10 feet. On the blooms are placed 
 the rail supports. 
 
 3. A drop-test machine consisting of a concrete or stone foundation, on which rests a 20,000- 
 pound anvil, to which the rail supports are securely fastened. 
 
 Up to within the last few years no rail mill has been equipped with a drop- 
 testing machine for rails that was built on thoroughly scientific principles, owing 
 to the lack of proper foundations or proper anvil, as well as many other essential 
 details. Further, no two rail mills had machines built on even comparatively 
 the same lines. Consequently, any exact determination of the loss of energy 
 of the falling weight which is dissipated by the machine would have had no 
 general application, and the results obtained from testing rails in the drop- 
 testing machines of any two mills were not comparable. 
 
290 STEEL RAILS 
 
 ' The Manufacturers' Committee, recognizing these defects, prepared speci- 
 fications and plans of a proposed standard drop-testing machine that will give 
 satisfactory and comparable results, and the Rail Committee of the American 
 Railway Engineering Association has, with certain small modifications and 
 additions, approved these plans and specifications. 
 
 Specifications for Drop-Testing Machine 
 
 (See Fig. 204.) 
 
 1. The machine shall be arranged to allow a 2000-pound tup to fall freely at least 25 feet 
 on the center of a rail resting on supports that can be adjusted to spans varying from 3 feet to 
 4 feet 6 inches. 
 
 2. The anvil shall be a solid casting, weighing, with the attachments that move with it, 
 20,000 pounds. It shall be free to move vertically independently of the lead columns and shall 
 be supported on 20 springs known as the standard " C " spring, without center coil, as employed 
 by the Master Car Builders' Association (their figure 5614). This spring has a free length of 
 8 1 inches, an outside diameter oi 5ts inches, and is made from a bar having a diameter of 1^ 
 inches. These springs are to be arranged in groups of five at each corner of the anvil and are to 
 be held in place by hubs raised on the top of the base plate and by circular pockets on the under 
 side of the anvil. Anvil to be guided in its vertical movement by removable finished wearing 
 strips, these strips to be suitably attached to the finished edges of the column base. 
 
 3. The base plate shall be of cast iron or cast steel, 8 inches thick in the area covered by 
 the anvil. It shall be firmly secured to the substructure by four bolts 2 inches in diameter. 
 
 4. The substructure shall consist of a timber grillage resting on a masonry foundation. The 
 grillage shall project 9 inches beyond the ends of the base plate, and clear the columns at the side. 
 It shall consist of one course of 12 inches by 12 inches sound oak or southern yellow pine, pref- 
 erably creosoted, laid close and well bolted together. The masonry, preferably concrete, shall 
 not be less than 5 feet deep below the grillage and be suitably supported on the subsoil. 
 
 5. The pedestals for supporting the test rail shall be substantial castings, and the surface of 
 the anvil between these pedestals shall be formed to receive a wooden block to absorb shock under 
 broken pieces. The rail supports shall be removable pieces of steel securely held in the pedestals 
 having an upper cylindrical bearing surface with a radius of 5 inches. The pedestals shall be 
 adjustable to spans, varying from 3 feet minimum to 4 feet 6 inches maximum between centers. 
 They shall be securely held together and so fixed to the anvil as to insure that the center of span 
 shall always coincide with the center between leads. 
 
 6. The leads shall be firmly connected to the base plate and well braced. They shall be 
 long enough to provide the prescribed free fall of the tup. They shall be provided with a con- 
 venient ladder and a plainly marked gauge, divided into one-foot intervals. The zero of this gauge 
 shall be 5 $ inches above the top of the rail support, and the specified height of drop shall be 
 measured from this zero, irrespective of the height of the rail being tested. One of the guides 
 shall have a removable section, 6 feet long at the bottom, so that the tup or tripping block can 
 be readily removed. 
 
 7. The tup shall weigh, with the accessories that drop with it, 2000 pounds. The striking 
 die shall be steel having a cylindrical striking face with radius of 5 inches and a length of 12 inches. 
 The guide grooves shall have finished surfaces. The tripping head shall allow a grip of the tongs 
 that will release at the exact height for which the tripping device is set and that will be safe from 
 accidental release while the test piece is being shifted. 
 
 8. The tongs and tripping device shall be arranged to release the tup automatically only; 
 no manual releasing will be allowed. The tripping device shall be easily adjustable at one-foot 
 intervals. 
 

 Fia. 204. — Standard Drop Testing Machine, as adopted by Committee of Rail Manufacturers of the United States, April 8, 1908. 
 Recommended by Committee on Rail of the American Railway Engineering Association at Meeting of June 26, 1908. 
 Adopted by the Association, Vol. 10, Pt. 1, 1909, pp. 369-373, 375, 395, 396; Vol. 11, Pt. 1, 1910, pp. 240, 252, 562. 
 Sellew." Steel Rails." 
 
STRENGTH OF THE RAIL 
 
 291 
 
 A diagram showing graphically the relation between results of tests as 
 between the old and new methods, furnished by Mr. Thomas H. Johnson, 
 is given in Fig. 205. 
 
 These tests were made on rails of the same grade of steel, viz., carbon, 
 about .50, and manganese, .90 to 1.00. All of the tests were made with a 
 
 Fig. 205. — Diagram of Tests with Drop-Testing Machines of Old and New Design. (Johnson.) 
 
 2000-pound tup, with a radius of striking face 5 inches, and span between 
 centers of supports 3 feet. Table LXXI gives the result of these tests. 
 
 Since 1908 a number of machines of this design have been installed at 
 various mills with satisfactory results. The introduction of a standard drop- 
 testing machine has been of such benefit to the manufacturers and consumers 
 of this country that the International Association for Testing Materials has pub- 
 lished a description and cuts of the American Standard drop-testing machine in 
 French, German, and English.* The seventh London Resolution of the Council 
 was as follows : 
 
 "That a standard drop-testing machine for rails be adopted in each country, as has already- 
 been done in the United States, in order to make tests comparative." 
 
 * Proceedings, Vol. 11, No. 4, May 20, 1911, p. 237. 
 
292 
 
 STEEL RAILS 
 
 If we consider what occurs in dealing the blow on the rail with the falling 
 weight, it will be seen that the work utilized by the rail in order to take a cor- 
 responding set is not the potential energy of the weight at the moment when 
 the weight touches the bar. The energy of the falling weight serves to deform 
 the rail, to compress the supports, the body of the drop weight, the anvil, and 
 the ground upon which the anvil is placed. 
 
 TABLE LXXI.— TESTS ON NEW AND OLD DROP-TESTING MACHINE OF P. R.R. 
 AND P. S. RAIL SECTIONS 
 
 (Am. Ry. Eng. Assn.) 
 
 
 New Standard Machine. 
 
 Old-Style Machine. 
 
 Height of 
 
 
 
 
 
 
 
 
 
 P. R.R. 
 
 P. S. 
 
 P. R.R. 
 
 P. S. 
 
 P. R.R. 
 
 P. s. 
 
 P. R.R. P. S. 
 
 
 85-Pound 
 
 85-Pound 
 
 100-Pound 
 
 100-Pound 
 
 85-Pound 85 
 
 Pound 
 
 100-Pound 100-Pound 
 
 
 Set. 
 
 Set. 
 
 Set. 
 
 Set. 
 
 Set. 
 
 Set. 
 
 Set. 
 
 Set. 
 
 Feet. 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 Inches. I 
 
 iches. 
 
 Inches. Inches. 
 
 .5 
 
 .02 
 
 .02 
 
 .02 
 
 .02 
 
 .0 
 
 
 
 .0 
 
 
 
 1 
 
 .08 
 
 .08 
 
 
 08 
 
 .06 
 
 .031 
 
 
 
 .031 
 
 
 
 1.5 
 
 .16 
 
 .17 
 
 
 14 
 
 .15 
 
 .093 
 
 062 
 
 .062 
 
 031 
 
 2 
 
 .25 
 
 .22 
 
 
 19 
 
 .19 
 
 .125 
 
 093 
 
 .125 
 
 093 
 
 3 
 
 .38 
 
 .37 
 
 
 28 
 
 .25 
 
 .250 
 
 171 
 
 .218 
 
 109 
 
 4 
 
 .49 
 
 .47 
 
 
 40 
 
 .38 
 
 .343 
 
 250 
 
 .281 
 
 250 
 
 5 
 
 .60 
 
 .53 
 
 
 50 
 
 .45 
 
 .406 
 
 343 
 
 .375 
 
 312 
 
 6 
 
 .76 
 
 .66 
 
 
 64 
 
 .62 
 
 .513 
 
 468 
 
 .437 
 
 375 
 
 7 
 
 .88 
 
 .76 
 
 
 74 
 
 .70 
 
 .700 
 
 600 
 
 .650 
 
 375 
 
 8 
 
 .99 
 
 .90 
 
 
 86 
 
 .81 
 
 .800 
 
 700 
 
 .700 
 
 600 
 
 9 
 
 1.12 
 
 1.01 
 
 
 98 
 
 .90 
 
 .900 
 
 800 
 
 .750 
 
 700 
 
 10 
 
 1.20 
 
 1.13 
 
 1 
 
 06 
 
 1.02 
 
 1.000 
 
 850 
 
 .800 
 
 800 
 
 Breuil* has shown that for the same amount of actual work the bending 
 curve of the impact test is the same as the slow-bending test, and Hattf has 
 observed that there is little difference in the total elongation and unit rupture 
 work whether the bar is ruptured in ten minutes or in from one to two one- 
 hundredths of a second.J 
 
 * Relation between the Effect of Stresses slowly applied and Stresses suddenly applied in the 
 Case of Iron and Steel. P. Breuil. 1904. 
 
 t Tensile Impact Tests of Metals. Hatt. Proceedings Am. Soc. for Test. Materials, 1904, 
 Vol. IV. 
 
 | As further evidence we may quote the following opinions (Report on Impact Tests of Metals. 
 Official report by G. Charpy, Montlucon, International Association for Testing Materials, 5th Con- 
 gress, Copenhagen, 1909. McGraw-Hill Book Company, New York) : 
 
 Captain Duguet writes: " The effect (of the duration of the stress) is very marked, especially 
 during the period of great deformations; that would in itself suffice to render any too detailed 
 investigation of the phenomena illusory. But we must not exaggerate the importance of this 
 point. The extreme deformations are, in the case of soft steel, submitted to bending, very sen- 
 sibly the same, whether they be produced slowly by hydraulic pressure or by the impact of a tup." 
 
 In his treatise on material testing, Professor Martens points out that, according to the 
 experiments of Kick, the velocity of fall has only an insignificant influence on the magnitude of 
 the deformation in impact bending tests. As regards tensile strength tests, Professor Martens 
 concludes: " From the impact tension tests so far conducted in the Charlottenburg Laboratory, 
 I have acquired the conviction that the deformations are produced exactly as by slow tension. 
 
STRENGTH OF THE RAIL 293 
 
 Fig. 206 shows the amount of energy dissipated in 90-pound A. R. A. type 
 " B " Bessemer rails when tested in the drop-testing machine.* 
 
 The weight of the tup was 2000 pounds and the distance between supports 
 three feet in both the dynamic and static tests. The anvil in the drop test 
 weighed 10 tons, spring supported. Calculating the work done on the rail in 
 the static test from the load deflection diagram, and in the drop test from 
 the height of drop and weight of tup, it would appear from the lower diagram 
 of the figure that about two-thirds of the energy of the falling tup is utilized 
 to deflect this rail. 
 
 The difficulty of comparing the values of the stresses in impact tests with 
 those which occur in static tests (i.e., where the momentum of the load does not 
 factor) lies in the difficulty of accurately determining the value of the force 
 acting between the hammer and the specimen in the former. Comparisons of 
 the total work required to rupture a specimen, or to produce a given deflection, 
 are comparatively simple. 
 
 The manner in which impact stresses are related to so-called static stresses 
 requires careful theoretical consideration before it can be clearly comprehended. 
 The author is indebted to the work of Mr. H. D. Tiemann f for the following 
 presentation of the subject. 
 
 The rail tested by impact is in reality in the nature of a cushion between 
 the two impacting bodies, namely, the tup A and the anvil B, and the anvil B 
 must be of such proportions that its relative velocity v b to that of the com- 
 mon center of gravity of itself and the tup, V, may be disregarded, as other- 
 wise a correction must necessarily be made, which not only complicates the 
 " subject, but on account of the nature of the foundation of the anvil is almost 
 impossible to apply. 
 
 In tension tests by means of several blows, we find often that the elongation was greater than 
 in the slow-tension tests." 
 
 M. Lebasteur (Annales des Ponts et Chaussees, 1890) had likewise arrived at the following 
 conclusions : 
 
 " 1. The elongations observed in the fracture of bars under high-drop impact tests are nearly 
 identical with the elongations of similar bars under slow tension. 
 
 "2. The appearances of the broken sections are absolutely the same in the two cases. 
 
 "3. The total intensities of the blows necessary to break bars under high-drop impacts are 
 proportional to the dynamic resistance to rupture (as determined from the area of the curve of 
 slow tensile stress);" and further on M. Lebasteur added that "the dynamic resistance to 
 rupture measures the strengths of metals equally well for dynamic as for static stresses." 
 
 * Report, M. H. Wickhorst to Rail Committee of Am. Ry. Eng. & M. of W. Assn., Proceedings, 
 Vol. 12, Part 2, p. 389-394. 
 
 t The Theory of Impact and its Application to Testing Materials, Journal of the Franklin In- 
 stitute, October and November, 1909. 
 
STEEL RAILS 
 
 sannod avoi 
 
 
 
 
 
 
 
 
 z 
 z9 
 - \n 
 
 
 
 
 
 en w 
 
 S. CD 
 
 
 
 
 
 zoj^ 
 
 
 
 
 
 Qz 
 
 
 
 
 
 
 
 
 Po 
 
 \ 
 
 \ 
 
 
 
 A 
 
 UJ 
 
 O 
 
 a. 
 
 V 
 
 J 
 
 
 \ 
 
 A 
 
 
 
 
 
 
 
 ■a 
 
 o2 g- 
 
 Z .9 
 Q 
 
 o o 
 
 (SI — 
 
 SNOJ. JLOOJ 
 
 O W O 10 
 
 CM — ~ 
 
 133J dOHQ dO 1H9I3H 
 
STRENGTH OF THE RAIL 295 
 
 If we consider what occurs at impact, it is seen that at the beginning of 
 contact a mutual repellant force F begins to come into play between A and B, 
 which produces a change in the relative velocities (y a - V) and (v b - V), where 
 v a = the velocity of A, 
 v b = " " " B, 
 
 V = " " " the mutual center of gravity of A 
 and B along a line joining their 
 respective centers of gravity. 
 
 This force F, starting from at first contact and increasing to a maximum 
 when both relative velocities of A and B are brought to 0, then when rebound 
 begins again decreasing until it becomes at departure, is counterbalanced 
 directly by the local compression of the material of both bodies at the point 
 of contact. 
 
 This force is made up of two parts, one being the elastic resistance of the 
 body to compression or deformation, /, and the other of the force necessary to 
 produce the local acceleration of the particles compressed, f x . 
 F = f + f, 
 At any instant F is exactly equal to the change in momentum produced 
 by its action divided by the time required to produce this change : 
 p m (v - vi) 
 F= t-t, • 
 
 t-, mdv d 2 s 
 
 F --df = m di> = ma - 
 
 m = mass, 
 s = space, 
 t = time, 
 v = velocity 
 a = acceleration. 
 
 Consequently, if the time-velocity curve can be determined, the force F can 
 be calculated for any instant. 
 
 Let us examine first the relations between the various quantities of an 
 impact test graphically, and then proceed to develop formulae of the time-space 
 curve and of the interrelations of the values. 
 
 Consider, first, a rail lying horizontally on the anvil B and supported 
 freely at its ends, and let it be struck at the center by a falling tup A. For 
 simplicity let us consider the rail as massless as well as weightless. 
 
 Or more exactly, 
 where 
 
296 
 
 STEEL RAILS 
 
 Take the velocities as relative to the anvil, as explained above, or assume 
 the mass of the anvil so great that its motion may be taken as zero. Let 
 the motion of the center of gravity of the tup A be plotted as in Fig. 207, with 
 space as ordinates and time as abscissae. If the tup start at some point J, and 
 fall freely the distance H before striking the rail, the curve JB will be a parab- 
 ola, or if its velocity be uniform its motion would give the straight line 
 JiB. Impact with the rail begins at B, and, assuming the rail as massless, 
 the only resistance offered to the momentum of the tup will be the bending 
 stress of the rail, which will be the force F. The motion curve then becomes 
 
 7 
 
 Fig. 207. — Time-deflection Curve, Massless Beam, within the Elastic Limit. 
 (Journal of the Franklin Institute. — Tiemann.) 
 
 BC. The resisting force, starting with at B, increases in proportion to the 
 deflection of the rail until the maximum value is reached at C. (In this case, 
 let the elastic limit of the rail be not exceeded.) It is this force F which 
 overcomes the momentum of the tup A by producing a negative accelera- 
 tion until the momentum is reduced to at C. Rebound then begins. If the 
 rail is perfectly elastic, the force F continuing to act will restore exactly the 
 same amount of momentum to A, and in the same manner, but in the opposite 
 direction that it had at the beginning of contact. The curve will be CD. At 
 D departure takes place (the rail being considered as massless). 
 
 The total mutual repellent force F acting on the rail, between the tup A 
 and the anvil B, is at any instant equal to 
 
 d 2 S , 
 
 dv 
 
 A, 
 
 where S is the space traversed by A and v is its velocity at the instant under 
 
STRENGTH OF THE RAIL 297 
 
 consideration. If S is in feet, t in seconds, and A = (weight in pounds -4- accel- 
 eration of gravity in feet), then F is given in pounds, by the last formula. 
 
 If the force of gravity is to be considered, as well as the initial velocity of 
 A at impact, then this formula should be written: 
 
 F-W a = ^A= d i t A, 
 
 at 2 at 
 
 where W a is the weight in pounds of the tup A. This could be avoided by 
 having the tup move horizontally instead of falling vertically. 
 
 Examining Fig. 207, it will be noted that the acceleration of the tup from 
 J to B is equal to g and is produced by the uniform force W a . From B to 
 D it is produced by the force W a — F and becomes negative as soon as the 
 value of F begins to exceed W a . This is at the point of reverse curvature, since 
 the acceleration 
 
 cN 
 dt 2 
 is zero, and evidently occurs at the point b or the deflection which would be 
 produced by the static load W a . In horizontal motion the change would occur 
 at contact, B. It should be remembered that while W a is a constant force, 
 F is a variable, ranging from at B, to a maximum at C, and again to at D. 
 Whenever 
 
 dh 
 dt 2 
 becomes a maximum the force F becomes a maximum, and this evidently occurs 
 at the sharpest part of the curve, which in this particular case is at C. 
 
 The value of F at any instant may, therefore, be determined from the curve 
 
 F = ^ x A + W a . 
 
 at 1 
 
 If the force F of the impulse becomes sufficient to cause complete failure 
 of the specimen, the conditions are those shown in Fig. 208. 
 
 The first part of the curve JB is the same as before. The velocity or 
 momentum of the tup is, in this case, not entirely overcome by the resistance 
 F of the rail, so that at failure the tup retains a portion of its velocity as 
 indicated by the tangent line DE X . If the tup works vertically in free fall, 
 instead of horizontally, then the curve DE is again a parabola of free fall. In 
 this case the force F becomes a maximum at some point C, when the curvature 
 is sharpest, and must be determined from the curve by 
 d 2 s 
 dt 2 ' 
 since there is no means of calculating it mathematically. 
 
298 
 
 STEEL RAILS 
 
 The curve can be conveniently obtained by some mechanical device by 
 means of which the falling tup makes a tracing on a uniformly revolving drum. 
 When F is thus determined the maximum strength values may be calculated. 
 
 To supplement the information furnished by the drop test engineers are turn- 
 ing their attention to other means of testing the physical properties of the rail. 
 
 Fig. 208. — Time-deflection Curve, Beam Stressed beyond the 
 Elastic Limit (Journal of the Franklin Institute. — Tiemann.) 
 
 * The Baltimore and Ohio Railroad 
 Company, in connection with its inves- 
 tigations on rail, has been making use of 
 the scleroscope (Fig. 209), an instrument 
 for determining the degree of hardness of 
 metals. 
 
 f The principle of the scleroscope 
 (Greek sclero = hardness) consists of drop- 
 ping a small plunger hammer from a fixed 
 height onto the surface of the material 
 whose hardness is to be measured. This 
 hammer after striking, by no other force 
 than its own weight, rebounds to variable 
 heights, depending on the hardness or 
 
 amount of resistance to penetration offered Fig. 209.— Scleroscope. (Am. Ry. Eng. Assn.) 
 
 -by the metal tested. The rebound of the hammer is used to measure the 
 hardness of the metal, and the scale shown on the glass tube is simply for 
 
 * General Information Concerning the Scleroscope and its Use on the Baltimore and Ohio Rail- 
 road. A. W. Thompson. Proceedings, Am. Ry. Eng. & M. of W. Assn., 1910, Vol. 11, Part I. 
 
 t See also The Scleroscope, Albert F. Shore, p. 490, Proceedings American Society for Testing 
 Materials, Vol. X, 1910. 
 
STRENGTH OF THE RAIL 
 
 299 
 
 comparative purposes and has no direct numerical value. This scale has 140 
 graduations, and a test of very hard steel has resulted in a rebound to the point 
 marked 110, while soft brass results in a rebound to the point marked 12, and 
 lead is about 2 per cent of hard steel. 
 
 Figs. 210, 211, and 212 present examples of tests. The numbers in these 
 figures indicate the degrees of hardness. 
 
 Fig. 210 is an A. R. A. section of open-hearth rail, as rolled by the Bethle- 
 hem Steel Company, and is a new section which has not been in the track. It 
 will be noted that the hardness on the top of the head of the rail is practically 
 
 Fig. 210. — Scleroscope Tests on Open Hearth 
 Rail (New.) (Am. Ry. Eng. Assn.) 
 
 Fig. 211. — Scleroscope Tests on Bessemer Rail. 
 (Am. Ry. Eng. Assn.) 
 
 the same as the steel in the section of the rail just below the surface. The center 
 of the head appears to be the hardest, as well as a line through the center of 
 the web and base. The upper corners of the head are comparatively soft, the 
 ends of the base, however, being very much softer than any of the rest of the rail. 
 Fig. 211 is a section of a Bessemer rail rolled at Buffalo in 1908. It is a 
 crop from the top end of a top rail. Although the specimen was from the top of 
 the ingot, there is a difference of but three points in the. readings throughout the 
 head. The section, where polished and etched, showed rather dimly marked 
 segregation. The head and base when planed into and etched showed some 
 dark streaks in the head and light streaks and fissures in the base. The top 
 of the head for finch depth was sound. The experimenter says: " The section 
 as a whole is more uniform than is usually to be found in top, middle, or bottom 
 rail of a Bessemer ' ingot.' " 
 
300 
 
 STEEL RAILS 
 
 ^ 
 
 111. 
 
 Fig. 212 shows the comparative hardness on different lines on experimental 
 titanium rails. This test indicates a skin of soft metal across the top of the 
 head, but as soon as this is penetrated the 
 hardness is reached, which compares favor- 
 ably with any part of the whole section. 
 
 * Experiments were made at the 
 laboratories of McGill University on the 
 value of the indentation test for steel rails 
 in regard to essential qualities desired in 
 service. The study of this method of 
 testing was suggested by tests made on a 
 large number of rail sections by the Chief 
 Engineer of the Canadian Pacific Railway, 
 a spherical punch .75 inch in diameter 
 
 Fig. 212.— Scleroscope Tests on New Titanium being used, with a load of 100,000 pounds 
 
 Ran. (Am. Ry. En g . Assn.) applied by an Emery testing machine for 
 
 10 seconds after commencing the load, and the indentation was measured by 
 an instrument reading to yoVo i ncn - 
 
 The tests, conducted by Mr. Dutcher, were on a set of bars of 2.5 by .75- 
 inch section containing known percentages of carbon, which were verified by 
 tests, and varied between .11 per cent and .96 per cent. The punches used 
 (in addition to the foregoing) were a 60° cone, a 90° cone, and a paraboloid. 
 
 The term " hardness factor " applied to the results was obtained by divid- 
 ing the projected area of the indentation on the surface of the specimen into the 
 load applied. It was found that the yield point (as determined by tensile 
 tests) varies directly as the hardness factor. The percentage elongation curve 
 is also fairly straight between 200,000 pounds hardness factor (.10 per cent 
 carbon steel) and 450,000 pounds hardness factor (about .70 per cent 
 carbon steel); and the percentage of carbon varies directly with the hardness 
 factor up to about .90 per cent. 
 
 f There have been several methods proposed to test the hardness of the 
 metal by ball-pressure tests. In the Brinell test J a hard ball of steel is forced 
 
 * Transactions of the Canadian Society of Civil Engineers, Dutcher, Vol. XXI, pp. 47-88. 
 
 t Hardness Tests. Official report by Dr. techn. P. Ludwik, of Vienna. International Associa- 
 tion for Testing Materials, 5th Congress, Copenhagen, 1909. McGraw-Hill Book Company, New York, 
 also various technical papers on Hardness Tests. Proceedings, American Society for Testing Materials, 
 Vol. XI, 1911, pp. 707-743. 
 
 % Compare P. Ludwik, " TJber Hartebestimmung mittelst der Brinnellschen Kugeldruckprobe 
 und verwandter Eindruckverfahren," " Zeitschrift des osterr. Ingenieur und Architekten-Vereines," 
 1907, Nr. 11 und 12 (Nr. 12, p. 205, extensive literature references). 
 
STRENGTH OF THE RAIL 
 
 301 
 
 by quiet pressure into the material to be examined ; the diameter of the spherical 
 impression is determined, as a rule, with the aid of a special microscope, and the 
 area of the cavity is calculated. The quotient of pressure (in kilograms) by the 
 area (in millimeters 2 ) is Brinell's hardness number H. 
 
 The cone-pressure test marks a transition from the ball-pressure methods 
 to scratch methods. It is the outcome of efforts to simplify the Brinell test, with 
 the further object of making the hardness number independent of the load and 
 of the dimensions of the impression. 
 
 In the Amsler-Laffon instrument (Fig. 213) a cylindrical steel center punch, 
 plane above, ground to a right-angled 
 cone below, is vertically mounted in a 
 casing of bronze, in which it is free to 
 turn ; it is balanced by a lateral spring. 
 The displacement of the cone (with 
 regard to the casing) is transferred to 
 a pointer by an elastic threaded bush- 
 ing and a toothed wheel; the pointer 
 allows of easily reading depths up to 5 
 millimeters within .01 millimeter. The 
 pointer is accurately adjusted by turn- 
 ing it with the aid of the milled edge 
 of the bushing, in case the top of the 
 specimen should not be perfectly plane. 
 The cone can easily be exchanged and 
 be reground. The whole instrument 
 
 & Fig. 213. — Amsler-Laffon Instrument for 
 
 Weighs .7 kilogram (1| pounds), and Measuring Hardness. 
 
 its height, from the upper pressure plate to the surface of the specimen, is 
 about 10 centimeters (4 inches). 
 
 The question, whether the hardness numbers of a material, obtained by 
 these methods, admit of any general conclusions respecting the strength of 
 the material, and in particular the yield point and the tensile strength, is of 
 high practical interest. 
 
 A direct constant relation between yield point and tensile strength on the 
 one side, and hardness on the other, can not exist, since that relation would, 
 among other factors, depend upon the shape of the impression and of the stress- 
 strain diagram. 
 
 This admission does not, however, at all exclude the possibility of deducing 
 from the hardness number, with the aid of a coefficient which will only hold 
 
302 STEEL RAILS 
 
 good between certain known limits, the yield point and the tensile strength with 
 an approximation which will frequently be sufficient for practical purposes.* 
 
 f Permanent-way materials have been tested by Ludwik's cone-pressure 
 method with two objects: to inquire into the suitability of the method for practi- 
 cal purposes, and to ascertain the relation between the cone-pressure hardness 
 and the tensile strength. 
 
 The experiments have been conducted in connection with the acceptance 
 tests of the materials supplied to the J. R. Austrian State Railways during the 
 year 1908, in the Trzynietz Iron Works of the Osterr. Berg- und Hiittenwerks- 
 Gesellschaft with the aid of an Amsler-Laffon cone-pressure hardness tester J 
 on a Mohr and Federhaff testing machine. 
 
 The material experimented upon comprised rails, railway ties, splices, and 
 steel crossings. 
 
 The specimens were not prepared in any way apart from being cleaned ; an 
 exception was made in the case of the steel points, in which the outer skin con- 
 taining coarse impurities had to be removed completely. 
 
 The following are the chief results : 
 
 The ratio of tensile strength to cone-pressure hardness had a mean value 
 of about .335, the range of deviation being ±6 per cent. 
 
 The lowest tensile strength of 65 kg./mm. 2 (42 tons per square inch), ad- 
 missible for rails, corresponded to a cone-pressure hardness of about 190. 
 
 The tensile strengths of ties and of smaller parts for the permanent way 
 varied between 39 and 47 kg./mm. 2 (24.75 and 29.8 tons per square inch) and 
 the corresponding hardness numbers between 117 and 144. The range of 
 variation is hence approximately the same for the tensile strengths as for the 
 hardness numbers. 
 
 Other methods of testing hardness have been used. The sclerometer of 
 Turner makes use of a diamond point which is drawn across the surface to be 
 measured. The weight required upon this point to make a barely visible scratch 
 determines the degree of hardness. This machine is sometimes used with a 
 series of standard weights and the width of a scratch made by one of these 
 
 * For instance, the Prussian Railway Department stipulates for rails of a minimum tensile strength 
 of 60 kg. /mm. 2 (38 tons per square inch), with balls of 19 mm. (finch) diameter and 50 tons loads, impres- 
 sion depths of from 3.5 to 5.5 mm. ; for rails of a minimum tensile strength of 70 kg./mm. 2 (44.5 tons per 
 square inch), impression depths ranging from 3 to 5 mm. Breaking tests and ball tests have to be 
 made in equal numbers. (Zentralblatt der Bauver-waltung, 1908, No. 77, p. 520.) 
 
 t The Cone-pressure Test for Determining the Hardness of Permanent-way Materials, by Dr. 
 techn. August Gefener, Vienna. International Association for Testing Materials, 5th Congress, Copen- 
 hagen, 1909. McGraw-Hill Book Company, New York. 
 
 t Compare P. Ludwik, " Die Kegelprobe, ein neues Verfahren zur Hartebestimmung von Materi- 
 alien." Berlin, 1908, Julius Springer. 
 
STRENGTH OF THE RAIL 303 
 
 measured under a microscope. The Keep test employs an instrument which 
 drills into the specimen and gives a measure of the work required to cut out 
 the metal, thus testing not only the surface, but also the interior. The Jaggar 
 instrument is similar, but uses a small diamond drill in connection with a micro- 
 scope. 
 
 Resistance to penetration was long tested by the United States Ordnance 
 Department by means of a weighted punch, and a somewhat similar result was 
 obtained by means of a series of needles of graded hardness, which were tried 
 in succession until one was found that would scratch the material under test. 
 
 While many inconsistencies are found in hardness tests, it is generally con- 
 ceded that the test affords an excellent comparison of metals of the same general 
 composition and treatment, and the results thus far seem to justify the expecta- 
 tion that it will in many cases be possible and advantageous to employ this 
 method in the testing of rails in place of the more elaborate and expensive tensile 
 strength tests which some foreign engineers require in addition to the drop test. 
 
 The magnetic laboratory of the Bureau of Standards is carrying on an investi- 
 gation on the relation between the magnetic and mechanical properties of steels. 
 The reluctance, or the ratio of the magnetizing force to the magnetic induction, of a 
 rail is greatly affected by changes in homogeneity, such as may be caused by segre- 
 gation, blowholes, or strains due to any cause whatever. By means of the mag- 
 netic data taken along the length of a rail it is possible to detect the presence of 
 these defects. 
 
 Special machines have been devised from time to time for testing different 
 properties of the rail, as the machine for testing rail wear illustrated in Fig. 151. 
 
 * The Pennsylvania Steel Company has a machine (Fig. 214) for testing 
 rail wear which enables specimens of rails to undergo wear similar to that im- 
 posed upon them by every description of traffic, but in a much shorter time than 
 if tried in the ordinary road. The rails are fixed to a 20-foot diameter circular 
 frame, three specimens being included in the circle. Two standard 33-incl 
 wheels having independent axles fixed at each end of a heavy casting, which is 
 pivoted at the center of the circle, run upon the rails at speeds up to 85 revolu 
 tions per minute, equivalent to a train speed of about 60 miles per hour, ^hert 
 are devices by which, through means of springs, pressure is exerted vertically and 
 also centrifugally on the rail, so that the action of the load can be imitated, as 
 well as that of its lateral pressure on the rail, and the effect produced by continu- 
 ous wear on the rails of different manufacture and composition can be estimated 
 in a comparatively short time. 
 
 * Railway and Engineering Review, Chicago, 1908, Vol. XI VIII, p. 868. 
 
304 
 
 STEEL RAILS 
 
 Fig. 214. — Machine for Testing Rail Wear at Pennsylvania Steel Company. 
 
 Extensive tests have been 
 made of the tensile strength of 
 the steel in the rail by Mr. M. H. 
 Wickhorst, Engineer of Tests 
 of the Rail Committee of the 
 American Railway Engineering 
 Association, covering rails rolled 
 at Gary and at the Lackawanna 
 Steel Company. 
 
 The rails from the Gary 
 works were open-hearth steel 
 and 100-pound, A. R. A., Type 
 B section. The ingots furnished 
 six rails, which were lettered 
 A, B, C, D, E, F, the A rails 
 coming from the top of the 
 ingot, etc. 
 
 Tensile tests were made of 
 pieces |-inch diameter by 2-inch 
 gauge length, cut from near the 
 top end of each rail, as shown 
 in Fig. 215. Five locations in 
 
 - Diagram of Round Test Pieces; Tensile Tests on 
 
 Rail steel. (Wickhorst.) the sections were selected as 
 
STRENGTH OF THE RAIL 305 
 
 shown, and tests were made in duplicate, the bar being cut sufficiently long to 
 make two test pieces. The yield point was determined by means of a Capp's 
 multiplying divider, which method gives a result somewhat above the elastic 
 limit, but which, however, is probably sufficiently definite to make it desirable 
 to determine it, and is not subject to the irregularities of the yield point as 
 determined by the drop of the beam of the test machine, or even by ordinary 
 dividers. 
 
 The test pieces were very close to J-inch diameter at the center, but toward 
 the ends of the gauge length most of them were from .002 to .004 inch larger in 
 diameter. This would tend to make the elongation less than if the diameter 
 were perfectly uniform. 
 
 The results of the tensile tests are shown in Table LXXII. 
 
 The duplicates agree well with each other except in a few cases where the 
 test pieces broke " short " as follows: One sample from base of the A rail, one 
 sample from the interior of the head of the B rail, and one sample from the 
 corner of the head of the C rail. One sample from the web of the A rail should 
 probably also be classed here. The duplicates from the lower rails of the ingot 
 agree particularly well, indicating a freedom from local irregularities. The 
 ratio of the yield point to the tensile strength averages about 51 per cent, and 
 most specimens differ but little from this figure. A comparison of the tensile 
 strength is interesting. Table LXXIII shows the tensile strength of the 
 sample in each pair that gave the greatest ductility. 
 
 The b samples from the interior of the head and the c sample from the 
 web represent what was originally the interior of the ingot, and in the A rail 
 these samples show strengths much higher than the samples from the other 
 parts of the section representing what was the outer part of the ingot. This is 
 also true of the B rail, but to a less extent, and also of the C rail to a slight extent. 
 As we continue down the ingot, however, conditions are reversed, and we find 
 in the D rail a little lower strength in the samples from the interior than in the 
 samples representing what was the outer portion of the ingot. This difference 
 is greater in the E rail and greatest in the F rail. 
 
 The a sample from the corner of the head and the d sample from the 
 flange would be metal of similar chemistry, but the flange has a considerably 
 lower finishing temperature and is also reduced differently. Table LXXIV 
 compares results from these two places. 
 
 The d sample from the flange of the A rail is abnormal, with a low strength 
 and high ductility, being evidently taken at a point of negative segregation of 
 carbon; but, except for this sample, the d samples from the flange show a little 
 
306 
 
 STEEL RAILS 
 
 TABLE LXXIL — TESTS ON STRENGTH OF RAIL STEEL 
 
 Tensile Tests on Open Hearth Rail Steel from Gary (Wickhorst) 
 
 Tensile Strength. 
 
 a-head, corner . . . 
 6-head, interior. 
 
 c-web 
 
 d-flange 
 
 e-base 
 
 a-head, corner. . . 
 6-head, interior. . 
 
 a-head, corner. . . 
 6-head, interior. . 
 
 c-web 
 
 d-flange 
 
 e-base 
 
 a-head, corner . . . 
 6-head, interior . . 
 
 c-web 
 
 d-flange 
 
 e-base 
 
 a-head, corner. . . 
 6-head, interior. . 
 
 c-web 
 
 d-flange . : 
 
 2-base 
 
 j-head, corner. . . 
 6-head, interior. . 
 
 c-web 
 
 d-flange 
 
 Pounds per Square 
 
 63,930 
 61,640 
 69,510 
 70,020 
 
 "73,060 ' 
 
 56,820 
 63,420 
 65,690 
 66,970 
 68,220 
 70,310 
 70,010 
 66,970 
 65,690 
 64,940 
 63,930 
 67,710 
 65,690 
 66,970 
 69,230 
 66,970 
 66,230 
 66,490 
 67,480 
 64,170 
 64,170 
 67,990 
 68,720 
 64,940 
 65,930 
 66,930 
 66,970 
 70,020 
 70,460 
 65,690 
 69,290 
 66,740 
 
 "63,920 " 
 64,180 
 64,940 
 
 69,510 
 65,210 
 63,420 
 60,900 
 66,500 
 64,680 
 64,180 
 70,020 
 67,710 
 63,670 
 68,950 
 
 124,800 
 125,800 
 134,400 
 133,900 
 134,900 
 141,000 
 121,700 
 123,300 
 109,600 
 120,200 
 128,800 
 128,300 
 
 135,900 
 137,000 
 137,000 
 132,900 
 129,800 
 129,400 
 126,800 
 130,900 
 
 132,900 
 
 133,400 
 
 133,900 
 
 133,900 
 
 132,900 
 
 133,400 
 
 128,300 
 
 128,800 
 
 132,400 
 
 130,900 
 
 129,300 
 
 128,300 
 
 130,800 . 
 
 131,900 
 
 133,900 
 
 134,400 
 
 130,900 
 
 132,400 
 
 127,900 
 
 125,300 
 
 124,800 
 
 124,300 
 
 128,900 
 
 128,900 
 
 131,800 
 
 132,400 
 
 129,800 
 
 130,400 
 
 128,400 
 
 128,300 
 
 120,800 
 
 123,700 
 
 125,300 
 
 125,800 
 
 132,400 
 
 131,900 
 
 127,300 
 
 128,300 
 
 11.5 
 12.5 
 12.5 
 10.5 
 
 13 
 
 11.5 
 
 12.5 
 
STRENGTH OF THE RAIL 
 
 TABLE LXXIII.— TESTS ON STRENGTH OF RAIL STEEL 
 
 Comparison of Strength in Different Parts of Open Hearth Rails from Gary (Wickhorst) 
 
 307 
 
 Rail. 
 
 a-Head, Corner. 
 
 6-Head, Interior. 
 
 c-Web. 
 
 d-Flange. 
 
 e-Base. 
 
 A 
 
 B 
 
 124,800 
 128,800 
 130,900 
 132,400 
 127,900 
 128,300 
 
 133,900 
 
 135,900 
 133,400 
 128,300 
 124,300 
 120,800 
 
 141,000 
 137,000 
 133,900 
 131,900 
 128,900 
 125,300 
 
 121,700 
 129,800 
 133,400 
 133,900 
 131,800 
 131,900 
 
 120,200 
 129,400 
 128,800 
 132,400 
 129,800 
 128,300 
 
 C 
 
 D 
 
 E 
 
 F 
 
 Average 
 
 128,800 
 
 129,400 
 
 133,000 
 
 130,400 
 
 128,200 
 
 TABLE LXXIV. — TESTS ON STRENGTH OF RAIL STEEL 
 
 Comparison of Strength and Ductility of Steel taken from the Corner of the Head and Flange of Open Hearth Rails (Wickhorst) 
 
 Rail. 
 
 Tensile Strength pounds per 
 Square Inch. 
 
 Elongation 
 
 in 2 Inches. 
 
 Reduction of Area. 
 
 
 « 
 
 d 
 
 a 
 
 d 
 
 « 
 
 d 
 
 A 
 
 B 
 
 124,800 
 128,800 
 130,900 
 132,400 
 127,900 
 128,300 
 
 121,700 
 129,800 
 133,400 
 133,900 
 131,800 
 131,900 
 
 Per cent. 
 
 10 
 
 11.5 
 
 11.5 
 
 12 
 
 11 
 
 12 
 
 14 
 
 12 
 
 12.5 
 
 13 
 
 12 
 
 12.5 
 
 15 
 20 
 22 
 19 
 19 
 20 
 
 Per cent. 
 26 
 
 23 
 
 C 
 
 D 
 
 E 
 
 F 
 
 27 
 28 
 26 
 21 
 
 Average B to F . . 
 
 129,700 
 
 132,200 
 
 11.6 
 
 12.4 
 
 20 
 
 25 
 
 higher strength and also a little greater ductility. As the difference in the 
 work of rolling is perhaps sufficient to account for this, the conclusion seems to 
 be that the difference in rolling temperature has not resulted in any important 
 difference in the tensile properties. 
 
 The b sample from the interior of the head and the c sample from the web 
 would be of similar chemistry, as representing the interior of the ingot, but the 
 web is thinner and gets more work in rolling and probably is finished at a lower 
 temperature. A comparison of these two locations is shown in Table LXXV. 
 
 TABLE LXXV. — TESTS ON STRENGTH ON RAIL STEEL 
 
 Comparison of Strength and Ductility of Steel taken from the Interior of the Head and Web of Open Hearth Rails (Wickhorst) 
 
 Rail. 
 
 Tensile Strength Pounds per 
 Square Inch. 
 
 Elongation in 2 Inches. 
 
 Reductio 
 
 r of Area. 
 
 
 b 
 
 • 
 
 5 
 
 • 
 
 6 
 
 « 
 
 A 
 
 133,900 
 135,900 
 133,400 
 128,300 
 124,300 
 120,800 
 
 141.000 
 137,Q0O 
 133,900 
 131,900 
 128,900 
 125,300 
 
 Per cent. 
 
 9 
 
 9 
 10 
 11 
 11 
 12 
 
 Per cent. 
 
 9 
 11 
 
 11.5 
 11.5 
 12.5 
 13 
 
 Per cent. 
 16 
 16 
 17 
 21 
 
 21 
 
 22 
 
 Per cent. 
 10 
 
 B 
 
 19 
 
 c 
 
 21 
 
 D 
 
 23 
 
 E 
 
 25 
 
 F 
 
 25 
 
 
 
 Average B to F . . 
 
 128,500 
 
 131,400 
 
 10.6 
 
 11.9 
 
 19.4 
 
 22.6 
 
308 
 
 STEEL RAILS 
 
 In the A rail it is probable that the carbon is higher in the web sample than 
 in the head sample, but in the other rails there probably are no great differences, 
 and the averages shown above are of the B to F rails inclusive. The tensile 
 strength decreases as we go down the ingot and the ductility increases. The 
 web samples, as compared with the head samples, show a little greater strength, 
 an average of 131,400 pounds, as against 128,500 pounds, and also a little greater 
 ductility, an average elongation of 11.9 per cent, as against 10.6 per cent, and 
 a reduction of area of 22.6 per cent, as against 19.4 per cent. This difference, 
 
 it would seem, is probably due to 
 the increased work in rolling that 
 the web gets. It is also interest- 
 ing to note that the top rails show 
 as good ductility in the head 
 samples as the lower rails, allow- 
 ing, of course, for the difference in 
 tensile strength, which would 
 make about 3 per cent decrease 
 in elongation for an increase in 
 tensile strength of 10,000 pounds. 
 The tensile tests on the rail 
 steel from the Lackawanna mill 
 were from titanium Bessemer 
 rails, 90-pound, A. R. A., Type B. 
 It was planned to make ten- 
 sile tests of pieces from near the 
 top ends of each rail by cutting 
 
 Fig. 216. — Test Pieces 16 inches long. Diagram of Flat five flat pieces f by 1 \ by 16 inches 
 Test Pieces. Tensile Tests on Rail Steel. (Wickhorst.) f rom ^^ ra ^ ag s h own i n Fig. 
 
 216. The test pieces were cut in this manner from the A and D rails, but 
 it soon appeared that the time required to prepare the pieces in this manner 
 would cause considerable delays, and then, too, the surfaces are apt to be finished 
 in a condition unsatisfactory for test. The plan was then changed so as to 
 obtain pieces \ inch in diameter and 2-inch gauge length, as shown by Fig. 215. 
 The pieces from the B and C rails were prepared in this manner. The results 
 of the tensile tests are shown in Table LXXVI. Care must be taken not to 
 compare the results of the flat test pieces with those from the round test pieces, 
 except, perhaps, as regards the tensile strength, although even here the shape 
 of the test piece would make some difference. 
 
STRENGTH OF THE RAIL 
 
 309 
 
 TABLE LXXVI.— TESTS ON STRENGTH OF RAIL STEEL 
 
 Tensile Tests on Titanium Bessemer Rail Steel from Lackawanna (Wickhorst) 
 
 a-head . . 
 6-head . . 
 c-web . . . 
 rf-fiange. 
 e-base . . 
 a-head. . 
 6-head . . 
 c-web . . . 
 d-flange 
 e-base . . 
 a-head . . 
 6-head. . 
 c-web . . . 
 d-flange . 
 e-base. . 
 a-head . . 
 6-head . . 
 c-web . . 
 d-flange . 
 e-base. . 
 
 Tens: 
 
 .■ Strom; I h 
 
 Per Square Inc 
 103,400 
 
 111,100 
 109,200 
 103,300 
 105,000 
 111,700 
 122,900 
 116,200 
 110,700 
 111,600 
 109,100 
 112,100 
 112,300 
 109,600 
 112,400 
 109,800 
 106,800 
 104,700 
 108,600 
 110,000 
 
 14.5 ' 
 11.7 ' 
 13.7 ' 
 19.0 ' 
 14.0 ' 
 17.5 ' 
 19.0 ' 
 18.0 ' 
 20.0 ' 
 16.0 ' 
 21.0 ' 
 20.0 ' 
 19.5 ' 
 12.5 
 10.0 ' 
 13.7 ' 
 10.0 ' 
 9.5 ' 
 
 The following is a re- 
 cord of tensile tests made 
 by Waterhouse* on 100- 
 pound A. S. C. E. rail taken 
 from stock. The steel was 
 made by the acid Bessemer 
 process and the ingots rolled 
 without reheating. From 
 this rail two adjacent pieces 
 8| inches long were cut with 
 a slow-speed cold saw, and 
 from these pieces tensile- 
 test bars were machined in 
 the positions shown in Fig. 
 217. The results of the 
 tensile tests are given in the 
 following table, the figures 
 being the average of those 
 obtained from the duplicate 
 pieces: 
 
 * Examination of 100-pound Rails, 
 p. 478. 
 
 Fig. 217. — Location and Numbers of Test Pieces used in 
 Waterhouse's Tests. (Railroad Age Gazette.) 
 
 G. B. Waterhouse. Railroad Age Gazette, July 10, 1908, 
 
310 
 
 STEEL RAILS 
 
 TABLE LXXVIL— TESTS ON STRENGTH OF RAIL STEEL 
 
 Tensile Tests on Bessemer Rail Steel (Waterhouse) 
 
 
 Elastic Limit. 
 
 Ultimate Stress. 
 
 Elongation in 
 2 Inches. 
 
 Reduction of 
 
 1 
 
 Pounds. 
 
 52,200 
 52,200 
 54,460 
 55,000 
 53,100 
 53,820 
 51,740 
 53,340 
 
 Pounds. 
 
 108,400 
 109,850 
 110,750 
 110,150 
 110,300 
 110,400 
 109,850 
 111,300 
 
 Per cent. 
 16.75 
 16.25 
 18.50 
 18.50 
 18.25 
 18.00 
 18.25 
 17.00 
 
 Per cent. 
 29.9 
 28.4 
 33.2 
 28.6 
 29.4 
 31.0 
 35.4 
 36.4 
 
 2 . 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 Table LXXVIII shows the chemical composition of the rails tested in the 
 three tests just mentioned. 
 
 TABLE LXXVIII.— CHEMICAL COMPOSITION OF RAIL STEEL 
 IN TENSILE TESTS OF WICKHORST AND WATERHOUSE 
 
 Carbon 
 
 Silicon 
 
 Manganese . . 
 
 Sulphur 
 
 Phosphorus . . 
 Copper 
 
 
 Per cent. 
 
 0.72 
 0.20 
 0.72 
 0.036 
 0.036 
 
 Per cent. 
 0.51 
 0.147 
 0.77 
 0.078 
 0.089 
 0.185 
 
 28. The Strength of the Rail and Proper Weight for Various 
 Conditions of Loading 
 
 In determining the proper stress to use for the rail, careful consideration 
 must be given to the exact meaning of the terms by which the strength of the 
 steel is shown. 
 
 * The elastic limit or yield point may be properly called the limit of pro- 
 portionality of stress to deformation, or more briefly, the limit of proportion- 
 ality. The limit of proportionality is sometimes called the " true " elastic 
 limit, and is frequently regarded as a measure of the load-carrying capacity of 
 a member or structure. 
 
 The absolute value of this limit cannot, in general, be determined even by 
 the most careful measurements of deformation and load. It has been the 
 experience of experimenters that any additional refinement of measurements 
 in stress-deformation tests results in the detection of the limit of proportionality 
 
 * Proceedings American Society for Testing Materials, 1910, Vol. X, pp. 243, 244. Moore. 
 
STRENGTH OF THE RAIL 
 
 311 
 
 at a stress lower than that determined by the less refined methods of measure- 
 ment. Perhaps the thermo-electric apparatus used by Turner * and Rasch f for 
 measuring deformation is the most delicate yet employed, and both of these 
 experimenters showed the existence of a limit of proportionality at stresses 
 far lower than those determined by extensometer measurements as ordinarily 
 made. 
 
 Members and structures become more and more nearly perfectly elastic 
 if subjected to repeated stresses in the same direction, even if these stresses 
 are so far beyond the limit of proportionality 
 that there is a small but well-defined per- 
 manent set upon release of the load. Fig. 218 
 shows the result of loading a beam several 
 times to a stress well beyond the elastic limit. 
 The first load applied to the beam produced 
 considerable permanent set, and during the 
 application and release of load considerable 
 energy was lost, presumably in heat. This 
 energy is shown by the shaded area to the 
 left of the figure. During the second cycle 
 of loading and release much less energy was 
 lost, as shown by the central shaded area; and 
 during the third cycle the beam behaved as if 
 almost perfectly elastic. It should be noted 
 that the above results would not be obtained 
 if the direction of the load were reversed. 
 
 The ultimate strength, or maximum capacity for resisting stress, has a 
 ratio to the maximum stress due to the working load, which, although less in 
 metal than in wooden or stone structures, is, nevertheless, made of considerable 
 magnitude in many cases. It is much greater under moving than under steady 
 " dead " loads, and varies with the character of the material used. For 
 machinery it is usually 6 or 8; for structures erected by the civil engineer, 
 from 5 to 6.| 
 
 In general, parts of structures are so proportioned as to carry their loads 
 without risk of exceeding their elastic limits; and in such cases the factor of 
 safety should probably always be referred to the elastic limit. 
 
 * Thermo Electric Measurements of Stress, Transactions American Society of Civil Engineers, 
 Vol. XXVIII, January, 1902, p. 27. 
 
 t Proceedings International Association for Testing Materials, No. 11, August 4, 1909, Article VII. 
 t Iron and Steel, Materials of Engineering, Thurston, 1909, p. 340. 
 
 One Division = 0.1 in. Deflection 
 Fig. 218. — Effect of Repeated Loads 
 on Beams. (Am. Soc. for Testing 
 Materials — Moore.) 
 
312 STEEL RAILS 
 
 The elastic limit is made the basis of estimates by nearly all French engineers, 
 while the ultimate strength is taken by German engineers, using a factor of 
 safety of larger magnitude. British and American engineers usually base all 
 calculations on the ultimate strength, although the former system is extending 
 in general practice, and the limit of working load is made to fall well within the 
 limit of elasticity. 
 
 The general practice at the present time, for railway and highway bridges, is 
 to use a unit strain of about one-half the elastic limit of the material. This factor 
 is considered correct in places where the load assumed is an absolute maximum, 
 as, for instance, where it consists of a definitely determined dead load only. 
 
 In the rail the maximum stress acts during a very short space of time and 
 its effect is not the same as the same load applied for a longer period. It is 
 possible to apply a much greater stress than the elastic limit of the metal, pro- 
 vided the stress be applied very quickly and then released. 
 
 The excellent service given by some of the rails of lighter section, exposed 
 to heavy wheel loads, gives evidence that a limited number of excessive stresses 
 in the rail will not cause injury when applied for the small fraction of a second, 
 as is the case of the stress caused by the wheel load. 
 
 An extreme fiber stress of 20,000 pounds per square inch as applied to 
 the base of the rail probably represents a satisfactory mean between the danger, 
 on the one hand, of not providing a sufficient margin of safety for the unknown 
 quantities of the problem and the liability, on the other hand, of taking too 
 great a factor of safety, and thus designing an uneconomical structure. 
 
 The following remarks of Professor Unwin are pertinent to this question : 
 
 " If an engineer builds a structure which breaks, that is a mischief, but one of a limited and 
 isolated kind, and the accident itself forces him to avoid a repetition of the blunder. But an engineer 
 who from deficiency of scientific knowledge builds structures which don't break down, but which 
 stand, and in which the material is clumsily wasted, commits blunders of a most insidious kind." 
 
 Any consideration of the strength of the rail should take account not only 
 of the stresses in the rail itself and the ability of the material of which it is com- 
 posed to resist them, but a proper proportion must be made of the rail in order 
 that it may distribute the wheel load to the ties in such a manner as not to over- 
 strain any part of the track structure. The fact that a rail will not break should 
 not be the determining factor in its selection, if, on account of lack of stiffness, 
 it will allow too great a proportion of the load to come on a tie. 
 
 The damaging effects of overloading the track, while much less apparent 
 than the results attending the use of too great loads in other engineering struc- 
 tures, are, nevertheless, of very real importance, and the lack of proper appre- 
 ciation of the fundamental principles underlying its design has brought about 
 
STRENGTH OF THE RAIL 313 
 
 conditions requiring excessive maintenance charges to keep the track in proper 
 condition. 
 
 As the real function of the heavier rails is to distribute the wheel load and 
 prevent too great a concentration of pressure on the track substructure, we 
 have two limiting conditions to consider: First, the rail should be stiff enough 
 to enable it to transfer the load in such a manner as not to exceed the maximum 
 bearing power of the track substructure, and second, the safe working stress 
 of the metal in the rail must not be exceeded. 
 
 Before investigating the proper weight of rail to use with any track struc- 
 ture, the weight and types of the locomotives and cars to be run over it should 
 be considered, and the maximum wheel pressures ascertained for each type. 
 
 The bearing power of the roadway or subgrade should next be examined. 
 The influence of the character of the roadway is well shown by the follow- 
 ing case reported by Mr. A. G. Wells, General Manager of the Atchison, Topeka 
 and Santa Fe:* 
 
 " From Seligman to Barstow our track is laid with eighty-five-pound rails; the density of 
 the traffic is practically the same over every foot of it. Between Yucca and Barstow, a distance 
 of 227 miles, the subgrade is sandy, porous, and well drained; between Yucca and Seligman, a 
 distance of 91 miles, the subgrade is largely clay, of a kind that holds water. From November, 
 1907, to October, 1908, we had eighty-three rail breakages on the territory first named, or a per- 
 centage of .001; on the other stretch we had in the same period seventy-two breakages, the per- 
 centage being .0025, or, in other words, where the subgrade was dense and more or less clay, the 
 s per mile were two and one-half times greater than where the subgrade was sandy." 
 
 The bearing power of the subgrade is such an important factor in pro- 
 portioning the track that it will prove profitable to examine what takes place 
 when the soil is subjected to pressure.! 
 
 As in any structure, good judgment must enter into the design; the formulas 
 which will be demonstrated must be used as guides only. These formulas will 
 depend upon the angle of repose of a homogeneous granular mass. For 
 ordinary earths for which the angles of repose are known, the results obtained 
 by the use of the formulas will compare very favorably with those obtained 
 from examples of the best practice. 
 
 When the angle of repose is not known it should be determined by test.f 
 
 * Railroad Age Gazette, April 9, 1909. 
 
 f The following discussion is based upon Retaining Walls for Earth, Howe, New York, 1896. 
 t This can conveniently be done by measuring the force required to cause slipping of two por- 
 tions of the earth past each other when subjected to a known pressure, and 
 
 <j> = angle of repose. 
 
 s = force required to cause slipping. 
 
 p = pressure on earth. 
 
314 
 
 STEEL RAILS 
 
 Earth which has an angle of repose of at least 27 degrees may be considered 
 as firm. From Table LXXIX it is seen that sand, gravel, and damp clay are 
 classed as firm soils; however, these may become so saturated with water that 
 their angles of repose will become considerably less than 27 degrees, hence 
 precaution must be taken against too much water by draining the ground in the 
 immediate vicinity of the roadbed. Particular care must be taken in the case 
 of clay, or sand which will become a kind of quicksand when saturated with 
 water. 
 
 The water which destroys the bearing power of such soils may come from 
 below by capillary attraction,* and the drainage should be carried to a depth 
 sufficient to prevent this. Semi-fluid soils, such as quicksand, alluvium, etc., 
 should be removed where practicable or the foundation carried to a lower 
 stratum. 
 
 TABLE LXXIX. — ANGLES AND COEFFICIENTS OF FRICTION 
 
 
 (Rankine's Applied Mechanics.) 
 
 
 
 
 tan*. 
 
 <f>. 
 
 sh 
 
 Earth on earth 
 
 Earth on earth, dry sand, clay, and 
 
 Tilxed earth. . . 
 
 0.25 to 1.0 
 0.38 to 0.75 
 
 1.0 
 
 0.31 
 
 0.81 
 
 Degrees. 
 14 to 45 
 21 to 37 
 
 45 
 
 17 
 39 to 48 
 
 4 to 1 
 2.63 to 1.33 
 1 
 
 Earth on earth, wet clay 
 
 Earth on earth, shingle, and gravel 
 
 3.23 
 1.23 toO. 9 
 
 Let Fig. 219 represent a section of the track, and 
 
 BALLAST 
 
 WW 
 
 7 r 1 
 
 S7777777777777777T. 
 
 SUBGRADE 
 
 Fig. 219. — Resistance of Sub-grade to 
 
 = the depth of ballast; 
 
 WVM//////MI* 
 
 e of the Track. 
 
 p = the maximum supporting power per square foot of thesubgrade; 
 p l the pressure exerted on subgrade midway between ties; 
 7 = the weight of one cubic foot of ballast; 
 (f> = the angle of repose of subgrade; 
 xj equals the vertical intensity of the pressure caused by the weight of the 
 
 * Movements of Ground Water, by F. H. King and C. S. Slichter, Government Printing Office, 
 Washington, D. C, 1899, p. 65. 
 
STRENGTH OF THE RAIL 
 
 315 
 
 ballast on the subgrade midway between the ties. This pressure is augmented 
 by the pressure transmitted from the tie, and, while this is much less between 
 the ties than immediately underneath a tie, it is, nevertheless, an important 
 factor in strengthening the surface of the roadbed. 
 
 If we assume this extra pressure on the roadbed midway between the ties 
 to equal in amount yx we will probably be on the safe side and can then write 
 Pi = 2 yx* 
 
 Now when the ballast is about to sink 
 
 v 1 + sin <t> 1 — sin 6 
 
 - = =-^ — ; — - or q = v \ : — - • 
 
 q 1 - sin (j) 1 + sm 
 
 But when the roadbed under the tie is on the point of sinking, the part of 
 
 the roadbed between the ties must be on the point of rising, or 
 
 (I _ 1 + sin 4> 
 
 Vx ~ 1 - sin 4> ' 
 
 and the supporting power of the subgrade, or 
 
 P=Pi 
 
 For convenience the values 
 > in Table LXXXI. 
 
 \ 1 + sin 4>\ 
 
 1 — sin <i>S 
 1 + sin 4> ) 2 
 1 — sin <j>S 
 
 1 - sin 4>) 
 are given in Table LXXX and for 
 
 TABLE LXXX 
 
 l+sin<* >) 
 sin<£ ) 
 
 Values of 
 
 (How 
 
 1 
 
 ( 1 + sin 4, 1 \ 
 
 * 
 
 j l + sin*)* 
 
 \ 1 - sin <)> i 
 
 1 1 - gin 4> 1 ' 
 
 
 
 1.00 
 
 23 
 
 5.21 
 
 5 
 
 1.42 
 
 24 
 
 5.62 
 
 6 
 
 1.52 
 
 25 
 
 6.07 
 
 7 
 
 1.63 
 
 26 
 
 6.56 
 
 8 
 
 1.75 
 
 27 
 
 7.09 
 
 9 
 
 1.88 
 
 28 
 
 7.67 
 
 10 
 
 2.02 
 
 29 
 
 8.30 
 
 11 
 
 2.16 
 
 30 
 
 9.00 
 
 12 
 
 2.32 
 
 31 
 
 9.76 
 
 13 
 
 2.50 
 
 32 
 
 10.59 
 
 14 
 
 2.68 
 
 33 
 
 11.50 
 
 15 
 
 2.88 
 
 34 
 
 12.51 
 
 16 
 
 3.10 
 
 35 
 
 13.62 
 
 17 
 
 3.33 
 
 36 
 
 14.84 
 
 18 
 
 3.59 
 
 37 
 
 16.18 
 
 19 
 
 3.86 
 
 38 
 
 17.67 
 
 20 
 
 4.22 
 
 39 
 
 19.64 
 
 21 
 
 4.48 
 
 40 
 
 21.16 
 
 22 
 
 4.83 
 
 
 
 * This apparently would be a safe assumption for a depth of gravel ballast under the tie of 18 
 inches and 12 inches of stone. For a less depth of ballast the pressure would be less and for a greater 
 depth the pressure would increase, the increase being more rapid in the case of the stone than of the 
 gravel ballast. 
 
316 
 
 STEEL RAILS 
 TABLE LXXXI. —VALUES OF 7 
 
 Name of Ballast. 
 
 Average Weight, in 
 
 Pounds per Cubic 
 
 Foot. 
 
 
 90 to 106 
 90 to 106 
 118 to 129 
 90 to 108 
 
 
 
 Stone, crushed 
 
 Considering first the weight of rail which will give a proper distribution of 
 pressure to the track, we may adopt a tentative system of classification for the 
 track structure based upon the kind of tie, tie spacing, depth and kind of 
 ballast, and character of subgrade. As previously noted, a tie spacing of 20 
 inches with 18 inches of gravel or 12 inches of stone under the tie, resting on a 
 roadbed capable of bearing 1.5 tons per square foot, will sustain safely a load of 
 700 pounds per linear inch under each rail. This grade of track we will desig- 
 nate as class A. 
 
 Class B and C will represent weaker structures, which may be brought 
 about by a departure in any or all of the elements from those found in class A 
 track. 
 
 A track would be graded as class B if it was capable of carrying only 600 
 pounds per linear inch under each rail. This might occur in several ways. 
 A tie spacing of 22 inches, but with all the other elements of class A track, would 
 diminish the strength of the structure on account of greater concentration of 
 the load on each tie and on the subgrade; similarly, a lesser depth of ballast 
 would affect the load on the subgrade. Evidently, a track with all the other 
 elements of class A, but resting on a soil having a lower bearing power, would 
 offer less resistance to the action of the wheel load. 
 
 Table LXXXII presents descriptions of different kinds of track in each of 
 these classes. It will be observed that the limiting factors may be considered 
 as being the supporting power of the roadbed and the bearing of the rail on the 
 tie. The bending stress in the tie and the bearing of the tie on the ballast are of 
 secondary importance. 
 
 Considering first the bearing under the rail. This is evidently a function 
 of the tie spacing and kind of wood of which the tie is made. For class A track, 
 capable of supporting 700 pounds per linear inch under each rail, a tie spacing 
 of 20 inches would give a load on each tie under the rail bearing of 14,000 
 pounds. This is about the limit of the strength of the woods shown for class A 
 track, and therefore excludes the use of weaker woods or greater tie spacing for 
 this class of track. 
 
STRENGTH OF THE RAIL 
 
 317 
 
 In class B track, having a supporting power of 600 pounds per linear inch 
 under each rail, the tie spacing may be increased to 22 inches for the woods 
 allowed in class A track. It is doubtful whether the group of woods shown in 
 the table under class C track should be used for class B track even with a tie 
 spacing of 20 inches. 
 
 TABLE LXXXII. — CLASSES OF TRACK 
 
 Class of 
 
 Tie. 
 
 Ballast, Depth 
 under Tie. 
 
 Subgrade 
 Bearing Power. 
 
 Bearing Power 
 of Track 
 
 Track. 
 
 Size. 
 
 Class of Woods Represented by- 
 
 » 
 
 Stone. 
 
 Gravel. 
 
 under Each 
 Rail. 
 
 At 
 
 A 2 
 
 7X9 
 7X9 
 
 7X8 
 7X8 
 7X8 
 7X8 
 
 7X8 
 
 7X8 
 
 Oak, locust, hard maple, hickory, 
 
 cherry; not tie plated. 
 Longleaf pine, black walnut, beech, 
 
 birch, elm, gum, hemlock, Douglas 
 
 fir; tie plated. 
 
 Inches. 
 
 20 
 20 
 
 22 
 20 
 20 
 22 
 
 22 
 20 
 
 12 
 12 
 
 12 
 9 
 
 12 
 6 
 
 12 
 
 8 
 
 18 
 18 
 
 18 
 14 
 18 
 10 
 
 18 
 12 
 
 Tons per Square 
 Foot. 
 
 1.0 to 1.5 
 1.0 to 1.5 
 
 1.0 to 1.5 
 l.Oto 1.5 
 0.8 to 1.2 
 1.0 to 1.5 
 
 0.5 to 1.0 
 0.5 to 1.0 
 
 Pounds per 
 Linear Inch. 
 
 700 
 700 
 
 600 
 
 B 2 
 
 
 600 
 
 Bi 
 
 
 600 
 
 c 2 
 
 Loblolly pine, shortleaf pine (at times 
 this wood is nearly equal to longleaf), 
 soft maple, catalpa, chestnut, white 
 pine, Norway pine; tie plated. 
 
 Do . . . 
 
 450 
 450 
 
 c 3 
 
 Do 
 
 450 
 
 
 
 
 The low-supporting power of class C track permits the use of weaker 
 woods. A 22-inch tie spacing for this grade of track gives a load at the rail 
 bearing of about 10,000 pounds. 
 
 The pressure transmitted to the subgrade is determined by the spacing of 
 ■the ties (or more properly by the distance between adjacent ties) and the depth 
 and kind of ballast used. In the table three characters of subgrade are con- 
 sidered, the firmest having a bearing power of from 1.0 to 1.5 tons per square 
 foot and the least firm a bearing power of from 0.5 to 1.0 tons per square 
 foot. 
 
 By applying the formula 
 
 [l + sind>) 2 
 
 v = 2 yX 
 
 we see that the firmer grade corresponds to a soil with an angle of repose of 
 from 23 degrees to 31 degrees for 12 inches of stone ballast or 18 inches of gravel 
 ballast under the tie, and from 28 degrees to 36 degrees with 6 inches of stone 
 or 10 inches of gravel ballast under the tie. Table LXXIX shows that these 
 angles of repose fall within the limits given for dry sand, clay, and mixed 
 earth. 
 
318 STEEL RAILS 
 
 The subgrade, having a supporting power of 0.5 ton per square foot, cor- 
 responds to a soil having angles of repose varying from 13 degrees to 20 degrees 
 under the conditions stated in the table. This agrees with the angle of repose 
 for wet clay shown in Table LXXIX. 
 
 Table XL VII (article 19) gives, for a subgrade capable of bearing 1.5 tons 
 per -square foot, an allowable load applied to the tie at the rail bearing of from 
 10,100 to 11,800 pounds, for the size of tie, depth and kind of ballast used in 
 class A track. Owing to the rapid application of the load it has been assumed 
 in article 19 that these values could be safely increased to 14,000 pounds on 
 account of the inertia of the track and roadbed. For 20-inch tie spacing this 
 gives a supporting power of 700 pounds per linear inch of rail. 
 
 For the track designated by Bi in Table LXXXII, owing to the increase in 
 distance between adjacent ties this value falls to 600 pounds per linear inch of 
 rail. In class B 2 track we find from equations Nos. 1 and 2, article 18, that the 
 allowable load at the rail bearing, as determined by a supporting power of 1.5 
 tons per square foot on the roadbed, is 9500 pounds in the case of 9 inches of 
 stone under the tie and 8500 pounds for 14 inches of gravel. Making the same 
 allowance for the inertia of the roadbed, as in the previous case, it is seen that a 
 supporting power of about 600 pounds per linear inch of rail is realized. 
 
 In class B 3 track, if we take the supporting power of the roadbed as 1.2 
 tons per square foot, we find values from 7700 to 9000 pounds for the load that 
 can be applied to the tie at the rail bearing, which agree fairly well with those 
 previously determined for class B 2 track. 
 
 The bearing power under each rail, as determined by the permissible load 
 on the roadbed for class C track, has been calculated in a similar manner. 
 
 It will be noticed that in each case the upper limits of the bearing power of 
 the roadbed have been worked to in determining the values given in the table. 
 This is a feature of the analysis which requires careful consideration of the kind 
 and volume of traffic over the track. 
 
 The inertia of the roadbed plays an important part in strengthening the 
 track when the maximum loads imposed upon it do not occur too frequently, as 
 in the case of high-speed passenger trains where the most destructive forces to be 
 provided for are those produced by the drivers of the locomotive. In the case 
 of dense freight traffic where the heavy loads imposed by the engine drivers are 
 followed by the passage of a long train, thus subjecting the track to a contin- 
 uing load lasting over a considerable interval of time, the inertia of the road- 
 bed is, in a great measure, overcome and a correspondingly lower value for the 
 allowable pressure on the roadbed must be used. 
 
STRENGTH OF THE RAIL 
 
 319 
 
 The all steel 70-ton coal cars, which are coming into use on some of the 
 large coal-carrying roads in the East, weigh over 50,000 pounds, and have a 
 capacity of 140,000 pounds. This weight is carried on four axles and a train 
 composed of these cars would prove very destructive to the roadbed unless an 
 ample provision was made for the effect of the repeated application of the heavy 
 wheel loads. 
 
 Examining now the weights of rail required to distribute the wheel loads 
 so as not to exceed the bearing power of the track. Table LXXXIII shows the 
 moments of inertia of standard rails * ; evidently the rails with the highest 
 moment of inertia will give the most favorable loading of the track structure. 
 
 TABLE LXXXIII. — MOMENTS OF INERTIA AND SECTION MODULI OF STANDARD 
 RAIL SECTIONS. 
 
 Moment of In- 
 
 sertion Modulus of 
 
 A.S.C.E 
 
 A.S.C.E 
 
 A.S.C.E 
 
 A.R. A. type A 
 
 A.R. A. type A 
 
 A.R. A. type B 
 
 A.R. A. type B 
 
 P.S., Pennsylvania R. R. System. 
 P.S., Pennsylvania R. R. System. 
 
 P.R.R., Pennsylvania R. R 
 
 P.R.R., Pennsylvania R. R 
 
 Dudley, New York Central Lines 
 
 Dudley, C. R. R. of N. J 
 
 Dudley, New York Central Lines 
 
 29.1 
 38.0 
 27.4 
 49.0 
 
 16.11 
 12.00 
 9.62 
 17.78 
 12.46 
 15.74 
 11.08 
 15.91 
 12.02 
 14.29 
 11.25 
 17.00 
 11.76 
 11.53 
 
 In Plate XXVI are given the dynamic wheel loads with different axle 
 "spacing for rails having moments of inertia of 20, 30, 40, and 50. The curves 
 shown on the diagrams are calculated by the method explained in the first part 
 of article 23 and show the allowable dynamic axle loads, as determined by the 
 safe bearing power of the different classes of track given in Table LXXXII. 
 
 An examination of these diagrams shows that for each class of track the 
 allowable wheel load increases with the axle spacing up to a certain point when 
 a maximum is reached and, as the spacing of the wheels is still further increased, 
 the allowable wheel load decreases. The most favorable axle spacing, as might 
 be expected, is greater for the heavier rails than for those of lighter section. 
 
 * If the moment of inertia of the section is not known it may be calculated by one of the following 
 methods : 
 
 Culmann, C. Centralellipse und kerneines schienenprofils, 5 p. 111., 1875. (In his Die graphische 
 statik, ed. 2, p. 475.) 
 
 Morely, Arthur. Graphical determination of moments, centroids, and moments of inertia of 
 areas, 6 p. 111., 1908. (In his Strength of Materials, p. 117.) 
 
 Sankey, C. E. P. Note on the graphical determination of moments of inertia, 1000 w. dr., 
 1910. (In Engineer, London, v. 110, p. 57.) 
 
320 STEEL RAILS 
 
 Plate XXVII presents diagrams showing the maximum bending moments 
 in the rails under the conditions given in Plate XXVI. In the table shown on 
 Plate XXVII are given bending moments corresponding to an extreme fiber 
 stress of 20,000 pounds per square inch in the base of the rail. In using this 
 table it should be borne in mind that the relation between the moment of 
 inertia and section modulus of different sections vary. The values given in 
 the table represent average conditions. 
 
 A comparison of Plates XXVI and XXVII shows that the wheel load is 
 determined by the bearing power of the track in the case of classes B and C 
 track, but with class A track the working stress of the steel may be exceeded 
 without overloading the track. The dotted lines in the diagram for class A ' 
 track on Plate XXVI show the correction necessary to apply to the curves of 
 this diagram in order to keep within the working stress of the metal of the rail. 
 
 Plate XXVI will now serve as a basis for determining the dynamic wheel 
 load corresponding to any section of rail and axle spacing. For a wheel arrange- 
 ment consisting of a series of wheels having the same spacing and each sup- 
 porting approximately the same load, the values of the load may be read directly 
 from the diagrams. This loading satisfies the condition of the calculations 
 which assumes that the tangents to the elastic curve of the rail under the wheels 
 and midway between them are horizontal. 
 
 When the wheel spacing is not the same for adjacent wheels, the average 
 of the loads given for each axle spacing should be taken. 
 
 In the case of the front and back drivers the conditions are more complex. 
 Here we may have a trailing truck and either a two-wheel or four-wheel leading 
 truck carrying loads much less than those on the drivers. The preparation of 
 charts for all of these conditions and for the case where the wheel spacing of 
 adjacent drivers is not uniform would appear to be a refinement which would 
 not be warranted by the data upon which the calculations must necessarily be 
 based. 
 
 It will be observed that little variation in the wheel load occurs after a 
 distance between adjacent wheels of 180 inches is reached, the curve becoming, 
 in most cases, nearly horizontal at this point. In other words the wheels are 
 so far apart as to have little or no effect upon each other. 
 
 The following formula is proposed to be used in determining the wheel 
 loads on the front and rear drivers, it is also applicable to the outside wheels of 
 the trucks under the cars. 
 
 W _W W"-W" L W" 
 2 t- 2 W"^ 2 
 
STRENGTH OF THE RAIL 
 
 321 
 
 where 
 
 W = dynamic load of outside driver. 
 W = dynamic load corresponding to the wheel spacing between the 
 
 outside and adjacent driver. 
 W" = dynamic load corresponding to the distance between the outside 
 
 driver and the center of truck. 
 W" = the value given below for different moments of inertia and classes 
 of track. 
 L = dynamic load on truck wheels (one side). 
 
 Moment of In- 
 ertia of Rail. 
 
 For Front Drivers. 
 Class of Track. 
 
 For Back Drivers. 
 Class of Track. 
 
 A. 
 
 B. 
 
 C. 
 
 A. 
 
 B. 
 
 C. 
 
 50 
 40 
 30 
 
 20 
 
 36,000 
 30,000 
 26,000 
 
 28,000 
 26.000 
 24,000 
 22,000 
 
 18,000 
 16,000 
 14,000 
 12,000 
 
 40,000 
 
 34,000 
 30,000 
 
 32,000 
 30,000 
 28,000 
 26,000 
 
 20,000 
 18,000 
 16,000 
 14,000 
 
 
 
 
 
 w 
 
 The term — in the formula represents the part of the driving-wheel load 
 
 which is supported within the driving-wheel base: 
 
 W" 
 
 the load carried by the 
 
 rail outside of the driving-wheel base if there is no leading truck, this is made 
 smaller than indicated by the diagrams of Plate XXVI in the case of the leading 
 driver on account of the probable extra stress set up in the rail when it is first 
 depressed by the weight of the locomotive. 
 
 The term 
 
 W" - W" 
 
 ™ is introduced to provide for the extra support 
 
 2 W" 
 
 afforded by the truck wheels. In the extreme case where W" = L, the wheels 
 ahead of the driver exert the pressure corresponding to their distance from the 
 driver, and the term — ^— drops out of the equation. When L = or there is 
 
 no leading truck the term 
 
 W" - W" 
 
 ; becomes equal to zero. 
 
 2 W" 
 
 The dynamic wheel loads, having been determined, the corresponding 
 static loading can be readily computed by the methods given in article 10 for 
 steam locomotives, article 11 for electric locomotives, and article 12 for cars. 
 
 Plate XXVIII presents diagrams showing approximately the static loading 
 that can be placed on the rail with different wheel arrangements. From these 
 diagrams can be obtained the weight of rail and design of track necessary to use 
 in connection with engines where the maximum axle load is fixed, or the diagrams 
 
322 STEEL RAILS 
 
 may be used in determining whether or not it is safe to run a certain weight of 
 equipment over an existing line. 
 
 On account of the variation in design of engines a separate examination 
 should be made in most cases, as the diagrams, of necessity, represent general 
 conditions which may be varied from in a considerable degree. 
 
 Fig. A, Plate XXVIII, gives the track diagrams for passenger engines of the 
 Atlantic and Pacific types. The main driver in the wheel arrangement of the 
 Pacific engine can carry more weight than when it is one of the outside wheels, 
 as in the case of the Atlantic engine, and for this reason the former engine is 
 generally the most favorable on the track. 
 
 The ten-wheel engine is used extensively for passenger and freight service 
 on branch lines. This engine does not have the trailing truck of the two former 
 types, and the rear driver has, therefore, less carrying power than in the Pacific 
 engine, although about the same as the Atlantic where the effect of the trailing 
 truck is offset by the fact that the rear driver is the main wheel. The diagrams 
 given on Fig. B, Plate XXVIII, show the axle loads of ten-wheel engines for 
 classes B and C track. 
 
 Fig. C shows diagrams for Mogul and Consolidation freight engines. The 
 wheel arrangement of these engines is quite similar to that of the ten-wheel 
 engine, and it will be found that the curves agree very closely with that of the 
 ten-wheel engine used in freight service. 
 
 On Figs. D and E are diagrams for cars. It should again be observed 
 that these smaller diameter wheels should not be loaded as heavily as the 
 drivers, and the diagrams for the loads on car axles are not extended beyond 
 axle loads of 45,000 pounds. 
 
 The diagrams of the figures on Plate XXVIII illustrate very clearly the 
 effect of the different wheels on the track, and emphasize the fact that the 
 entire wheel arrangement must be considered in determining the maximum 
 load on any one wheel. 
 
 The assumption made by most foreign writers on this subject, that the 
 strains produced by the loads are independent of the position of the wheel is 
 obviously incorrect, as has been shown experimentally by Dr. Dudley's strem- 
 matograph tests. 
 
 The diagrams of Plate XXVIII are not extended beyond 60,000 pound 
 axle loads for drivers or 45,000 pounds for car wheels. With carbon steel 
 rails the use of very heavy loads should be approached cautiously until 
 further evidence is obtained in regard to the effect of such wheel loads on the 
 intensity of the stress at the contact of the wheel and the rail. While axle 
 
STRENGTH OF THE RAIL 323 
 
 loads of nearly 70,000 pounds are in service on experimental locomotives, they 
 have not been used in sufficient numbers to demonstrate fully their effect on 
 the rail. 
 
 The lack of proper experimental data presents many difficulties in accu- 
 rately determining the functions performed by the rail. Within certain limits, 
 however, the duty of the rail can be calculated with a considerable degree of 
 accuracy and more attention should undoubtedly be given to the effect of dif- 
 ferent loadings on the rail in the design of the engines and cars which run 
 over it. 
 
 It must be constantly borne in mind in dealing with the design of the track 
 that in many cases the strength of the rail is not the first consideration in the 
 selection of the section to be used, and that the question of obtaining a rail of 
 the requisite stiffness is of the greatest importance. The sudden failure of 
 any part of the track is not to be anticipated within the limits of customary 
 practice, but rapid deterioration of the structure may take place which will 
 eventually result in failure. 
 
 Economy of train service has become so important that it is safe to say 
 that there will be no return to lighter loads the tendency is, and will be constantly, 
 in the opposite direction. The importance therefore of giving to the design of the 
 track the same careful investigation as is considered essential in the design of a 
 bridge cannot be over-estimated. The track is, in fact, a continuous girder con- 
 necting termini over which pass the same loads as over the bridges. 
 
 The discussion of stresses in the ties, ballast and subgrade which has been 
 made in the preceding pages while sufficient to enable the allowable bearing power 
 of the supports of the rail to be determined within reasonable limits, is far from ex- 
 haustive enough in its character to serve as a basis for the general design of the 
 track and the proper proportioning of all the elements entering into its construc- 
 tion. Such an analysis would be clearly outside the limits of the present work and 
 while the various tables and formulae that have been developed appear sufficient 
 for the purpose intended, any general conclusions based upon their evidence alone 
 should be avoided. 
 
CHAPTER VI 
 INFLUENCE OF DETAIL OF MANUFACTURE 
 
 The evidence of the failure of rails of heavy section rolled within the 
 last few years, equaling and at times exceeding that of lighter rails of earlier 
 manufacture exposed to similar conditions of traffic and roadbed, points un- 
 mistakably to defective material in some of the later rails. These rails apparently 
 do not fail in the majority of cases due to too great a stress of the metal, and it is 
 this irregular failure of individual rails due to defects in manufacture which has 
 given rise to such just feelings of dissatisfaction and alarm. 
 
 Inferior quality of the metal in the rail may be attributed to two causes: 
 first, the use of imperfect methods of manufacture; second, the influence of the 
 form of the section upon the detail of manufacture. 
 
 First, let us examine the methods employed in the manufacture of the earlier 
 rails, which gave such satisfactory results, and which have been so constantly 
 presented to rail makers as representing that which they ought to do. 
 
 * The first steel rails were rolled in mills which had been designed for 
 iron rails. Other rolls were used and the number of passes was increased, 
 making the reduction very gradual. All blooms were allowed to cool before 
 being charged into the reheating furnace. After the drawing of one heat and 
 before the charging of another, the furnace was cooled down, then the heat 
 was brought up very gradually and plenty of time was taken to allow the steel 
 to " soak." In the converting house, all the possible practices of crucible steel 
 teeming were introduced. The ingot molds were carefully brushed out, heated 
 and smoked before being used. When the steel was teemed all doors and 
 windows of the casting house were closed and time was not spared on any of 
 the details. It was expected to produce but 50 per cent as much steel as iron 
 rail, and all employees working by the ton were paid twice as much for steel 
 as iron. The constant demand for cheaper prices (Fig. 220) and increased 
 tonnage rapidly changed these conditions. 
 
 Many of these practices, it is felt to-day, were entirely without reason, and it 
 is difficult to say as a general proposition that the steel produced was better than is 
 
 * See paper on Steel Rails, and Specifications for Their Manufacture, R. W. Hunt. Trans. 
 American Institute of Mining Engineers, Vol. XVII (1888-89), p. 226. 
 
 324 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 325 
 
 obtainable at the present time. Mr. Bumngton stated positively to the Indiana 
 Railway Commission, at its hearing, that the quality of the metal is to-day much 
 better than it ever was before, owing to the increased knowledge and better machines 
 and mechanical appliances than formerly existed. 
 
 Mr. James E. Howard, in his report of the accident on the Great Northern 
 Railway, near Sharon, N. D., on December 30, 1911, states: "It is important to 
 consider whether an improvement in the structural condition of rail steel is attain- 
 able. Such seems to be the case, since experimental rollings have furnished rails 
 $I50,_ , , , , , ,$150 
 
 1850 1850 1870 1880 1890 1900 I91C 
 
 YEAP 
 
 Fig. 220. — Prices of Iron and Bessemer Steel Rails, 1855 to 1910. 
 
 which, so far as could be ascertained, were free from streaks . . . It is believed to 
 be metallurgically feasible to produce better steel than has at times been offered 
 and accepted." 
 
 No doubt the failures which have their origin in defective metal are consider- 
 ably augmented by the character of the stress in the rail. On account of the con- 
 centration of the load on a small area, the stress is not distributed, and consequently 
 a metal of a great degree of uniformity is required. 
 
 With the large wheel loads now in use the injurious effect of inferior material 
 in the rail is much more apparent than in other structures not subjected to such 
 highly localized stresses. The situation calls for a refinement of manufacture not 
 generally realized in practice, and is further complicated by the desire for high 
 carbon to resist the head stresses, with the need for physical properties in the other 
 parts of the rail which could best be obtained by the use of a much milder steel. 
 
STEEL RAILS 
 
 29. Chemical Composition 
 It was supposed that the chemical character of the steel in the earlier 
 rails accounted for their excellent wear. Among the makers of these rails, 
 Sir John Brown & Co., of Sheffield, England, sent to this country those which, 
 from their excellent service, were considered by railroad engineers as the type 
 of what rails should be. Accepting the chemical theory, rail makers expected 
 that the analyses of these celebrated rails would present steel of exceptional 
 uniformity and purity. The contrary was proved to be the case. Carbon 
 varied from .24 to .70; silicon, .032 to .306; phosphorus, .077 to .156; sulphur, 
 .050 to .181; manganese, .312 to 1.046. The following is the variation found in 
 thirteen rails made by John Brown & Co., England, all of which had given 
 good service: 
 
 Per cent. Per cent. 
 
 Carbon 0.24 to 0.70 
 
 Manganese 312 to 1.046 
 
 Phosphorus 077 to . 156 
 
 Sulphur 050 to .155 
 
 Silicon 032 to .306 
 
 Below is given the analysis of some of the earliest English rails, imported 
 between 1860 and 1870. These rails, low in carbon and all other hardening 
 constituents, have given from thirty to thirty-five years' service before wearing 
 out, not breaking. 
 
 
 Rails Used on 
 Southern Railway. 
 
 Rails Used on P., C, C. & St. L. 
 Railway. 
 
 
 Per cent. 
 0.158 
 
 .77 
 .114 
 .067 
 .490 
 
 Per cent. 
 0.273 
 
 .28 
 .05 
 .04 
 .025 
 
 0.22 
 .21 
 .05 
 .031 
 .035 
 
 
 Phosphorus 
 
 Silicon 
 
 
 Fig. 221 shows the performance of two rails of very similar chemical com- 
 position, which, however, possessed quite different wearing qualities. 
 
 * In 1881, Dr. C. B. Dudley, the chemist of the Pennsylvania Railroad, 
 made an investigation to determine the relative relation between the chemical 
 and physical characteristics of steel rails and their power to resist wear. Dr. 
 Dudley found for the average of 32 slow- wearing rails the following composition: 
 
 Per cent. 
 
 Carbon 0.334 
 
 Phosphorus 0.077 
 
 Silicon 0.060 
 
 Manganese . 491 
 
 * The Wearing Capacity of Steel Rails in Relation to Their Chemical Composition and Physical 
 Properties, C. B. Dudley, Trans. American Institute of Mining Engineers, Vol. IX (1880-81), p. 321. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 327 
 
 and proposed a formula for the correct composition of steel rails, as follows: 
 
 Carbon, between .25 and .35; aim at . 30 
 
 Phosphorus, not above 0.10 
 
 Silicon, not above . 04 
 
 Manganese, between .30 and .40; aim at 0.35 
 
 Fig. 221. — Comparative Wear of Rails of Similar Chemical Composition. (Trans. A. S. C. E. 1889.) 
 
 Chemical 
 Composition 
 
 Year Rolled 
 Time in Service 
 
 Tonnage over Rail 
 
 0.322 per cent Carbon 0.355 per cent. 
 
 0.026 per cent Silicon 0.029 per cent. 
 
 0.077 per cent Phosphorus 0.108 per cent. 
 
 0.492 per cent Manganese 0.490 per cent. 
 
 1871 1876 
 
 August, 1871 to July, 1879 August, 1876 to July, 1879 
 
 7 years, 11 months. 2 years, 11 months. 
 
 On high side of 5° Curve On high side of 5° Curve 
 
 Grade 39.6 ft. per mile. Grade 39.6 ft. per mile. 
 
 40,061,230 tons. 21,504,824 tons. 
 
 Table LXXXIV gives the result of his experiments. 
 
 TABLE LXXXIV. — WEAR OF RAILS 
 
 (Dudley.) 
 
 Level tangents 
 
 Grade tangents 
 
 Level curves 
 
 Grade curves 
 
 Low side level curves. . . 
 High side level curves. . . 
 Low side grade curves. . . 
 High side grade curves. . 
 
 Tangents 
 
 Curves 
 
 Levels 
 
 Grades 
 
 Low side curves 
 
 High side curves 
 
 All conditions 
 
 2 slower wearing 
 
 32 faster wearing 
 
 .0701 
 .0706 
 .1277 
 .0500 
 .0911 
 .0801 
 .1754 
 .0542 
 .0992 
 .0545 
 .0989 
 .0650 
 .1332 
 .0767 
 .0506 
 
 114.1 
 113.3 
 62.7 
 160.0 
 
 123.1 
 
 60.1 
 104.3 
 158.1 
 
 77.8 
 
328 
 
 STEEL RAILS 
 
 Wellington states that the result of this investigation, which showed, or 
 seemed to show, that very hard rails did not wear so well as softer and tougher 
 rails, was taken to indicate that softness in itself was a desirable quality in a 
 rail ; and the painstaking character of the investigation and high reputation of 
 the road having given these conclusions wide dissemination, manufacturers for 
 many years took them as a guide, and produced rails that were too soft. 
 
 Table LXXXV shows the gradual increase of the hardening constituents in 
 the steel for rails since Dr. Dudley's investigation. 
 
 TABLE LXXXV. — COMPARISON OF THE CHEMISTRY OF EARLY AND RECENT RAILS 
 
 
 Date. 
 
 Weight 
 of Rail. 
 
 Lbs. 
 
 Chemistry. 
 
 Name. 
 
 Carbon. 
 
 Manganese. 
 
 Silicon. 
 
 Sulphur. 
 
 Phosphorus. 
 
 
 Des'd: 
 
 Min. 
 
 Max. 
 
 Min. 
 
 Max. 
 
 Min. 
 
 Max. 
 
 Min. 
 
 Max. 
 
 Min. 
 
 Max. 
 
 
 1881 
 1885 
 1892 
 1896 
 1899 
 
 1908 
 1909 
 
 
 Per 
 
 .30 
 
 Per 
 
 Per 
 
 Per 
 .30 
 
 Per 
 
 .40 
 
 Per 
 
 Per 
 .04 
 
 Per 
 
 Per 
 
 Per 
 
 Per 
 
 
 
 30 
 
 42 
 15 
 55 
 35 
 38 
 
 43 
 45 
 45 
 
 
 50 
 47 
 55 
 63 
 45 
 48 
 
 53 
 55 
 55 
 
 
 
 
 
 
 75 
 80 
 85 
 50- 60 
 60- 70 
 70- 80 
 80- 90 
 90-100 
 
 90 ' 
 
 ::: :56 :: 
 
 
 
 
 
 
 .05* 
 .07 
 .04 
 
 
 
 
 .85 
 
 >0 
 .70 
 .75 
 
 ^80 
 .80 
 
 1.15 
 1.10 
 
 LOO 
 1.05 
 1.10 
 1.10 
 1.20 
 1.05 
 
 .10 
 
 .10 
 
 .05 
 
 .20 
 
 .20 
 .20 
 
 
 
 
 
 The Carnegie Steel Company 
 
 Pennsylvania R.R. (Bessemer) 
 
 .10 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * Metal from which steel is 
 
 It has often been stated that the reason why the earlier rails seem to last 
 so well was due to the elimination of the poorer quality of rails by the service 
 in the track. This statement is not the complete explanation. The older 
 rails were cold rolled by the light wheel loads until the surface was sufficiently 
 hardened to bear the recent heavier loads without much increased abrasion. 
 New rails of the same section and practically the same physical properties 
 would, when subjected to heavier wheel loads, lose more of the metal by wear, 
 before the surface was rolled as hard as the former sections, and their rate 
 would be much faster.* 
 
 The light earlier sections could carry from 60 million to 75 million tons 
 before they were rough and unsuitable for passenger traffic, and, when in a 
 location where the tonnage was only 250,000 to 500,000 tons per month, 
 would last in the track many years. 
 
 The six-inch 100-pound Dudley sections of 0.06 phosphorus and 0.65 car- 
 bon laid in 1895 on the New York Central and Hudson River Railroad, when 
 taken up in 1907 had carried 375 million tons with a loss of about one-eighth of 
 an inch in depth on the head of the rail. 
 
 * Paper by Dr. P. H. Dudley before the Railroad Commission of Indiana, February 20, 1912. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 329 
 
 0.05 
 
 0.1 
 
 0.2 
 
 0.4 
 
 0.6 
 
 0.8 
 
 1.0 
 
 1.3 
 
 25.00 
 
 26.0 
 
 31.0 
 
 36.0 
 
 43.0 
 
 58.0 
 
 60.0 
 
 44.0 
 
 Garbon is the most important element, except iron, in steel. The me- 
 chanical properties of iron-carbon alloys are closely connected with the relative 
 amounts of the two elements. The relation between the percentage of carbon 
 in an alloy and the tenacity in tons per square inch is indicated* in the following 
 table : 
 
 Per cent of carbon 
 Tenacity, tons per sq. in. 
 
 The results are shown graphically in Fig. 222. 
 
 Silicon in small proportions hardens the steel and stands intermediate 
 between carbon and phosphorus in this respect. It is used to prevent unsound 
 or honey-combed ingots, but when so used 60p 
 tends to render the steel unduly hard. Silicon 
 as high as .2 per cent, in high-carbon steel of 
 .5 and .6 per cent carbon, probably has no 
 injurious effect. 
 
 Phosphorus hardens steel more rapidly 40 - 
 than either carbon or silicon. It increases 
 its rigidity but impairs its power to resist 
 impact. Small proportions render the metal 
 harder without materially affecting its 
 tenacity, but makes the metal at the same 20 
 time decidedly cold-short. An excess of phos- Fig. 222.- 
 phorus also renders the steel sensitive to high 
 
 heat. Mr. Robert W. Hunt, in his experiments in trying to make high phosphorus 
 steel in the Clapp-Grimth converter, found that it was necessary to be very careful 
 not to over-heat the steel. 
 
 | Owing to the exhaustion of the available low-phosphorus ores, Bessemer 
 rail steel is now of necessity a high-phosphorus and low-carbon alloy, the mean 
 carbon being about 0.50 J per cent, while the impurity of phosphorus is limited 
 to 0.10 per cent. 
 
 Plain basic open-hearth rail steel is usually a low-phosphorus and medium 
 unsaturated carbon alloy, as most of the phosphorus has been reduced by this 
 
 * H. M. Howe, Engineering and Mining Journal, p. 241, 1887. See also Steel by Harbord and 
 Hall, London, 1911, pp. 347, 348, and a Study of the Elastic Properties of a Series of Iron-Carbon 
 Alloys, Jones and Waggoner. Proceedings American Society for Testing Materials, Vol. XI, 1911, 
 pp. 492^99. 
 
 t See Proceedings American Society for Testing Materials, Vol. XI, 1911, Dudley, Ductility in 
 Rail Steel. 
 
 t The chemical composition refers, in this and the following example, to 100-pound rails. 
 
 rP 
 
 0-4 0-8 1-2% Carbon 
 
 —Tenacity of Iron-Carbon Alloys. 
 
330 STEEL RAILS 
 
 process from its content in the ores and iron to 0.04 per cent or under. This 
 permits, in this class of steel rails, carbon of 0.63 to 0.75 per cent. 
 
 Sulphur has little influence on the tensile strength or ductility. The 
 real effects of sulphur, however, are seen during the rolling, a very small per- 
 centage causing a great red-shortness. Its presence in excess of .06 or a maxi- 
 mum of .08 per cent tends to cause cracks to develop during the rolling, which, 
 while they close up and are almost imperceptible in the finished rail, neverthe- 
 less remain as flaws and may form starting points for rupture when the rail 
 is subjected to any sudden stress. With sulphur, it is necessary to work the metal 
 at a high heat to avoid its cracking during manipulation. The " red-short" term 
 means that as the heat approaches the red color the tendency to crack becomes 
 intensified, while the effect of phosphorus on heated metal is to make it hot 
 short or short under high heat; in other words, it will work at a low tempera- 
 ture, but is sensitive to a high one. 
 
 Apparently the greater part of the sulphur unites with the manganese 
 forming manganese sulphide, which is occluded by the metal as a foreign sub- 
 stance, preventing its welding and breaking up the continuity of its structure. 
 The impurity of sulphur was limited formerly to 0.075 or 0.08 per cent. The 
 manufacturers now charge for this limitation of sulphur five cents extra per 
 hundred pounds, and it is, therefore, being omitted from some specifications, 
 although in most cases it is required that its content be reported. 
 
 Manganese has a general tendency to increase the tensile strength and 
 reduce the ductility; this influence varying with the amount of carbon pre- 
 sent in the steel and becoming more marked in the case of high-carbon than 
 low-carbon steels. It is possible to keep the manganese down, by the use of a low 
 manganese spiegle, and with low-sulphur steel its presence in excess of .8 per cent, 
 or its use to bring up the tensile strength in place of carbon, is dangerous on account 
 of its very distinct hardening effect when above .6 per cent. In the commercial 
 run of iron, where the sulphur varies, the practice is to allow the manganese to go 
 as high as 1.1 per cent, and some authorities do not consider it dangerous unless 
 above 1.0 per cent even with low sulphur. Manganese tends to neutralize the effect 
 of sulphur and prevent the metal becoming red-short, and, to a limited degree, the 
 cold-shortness produced by phosphorus. 
 
 The above elements are those generally considered in rail steel, and speci- 
 fications rarely refer to the other elements which may be contained in the ore, 
 and which either from design or accident are present in the finished product. 
 The most important of these are arsenic and copper. 
 
 The effect of arsenic upon steel was quite fully investigated several years 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 331 
 
 ago by Harbord and Tucket.* The conclusions given by them may be sum- 
 marized as follows: 
 
 Arsenic, in percentages not exceeding .17, does not appear to affect the 
 bending properties at ordinary temperatures, but above this percentage cold- 
 shortness begins to appear and rapidly increases. In amounts not exceeding .66 
 per cent the tensile strength is raised very considerably. It lowers the elastic 
 limit and decreases the elongation and reduction of area in a marked degree. 
 
 Messrs. Ball and Wingham f have investigated the influence of copper on 
 the tensile strength of iron and steel. An alloy containing : 
 
 Per cent. 
 
 Copper 7.550 
 
 Carbon 2.720 
 
 Manganese . 290 
 
 Silicon . 036 
 
 Phosphorus . 130 
 
 Sulphur 0. 190 
 
 was bright, white in color, crystalline, and very hard, but did not offer any 
 great resistance to impact. Varying quantities of the alloy were melted down 
 with Bessemer steel, and test pieces 1 inch by | inch by ^ inch were annealed 
 before being tested. The following table shows the results: 
 
 Number. 
 
 Copper. 
 
 Carbon. 
 
 Tensile 
 
 Strength. 
 
 1 
 
 2 
 3 
 4 
 
 Per cent. 
 0.847 
 
 2.124 
 3.630 
 7.171 
 
 Per cent. 
 
 0.102 
 
 0.217 
 0.380 
 0.712 
 
 Tons per sq. in. 
 
 18.3 
 36.6 
 47.6 
 56.0 
 
 From these experiments it is clear that copper increases the tensile strength 
 of iron. % The simultaneous presence of carbon tends to prevent the more 
 intimate association of copper with iron. In test piece No. 1, the fractured 
 surface was somewhat fibrous, while No. 2 and the others were highly crystalline. 
 Even in the absence of carbon, copper makes iron extremely hard. Mr. F. 
 Stubbs states that the presence of \ per cent of copper in steel gives to it the 
 property of preventing the oxidation of the steel on being subjected to a burning 
 heat. 
 
 Copper in steel rails in small quantity does not materially affect the me- 
 chanical properties, but in steels, in which high ductility is required, especially 
 in those with high carbon, copper is objectionable. Steel with copper, say up 
 to 1 per cent, appears to resist corrosion better than the same steel without 
 
 * On the Effect of Arsenic on Mild Steel, Journal Iron and Steel Inst., Vol. 1, 1 
 
 t Iron and Steel Inst., No. 1, 1889. 
 
 t Steel and Iron for Advanced Students, Hiorns, London, 1903, p. 322. 
 
 ;, p. 183. 
 
332 STEEL RAILS 
 
 copper. * Campbell states that 1 per cent may be present without injuring 
 the steel, provided there be but little sulphur; but that if the sulphur be up to 
 .08 or .10, the metal will be red-short, and that copper also reduces the welding 
 power of the metal, especially if sulphur be present; but he adds: " In all cases 
 the cold properties seem to be entirely unaffected." f Richards states that 
 "copper causes red-shortness, but much less than sulphur. Five-tenths per 
 cent may be allowed in rails and its effect is overcome by manganese." 
 
 X The influence of copper on steel was formerly greatly exaggerated. 
 Whereas it was considered to be very harmful, it is now known, when present 
 in small quantities, to have no serious influence on the physical properties of steel. 
 
 Mr. H. J. Force reports a case of an 80-pound rail made by the Lackawanna 
 Steel Company in 1895, which had given very good service. An analysis showed 
 about .40 per cent carbon and about .60 per cent copper. § 
 
 According to a statement in Professor Howe's " Metallurgy of Steel " || an 
 American firm of steel-rail makers habitually made Bessemer tee rails with .51 
 to .66 per cent copper and they were so slightly red-short that in spite of the 
 thin flanges and low finishing temperature only from 1.25 to 2.5 per cent of 
 them were so defective as to be classed as second quality. 
 
 Mr. R. W. Hunt states that in the early days of the steel industry excellent 
 rails were produced from Cornwall irons. A large number of these rails con- 
 tained .5 per cent of copper. The Pennsylvania Steel Company, as well as 
 the Bethlehem and the Troy Works, used Cornwall iron containing low phos- 
 phorus and high copper as their basis for a long time. 
 
 Clamer H has found that the addition of copper and nickel in combination 
 seems to have the same effect upon the steel as if they were individually added, 
 the copper in its effect really being about the same as so much added nickel. 
 It is possible, therefore, to replace part of the nickel in nickel steel by copper, 
 without materially altering its physical properties. Recently Messrs. Burgess 
 and Aston, working quite independently of Clamer, have confirmed these results. 
 
 The attention of railroad engineers is being directed toward the develop- 
 ment of alloy steel, or steel containing a percentage of various materials intro- 
 duced to give it special mechanical qualities. In general, however, on account 
 of the higher cost of production, these steels are confined to use in special locali- 
 
 * Metallurgy of Iron and Steel, A. Humboldt Sexton, Manchester, 1902, p. 247. 
 
 t Notes on Iron, Robert H. Richards, 1895. 
 
 J Metallurgy of Steel, Harbord, London, 1911, p. 375. 
 
 § Proceedings American Society for Testing Materials, Vol. X, 1910, p. 279. 
 
 II Metallurgy of Steel, New York, 1891, p. 83. 
 
 1 Proceedings American Society for Testing Materials, Vol. X, 1910, Clamer on Cupro- 
 
INFLUENCE OF DETAIL OF MANUFACTURE 333 
 
 ties where the conditions are especially severe, as on sharp curves under heavy 
 traffic or in tunnels where it is a troublesome matter to inspect or renew the 
 rails. 
 
 The requirements of steel alloy may be summarized as follows: 
 
 (1) High resistance to shock; 
 
 (2) High elastic limit; 
 
 (3) Resistance to abrasion. 
 
 Some of the alloys best known are manganese, nickel, chromium, and titanium. 
 * The record of the chrome nickel on the Central Railroad of New Jersey, 
 and of the plain nickel on the Pennsylvania Lines, Northwest System, is not 
 very good. In a period of six months there were 112 failures per 10,000 tons 
 laid of 85-pound A. S. C. E. section of nickel steel from the Carnegie Steel Com- 
 pany on Pennsylvania Lines, Northwest System, the chemical composition 
 being: 
 
 Per cent 
 
 Carbon 44 
 
 Phosphorus 09 
 
 Manganese 80 
 
 Silicon. . . , 10 
 
 Sulphur 03 
 
 Nickel 3.42 
 
 The 90-pound A. S. C. E. section chrome nickel steel from the Bethlehem 
 Steel Company on the Central Railroad of New Jersey for the same period 
 showed failures of 41 per 10,000 tons laid, nearly all of which were broken rails. 
 
 The same committee reported for the year, ending October 31, 1910, that 
 the record for 90-pound A. S. C. E. open-hearth rail with chromium and nickel 
 on the Central Railroad of New Jersey has been very bad so far as failures are 
 concerned, there having been 1,129 per 10,000 tons of rail laid, mostly break- 
 ages. This rail is 1909 manufacture. Small lots have also been tested on the 
 Baltimore and Ohio and the Erie with a large number of failures. The amount 
 of nickel is 2 per cent to 2 J per cent, and the chromium 0.5 per cent to 0.9 per 
 cent. In most cases these rails showed a very marked resistance to flange 
 wear as compared with ordinary carbon steel rails. 
 
 It has been found desirable to lower the carbon when the other hardening 
 elements are added. A rail with carbon 0.40, chrome 0.50, and nickel 1.25 is about 
 equal to a 0.60 carbon ordinary rail. 
 
 Manganese steel with C 0.77, P 0.06, Mn 9.93, Si 0.25, and S 0.038 showed 
 about one-third as much abrasion of the head as ordinary Carnegie Bessemer in 
 a test, on the Norfolk and Western, lasting nineteen months. 
 
 * Proceedings Am. Ry. Eng. & M. of W. Assn., Vol. 11, Part 1, 1910, p. 315. 
 
334 STEEL RAILS 
 
 Chrome steel, which usually contains about 2 per cent of chromium and 
 .80 to 2 per cent of carbon, owes its value to combining, when in the " hardened" 
 or suddenly cooled state, intense hardness with a high elastic limit, so that it 
 is neither deformed permanently nor cracked by extremely violent shocks. 
 
 * The tensile strength rises with increase in the percentage of chromium 
 till with about 5 per cent it is about 74 tons unannealed, or 55 tons annealed, 
 the elongation being 13 per cent in the latter and 8 per cent in the former case. 
 The limit of elasticity was 40 tons in the first and 20 tons in the second case. 
 As the quantity of chromium is increased the metal becomes harder, and with 
 about 9 per cent can hardly be touched with the file. In the absence of carbon 
 its hardening influence is not so marked. Forging makes the metal hard and 
 brittle, but the latter property is removed by annealing, and it is rendered 
 excessively hard by quenching. It has a high resistance to shock, and is there- 
 fore suitable for the manufacture of rails. 
 
 t On low-carbon steels not annealed, the addition of each 1 per cent of 
 nickel up to 5 per cent causes, approximately, an increase of 5000 pounds per 
 square inch in the elastic limit and 4000 pounds in the ultimate tensile strength. 
 The influence of nickel on the elastic limit and ultimate strength increases with 
 the percentage of carbon present, high-carbon nickel steels showing a greater 
 gain than low-carbon nickel steels.J 
 
 The addition of nickel to steel raises the proportion of elastic limit to ulti- 
 mate strength and adds to the ductility of the steel. This effect of nickel in 
 increasing the ratio of the elastic limit to tensile strength, without sacrifice to 
 ductility, accounts for the increase in the working efficiency of nickel steel over 
 carbon steel; in other words, its increased resistance to molecular fatigue. 
 
 The exhaustive series of experiments made by Wedding and Rudeloff 
 show that the resistance to compression of nickel-iron alloys increases steadily 
 with the per cent of nickel present, until 16 per cent of nickel is reached. Had- 
 iield has also made a very complete series of experiments on the resistance of 
 nickel steel to compression. He has found that a steel containing .27 per 
 cent nickel shortened, under a compression of 100 tons (224,000 pounds) per 
 
 * Metallurgy of Iron and Steel, A. H. Sexton, Manchester, 1902, p. 517. 
 
 t Nickel Steel: Its Properties and Applications. Colby. Proceedings American Society for 
 Testing Materials, Vol. Ill, 1903. 
 
 t The subject of nickel steel has received considerable attention, notably by D. H. Browne, 
 Trans. American Institute of Mining Engineers, Vol. 29, 1899, p. 569, and A. L. Colby, "A Comparison 
 of Certain Physical Properties of Nickel and Carbon Steel," Bethlehem Steel Company, 1903. See also 
 Guillet, Journal, Iron and Steel Institute, Vol. 2, 1908, p. 177; Waterhouse, Proceedings Am. Soc. 
 for Testing Materials, Vol. VI, 1906, pp. 249-258; Campbell and Allen, ibid, Vol. XI, 1911, pp. 428- 
 
INFLUENCE OF DETAIL OF MANUFACTURE 335 
 
 square inch, 49.90 per cent in a length of 1 inch; a steel with 3.82 per cent nickel 
 shortened 41.38 per cent; with 5.81 per cent nickel, 37.76 per cent; and with 
 11.30 per cent nickel, only 1.05 per cent. He states that an ordinary mild carbon 
 steel without nickel, under similar conditions, would be shortened 60 per cent 
 to 65 per cent. He argues that the toughening action of nickel when added to 
 steel is caused in a very intimate combination of the molecular structure, and 
 that this advantage is further enchanced by the fact that the nickel does not 
 show a disposition to segregate in steel like other elements; in other words, it 
 appears to be more intimately combined. 
 
 Mr. Campbell, of the Pennsylvania Steel Company, made a series of tests 
 to prove what he states to be the current impression among manufacturers 
 of nickel steel, — that the presence of this element prevents segregation. His 
 conclusion is, that there seems to be good ground for the assumption that nickel 
 prevents the separation of the metalloids, but that it does not prevent it alto- 
 gether, and he states that it is not probable that any other agent will ever be 
 found competent for this task. 
 
 * Howe states that nickel steel, which usually contains from 3 to 3.50 
 per cent of nickel and about .2.5 per cent of carbon, combines very great 
 tensile strength and hardness, and a very high limit of elasticity, with great 
 ductility. 
 
 The combination of ductility, which lessens the tendency to break when 
 overstrained or distorted, with a very high limit of elasticity, gives it great value 
 for shafting, the merit of which is measured by its endurance of the repeated 
 stresses to which its rotation exposes it whenever its alignment is not mathe- 
 matically straight. The alignment of marine shafting, changing with every 
 passing wave, is an extreme example. In a direct comparative test the pres- 
 ence of 3.25 per cent of nickel increased nearly sixfold the number of rotations 
 which a steel shaft would endure before breaking. 
 
 As has been seen, nickel steel has been used tentatively for railroad rails; 
 but while it has the stiffness and resistance to wear which they require, too many 
 rails have broken in use. We may hope that this treacherousness will be pre- 
 vented. It is quite possible that a change in the percentage of nickel may give an 
 entirely different record. The Mayari ore used by the Maryland Steel Company 
 contains a natural percentage of chromium and nickel, and the results with rail 
 made from this ore seem, so far, to be pretty good. 
 
 Figs. 223 and 224 give the tensile strength and the ductility of many speci- 
 mens of nickel steel from various sources, chiefly, however, from M. Dumas' 
 
 * Iron, Steel, and Other Alloys, Howe, 1903, pp. 316-324. Contains report of M. Dumas' work. 
 
336 
 
 STEEL RAILS 
 
 important monograph.* The curves here given are taken from his work (pages 
 18 and 19). A rough resemblance to the manganese steel curves (Figs. 225 and 
 226) may be noticed. The great increase of ductility in case of manganese steel 
 in the 13 per cent manganese region is reflected in case of nickel steel by a like 
 and very abrupt rise at about 25 per cent of nickel. 
 
 260,000 
 
 
 h 
 
 
 
 
 
 
 H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 h 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 H 
 
 H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 f 
 
 
 $ 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 '** 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " 
 
 
 
 
 <" 
 
 
 
 K- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 A 
 
 
 1 '^ 
 
 140,000 
 120,000 
 100,000 
 80,000 
 60,000 
 40,000 
 20,000 
 
 ° 
 
 
 $ 
 
 » 
 
 
 
 
 
 
 
 
 
 ^ffi H 3 
 
 
 
 
 
 
 
 
 
 
 
 *i n 
 
 $ h 
 
 
 
 
 
 h 
 
 
 
 • 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 i 
 
 
 if 
 
 
 
 
 
 
 
 
 * 
 
 iHr a H 
 
 
 
 
 
 
 
 
 
 
 V 
 
 $3 
 
 
 
 
 
 
 o'h \ 
 
 
 
 
 H 
 
 h£ h 
 
 $ \ 
 
 H 
 
 
 
 ^ 
 
 
 
 
 H 
 
 »! 
 
 '% 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 H 
 
 ^T 
 
 
 
 
 
 
 
 
 m 
 
 , H 
 
 i 
 
 
 
 
 
 $ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 $ 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 i 
 
 1 
 
 2 1 
 
 I 1 
 
 6 1 
 
 8 2 
 
 2 
 
 2 2 
 
 4 26 28 30 32 34 36 38 40 42 44 
 
 O = -The Steel has not n 
 
 PERCENTAGE OF NICKEL 
 
 ie Steel hi 
 
 >» cooled by extremely low 
 3- r- 78°C-»r78°C T efo). 
 fany!, not given. 
 
 Fig. 223. — Influence of the Proportion of Nickel and Varying Heat-Treatment upon the Tensile Strength 
 of Nickel Steel. (Dumas.) 
 
 As actually made, manganese steel contains about 12 per cent of manganese 
 and 1.50 per cent of carbon. Although the presence of 1.50 per cent of manga- 
 nese makes steel brittle, and although a further addition at first increases this 
 brittleness, so that steel containing between 4 and 5.5 per cent can be pulverized 
 under the hammer, yet a still further increase gives very great ductility, accom- 
 panied by great hardness, — a combination of properties which, so far as known, 
 v/as not possessed by any other known substance when this remarkable alloy, 
 known as Hadfield's manganese steel, was discovered. 
 
 Its ductility, to which it owes much of its value, is profoundly affected by 
 the rate of cooling. Sudden cooling makes the metal extremely ductile, and 
 slow cooling makes it brittle; its behavior in this respect is thus the opposite of 
 
 * " Recherches sur les Aciers au Nickel a Hautes Teneurs," M. L. Dumas, Paris, 1902. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 337 
 
 60 
 56 
 52 
 48 
 44 
 | 40 
 § 36 
 
 z 32 
 
 j§28 
 3 24 
 S 20 
 16 
 12 
 8 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ** 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 « 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 } 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 i j 
 
 
 
 j 
 
 
 
 
 
 
 
 N i 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 s 5 ? 
 
 
 
 
 
 
 
 
 :•*■ 
 
 is 
 
 
 8 
 
 
 
 
 
 
 
 
 
 
 
 ® N 
 
 
 
 
 
 
 
 
 *tf 
 
 , 
 
 S 
 
 
 5 
 
 
 
 
 
 
 
 
 „ * 
 
 
 
 
 
 
 
 
 
 
 i. j 
 
 
 <!> 
 
 N< 
 
 I 
 
 
 
 
 
 
 
 i 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 i .; 
 
 r* 
 
 f 
 
 
 \ 
 
 
 
 
 
 * 
 
 ( 
 
 r 
 
 
 
 ® 
 
 
 
 
 
 
 
 j 
 
 *f 
 
 
 ? 
 
 
 V 
 
 
 
 
 / 
 
 *~~ 
 
 ji- 
 
 -k! 
 
 
 
 j 
 
 
 
 
 
 
 
 
 
 * 
 
 j 
 
 
 I 
 
 1 
 
 • t 
 
 /f 
 
 / 
 
 \ 
 
 ? 
 
 H 
 
 
 
 i 
 
 
 
 
 
 
 
 
 T 
 
 t 
 
 
 
 
 \ 
 
 ^ 
 
 r' 
 
 
 
 
 I 
 
 8 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ♦ 
 
 
 
 
 
 
 i 
 
 • 
 
 
 
 
 
 
 
 
 10 12 14 16 18 20 22 24 26 28 30 32 34 
 Legend!- PERCENTAGE OF NICKEL 
 
 2="% Elongation in 2 inches. g=% Elongation in 9 or 10 inches. 
 
 4=#> a »4 « x _] 4 ,, , not given, -inches. 
 
 J" = The Steel has received heat-treatment ( annealing! hardening, etc)". 
 = n n ,j> been subjected to extremely low temperature.. 
 
 Fig. 224. — Influence of the Proportion of Nickel and Varying Heat-Treatment upon the Ductility of 
 Nickel Steel. (Dumas.) 
 
 200,000 
 ^ 190,000 
 ~ 180,000 
 §" 170,000 
 £ 160,000 
 °"- 150,000 
 2 140,000 
 ~ 130,000 
 ^E 120,000 
 i 110,000 
 g 100,000 
 w 90,000 
 ^ 80,000 
 £ 70,000 
 £ 60,000 
 
 50,000 o~j~2~3~4." 5 6 7 8 9 1011121314151617 1819202122 
 PERCENTAGE OF MANGANESE 
 Fig. 225. — Influence of the Proportion of Manganese < 
 (Howe.) 
 Legend: 
 
 • = Slowly Cooled Manganese Steel. 
 + = Water-toughened or Suddenly Cooled Manganese Steel. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 + 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 , + 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 /* 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i the Tensile Strength of Manganese Steel. 
 
338 
 
 STEEL RAILS 
 
 that of carbon steel. Its great hardness, however, is not materially affected 
 
 by the rate of cooling. 
 
 The fact that when cold it is unalterably hard has, however, limited its use, 
 
 because of the great difficulty of cutting it to shape, which has in general to 
 
 be done with emery wheels instead of the usual iron-cutting tools. Another 
 
 defect is its relatively low elastic limit. 
 
 Fig. 225 shows the remarkable increase of tensile strength which occurs 
 
 when the manganese rises from 7 to 13 per cent, and the decline of tensile strength 
 
 as the manganese increases still 
 further. By the contrast be- 
 tween the position of the crosses 
 and the black dots it shows also 
 the remarkable effect of sudden 
 cooling. 
 
 Fig. 226 shows the corre- 
 sponding changes in ductility. 
 To show that the maxima for 
 tensile strength and ductility 
 coincide, the tensile-strength 
 curve sketched by eye in Fig. 
 225 is reproduced in Fig. 226. 
 
 In Fig. 227 is shown the 
 degree to which manganese steel 
 combines tensile strength with 
 ductility, and in Fig. 228 the 
 degree to which it combines 
 ductility with elasticity. These 
 combinations are often taken 
 as a rough measure of the 
 general degree of excellence of 
 
 • = Slowly Cooled Manganese Steel. a metal for engineering pur- 
 
 + = Water-toughened or Suddenly Cooled Manganese Steel. ,-, . , , 
 
 poses. For comparison the 
 corresponding properties of carbon steel are shown by small black dots, which 
 fall in a pretty well-defined band, much below the manganese-steel crosses. 
 
 These comparisons may, however, give a false idea of the ductility of man- 
 ganese steel. If two metals elongate in a like manner, the extent of their elon- 
 gation may be a fair comparative measure of their ductility; not necessarily so, 
 however, when their mode of elongating is unlike in kind. A bar of carbon steel 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -t) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 | 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 | 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2: 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ..'; 
 
 4 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 of 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 V 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 + 
 
 
 
 V 
 
 7> 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 ^ 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 IS 
 
 + 
 
 
 ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 & 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 / 
 
 # 
 
 
 
 
 
 
 
 
 % 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 v? 
 
 
 
 
 
 
 k 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 /<. 
 
 $ 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 + 
 
 
 
 
 
 
 
 
 
 
 
 '/ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 0123456789 10 11 12 13 H15 16 17 18192021 22 
 PERCENTAGE OF MANGANESE 
 Fig. 226. — Influence of the Proportion of Manganese on 
 the Ductility of Manganese Steel. (Howe.) 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 339 
 
 habitually yields by " necking " when pulled in two, contracting greatly just 
 about the place where rupture occurs, while a bar of manganese steel or of brass 
 elongates far more uniformly over its whole length. 
 
 The use of manganese frogs in severe service on steam roads and for rails 
 on curves of 75 feet radius or less for permanent street railway work has 
 been found preferable to ordinary carbon open-hearth or Bessemer material. 
 
 Titanium steel, while not strictly an alloy steel, may be conveniently treated 
 under this head. This 
 metal, like vanadium, alu- 
 minum or silicon, produces 
 a sounder ingot, and under 
 the usual practice the 
 titanium goes into the slag 
 and ordinarily there is no 
 intention of producing ti- 
 tanium alloy steel. 
 
 A progress report of 
 the Baltimore and Ohio 
 shows that the titanium 
 rail with .70 carbon on 
 Kessler's curve is only 
 wearing one-third as fast 
 as the Bessemer steel 
 with .50 carbon, with 
 which it is compared. 
 The results of six months' 
 service on a New York 
 Central crossover carry- 
 ing a heavy tonnage show 
 that the flange wear of 
 titanium rails was very 
 much reduced as com- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 + 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 !■+ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 y. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 + 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 + 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 TENSILE STRENGTH- POUNDS PER SQUARE INCH 
 Fig. 227. — Tensile Strength and Ductility of Carbon Steel and of 
 
 Manganese Steel. (Howe.) 
 Legend: 
 
 • = Carbon Steel. 
 + = Water-toughened or Suddenly Cooled Manganese Steel. 
 
 pared with that of ordinary Bessemer rails.* 
 
 t The effect of titanium on steel as understood to-day is to give the 
 metal greater density and strength. Recent tests on titanium rail steel made by 
 
 * Iron Age, March 25, April 29, and August 5, 1909. 
 
 t The Use of Titanium Rail on the Baltimore & Ohio Railroad. A. W. Thompson. Proceedings 
 Am. Ry. Eng. & M. of W. Assn., Vol. 11, Part 1, 1910, and Railroad Age Gazette, November 12, 1909. 
 
340 
 
 STEEL RAILS 
 
 Dr. Waterhouse show the elastic limit to be raised about 6000 pounds above the same 
 steel to which the titanium alloy had not been added. In the 150 rails examined 
 the titanium steel from different parts of the ingot showed a remarkable degree of 
 uniformity.* 
 
 The first experiments with titanium alloy in rail manufacture were made 
 
 by the Maryland Steel Company 
 in November, 1907, and this was 
 followed in 1908 by the Duquesne 
 Works of the Carnegie Steel Com- 
 pany, the Cambria Steel Com- 
 pany, and the Lackawanna Steel 
 Company. During the year 1909 
 the process passed the experi- 
 mental stage and has since been 
 used in a large number of rails and 
 may be regarded as firmly fixed on 
 a commercial basis. 
 
 During the month of June, 
 1908, 19 rails were rolled by the 
 Maryland Steel Company, of the 
 usual composition, to which was 
 added 1.5 per cent titanium alloy. 
 This alloy was claimed to increase 
 the elastic limit, ultimate strength, 
 and remove a large percentage of 
 the slag; also to make the rail less 
 brittle and avoid extreme segrega- 
 tion and blowholes, leaving the 
 metal homogeneous, tough, and 
 The use of the alloy 
 resulted in a rail with a composi- 
 tion high in carbon and phosphorus, which even then successfully passed the 
 physical test. 
 
 The analysis made by the Maryland Steel Company of this rail shows 
 the following: 
 
 C. Mn. P. . S. Si. N. O. 
 
 0.701 0.92 0.086 0.048 0.079 0.004 Nil 
 
 * See also C. V. Slocum, Mechanical Engineer, Vol. XXIII, 1909, pp. 336-337. 
 
 _fc± 
 
 ELASTIC LIMIT- POUNDS PER SQUARE INCH 
 Fig. 228. — Elasticity and Ductility of Carbon Steel 
 
 and of Manganese Steel. (Howe.) 
 Legend: 
 
 • = Carbon Steel. 
 + = Water-toughened or Suddenly Cooled Manganese ■ft ne _g ra m e cl 
 Steel. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 341 
 
 The addition of ferrotitanium to the ladle has an important influence on tho 
 mechanical structure of the steel by acting as a flux or scavenger, and the cleansing 
 effect results in increased solidity and purity of the metal. 
 
 Ferrotitanium contains 10 to 15 per cent of titanium, 5 to 10 per cent of car- 
 bon, and of other impurities less than 5 per cent; the balance is pure iron which 
 has been electrically refined. In Bessemer steel rail manufacture it is the practice 
 to add from ^ to 1 per cent crushed ferrotitanium in the ladle as the steel is poured 
 from the converter, and then hold the heat in the ladle about three minutes before 
 pouring the ingot. Sulphur and phosphorus do not appear to be reduced ; but in 
 combining with oxygen and nitrogen, forming oxides and nitrides, titanium has an 
 important action in removing these impurities, forming a stable combination of 
 them, which passes into the slag. 
 
 The New York Central Lines, prior to 1911, obtained rails from the Lacka- 
 wanna Steel Company with ^ of 1.0 per cent of ferrotitanium alloy added to the 
 ladle. The addition to the cost for plain Bessemer was 25 cents per ton for hold- 
 ing the metal in the ladle three minutes after the ferrotitanium was added and 
 $1.05 per ton for the alloy, or a total of $1.30 per ton. In 1911 T V of 1.0 per cent 
 of metallic titanium was added to the metal and the price subsequently reduced, 
 owing to a reduction in the cost of ferrotitanium. It is claimed that this small 
 proportion of ferrotitanium is sufficient to remove the bulk of blowholes and segrega- 
 tion usually found in Bessemer ingots and produce a clean, solid, good-wearing rail. 
 
 The following table * gives the production of alloy rails in the years 1909 
 and 1910. It appears that greater effort has been made to improve the Besse- 
 ' mer rail by the use of alloys than the open-hearth rail. 
 
 
 1909 
 
 1910 
 
 
 Tons. 
 
 35,945 
 1,028 
 
 12,287 
 1,245 
 
 Tons. 
 
 195,940 
 390 
 
 4,210 
 81 
 
 
 
 
 
 Nickel, chrome, and vanadium rails 
 
 Total 
 
 50,505 
 37,809 
 12,696 
 
 200,621 
 174,822 
 25,799 
 
 
 
 
 Table LXXXVI gives the specification of chemical composition adopted 
 as recommended practice by the American Railway Engineering Association 
 March, 1912, for carbon steel rails. Table LXXXVII presents the chemical 
 specification adopted January 1, 1909, by the Association of American Steel 
 
 * Railway Age Gazette, March 16, 1910 (daily edition), and the Iron Age, February 23, 1911, 
 p. 461. 
 
342 
 
 STEEL RAILS 
 
 Manufacturers for standard Bessemer and open-hearth steel rails for A. S. C. E. 
 sections. 
 
 TABLE LXXXVI 
 
 CHEMICAL COMPOSITION OF RAILS — American Railway Engineering Association 
 
 The chemical composition of the steel shall be within the following limits: 
 
 BESSEMER PROCESS 
 
 
 70 lbs. and over, 
 but under 85 lbs. 
 
 85-100 lbs. inclusive 
 
 Carbon 
 
 Manganese 
 
 Per cent 
 
 0.40 to 0.50 
 0.80 to 1.10 
 
 0.20 
 
 0.10 
 
 Per cent 
 
 0.45 to 0.55 
 0.80 to 1.10 
 
 0.20 
 
 0.10 
 
 Phosphorus, not to exceed 
 
 (When lower phosphor 
 
 n be secured, a proper proportionate incre 
 OPEN-HEARTH PROCESS 
 
 n carbon should b 
 
 
 70 lbs. and over, 
 but under 85 lbs. 
 
 85-100 lbs. inclusive 
 
 Carbon 
 
 Manganese 
 
 Per cent 
 0.53 to 0.66 
 0.60 to 0.90 
 
 0.20 
 
 0.04 
 
 Per cent 
 
 0.63 to 0.76 
 0.60 to 0.90 
 
 0.20 
 
 0.04 
 
 Phosphorus, not to exceed 
 
 (When higher phosphor 
 
 , a proper proportionate reduction ir 
 
 TABLE LXXXVII. 
 
 arbon should be made.) 
 
 
 81 to 90 Pounds. 
 
 91 to 100 Pounds. 
 
 
 Per cent. 
 
 0.43 to 0.53 
 
 0.10 
 
 0.20 
 0.80 to 1.10 
 
 Per cent. 
 0.45 to 0.55 
 
 0.10 
 
 0.20 
 0.84 to 1.14 
 
 
 
 
 
 OPEN-HEARTH 
 
 STEEL RAILS 
 
 
 
 81 to 90 Pound, 
 
 91 to 100 Pounds. 
 
 
 Per cent. 
 
 0.59 to 0.72 
 
 0.04 
 
 0.20 
 
 . .0.60 to 0.90 
 
 Per cent. 
 
 0.62 to 0.75 
 
 0.04 
 
 0.20 
 0.60 to 0.90 
 
 
 Silicon, not over 
 
 
 
 Table LXXXVIII gives the Pennsylvania specifications revised January 
 10, 1912, for 85-pound and 100-pound carbon steel rails. 
 
 TABLE LXXXVIII 
 
 CHEMICAL COMPOSITION OF RAILS. — Pennsylvania Railroad System. 
 
 BESSEMER STEEL RAILS 
 
 
 Lower Limit. 
 
 Desired Com- 
 
 Upper Limit. 
 
 Carbon 
 
 0.45 
 
 0.80 
 0.05 
 
 Per cent. 
 0.50 
 
 1.00 
 0.12 
 
 Per cent. 
 
 0.55 
 1.20 
 0.20 
 0.10 
 
 Silicon 
 
 
 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 OPEN-HEARTH STEEL RAILS 
 
 
 Caseation A. 
 
 Classification B. 
 
 
 Lower Limit. 
 
 Desired Com- 
 
 Upper Limit. 
 
 Lower Limit. 
 
 Desired Com- 
 
 Upper Limit. 
 
 
 Per cent. 
 
 0.70 
 
 Per cent. 
 0.75 
 
 Per cent. 
 0.83 
 0.80 
 
 0.20 
 0.03 
 
 Per cent. 
 0.62 
 
 0.70 
 
 Per cent. 
 75 
 
 
 80 
 
 
 0.05 
 
 0.12 
 
 0.05 
 
 0.12 
 
 0.20 
 
 
 04 
 
 
 
 
 
 
 
 BIBLIOGRAPHY 
 
 General 
 
 Luty, B. E. V. — Mathematical relations between increases in cost and increases in durability 
 in steel rails. 1500 w. 1911. (In Railway Age Gazette, Vol. 51, p. 1223.) 
 
 Dudley, P. H. — Ductility in rail steel. 1800 w. 1911. (In Railway Age Gazette, Vol. 51, 
 p. 289.) Paper before the American Society for Testing Materials. 
 
 Considers varying composition of rail steel and its influence on the wear. 
 
 Sandberg, Christer P. — Chemical composition of steel rails and latest developments. 
 4000 w. 1908. (In Bulletin of the International Railway Congress, Vol. 22, Part 1, p. 13.) 
 
 The same. (In Engineering, Vol. 83, p. 827.) 
 
 Discusses effect of different elements on quality and wear of metal. 
 
 Manganese Steel 
 
 Rolled manganese steel rail. 1200 w. 111. 1908. (In Railroad Age Gazette, Vol. 45, p. 1536.) 
 
 Gives results of tests and shows comparative life of rails. 
 
 Rolled manganese steel rails. 1600 w. 111. 1909. (In Iron Age, Vol. 83, Part 2, p. 1261.) 
 
 Discusses wear of Manard rail of the Pennsylvania Steel Company. 
 
 Steward, H. M. — Life of manganese steel rail on curves from service tests made on the 
 elevated division of the Boston Elevated Railway Company. 1500 w. 1908. (In Proceedings of 
 the American Street and Interurban Railway Engineering Association, Vol. 6, p. 333.) 
 
 The same. (In Electric Railway Journal, Vol. 32, p. 1196.) 
 
 Titanium Steel 
 
 Dudley, P. H. — Use of ferro-titanium in Bessemer rails. 3000 w. 111. 1910. (In Journal of 
 Industrial and Engineering Chemistry, Vol. 2, p. 299.) 
 
 Gives ductility tests of ferro-titanium rails showing them to average several per cent higher 
 than ordinary Bessemer rails in ductility. Believes that range of ductility can be prescribed by 
 proper study of chemical composition. 
 
 Maltitz, Ed. von. — Der einfluss des titans auf stahl, besonders auf schienenstahl. 6000 w. 
 111. 1909. (In Stahl und Eisen, Vol. 29, Part 2, p. 1593.) 
 
 The same, condensed. 1200 w. (In Iron Age, Vol. 84, Part 2, p. 1790.) 
 
 Gives results of experiments on effect of additions of titanium to Bessemer rail steel. 
 
 Slocum, Charles V. — Titanium alloy in rails and car wheels. 6000 w. 111. 1909. (In 
 Proceedings of the Railway Club of Pittsburg, Vol. 8, p. 176.) 
 
 Emphasizes the increased wear and soundness of titanium rails. 
 
 Slocum, Charles V. — Use of titanium in steel for rails, car wheels, etc. 2000 w. 111. 
 1909. (In Electrochemical and Metallurgical Industry, Vol. 7, p. 128.) 
 
 Shows the increased durability and strength of titanium steel and its products. 
 
 Springer, J. F. — Titanium steel. 2000 w. 111. 1911. (In Cassier's magazine, Vol. 40, p. 483). 
 
 Considers especially its properties and importance as rail steel. 
 
 Thompson, A. W. — Use of titanium rail on the Baltimore and Ohio Railroad. 2500 w. 
 1909. (In Canadian Engineer, Vol. 17, p. 238.) 
 
344 STEEL RAILS 
 
 Gives properties and tests. 
 
 Waterhouse, G. B. — Influence of titanium on segregation in Bessemer rail steel. 3000 w. 
 111. 1910. (In Proceedings of the American Society for Testing Materials, Vol. X, p. 201.) 
 
 Results indicate that presence of titanium in rail steel lessens segregation and promotes 
 uniformity. 
 
 30. Extraction of the Iron from Its Ore 
 Before the process of reduction or " smelting " is attempted at the blast 
 furnace the ore is usually subjected to some preliminary treatment.* 
 The preparatory processes are: 
 
 (a) " Grading " the ore; 
 
 (b) Calcination or roasting; 
 
 (c) Mixing to make up the desired proportions of ore charge. 
 
 The grading of the ore is not necessary at the furnace when it has already 
 been properly done at the mine. When the sorting at the mines has not been 
 carefully done, or when a greater number of grades than usual are required, 
 sorting is also practiced at the furnace, and the ore is then distributed to the 
 several bins of the stock house, which building is erected as near the furnace 
 stack as possible. 
 
 f The purpose of roasting is to remove sulphur, carbonic acid, and water 
 and to increase the porosity of the ore. It is accomplished in two ways, — ■ 
 by roasting the ore in a heap, or in a kiln using wood, coal, or gas for fuel. 
 Fig. 229 shows the ore roasters used at the Norway furnace, Bechtelsville, Pa., 
 in 1883.J 
 
 Lake Superior ores require no roasting, and for this reason very little roasting 
 of the ore is necessary at the present time. The iron ores in the vicinity of 
 Johnstown, which were formerly used by the Cambria Steel Works, contain 
 high sulphur and phosphorus content. The iron content in the ore was but 
 30 per cent, necessitating roasting before charging into the furnace. These 
 works now use Lake Superior ores having an iron content of from 50 to 65 per 
 cent, and the process of roasting is not necessary. 
 
 Making up the furnace charge is an operation which demands both a knowl- 
 edge of the chemistry of the blast furnace and of ores. The proportions of 
 the charge are determined by the character of the ore, the fuel, and the flux, by 
 the size and method of working the furnace, and by the character of product 
 required. 
 
 * Iron and Steel, Materials of Engineering, Thurston, Part 2, 1909, p. 91. 
 t Notes on Iron, Richards. 
 
 t Roasting Iron-Ores, by John Birkinbine. Trans. American Institute of Mining Engineers, 
 Vol. XII (1883-4), pp. 361-379. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 345 
 
 The location of the plant is usually chosen according to the cost of assembling 
 these materials and getting the product to the market. Other things being 
 equal, that furnace will be most economically located which is placed near the 
 mines. Where the ores and fuel are widely separated, location is often deter- 
 mined by the facilities for marketing the iron, and the furnace is so placed that 
 the total of all the costs of transportation and of working shall be a minimum. 
 
 Fig. 229. — Ore Roasters, Norway Furnace, 
 
 (Am. Inst, of Mining Engineers.) 
 
 If the quantities transported are 0', 0", 0'" respectively, and the cost 
 of carriage is c dollars per ton, the distance for each being S', S", S'", the total 
 cost (Thurston), 
 
 K = cO'S' + cO"S" + cO'"S'", 
 
 should, other things being equal, be made a minimum. 
 
 The notable present tendency in the iron industry is the lower average iron 
 content in the ores used. * This tendency will undoubtedly continue in the 
 future as the more easily accessible portions of the richer deposits are worked out. 
 As a corollary to this is the observed tendency toward a decentralization of the 
 
 * Iron Ores of the United States. Report of the National Conservation Commission, Vol. Ill, 
 p. 483, February, 1909. Government Printing Office, Washington. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 347 
 
 .a h 
 1.« 
 
 M 
 
348 
 
 STEEL HAILS 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 349 
 
 iron industry, and with a decrease in the iron content of the ore used, involving 
 a corresponding increase in cost of transportation per unit of iron, there will be 
 an increase in the proportion of fuel which goes to the region producing the ore. 
 Sir I. Lowthian Bell in 1884 stated * that while "Wages (in America) are 
 high . . . the geographical position of the ore and coal and of the markets 
 themselves constitute obstacles of a far more insurmountable description. The 
 distances over which ore is conveyed are sometimes very great; as an example, 
 
 Fig. 233. — Steamer " Augustus B. Wolvin," 560 ft. in length, capacity about 12,000 tons 
 
 the produce of the Lake Superior mines is carried to Pittsburg, involving car- 
 riage of 790 miles. The cost of transport on the minerals consumed for 
 each ton of pig iron I have calculated f to average 10s., 9d., at the eight chief 
 seats of the iron trade in Great Britain; whereas, in the United States the mean 
 charge at fourteen of the large centers is 25 s., 8 d." The introduction of improved 
 methods for handling the ore in transport and the deepening of the waterways of 
 the Great Lakes J has in a measure overcome the adverse conditions mentioned 
 above. 
 
 * Manufacture of Iron and Steel, Bell, London, 1884, p. 473. 
 
 t Report to Her Majesty's Government on Iron Manufacture of the United States compared 
 with that of Great Britain. 
 
 t William Chandler, History of St. Mary's Falls Ship Canal, 1877. The Great Lakes and Our 
 Commercial Supremacy, John Foord, North Am. Review, Vol. 167, p. 155. Saint Mary's Falls Canal 
 Semicentennial, History of the Canal, John H. Goff, 1907. 
 
350 
 
 STEEL RAILS 
 
 General View of the Dock. 
 
 Side View of the Dock with Ore Cars on the Structure. 
 
 Fig. 234. — Great Northern Railway Ore Dock at Allouez Bay, Superior, Wis. 
 (From Science Conspectus.) 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 of the Lake Superior ore 
 
 Figs. 230 to 234 illustrate some of the features 
 industry. Figs. 230 and 231 show the 
 method of mining the ore by steam 
 shovels employed in northern Minne- 
 sota. The shovels are large, with 
 about 5-ton dippers. The amount of 
 stripping required at these mines is 
 often heavy, amounting in some cases 
 to as much as 100 feet and costing 
 from $0.25 to $0.40 per cubic yard of 
 material removed from on top of the 
 bed of ore. It is generally considered 
 profitable to strip up to a maximum 
 depth which does not exceed the thick- 
 ness of the layer of ore uncovered. 
 
 Figs. 232, 233, and 234 show the 
 ore docks and the type of vessels used 
 in transporting the ore. * Down to 
 late in the fifties the ore product of 
 Lake Superior was handled over a 
 mule-tram road to Marquette, and as 
 late as 1870 a 700-ton ship was an 
 enormous craft, the loading of which re- 
 quired two days and the unloading be- 
 ing seldom accomplished in that time. 
 
 In 1871 the largest ore barge 
 carried 1050 tons, now the cargoes 
 reach 14,000 tons. 165,000 tons of 
 ore has been loaded into sixteen steam- 
 ships in one day at the docks of the 
 Duluth, Missabe and Northern Rail- 
 way. The loading of the steamer " H. 
 E. Corey " of 10,000 tons capacity at 
 the Duluth and Iron Range Steel 
 Ore Dock, at Two Harbors, Minn., 
 was accomplished in 39 minutes, f 
 
 * The Development of Lake Superior Iron Ores. Bacon. Trans. American Institute of Mining 
 Engineers, Vol. XXVII (1897), p. 341. f Scientific American, December 11, 1909. 
 
352 
 
 STEEL RAILS 
 
 The construction of a special type of ship of large tonnage for ore trade, 
 coupled with the invention of unloading machinery of great capacity at the 
 
 terminal ports, has brought the 
 cost of transportation down to 
 a very low figure. Thus, a ton 
 of ore is now hauled one hun- 
 dred miles by rail from the 
 most distant mines in the Lake 
 Superior range to a Lake Su- 
 perior port, is loaded into cars 
 or into the stock pile at a Lake 
 Erie port at a cost of less than 
 $1.80 per ton. 
 
 Fig. 235 presents an in- 
 board profile and cross section 
 of the " Wolvin," a representa- 
 • tive of the type of present ore 
 steamers. This vessel is 560 
 The largest single cargo of ore 
 carried by the " Wolvin " was 11,536 tons, a feat which she performed in 1904. 
 
 - Ten-ton Bucket of Unloader in Hold of the " 
 vin." (Scientific American.) 
 
 feet in length, 56 feet beam and 32 feet deep. 
 
 Fig. 237. — General View of Ore Unloader with Steamer at the Dock. (Railroad Age Gazette.) 
 
 The " E. H. Gary " in 1905 carried a single cargo of 12,368 tons. In 1906 the 
 "J. P. Morgan" carried a single cargo of 13,272 tons of ore and in 1907 she 
 carried 13,800 tons. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 353 
 
354 
 
 STEEL RAILS 
 
 Fig. 236 shows the bucket of the Hulett ore unloader. Four of these 
 machines located at the docks at Conneaut, Ohio, are credited with having 
 taken out of the "Wolvin" 7257 gross tons of ore in four hours and six minutes.* 
 The Hulett unloaders at Gary are showing an average rate of 300 tons per hour 
 
 Fig. 239. — ■ Blast Furnace with Stoves and Buildings. (Thurston.) 
 
 for each machine. On July 10, 1912, the "Morgan" discharged 10,091 tons of ore 
 in three hours and ten minutes at Conneaut. This was apparently the fastest time 
 ever made in unloading, but on July 24, 1912, it was surpassed when the "Wm. 
 P. Palmer" was relieved of 11,044 tons in three hours and seventeen minutes, or at 
 the rate of 56 tons per minute. 
 
 * Saint 
 1907, p. 201. 
 
 Gary's Falls Canal Semicentennial, Commerce of the Great Lakes, Ralph D. Williai 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
356 
 
 STEEL RAILS 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 357 
 
 * The coal and ore docks of the Baltimore and Ohio Railroad at Lorain, 
 Ohio, are among the largest on the Great Lakes. The machinery for unload- 
 ing the steamers is the latest design of Brown hoist unloader, driven by elec- 
 tricity and equipped with three grab buckets having a total capacity of 1000 
 tons of ore an hour. Figs. 237 and 238 illustrate a steamer being unloaded at 
 the dock. 
 
 The large grab buckets employed scoop up from seven to ten tons of ore 
 each time they are lowered into the hold of the vessel, after which they are 
 hoisted and carried in over the dock on a movable girder, or ram, carried in a 
 heavily braced portal frame, which is itself movable lengthwise of the dock. 
 
 The buckets may either be dumped into a 75-ton weighing hopper, from 
 which the ore is discharged directly into cars on any one of the four tracks 
 spanned by the unloader, or dropped 
 into the trough space, which has a 
 capacity of 100,000 tons, and is sepa- 
 rated by a concrete wall from the 
 tracks. Once deposited in the trough, 
 the ore may either remain in tempo- 
 rary storage, or be conveyed to the 
 larger storage space covered by the 
 ore bridge. 
 
 The combination of fast unload- 
 ing plants on the dock front with 
 buckets moving at high speed over a 
 short travel, with a storage bridge of 
 long span, carrying a larger bucket 
 over the storage space, is found on 
 all modern lake docks. 
 
 The blast furnace is shown by 
 Figs. 239, 240, and 241. It is a brick 
 structure, usually circular in section 
 and built in two parts; the upper part resting on columns, while the lower 
 portion rests directly on the foundation. The upper portion is sheathed with 
 boiler plates. Fig. 242 shows the top rigging of a modern blast furnace. The 
 charge is automatically elevated and dumped into the hopper. 
 
 In the United States furnaces are worked up to 100 feet high. The best 
 modern practice is, however, about 90 feet high, with a product of 400 to 500 
 
 * Railway Age Gazette, July 28, 1911, p. 178. 
 
 Fig. 242. — Top Rigging of Blast Furnace. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 359 
 
 tons per day. The following dimensions of the Gary furnaces are typical of 
 the best practice. The blast furnaces (Fig. 243) are 88 feet in height from the 
 tap hole to the top of the furnace lining, and the capacity of each is 450 tons- per 
 day. Each furnace has four blast stoves. The interior diameter of the blast 
 furnace is 15 feet at the hearth, 21J feet at a height 13 to 21 feet above the hearth, 
 and 16 feet at the top. 
 
 Fig. 244. — The Whitwell Hot-blast Stove. (Thurston.) 
 
 The earlier blast furnaces were blown with cold air, but later a hot blast 
 was used with an aim to saving fuel, and the air from the blowing engines 
 passed through stoves which were heated by the waste gases from the furnace. 
 Fig. 244 shows the Whitwell stove and Fig. 245 a more modern stove. 
 
 The first stoves in use were of cast-iron. The gases were burned around 
 and circulated among U-shaped cast-iron pipes enclosed in a fire-brick struc- 
 ture. This process was continuous — a recuperative process. However, it was 
 
360 
 
 STEEL RAILS 
 
 subject to a number of defects, among which was the burning out of the tubes, 
 making it impossible to obtain more than 900° F. in the blast. This type of 
 stove was followed by the fire-brick stove operated on the regenerative princi- 
 ple, and by its use a hot-blast tem- 
 perature of approximately 1500° F. can 
 be obtained. 
 
 The atmosphere is the most vari- 
 able element involved in the blast- 
 furnace process, which consumes air 
 in large quantities. In furnaces 
 using ore from the Lake Superior 
 district the raw material, amounting 
 to about 7200 pounds per ton of iron, 
 varies in composition within 10 per 
 cent, but the atmosphere, of which 
 11,700 pounds are consumed per ton 
 of iron, varies in its content of 
 moisture from' 20 to 100 per cent 
 from day to day and often in the 
 same day. 
 
 Many experiments have been 
 made to determine the most feasible 
 method for extracting the moisture 
 from the air. Various schemes for 
 its direct absorption were worked out 
 and in turn abandoned, and finally 
 Mr. Gayley* designed and put in 
 successful operation the dry-blast 
 process which bears his name. This 
 consists in freezing the moisture out 
 of the air. The Gayley process not only reduces the cost of producing the 
 pig iron, but, which is very much more important, gives a more effective 
 control of the operation and product of the furnace. 
 
 The product was first put in operation on the Isabella furnaces of the 
 
 * The Application of Dry-air Blast to the Manufacture of Iron. James Gayley. Trans. American 
 Institute of Mining Engineers, Vol. XXXV (1905), p. 746. 
 
 The Application of Dry-air Blast to the Manufacture of Iron — Supplementary Data. James 
 Gayley. Ibid, Vol. XXXVI (1906), p. 315. 
 
 Gayley's Invention of the Dry Blast. R. W. Raymond. Ibid, Vol. XXXIX (1908), p. 695. 
 
 Fig. 245. — Julian Kennedy Stove. 
 (Harbison- Walker Refractories Co.) 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 361 
 
 J" 
 
 Carnegie Steel Company, situated at Etna, Pa., a suburb of Pittsburg, on August 
 11, 1904. The lines and dimensions of this furnace, shown in Fig. 246, repre- 
 sent the usual construction of furnaces in the Pittsburg 
 district. 
 
 Fig. 247 shows graphically the operation of each 
 day, averaged with all the preceding days from August 1 
 to September 9, 1904, nclusive; the increase in output 
 and reduction in coke consumption corresponding to the 
 increase in burden; the varying conditions of humidity 
 from day to day, which represent the average humidity 
 for each twelve-hour period ; and the change in humidity 
 after treatment in the dry-blast apparatus. 
 
 The materials for smelting are iron ores, limestone 
 (flux), and fuel. Charcoal was first used and the iron 
 from this fuel was of excellent quality on account of the 
 low ash and sulphur of the charcoal and its great 
 porosity. It has so little strength, however, that its use 
 in the modern high furnaces is prohibited. 
 
 Coke is now generally used. Anthracite as a blast- 
 furnace fuel is inclined to decrepitate and give trouble 
 from its fineness. Bituminous coal is not used, as it 
 cakes and absorbs heat for distillation of volatile consti- 
 tuents. 
 
 * At Gary, Plate XXIX, between the stock pile and 
 the furnace is a line of elevated storage bins arranged in 
 two parallel rows. One row is for coke and the other 
 for ore and limestone. Above the bins are four tracks 
 on which travel two 60-ton electric transfer cars. The 
 ore is loaded into the transfer cars by the buckets of the overhead ore bridges. 
 The coke and limestone are brought up over the bins by rail and deliver their 
 load directly by gravity. 
 
 At the bottom of the bins are spouts controlled by electrically operated 
 gates, and below these are tracks which run the full length of the bins. Traveling 
 on these tracks are electrically operated lorries into which the ore, coke, and 
 limestone are delivered froii the bin spouts. The lorries carry the materials 
 to what are known as the " furnace skips," of which there is a pair to each 
 furnace. The skips run upon an inclined railway which runs direct from a pit 
 
 * Scientific American, December 11, 1909. 
 
 JB£f£_ 
 
 Fig. 246. — Isabella Fur- 
 nace, Carnegie Steel 
 Company. (Am. Inst, 
 of Mining Engrs') 
 
362 
 
 STEEL RAILS 
 
 
 
 
 0. j 
 
 4- -1 JL JL 
 
 Ol 
 
 eo P 
 
 t 4 14 
 
 CO 
 
 p 
 
 4 ^4 4 
 
 r- 
 
 S> 4 
 
 4 S I 
 
 <D 
 
 t^ 
 
 it _j l f 
 
 •0 
 
 ^ U 
 
 \ y } \ 
 
 * 
 
 CO P 
 
 t t v \ 
 
 <0 
 
 U 
 
 rj r t 
 
 d 
 
 m- [ 
 
 t \ \ i 
 
 
 "s 4 
 
 IT j 1 4 
 
 5 
 
 8 4 
 
 4 4 i- 5 
 
 8 
 
 s 4 
 
 I 4lZL ■ 
 
 SI 
 
 s 4 
 
 I Ut 4 
 
 s 
 
 s 4 
 
 t >v 4 
 
 5 
 
 s 4 
 
 4 -^ x 4 4^ 
 
 s 
 
 s 4 
 
 v X I t 
 
 8 
 
 s 4 
 
 7 -tl t 
 
 S 
 
 s 4 
 
 r ^rt t 
 
 a 
 
 s 4 
 
 E i'jt i 
 
 a 
 
 s t _J 
 
 ^5 5 -t 
 
 s 
 
 S t -y 
 
 vtt t 
 
 8 
 
 2 I 1 
 
 ^ f 4 
 
 2 
 
 
 
 2 
 
 s r r 
 
 t p r 
 
 £ 
 
 2 i -l 
 
 4 H I 
 
 2 
 
 
 
 12 
 
 s - j 
 
 2 II 4 
 
 * 
 
 2 4 
 
 4 5 V 
 
 2 
 
 « Ct— fv 1 
 
 \k X 
 
 2 
 
 
 / 
 
 v 
 
 2 ^ 1 
 
 14 
 
 
 
 IT 
 
 i o 
 
 » it 1 4 
 
 \ ^ 
 
 <x> 
 
 
 i i> 
 
 "• 
 
 
 =3 
 
 <<o 
 
 «> §£ £ 4 
 
 14 
 
 So 
 
 * >1 g T 
 
 11 
 
 1 * 
 
 It TT 
 
 &4 
 
 5<m 
 
 14 -IZ 
 
 12 
 
 S'CM 
 
 g- ns* 2- 
 
 o o oooooc 
 
 oow mooooooc 
 OOk."^™ oooooc 
 
 3 ri 
 
 8 
 
 'now jo Nouonaoaa noi uad 3xoo d< 
 
INFLUENCE OF DETAIL OF MANUFACTURE 363 
 
 below the transfer cars to the charging platform at the top of the blast 
 furnaces. 
 
 The operation is entirely automatic. Each trip of the skip is made in about 
 sixty seconds, and its average load consists of about 7000 pounds of ore, or 
 6000 pounds of limestone, or 3600 pounds of coke. 
 
 At Cambria* from the port of entry the ore is hauled to the works and there 
 unloaded by a car-dumping machine, and again handled and stacked by a travel- 
 ing gantry. From the stock pile it is brought back to special bin cars that are 
 run into the charging house, where they take the place of the usual bins and 
 discharge into hoppers that empty into the loading skips, which are hoisted to 
 the charging door at the top of the furnace. 
 
 Slag is run off either continuously or at short intervals. The iron is tapped 
 regularly into a runner, through which it flows either into the molds of the pig- 
 bed or else into the direct metal ladles. 
 
 To control the kinds of pig iron produced by a furnace, we can vary the 
 composition of the slag, and change the burden. The burden is made heavy by 
 increasing the amount of ore and flux to a charge. 
 
 Mr. Wickhorstf gives the following as the day's burden of "A" blast fur- 
 nace at the Maryland Steel Company: 
 
 Tons. 
 
 El Cuero ore 346.299 
 
 Nicolaieff ore 60.676 
 
 Sierra Morena ore 17.759 
 
 Coke 346.875 
 
 Limestone 82.366 
 
 Dolomite 82.266 
 
 The records of the ore analyses were as follows : 
 
 
 El Cuero. 
 
 Nicolaieff. 
 
 Iron, natural 
 
 Iron, dried at 212° F 
 
 Per cent. 
 58.31 
 59.52 
 2.03 
 9.52 
 .32 
 .017 
 
 Per cent. 
 
 65.75 
 66.86 
 1.66 
 2.16 
 .13 
 .017 
 
 
 Silica.. 
 
 
 Sulphur 
 
 
 No nickel, cobalt, copper or chromium in either one. 
 
 The metal from the blast furnace was poured into an 85-ton receiver, from 
 which it was weighed and poured into an 18-ton converter. 
 
 The same authority gives the following for the blast-furnace practices at 
 
 * Railway Age Gazette, August 19, 1910. 
 
 f Report to Rail Committee, Proceedings Am. Ry. Eng. & M. of W. Assn., Vol. 12, Part 2, 1911. 
 
STEEL RAILS 
 
 the Gary works. Lake Superior ore was used, reduced in blast furnaces using 
 ordinary air and having the following average charge: 
 
 Coke.. 
 Stone.. 
 Ore ... . 
 
 Pounds. 
 12,000 
 
 5,100 
 26,500 
 
 Shell 
 
 PLAN. SECTION. 
 
 Fig. 248.— 300-ton Mixer. E, Filler; H. Pouring Spout. (Harbord and Hall.) 
 
 A mixture of ores was used, but consisted largely of Chapin ore, showing 
 analysis as follows: 
 
 Per cent. 
 
 Iron 54 . 63 
 
 Aluminum (about) 2 . 00 
 
 Manganese .21 
 
 Phosphorus 059 
 
 Sulphur trace 
 
 Silica (SiO») 5.51 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 365 
 
 The limestone used analyzed about 
 
 aS follOWS: Percent. 
 
 Calcium oxide 54.67 
 
 Magnesium oxide 
 
 Iron and aluminum oxides . 
 Silica 
 
 .53 
 .74 
 
 The iron from the blast furnaces 
 was poured into a mixer, from which 
 it was weighed. 
 
 The direct process is one by which 
 the steel is made from the ore in one 
 operation. In ordinary coke blast- 
 furnace practice successive casts vary 
 too much in Si and S to allow of taking 
 the metal as it flows from the furnace 
 into ladles and from them to the con- 
 verter or open-hearth furnace. It is, 
 therefore, poured first into large reser- 
 voirs or mixers, the casts from different 
 furnaces being mixed together. 
 
 Capt. Wm. R. Jones was the first 
 to use the mixer in anything like the 
 form which has now become universal 
 practice. He built and successfully 
 operated his mixer at the Edgar Thom- 
 son Works, and although it may have 
 previously been used in some modified 
 form the introduction and practical use 
 of the mixer in connection with the Bes- 
 semer process is apparently due to his 
 efforts. 
 
 The larger the mixer the better 
 are the results obtained, both with re- 
 spect to the purification and also in 
 retaining the available heat of the 
 metal. Mixers capable of holding 
 600 tons of metal are now in use. 
 Fig. 248 shows the general arrange- 
 ment of a 300-ton metal mixer in use 
 at the Cambria Iron Company's Works. 
 
 -Ten-foot Iron Cupola, Maryland 
 Steel Company. 
 
366 STEEL RAILS 
 
 The molten metal for supplying the converter when not taken direct from 
 the blast furnace or the mixer is melted in cupolas. The modern cupola is 
 really a small blast furnace, as shown in Fig. 249. 
 
 31. Conversion of the Steel 
 The principal points in connection with the conversion are given below: 
 
 1. The temperature of the conversion must be controlled; 
 
 2. The recarbonizer must be thoroughly mixed; 
 
 3. Time and opportunity must be allowed for the escape of gases im- 
 
 prisoned in the molten steel. 
 
 There are three methods of converting the metal from the blast furnace 
 into steel: the Bessemer converter, the Open-hearth furnace and the Electric 
 furnace. The first or Bessemer process has had an important influence upon 
 railroad history. The out-growth of an attempt to make wrought iron cheaply, 
 it came in just at the time when the wrought-iron rail was beginning to demon- 
 strate its unfitness to stand the increased wheel loads then coming into use. It 
 perhaps is not too much to say that the Bessemer steel rail has made the modern 
 railroad possible, and that without it, or its equivalent, the world's development 
 would be half a century behind its present advanced position. 
 
 The pneumatic method of steel making, generally known as the Bessemer 
 process, was until the last few years the most extensively practiced and the most 
 productive, by far, of all known methods of making ingot metal. 
 
 * It has been known since the time of Cort that the agitation of molten 
 cast iron in presence of oxygen will produce combustion and removal of carbon, 
 and the reduction of the cast iron to the state of malleable iron or of steel. 
 
 The pneumatic process secures such an agitation and a very thorough inter- 
 mixture of the fluid iron with the oxidizing atmosphere, by causing the latter to 
 stream up through the molten mass in innumerable minute bubbles; the rapid 
 combustion thus secured is sufficient to supply all heat needed, not only to retain 
 the metal in a fused condition, but, also, so rapidly and so greatly to elevate its 
 temperature during the operation that the product, even when entirely deprived 
 of carbon, remains a perfectly fluid wrought iron in the converting vessel. 
 Fig. 250 shows earlier experiments of blowing air through the bath.f 
 The process was invented independently by Henry Bessemer, in Great 
 Britain, and by William Kelly, in the United States. 
 
 * Iron and Steel, Materials of Engineering, Thurston, New York, Part 2, 1909, p. 241. 
 
 t Sketch of the Origin of the Bessemer Process, by Sir Henry Bessemer. Trans. American 
 Society of Mechanical Engineers, Vol. XVIII, 1897, p. 455. 
 
INFLUENCE OF DETAIL OF MANUFACTURE. 
 
 367 
 
 * It is even claimed for America that it was the birthplace of the pneu- 
 matic process of steel making, Kelly having begun a series of experiments based 
 upon this theory as early as 1851. As Kelly, soon after Beessmer's patent 
 was taken out, succeeded in showing that he had previously had similar views, 
 Bessemer's patent rights in the United States became limited to certain me- 
 chanical arrangements, and a lawsuit arose between the company which bought 
 
 Fig. 250. — Early Experiments of Blowing Air through Bath. (Am. Soc. M. E.) 
 
 Bessemer's patent rights for that country and that which took over both Kelly's 
 right and Munshet's patent for taking away the red-shortness of the final 
 product by the addition of spiegeleisen. This lawsuit, together with the Civil 
 War, prevented the development of the Bessemer process in the United States 
 up to the year 1866, when an agreement was at last entered into between the 
 two companies. Both these companies had, indeed, before this time con- 
 structed their experimental works; but it was only after the compromise was 
 concluded between the two companies that there could be any steps taken for 
 erecting Bessemer works on a larger scale. 
 
 * Steel: Its History, Manufacture, Properties, and Uses, J. S. Jeans, London, 1880, p. 144. 
 
368 
 
 STEEL RAILS 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 369 
 
 
 Fig. 252. — Plan. 
 
 American 5-ton 
 
 Fig. 253. — Section on Line HK. 
 Plant. (Thurston.) 
 
 The plant shown by Fig. 251 may be taken as illustrative of an efficient 
 arrangement. The general arrangement of the Bessemer plant is shown in the 
 accompanying drawings. Fig. 252 represents the ground plan as designed by 
 
 
 - ' nHffij 
 
 ^3r 1- 
 
 
 Fig. 254. — Arrangement of Converters at Maryland Steel Company. (Am. Soc. of Mech. Engrs.) 
 
370 STEEL RAILS 
 
 Holly, and Fig. 253 is a section laterally on the center line of the pit surround- 
 ing the converter. 
 
 The pig iron is melted in cupolas, A, A, A, A, Fig. 252 in plan, and seen 
 in elevation in the section resting upon the second floor of the converting house, 
 at the right of the converters, C, C. Materials are hoisted from the lower levels 
 by hydraulic elevators placed at each end of the charging floor, the one for fuel, 
 the other for metal. 
 
 Fig. 255. — 18-ton Converter, Maryland Steel Company. (Am. Soc. of Mech. Engrs.) 
 
 Figs. 254 and 255 show 18-ton converters at the Maryland Steel Works, 
 at Sparrows Point, Md., in 1897. Fig. 256 illustrates a typical English Bessemer 
 converter, while Figs. 257 and 258 show converters in operation. 
 
 The blow generally requires about ten minutes. The molten iron is 
 poured into the converter when it is lying on its side, as shown by Fig. 257, the 
 converter is then placed in a vertical position and the air, compressed to about 
 20 pounds per square inch, is turned on, the pressure of the blast being suf- 
 ficient to prevent the molten metal entering the tuyeres in the bottom of the 
 converter. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 ? 3 i 1-2 1 
 
 3 O <u *= ^ T3 
 
 
 £ 
 
 > 'S S -5 a S 
 
 £3 
 
 || f 
 
 !_ '' i 
 
 
 
 u 
 
 
372 
 
 STEEL RAILS 
 
INFLUENCE OF DETAIL OF MANUFACTURE 373 
 
 Fig. 258. — Bessemer Converter in Full Blast. (Am. Tech. Soc.) 
 
374 STEEL RAILS 
 
 The molten pig iron contains a large proportion of carbon which is almost 
 burned out during the blow. The combustion of this carbon increases the heat 
 of the metal and the flame, shown in Fig. 258, is at first red, but rapidly be- 
 comes brighter until it can hardly be looked upon by the naked eye. The sudden 
 dropping of the flame after nine or ten minutes gives evidence that the carbon 
 is almost burned out, and the operator turns the converter down and shuts 
 off the blast. Spiegeleisen or ferromanganese is then added to recarbonize the 
 metal. 
 
 Mr. Wickhorst* gives the following description of the process of making 
 Bessemer steel at the Maryland Steel Company: The metal from the blast 
 furnace was poured into an 85-ton receiver, from which it was weighed and 
 poured into an 18-ton converter. In addition to the hot metal from the blast 
 furnace, cupola metal was used, which ordinarily is the same metal that has 
 been run into pigs and then remelted in a cupola, this being necessary when 
 the Bessemer plant cannot take care of all the metal from the blast furnaces. 
 In the case of this heat, two-thirds of the cupola metal was Lebanon iron. The 
 converter charge was as follows: 
 
 Pounds. 
 
 Metal from No. 2 receiver 22,500 
 
 Cupola metal 18,000 
 
 Scrap steel 1,000 
 
 After blowing, 4300 pounds of spiegel was added to the converter and 260 pounds 
 ferromanganese and 30 pounds ferrosilicon added to the ladle during the pour- 
 ing. The analyses of the metal in the converter before starting to blow, and 
 before the addition of scrap and of spiegel, were as follows, special samples 
 being taken for these analyses: 
 
 Carbon 
 
 Phosphorus 
 
 Sulphur 
 
 Manganese 
 
 Silicon 
 
 Copper 
 
 Nickel or chromium 
 
 .064 
 trace 
 3.96 
 
 The basic open-hearth is rapidly supplanting the Bessemer process. This is 
 probably due to the supply of low phosphorus ores being exhausted and the re- 
 duced price of scrap, as on account of the great capacity of the Bessemer process 
 the open-hearth would otherwise have little chance. 
 
 * Report to Rail Committee, Proceedings Am. Ry. Eng. & M. of W. Assn., Vol. 12, Part 2, 1911. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 375 
 
 The first experiments which eventually led to the development and per- 
 fection of the open-hearth process were carried on by Josiah Heath about 1845. 
 Siemens began his experiments about 1861, while at the same time Martin was 
 independently working on the same problem in France. The first open-hearth 
 furnace introduced into the United States for the production of steel was built 
 by Frederick J. Slade for Cooper, Hewitt and Company, then proprietors of the 
 New Jersey Steel and Iron Company, at Trenton, N. J. 
 
 This method is sometimes regarded as one of decarburization of cast iron 
 by the addition of uncarborized metal ; it must not be forgotten, however, that the 
 carbon and silicon of the molten pig metal are not entirely taken out or neutralized 
 by the addition of the uncarborized metal, but that the oxidizing flame from the 
 gas which is burned in the furnace plays an important part. One of the principal 
 objects in adding a large amount of scrap is to save time and cost of fuel in the de- 
 carborizing and desiliconizing process as well as to also save the lining of the furnace. 
 
 * The furnace (Fig. 259) consists of a rectangular bath, hearth, or basin, 
 open at each end for the admission of gas and air at the ports. This hearth is 
 arched by a roof from 9 inches to 12 inches in thickness. At each end of the 
 furnace are two checker chambers, one for the preheating or regeneration of 
 the air, the other of the gas. Before starting the furnace a wood fire is built 
 in one set of chambers (or in the furnace) and after these have attained a dull- 
 red heat the gas and air are passed through them, entering at one end of the 
 furnace, are deflected downward by the direction of the ports, unite in com- 
 bustion over the hearth, and the gases, the products of combustion, leave the 
 furnace through the ports at the opposite end, passing downward through the 
 checkers or regenerative chambers, there giving up their heat to the checkers, 
 thence through the flues to the stacks. 
 
 At frequent intervals, say from 15 to 20 minutes, dependent on the quality 
 and amount of fuel, charge, working of furnace, etc., the currents of gas and 
 air are reversed, now entering the furnace at the opposite ends and having 
 passed through the checker chambers, heated up during the previous period, 
 take this stored-up heat to create a more intense flame over the bath. These 
 waste gases in turn pass out through the chambers, giving up their heat. This 
 reversal is maintained with regularity until the charge is ready to tap. 
 
 Fig. 260 illustrates the general arrangement of an open hearth plant. 
 
 The Talbot continuous open-hearth process employs a tilting furnace 
 which may be operated at a capacity of from 20 tons upward; 100 to 150 or 
 even 200 tons are entirely practicable. The charge is run in from the cupola, 
 
 * A Study of the Open Hearth, Harbison-Walker Refractories Company, Pittsburg, 1909. 
 
376 
 
 STEEL RAILS 
 
 blast furnace, or mixer, desiliconized wholly and decarbonized largely by a 
 blanket of slag rich in oxides, and reduced ultimately in the usual way. The 
 
 GAS 
 
 TRANSVERSE SECTION OF LONGITUDINAL SECTION AT TRANSVERSE SECTION AT 
 
 REGENERATIVE CHAMBERS CENTER LINE OF FURNACE CENTER LINE OF FURNACE 
 
 Fig. 259. — Modern Open-hearth Furnace. (Harbison- Walker Refractories Co.) 
 
 charge is run off and its place supplied by a new charge, the bottom being at 
 no time allowed to become exposed. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 377 
 
 Figs. 261 and 262 show tilting open-hearth furnaces. Fig. 261 shows a 
 Wellman tilting open-hearth furnace and Fig. 262 was taken at the Jones and 
 
 Laughlin Steel Company's plant, where the Talbot open-hearth process is 
 employed. More Talbot tilting open-hearth furnaces have been installed, both 
 in this country and in England, than the Wellman furnace. 
 
378 
 
 STEEL RAILS 
 
 The tilting furnaces do away with a great portion of the tap-hole troubles, 
 the taphole being above the metal and slag lines with the furnace in the normal 
 position, and it is consequently only necessary to fill the tap hole with a very 
 light tamping. They also enable the melter to thoroughly drain the furnace 
 bottom of any slag or metal, it being in the stationary furnace often a difficult 
 matter to rabble or splash out all depressions, and any portion of the heat left 
 
 ""™'«Lani™— . — — -■■ "i '■' , : ■'''"S^"-*. <*£& f S/jf Wm 7 ~-"-~~'. 
 
 Fig. 261. — Wellman Tilting Open-hearth Furnace. (Am. Tech. Soc.) 
 
 in such a hole very soon tends to permeate and disintegrate the surrounding 
 bottom 
 
 The process of open-hearth steel production* at the Gary works is illus- 
 trated by the following description of an open-hearth heat. This consisted of 
 charging limestone and ore into a basic open-hearth furnace heated with pro- 
 ducer gas and piling on scrap. The charging was started at 7.29 a.m. After 
 2| hours liquid mixer metal was added, and the whole was melted down until 
 
 * Report of Tests of Open-hearth Rails 
 c M. of. W. Assn., Vol. 12, Part 2, 1911. 
 
 - Gary Works. Wickhorst. Proceedings Am. Ry. Eng. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 379 
 
 the tapping test showed carbon .26 per cent by fracture. During the melting 
 small quantities of fluor spar were added at intervals to make the slag more 
 fluid and assist in the melting. The additions of fluor spar started at 2.50 p.m. 
 and amounted to about 1300 pounds total. At 3.00 p.m. a furnace sample was 
 
 
 p 
 
 v .'"■/. 
 
 >j^ tiit- /^$ ■■ 
 
 
 Sir 
 
 4: 
 
 Fig. 262. — Pouring Steel into Ladle of Open-hearth Furnace. (Copyright, Keystone View Co.) 
 
 taken and the phosphorus found to be .012 per cent. At 4.25 P.M. 150 pounds 
 of ore was added. Mixer metal to recarbonize was added to the furnace at 
 5.02 p.m. The furnace was tapped at 5.12 p.m. and ferromanganese (80 per 
 cent) and ferro-silicon (50 per cent) were added to the ladle, shortly after 
 tapping. 
 
380 STEEL RAILS 
 
 The amounts of the various materials used were as follows: 
 
 Pounds. 
 
 Limestone 25,000 
 
 Ore (Chapin) 15,000 
 
 Scrap steel 60,400 
 
 Mixer metal, first charge 110,900 
 
 Mixer metal to recarbonize 24,000 
 
 Ferromanganese 1,000 
 
 Ferrosilicon 500 
 
 A sample of the mixer metal used to charge the furnace gave the following 
 
 analysis : 
 
 Per cent. 
 
 Carbon 3.85 
 
 Silicon 1 . 73 
 
 Manganese 1.50 
 
 Phosphorus 215 
 
 Sulphur 030 
 
 (The silicon in the mixer metal ordinarily averaged about 1.25 per cent 
 instead of 1.73, as shown above.) 
 
 The ferromanganese and ferrosilicon had compositions about as follows: 
 
 ili 
 
 Carbon. . . . 
 Phosphorus 
 Sulphur 
 
 * The hot metal is tapped from the blast furnaces into 40-ton ladles, in 
 which it is hauled to two 300-ton mixers. The metal is poured from the 
 mixers into 60-ton charging ladles, in which it is conveyed to the open-hearth 
 furnaces on electric transfer cars. From these cars the ladles are picked up 
 by a 75-ton traveling crane and the metal is poured into the open-hearth fur- 
 naces through a runner (Fig. 263) . 
 
 The fact that the Bessemer process has already passed the zenith of its 
 growth is one which has now become well recognized by metallurgists generally. 
 
 Mr. Talbot has given a very clear presentation of this subject and the 
 author is indebted to him for the abstract of his paper which follows.f 
 
 The three main causes bringing about the supersession of the Bessemer 
 process are: 1. The ever-growing scarcity of iron ores suitable either for the 
 acid or basic Bessemer process; 2. The superiority of the product obtained 
 
 * Scientific American, December 11, 1909. 
 
 f Benjamin Talbot, in the London Times Engineering Supplement, February 13, 1907. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 381 
 
 by the open-hearth processes of manufacture; 3. The cheapening of the pro- 
 duction of the steel ingot by modern open-hearth methods of manufacture.* 
 
 The increasing scarcity of iron ores suitable for use in the acid Bessemer 
 process is, perhaps, the most cogent of the three causes named. In the United 
 
 Fig. 263. — Charging Platform of the Open-hearth Furnaces at Gary. (Scientific American.) 
 
 States, apart from the Southern States and the northern portion of New York 
 State, there are practically no ores at present available for the manufacture of 
 pig iron suitable for the basic Bessemer process. The rivalry is, therefore, 
 between the acid Bessemer and the basic open-hearth. The former had a 
 long lead, but the growth of the open-hearth rail manufacture has been rapid 
 and in the immediate future the development of the open hearth will be out of 
 all proportion to the further development of the acid Bessemer process. 
 
 f The total annual capacity for the production of open-hearth rails in the 
 
 * The cheapening of the cost of scrap, while not mentioned by Mr. Talbot, has been an impor- 
 tant factor. It should also be observed that open-hearth rails are subject to the same mechanical de- 
 fects as Bessemer rails, and it has not yet been proved that their superiority over Bessemer rails is very 
 marked; in fact, some extremely poor open-hearth rails have been made. 
 
 t Railway Age Gazette, March 16, 1910, daily edition. 
 
382 
 
 STEEL RAILS 
 
 United States is now about 1,500,000 tons, the principal mill being that at 
 Gary, which is turning out 500,000 tons; then Ensley, with 400,000 tons; Bethle- 
 hem, 200,000 tons; Colorado, 200,000 tons; Lackawanna and others of smaller 
 capacity, 200,000 tons. The total capacity of all mills making open-hearth rails 
 up to 1907 was less than 200,000 tons, and in that year production reached 
 253,629 tons. In 1908 it reached 571,841 tons, and in 1909, 1,256,674 tons. 
 
 The production * of open-hearth steel rails in 1910 was 1,715,899 tons, 
 against 1,256,674 tons in 1909. The increase in 1910 over 1909 was 459,225 
 tons or more than 36.5 per cent, while the increase in 1909 over 1908 was 
 684,833 tons or over 119 per cent. 
 
 The production of Bessemer steel rails in 1910 amounted to 1,917,900 tons, 
 against 1,767,171 tons in 1909, an increase of 150,729 tons or over 8.5 per cent. 
 Included in the total for 1910 is 68,497 tons of re-rolled rails. In 1911 the produc- 
 tion of open-hearth steel rails was less than in the previous year, but on account 
 of the smaller tonnage of Bessemer steel rails rolled than in 1910 more rails were 
 made from open-hearth steel than Bessemer, f 
 
 Table LXXXIX gives the production from 1907 to 1911. 
 
 TABLE LXXXIX. 
 
 -PRODUCTION OF STEEL RAILS IN THE UNITED STATES 
 FROM 1907 TO 1911. 
 
 
 1907. 
 
 1908. 
 
 1909. 
 
 1910. 
 
 1911. 
 
 
 Tons. 
 
 3,380,025 
 
 253,629 
 
 3,633,654 
 
 Tons. 
 
 1,349,153 
 
 571,841 
 
 1,920,994 
 
 Tons. 
 
 1,767,171 
 1,256,674 
 3,023,845 
 
 Tons. 
 
 1,917,900 
 1,715,899 
 3,633,799 
 
 Tons. 
 
 1,138,633 
 1,676,923 
 2,815,556 
 
 
 Total 
 
 t In any consideration as to the future of the acid Bessemer process in 
 the United States a thorough understanding of the ore situation is essential. 
 As is well known, the Lake Superior, particularly the Mesaba, ores are the 
 mainstay of pig-iron production in the north. Each year this ore becomes 
 leaner, and there is a difficulty in keeping the phosphorus content of the pig 
 iron manufactured from it below the .1 per cent of phosphorus which is the 
 standard for Bessemer steel in the United States. Steel made from such pig 
 is dangerously near the limit of safety for some purposes, when it is manu- 
 factured by the acid Bessemer process, but when treated in any form of the 
 basic open-hearth process such pig produces a metal of most excellent quality, 
 with phosphorus, when desired, down to .02 per cent, or even less. The carbon 
 content of the steel can also, in the latter class of process, be varied within very 
 
 * The Iron Age, February 23, 1911, p. 461. 
 
 t Railway Age Gazette, July 19, 1912, p. 125. 
 
 t Benjamin Talbot, London Times Engineering Supplement, February 13, 1907, daily edition. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 383 
 
 wide limits, while it is not so easy to produce .6 to .7 per cent carbon steel 
 in the acid Bessemer process, and even if made, steel with such high carbon 
 and with .1 per cent phosphorus, or thereabouts, is certainly not a material 
 that should be looked upon with favor for rail purposes. 
 
 All the facts point in one direction. The Bessemer process, while the 
 actual cost of conversion, apart from the question of waste, is perhaps the 
 cheapest, is yet one which requires, either for acid or basic working, a special 
 quality of pig iron, — a quality which is ever tending to become dearer. The 
 waste of metal in the Bessemer must of necessity be higher than in any form 
 of the open-hearth process, and this fact accentuates the importance of the ques- 
 tion of the cost of the pig iron; the higher the price, the greater the cost due 
 to waste. Roughly speaking the loss in a Bessemer is from 8 to 10 per cent and 
 in the open-hearth from 3 to 6 per cent. 
 
 The margin for economies in the Bessemer process is less than any which 
 can be made in the basic open-hearth process. Unless a radical change is 
 effected in the operation of the Bessemer furnace, only small further savings 
 appear possible. It is true that in some Bessemer the blowing power is still 
 raised by steam obtained from coal burnt under boilers, but even in cases in 
 which the blowing power is obtained from surplus blast-furnace gas, products 
 are absorbed which could otherwise be economically and usefully employed in 
 creating power for other purposes, if the open-hearth process were employed. 
 
 * The electric furnace is rapidly coming into use as an important factor 
 in steel manufacture, and where water power is abundant and fuel is scarce 
 
 ' it is extending the boundaries which have for a long time confined the iron 
 and steel districts. Experience with the electric furnace in foreign countries 
 has shown that it will purify the metal to a larger extent than the gas furnace 
 or the Bessemer converter, and it is proposed to use it as an adjunct to the 
 ordinary processes of steel manufacture for the purpose of reducing the amount 
 of phosphorus and sulphur and to deoxidize the bath. 
 
 After a careful investigation by its metallurgists, the United States Steel 
 Corporation has decided to use a 15-ton Heroult electric furnace at the South 
 Chicago works. Three-phase alternating current will be used, and it is pro- 
 posed to refine the blown metal from the Bessemer converter in the Heroult 
 furnace, reducing the percentage of phosphorus and sulphur, and to use the 
 product for high-grade steel rails. The capacity of one furnace is sufficient for 
 the production of 500 tons of steel in 24 hours. 
 
 * Railroad Age Gazette, March 12, 1909, and Composition and Heat Treatment of Steel, E F. 
 Lake, 1910, pp. 42-63. 
 
384 
 
 STEEL RAILS 
 
 The Heroult steel-refining furnace is of the crucible type with a tilting rack. 
 The heating is initially effected by means of the electric arc which forms between 
 the surface of the slagging materials which float on the metal bath and the two 
 
 massive carbon electrodes which are suspended above it. The impurities of 
 the steel are removed by renewing the slag. The refining operation thus be- 
 comes a " washing out " one. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 385 
 
 The lining is the same as the basic open-hearth and the phosphorus is first 
 reduced and then the sulphur. Recarbonizing is done in the bath by adding 
 crushed electrodes which are 98 per cent pure 
 carbon. 
 
 Fig. 264 shows a transverse section 
 through the pouring spout of the Heroult 
 furnace at La Praz, and also longitudinal sec- 
 tions of the furnace through the roof. The 
 electrodes, of which there are two passing 
 through the roof, are shown at E on the figure. 
 An alternating current at 110 volts is used. 
 
 In the Stassano arc furnace the neces- 
 sary heat is obtained by direct radiation 
 from the arc. It is shown in Fig. 265. 
 Various experiments have been made in this 
 furnace to produce steel direct from the ore, 
 but, owing to the difficulty of controlling the 
 composition of the slags with average ores, 
 the production of steel of any required grade 
 is far from easy.* 
 
 t More than 5000 tons of rails have been 
 made from steel from the electric furnace 
 at the Roechling Iron and Steel Company, 
 Voelklingen, Germany. The furnace is of a 
 special combination electrode and induction 
 type, known as the Roechling Rodenhauser, 
 and takes three-phase current at 25 periods. 
 The pig iron is blown in a basic lined Bes- 
 semer converter, then transferred to the 
 electric furnace for refining at an expenditure 
 of power of 125 kilowatt hours per ton. Re- 
 cently some tests have been published, made 
 
 Fig. 265. — Stassano Electric Furnace. 
 
 The furnace is inclined to the vertical and 
 rotated by the mechanism shown below. 
 
 November 27, 1908. The analysis of the three pieces then tested was as follows: 
 
 Per cent. 
 
 Carbon . 75 
 
 Silicon 0. 10 
 
 Manganese . 67 
 
 Sulphur 0.044 
 
 Phosphorus . 023 
 
 The rails were of flange section, 82.65 pounds to the yard. 
 
 * Steel, Harbord and Hall, London, 1911, pp. 261-283. f Railroad Age Gazette, July 2, 1909. 
 
STEEL RAILS 
 
 Physical tests were made on these rails; the pieces have a length between 
 punch marks of 7.94 inches and a diameter of almost 1.0 inch, being .975, 
 .966, .984 respectively. These results are given below: 
 
 Number. 
 
 Ultimate 
 Stress. 
 
 Elongation. 
 
 Reduction of 
 
 1 
 
 123,341 
 126,172 
 122,765 
 
 Per cent. 
 
 12.25 
 12.25 
 
 13.80 
 
 Per cent. 
 21.00 
 12.60 
 
 20.40 
 
 2 
 
 3 
 
 
 They show excellent ductility, in conjunction with tenacity. 
 
 Fig. 266. — • Roechling-Rodenhauser Furnace. (Lake.) 
 
 The latest development * in connection with the furnace is its operation 
 by a three-phase current, with a frequency of 50 periods for a 15-ton furnace. 
 Fig. 266 shows this furnace in sectional elevation and plan. It is claimed that 
 a special feature of the furnace is the rotation of the charge due to the presence 
 of a rotatory field, as in an induction motor, which insures an automatic circu- 
 lation in the bath. The furnace is essentially a transformer with a primary 
 winding A round both iron cores H of the transformer. The secondaries are 
 two in number; one is the molten bath in the form of an 8, the channel D 
 
 * Steel by Harbord and Hall, London, 1911, pp. 261-283, and The Report of the Canadian Com- 
 mission appointed to investigate the Different Electro-Thermic Processes for the Smelting of Iron 
 Ores and the Manufacture of Steel in Operation in Europe. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 387 
 
 between the two cores being very broad. The other secondary is the copper 
 winding B, which is connected with the metal plate E. 
 
 The electric furnaces, just described, show the three distinct types which 
 are claiming the serious attention of metallurgists. In the present state of 
 
 Fig. 266. — Roechling-Rodenhauser Furnace. (Lake). (Continued). 
 
 development of the electric furnace it cannot compete as regards cost of pro- 
 duction with a modern large open-hearth furnace for the manufacture of rails, 
 and it is only when a superior quality is required that it can be employed. 
 
 The steel from different parts of the ingot shows a very regular compo- 
 sition and is remarkably free from segregation of impurities. The mechanical 
 properties are extremely good. 
 
388 STEEL RAILS 
 
 Some recent experiments in the United States show that steel made in the 
 electric furnace has a greater density, and within the range of .08 to .75 carbon, 
 shows 10 per cent greater strength than open-hearth steel of the same chemical 
 composition. 
 
 * The duplex process is a combination of the Bessemer and the open- 
 hearth, and is particularly applicable to pig iron containing too high silicon 
 for advantageous working in either basic Bessemer or basic open-hearth. 
 
 In the acid Bessemer converter the preliminary blast removes the silicon, 
 together with a considerable portion of the manganese and a certain amount 
 of the carbon. The desiliconized metal is then transferred to the basic open- 
 hearth, where the phosphorus and the remainder of the carbon is eliminated 
 in accordance with the usual practice. 
 
 f The Jones and Laughlin Steel Company has been experimenting for some 
 time with the duplex process in its present Pittsburg plant, with the idea of 
 using a portion of its Bessemer capacity for preparing metal for its open-hearth 
 furnaces, thus decreasing its output of Bessemer steel and correspondingly 
 increasing the open-hearth output. The Maryland Steel Company has com- 
 pleted five open-hearth furnaces and is using a considerable portion of its 
 Bessemer capacity to duplex with the new open-hearth furnaces. 
 
 These moves in the direction of duplexing represent a distinct desire to 
 find a new use for Bessemer capacity because there is not sufficient employment 
 for it in its old function. 
 
 The casting ladle, or the ladle which receives the finished steel for casting 
 into molds, is shown in Fig. 267. If slag is allowed to pass into the ingot molds 
 with the steel the latter is liable to be spoiled, and in consequence the steel can- 
 not be poured from a lip into the molds, but has to be tapped or teemed from 
 a hole in the bottom of the ladle. 
 
 The time allowed after the conversion of the steel and when it is held 
 in the converter or casting ladle exercises considerable influence upon the 
 finished product. The thorough mixing of the recarbonizer, and the oppor- 
 tunity for the impurities to separate from the metal and the gas to escape from 
 the molten steel are of importance. Dr. P. H. Dudley requires a definite in- 
 terval of time between the additions of the spiegel and the teeming of the steel. 
 
 He says:{ " Restricting the ingots to three-rail lengths and holding the steel 
 three minutes after recarbonizing, in connection with the dry blast at South 
 
 * A Study of the Open-hearth, by Harbison- Walker Refractories Company. 
 
 t Railway Age Gazette, March 18, 1910, daily edition. 
 
 t Proceedings American Society for Testing Materials, Vol. VIII, 1908, p. 112. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 389 
 
 Chicago, shows a marked reduction in seams and cracks in the bases of rails. 
 In a lot of 2500 tons of these rails hardly a trace of seam has been found." 
 
 The dry blast referred to is the Gayley process of furnishing air, practically 
 free from aqueous vapor, to the converters while blowing the charge. This 
 
 & 
 
 Fig. 267. — Details of Casting Ladle. (Harbord and Hall.) 
 A, Goose-neck; B, stopper rod; L, sliding bar carrying goose-neck; M, M 1 , M 2 , bolts for attaching 
 lever, P; P, lever bar; Q 1 , Q 2 , brackets on ladle to guide sliding bar, L; K, screw bolt for holding 
 sliding bar rigidly in position previous to teeming; F, fire-clay sleeves threaded on steel stopper rod, B; 
 Z, teeming nozzle; E, fire-clay stopper head; G, nozzle box; H, trunnion; C 1 , C 2 , C 3 , C 4 , cotter pins; S, 
 forged head on sliding bar through which end of goose-neck is passed, and is fixed by cotter pin, C 4 . 
 
 decreases the amount of iron oxide in the bath which von Maltitz claims to be 
 a principal agent in producing blowholes. 
 
 That time should be given for the necessary chemical reaction after the 
 addition of the recarbonizer and before casting the metal into ingots, has been 
 known for at least thirty years. It was known that such time was also impor- 
 tant for the escape of the occluded gases, and the value of this latter knowledge 
 was manifested by the several devices for accelerating their escape which, years 
 ago, were either proposed or actually used. These ranged from the thrusting 
 into the ladle full of molten steel a wooden pole, or placing in the ladle before 
 it received the steel from the converter pieces of wood saturated with kerosene, 
 to more elaborate devices, for agitating the steel while in the casting ladle by 
 
390 STEEL RAILS 
 
 power-driven, refractorily protected screws. It was the practice at some of 
 the Bessemer works to again put on the blast after the introduction of the 
 recarbonizer, and then to partially turn the vessel up, thus agitating the charge. 
 
 Mr. Robert Forsyth sought to accomplish the desired results through his 
 transferring ladle arrangement, by which the ladle after receiving the steel from 
 the converter was transferred by a hydraulic ram from the receiving crane to 
 the ladle or casting crane. This he did when remodeling the Union Steel Plant, 
 Chicago, about 1886. Later he put the same arrangement in the South Works 
 of the Illinois Steel Company. Later Mr. William R. Walker carried this further 
 by pouring the steel over the top of the receiving ladle into the casting ladle 
 through a nozzle in the bottom of which it was cast in the usual way. 
 
 Prof. Henry Fay * has observed the moon-shaped fractures in the base and 
 the thermal cracks in the head of the rail which he believes to be generally 
 found along a streak of manganese sulphide, extending in the direction of the 
 rolling. The cause of the manganese sulphide being in the steel he infers is 
 probably the lack of time given for the steel to purify itself after the addition 
 of the recarbonizer. When the spiegel or ferromanganese is added to the 
 bath, the manganese combines with the sulphur, and, given time enough, man- 
 ganese sulphide, having a specific gravity less than that of steel, will rise to the 
 surface. 
 
 Manganese sulphide melts at 1162° C. Its specific gravity is 3.96, or 
 about half that of steel. It is a glassy, hard, and extremely brittle material. 
 The steel from which the rail is made solidifies at about 1450° C, and the 
 manganese sulphide will not solidify until it reaches 1162° C. Therefore, 
 the manganese sulphide is in a fluid state some time after the steel solidifies. 
 If the rolling of the rail starts, we will say, at a temperature above 1162° C., 
 this material will be rolled out in thin strips in the direction of rolling. It is 
 plastic below the melting point and it is capable of being rolled out while in the 
 plastic condition into long, thin strips. 
 
 He states: f " That manganese sulphide when existing in certain forms is a 
 harmful constituent of steels can no longer be doubted. The remedy seems to 
 be a very simple one. Specifications should be so drawn as to limit the amount 
 of sulphur in the steel. At the present time most of the specifications do not 
 even mention sulphur. Having done this, the next step is to allow the metal 
 to stand a longer time after the addition of the ferromanganese. With the 
 
 * Journal Association of Engineering Societies, July, 1908, p. 28. 
 
 t A Microscopic Investigation of Broken Steel Rails; Manganese Sulphide as a Source of Danger. 
 Fay. Vol. VIII (1908), Proceedings American Society for Testing Materials. Further Investigations 
 of Broken Steel Rails, Fay and Wint, ibid., Vol. IX. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 391 
 
 specific gravity of manganese sulphide 3.966 and steel 6.82, it should rise to the 
 surface and be skimmed off with the slag if given sufficient time. Usually this 
 time interval between charging of the ferromanganese and the pouring of the 
 ingot is very short. The desire of the manufacturer to increase his output 
 has led him to cut down this interval to the shortest possible limit, with the 
 natural consequence of a large number of defective rails. A longer time interval 
 will allow the metal to purify itself." 
 
 Mr. E. von Maltitz* found that where recarbonizing is done in the ladle 
 and insufficient time allowed for the complete reduction of the iron oxide in 
 the bath, an excessive number of gas holes may be formed. The presence 
 of gas seams tends to cause unsound metal. Mr. Robert Job has pointed out 
 that in nearly every case of failure due to crushed heads the section shows 
 marked unsoundness, and the vertical flaws of the gas seams weaken the head 
 greatly. 
 
 The fact should be emphasized that it is not alone sulphide of manganese 
 which is a source of danger, but other forms of slag are also to be looked upon 
 with suspicion. f These may be summed up as follows: 
 
 1. Excessive slag (manganese sulphide, silicate, etc.). 
 
 2. Segregation of slag concentric with the section. 
 
 3. Remnants of slag in the large split portion of the head. 
 
 4. Slag in those areas where flow of metal has occurred, or where micro- 
 
 scopic cracks have developed. 
 
 The rails illustrated in Figs. 268 and 269 show the effect of unsound metal in 
 the head of the rail. Referring to Fig. 268, view 1 shows a section of the rail 
 which has been polished and etched with acid. It shows some segregation and 
 flow of metal, as expected from the rolling. The rail was slivered, breaking off 
 the right of the head as indicated. This view shows also the portions into which 
 this section was cut and the marks by which they are identified; these were in 
 part polished and examined microscopically for defects as shown in the other 
 views of this figure. 
 
 View 2 shows the grain size at the center of the head and view 3 shows the 
 finer grain near the surface and the distortion of the grains by wheel action. 
 The photograph is taken at the end of a crack, and shows, besides this, some 
 
 * Blowholes in Steel Ingots. E. von Maltitz. Trans. American Institute of Mining Engi- 
 neers, Vol. XXXVIII (1907), p. 412-447. 
 
 t Iron and Steel Magazine, August, 1905 (Job) ; and Journal of the Iron and Steel Institute, 
 p. 301, 1905 (Captain Howorth). 
 
392 
 
 STEEL RAILS 
 
 small cavities, which were more noticeable before etching, and which might have 
 been minute oxide pits or pockets. 
 
 View 4 shows the metal, which is the white ground mass, to be badly con- 
 taminated by slag, which is extended longitudinally by the process of rolling. 
 These slag lines were found to some extent over the whole surface of this piece, 
 but were worst in the neighborhood of the point indicated by the dot By, view 1. 
 This is shown on the picture, which was taken at the edge of the localized portion. 
 
 leSSIM' 
 
 
 
 Cross Section at point C 2 , 1 
 68,000 grains per sq. in. 
 
 - Crushed Head. 
 
 Fig. 269 illustrates another example of a crushed head due to unsound 
 metal. View 1 shows a section of the head taken at the point of greatest dis- 
 tortion. The cavity in the top of this view is a drilled hole. On one side of 
 the head a cavity which did not show on the surface, but indicated marked 
 breaking down of the metal, was revealed. The metal, however, is more uni- 
 form throughout this rail than was the case in the rail of the preceding figure. 
 
 View 2 shows the grain at the center of the head; view 3, like view 3 of 
 Fig. 268, is taken at the end of the crack. It shows the finer grain and distor- 
 tion of the same, and shows as well the further distortion of the metal at the 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 393 
 
 end of the crack as a sort of tearing action. The end of the crack is at the 
 corner of the picture; the further direction of progress of the failure is shown 
 by the black defects extending across the photograph. 
 
 A longitudinal section of this rail made on portion G showed slag lines, as 
 in the rail of Fig. 268, somewhat most abundant at the point q, though the num- 
 ber was not so great as in the rail of the preceding figure. 
 
 
 1 1 
 
 J 
 »'« 
 
 ♦ ill . ih! 
 
 << 
 
 
 i 
 
 •1 
 
 i' 
 
 IV 
 
 • I 
 
 i«i 
 
 v r .' 
 
 ■r 
 
 Ir" 
 
 View 3. Cross Section of point Hu,, Mag. 100. 
 
 View 4. Longitudinal Section on Top of Portion B, 
 Mag. 50, Unetched. 
 Fig. 268. — Crushed Head. (Continued.) 
 
 * The deleterious influence of slag inclosures in steel has perhaps escaped 
 attention to some extent owing to the fact that in ordinary tensile tests, taken 
 in a direction parallel to that of rolling, these inclosures only occupy a very 
 small proportion of cross-sectional area and possess a tapered shape which allows 
 of gradual distribution of the stresses imposed on the material. If, however, we 
 consider the case of transverse stresses, or of shock or vibration, it will be seen 
 that these inclosures will be fractured as soon as the metal undergoes any material 
 deformation, and then each such inclosure practically represents an internal 
 
 * " Slag Inclosures " in Steel, by Walter Rosenhain. International Association for Testing 
 Materials, 5th Congress, Copenhagen, 1909. McGraw-Hill Book Company, New York. 
 
394 
 
 STEEL RAILS 
 
INFLUENCE OF DETAIL OF MANUFACTURE 395 
 
 fissure which is ready to extend — and actually does extend — in any direction 
 compatible with the applied stresses. 
 
 Fig. 270. — Teeming Ingots at Bessemer Converter. (Copyright, Keystone View Co.) 
 
 32. The Ingot 
 The principal points in connection with this part of the process are as follows: 
 
 1. Care must be exercised in casting the ingot. 
 
 2. The ingot must be in a perpendicular position until the interior has 
 
 had time to solidify. 
 
 3. The steel must not be overheated in the heating furnace or soaking pits. 
 
 4. Defective material from the top of the ingot must be excluded from the 
 
 finished product. 
 
STEEL RAILS 
 
 From the casting ladles the steel is run into cast-iron ingot molds located 
 on cars, as illustrated in Figs. 270 and 271. Table XC presents data showing 
 the time required to pour the steel at different mills. After solidifying, the 
 
 Fig. 271. — Teeming Ingots at Open-hearth Furnace. (Copyright, Keystone View Co.) 
 
 ingot mold cars are run under the stripper, shown in Fig. 272, from which hooks 
 are lowered and engage the lugs on either side of the mold and lift it off the 
 ingot. 
 
 The ingot is then taken up by a traveling crane and conveyed to the re- 
 heating furnaces or soaking pits, shown by Figs. 273 and 274, to allow the tem- 
 perature in all parts of the ingot to become equalized before rolling. 
 
Fig. 273. — Soaking Pits — Gary. (Scientific American.) 
 
398 
 
 STEEL RAILS 
 
 O .3 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 399 
 
 TABLE XC. — TEEMING PRACTICE AT AMERICAN RAIL MILLS 
 
 (Compiled by Committee on Rail, Am. Ry. Eng. Assn., 1909, and revised by the author 1912) 
 
 fAlgoma Steel Co 
 
 *Bethlehem Steel Co 
 
 *Cambria Steel Co 
 
 Carnegie Steel Co 
 
 ♦Illinois Steel Co 
 
 jlndiana Steel Co 
 
 ♦Lackawanna Steel Co 
 
 ♦Maryland Steel Co 
 
 Tenn. Coal, Iron & R.R. Co.. 
 
 Canadian Soo, Can. \ 
 
 Bethlehem, Pa 
 
 Johnstown, Pa 
 
 Braddock, Pa 
 
 South Chicago, 111. . 
 
 Gary, Ind 
 
 Buffalo, N.Y 
 
 Sparrows Point, Md. 
 Birmingham, Ala — 
 
 J mm. 
 Omin. 
 8 min. 
 
 2.15 
 
 2.07 
 
 2.1 to 1.18 
 
 Note. — Information froi 
 t Compiled by author. 
 
 t. W. Hunt & Co., except that marked (*), which was obtained direct from manufacturers by the 
 
 The unsoundness of the ingot results from several causes: 
 
 1. A funnel-shaped cavity or pipe at the top of the ingot. 
 
 2. Dispersed cavities or blowholes throughout the ingot. 
 
 3. Segregation of the impurities of the steel, as silicon, phosphorus, man- 
 
 ganese, etc., from the mass of the metal and their concentration 
 in different parts of the ingot. 
 
 The pipe is due to the contraction of the interior of the mass after the out- 
 side has set. After molten steel has been cast into an iron mold, the metal in 
 contact with the bottom and the sides begins first to solidify. 
 After a relatively short while the top of the ingot, which is 
 exposed to the cooling action of the air, also becomes solid 
 and the ingot now consists of a rigid metallic shell holding 
 a mass of molten steel, as shown in Fig. 275. As the cooling 
 proceeds this solid shell increases in thickness; but since steel, 
 like most substances, undergoes a considerable contraction in 
 passing from the liquid to the solid state, the mass of metal 
 which when liquid was sufficient to fill the space within the 
 solid shell will, after it has in turn solidified, be unable to fill 
 it and a cavity must necessarily be formed in the upper part 
 of the ingot. 
 
 The piping of ingots has been known for a number of years.* Robert 
 Forsyth at the Union Steel Works in 1888 demonstrated the relation of the 
 length of the pipe to the position of the ingot while its interior metal was solidi- 
 
 * The Manufacture of Bessemer Steels by R. W. Hunt, Lecture delivered at The Franklin In- 
 stitute, January 21, 1889. See Journal of the Franklin Inst., May, 1889. 
 
STEEL RAILS 
 
 fying, by breaking a number of ingots which had been differently handled — 
 some placed in a horizontal position as soon as possible after being cast, and 
 
 ,. . 
 
 Fig. 276. — Section of Ingot, 17 ins. Square at Top, 19 
 ins. Square at Base, and 50.5 ins. long, Containing 
 Cavity of 128 cubic inches. (Am. Inst, of Mining 
 
 Fig. 277. — Bloom from an Ingot of same 
 Heat and of same Size as Fig. 276, show- 
 ing Reduction of Cavity. (Am. Inst, of 
 Mining Engrs.) 
 
 others so placed at varying intervals up to having been kept vertically until 
 all of the steel was thoroughly set. 
 
 The best modern practice is to charge the hot ingots into the reheating 
 furnaces to equalize their heat for blooming as soon as possible after they are 
 teemed, stripped, and weighed. 
 
 An interesting experiment was tried by Dr. P. H. Dudley to determine 
 
INFLUENCE OF DETAIL OF MANUFACTURE 401 
 
 the relation between the pipe in an ingot which had been allowed to get cold 
 and one which had been promptly charged into the reheating furnace.* 
 
 Fig. 276 is a photograph of a three-rail ingot, for 100-pound rails, teemed in 
 a mold 19 inches square on the base, 17 inches square on the top, and 66 inches 
 long. The ingot, poured 50.5 inches long, was well deoxidized, and therefore 
 had a large cavity. The ingot had a volume of 7.4 cubic feet, inclosing a shrink- 
 age cavity of about 128 cubic inches, practically 1 per cent of its volume. This 
 is a larger percentage than would be found in rail steel not so well deoxidized, 
 or which contained numerous blowholes. 
 
 Fig. 277 is a photograph of the bloom of an ingot of the same heat and 
 length, cut for a 9 per cent mill discard. The ingot, after stripping and a subse- 
 quent ride of 500 feet, was charged directly into the reheating furnace without 
 allowing the temperature to fall below the recalescence point, while the bulk 
 of the steel was several hundred degrees above, and in about 2 hours the ingot 
 was drawn and bloomed. The cavity was small and less than one-tenth of that 
 of the cold ingot of the same heat. 
 
 Blowholes generally form in the upper half of the ingot, which is permeated 
 by honeycombs or dispersed cavities, due to the liberation of imprisoned gases, 
 principally hydrogen, as well as nitrogen and carbon monoxide. These gases 
 are absorbed, dissolved, or occluded in the molten steel, but are wholly or partially 
 evolved and collect into bubbles when the metal begins to solidify. These 
 bubbles are generally more numerous towards the side of the ingot. 
 
 The evolution of the gases in the mold seems probably to be due to two 
 causes: First, by the reduction of the temperature, the solvent power of the 
 steel for the gases is lowered and, consequently, certain proportions of the gas 
 are liberated; and, second, an evolution of carbon monoxide (CO) or carbon 
 dioxide (C0 2 ), due to chemical action. 
 
 t According to Howe, blowholes may be lessened or even wholly prevented 
 by adding to the molten metal shortly before it solidifies either silicon or 
 aluminium, or both. An addition of manganese has a like effect. J These ad- 
 ditions seem to act in part by deoxidizing the minute quantity of iron oxide and 
 carbonic oxide present, in part by increasing the solvent power of the metal for 
 gas, so that even after freezing it can retain in solution the gas which it had 
 dissolved when molten. But, since preventing blowholes increases the volume 
 of the pipe, it is often better to allow them to form, but to control their posi- 
 
 * Discussion of Henry M. Howe's paper on Piping and Segregation in Steel Ingots, Trans. Ameri- 
 can Institute of Mining Engineers, Vol. XL (1909), pp. 821-830. 
 
 t Iron, Steel, and Other Alloys, Howe, 1903, pp. 369-372. Contains record of Brinell's experiments. 
 t Titanium deoxidizes the steel in a very marked manner, as shown in Fig. 285. 
 
STEEL RAILS 
 
 tion, so that they shall be deep-seated. In case of steel which is to be forged 
 or rolled, this is done chiefly by casting the steel at a relatively low tempera- 
 ture, and by limiting the quantity of manganese and silicon which it contains. 
 
 Brinell finds that, for the conditions which are normal at his works at 
 Fagertsa, Sweden, if the sum of the percentage of manganese plus 5.2 times 
 that of the silicon is as great as 2.05, the steel will be so completely free from 
 blowholes as to have an undesirably large pipe. If this sum is 1.66, there will 
 be just that small quantity of minute, hardly visible blowholes which, while 
 sufficient to prevent any serious pipe, is yet harmless. If this sum is less than 
 1.66, blowholes will occur and will be injuriously near the surface unless this 
 sum is reduced to .28. He thus finds that this sum should be either about 
 1.66, so that the quantity of blowholes shall be harmlessly small, or as low as 
 .28, so that they shall be harmlessly deep-seated. 
 
 These numbers must be varied with the variations in other conditions. 
 In general, either a higher casting temperature, or a smaller cross section of 
 the ingots, or the use of hot or that of thin-walled molds, calls for a smaller 
 quantity of silicon and manganese. 
 
 Brinell also finds that an addition of .0184 per cent of aluminum is ap- 
 proximately equivalent to the presence of manganese and silicon in the pro- 
 portions Mn + 5.21 Si = 1.66 per cent; i.e., it unaided gives rise to structure 
 B (Table XCI). Naturally, little or none of this aluminum remains in the 
 steel. It oxidizes to alumina, which rises to the surface of the molten metal, 
 or is found lining the walls of the pipe. 
 
 Table XCI and Figs. 278 to 284 give some of Mr. Brinell's results. 
 
 TABLE XCI. — INFLUENCE OF MANGANESE AND SILICON UPON BLOWHOLES 
 AND PIPES 
 
 (Brinell, loc. eit., p. 370.) 
 
 Mn + 5.2 X Si. 
 
 2.05 
 1.66 
 1.16 
 
 Cast too hot 
 Cast too colcN 
 
 No blowholes, but a small pipe. 
 No visible blowholes, no pipe. 
 External blowholes, no pipe. 
 Fewer blowholes and somewhat 
 
 deeper seated. 
 The blowholes are very deep-seated. 
 Many external blowholes and a pipe, 
 Many blowholes, both external and 
 
 internal. 
 
 Injured by the pipe. 
 Just compact enough; excellent. 
 Injured by the external blowholes. 
 Blowholes still harmfully near the 
 
 surface. 
 Excellent. 
 
 Injured by the external blowholes. 
 ^Injured by the external blowholes. 
 
 The structures and H are those induced by too high and too low a casting 
 temperature respectively. The steel which here has structure would, if cast 
 at a normal temperature, have had structure A. It was thought that the reason 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 403 
 
 Fig. 278. — Structure A. — Brinell's Tests. Fig. 279. — Structure B. — Brinell's 
 
 (^L- 
 
 1 * J?2 
 
 ^\\ % '\J\® 
 
 Fig. 280. — Structure C. — Brinell's Tests. Fig. 281. — Structure D. — Brinell's Tests. 
 
 Fig. 282. — Structure E. — Brinell's Tests. 
 
 Fig. 283. — Structure 0. — Brinell's Tests. Fig. 284. — Structure H. — Brinell's Tests. 
 
404 STEEL RAILS 
 
 why the excessively high temperature caused these external blowholes was that it 
 caused the carbon of the molten steel to react on the iron oxide on the surface 
 of the mold, with the formation of carbonic oxide gas, which itself forms these 
 blowholes. 
 
 Von Maltitz* gives the following as the means for the prevention of blow- 
 holes in steel ingots: 
 
 1. Medium temperature of the heat during the last period of the process 
 in the converter or open -hearth. 
 
 2. Careful avoidance of overblowing or overoreing of the heat; careful 
 boiling out of the last portion of ore added to the bath. 
 
 3. A finishing slag not too rich in oxygen and having the proper degree 
 of fluidity. 
 
 4. The destruction, by stirring the heat before tapping, of the ferrous 
 oxide formed. 
 
 5. Addition of sufficient deoxidizing material to the heat, and the allowance 
 of sufficient time for the complete separation of the manganese protoxide, 
 silicate of manganese or alumina thus formed, into the slag. 
 
 f Howe maintains that the gas contained in the blowholes is partially 
 absorbed and the blowhole walls to some extent weldable during the process 
 of rolling. This action probably is less favorable in direct rolling of rails (i.e., 
 rolling direct from ingot to rail at a single heat) than in reheating practice, in 
 which the bloom into which the ingot is rolled is held in a special bloom-heating 
 furnace before rolling into a rail. 
 
 Segregation is one of the important questions before the steel maker. It 
 is, therefore, natural that for many years it should have engaged the attention 
 of iron and steel metallurgists in different countries, and have given rise to an 
 important literature. 
 
 Steel contains different impurities, as silicides, phosphides, carbides, sul- 
 phides, etc., whose freezing or solidifying points vary, and all have a lower 
 melting point than the metallic iron, consequently those having the lowest 
 melting point will tend to gradually segregate from the iron and concentrate 
 in the hottest part of the ingot. The top and center of the ingot always con- 
 tains the larger proportions of impurities. 
 
 All steels do not necessarily exhibit excessive concentration of impurities. 
 
 * Blowholes in Steel Ingots, E. von Maltitz, Trans. American Institute of Mining Engineers, 
 Vol. XXXVIII (1907), p. 445. 
 
 t The Welding of Blowholes in Steel, Henry M. Howe, Proceedings American Society for Test- 
 ing Materials, Vol. X, 1910, p. 168. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 405 
 
 The highly segregated portions of an ingot are often very small isolated areas 
 in the interior of the mass. 
 
 It is highly probable that a large part of the segregates in steel ingots is 
 directly traceable to the formation of blowholes. The pressure of the cooling 
 gas on the mixture of pure solids and impure liquid in which it forms must 
 squeeze out some of the impure liquid, which passes outward, ascends to the 
 top of the ingot, or finds its way into previously formed blowholes. 
 
 Below are presented the results recently obtained by Waterhouse* on 
 segregation in acid Bessemer rail steel. They point in the same direction as 
 those published in 1905 by Talbot, which showed the important influence of 
 aluminium in greatly retarding segregation in certain cases. 
 
 In the present instance, titanium, when rightly applied in the proper amount, 
 was also found to retard segregation of sulphur, phosphorus, and carbon, in what 
 is normally quiet, quick-setting steel. 
 
 The ingots used were from an ordinary rail-steel heat, and from a heat to 
 which had been added 64 pounds of ferrotitanium, which amounted to only 
 .25 per cent of the weight of the heat. The ordinary steel was made in the usual 
 way. After the heat was turned down, the proper amount of molten spiegel 
 was poured into the vessel. The heat was held there a short time, poured 
 into the ladle, and then through a lj-inch nozzle into the ingot molds. 
 
 There were six molds. After three ingots had been poured a ladle test 
 was taken. The third ingot of the six was allowed to cool while still standing 
 on its stool, and was then cut through longitudinally. 
 
 The heat in which ferrotitanium was used was made in exactly the same 
 way. As the steel was poured into the ladle, the alloy was shoveled in; the heat 
 was then held in the ladle for three minutes before pouring the ingots. In this 
 case also the third ingot of the heat was cooled in an upright position and cut 
 through longitudinally. An analysis of the ferrotitanium gave: 
 
 Per cent. 
 
 Carbon 10.50 
 
 Titanium 11.60 
 
 Iron 74. 12 
 
 Silicon 1 .60 
 
 Manganese 0.30 
 
 Calcium trace 
 
 Photographs of the ordinary and titanium ingots are shown in Fig. 285. 
 The noticeable feature is the increased soundness of the titanium steel, due to 
 the concentration of the blowholes in the pipe cavity. 
 
 * The Influence of Titanium on Segregation in Bessemer Rail Steel, G. B. Waterhouse, Proceed- 
 ings American Society for Testing Materials, Vol. X, 1910, p. 201. 
 
406 STEEL RAILS 
 
 The analyses of the ladle tests from the two heats were as follows: 
 
 
 Carbon. 
 
 Sulphur. 
 
 Phosphorus. 
 
 Silicon. 
 
 Manganese. 
 
 Ordinary steel 
 
 Titanium steel 
 
 Per cent. 
 0.44 
 
 0.47 
 
 Per cent. 
 
 0.098 
 0.068 
 
 Per cent. 
 
 0.088 
 0.093 
 
 Per cent. 
 0.117 
 0.118 
 
 Per cent. 
 
 0.91 
 0.95 
 
 W^ 
 
 3 v : 
 
 N! 
 
 m 
 
 « 
 
 teel. Titanium Steel. 
 
 Fig. 285. — Ordinary Steel Ingot and Titanium Steel Ingot. (Am. Soc. for Testing Materials.) 
 
 No trace of titanium could be found in the latter steel, so that it is not, 
 strictly speaking, a titanium steel, but will be called so for the purpose of dis- 
 tinction. 
 
 In the accompanying diagrams, Figs. 286 to 291 inclusive, are shown 
 the results obtained from determinations for sulphur, phosphorus, and carbon. 
 A if-inch drill was used, and drillings were taken to a depth of f inch. The 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 407 
 
 diagrams are drawn to scale, so that the location of the drillings can be readily 
 seen. The methods used were the same for all the samples and all the deter- 
 minations were made by one man. The results are briefly considered in order. 
 Sulphur segregates the most. In the ordinary steel it has risen from .098 
 per cent in the ladle test to a maximum of .223, and there is a considerable 
 
 '?£"■ 
 
 
 • if % 3 % s . 
 ■'¥ '? . '%* V s 
 ■% J if y i s •'? ■«" 
 ■"i 'if J i° 'i 3 % s 
 
 
 -/7^ : 
 
 -w 
 
 %-S 
 
 ■076 .09/ ,C*5 JOSS .07/ 
 .OSS .07S .OSS 
 
 h)|^ 
 
 .067 .OG3 O*/ i, 
 
 *• — uc 
 
 1,^. 
 
 Fig. 286. — Normal Ingot, Half Section. Fig. 287. — Titanium Ingot, Half Section. 
 
 Sulphur in Ordinary Steel. Sulphur in Titanium Steel. 
 
 (Am. Soc. for Testing Materials — Waterhouse.) 
 
 area with more than .147, which is 50 per cent more than the ladle test. In 
 the case of the titanium ingot, the contrast is very remarkable. The greatest 
 result is .101, which is not quite 50 per cent more than the ladle test, .068, 
 and the segregated area is very much smaller than in the case of the ordinary 
 ingot. It is true that there are two factors which may partly account for this 
 difference in results: the titanium ingot is somewhat smaller, and there is less 
 sulphur in the steel as a whole. The fact remains, however, that there is much 
 less segregation of sulphur in the titanium than in the ordinary steel. 
 
408 
 
 STEEL RAILS 
 
 In the ordinary steel the phosphorus has risen from .088 per cent in the 
 ladle test to a maximum of .167, and here also the segregated area is seen to 
 be considerable. In the titanium steel the maximum is .124, starting with a 
 ladle test of .093, and the segregated area is more restricted than in the case 
 of the ordinary steel. The ordinary steel is in a better condition to start with 
 
 . 1 
 
 <?£"- 
 
 fi 
 
 OZ3 ./JO -/3S /*/ Od9 
 
 O06 .069 
 
 vhW 
 
 -ssg- 
 
 p ",i 1 
 
 **\ 
 
 • • 
 
 .oee 
 
 
 03/ 094 ./24 ./// joo 
 
 .034 
 
 .//2 ./OO //? .096 
 
 .«£> ./o, .,/+ 
 
 o** 
 
 • "• * °Z 7 ' C9 i 
 
 
 5? 
 
 
 i „ 
 
 
 .OJ6 
 
 
 
 • • 
 
 • • • • • 
 
 L /8i ". * 
 
 rf' 
 
 Fig. 288. — Normal Ingot, Half Section. Fig. 289. — Titanium Ingot, Half Section. 
 Phosphorus in Ordinary Steel. Phosphorus in Titanium Steel. 
 
 (Am. Soc. for Testing Materials — Waterhouse.) 
 
 than the titanium steel, as it has slightly less in amount (.088 as compared 
 with .093), so that the behavior of the phosphorus is a good test of the pre- 
 ventive power of the titanium. The results given in the diagrams Figs. 288 
 and 289 show it to have been effective. 
 
 The titanium steel shows less segregation of carbon than the ordinary steel, 
 and it must be remembered that it starts with .47 as compared with .44 per 
 cent. The highest results found are .67 in the one case and .69 in the other, 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 409 
 
 and the diagrams show that the ordinary steel again has the larger segregated 
 area. 
 
 The results of the silicon and manganese determinations are not given; 
 they are somewhat erratic in each case, but do not exhibit segregation. 
 
 Mr. Henry M. Howe presented a very complete discussion on Piping and 
 /rj-" 
 
 
 
 St ,66 .SO 
 
 r r •*' v 
 
 ,ss .so .sz 
 
 .SO STa 
 
 
 .SO .SS .S2 
 
 .S6 .S2 .67 .S6 .S3 
 
 .SS .S6 .60 S6 .S9 
 
 .S7 .60 .S0 
 
 .S7 
 
 .S/ .4-9 .S7 >, 
 
 * * ' * ' T J =» 
 
 Fig. 290. — Normal Ingot, Half Section. Fig. 291. — Titanium Ingot, Half Section. 
 
 Carbon in Ordinary Steel. Carbon in Titanium Steel. 
 
 (Am. Soc. for Testing Materials — Waterhouse.) 
 
 Segregation in Steel Ingots, at the London meeting (July, 1906) of the American 
 Institute of Mining Engineers,* of which the following is an abstract. The 
 first part of this paper treats of the causes and the restraining of piping in steel 
 ingots, the second considers the causes and the restraining of segregation, and 
 the third proposes certain precautions in engineering specifications concerning 
 these two defects. An article coming from such a source is necessarily of value 
 
 * Piping and Segregation in Steel Ingots. Howe, Trans. American Institute cf Mining En- 
 gineers, Vol. XXXVIII (1907), p. 3-108. 
 
410 STEEL RAILS 
 
 in any consideration of these subjects, and it will be pertinent to briefly review 
 it. 
 
 Professor Howe infers that the pipe is chiefly due to what may be called 
 the virtual expansion of the outer walls of the ingot in the early part of the 
 freezing, and finds that the upper and smooth-faced part of the pipe probably 
 forms while the interior is still molten, but that the lower, steep, and crystal- 
 faced part probably forms in metal which is already firm. Of the causes 
 which may cooperate to limit the depth of the pipe, it is suggested that three, 
 namely, blow holes, sagging, and the progress of freezing from below upwards are 
 usually effective. 
 
 The pipe may be lessened by casting (1) in wide ingots; (2) in sand molds; 
 (3) at the top instead of at the bottom;* (4) slowly; (5) and with the large end up; 
 (6) by the use of a sinking-head or other means of retarding the cooling of the 
 top; (7) by permitting blowholes to form; (8) and by liquid compression. 
 
 It is believed that although the reasons why (1) casting in wide ingots 
 and (2) in sand or clay-lined molds shortens the pipe do not apply to show that 
 they should raise the segregate, yet the position of the segregate should be 
 raised by the six other means by which the pipe is shortened (see 3, 4, 5, 6, 7, 
 and 8 in preceding paragraph). 
 
 The means proposed for lessening the degree of segregation, as distinguished 
 from raising the position of the segregate, are next considered. 
 
 These are: 
 9. Quieting the steel by adding aluminium. 
 
 10. Casting in small instead of in large ingots, and hastening the solidifica- 
 
 tion, not only by casting in small ingots but also 
 
 11. By casting at a low temperature. 
 
 12. By casting in thick-walled iron molds (i.e., those of high thermal con- 
 
 ductivity) ; and 
 
 13. By casting slowly. 
 
 It is pointed out that quieting the steel has materially lessened segregation 
 in certain cases, and that segregation is probably much less in small than in 
 large ingots. 
 
 The effectiveness of the different methods of fluid compression (see Figs. 
 294-297) is considered, and it is concluded that the beneficial lifting effect on 
 the segregate should be the greatest in Williams' system, which compresses the 
 
 * Ingots are now practically all cast from the top, absolutely so in regard to rail steel. About 
 twenty years ago there was a great deal of bottom casting practiced, in casting the ingots at the Penn- 
 sylvania Steel Company's Works, which were rolled into rails at the Cambria Company's Works, Mr. 
 A. L. Holley first used bottom casting. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 411 
 
 ingot chiefly in the middle of its length; it should be the least in Whitworth's 
 system, which compresses the ingot more at its top than elsewhere; and it should 
 be intermediate in the systems of Illingworth and Harmet, which compress the 
 ingot equally in all parts of its length. 
 
 Finally, stress is laid on inspection at the rolls and shears, and especially 
 on axial drilling of the billets or other products. 
 
 Figs. 292 and 293 * present experimental verification of the following pre- 
 dictions made in this paper: 
 
 A. That the pipe is shortened and the segregate raised: 
 
 1. By slow casting. 
 
 2. By casting with the large end up instead of down. 
 
 3. By retarding the cooling of the top, e.g., by means of a sinking 
 
 head. 
 
 B. That the pipe is shortened by slow cooling. 
 
 C. That the pipe and segregate lie in the last freezing part. 
 
 The procedure was to cast ingots of wax containing a little bright-green 
 copper oleate (usually 1.5 per cent) under varying conditions; to saw each ingot 
 open along a longitudinal plane passing through its axis; and to examine the 
 longitudinal section thus laid bare. The segregated or enriched parts are shown 
 by the darker areas in the photographs, indicating the green of the segregated 
 copper oleate. 
 
 Taking up the evidence in detail, the influence of the rate of casting is 
 shown in ingots Nos. 1, 2, and 3, Figs. 1, 2, and 3. The casting of No. 1 was 
 finished in 30 seconds and that of No. 2 was so slow that, though it was continuous 
 except for momentary interruptions for heating the wax, it lasted 1 hour and 
 13 minutes. 
 
 The pipe in the fast-poured No. 1 stretches down 90 per cent of the ingot's 
 length, and, except for some very thin bridges, is practically continuous for 
 49 per cent; whereas in the slowly cast ingot the pipe stretches down only 14 per 
 cent of the length of the ingot. In this particular ingot (No. 2) there is a second 
 rudimentary pipe near the bottom, caused by the accidental pouring at first 
 faster than was intended. In ingot No. 3, which was poured slowly from the 
 start, this second pipe is absent. 
 
 The segregate in the fast-poured ingot No. 1 can be traced at A near the 
 bottom of the ingot. The slow-cast ingot No. 2 has a succession of local axial 
 
 * The Influence of the Conditions of Casting on Piping and Segregation, as shown by Means of 
 Wax Ingots. Howe and Stoughton, Trans. American Institute of Mining Engineers, Vol. XXXVIII 
 (1907), p. 109. 
 
412 
 
 STEEL RAILS 
 
 horizontal segregates, and in the still more slowly cast ingot No. 3 these local 
 segregates are so small as almost to escape notice 
 
 The effect of casting with the large end up instead of down is shown in 
 
 Figs. 4 and 5, which represent two ingots cast in immediate succession and under 
 otherwise like conditions. The pipe stretches down only 30 per cent of the in- 
 got's length when the large end is up, but 82 per cent when the large end is down. 
 The segregate lies well above the center in the ingot with the large end up, but 
 very near the bottom in that with the large end down. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 413 
 
 The effect of retarding and of hastening the cooling of the top of the ingot 
 is shown in Figs. 6 and 7 and in Figs. 8 and 9. The depth to which the pipe reaches 
 as a nearly continuous cavity is only 26 per cent of the ingot's length in the hot 
 
 topped ingot No. 6, but 85 per cent in the cold-topped No. 7. The pair of ingots 
 Nos. 8 and 9 which were slow cooling do not show so well the influence of the 
 distribution of temperature on the position of the pipe. 
 
 The effect of the rate of cooling is shown in Figs. 8, 9, and 10. Figs. 11 and 
 
STEEL RAILS 
 
 13 illustrate what Professor Howe calls "surface tension bridges." In the middle 
 and lower part of No. 11 there are five bridges from F to J and at K and L, as 
 well as at M of Fig. 13, there are the remains of bridges. These are in each case 
 greener than between the bridges and appear to have been formed on account of 
 
 Fig. 294. — Illingworth's Press for Compressing Steel Ingots Horizontally while Solidifying, Sectional 
 Plans. 
 In I, the mold is shown ready for receiving the molten steel. Two distance-bars, DD, are set between 
 the halves of the split mold B and C. After the steel has been poured into the mold, these distance-bars 
 are pulled out lengthwise, and the two halves of the mold are then forced towards each other by means 
 of the ram F , shown in II. The convex edges of the distance-bars are for the purpose of making an 
 initial depression in the side of the ingot lest part of its side should be forced out as a fin or welt into the 
 crevice between the two halves of the mold. 
 
 (Trans. Am. Inst, of Mining Engrs., Vol. XXXVIII.) 
 
 the local enrichment of the oleate, making the wax so fusible and plastic that it 
 stretches instead of cracking open when the pipe is being formed. The same phe- 
 nomena was noticed at E, F, G, of Fig. 12. The ingot of this figure had its cooling 
 
 Fig. 295. — Williams' Abdominal Liquid Compression of Solidifying Steel Ingots. 
 
 The ingot is cast in its mold as shown at I. After its outer crust has solidified the mold is opened, as 
 
 shown at II, and a liner B is slipped between mold and ingot. A strong cap A is then fastened down, 
 
 and by means of pressure applied through the ram C the abdominal protuberance on the ingot is forced 
 
 in, so as to close the pipe and lift the segregate into it, as shown at III. 
 
 (Trans. Am. Inst, of Mining Engrs., Vol. XXXVIII.) 
 
 hastened on the right-hand side and retarded on the left-hand side, which resulted 
 in shifting the pipe distinctly to the left of the axis. 
 
 Several methods have been used to produce sound ingots, as stirring the 
 steel in the casting ladle to allow the gases to escape; casting on a turntable 
 that is made to revolve and the metal run into a mold at its center; bottom 
 casting, with a closed top instead of the ordinary open-topped molds; and casting 
 under pressure. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 415 
 
 There are different methods of exerting the pressure on the fluid metal 
 in the mold. At the Krupp mills, fluid compression was tried by applying 
 
 the pressure exerted by carbonic anhydride , . 
 
 in its fluid state, the ingot mold being 
 capped after the ingot was cast and con- 
 nected to a reservoir containing the car- 
 bonic anhydride. 
 
 At the Edgar Thompson works the same 
 principle was applied, using steam under a 
 pressure of about 200 pounds per square 
 inch. Under this steam pressure the ingot 
 of 5 or 6 inches diameter shortened in length 
 from 1| to 2 inches. 
 
 Illingworth's process (Fig. 294) consists 
 of casting in vertical molds, split lengthwise. 
 The two halves are separated during the 
 casting, but when the crust is formed they 
 are brought together by a ram. 
 
 Williams' system (Fig. 295) employs the 
 split mold, and the two sides are pressed to- 
 gether with a liner between. 
 
 The Whitworth process (Fig. 296) con- 
 sists of using a steel mold which is placed 
 in a hydraulic press and the fluid steel sub- 
 jected to a pressure of about 6 or 7 tons 
 per square inch of horizontal section. Un- 
 der this pressure the ingot shortens about 
 1| inches per foot of its length. This pro- 
 cess produces an ingot of uniform quality 
 throughout and in a great measure over- 
 
 Fig. 296. — Whitworth's Hydraulic Press 
 for the Compression of Steel Ingots while 
 Solidifying. 
 
 A, main compression-cylinder. B, its 
 plunger. C, the carriage on which the 
 mold or flask sits. G, boss against which 
 the steel in the mold is forced. KK, steel 
 jackets for the mold. LL, the mold 
 proper. MM, perforated cast-iron lag- 
 ging. NN, inner sand lining. 
 
 comes the difficulty experienced from the ( Trans - Am - Inst - of Minin s En & s -> VoL 
 
 * *■■ * w u i 'J • • XXXVIII.) 
 
 formation of blowholes and piping. 
 
 The Harmet process (Fig. 297) consists in using a tapering mold and com- 
 pressing the fluid by means of a hydraulic ram acting on the open end of the 
 mold. The effect of the tapering mold is to exert a lateral pressure which tends 
 to close up any axial pipes. The French Government require 28 per cent crop 
 in uncompressed ingots and only 5 per cent in compressed ingots made at the 
 St. Etienne works by the Harmet process. 
 
416 STEEL RAILS 
 
 The test pieces cut from the compressed ingots show, without forging 
 or rolling, as good results under tensile and impact tests as test pieces cut from 
 ingots of the same composition which had been forged with a reduction of two 
 times in the cross-sectional area. There is a very marked diminution of segre- 
 gation, the chemical analyses at the top and bottom of the ingot being sub- 
 stantially the same. The compressed ingot has a grain of a visibly finer structure 
 and the large cleavages often found in sections cut 
 from uncompressed ingots are not found. The 
 metal is sound and thoroughly homogeneous. 
 
 Fluid compression of the steel in the ingot, when 
 in proper hands, according to our present evidence, 
 prevents pipes, blowholes, and cracks almost com- 
 pletely, and, to a limited extent, segregation. 
 While its introduction would certainly lead to 
 complication at the mills, the benefits to be derived 
 warrant a trial of this method. 
 
 As has been seen the greatest defects are found 
 
 Fig. 297.-Harmet : s Liquid com- in the upper part of the ingot, and to obtain a 
 
 pression by Wire Drawing. sound ingot it was generally specified that a certain 
 
 The ingot, a A .is cast in a strong discard or crop snou i d be made from the top of 
 
 conical mold, 27, reinforced 
 
 with hoops, 28. strong pres- the ingot, which it was supposed would contain all, 
 sure at the base of the ingot, 26, or near i y a u f the imperfect metal. 
 
 forces it lengthwise of the mold, 
 
 thus compressing it radially. To get definite knowledge on this point Dr. P. 
 
 (Trans. Am. Inst, of Mining H. Dudley experimented by lettering the rails formed 
 from different parts of the ingot. The rails were let- 
 tered "A" for the top rail, " B" for the second rail, and " C " for the third, or rail from 
 the bottom part of the ingot. These letters could be found in the tracks, and, as 
 was to be expected, the "A" rails have a larger percentage of impurities than the 
 " B " or " C " rails. They wear faster, developing more surface defects, and at several 
 points upon the road (the New York Central), under heavy traffic, after 10 or 12 
 years' service, have become practically worn out for main-line traffic, while the 
 "B" or "C" rails are still good. 
 
 The Committee on Standard Rails of the American Railway Association 
 reported at the meeting of the Association, April 22, 1908, on this subject, 
 as follows: 
 
 All rails are to be branded with the name of the maker, the weight of the rail, and the month 
 and year; and the number of the heat, and a letter indicating the portion of the ingot from which 
 the rail was made, shall be plainly stamped on the web of each rail, where it will not be covered 
 
INFLUENCE OF DETAIL OF MANUFACTURE 417 
 
 by the splice bars. Rails to be lettered consecutively " A," " B," " C," etc., the rail from the 
 top of the ingot being " A." In case of a top discard of 20 or more per cent, letter " A " will be 
 omitted. All rails marked " A " shall be .kept separate and be shipped in separate cars. 
 
 While railway engineers formerly specified a definite percentage of discard 
 from the ingot, they now unanimously agree that the specifications should not 
 state definitely how much should be sheared from the bloom, but that sufficient 
 material should be discarded from the top of the ingots to insure sound rails. 
 Records of rail failures which have been kept for a number of years disclose 
 the incontrovertible fact that, where a 15 per cent discard might do for one 
 ingot, 50 per cent would not be adequate for another.* 
 
 In its report the Committee of the American Railway Association says: 
 With regard to the discard question, the Committee has always held that it 
 would be preferable to test the finished product rather than specify as to the 
 details of mill manufacture, and the Committee arranged for a trial lot of rails 
 to be rolled from the ingot without any discard whatever, except such as was 
 necessary to enable the bloom to enter the rolls, and after these rails had been 
 cut into small pieces they were broken under the hammer and the fracture ex- 
 amined. This test proved to the satisfaction of the Committee that if "pipes" 
 or other physical defects were present they could be detected by this means. 
 The test also proved quite conclusively that it is possible so to conduct the 
 process of manufacture that the "pipes" or other physical defects will be re- 
 duced to a minimum, and that these defects may not occur at all, even in rails 
 rolled from the top portion of the ingot. 
 
 In order to avoid an unnecessary waste of good material, the Committee 
 set about to devise means by which the rejection of defective material could 
 be insured without requiring an arbitrary and definite percentage of discard 
 in every case, and a committee of the Pennsylvania Railroad, pursuing the 
 same line of investigation, adopted a tentative specification which provided 
 for a physical test of this nature, and which further provided that when physical 
 defects were discovered all top rails of the heat should be rejected. A trial 
 lot of rails, of a section corresponding to " type B" (Plate VIII),was rolled un- 
 der this specification as to discard, and the results convinced the committee 
 that a development of this idea would prove the best solution of the discard 
 problem. 
 
 Some of the advocates of a fixed and arbitrary discard have argued that 
 the mere provision of a discard to insure the elimination of " piped " rails, or 
 rails containing physical defects, was not sufficient, and urged the rejection of 
 
 * See Paper by W. C. Cushing before the Indiana Railroad Conmission, February 20, 1912. 
 
418 STEEL RAILS 
 
 a fixed percentage from the top of the ingot, because of the well-known fact that 
 segregation occurs in the upper portion. This question of segregation was given 
 careful consideration by the Committee, and while it is a fact that, due to the 
 rearrangement of the constituent parts of the metal during the process of cooling 
 and solidification in the ingot mold, any analysis of the metal in the finished 
 rail will often show a wide departure from the analysis required by the speci- 
 fications, it is also true that an analysis of the metal taken from the different 
 parts of the finished rail will frequently show similar wide variation. This 
 discrepancy is due to the fact that the test ingot referred to in the specifications, 
 and upon which the chemistry specification is based, is taken from the ladle before 
 the metal is poured into the ingot mold, and, consequently, before the segrega- 
 tion takes place. 
 
 It has been assumed that, because of this variation from the standard 
 composition of the metal in the finished rail, the rejection of all segregated 
 metal would be warranted. But, on this assumption, it would be necessary 
 to discard more than a third of the upper part of the ingot to be on the safe 
 side, as the segregation frequently extends that far; and, while our knowledge 
 of the subject is not so complete as we could wish it to be, we have a great deal 
 of evidence that rails of good physical condition can be made from the upper 
 portion of the ingot. Furthermore, the analyses of a large number of rails, 
 taken after years of service, indicate that these wide variations in chemical 
 composition may occur without apparently affecting the safety or wearing 
 quality of the rail; and, since it is impossible to check the analyses of the finished 
 rail with that of the test ingot, the question arises as to what limits should be 
 placed on the variation which will be permissible. None of the experts con- 
 sulted are ready to say what this limit should be, and all admit that no facts 
 are available as the results of actual experience which would warrant the adoption 
 of any fixed limit to govern the rejection of material. 
 
 The provision in the new specifications for stamping the rails to show 
 their position in the ingot will enable us to obtain more definite information 
 on this point in the future. 
 
 BIBLIOGRAPHY 
 
 Beikirch, F. O. — Verfahren zur verhiitung der lunkerbildung in schweren rohstahlblocken, 
 1800 w. 111. 1905. (In Stahl und Eisen, Vol. 25, Part 2, p. 865.) 
 
 Describes use of sinking-head, afterwards cut off. Piped portion is thus removed. 
 
 Daelen, R. M. — Die verfahren zur verhiitung der lunkerbildung in stahlblocken. 1800 w. 111. 
 1905. (In Stahl und Eisen, Vol. 25, Part. 2, p. 923.) 
 
 The same. (In Zeitschrift des Vereines Deutscher Ingenieure, Vol. 49, Part 2, p. 1398.) 
 
 Describes fluid compression methods, methods of heating moulds and continuous process, proposed 
 by writer. 
 
 Dormus, A. V. — Die blasen- und lungerbildung des flusseisens. 1000 w. 111. 1902. (In Zeit- 
 schrift des Osterreichischen Ingenieure- und Architekten- Vereines, Vol. 54, Part 1, p. 279.) 
 
INFLUENCE OF DETAIL OF MANUFACTURE 419 
 
 The same, translated. (In Sibley Journal of Engineering, Vol. 17, p. 128.) 
 
 Describes formation of blow-holes and pipes in steel and gives mechanical and chemical means for 
 prevention. 
 
 Dudley, P. H. — Dark carbon streaks in segregated metal in split heads of rails. 2000 w. 111. 
 
 1909. (In Proceedings of the American Society for Testing Materials, Vol. 9, p. 98.) 
 The same. (In Railway and Engineering Review, \ ol. 49, p. 626.) 
 
 Heroult, P. L. T. — Presence and influences of gases in steel. 800 w. 1910. (In Transactions 
 of the American Electrochemical Society, Vo . 17, p. 135.) 
 
 Concludes that blow holes are the result of disengagement of carbon monoxide in steel. 
 
 Howe, Henry M. — Does the removal of sulphur and phosphorus lessen the segregation of carbon? 
 2500 w. 1907. (In Proceedings of the American Society for Testing Materials, Vol. 7, p. 75.) 
 
 Results from examination of many cases indicate that removal of sulphur and phosphorus does 
 not lessen segregation of carbon. 
 
 Howe, Henry M. — • Further study of segregation in ingots. 5000 w. 111. 1907. (In Engineer- 
 ing and Mining Journal, Vol. 84, p. 1011.) 
 
 Suggests that quiet, resulting from increased ingot-size and slow cooling, may explain why these 
 conditions do not always favor segregation. 
 
 Howe, Henry M. — Influence of ingot-size on the degree of segregation in steel ingots. 800 w. 
 111. 1909. (In Transactions of the American Institute of Mining Engineers, Vol. 40, p. 644.) 
 
 Howe, Henry M. — Influence of top-lag on the depth of the pipe in steel ingots. L000 w. 1909. 
 (In Transactions of the American Institute of Mining Engineers, Vol. 40, p. 804.) 
 
 Howe, Henry M. — Piping and segregation in steel ingots. 106 p. 111. 1906. (In Transactions 
 of the American Institute of Mining Engineers. Vol. 38, p. 3.) 
 
 Discussion, Vol. 39, p. 818. 33 p. 
 
 Considers causes and prevention of piping and of segregation and precautions to be taken regarding 
 these in specifications. 
 
 Howe, Henry M. — Segregation in steel ingots. 1000 w. 1908. (In School of Mines Quarterly, 
 Vol. 29, p. 238.) 
 
 Summarizes results of author's investigations. 
 
 Howe, Henry M., and Stotjghton, Bradley. — Influence of the conditions of casting on piping 
 and segregation, as shown by means of wax ingots. 3000 w. III. 1907. (In Transactions of the 
 American Institute of Mining Engineers, Vol. 38, p. 109.) 
 
 Verifies by experiments views expressed by Howe in earlier paper. 
 
 Huston, Charles L. — Experiments on the segregation of steel ingots in its relation to plate 
 specifications. 3000 w. 111. 1906. (In Proceedings of American Society for Testing Materials, Vol. 6, 
 p. 182.) 
 
 Illustrates and explains cases of segregation. 
 
 Job, Robert. — Investigation of defective open-hearth steel rails. 1500 w. 111. 1909. (In Pro- 
 ceedings of the American Society for Testing Materials, Vol. 9, p. 90.) 
 
 Discussion, p. 106 
 
 The same. (In Railway Age Gazette, Vol. 48, p. 523.) 
 
 Shows failures to be due to unsoundness of metal. 
 
 Knight, S. S. — Observations on segregation phenomena as applied to cast steels. 3000 w. 111. 
 
 1910. (In Iron Trade Review, Vol. 46, p. 475.) 
 
 Paper before Philadelphia Foundrymen's Association. 
 
 Illustrates examples of segregation and emphasizes need of further investigations. 
 
 Lilienberg, N — Piping in steel ingots. 2000 w 111. 1906. (In Transactions of the American 
 Institute of Mining Engineers, Vol. 37, p. 238.) 
 
 Considers methods for prevention of piping, with special attention to Illingworth's side-compression 
 method. 
 
 Mathesius, W. — Herstellung von poren- und lunkerfreiem grauguss, stahlguss und schmiede- 
 stucken durch anwendung von thermit 2500 w. 111. 1903. (In Stahl und Eisen, Vol. 23, Part 2, p. 925.) 
 
 Riemer, Julius. — Ein neues verfahren zum verdichten von stahlblocken in flussigem zustande. 
 2000 w. 111. 1903. (In Stahl und Eisen, Vol. 23, Part 2, p. 1196.) 
 
 The same, translated. (In Iron and Coal Trades Review, Vol. 67, p. 1776.) 
 
 New method consists in pressing of ingot from below, allowing gases to escape at the top. 
 
 Sauveur, Albert, and Whiting, Jaspar. — Casting of pipeless ingots by the Sauveur overflow 
 method. 1500 w. 111. 1903. (In Proceedings of the American Society for Testing Materials, Vol. 3, 
 p. 129.) 
 
 Segregation in soft steel ingots; its effect on rolled material as shown in tensile and shock tests. 
 2500 w. 111. 1910. (In Iron Age, Vol. 86, p. 730.) 
 
 Gives results and conclusions reached by Wiist and Felser in paper in " Metallurgies Shows 
 segregation to have greatest influence in shock tests, as segregated material is very brittle. 
 
 Springer, J. F. — Piping in steel ingots; methods for its reduction and elimination. 4000 w. 
 111. 1909. (In Cassier's Magazine, Vol. 35, p. 426.) 
 
 Springer, J. F. — Thermal treatment of steel ingots. 1200 w. 111. 1910. (In Scientific American, 
 Vol. 116, pp. 262, 269.) 
 
 Describes piping and segregation phenomena and briefly reviews methods for prevention. 
 
 Stead, J. E. — Crystallization and segregation of steel ingots. 53 p. 111. 1906. (In Proceedings 
 of the Cleveland Institution of Engineers, 1905-06, p. 163.) 
 
 Review of work done on the subject, with conclusions. 
 
420 STEEL RAILS 
 
 Talbot, Benjamin. — Segregation in steel ingots. 44 p. 111. 1905. (In Journal of the Iron and 
 Steel Institute, Vol. 68, p. 204.) 
 
 Includes extensive tabulated data on the effect of additions of small amounts of aluminum to 
 the ingot. 
 
 Wahlberg, Axel. — Influence of chemical composition on soundness of steel ingots. 38 p. Ill 
 1902. (In Journal of the Iron and Steel Institute, Vol. 61, p. 333.) 
 
 Wedding, H. — Untersuchung iiber den ursprung eines blasenraumes in einen flusseisenblocke 
 2000 w. 111. 1905. (In Stahl und Eisen, Vol. 25, Part 2, p. 832.) 
 
 Considers probable origin of a blow-hole in a 2-ton ingot. Faulty shape of mould considered to 
 be explanation. 
 
 Weitere entwicklung des Riemerschen verfahrens zur herstellung dichter stahlblocke. 1000 w. 
 111. 1904. (In Stahl und Eisen, Vol. 24, Part 1, p. 392.) 
 
 Illustrates great reduction in piping from use of new method of Riemer. 
 
 Wickhorst, M. H. — Low-carbon streaks in open-hearth rails. 1200 w. 111. 1910. (In Proceed- 
 ings of the American Society for Testing Materials, Vol. 10, p. 212 ) 
 
 Studies rails that developed a peculiar kind of failure, shown to be due to streaks of metal low in 
 carbon. 
 
 Wickhorst, M. H. — Segregation and other rail properties as influenced by size of ingot 97 p. 
 111. 1911. (In Proceedings of the American Railway Engineering Association, Vol. 13, p. 655.) 
 
 33. Influence of Mechanical Work 
 
 The principal points in connection with the rolling are given below: 
 
 1. Resistance to wear is a function of fineness of grain. 
 
 2. Fineness of grain is principally a result of mechanical treatment at 
 proper temperature.* 
 
 3. Work done on steel above 950°-1050° C. (1742° F.-1922 F.) has less 
 effect on changing the size of grain from the normal crystallization of the ingot 
 than when the rolling is done at a lower temperature. 
 
 Fig. 298 shows the steel entering the rolls. Figs. 299, 300, 301, and 302 
 illustrate views taken by Mr. Howard and show the gradual reduction of the 
 bloom to the finished rail as it passes through the successive rolls. 
 
 In 1909 a further investigation was made of the steel at different stages of 
 the rolling by James E. Howard at the Watertown Arsenal. f In these tests, 
 beginning with the ingot, the structural state of the metal was examined by 
 taking cross sections and longitudinal sections. This method was carried 
 through the various successive derivative shapes, and the results obtained are 
 shown in the large number of illustrations which form the body of the report. 
 
 The greater part of the work was devoted to Bessemer rail steel, five acid 
 Bessemer heats being made for this series of tests, each heat furnishing six 
 ingots about 19| by 20| inches at the bottom and about 5 feet high. 
 
 One of the most important results of the tests was to throw light on the 
 question of the amount of work or reduction necessary in rolling to develop 
 the full physical qualities of the steel. Mr. Wickhorst draws the following 
 
 * This should not be interpreted as meaning that resistance to wear is not a function of the chem- 
 ical composition. 
 
 t Tests of Metals, etc., 1909, Vol. 1 and Vol. 2, Government Printing Office, Washington. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 421 
 
STEEL RAILS 
 
 conclusion from the tensile tests made of specimens taken at various stages 
 from the ingot to the finished rail. 
 
 s Section of 8 by 8 in. Rail Bloom Rolled from an Ingot 20 ins. Square. 
 (Am. Ry. Eng. Assn. — Howard.) 
 
 The results indicate that the metal in the walls of the ingot takes com- 
 paratively little work or reduction to impart to it what may be called its full 
 
 Fig. 300. — Rail from an Early Pass in Roughing Rolls, Rolled from Bloom Shown in Fig. 299. 
 (Am. Ry. Eng. Assn. — Howard.) 
 
 physical properties of tensile strength and . ductility. These are reached in the 
 bloom, except at the top end. The axial metal at the bottom of the ingot also 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 423 
 
 Fig. 301. — Same Rail as Shown in Fig. 300 after Further Reduction. 
 (Am. Ry. Eng. Assn. — ■ Howard.) 
 
 soon reaches its full physical properties, but in the upper half of the ingot it 
 must be carried well toward the finished rail before these properties are fully 
 developed. 
 
 Fig. 302. — Finished Rail from Same Ingot as Bloom and Pieces from Roughing Rolls. 
 (Am. Ry. Eng. Assn. — Howard.) 
 
424 STEEL RAILS 
 
 Where the metal is of fairly even composition and free from sponginess, it 
 reaches its full physical qualities of tensile strength and ductility at about ten 
 reductions or a reduction to one-tenth of the original cross section of the ingot, 
 but the interior portion of the upper part of the ingot requires twenty-five or 
 more reductions to have its full physical qualities developed, that is, the cross- 
 sectional area must be reduced below one twenty-fifth of its original amount. 
 
 Table XCII gives the result of tests by Sauveur on the relation between 
 the size of the grain and the physical properties of the same piece of steel.* 
 
 TABLE XCII. — RELATION BETWEEN SIZE OF GRAIN AND 
 PHYSICAL PROPERTIES OF STEEL 
 
 Size of Grain. 
 
 Tensile Strength. 
 
 Elongation 
 Per cent of 
 Length. 
 
 Reduction of 
 Area, Percent. 
 
 In 0.0001 
 Millimeter. 
 
 Number per 
 Approximately. 
 
 Kilograms per 
 
 Square 
 
 Millimeter. 
 
 Pounds per 
 Square Inch. 
 
 148 
 
 118 
 62 
 
 44,000 
 54,000 
 104,000 
 
 69.6 
 70.3 
 
 77.7 
 
 99,000 
 100,000 
 110,000 
 
 15.0 
 19.0 
 22.5 
 
 20 
 22 
 35 
 
 Mr. Robert Job, chemist of the Philadelphia & Reading Railroad, states 
 that " in a lot of over 75,000 tons of rail observed during a period of five years, 
 we have 15 times as many fractures in service from rails of coarse grain, or 
 19,600 cells per square inch, as from rails of medium fine structure, 48,400 cells 
 per square inch, and there is also a marked difference in capacity for wear in 
 favor of the finer structure rail. Out of several thousand tons of rail now 
 (1905) in our tracks, made with a clause in the specifications requiring more 
 than 40,000 cells per square inch, only one rail has fractured in service, and that 
 owing to pipes in the steel in process of manufacture." 
 
 Dr. P. H. Dudley states that rails of 100-pound section with 48,000 to 70,000 
 cells per square inch after having sustained 250,000,000 tons in the track were 
 still in good condition. 
 
 The rolling tends to break down the grain and give a finer structure, but im- 
 mediately after the work stops the grain commences to grow again, consequently 
 the lower the finishing temperature the smaller the grain size. If the steel is 
 worked below the critical point, strains are developed which injure the metal 
 and may rupture it. Work at too low a temperature distorts the grain or 
 flattens and elongates the crystals in the direction of the rolling. 
 
 * N. Ljamin, Chem. Zeit, 1899, Baumaterialien, 1899, finds the tenacity in different steels varies 
 directly as the size of the pearlite grains, at the same finishing temperature. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 425 
 
 To thoroughly understand the effect of the rolling, it is necessary to study 
 the structural changes that take place in the cooling metal.* 
 
 Let us first consider the cooling curve of a copper bar, shown in Fig. 303. 
 We find here no evidence of any sudden change in the nature of the cooling copper. 
 
 V- 
 
 
 
 
 V 
 
 Freezing 
 
 
 
 
 
 
 
 
 
 
 
 20 40 60 Time 
 
 Fig. 303. — Cooling Curve of Solid Copper. 
 (J. W. Mellor.) 
 
 '"0 20 40 60 Time 
 
 Fig. 304. — Cooling Curve of Water. 
 
 (J. W. Mellor.) 
 
 If, however, a curve is drawn for water cooling down from 20° to -20° C, 
 we get a terrace in the cooling curve, as shown in Fig. 304. This tells us that 
 some change has taken place in the nature of the substance at 0°. We see at 
 
 ■■ 
 
 Fig. 305. — Recalescence. (J. W. Mellor.) 
 
 once that this change corresponds with the passage of water from the liquid 
 to the solid state. 
 
 When a steel bar is cooling, an evolution of heat occurs at about 690° C. 
 The amount of heat evolved is so great that the metal visibly brightens in color. 
 The phenomenon is called " recalescence." The cooling curve is shown in 
 Fig. 305. 
 
 * See The Crystallization of Iron and Steel, J. W. Mellor, 1905. 
 
426 
 
 STEEL RAILS 
 
 The cooling curve of iron from the molten condition is shown in Fig. 306. 
 The iron was practically pure, containing only .01 per cent of carbon. 
 
 Osmond maintains that the existence of the transition points, Ar 3 and Ar 2 , 
 in the cooling curve of the solidified metal points to the existence of three 
 allotropic modifications of solid iron. Each critical point is found to be associ- 
 ated with a change in the mechanical properties, the microscopic appearance, 
 and the specific gravity of the metal. 
 
 1800°(— { 1 1 1 1 900°rr 
 
 
 Freezing 
 
 
 
 
 
 
 
 
 860 
 
 vAr 3 
 750° %■ 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 Ac, 
 
 -^730° 
 
 69 
 
 
 \& 
 
 
 
 
 vQ? 
 
 
 20 4-0 60 Time 
 
 Fig. 306. — Cooling Curve of Iron. 
 (J. W Mellor.) 
 
 20 40 60 Time 
 
 Fig. 307. — Cooling and Heating Curves of 
 
 Steel. (J. W. Mellor.) 
 
 The changes which occur during the cooling of a substance are reversed 
 when the substance is heated. The cooling curve of steel, with 1.2 per cent 
 of carbon, shown in Fig. 307, is reversed on heating, as shown by the heating 
 curve in the same diagram. 
 
 The critical points on the heating curve of mild steel are generally a few 
 degrees higher than the corresponding points on the cooling curve. 
 
 Let us now consider what takes place when steel containing 0.6 per cent 
 carbon cools from 900° C. The cooling curve shows nothing very remarkable 
 until a temperature of about 720° C. is attained. Here the critical points 
 Ar s and Ar 2 of pure iron coalesce into one. At this point pure iron, or 
 ferrite, separates from the solid solution. The separation of ferrite goes on 
 along the curve AP (Fig. 308) until the temperature reaches about 690° C, 
 when another recalescence point occurs (An). No other essential change, as 
 far as we are concerned, occurs as the system cools down to the normal tem- 
 perature of the atmosphere. 
 
 Fig. 308 is derived from Roozeboom's diagram,* the carbon-iron diagram, 
 
 * H. W. B. Roozeboom, Zeit. Phys. Chem., 34.437, 1900; improved in Zeit. Elektrochem., 
 10.489, 1904; Metallographist, 3.293, 1900; H. le Chatelier, ibid., 3.290, 1900; 4.161, 1901; F. Osmond, 
 4.150, 1901; H. Jiiptner von Jonstorff, ibid., 5.210, 1902. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 427 
 
 
 
 
 B 
 
 A 
 
 
 
 Ar>4 
 
 
 
 
 
 Pearlit^ Line 
 
 
 
 
 
 
 
 Fig. 308. - 
 
 - Cooling of "Solid Steel." 
 (J. W. Mellor.) 
 
 given by Howe,* based upon later researches, shows the temperatures some- 
 what higher than those of the figure. 
 
 When the temperature is above the line APB the iron is in a form known 
 as "austenite." Whatever carbon is present is dissolved in this austenite, 
 which is what is called a "solid solution" as distinguished from a mechanical 
 mixture or conglomerate, just as salt and water, when brought in contact, merge 
 in each other and pass from the condition 1200V 
 of a mixture or conglomerate to that of a 
 single substance. 
 
 As the iron with 0.6 per cent carbon 
 cools and the line AP is reached the 
 austenite begins expelling from itself part 
 of its iron in the form of ferrite. As the 
 ferrite thus expelled is nearly or quite free 
 from carbon, the remaining austenite be- 
 comes relatively richer in carbon, until, 
 when the temperature reaches An, it 
 contains 0.9 per cent carbon which is the 
 carbon content of pearlite. On cooling 
 past this point all the austenite changes into pearlite, with no change in the 
 ferrite which it has generated in the passage along the line AP, so that the 
 steel now consists of a conglomerate of ferrite and pearlite. This conversion of 
 the austenite into pearlite is accompanied by a considerable evolution of heat, 
 and is shown by the recalescence curve of Fig. 305. 
 
 Steels containing just 0.90 per cent carbon, and hence consisting of pearlite 
 alone, are called "eutectoid" steels. Those containing less than this are called 
 "hypo-eutectoid," and those more than this, " hyper-eutectoid " steels. 
 
 As previously stated, work below the Ar x point distorts the grain or flattens 
 and elongates the crystals in the direction of the rolling. 
 
 The result of work above the Ar x point is to retard the growth of the 
 grains. Howe f explains the relation between the temperature of the hot work 
 and the size of the grain as follows: 
 
 The mechanical distortion in rolling elongates these grains in the direction 
 of rolling and shortens them in the plane of the pressure; this appears to throw 
 the metal crystallographically into unstable equilibrium, with the result that 
 
 *'Life History of Network and Ferrite Grains in Carbon Steel, H. M. Howe. Proceedings, 
 American Society for Testing Materials, Vol. XI, 1911, p. 266. 
 t H. M. Howe, Iron, Steel, and Other Alloys, 1903, p. 262. 
 
428 
 
 STEEL RAILS 
 
 the old grains thus distorted break up, and that the metal rearranges itself into 
 new and equiaxed grains. 
 
 But these new grains assume a size normal, not to the temperature at which 
 the old ones had formed, but rather to the temperature now existing; during the 
 rolling the temperature is constantly falling; each pass through the rolls tends 
 more or less fully to break up the preexisting grain, and to substitute for it a 
 new grain of a size more nearly normal to the 
 now lower temperature. To speak more accu- 
 rately, the new grain size approaches that 
 normal for the existing temperature; but the 
 result is much the same. For if each of a 
 succession of passes through the rolls breaks 
 up the existing grain, and substitutes for it a 
 new one, then each new grain will be smaller 
 than the preceding, because the normal towards 
 which it tends is smaller than the normal 
 towards which its predecessor tended at the 
 higher temperature then existing. 
 
 Fig. 309 attempts to express this condi- 
 tion of affairs graphically. Here ordinates 
 represent temperature and abscissae coarseness 
 of grain. The line Ac±A may be taken as 
 representing roughly the normal size of grain, 
 D", which steel of given composition tends to 
 assume with varying temperature, or the line 
 of normal coarseness of grain. If the grain is 
 smaller than the normal for existing tempera- 
 ( Howe -) ture, it always tends to grow and to approach 
 
 that normal. If it is coarser than that normal, it does not tend to shrink 
 back towards the normal, except when the temperature is rising past Ac s . 
 
 Let us suppose that we cease rolling a piece of steel while its temperature 
 is at B, the mechanical work of the rolls having broken the grain down. During 
 subsequent cooling the grain will grow, somewhat as sketched in the line BCE. 
 If, however, we resume rolling when the grain has reached C, we will break down 
 the grain, and drive it back, say to D. And so, keeping on, between passes the 
 grain grows and the temperature simultaneously falls, while at each pass the 
 squeeze which we give the metal breaks up the grain, and the curve of grain 
 and temperature follows the zigzag line BCDG. 
 
 
 I 
 
 I 
 
 P E 
 
 Fig. 309. — The Influence of the Finish- 
 ing Temperature on the Size of Grain. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 429 
 
 If we cease rolling when the temperature has fallen to G, then the grain 
 will grow as the metal cools, till the line of the actual size of grain intersects 
 that of the normal size, the line Ac x A; with further cooling no further growth 
 ensues, and the final size of grain is OP. If we had quenched the metal while 
 at G, the final size of grain would have been OH. If we had ceased rolling 
 when the temperature was at B, the final size of grain in the cooled steel would 
 have been OE. 
 
 Generally speaking, the grain size will be the coarser the higher the finish- 
 ing temperature. Fig. 310 illustrates this principle. This shows the micro- 
 structure of two like bars of the same steel, of which each had first been heated 
 
 Cooled to 963° C. Cooled to 837° C. 
 
 Fig. 310. — Influence of Finishing Temperature on the Size of the Grain of Steel 
 
 of 0.50 per cent Carbon (Howe.) 
 
 to 1394° C, then cooled slowly to the temperature indicated in the figure, then 
 rolled, and then cooled slowly, so that these temperatures are the " finishing 
 temperatures." Note how much coarser the meshes are in A, finished at 
 963° C, than in B, finished at 837°. 
 
 Professor Sauveur's micrographs of rail structure show that in the section 
 of a given rail, the network, or size of the walled cells, is the coarser the higher 
 the temperature at which the rolling is finished.* 
 
 According to Howe, when a piece like a rail, which is highly heated, is 
 rolled with such heavy reduction as to distort the austenite grains greatly, the 
 distorted and hence unstable grains immediately shatter, and their remains im- 
 mediately begin growing again by coalescence. This is repeated as often as the 
 piece is greatly reduced by the rolling. Each of the grains of austenite, formed 
 by coalescence after the last of these reductions, in cooling down to the Ar t 
 point, gives birth to a walled cell by ejecting to its outside the ferrite which it 
 generates. Hence it is the size which these austenite grains reach after this last 
 
 * Trans. American Institute of Mining Engineers, Vol. XXII, 1893, pp. 546-557, and especially 
 Plates TV and V. 
 
430 
 
 STEEL RAILS 
 
 effective reduction that determines the network size of the cold steel, and as 
 regards the opportunity for network growth, the finishing temperature is the 
 equivalent of the highest temperature reached by objects not rolled. 
 
 This well-marked network common to rail steels is probably due to their 
 large manganese content, as otherwise, on account of their slow cooling, the 
 ferrite would coalesce and break up the network.* 
 
 -^ 
 
 
 SIZE OF GRAIN 
 
 Fig. 311. — Diagram of Results of Experiments on Rolling at Different Temperatures. 
 
 This principle of governing the grain size by means of the finishing tempera- 
 ture is of very great importance. In general, we should be inclined by considera- 
 tions of economy of power to roll steel as hot as we dare, because the hotter it is 
 the softer it is, and the less power is consumed in rolling. But this would natu- 
 rally lead to a high finishing temperature, and thus to coarseness of grain and 
 brittleness. Hence a high temperature is desirable as regards power consump- 
 tion, but undesirable as regards the quality of the steel. 
 
 f Fig. 311 shows graphically the results of experiments made at the Spar- 
 
 * For illustration of this cellular structure in rails see Job, The Metallographist, Vol. 5, 1902, 
 pp. 177-191; P. H. Dudley, ibid., Vol. 6, 1903, p. 111. 
 
 t The Manufacture and Properties of Iron and Steel, H. H. Campbell, 1904, p. 410. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 431 
 
 rows Point plant of the Maryland Steel Company. As in Fig. 309 ordinates 
 represent temperature and abscissae coarseness of grain, the grain growing 
 coarser from left to right. An ingot was rolled into blooms and two adjacent 
 blooms, "A" and " B," were rolled into rails without further heating, the first, 
 "A," being held before rolling in order to allow it to cool so that all work 
 should be done at as low a temperature as possible, without, of course, reaching 
 the lower critical point, while the second, " B," was rolled as quickly as possi- 
 ble through all the passes, except the last, but was then held at the finishing 
 pass If minutes; the result being that both pieces went through the finishing 
 pass at the same temperature, which was about 750° C. (1382° F.). 
 
 Fig. 312. — Rail "B" Near Surface, 
 46 Dia. — Campbell. 
 
 Fig. 313. — Rail "A" Near Surface, 
 46 Dia. — Campbell. 
 
 Fig. 312 shows " B " rail near the surface. 
 Fig. 313 shows " A " rail near the surface. 
 
 Fig. 315. — Rail 
 46 Dia. ■ 
 
 Center of Head, 
 Campbell. 
 
 Fig. 314 is from the center of the head of " B " rail. 
 Fig. 315 is from the center of the head of " A " rail. 
 While Figs. 312 and 313 appear similar, Figs. 314 and 315 show the real 
 difference between the two rollings. The last pass does very little work; there- 
 
432 
 
 STEEL RAILS 
 
 fore, holding the rail before the last pass does little good, except on the outer 
 surface of the rail, and a low shrinkage or finishing temperature does not neces- 
 sarily mean that the rail will have a good grain throughout. 
 
 Fig. 318. — Side View at Top of Head, 70-lb. Rail, 
 50 Dia. (Am. Ry. Eng. Assn.) 
 
 Fig. 319. — Side View at Center of Head, 
 70-lb. Rail, 50 Dia. (Am. Ry. Eng. Assn.) 
 
 Figs. 316 to 321 presented by Wickhorst* illustrate the different grain found 
 in the top and center of a new 70-pound Bessemer rail. A section about \ inch 
 
 * Flow of Rail Head under Wheel Loads, M. W. Wickhorst, Am. Ry. Eng. & M. of W. Assn., 
 Vol. 12, Part 2, 1911, p. 535. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 thick was taken from the rail and two pieces cut from it for microscopic test, 
 as shown in Fig. 322. These pieces were polished on three sides, etched with 
 
 Fig. 320.— Transverse View at Top of Head, 70-lb. 
 Rail, 50 Dia. (Am. Ry. Eng. Assn.) 
 
 Fig. 321. — Transverse View at Center of 
 Head, 70-lb. Rail, 50 Dia. (Am. Ry. 
 
 Eng. Assn.) 
 
 10 per cent solution of nitric acid in alcohol, and microphotographs made, mag- 
 nified 50 diameters. Thus, horizontal, vertical transverse, and vertical longi- 
 tudinal sections were obtained at 
 the top and at the center of the 
 head. 
 
 There will always be some 
 difference between the structure of 
 the center of the head and the por- 
 tion near the surface, but when the 
 rail is rolled at a proper temperature 
 during the passes, when considerable 
 work is put upon the piece, this 
 difference will not be serious. 
 
 The effect of finishing temper- 
 ature is not fully agreed upon, and 
 many rolling-mill men feel that 
 the properties of the steel depend 
 quite as much on the amount of 
 reduction in the rolls as upon the finishing temperature. 
 
 Fig. 322. — Pieces for Microscopic Views shown in 
 Figs. 316 to 321. 
 
434 STEEL RAILS 
 
 * To try to arrive at some conclusion in this matter, a number of tempera- 
 ture readings were taken with both the Fery and Wanner pyrometers and 
 checked against a thermo-couple at one or two large plants. For steel rails the 
 finishing temperatures as indicated by the optical pyrometers averaged 1050 
 to 1100° C.;f for structural steel, 950 to 1000° C. And yet such material is not 
 coarse-grained. (Reheating to such a temperature would give rail steel a very 
 coarse grain.) A difference of over 100° C. in finishing temperature could not 
 be detected in the size of grain, but a difference in section could very soon be 
 noticed. Similar results can be reached experimentally by rolling out small 
 sections at different temperatures. A section heated to 1300° C. and rolled 
 out with 30 per cent reduction showed about the same sized grain as one heated 
 to 1300° C, cooled to 900°, and rolled out, the finishing temperature being about 
 700° C. 
 
 The question of rolling at a low temperature is one that has occupied the 
 minds of engineers for a long time, and the fact that at present no solution has 
 been made which is satisfactory to both the manufacturer and the consumer 
 is evidence that it is not easily put aside. 
 
 In rolling early rails it was recognized that mechanical defects would be de- 
 veloped to a greater or less extent by the rolling process, and therefore the bloom, 
 about 7 inches square, was conveyed from the blooming rolls to a steam hammer 
 by which all visible cracks or defects were chipped out, care being taken to cut 
 to the bottom of the imperfection, and not leave any pronounced shoulders at 
 the edges of the resulting depressions; and until the adoption of automatically 
 operated tables attached to the rail rolls, if the partially formed rails still showed 
 defects, the operation of rolling was halted, while such places were chipped out 
 by hand. These were usual practices, and were not abandoned because of their 
 results being unsatisfactory, but on account of the time consumed and the 
 expense incurred. 
 
 In the formation of the grooves in the rolls much damage can be and often 
 is done to the steel. With the object of increasing the product of a given mill, 
 the ingot is rolled off at one heat, with heavy reductions in each pass so as to 
 reduce the number of passes and consequently the time taken in rolling. 
 
 It is interesting to turn to the following review of the English practice by 
 Mr. Talbot: f " Our practice "s to take large ingots and have a furnace between 
 
 * Some Practical Applications of Metallography, Campbell. Proceedings American Society for 
 Testing Materials, Vol. VIII, 1908, p. 353. 
 
 t These temperatures are from 100 to 200 degrees higher than the usual practice. 
 t On Rail Steel as Manufactured by the Continuous Open-hearth Process, Talbot. Proceedings 
 American Society for Testing Materials, Vol. VII, 1907. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 435 
 
 the cogging and finishing mills,* which has the effect of acting as an equalizer 
 so that the blooms are delivered to the finishing mill at an even temperature, 
 making the bar more easily shaped, and the flange of the rail is sent out of the 
 finishing groove at a temperature nearer to that of the head than has hitherto 
 been possible. This, no doubt lessens the strains set up in cooling on the hot 
 banks. Our practice is to increase the number of passes, decrease the amount 
 of reduction per pass, and get the product by increased speed of the rolls and 
 not by digging into and tearing the metal, as is done in the case where too few 
 passes and heavy drafts are adopted. 
 
 " With regard to rolling temperature, we may say we roll a 100-pound rail 
 in lengths which give, after crops are cut off, three lengths of 10 meters. The 
 first length is cut within 15 to 20 seconds after leaving the finishing groove of 
 the mill, and on this we allow 7f inches shrinkage. The next length is cut within 
 35 to 50 seconds from leaving groove, and here 1\ inches is allowed. The third 
 is within 60 to 80 seconds, and 7 inches is allowed for shrinkage. Of course these 
 allowances only apply to 100-pound rails; less allowance is made in lighter rails." 
 
 | At one of the large rail mills they formerly had a table on which the rails 
 were held before they went through the finishing pass. The scheme of holding 
 the rail before the finishing pass, where only a small amount of work is put upon 
 the metal, as illustrated in Figs. 312, 313, 314, and 315, while giving a low 
 finishing temperature, does not necessarily decrease the size of the grain. 
 
 Starting with such a scheme as a basis and designing the rolling mill to 
 hold the pieces to allow them to cool before the passes where the most work 
 is done, and also arranging for the sorting out of the blooms to equalize the 
 finishing temperature, would give a better arrangement and would have merit 
 if applied to mills where the rails are finished too hot as a direct result of pro- 
 ducing large tonnage. However, even with the most careful manufacture, a 
 certain amount of heat on the bloom is necessary in order to take the A. S. C. E. 
 sections through with the flanges properly filled out, and it will not be possible 
 to reduce this temperature without the danger of the breakage of rolls, improper 
 filling out of the flange, and additional strains in the steel, which are necessarily 
 detrimental, unless the design is changed to make all parts of the section more 
 nearly in balance as to the temperature of finish. 
 
 An interesting experiment may be tried on a tee rail, which has been finished 
 and straightened. Take 6 or 8 feet of rail and place it on a planing machine 
 and cut the head off the web at the point where the web joins the head, and both 
 
 * This is customary at a number of rolling mills in America. 
 
 t Kennedy-Morrison Process. The Iron Age, December 20, 1900, pp. 16-18. 
 
436 
 
 STEEL RAILS 
 
 the head and bottom portion will spring out of a straight line, sometimes to a 
 very marked extent, thus showing that great internal strains are there. This 
 is a condition that cannot be avoided by the manufacturer without some help 
 from the rail designer. 
 
 The better distribution of metal in the new American Railway Association 
 sections gives a rail that can not only be rolled at a lower temperature, but 
 which is much less liable to injury in the straightening press. 
 
 The rolling of these rails has developed some surprises. It was expected 
 that the rails could be rolled at a lower temperature than the old sections, and 
 that the shrinkage allowance could be reduced ; but it was found that under 
 the same conditions the new section would require a greater shrinkage allowance 
 than the old. The rails were unquestionably rolled colder than the old section, 
 with the exception of the thin flange, but it was this thin flange that determined 
 the shrinkage of the old rails. In going through the cambering wheels the head 
 was stretched, giving the hot head a greater length for shrinkage than the base. 
 In the new section the temperature is nearly uniform and much colder than the 
 head of the old rail was, but no part of the new rail is as cold as the thin base 
 of the old rail ; consequently a greater shrinkage allowance is required. 
 
 At Gary* the ingots are bloomed to 8 by 8 inches in 9 passes and finished 
 in 9 passes, making a total of 18 passes from ingot to rail. The reduction of 
 the rail is from 400 square inches at the bottom of the ingot to 10.1 square inches, 
 or a reduction of 39 times. The areas of the various passes, as furnished by the 
 Steel Company at Gary, of their section 10030, which is the A. R. A. type " B " 
 100-pound rail, are as follows: 
 
 Pass Number. 
 
 Area. 
 
 Pass Number. 
 
 Area. 
 
 
 Square Inches. 
 
 
 Square Inches. 
 
 Ingot 
 
 400 
 
 10 
 
 43.2 
 
 1 
 
 376.6 
 
 11 
 
 32.9 
 
 2 
 
 282.4 
 
 12 
 
 25.2 
 
 3 
 
 214.9 
 
 13 
 
 21.5 
 
 4 
 
 164.8 
 
 14 
 
 17.8 
 
 5 
 
 130.3 
 
 15 
 
 16.4 
 
 6 
 
 107.9 
 
 16 
 
 13.2 
 
 7 
 
 88.9 
 
 17 
 
 10.7 
 
 8 
 
 70.8 
 
 18 
 
 10.1 
 
 9 
 
 58.9 
 
 
 
 At the Maryland Steel Company f the blooms remain in the soaking pit 
 about 1 hour and 25 minutes, and are rolled to 7| by 7f -inch blooms in 13 passes, 
 the top end of the ingot forward, and turned after each two passes. The blooms 
 
 k Report to Rail Committee, Proceedings Am. Ry. Eng. & M. of W. Assn., Vol. 12, Part 2, 
 
 1911, p. 
 
 t Report to Rail Committee, Proceedings Am. Ry. Eng. Assn., Vol. 12, Part 2, 1911, p. 3 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 437 
 
 are rolled directly into rails without reheating, in 11 passes, making a total of 
 24 passes from ingot to rail. The area of cross section in square inches in the 
 various passes for the A. R. A. type " B " 90-pound rail, Maryland Steel 
 Company, section No. 162, is about as follows: 
 
 TABLE XCIIL- 
 
 - REDUCTION OF AREA IN 90-LB. A. R. A. TYPE "B" RAILS ROLLED 
 AT MARYLAND STEEL COMPANY 
 
 (A. R. E. Assn.) 
 
 Operation. 
 
 Number of 
 Pass. 
 
 Area. 
 
 *— ■ 
 
 Number of 
 Pass. 
 
 Area. 
 
 Operation. 
 
 Number of 
 Pass. 
 
 w 
 
 In ot 
 
 
 Sq. In. 
 416 
 
 374 
 336 
 300 
 260 
 234 
 205 
 172 
 150 
 
 Blooming. 
 Blooming. 
 Blooming. 
 Blooming. 
 . Blooming. 
 Roughing. 
 Roughing. 
 Roughing. 
 Roughing. 
 
 9 
 10 
 11 
 
 12 
 13 
 1 
 2 
 3 
 4 
 
 Sq. In. 
 
 132 
 
 117 
 96 
 74 
 58 
 
 46.2 
 34.4 
 31.3 
 23.1 
 
 Roughing 
 
 Roughing 
 
 Intermediate... 
 Intermediate... 
 Intermediate.. . 
 Intermediate... 
 Finishing 
 
 5 
 6 
 
 7 
 8 
 9 
 
 10 
 11 
 
 Sq. In. 
 
 21.5 
 
 Blooming. . . 
 Blooming... 
 Blooming. . . 
 Blooming. . . 
 Blooming. . . 
 Blooming. . . 
 Blooming. . . 
 Blooming. . . 
 
 1 
 
 2 
 3 
 
 4 
 5 
 
 6 
 
 7 
 8 
 
 17.8 
 16.0 
 15.2 
 12.3 
 9.7 
 9.0 
 
 Table XCIV presents data on American rolling-mill practice. Fig. 323 
 shows the arrangement of the rail mill of the Algoma Steel Company's plant. 
 This consists of one 32-inch rev. bloom mill, three Seimens heating furnaces 
 for reheating the blooms, and one 23-inch three-high rail mill. 
 
 Plate XXX illustrates a reversing cogging mill and Plate XXXI and 
 Fig. 324 three-high mills. The left hand illustration of Fig. 325 shows a set 
 of rolls for use in a three-high mill, according to common English practice, in 
 which the bottom and middle rolls are grooved to receive the rail, while the 
 "closers" are on the middle and top rolls, and the guards to peel the bars out 
 of the grooves rest by gravity on the bottom and middle rolls. 
 
 In the American three-high mill, shown on the same figure, the top and 
 bottom rolls are grooved and the middle roll serves as the "closer" for each of 
 the others, itself carrying no grooves at all; to enable the bars to be got out of 
 the top groove, the upper guard must be placed on the upper side of the bar as 
 it issues from the roll, and as it will not lie there by gravity, the guard has to 
 be kept up to the roll by a counterweight or spring, and is known as a balanced 
 guard.* 
 
 Fig. 326 shows the three-high rolls in the rail mill at Gary, f This mill is 
 equipped with 12 sets or stands of roll trains, all operated at varying speeds by 
 General Electric alternating-current motors, some of which are of the largest 
 
 * Steel, by Harbord and Hall, London, 1911, p. 625. 
 
 t Railroad Age Gazette, November 13, 1908, and The Iron Age, April 1, 1909. 
 
438 STEEL RAILS 
 
 TABLE XCIV. — DATA ON AMERICAN ROLLING-MILL PRACTICE 
 
 (Compiled by Committee on Rail, Am. Ry. Eng. Assn., 1909, and revised by the author 1912) 
 
 
 
 
 
 
 Number of 
 
 1 
 
 S 
 
 ms3| 
 
 
 
 
 Blooming Mill. 
 
 Passes in 
 
 fS . 
 
 | 
 
 |g£J 
 
 
 
 
 
 
 
 Rail Train. 
 
 g£ 
 
 (2 
 
 
 
 
 
 
 
 
 
 c ./ 
 
 
 
 
 
 
 
 
 
 
 
 Rail Mill. 
 
 Location. 
 
 !" 
 
 .S ™ o 
 
 s 
 -J 
 
 | 
 
 1 
 
 E 
 
 Is 
 
 If 
 1° 
 
 - 
 
 r - i ' 
 
 llll 
 
 Remarks. 
 
 
 
 
 m. s. 
 
 
 
 
 
 Sq. In 
 
 
 m. s. 
 
 
 Algoma Steel Co. 
 
 Canadian Soo, 
 Canada. 
 
 m"xw 
 
 t2 30 
 
 t8"X8" 
 
 19 
 
 8 
 
 3 
 
 
 30 
 
 
 Blooms reheated. 
 
 tBethlehem Steel 
 
 Bethlehem, 
 
 19"X23" 
 
 1 20 
 
 8"X8" 
 
 15 
 
 10 
 
 
 
 
 6 
 
 Ingots are reheated after 
 
 Co. 
 
 Pa. 
 
 
 
 
 
 
 
 
 
 
 being taken from molds 
 and are rolled direct with- 
 out further reheating. 
 
 "Cambria Steel Co. 
 
 Johnstown, Pa. 
 
 20"X23" 
 
 
 8J"X10" 
 
 15 
 
 
 2 
 
 25.00 
 22 00 
 
 27 
 
 6 45 
 
 Ingots are reheated, rolled 
 into blooms, again re- 
 heated and rolled into 
 rails. No rest is given 
 prior to finishing pass. 
 
 Carnegie Steel Co. 
 
 Braddock, Pa. 
 
 18"X20" 
 
 
 9*"X9F 
 
 7 
 
 fl2 
 
 
 38.53 
 
 f20 
 
 4 41 
 
 Ingots are reheated, rolled 
 into blooms, again re- 
 heated and rolled into 
 rails. Rails are given a 
 rest immediately prior to 
 finishing pass of from 
 
 Carnegie Steel Co . 
 
 Braddock, Pa. 
 
 16rX18i" 
 
 
 
 7 
 
 
 
 
 
 
 Old mill. Light rails 360 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 Rails not held between 
 
 tCarnegie Steel Co. 
 
 Youngstown, 
 
 Ohio. 
 Pueblo, Colo. 
 
 19"X21" 
 
 50 
 
 8'X8° 
 
 9 
 
 10 
 
 1 
 
 35 50 
 
 20 
 
 4 50 
 
 passes. 
 
 Colorado Fuel & 
 
 20"X18" 
 
 
 
 17 
 
 7 
 
 5 
 
 
 29 
 
 
 
 Iron Co. 
 
 
 
 
 tDominion Iron & 
 
 Sydney, 
 
 18"X21" 
 
 
 
 17 
 
 10 
 
 
 
 28 
 
 
 
 Steel Co. 
 
 Canada. 
 
 
 
 
 
 
 
 
 
 
 
 ♦Illinois Steel Co. 
 
 S.Chicago, 111. 
 
 tl8"X19£' 
 
 t50 
 
 8"X8" 
 
 9 
 
 4 
 
 5 
 
 32.89 
 
 18 
 
 
 Rails held 30 to 60 seconds 
 before finishing pass. 
 First rails shipped Feb. 26, 
 
 Indiana Steel Co. 
 
 Gary, Ind. 
 
 20"X24" 
 
 
 
 9 
 
 8 
 
 1 
 
 
 18 
 
 
 
 
 
 1909. 
 
 •Lackawanna Steel 
 
 Co. 
 'Maryland Steel 
 
 Buffalo, N. Y. 
 
 19"X19" 
 
 2 45 
 
 8"X8" 
 
 6 
 
 4 
 
 5 
 
 49 5 
 
 15 
 
 6 5 
 
 No rest prior to finishing 
 
 Sparrows Pt., 
 
 20"X21" 
 
 2 10 
 
 7|''X7r' 
 
 13 
 
 6 
 
 5 
 
 27 7 
 
 24 
 
 5 50 
 
 Ingots are reheated after 
 
 Co. 
 
 Md. 
 
 
 
 
 
 
 
 
 
 
 being taken from molds 
 and are rolled direct with- 
 out further heating. No 
 rest is given except that 
 which occurs when bloom 
 is being sheared. 
 
 •Pennsylvania 
 
 Steelton, Pa. 
 
 18| "X18^" 
 
 1 23 
 
 
 9 
 
 6 
 
 5 
 
 
 20 
 
 3 
 
 
 Steel Co. 
 
 
 
 
 
 
 
 
 
 
 
 bria Steel Co.f 
 
 tTennessee Coal, 
 
 Birmingham, 
 
 19"X23" 
 
 1 15 
 
 8"X8" 
 
 13 
 
 
 
 
 
 5™ to 6"° 
 
 
 Iron & R.R. Co. 
 
 Ala. 
 
 
 
 
 
 
 
 
 
 
 
 t Compiled by author. 
 
 sizes ever constructed for industria service. These are housed in a separate 
 bay, or lean-to, running parallel with the rolls. The rotors are 20 feet in 
 diameter and have a speed of 83 revolutions per minute. All of the motors are 
 connected directly to the roll trains by regular mill couplings. Although the 
 motors are provided with flywheels and run in one direction, provision is made 
 for reversing in case of necessity. The control system has been worked out 
 with the greatest nicety, all operations being under the instant control of the 
 operator by means of a master controller. 
 
 The first group of rolls consists of four stands of continuous 40-inch mills, 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 439 
 
440 
 
 STEEL RAILS 
 
 
 •® 
 
 "4 
 
 m 
 
 Fig. 324. — Housing far 28-inch Three-high Mill. (Harbord and Hall.) 
 
 English Three-high Rail Mill. American Three-high Rail Mill. 
 
 Fig. 325. — Rolls used in Three-high Rail Mills. (Harbord and Hall.) 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 441 
 
 each two of which are driven by a 2000-h.p. motor. They are arranged in 
 tandem, requiring no manipulation from stand to stand. Here, as elsewhere 
 through the plant, sufficient distance is left between successive sets of rolls to 
 enable a quarter turn of the ingot or bloom to be made, so that it is worked 
 equally on all sides. The first two mills are equipped with 42-inch rolls, 
 enabling 20-inch by 24-inch ingots to be used. After passing these four mills 
 the ingot is sent to a 40-inch three-high blooming mill equipped with lifting 
 
 sILiH 
 
 Fig. 326. — Three-high Rolls in the Rail Mill at Gary. (Scientific American.) 
 
 tables and arranged with a combined hydraulic and pneumatic balancing device. 
 This mill, which is driven by a 6000-h.p. motor, gives the ingot five passes. 
 
 After being bloomed the ingot is sheared in a 10-inch by 10-inch horizontal 
 blooming shear, and the crop ends or butts are taken outside of the mill by a 
 butt conveyor of unusual construction, which was designed and built by the 
 engineers of the Indiana Steel Company. Each bloom then goes through a 
 28-inch roughing mill, which is three-high and equipped with tilting tables. 
 This mill has three stands or rolls. The roughing stand, however, is the only 
 one that is three-high, the other two stands being two-high. The mill is driven 
 
STEEL RAILS 
 
 by a 6000-h.p. motor and gives the bloom three passes. After leaving the rough- 
 ing mill the bloom goes through a two-high 28-inch forming mill driven by a 
 2000-h.p. motor, receiving but one pass. Then it is sent to finishing mills, which 
 consist of five stands of 28-inch mills driven by two 6000-h.p. motors. 
 
 After the dummy pass, the bloom is transferred to the first edging, which 
 is in this same mill but the second stand, and turns back on an elevated table 
 to the second edging, which is in line with the 28-inch roughing mill. It then 
 travels by chain transfer to the lower tables and on the leading pass goes through 
 
 FINISHING PASS 
 £8"x69/*" 
 
 k 
 
 DUMMY ZB"x40" 
 
 
 ?x 
 
 ■ ■ 
 
 BLOOMING PASS 
 40" x 90" 
 
 2N. D ROUGHING PASS 
 28" x 56" 
 
 ROUGHING PASS 
 28"x60" 
 
 Fig. 327. — Pass Diagram, Rail Mill, Illinois Steel Company, South Works. 
 
 a stand, which is also in line with the roughing mill and driven by the same 
 motor, and continues on to the third stand of the 28-inch finishing mill, this 
 being the eighteenth and last pass. After the finishing pass the rail travels 
 through to the saws, of which there are five provided, thus cutting four rails to 
 length. These four rail lengths consist of half the ingot. As the capacity of 
 this mill is 4000 gross tons per 24 hours, it will be seen that there must be a 
 four-rail length sawed about every half-minute. The saws have 42-inch 
 blades, arranged to be raised and lowered in unison by one controller from 
 the hot-saw operator. 
 
 The pass diagram of the rail mill at the South Works plant of the Illinois 
 Steel Company is illustrated in Fig. 327. The Bessemer ingot is 18 inches by 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 443 
 
 -lH3^-<^^&-|^ 
 
 r 
 
 
 °£* 
 
 
 o<> 
 
 3= LD 
 
 
 
 
 
 £ 
 
 $ 
 
 
444 
 
 STEEL RAILS 
 
 19| inches, the heating capacity is 192 ingots (24 single-hole pits). The ingot 
 is worked direct to rail without reheating. The blooming mill is 40-inch pitch 
 diameter three-high, and the ingot is given 9 passes and reduced to an 8-inch by 
 8-inch or 8-inch by 8^-inch bloom. The number of rail lengths rolled are three 
 and four. 
 
 The finishing mill consists of one stand 28-inch P. D. three-high first rough- 
 ing rolls, one stand 28-inch P. D. two-high second roughing rolls, one stand 28- 
 inch P. D. two-high dummy rolls,one stand 28-inch P.D. three-high finishing rolls. 
 
 The number of passes from ingot to rail is as follows: 
 
 Passes. 
 
 Blooming 9 
 
 First roughing 3 
 
 Second roughing 1 
 
 Finishing 4 
 
 Total 18 
 
 Fig. 328 presents the general arrangement of the rail mill. 
 Table XCV shows the shrinkage allowed by American rail mills and 
 Fig. 329 the lengths of saw runs. 
 
 TABLE XCV. — SHRINKAGE ALLOWED BY AMERICAN RAIL MILLS 
 
 (Compiled by Committee on Rail, Am. Ry. Eng. Assn., 1909, and revised by the author 1912) 
 
 No. of 
 
 Rail Mill. 
 
 Location. 
 
 Time 
 of 
 
 Saw 
 
 Shr 
 
 nkage Allowed by Mills on 33-foot Rails. 
 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 1 
 
 Algoma Steel Co. . . . 
 *Bethlehem Steel Co. 
 *Cambria Steel Co. . . 
 
 Carnegie Steel Co. . . 
 
 Carnegie Steel Co. . . 
 
 Carnegie Steel Co. . . 
 
 Colo. Fuel & Iron Co. 
 
 Dominion Iron & 
 
 . Steel Co 
 
 ♦Illinois Steel Co 
 
 Indiana Steel Co. . . . 
 *Lackawanna Steel 
 
 Co 
 
 ♦Maryland Steel Co... 
 ♦Pennsylvania Steel 
 Co 
 
 Tenn. Coal, Iron & 
 R.R. Co 
 
 Canadian Soo, Canada. 
 
 Bethlehem, Pa 
 
 Johnstown, Pa 
 
 35 
 
 20 
 
 17-19 
 12-14 
 
 12-17 
 32 
 
 17 
 10-20 
 
 16 
 
 26 
 
 5 
 
 27 
 
 
 
 6| 
 6i 
 
 6& 
 
 6i 
 
 6& 
 ' 61 ' 
 
 
 
 
 2 
 3 
 
 4 
 
 6 
 
 5f 
 
 6iV 
 
 6i 
 61 
 
 6A 
 
 61 
 51 
 6| 
 
 5 
 
 Braddock, Pa 
 
 Youngstown, Ohio. . . . 
 Pueblo, Colo 
 
 7! 
 
 6 
 
 6^ 
 
 ' 6 " 
 
 6£ 
 
 6f 
 6& 
 
 
 
 6 
 
 6| 
 6| 
 
 6! 
 
 6} 
 
 6J 
 6A 
 
 6| 
 
 5J 
 
 6| 
 6H 
 
 6A 
 6f 
 
 61 
 6| 
 
 6f 
 
 61 
 
 
 
 7 
 
 
 
 8 
 
 Sydney, Canada 
 
 South Chicago, 111. . 
 Gary, Ind 
 
 Buffalo, N. Y 
 
 Sparrows Point, Md. 
 
 Steslton, Pa 
 
 Birmingham, Ala 
 
 
 6* 
 
 6A 
 
 9 
 10 
 
 6& 
 
 6A 
 
 61 
 6 
 
 6| 
 
 61 
 
 
 11" 
 
 12 
 13 
 
 14 
 
 
 6* 
 6 
 
 6| 
 
 
 
 
 Note. — Information from R. W. Hunt & Co., except that marked (*), wh 
 seconds consumed from the time rail leaves finishing rolls till ss 
 
 s obtained direct from manufacturers by the 
 
 After leaving the saw the rails pass through the cambering machine and 
 are given a head sweep (Fig. 330) of from 3 to 8 inches, the A. S. C. E. section 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 SAW RUNS 
 
 2 I 2 RAIL STOP 
 
 +-H 
 
 ^_^_ 
 
 143- 9" 
 
 SAW NO. 1 1 
 
 -b^f" 
 
 - 193'- 3%- 
 
 SAW NO. I | ,1 RAIL STOP 
 
 — 105- O" 
 
 SAW NO, 
 
 H- 
 
 I RAIL STOP 
 
 — 107-0" — 
 SAW NO. I 
 
 197"-0"- 
 
 SAW NO. I 
 
 SANA/ NO.I I 21 .31 
 
 il Is I 3ll 
 
 is I 3IRAILSTOP 
 
 —212-0" 
 
 SAW NO. I 
 
 24K-0" 
 
 3Q'AND 6Q'R 
 
 *H— *!- 
 
 I 69 -O" — >• 
 
 SAW NO.I I 2 I 5 I 
 
 < | I6 '_ B " __^| I |- 
 
 Fig. 329. — Saw Runs of American Rail Mills. (See Table XCV for number of mills.) 
 
 requiring a greater ordinate than the A. R. A. rails. The rails then pass to the 
 hot beds and after being allowed to cool are transferred to the gagging or cold- 
 straightening presses (Fig. 331). 
 
 Fig. 330. — Head Sweep. 
 
 In the heavier A. S. C. E. sections the gagging or straightening of the rail 
 after it leaves the rolls and is cooled in the hot beds tends to develop injurious 
 strains in the base and web. In some of the most modern English plants the 
 
446 
 
 STEEL RAILS 
 
 rails are run while still hot through straightening rolls, thus much reducing 
 the labor of final straightening.* 
 
 Fig. 331. — Cold Straightening Press, Maryland Steel Company. 
 
 After straightening, the rails are inspected, drilled, reinspected, and loaded 
 on cars for shipment. Many rails were formerly damaged in loading, but at 
 several mills the use of a magnetic crane is now employed for loading the rails. 
 
 The brand on rails gives the name of the manufacturer, a number or abbrevia- 
 tion by which the rail section is designated, the month and year of manufacture, 
 
 * S. von Schukowski 
 schienen im katten und 
 
 vocates straightening rails while still hot. Das richten von eisenbahn- 
 zustande. Stahl und Eisen, Vol. 27 (1907), Pt. 1, p. 797. 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 447 
 
 and, if the metal is open-hearth steel, the letters " 0. H." are also added. Some- 
 times the letters "F. T." are added to signify ferro-titanium steel. Square 
 block letters and figures about one inch high are commonly used, and, as these 
 are cut into one of the rolls of the last pass, the brand will always appear slightly 
 raised at regular intervals on the web of the rail. The month is generally 
 shown by Roman numerals, as VII for July, and sometimes by a series of I's, 
 as IIIII for May. 
 
 
 1 
 
 
 ' 3 \ 
 
 
 
 
 
 
 
 
 IO 
 
 E 
 
 ?>eo 
 
 Q) 40 
 
 ^ 20 
 
 O 
 
 1 
 
 
 1 
 
 l" 
 
 
 
 
 
 
 
 
 
 
 
 ■1 
 
 \ 
 
 
 
 \ 
 
 
 
 
 
 X 
 
 
 
 •s 
 
 \^-A 
 
 
 
 V 
 
 
 
 20 
 
 
 
 23 
 
 
 HS ^ 
 
 
 
 21 
 
 
 " 
 
 1 
 
 00 
 
 IC 
 
 60 10 
 
 20 II 
 
 00 10 
 
 60 IC 
 
 20 9 
 
 30 1 
 
 OO 1060 I020 980 
 
 TEMPERATURE IN < 
 
 Fig. 332. — Value of V/E for Tables XCVI, XCVII and XCVIII. 
 
 The number representing the heat, blow or melt of steel, and the letter 
 to indicate the position of the rail in the ingot is stamped on the web of the 
 rail with dies while it is still red hot, but after it has been completely rolled and 
 sawed to length. As the brand always appears in raised letters, and the heat 
 number and letter is stamped on no confusion of the two should exist.* 
 
 A series of experiments were made in Germany by Dr. Puppef to determine 
 the power required to roll different sections. As this investigation presents 
 many features connected with the design of the section and rolls and the effect 
 of the temperature on rolling, it will be of interest to briefly review Dr. Puppe's 
 work as it relates to the heavier weights of rails examined. 
 
 The reversing mill on which these tests were made consisted of one cogging 
 mill housing and three finishing mill housings. A flywheel converter set of 
 the Ilgner system served to equalize the fluctuations of the power taken. The 
 mill was driven by three motors rigidly coupled together, and connected in i 
 
 * R. W. Hunt and Company, 1121, The Rookery, Chicago, 111., have published in convenient 
 form full information in regard to the practice of branding and stamping at the different mills. 
 
 t Experimental Investigations on the Power Required to Drive Rolling Mills, J. Puppe, London, 
 1910. See also for a full treatise of this subject Steel, Harbord and Hall, London, 1911, pp. 666-741. 
 
448 
 
 STEEL RAILS 
 
 series electrically; they had an aggregate output of 3600 H.P. normal and 
 10,350 H.P. maximum at a speed of 110 r.p.m. 
 
 Tables XCVI, XCVII and XCVIII contain the results" of the tests with 
 the heavier rails. The second line of the tables gives the time taken by the 
 pass (in seconds) determined from the rise and fall of the current and power 
 curves. 
 
 Line 3 gives the time intervals (in seconds) between successive passes which 
 were determined in the same way as the times of the passes. The time interval 
 between the first and second passes is given in the column headed " pass 1," 
 that between second and third passes in the column headed " pass 2," and so on. 
 The last column contains the sum of all the figures in the preceding columns. 
 
 2 
 
 I 
 
 k- 
 
 
 4 
 
 
 5 
 
 
 
 US 
 
 r 1 
 
 ' i 
 
 
 
 ~ 
 
 X 
 
 ~~ 
 
 f 300 
 
 j 
 
 < 260 * 
 
 
 ( — 270—9 
 
 
 *-186—> 
 
 
 Fig. 333. — Diagram of Cogging Rolls, Tables XCVI, XCVII and XCVIII, Dimensions in Millimeters. 
 
 The energy given up, or stored by, the flywheel and other rotating masses 
 was calculated from the speed curves and the moments of inertia. 
 
 To arrive at the light-load loss, the average speed was determined, and the 
 light-load loss was obtained by multiplying the power taken to drive the mill 
 light at this speed by the time. 
 
 The total work done in rolling the bar up to any pass is the sum of the work 
 done in each individual pass up to that point, taking due account of the motor 
 efficiency and leaving out the work expended in accelerating the moving parts 
 or in light-load losses, copper losses, etc. Frequently, the section of the ingot 
 during the first few passes could be determined only roughly, and then the length 
 of the bar after the second or third pass was taken as unity, and the summation 
 of the work done in rolling commenced correspondingly later. 
 
 The areas of the cross sections of the bar or ingot were obtained wherever 
 possible by cutting trial pieces off the bar, and measuring the area directly by 
 a planimeter. It was not always possible to cut samples from the bars during 
 the roughing passes, as the bar was too large for the bloom shears or hot saws, 
 and in such cases the cross-sectional area was obtained from the scale showing 
 the position of the rolls, while sections were cut from the bar in the last two 
 passes, and the area measured as a check. It is often found that billets are 
 
STEEL RAILS 
 
 
 <! (§ 
 
 *1 
 
 Ssl22tsisS s2 
 
 «oS !*»S3S 
 
 SSSmSSSSraotS" S Sj§2 
 
 
 
 
 
 
 Ssw". 
 
 ;p 
 
 °-Jf ■si'J'Sll - -si I J • ■ if 
 
 -a li 
 
 s s s „- 
 
 S i n ^ »- 
 S ^ - 
 
 ilSi 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 451 
 
 f-|2 3*l> 
 
 
 I s "! K 
 
 ;r 
 
 iw'°l 
 
 :2^S-g--^ .2 JVSjSg .._- s 
 
 . --- = o 3 .^.Sol- 
 
 -■■■:.■■ 
 
 ■g > £ l- " - ! r - ; v ■ ■■■' 
 
 SS 4 ^ | | % I g ., . : 
 
 Q.5 tf <HW<5'- - SsSJHalQ 
 
 - o , o.2 : 5 ' S 
 
STEEL RAILS 
 
 « s 
 
 W « 
 
 53SK==3^S=- 
 
 i§; 5§s£S3S$'° 
 
 SCSSSSSSSSS^K XgS^K I 
 
 „ Sr-oSgu, ■ 
 
 
 
 •/? — --.- — r, •= ° — 33 o ^ ~ i ~ 
 
 8§ C.52«aS?;^5.-„l:'5 a ^S d ?. 
 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 11 
 
 
 
 tgS&g|Sg3o§S|§2SS2S 
 
 !! 
 1 1 
 
 
 ! e 1 
 
 ^urn* 
 
 £ s .1 
 
 
 iSlliSilsii filing 1.1 
 
 
STEEL RAILS 
 
 I* 
 
 1 I 
 
 «-« SmSsJSw « 
 
 S S m S 2 Jo S S S§8_ rt S gSS 
 
 ggSSSSSSSISSSSSS Xgs«-§ 3 ; 
 
 SS^SScJSSSSSS^S ij^m 
 
 
 
 
 
 
 21s 
 
 Hlslfeg 
 
 i 
 
 :■> 
 
 : jl-a-lililiglil 
 
 I _ „ 
 
 li 
 
 
 .a|S 
 
 - *'| 
 S8 8 
 
 a d-s 
 If 1 
 
 lis 
 
 fli 
 I s s 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 455 
 
 §ISSI11 S 
 
 
 =11 
 
 § ll3 S : • o >■ ° fc £ J §.— ' 2 Z* (v c e 
 
 IjiIi|||IllillKlll|||l 
 
456 
 
 STEEL RAILS 
 
 rolled to different cross sections in the same grooves in the cogging mill, but this 
 is because the position of the screwing-down gear is altered. 
 
 The cross sections of the ingot in each pass with the reversing cogging mills 
 were determined from the position of the top roll. A special pointer was fixed 
 to the top roll, which indicated the position of this roll on a vertical scale pro- 
 vided for the purpose. The length of ingot or bar was calculated from its cross 
 section and weight. The crop ends were frequently cut off by bloom shears 
 after the roughing passes, in which case a corresponding allowance was made 
 in the calculations for the finishing passes. 
 
 Fig. 335. — Sections in which only "Direct Pressure" occurs in the Process of Rolling. (Puppe.) 
 
 The elongation was calculated from the cross sections and the lengths, such 
 calculations often being based on the cross section after the second or third pass, 
 in order to obtain greater accuracy, as already mentioned. The " volume dis- 
 placed " is obtained by multiplying the reduction in cross section by the length 
 of the billet before the pass in question, i.e., (Q x - Q 2 ) X L qi , where Qi = the 
 cross section of the ingot or billet before the pass, Q 2 = the cross section after 
 the pass, and L Qi = length of billet before the pass. 
 In rolling there are two means by which the 
 pressure is applied to the bar, viz., " direct " and 
 " indirect " pressure. By direct pressure " is 
 meant pressure which exists between the surfaces 
 of two rolls, as, for example, with rolls grooved 
 as shown in Fig 335. 
 
 The term ' indirect pressure " will be used 
 to denote the pressure usually existing between 
 the groove sides of one and the same roll, which 
 produces principally a reduction in a horizontal 
 direction (width) and not in a vertical direction 
 (height), as is the case with the sections shown 
 Fig. 336. — illustration of "indirect in Fig. 335. The following example will make 
 
 Pressure." (Puppe.) this dear: 
 
 If the flange of a rail (Fig. 336) be pressed in such a manner that its thick- 
 ness a-b and a'-b' is reduced, then this must be due to indirect pressure along 
 the whole line from c to d and c' to d'. It will be noticed that indirect pressure 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 457 
 
 takes place here both between the sides of the groove of the same roll as also 
 between the two rolls. 
 
 Where the particles move solely in an axial direction along the bar, the 
 energy required to accomplish this displacement is directly proportional to the 
 product (Qi — Q 2 ) X L qi , in which Qi and Q 2 represent the cross sections of two 
 consecutive passes, and L S1 the length of bar corresponding to the cross section 
 Qx. For shortness, we will denote this product, which represents the volume 
 
 displaced, by V. The fraction - 
 
 - Qg) X Lq, (in cu. mm.) 
 
 depends on the 
 
 work done in rolling (in m. kg.) 
 plasticity of the material, and, therefore, to a large extent on the temperature. 
 By plotting a curve with the above fraction calculated for each pass as ordi- 
 nates, and the corresponding temperatures as abscissae, we can obtain from this 
 the number of cubic millimeters of material which will be displaced per m. kg. 
 at various temperatures. 
 
 Let us now investigate the case in which the particles of the metal do not 
 move mainly in an axial direction along the bar, but also at right angles thereto. 
 A simple example is that of a bar which is flattened out by rolls having no grooves 
 to restrict the movement sideways. 
 
 The angle of incision into the bar should be such as not to cause excessive 
 spreading. Incisions at acute angles naturally cause greater spreading than 
 incisions at obtuse angles, but very often the proper form at the first passes can 
 only be obtained by incision of the bar at an angle from about 40° to 60°,* in 
 which case a considerable amount of lateral 
 spreading cannot be avoided. It is then 
 advisable to facilitate lateral spreading and 
 not to hinder the movement too much by 
 the sides of the grooves. 
 
 When designing rolls, successive profiles 
 should be arranged in such a way as to 
 obtain a smooth curve for V/E free from 
 such sudden jumps as occur, for instance, 
 in the curves in Fig. 332. It need hardly 
 be said that the rolling of simple sections, 
 such as flats, squares, rounds, etc., requires less power than the more compli- 
 cated rail sections in which the flow of the metal is brought about by indirect 
 pressure, and where the loss due to friction against the sides of the grooves is 
 great. This loss should, of course, be kept as low as possible, and equally dis- 
 
 Fig. 337. — Effect of Inclination of Inner 
 Surface of the Rail Flange on Energy re- 
 quired in Rolling. (Puppe.) 
 
 ' Bartholme, Stahl und Eisen, 1907, p. 58. 
 
458 STEEL RAILS 
 
 tributed amongst the various passes. It can be reduced to a minimum by 
 making the angle oo (Fig. 337) as small as possible, so as to obtain a large com- 
 ponent a. 
 
 From the point of view of economy in energy consumption when rolling 
 rails, it is very desirable that the inner surfaces of their flanges be inclined as 
 much as possible. If, however, the inclination of the inner surfaces, or the 
 thickness of the finished flange, has been fixed, the amount of indirect pressure 
 required for the formation of the flange can be reduced to a minimum by 
 working it as thin as possible at the first forming pass. 
 
 To sum up, the power required for complicated sections is greater than for 
 simpler sections of the same final cross-sectional area, but this extra power 
 depends to a very large extent on the skill with which the rolls are designed. 
 If the rolls have the most favorable shape, the values for V/E for the various 
 passes will be consistent with one another. 
 
 Plate XXXII contains curves which are calculated from the tests. The 
 shaded areas represent the energy supplied by the motors to the mill, and cover, 
 therefore, not only the work required for actual rolling, but also the power 
 required for running the mill light and for accelerating. The power taken by 
 the cogging mill generally increases towards the end of the pass, owing to the 
 acceleration of the rotating masses and the increase in the no-lead losses. 
 
 The same holds good for a few of the roughing and finishing passes. At the 
 last passes, however, the power taken decreases when the maximum speed has 
 been attained, and the rotating masses are no longer accelerated. The peaks 
 at the beginning of the curves for the last passes are due to this cause. The 
 relatively high peak in the lower curve for pass No. 23 may be explained by the 
 fact that maximum speed was reached in a very short time, as indicated by 
 the sharp rise in the speed curve. 
 
 The speed curve (Plate XXXII) usually drops rapidly when the ingot enters 
 the rolls, and sometimes it even falls to zero, and then rises again. Frequently 
 the speed increases suddenly at the end of the pass, especially in cogging mills. 
 These irregularities in speed are due to careless manipulation of the driver's 
 lever. In reversing mills the kinetic energy of the rotating masses is seldom 
 used to assist the motor, but it does happen occasionally. 
 
 From the acceleration curve (Fig. 338), it will be seen that a very large 
 amount of energy is required to accelerate the rotating masses of a reversing 
 mill; for instance, about 5300 h.p.-minutes are required to bring the rotating 
 masses of this reversing mill up to a speed of 120 r.p.m. Assuming that the 
 average speed is only 80 r.p.m., and that there are 20 passes, this means that 
 
INFLUENCE OF DETAIL OF MANUFACTURE 
 
 459 
 
 48,000 h.p.-minutes are expended in accelerating the rotating masses per ingot. 
 This is a large percentage of the total energy required for rolling. It must be 
 noted, however, that in the case of electrically driven reversing mills the energy 
 expended in accelerating the rotating masses is largely returned again in the 
 form of electric energy when the speed decreases, so that the net amount of 
 energy required for acceleration purposes is not considerable. Where reversing 
 
 1 11 11 1 1 II 1 HI 11 II 1 1 H 1 1 1 1 
 
 ~1 
 
 t 
 
 j 
 
 
 / 
 
 7 
 
 _r 
 
 1 
 
 r 
 
 I z 
 
 _J 
 
 ~l_ 
 
 « I J. 
 
 J f 
 
 * 7 
 
 1 _f 
 
 
 / 
 
 / 
 
 
 ^ 
 
 r ~ /- <— 
 
 X 
 
 y 
 
 jf~ , 
 
 <^ 
 
 ^ n 
 
 x 5 ^ 
 
 /7— -=— "'^ 
 
 is per Minute. 
 - Work done in Accelerating the Rotating Masses in Reversing Mil) 
 
 (Puppe.) 
 
 steam engines are employed, however, the work done in accelerating the rotating 
 masses is lost, and this should be borne in mind when deciding whether for a 
 given mill a reversible engine or a flywheel engine is best. 
 
 Rails after becoming so worn in the track as to be unfit for further service may 
 be rerolled into new sections. A considerable tonnage of such rails has been re- 
 rolled by the McKenna process at a cost of about $7.00 per ton. The advantages 
 claimed for this process are: 
 
 First. That the worn rail is selected material, as the imperfect rails have been 
 eliminated to a large extent during the time the rails have been in service. 
 
460 
 
 STEEL RAILS 
 
 Second. The rerolling puts additional mechanical work upon the material, 
 which should improve its quality. 
 
 The practical difficulties in connection with the process may be summarized 
 as follows: 
 
 First. The rail must not be worn beyond a certain degree, as there must re- 
 main in the head of the rail a sufficient amount of metal to form the new head. 
 The metal from the base and flange cannot be made to flow into the head to make 
 up for its worn condition. In general it is not desirable to remove rails from the 
 main track, which show only the amount of wear best adapted to rerolling; how- 
 ever, when it is necessary to remove rails from some cause other than reduction of 
 section on account of wear it may prove economical to reroll them into sections of 
 lesser weight. 
 
 Second. It is desirable that the rerolled rails may conform to standards already 
 in use. This seems to be in a practical way difficult to accomplish, owing to the 
 varying shape of the head of the rerolled rail, as the rail is much or little worn. 
 
 The following figures are taken from a record of a mile of 80-pound rail 
 rerolled by the American McKenna Process Company. 
 
 10,560 ft. of rail taken out of track Tons. 
 
 Weight when new 125 .71 
 
 Loss in track after 10 years' use 5 . 33 
 
 Weight sent to mill 120 .38 
 
 10,151 ft. of rail received from mill 103 .20 
 
 Scrap received from mill 17 . 18 
 
 The sections given in Fig. 339 and Plates VII and VIII of the rails used 
 
 Pennsylvania, 100-lb. 
 
 Pennsylvania, 85-lb. 
 
 Fig. 339. — Recent Rail Sections. (Railroad Age Gazette.) 
 
INFLUENCE OF DETAIL OF MANUFACTURE 461 
 
 I* 2 J2 *| 
 
 Canadian Pacific, 85-lb. 
 
 Baltimore & Ohio, 90-Ib. 
 
 K- -2k M 
 
 Santa Fe, 85-lb. 
 
 Burlington, 90-Ib. 
 
 Section. 
 
 
 Area of 
 
 
 Head, 
 
 Web, 
 
 Base. 
 
 Total, 
 
 
 pCTcen . 
 42.2 
 
 36.8 
 37.0 
 36.2 
 
 per cent. 
 
 18.6 
 
 17.8 
 22.2 
 22.8 
 24.0 
 
 P 40 Ce 4 ' 
 40.0 
 
 ' 41.0 
 40.2 
 39.8 
 
 ioo 6 
 
 
 100 
 
 
 100 
 
 Santa Fe, 85-lb 
 
 100 
 
 Burlington, 90-lb 
 
 100.0 
 
 
 
 Fig. 339 (continued). — Recent Rail Sections. (Railroad Age Gazette.) 
 
462 STEEL RAILS 
 
 at the present time show clearly the willingness on the part of the railroads to 
 meet the criticism of the manufacturers in regard to the faults in the design of 
 the heavier A. S. C. E. sections. 
 
 These new thick-base sections, adopted after the studies of 1907, cool with 
 less curvature than the former thin-base types and require less cold-straighten- 
 ing. There appears to have been a material reduction in the number of base 
 failures in the new sections as compared with the A. S. C. E. design. 
 
 A number of students of this subject think that there is room for still 
 further improvements along this line to eliminate to an even greater extent the 
 failures in the base of the rail. This is evidenced by the new Chicago and 
 North-Western section for 100-pound rail in which the base is ff inch thick 
 at the outer edge. Dr. P. H. Dudley has designed a new section giving to the 
 fillet between the base and the web a radius of 1 inch. The Illinois Steel 
 Company at its South Works plant is also experimenting with a new section 
 of 110 pounds in weight which is similar to that used on the foreign railways 
 in that the upper surface of the base is broken at two angles. 
 
CHAPTER VII 
 RAIL SPECIFICATIONS 
 
 34. Comparison of American Specifications 
 
 The rail committee of the American Railway Engineering and Maintenance 
 of Way Association revised their specifications in the latter part of the year 
 1909; these were subsequently withdrawn and in March, 1912, the committee 
 presented to the American Railway Engineering Association, which was now 
 the name of the association, specifications for carbon steel rails. 
 
 Some paragraphs, such as those relating to carbon under remarks in Table 
 XCIX and Nos. 14 and 15, relating to physical requirements, were not considered as 
 final, it being thought that the committee did not have sufficient information in 
 its possession to make these sections in the specifications mandatory. The re- 
 quirements in section 14 for ductility were somewhat lower than some of the 
 members thought desirable. Paragraph No. 15, referring to deflections as a 
 method of classifying rails, is also tentative, and it is the intention, when suffi- 
 cient data is at hand, to prescribe maximum and minimum limits for deflections 
 under the drop test. The committee will continue its investigations and the 
 specifications in these respects will be subject to change. 
 
 On January 10, 1912, the Pennsylvania revised its specifications for 
 Bessemer and open-hearth rails. 
 
 On January 1, 1909, the Association of American Steel Manufacturers 
 issued standard specifications for Bessemer and open-hearth rails. These rail 
 specifications, adopted by the steel manufacturers of the United States and 
 Canada, which are practically the same for the different companies, indicate 
 the views of the rail makers as to the proper tests and chemical composition for 
 securing good rails. The A. S. C. E. sections are still officially regarded as 
 standard practice. 
 
 Standard specifications for Bessemer and open-hearth steel rails were 
 adopted by letter ballot on August 16 1909, by the American Society for Testing 
 Materials. 
 
 In October, 1909, the Harriman Lines adopted standard specifications for 
 Bessemer and open-hearth steel rails; the open-hearth specifications were subse- 
 quently revised in February, 1910. 
 
464 STEEL RAILS 
 
 The above specifications show the development during recent years of rail 
 specifications in this country and an examination of their requirements will prove 
 of interest. The American Railway Engineering Association specifications of 1912 
 reflect the latest thought and are noticeable for the increase in the number of 
 physical tests over those required in earlier specifications. A great many defects, 
 such as piping of the ingot, can be adequately guarded against by proper physical 
 tests, and in general it would appear desirable to leave the producer free in 
 such cases to adopt his own methods of manufacture. Within certain limits, 
 however, the specifications may well be drawn to exclude practice which is known 
 to result in defective material. The desirability of doing this is emphasized by the 
 great difference in quality found in rollings from different mills and in some cases 
 for rails from the same mill, but rolled in different years. The specifications given 
 in Article 35 are a good example of specifications drawn with a view to eliminating 
 defective practice at the mill. 
 
 The trend of recent specifications is to increase the amount of inspection which 
 is being given the rail at the mills. The plan of the R. W. Hunt and Company of 
 placing inspectors throughout the mill to watch the entire process of manufacture 
 is evidence of this. 
 
 Inspection 
 Am. Ry. Eng. Assn. : 
 
 1. Inspectors representing the purchaser have free entry to the works of the 
 manufacturer at all times while the contract is being executed, and shall have all 
 reasonable facilities afforded them by the manufacturer to satisfy them that the 
 rails have been made in accordance with the terms of the specifications. 
 
 2. All tests and inspections shall be made at the place of manufacture, prior to 
 shipment, and shall be so conducted as not to interfere unnecessarily with the oper- 
 ation of the mill. 
 
 All of the specifications are substantially the same as the above. 
 
 Material 
 
 Am. Ry. Eng. Assn. : 
 
 3. The material shall be steel made by the Bessemer or open-hearth process 
 provided by the contract. 
 
 The clause in reference to material in the Pennsylvania Specification is the 
 same, but in the other specifications it is omitted and a separate specification 
 written for each class of material, i.e., Bessemer or open-hearth steel. 
 
RAIL SPECIFICATIONS 4G5 
 
 Chemical Requirements 
 
 Am. Ry. Eng. Assn.: 
 
 4. The chemical composition of the steel from which the rails are rolled, de- 
 termined as prescribed in Section 7, shall be within the following limits: (See Table 
 XCIX.) 
 
 Table XCIX presents a comparison of the chemical requirements of the 
 different specifications. 
 
 The Committee on Standard Rail and Wheel Sections of the American 
 Railway Association, in its report of March 23, 1908, to the association, stated : 
 " In the matter of chemistry specifications for Bessemer steel rail, statistics 
 were obtained from the officers of the Ore Producers' Association which con- 
 vinced the committee that it would be impossible for the mills to furnish more 
 than a small percentage of the total rail requirements of the railways with a 
 phosphorus specification less than .10. 
 
 " The optional specification for .085 phosphorus prepared by the joint 
 committee of manufacturers and railway men is now in the hands of all members, 
 and is, therefore, available for use by those who are able to obtain low-phos- 
 phorus Bessemer rails. It is not considered proper, however, to require less than 
 .10 phosphorus in a specification intended for general use. Members desiring 
 to obtain' a low-phosphorus rail will have the further option of using open- 
 hearth steel. 
 
 The committee conferred with a number of disinterested experts on the 
 phosphorus question, and among the principal authorities consulted were 
 William Metcalf, of Pittsburg, Robert Forsyth, of Chicago, and Henry M. 
 Howe, of Columbia University. These gentlemen all agreed that it would be 
 unreasonable to require less than .10 phosphorus in a specification for Bessemer 
 rails intended to cover purchases for all American railways." 
 
 The Pennsylvania specification for open-hearth rails makes the upper limits 
 for classification A, phosphorus .03 and carbon .83; for classification B, .04 
 phosphorus and .75 carbon. The desired carbon for the two grades is .75 for 
 the lower phosphorus, and .70 for the higher. The reason for making two 
 classifications for open-hearth rails relates principally to the cost of manufacture. 
 It was thought desirable to specify phosphorus as low as .03 so that high carbon 
 could be used and the wearing quality of the rails, particularly on curves, would 
 be materially improved. But the extra time required in the open-hearth furnace 
 to reduce phosphorus from .04 to .03 results in some increase in the cost of 
 manufacture, and a slight addition to the normal price per ton is added for the 
 class A rails. 
 
STEEL RAILS 
 
 
 
 1 
 
 1 
 
 
 ion of Carbon for Low Phos- 
 — When the material used 
 mill is such that the average 
 orus content of the ingot metal 
 the Bessemer process is run- 
 elow 0.08 and in the Open- 
 process is running below 0.03, 
 t seems mutually desirable, 
 bon may be increased at the 
 0.035 for each 0.01 that the 
 orus content of the ingot metal 
 rerages below 0.08 for Besse- 
 eel, or 0.03 for Open-hearth 
 
 i! a Si 
 
 j-stjl 
 
 Ills! 
 
 lllll 
 
 |,|.a 
 
 ill 
 
 IK 
 
 
 <J 
 
 1 
 
 nills cannot furnish steel rail 
 phosphorus content less than 
 cent carbon shall be as follows: 
 
 65 lb. .38-. 48 % 
 
 75 1b. .42-. 52% 
 
 901b. .50-.60% 
 
 
 
 
 
 lllli^FlIli »? Ill: 1 Mi!!! 
 
 l^la 
 
 
 
 
 
 
 
 
 ^ Res as n-fl Sta k O.S £ 
 
 o^ocsa «.s»aa5 
 
 o"o h 
 
 g-So 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •XB W 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 a 
 
 •«!W 
 
 
 
 
 — 
 
 
 
 A" 
 
 
 S 
 
 S 
 
 
 ss 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ a 
 
 ui M 
 
 
 
 
 ->_ 
 
 
 
 
 
 
 s 
 
 
 
 
 
 
 
 H 
 
 •XB M 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 •u !M 
 
 
 8 
 
 
 ■xb k 
 
 S 
 
 g 
 
 
 s 
 
 S 
 
 
 
 
 !§ 
 
 
 
 
 
 
 
 
 
 s 
 
 s 
 
 
 
 
 
 
 
 s c 
 
 •m M 
 
 -°_ 
 
 ° 
 
 
 _-, 
 
 ,~°_ 
 
 
 
 
 
 
 
 
 sss 
 
 
 
 
 
 •XB M 
 
 
 
 
 
 
 
 
 g 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •p.saa 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •ai W 
 
 do 
 
 
 
 do' 
 
 ood 
 
 
 
 
 
 — — 
 
 , 
 
 
 — ~ 
 
 ■~~~ 
 
 
 § 
 
 
 g 
 
 •xb W 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 cc 
 
 •ni M 
 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 
 
 
 
 
 
 
 
 
 
 A 1 
 
 •xb W 
 
 2 
 
 s 
 
 
 s 
 
 1 
 
 
 
 
 f^a. 
 
 
 
 
 
 
 
 
 ■m M 
 
 
 
 
 
 « 
 
 
 
 
 
 
 
 1 
 
 | 
 
 •xb M 
 
 d 
 
 
 d 
 
 d 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 •ai W 
 
 
 
 
 
 
 
 
 B 
 
 
 
 
 
 
 
 
 
 
 •xb W 
 
 s 
 
 o sss 
 
 
 s 
 
 
 
 
 
 
 H 
 
 
 
 
 
 
 
 
 
 
 ° sss 
 
 
 s 
 
 
 
 
 
 s c 
 
 •"!W 
 
 ~°_ 
 
 o ooo 
 
 
 
 
 -°_ 
 
 
 
 ■XBW 
 
 ss 
 
 s sax 
 
 
 s 
 
 gss 
 
 
 
 
 § 
 
 
 
 
 
 
 
 
 ■p.saa 
 
 
 
 
 g 
 
 
 
 
 
 d 
 
 
 
 
 
 
 
 
 
 °« 
 
 m §S§ 
 
 
 § 
 
 sss 
 
 
 
 
 
 ui W 
 
 ^2-, 
 
 ^o_ooo 
 
 
 
 
 000 
 
 
 
 £ 
 
 o 
 
 
 
 00008 
 
 
 s 
 
 ss§ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 
 
 S § 
 
 
 S2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 | 
 
 I b 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 fa* 
 
 u 111 
 
 1 
 
 .9 
 
 a 
 
 
 
 
 
 
 
 
 
 <i 
 
 < 
 
 s 
 
 g 
 
 
 w 
 
 
RAIL SPECIFICATIONS 467 
 
 Sulphur is not generally mentioned in the chemical requirements, but the 
 trend of modern specifications is to require that this element be reported. This 
 is illustrated by the American Railway Engineering Association specification 
 for analyses given below. 
 
 Am. Ry. Eng. Assn.: 
 
 7. In order to ascertain whether the chemical composition is in accord- 
 ance with the requirements, analyses shall be furnished as follows : 
 
 (a) For Bessemer process the manufacturer shall furnish to the inspector, 
 daily, carbon determinations for each heat before the rails are shipped, and two 
 chemical analyses every twenty-four hours representing the average of the 
 elements, carbon, manganese, silicon, phosphorus, and sulphur contained in the 
 steel, one for each day and night turn respectively. These analyses shall be 
 made on drillings taken from the ladle test ingot not less than one-eighth inch 
 beneath the surface. 
 
 (&) For Open-hearth process, the makers shall furnish the inspectors with 
 a chemical analysis of the elements, carbon, manganese, silicon, phosphorus, and 
 sulphur, for each heat. 
 
 (c) On request of the inspector, the manufacturer shall furnish drillings 
 from the test ingot for check analyses. 
 
 The Pennsylvania specifications are the same as the above. The other speci- 
 fications only require one complete chemical analysis every twenty-four hours in 
 the Bessemer process. The manufacturers and American Society for Testing 
 Materials require that the analyses shall be made on drillings taken not less than 
 one-fourth inch beneath the surface of the test ingot. The Harriman Lines speci- 
 fication does not give the depth at which the drillings should be taken. 
 
 Physical Requirements 
 Table C compares the physical requirements of the different specifications. 
 
 Am. Ry. Eng. Assn. : 
 
 Physical Qualities. 
 
 8. Tests shall be made to determine: 
 
 (a) Ductility or toughness as opposed to brittleness. 
 
 (b) Soundness. 
 
 Method of Testing. 
 
 9. The physical qualities shall be determined by the Drop Test. 
 
 Drop Testing Machine. 
 
 10. The drop-testing machine used shall be the standard of the American Railway Engineering 
 Association. 
 
 (a) The tup shall weigh 2000 pounds, and have a striking face with a radius of five inches. 
 (6) The anvil block shall weigh 20,000 pounds, and be supported on springs. 
 
 (c) The supports for the test pieces shall be spaced 3 feet between centers and shall be a part of, 
 and firmly secured to the anvil. The bearing surfaces of the supports shall have a radius of 5 inches. 
 
468 
 
 STEEL RAILS 
 
 Pieces for Drop Test. 
 
 11. Drop tests shall be made on pieces of rail not less than 4 feet and not more than 6 feet long. 
 These test pieces shall be cut from the top end of the top rail of the ingot, and marked on the base or 
 head with gage marks 1 inch apart for 3 inches each side of the center of the test piece, for measuring the 
 ductility of the metal. 
 
 Temperature of Test Pieces. 
 
 12. The temperature of the test pieces shall be between 60 and 100 degrees Fahrenheit. 
 
 Height of Drop. 
 
 13. The test piece shall, at the option of the inspector, be placed head or base upwards on the 
 supports, and be subjected to impact of the tup falling free from the following heights: 
 
 For 70-pound rail ■ 16 feet 
 
 For 80-, 85-, and 90-pound rail 17 feet 
 
 For 100-pound rail 18 feet 
 
 Elongation or Ductility. 
 
 14. Under these impacts the rail under one or more blows shall show at least 6 per cent elongation 
 for 1 inch or 5 per cent each for two consecutive inches of the 6-inch scale, marked as described in 
 Section 11. 
 
 Permanent Set. 
 
 15. It is desired that the permanent set after one blow under the drop test shall not exceed that 
 in the following table, and a record shall be made of this information. 
 
 
 
 
 
 Rail. 
 
 
 
 Permanent Set, m 
 Ordinate in Inches 
 
 asured by Middle 
 in a Length of 3 ft. 
 
 Section. 
 
 Weight per 
 Yard. 
 
 Moment of Inertia. 
 
 Bessemer Process. 
 
 O. H. Process. 
 
 A.R.A.-A 
 
 
 100 
 
 100 
 90 
 90 
 80 
 80 
 70 
 70 
 
 48.94 
 41.30 
 38.70 
 32.30 
 28.80 
 25.00 
 21.05 
 18.60 
 
 1.65 
 2.05 
 1.90 
 2.20 
 2.85 
 3.15 
 3.50 
 3.85 
 
 1.45 
 
 A.R.A.-B 
 A.R.A.-A 
 
 
 
 
 
 1.80 
 1.65 
 
 A.R.A.-B 
 
 
 
 
 
 2.00 
 
 A.R.A.-A 
 A.R.A.-B 
 
 
 
 
 
 2.45 
 2.85 
 
 A.R.A.-A 
 A.R.A.-B 
 
 
 
 
 
 3.10 
 3.50 
 
 Test to Destruction. 
 
 16. The test pieces which do not break under the first or subsequent blows shall be nicked and 
 broken to determine whether the interior metal is sound. 
 
 Bessemer Process Drop Tests. 
 
 17. One piece shall be tested from each heat of Bessemer steel. 
 
 (a) If the test piece does not break at the first blow and shows the required elongation (Section 14), 
 all of the rails of the heat shall be accepted, provided that the test piece when nicked and broken does 
 not show interior defect. 
 
 (6) If the piece breaks at the first blow, or does not show the required elongation (Section 14), 
 or if the test piece shows the required elongation, but when nicked and broken shows interior defect, 
 all of the top rails from that heat shall be rejected. 
 
 (c) A second test shall then be made of a test piece selected by the inspector from the top end of 
 any second rail of the same heat, preferably of the same ingot. If the -test piece does not break at the 
 first blow, and shows the required elongation (Section 14), all of the remainder of the rails of the heat 
 shall be accepted, provided that the test piece when nicked and broken does not show interior defect. 
 
 (d) If the piece breaks at the first blow, or does not show the required elongation (Section 14), or 
 if the piece shows the required elongation, but when nicked and broken shows interior defect, all of the 
 second rails from that heat shall be rejected. 
 
 (e) A third test shall then be made of a test piece selected by the inspector from the top end of 
 any third rail of the same heat, preferably of the same ingot. If the test piece does not break at the 
 first blow and shows the required elongation (Section 14), all of the remainder of the rails of the heat 
 shall be accepted, provided that the test piece when nicked and broken does not show interior defect. 
 
 (/) If the piece breaks at the first blow, or does not show the required elongation (Section 14), or 
 if the piece shows the required elongation, but when nicked and broken shows interior defect, all of the 
 remainder of the rails from that heat shall be rejected. 
 
RAIL SPECIFICATIONS 469 
 
 Open-hearth Process Drop Tests. 
 
 18. Test pieces shall be selected from the second, middle, and last full ingot of each open-hearth 
 heat. 
 
 (a) If two of these test pieces do not break at the first blow and show the required elongation 
 (Section 14), all of the rails of the heat shall be accepted, provided that these test pieces when nicked 
 and broken do not show interior defect. 
 
 (6) If two of the test pieces break at the first blow, or do not show the required elongation, or if 
 any of the pieces that have been tested under the drop when nicked and broken show interior defect, 
 all of the top rails from that heat shall be rejected. 
 
 (c) Second tests shall then be made from three test pieces selected by the inspector from the top 
 end of any second rails of the same heat, preferably of the same ingots. If two of these test pieces do 
 not break at the first blow and show the required elongation (Section 14), all of the remainder of the 
 rails of the heat shall be accepted, provided that the pieces that have been tested under the drop when 
 nicked and broken do not show interior defect. 
 
 (d) If two of these test pieces break at the first blow or do not show the required elongation 
 (Section 14), or if any of the pieces that have been tested under the drop when nicked and broken show 
 interior defect, all of the second rails of the heat shall be rejected. 
 
 (e) Third tests shall then be made from three test pieces selected by the inspector from the top 
 end of any third rails of the same heat, preferably of the same ingots. If two of these test pieces do not 
 break at the first blow, and show the required elongation (Section 14), all of the remainder of the rails 
 of the heat shall be accepted, provided that the pieces that have been tested under the drop when 
 nicked and broken do not show interior defect. 
 
 (/) If two of these test pieces break at the first blow or do not show the required elongation 
 (Section 14), or if any of the pieces, that have been tested under the drop when nicked and broken 
 show interior defect, all of the remainder of the rails from that heat shall be rejected. 
 
 The drop-testing machine has been standardized, and it is claimed that the 
 lower drop called for under the new conditions is equivalent to the higher drop 
 of 22 feet previously specified. 
 
 The provision in the American Railway Engineering Association and Penn- 
 sylvania specifications that drop testing shall be continued to the destruction of 
 the specimen is a precaution which should result in a material benefit to the railway 
 by reducing the number of piped rails, which, in spite of the usual inspection and 
 tests, get into the main track. It seems far more desirable not to specify any defi- 
 nite discard, but to test to destruction a number of rail butts representing a certain 
 proportion of the total output, and to base rejections on the results of these 
 tests. 
 
 The Cambria Steel Company rolled a considerable tonnage of rails under 
 these specifications, and the testing to destruction unquestionably detected the 
 pipes. To find to what depth the pipes extended, they polished the ends of the drop- 
 test pieces and cut the top rail adjoining the test pieces into small lengths, examin- 
 ing carefully each cut for pipes. It was found that of the heats showing pipes in 
 the drop-test piece when tested to destruction, sixty per cent contained pipes so 
 short that they were confined entirely to the crop end. Of the remaining forty 
 per cent which extended into the top rail, nearly half showed a pipe extending less 
 than four feet. 
 
 Most of the specifications require that two test pieces be tested in the open- 
 hearth process but only one in the Bessemer process. This is because of the greater 
 tonnage of metal in an open-hearth heat as compared to a Bessemer heat, there 
 being about five times as much metal. All the basic open-hearth rail steel now 
 
STEEL RAILS 
 
 & s 
 
 *!*! 
 
 
 
 i 
 
 i 
 
 
 III 
 
 
 
 
 
 I s « 
 
 
 
 1 
 
 § 
 
 
 2SS 
 
 
 
 
 
 
 
 
 s 
 
 iS 
 
 
 111 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s|S 
 
 
 
 3 
 1 
 
 I 
 
 
 §.aa 
 ssg 
 
 
 
 
 
 E S g-2 
 
 
 
 
 dS 
 
 
 £££ 
 
 
 
 
 
 ■fifl 
 
 
 
 IllS'P'ajIf't 
 
 |-s|ffl||^:itd 
 
 111 
 
 1 
 
 jU =t~COraO 
 
 
 
 
 
 ■||sj? 
 
 
 
 
 IsJJ 
 
 1 
 
 Is-sII 
 
 
 
 
 
 llll 
 
 
 
 flllllillllll 
 
 Q^J§ 
 
 !§ 
 
 illllf 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 If 
 
 
 
 
 
 
 
 8 
 
 
 !! 
 
 
 SJ 
 
 
 
 
 
 
 
 £ 
 
 
 22 
 
 
 1" _ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Sw.1 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 III 
 
 
 
 s 
 
 
 
 
 i 
 
 
 
 
 
 
 l«"a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 m 
 
 
 
 
 ^ 
 
 
 
 
 
 
 '"S*M 
 
 
 
 1 
 
 
 
 
 c 
 
 
 
 
 
 
 <C.I o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 fa 
 
 
 
 
 
 
 JSS 
 
 
 
 1 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 
 
 
 
 
 i 
 
 1 
 
 
 <§ 
 
 < 
 
 Ik 
 
 
 
 
 
 
 
 
 1 
 
 
 s 
 
 a 
 
 
 1 
 
 11 
 
 
 ||| 
 
 
 B 
 
 
 
 <! 
 
 < 
 
 g 
 
 SS 
 
 
 
 
 < 
 
 
 
 § 
 
 3 
 
 Is 
 
 i 
 
 3 
 
 
 
 "°C 3 
 
 
 S 
 
 
 
 I 
 
 1 
 
 II 
 a- 
 
 && 
 
 
 
 
 I 
 <8 
 
 1 
 J 
 
 1 
 
 1 
 
 1 1 1^ 
 
 f f b| 
 
 1 
 
 1 
 
 1 
 1 
 1 
 
 11 
 
 Its 
 t«J 
 
 If 
 
 & o 
 
 si 
 
 Si 
 
 Is 
 ll 
 
 flL 
 S-SJJ 
 
 G.-2 
 
 1 
 6 
 
 2SSSSS2S 
 
 if 
 
 ll 
 
 SB 
 
 ll 
 
 |° 
 
 .it 
 
 
 
 
 H3SBS 
 
 
 °d 
 
 
 ■J 
 
 1 
 
 I 
 
 2 
 
 .11 
 
 BR 
 
 If 
 H|l 
 
 1 
 1 
 
 1 
 
 1 
 ll 
 
 1 J !!|S 
 
 sift 
 
 1 
 
 li 
 
 31 
 
 
 
 fig 
 
 iS 
 
 •. It: s 
 
 11 
 
 B 
 
 1 
 
 1 £ 
 
 1" 
 
 3:?£SSSSS2 
 
 si 
 
 ill 
 
 III 
 
 l s ll 
 
 |j 
 
 
 gggggggs 
 
 S.- a £ !„c = 
 
 =>Jg|<±6§ 
 
 1 
 
 
RAIL SPECIFICATIONS 
 
 471 
 
 I 
 
 1 
 
 Three test pieces shall 
 be selected from each 
 irelt at approximately 
 equal intervals from the 
 melt as poured. Two of 
 these test pieces shall be 
 tested, and if both meet 
 the requirements, all the 
 rails from the melt which 
 they represent shall be 
 accepted, provided they 
 conform to the other re- 
 
 ifications. Should both 
 of these test pieces fail, 
 all rails from the melt 
 will be rejected. Should 
 
 pieces fail, the'liini'sh-.ll 
 be tested, and if this third 
 test meets the require- 
 ments, all the rails from 
 1 he melt shall be accepted 
 
 the other requiieii enls 
 of these specifications. 
 Should the third lest fail, 
 all the rails from the melt 
 shall be rejected. 
 
 
 One piece shall be tested 
 from each heat of Besse- 
 mer steel. 
 
 If any rail breaks when 
 subjected to the drop test, 
 twoadditional tests will be 
 made of other rails from 
 the same blow of steel , and 
 if either of these latter 
 tests fails, all the rails of 
 the blow which they rep- 
 resent will be rejected; but 
 if both of these additional 
 test pieees meet the re- 
 quirements, all the rails of 
 the blow which they rep- 
 resent will be accepted. 
 
 & O «3 
 
 H al1 
 
 One piece shall be tested from each heat of Bessemer steel. 
 
 (a) If the test piece does not break, all of the rails of the heat 
 shall be accepted as No. 1 or No. 2 classification, according as 
 the deflection is less or more, respectively, than the prescribed 
 limit, provided that the test piece when nicked and broken does 
 not show interior defect. 
 
 (The words " interior defect " shall be interpreted to mean: 
 seams, laminations, cavities or interposed foreign matter, made 
 visible by the destruction test, the saws, or the drills.) 
 
 (6) If the test piece breaks, or if the test piece does not break 
 but when nicked and broken shows interior defect, all of the 
 top rails from that heat shall be rejected. 
 
 (c) A second test shall then be made of a test piece selected by 
 the inspector from the top end of any second rail of the same heat, 
 preferably of the same ingot. If the test piece does not break, 
 all of the remainder of the rails of the heat shall be accepted as 
 No. 1 or No. 2 classification, according as the deflection is lessoi 
 more, respectively, than the prescribed limit, provided thai I he 
 test piece when nicked and broken does not show any interior 
 
 (d) If the test piece breaks, or if the test piece does not break 
 but when nicked and bioken shows interior defect, all of the 
 second rails from that heat shall be rejected. 
 
 (e) A third test shall then be made of a test piece selected bv 
 the inspector from the top end of any third rail of thesan e heal, 
 preferably of ii. .-..,, ,,;>'"■ If the test piece does not break, all 
 of the remainder of the rails of the heat shall be accepted as No. 1 
 or No. 2 classification, according as the deflection is less or more, 
 respectively, than the prescribed limit, provided that the test 
 
 i i ' 11 n i loes not show interior defect. 
 (/) If the test, piece breaks, or if the test piece does not break 
 but when nicked and broken shows interior defect, all of the 
 remainder of the rails from that heat shall be rejected. 
 
 Test pieces shall be selected from the second, middle and last 
 full ingot of each open-hearth heat. 
 
 (a) If two of these test pieces do not break, all of the rails of 
 the heat shall be accepted as No. 1 or No. 2 classification, accoid- 
 iti» as I he deflectioi is less or more, respectively, than the pre- 
 scrihed limit, provided that these test pieces when nicked and 
 broken do not show interior defect. 
 
 (b) If two of the test pieces break, or if anv of the test pieces 
 that have bee:: • ■-■--'. ,ui!n i he drop when nicked and broken 
 si, on interior defect, all of the top rails from that heat shall be 
 rejected. 
 
 (c) A second test shall then be made from three test pieees 
 selected by the inspector from the top ends of any second rails of 
 the same heat, preferably the same ingots. If two of these test 
 pieces do not break, all of the remainder of the rails of the heat 
 shall be accepted as No. 1 or No. 2 classification, according as 
 (he deflection is lessor more, respectively, than the prescribed 
 limit , provided that the test pieces that have been tested under 
 the drop when nicked and broken do not show interior defect. 
 
 (d) If two of these test nieces break, or if any of the pieces that 
 have been tested under the drop when nicked and broken show 
 •'nterior defei 1 i id rulsof the heat shall be rejected. 
 
 0) A third test shall then he made from three test pieces 
 selected bv the inspector from the top ends of anv third rails of 
 the s.,i„e heat, preferably the same ingots. If two of these test 
 pieces do not break, all of the remainder of the rails of (he heal 
 shall be accepted as No. 1 or No. 2 classification, according as 
 the deflection is less or more, respectively, than the prescribed 
 limit, provided that the test pieces when nicked and broken do 
 
 ||j 
 
 fj"3 
 
 lis 
 
 §ii 
 
 111 
 
 S?'Jj=; 
 
 lirssfnirr Process Drop Tests: 
 
 17. One piece shall be tested from each heat of Bessemer steel. 
 
 (a) If the test piece does not break at the first blow and shows the required elon- 
 gation (Section 14), all of the rails of the heat shall be accepted, provided that the 
 test piece when nicked and broken does not show interior defect. 
 
 (6) If the piece breaks at the first blow, or does not show the required elongation 
 (Section 14), or if the test piece shows the required elongation, but when nicked and 
 broken shows interior defect, all of the top rails from that heat shall be rejected. 
 
 (c) A second test shall then be made of a test piece selected by the inspector 
 from the top end of any second rail of the same heat, preferably of the same ingot. 
 If the test piece does not break at the first blow, and shows the required elongation 
 (Section 14), all of the remainder of the rails of the heat shall be accepted, pro- 
 vided that the test piece when nicked and broken does not show interior defect. 
 
 (d) If the piece breaks at the first blow, or does not show the required elonga- 
 tion (Section 14), or if the piece shows the required elongation but when nicked and 
 broken shows interior defect, all of the second rails from that heat shall be rejected. 
 
 (e) A third test shall then be made of a test piece selected by the inspector from 
 the top end of any third rail of the same heat, preferably of the same ingot. If the 
 test piece does not break at the first blow and shows the required elongai ion (Sec- 
 tion 14), all of the remainder of the rails of the heat shall be accepted, provided 
 that the test piece when nicked and broken does not show interior defect. 
 
 (/) If the piece breaks at the first blow, or does not show the required elonga- 
 tion (Section 14), or if the piece shows the required elongation but when nicked and 
 broken shows interior defect, all of the remainder of the rails from that heat shall 
 be rejected. 
 
 18. Test pieces shall be selected from the second, middle, and last full ingot of 
 each open-hearth heat. 
 
 (a) If two of these test pieces do not break at the first blow and show the re- 
 quire 1 elongation (Section 14), all of the rails of the heat shall be accepted, pro- 
 vided that these test pieces when nicked and broken do not show interior defect. 
 
 (6) If two of the test pieces break at the first blow, or do not show the required 
 elongation, or if am oi t he pieces t hat have been tested under the drop when nicked 
 and broken show iniei ior deleci , all ol 1 he top rails from 1 hat heat shall be rejected. 
 
 (c) Second tests shall then be made from three test pieces selected by the in- 
 spector from the top end of any second rails of the same heat, preferably of the 
 same ingots. If two of these test pieces do not break at the first blow and show 
 the required elongation Section 11. all of i he remainder of the rails of the heat 
 shall be accepted, pmvidod I hat. t he pieces t hat have been tested under the drop 
 when nicked and broken do nol show interior defect. 
 
 (d) If two of these test pieces break a I the first blow or do not show the required 
 elongation (Section 1 1), or it anv of i he pieces thai have been tested under the drop 
 when nicked and broken show interior defect, all of the second rails of the heat 
 shall be rejected. 
 
 (e) Third tests shall then be made from three test pieces selected bv the in- 
 spector from the top end of an v third rails ol the same heal, preferably of the same 
 ingots. If two of these test pieces do not break at i he first blow, and show the re- 
 quired elongation (Section 14), all of the remainder of the rails of the heal shall lie 
 uccepted, provided that the pieces that have been tested under the drop when 
 nicked and broken do not show interior defect. 
 
 (/) If two of these test nieces break ai the first blow or do not show the required 
 elongation i sect ion 1 1 1, or il any ol t fie pieces ;lc, have been tested under the drop 
 when nicked and broken show interior defect, all of the remainder of the rails from 
 that heat shall be rejected. 
 
 
472 STEEL RAILS 
 
 being made in the United States is melted in furnaces of a minimum capacity of 
 40 tons, and the majority of it is made in 50-ton or 80-ton furnaces. 
 
 The feeling is growing among railway engineers that present specifications do 
 not go far enough in specifying tests of a certain number of ingots from each heat, 
 but that specimens should be tested from each individual ingot. On account of the 
 uncertainty attending the formation of the pipe in the ingot an increase in the 
 number of tests per heat would appear desirable. 
 
 The Pennsylvania is the only specification that provides for a limiting deflec- 
 tion, although the American Railway Engineering Association give limits that it is 
 desired not to exceed and state that it is proposed to prescribe the requirements in 
 regard to deflection as soon as proper limits have been decided on. In the absence 
 of a tension test it would seem desirable to make provisions for fixing maximum 
 and minimum limits for the deflection. 
 
 The American Railway Engineering Association specifications are the only 
 specifications which call for a ductility test. This test has been used for some 
 time by Dr. P. H. Dudley, as shown in the New York Central Lines specifications 
 given in Article 35. 
 
 No. 1 and No. 2 Rails 
 Am. Ry. Eng. Assn. : 
 
 19. No. 1 classification rails shall be free from injurious defects and flaws 
 of all kinds. 
 
 20. (a) Rails which, by reason of surface imperfections, or for causes men- 
 tioned in Section 30 hereof, are not classed as No. 1 rails will be accepted as No. 2 
 rails, but No. 2 rails which contain imperfections in such number or of such char- 
 acter as will, in the judgment of the inspector, render them unfit for recognized 
 No. 2 uses will not be accepted for shipment. 
 
 (6) No. 2 rails to the extent of 5 per cent of the whole order will be received. 
 All rails accepted as No. 2 rails shall have the ends painted white and shall have 
 two prick punch marks on the side of the web near the heat number near the end 
 of the rail, so placed as not to be covered by the splice bars. 
 
 All of the specifications state that "No. 1 rails shall be free from injurious 
 defects and flaws of all kinds." 
 
 The Pennsylvania specifications for No. 2 rails are the same as the above with 
 the clause added that rails which exceed the prescribed limits of deflection in the 
 drop test may be accepted as No. 2 rails. The manufacturers' specifications are 
 somewhat different and are given below. 
 
 Rails which, by reason of surface imperfections, are not classed as No. 1 rails 
 
RAIL SPECIFICATIONS 473 
 
 shall be considered No. 2 rails, but No. 2 rails shall not be accepted for shipment 
 which have flaws in the head of more than \ inch, or in the flange of more than 
 | inch in depth, and these shall not, in the judgment of the inspector, be, in any in- 
 dividual rail, so numerous or of such a character as to render it unfit for recognized 
 No. 2 rail uses. Both ends of No. 2 rails shall be painted white. 
 
 The Harriman Lines specifications are the same as the manufacturers' with 
 the additional clauses that No. 2 rails will be accepted up to 5 per cent of the 
 whole order, and in common with the American Society for Testing Materials they 
 will not accept rails as No. 2 from heats which failed under the drop test. 
 
 Quality of Manufacture 
 Am. Ry. Eng. Assn. : 
 
 21. The entire process of manufacture shall be in accordance with the best 
 current state of the art. 
 
 This section is common to all specifications except the manufacturers'. 
 
 Bled Ingots 
 Am. Ry. Eng. Assn. : 
 
 22. Bled ingots shall not be used. 
 
 This is specified by all except the manufacturers', with an additional clause 
 providing that the ingots be kept in a vertical position until ready to be rolled, 
 or until the metal in the interior has had time to solidify. 
 
 Discard 
 Am. Ry. Eng. Assn.: 
 
 23. There shall be sheared from end of the bloom formed from the top of 
 the ingot, sufficient metal to secure sound rails. 
 
 The Pennsylvania requirements are the same as the American Railway 
 Engineering Association, while the Manufacturers' specifications do not refer 
 to this part of the process. 
 
 The American Society for Testing Materials and the Harriman Lines speci- 
 fications call for a definite discard as follows: 
 
 A.S.forT. M.: 
 
 There shall be sheared from the end of the blooms formed from the top of 
 the ingots not less than x per cent,* and if, from any cause, the steel does not 
 then appear to be solid, the shearing shall continue until it does. 
 
 * The percentage of minimum discard in any case to be subject to agreement, and it should be 
 recognized that the higher this percentage the greater will be the cost. 
 
474 
 
 STEEL RAILS 
 
 Harriman Lines: 
 
 1. (d) There shall be sheared from the end of the bloom and rail formed 
 from the top of the ingot a total discard of not less than nine (9) per cent of the 
 weight of the ingot, and if, from any cause, the steel does not then appear to 
 be solid, the shearing shall continue until it does. If, by the use of any improve- 
 ments in the process of making ingots, the defects known as " piping " shall be 
 prevented, the above shearing requirements may be modified. 
 
 On this system, by excluding the top, or A, rails from main-line use, the 
 effect of an additional 30 per cent discard is obtained, and at the same time 
 the discard is saved for use on sidings and other locations where second-hand 
 rails usually are used; the road thus doing its own discarding beyond the 
 manufacturers' allowance, and saving the product without risk to the quality 
 of the main-track rails. 
 
 In view of the uncertainties, as to the length of the pipe it appears that 
 the position taken by the American Railway Engineering Association and the 
 Pennsylvania System in their specifications is the most reasonable one, viz., to 
 leave the discard to the manufacturers, and to safeguard the product by proper 
 tests, especially by choosing the test piece from such a location, and making the 
 rejections such, that it will be to the interest of the manufacturer to volun- 
 tarily discard the metal which will not stand test. 
 
 Am. Ry. Eng. Assn.: 
 
 24. The standard length of rails shall be 33 feet, at a temperature of 60 de- 
 grees Fahrenheit. Ten per cent of the entire order will be accepted in shorter 
 lengths varying by 1 foot from 32 feet to 25 feet. A variation of J inch from the 
 specified lengths will be allowed. No. 1 rails less than 33 feet long shall be painted 
 green on both ends. 
 
 TABLE CI.— LENGTH OF RAILS IN STANDARD AMERICAN SPECIFICATIONS 
 
 Specifications. 
 
 Standard 
 Lengths. 
 
 Shorter Lengths that 
 will be Accepted. 
 
 
 Feet. 
 33* 
 
 30 or 33 
 
 30 or 33 
 
 33 
 
 33* 
 
 Feet. 
 32, 31, 30, 29 
 28, 27, 26, 25 
 1 Varying by even 
 > feet to twenty- 
 ) four (24) feet. 
 30, 27|, 25 
 
 Manufacturers' 
 
 A. S. for T. M 
 
 
 
 
 * At a tempers 
 
 ture of 60° F. 
 
 
 All of the specifications agree in allowing ten per cent of the entire order to 
 be shorter lengths than the standard and permit of a variation of I inch in length 
 from that specified. All call for the short-length rails to be painted green on both 
 
RAIL SPECIFICATIONS 
 
 475 
 
 ends. The standard length and short lengths that will be accepted, however, 
 vary, and the requirements given in the various specifications are shown in Table CI. 
 
 Shrinkage or Control of Finishing Temperature 
 
 Am. Ry. Eng. Assn.: 
 
 25. The number of passes and speed of train shall be so regulated that, on 
 leaving the rolls at the final pass, the temperature of the rail will not exceed 
 that which requires a shrinkage allowance at the hot saws, for a rail 33 feet in 
 length and of 100 pounds' section, of 6f inches, and | inch less for each 10 pounds' 
 decrease in section. 
 
 .6* 
 
 6± 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 MANU 
 
 FACTUR 
 A.S. FO 
 
 ERS \ 
 
 RT.MJ 
 
 
 
 
 
 
 
 
 
 
 
 A.R.E.&. 
 
 A.R.A.\ 
 M.W.A, J 
 
 3^2 
 
 ^^ 
 
 
 
 
 
 
 *^^ £+ 
 
 ^ — 
 
 A.S.C.E. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 TO 80 90 IOO 
 
 WEIGHT OF RAIL. LBS. PER- YD. 
 
 Fig. 340. — Shrinkage Allowed in American Specifications in 1909. 
 
 26. The bars shall not be held for the purpose of reducing their tempera- 
 ture, nor shall any artificial means of cooling them be used after they leave 
 the finishing pass. Rails, while on the cooling beds, shall be protected from 
 snow and water. 
 
 The other specifications are the same, except that in the Manufacturers' 
 American Society for Testing Materials and Harriman Lines, the statement that 
 rails, while on the cooling beds, shall be protected from snow and water, is omitted. 
 
 A greater shrinkage is allowed in present specifications than was formerly the 
 case. Fig. 340 shows the allowance of different specifications in 1909; it will be 
 
476 STEEL RAILS 
 
 noted that the shrinkage has been increased to agree with the higher figure given by 
 the manufacturers. In the Pennsylvania requirements of that date, it was pro- 
 vided that the shrinkage allowance be decreased at the rate of t }q inch for each 
 second of time elapsed between the rail leaving the finishing rolls and being sawed. 
 The A. S. C. E. specifications called for -gV inch in place of i| ¥ inch- 
 
 The control of the finishing temperature by the amount of contraction which 
 the rail undergoes in cooling from the finishing to the atmospheric temperature 
 appears to be the only method practical to use. Other efforts have been made 
 to determine the finishing temperature by the use of pyrometers and by the ex- 
 amination of the microstructure of the rails. The use of pyrometers naturally 
 suggested itself at first as the most promising means of accomplishing that pur- 
 pose, but it was soon found that no pyrometric device existed which could be 
 applied in a practical way to the detection of the temperature of quickly moving 
 rails. The micro-test, although attractive and useful, can only be applied to a 
 very small percentage of the rails manufactured, and this is its greatest 
 weakness. 
 
 Section 
 Am. Ry. Eng. Assn.: 
 
 27. The section of rails shall conform as accurately as possible to the tem- 
 plet furnished by the railroad company. A variation in height of ^ inch less 
 or 3V inch greater than the specified height and T \ inch in width of flange will 
 be permitted; but no variation shall be allowed in the dimensions affecting 
 the fit of the splice bars. 
 
 The other specifications are substantially the same as those given above. The 
 Manufacturers' specify A. S. C. E. sections; the Harriman Lines, A. R. A. section 
 "A" 90-pound, Railroad Company's "Common Standard" 75-pound, A. S. C. E. 
 section, 65-pound; and the American Society for Testing Materials say, unless 
 otherwise specified, A. S. C. E. sections. 
 
 Weight 
 Am. Ry. Eng. Assn.: 
 
 28. The weight of the rails specified in the order shall be maintained as 
 nearly as possible, after complying with the preceding section. A variation of 
 one-half of one per cent from the calculated weight of section, as applied to an 
 entire order, will be allowed. 
 
 29. Rails accepted will be paid for according to actual weights. 
 The other specifications are substantially the same as the above. 
 
RAIL SPECIFICATIONS 477 
 
 Straightening 
 Am. Ry. Eng. Assn.: 
 
 30. The hot straightening shall be carefully done, so that gagging under 
 the cold presses will be reduced to a minimum. Any rail coming to the straight- 
 ening presses showing sharp kinks or greater camber than that indicated by a 
 middle ordinate of 4 inches in 33 feet, for A. R. A. type of sections, or 5 inches 
 for A. S. C. E. type of sections, will be at once classed as a No. 2 rail. The 
 distance between the supports of rails in the straightening presses shall not be 
 less than 42 inches. The supports shall have flat surfaces and be out of wind. 
 
 All of the specifications are substantially the same. 
 
 Drilling 
 Am. Ry. Eng. Assn.: 
 
 31. Circular holes for joint bolts shall be drilled to conform accurately in 
 every respect to the drawing and dimensions furnished by the Railroad Company. 
 
 Substantially the same for all specifications. 
 
 32. (a) All rails shall be smooth on the heads, straight in line and surface, 
 and without any twists, waves or kinks. They shall be sawed square at the 
 ends, a variation of not more than one-thirty-second inch being allowed; and 
 burrs shall be carefully removed. 
 
 (&) Rails improperly drilled or straightened, or from which the burrs have 
 not been removed, shall be rejected, but may be accepted after being properly 
 finished. 
 
 Substantially the same for all specifications. 
 
 Branding 
 Am. Ry. Eng. Assn.: 
 
 33. The name of the manufacturer, the weight and type of rail, and the 
 month and year of manufacture shall be rolled in raised letters and figures on 
 the side of the web. The number of the heat and a letter indicating the portion 
 of the ingot from which the rail was made shall be plainly stamped on the web 
 of each rail, where it will not be covered by the splice bars. The top rails shall 
 be lettered "A," and the succeeding ones "B," "C," "D," etc., consecutively; 
 but in case of a top discard of twenty or more per cent, the letter "A" will be 
 omitted. Open-hearth rails shall be branded or stamped "0. H.* w All mark- 
 ings of rails shall be done so effectively that the marks may be read as long as 
 the rails are in service. 
 
 The Pennsylvania specifications are the same as the above. 
 
478 
 
 STEEL RAILS 
 
 All of the other specifications omit the clause in the American Railway Engi- 
 neering Association specifications in reference to omitting the letter "A" in case of 
 a top discard of twenty per cent or more, but are substantially the same in other 
 respects. The Harriman Lines specify that all "A" rails shall have the top of the 
 flange at each end painted yellow, and the American Society for Testing Materials 
 only require rails weighing 70 pounds per yard or over to be stamped with a letter 
 to indicate the portion of the ingot from which the rail was rolled. 
 
 Separate C 
 Am. Ry. Eng. Assn. : 
 
 34. All classes of rails shall be kept separate from each other. 
 
 Loading 
 Am. Ry. Eng. Assn. : 
 
 35. All rails shall be loaded in the presence of the inspector. 
 
 The Pennsylvania specifications are the same for both of the above clauses. 
 The Harriman Lines specifications contain the following clause: 
 
 The following classes of rail shall be loaded separately as far as practicable, 
 excepting at the finishing of an order or the end of a rolling. In this case the differ- 
 ent classes shall be kept separate by placing strips of wood between each class, and 
 each shipping notice shall contain full information as to the contents of each car: 
 
 No. 1 rails, B, C, D, etc., full lengths. 
 
 No. 1 rails, B, C, D, etc., short lengths. 
 
 No. 1 "A" rails; that is, rails from the top of the ingot, full length. 
 
 No. 1 "A" rails, short length. 
 
 No. 2 rails, all lengths. 
 
 35. Specifications (New York Central Lines) for Basic Open-hearth 
 Rails 
 i st — Chemical Composition: 
 
 
 80 Lb. 
 
 90 Lb. 
 
 100 Lb. 
 
 
 .55 to .68 
 
 .70 to 1.00 
 
 .10 to .20 
 
 .04 
 
 .60 to .73 
 
 .70 to 1.00 
 
 .10 to .20 
 
 .04 
 
 .62 to .75 
 
 .70 to 1.00 
 
 .10 to .20 
 
 .04 
 
 
 
 
 
 To adjust the chemical composition to the special conditions of manu- 
 facture at each mill, the engineer representing the railroad company, from the 
 inspection of the ingots, their heating, blooming, and rolling into rails, shall 
 have the right to select the lower or average limit of either the silicon or man- 
 
RAIL SPECIFICATIONS 479 
 
 ganese, or both, with the average carbon content as the working basis for mak- 
 ing the steel, as he may find requisite for good setting ingots with freedom from 
 pipes and rolling into tough steel by the plant of the manufacturer. 
 
 2nd — Process of Manufacture : The entire process of manufacture and 
 testing shall be in accordance with the best current state of the art, and special 
 care shall be taken to conform to the following instructions: 
 
 (a) Excessive use of material thrown into the teeming ladle to set the 
 stopper must be avoided. 
 
 (6) The steel must be well deoxidized and the waste products eliminated 
 before the ingots are teemed. 
 
 (c) The steel must be made to set quiet by the chemical composition in the 
 
 molds without the addition of aluminum, either in the ladle or molds. 
 
 (d) Spattering the interior sides of the molds in pricking the melts and 
 
 teeming the ingots to be avoided as much as possible. 
 
 (e) Time must be allowed for the tops of the ingots to set without spray- 
 
 ing with water. 
 
 (/) The ingots should be stripped as soon as the metal caps over on top, 
 then sent to the scales to be weighed, then sent to the reheating 
 furnaces to be charged promptly, to avoid the cooling of the interior 
 metal and thus check the large shrinkage which occurs in it from 
 unnecessary loss of temperature due to delays. It has been found, 
 in good practice, possible in this way to confine the interior shrinkage 
 to 0.05 to 0.1 of one per cent per cubic foot of the metal of rail 
 ingots. The total shrinkage of an ingot depends upon its volume, 
 chemical composition and loss of temperature at the time it is charged, 
 yet in fair practice it may be confined to such small limits that it is 
 removed in the usual discard of the bloom. Piped rails come from 
 cold ingots or those which have been unduly delayed before charging 
 into the reheating furnace. 
 
 (g) Cast Iron Cut Out of the Ingot Stools: Care to be taken in teeming the 
 ingots to prevent cutting out of the cast iron of the stools or ingot 
 molds by the falling stream of hot metal from the ladle, avoiding a 
 frequent cause of carbon streaks found in the segregated steel of 
 " split heads." The most disturbing factor of the small amount of 
 ordinary segregation in rail steel is the diffused cast iron in some 
 ingots cut out from the stools. 
 
 (h) Ingots shall be kept in a vertical position on the ingot cars and in the 
 reheating furnaces until their heat is equalized ready to be rolled. 
 
480 STEEL RAILS 
 
 (i) Bled ingots shall not be used. ("Bled ingot" — one from the center 
 
 of which the liquid steel has been permitted to escape.) 
 (j) There shall be sheared from the end of the bloom formed from the top 
 of the ingot sufficient discard to secure sound rails. (All metal from 
 the top of the ingot, whether cut from the bloom or the rail, is the 
 "top discard.") 
 (k) One-hundred-pound (100 lb.) rails not to be rolled from blooms exceed- 
 ing three (3) thirty- three-foot (33') lengths in a continuous bar; 
 eighty -pound (80 lb.), or lighter, rails in not over four (4) lengths 
 of thirty -three feet (33 ') in a continuous bar, when inserted in the 
 contract. 
 3rd — Shrinkage : The number of passes and speed of train shall be so reg- 
 ulated that, on leaving the rolls on the final pass, the temperature of the rails 
 will not exceed that which requires a shrinkage allowance at the hot saws for 
 a 33-foot rail of 100 pounds section of 6f inches, and T V inch less for each five 
 pounds decrease of section. No artificial means of cooling the steel shall be 
 used between the leading and finishing passes, nor after the rails leave the 
 finishing rolls; neither shall rails be held before sawing for the purpose of 
 reducing their temperature. 
 
 4th — Drop and Ductility Tests : A drop test to be made of a crop from 
 the top bar of the second, the middle and the last full ingot of the melt. The 
 crop 4 to 6 feet long to be stamped with a spacing bar of six one-inch spaces on 
 the base, head or side as desired. 
 
 Each butt must show under a single blow of the drop, of 18-foot, for the 
 80-pound or 90-pound section, and 20-foot for the 100-pound section, at least 
 six per cent elongation for one inch or five per cent each for two consecutive 
 inches before fracture for acceptance of the melt. 
 
 The crop or butt is liable to be chilled accidentally in entering the rolls 
 several times, or it may be caused by other delays, and should it break under 
 a single blow without showing the percentage of elongation specified, it shall 
 "be considered as indicating deficient ductility or chilled metal, and the results 
 must be rejected. 
 
 The Inspecting Engineer representing the Railroad Company must then 
 take a duplicate test from the same ingot at the top end of the "A " or " B " 
 rail, according to the nine or greater percentage of discard, and the results 
 taken in lieu of those from the first crop or test to determine whether or not 
 the piece had the requisite ductility in accordance with the specifications. 
 
 The distinction between a chilled test crop and those of inadequate ductility 
 
RAIL SPECIFICATIONS 481 
 
 must be ascertained according to above prescribed tests before rejections are 
 made or rails accepted. 
 
 Should any test piece under the first blow of the drop not break, but fail 
 to show the percentage of elongation specified, the test piece shall be subjected 
 in the same position to a second blow and the results so obtained govern in 
 passing the test. 
 
 The ductility of at least one specimen of each melt to be exhausted by one 
 or more blows of the drop, and a record made of the respective elongations of 
 each test. 
 
 The drop-testing machine shall have a tup of 2000 pounds weight, the 
 striking face of which shall have a radius of not more than five inches (5"), 
 and solid supports, centers three feet (3') apart, for the test butts. The anvil 
 block shall weigh at least 20,000 pounds and the supports shall be part of or 
 firmly secured to the anvil. The report of drop test shall state the atmospheric 
 temperature at the time the test was made. The testing shall proceed con- 
 currently with the operation of the mill. The temperature of the test butts 
 to be between 40 degrees and 100 degrees Fahr. 
 
 5th — Section : The section of rail shall conform to the dimensions fur- 
 nished by the purchaser as accurately as possible consistent with the paragraph 
 relative to specified weight. 
 
 A variation in height of rails of 3V of an inch over or e 1 ? of an inch under, 
 also yq of an inch in width of flange will be permitted, but no variation will 
 be allowed in dimensions affecting the fit of the splice bars. 
 
 6th — Weight: The weight of the rail shall be maintained as nearly as 
 possible, after complying with the preceding paragraph, to that specified in the 
 contract. 
 
 A variation of one-half of one per cent, from the calculated weight of sec- 
 tion, on the entire order, will be allowed. 
 
 Rails will be accepted and paid for according to actual weight. 
 
 7th — Length: The standard length of rails shall be thirty-three feet (33'). 
 Ten per cent of the entire order will be accepted in shorter lengths varying as 
 follows: Thirty feet (30'), twenty-eight feet (28'), twenty-six feet (26') and 
 twenty -four feet (24'). A variation of f of an inch from the specified length 
 will be allowed. 
 
 Three rails in every 100 tons to be thirty-two feet and six inches (32' 6") 
 long, the ends painted red, when inserted in the New York Central Contract. 
 
 All other No. 1 rails less than thirty-three feet (33') long shall be painted 
 green on both ends. 
 
482 STEEL RAILS 
 
 8th — Branding : The name of the maker, the weight of the rail and the 
 month and year of manufacture, together with "0-H," shall be rolled in raised 
 letters on the side of the web, and the number of the melt and letter to desig- 
 nate the position of the rail in the ingot shall be so stamped on each rail as not 
 to be covered by the splice bars. 
 
 When the rails are to be rolled with twenty per cent (20%) discard the 
 first rail in the ingot shall commence with the letter "B," the second "C, " 
 the third "D" and the fourth "E." 
 
 When the "A" rails are to be taken they are to be loaded separately upon 
 cars for shipment and the flanges at the ends painted yellow, when inserted 
 in the contract. 
 
 9th — Drilling : Circular holes for splice bars shall be drilled in accordance 
 with specifications of purchaser. They shall in every respect accurately con- 
 form to drawing and dimensions furnished and shall be free from burrs. 
 
 ioth — Straightening: Care must be taken in cambering the rails and 
 with the hot-bed work, which must result in the rails being left in such con- 
 dition that they shall not vary throughout their entire length more than four 
 inches (4") for the "A. R. A." thick bases and not more than five inches (5") 
 for the "DUDLEY" section or "A. S. C. E." sections from a straight line in 
 any direction when delivered to the cold-straightening presses. Those which 
 vary beyond that amount, or have short kinks, shall be classed as second qual- 
 ity rails and be so marked. Rails while on the "hot-beds" shall be protected 
 from coming in contact with water or snow. The distance between supports 
 of rails in the gagging press shall not be less than forty-two inches (42"); 
 supports to have flat surfaces. 
 
 Rails shall be straight in line and surface and smooth on head when fin- 
 ished — final straightening being done while cold. They shall be sawed square 
 at ends, variations to be not more than ^ of an inch, and prior to shipment 
 shall have the burr caused by the saw cutting removed and the ends made 
 clean. 
 
 nth — Inspection: The inspector representing the purchaser shall have 
 free entry to the works of the manufacturer at all times while his contract is 
 being executed, and shall have all reasonable facilities afforded him by the 
 manufacturer to satisfy him that the rails are being made in accordance with 
 the terms of the contract. All tests and inspection shall be made at the place 
 of manufacture prior to shipment, and shall be so conducted as not to unneces- 
 sarily interfere with the operation of the mill. 
 
 The manufacturer shall furnish the inspector with a chemical analysis of 
 
RAIL SPECIFICATIONS 483 
 
 each melt of steel covering the elements specified in the section No. 1 hereof, 
 and also report sulphur and copper. 
 
 Analysis shall be made on drillings taken from small test ingots, the drill- 
 ing being taken at a distance of not less than | of an inch beneath the surface 
 of said test ingots. On request of the inspector the manufacturer shall furnish 
 drillings for check analysis. 
 
 12th — No. 2 Rails : Rails which by reason of surface imperfections are 
 not classed as No. 1 rails shall be considered No. 2 rails, but No. 2 rails shall 
 not be accepted for shipment which have flaws in the head of more than | of 
 an inch; or in the flange of more than \ of an inch in depth; and these shall 
 not, in the judgment of the inspector, be, in any individual rail, so numerous 
 or of such a character as to render it unfit for recognized No. 2 rail uses. 
 
 13th — Designation of No. 2 Rails and Short Lengths of No. 1 Rails : 
 Both ends of all No. 2 rails shall be painted white. 
 
 Both ends of all short lengths No. 1 rails shall be painted green, except 
 the 32-foot and 6-inch rails, which are to be painted red. 
 
 (Sgd.) P. H. DUDLEY, 
 New York Central Lines. 
 (Specifications of Oct. 1st, 1909. 
 Revised Jan. 11th, 1911, to con- 
 form to Manufacturers' sale per 
 100 pounds.) 
 
 Note 1. "Process of Manufacture" (b): The elimination of the deoxidation 
 products and impurities from the bath of metal is more important than has 
 yet been appreciated. This prevents minute portions of the deoxidation prod- 
 ucts from becoming entrained in the setting metal and therefore will avoid 
 their being rolled in the steel, where in the rail head or base they would be 
 subjected to alternate unit fiber strains under moving trains and contribute 
 the needed factor to develop the interior transverse checks recently observed 
 in a few rail heads. 
 
 Time is required for the deoxidation products and impurities to rise after 
 the steel is tapped into the ladle. 
 
 These heterogeneous portions of the deoxidation products or impurities 
 in the steel, as well as small flaws and interior cavities, are theoretically and 
 practically known to be zones of weakness, and interrupt the normal unit strains 
 and increase them in the surrounding metal, which often result in detailed 
 fractures. 
 
 Note 2. "Process of Manufacture" (/): The percentage of interior shrink- 
 
484 
 
 STEEL RAILS 
 
 Chemical 
 Composition. 
 
 Chemical 
 Analysis. 
 
 age per cubic foot of the metal of the ingots there mentioned was reduced the 
 past year by good mill practice and well organized train service. The latter 
 was to transport promptly the ingots after they were teemed and stripped 
 so that they could be charged with the least possible delay into the reheating 
 pits and then as soon as the heat of the metal was properly equalized, they 
 were bloomed, which restricted the reduced cavity to the discard. 
 Note 1 and 2 added for information. 
 P.H.D. 
 1/3/12 
 
 36. British Standard Specifications of Bull Head Railway Rails 
 
 (Report No. 9, Revised July, 1909.) 
 Issued by The Engineering Standards Committee 
 
 Supported by: The Institution of Civil Engineers; The Institution of Mechanical Engineers; The 
 Institution of Naval Architects; The Iron and Steel Institute; The Institution of Electrical Engineers. 
 (Reprinted by permission of the Committee) . 
 
 1 . The steel for the Rails shall be of the best quality made by the Bessemer, Siemens-Martin, 
 or other process, as may be approved by the Engineer (or by the Purchaser). 
 
 The Rails shall show on analysis that in chemical composition they conform to the following 
 limits: 
 
 Carbon from 0.35 to 0.5 per cent. 
 
 Manganese " .7 to 1.0 " " 
 
 Silicon not to exceed 0.1 " " 
 
 Phosphorus " 0.075 
 
 Sulphur " " 0.08 " " 
 
 2. The Manufacturer shall make and furnish to the representative of the Engineer (or of 
 the Purchaser) carbon determinations of each cast. 
 
 A complete chemical analysis, representing the average of the other elements contained 
 in the steel, shall be similarly given for each rolling. Such complete analysis shall be made from 
 drillings taken from the rail or tensile test piece or pieces. When the rolling exceeds 200 tons, 
 an additional complete analysis shal be made for each 200 tons or part thereof. 
 
 Should the Engineer (or the Purchaser) desire to make independent chemical determinations, 
 the necessary specimens and samples shall be furnished by the Manufacturer. For this purpose 
 not more than two rails in every hundred tons manufactured shall be selected by representative 
 of the Engineer (or of the Purchaser) and drillings taken with a drill of 2 inches diameter from 
 the top of the head of the rail, unless otherwise specified by him, and if, upon being subjected to 
 the specified tests, either fail to comply therewith, then all the Rails in the cast of which the test 
 pieces form a part may be rejected. 
 
 The representative of the Engineer (or of the Purchaser) may then take similar samples 
 from a further two rails out of the same 100 tons, and should either fail to comply with the 
 specified analysis the whole 100 tons may be rejected. 
 
 In case of difference between the Engineer (or between the Purchaser) and the Manufacturer 
 as to the accuracy of any analysis, either party shall have the right to have samples of the steel 
 analyzed by an independent metallurgist, to be mutually agreed upon. The expenses attendant 
 upon such independent analysis shall be borne by the party adjudged to be in the wrong. 
 
 3. Each Rail shall be made from an ingot not less than 12 inches square at the smaller and 
 14 inches square at the larger end, and must be cogged down into blooms, and sufficient crop 
 then sheared from each end to ensure soundness. 
 
RAIL SPECIFICATIONS 
 
 485 
 
 All straightening shall be done by pressure and not by hammering. 
 4. A rolling margin of i per cent under to | per cent above the calculated 
 permitted, but the calculated weight only will be paid for. 
 
 Permissible 
 will be Variation 
 in Weight. 
 
 TABLE OF GENERAL DIMENSIONS AND WEIGHTS OF 
 
 (See Plate XIV) 
 
 B. S." RAILS 
 
 Number of "B.S." 
 
 
 
 
 Sl.tI inn mill Nominal 
 
 Weight per Yard 
 
 in lbs. 
 
 Height of Rail. 
 
 Width of Head. 
 
 of Rail. 
 
 Pounds. 
 
 Inches. 
 
 Inches. 
 
 Pounds per Yard. 
 
 60 
 
 4| 
 
 2A 
 
 59.79 
 
 65 
 
 4J 
 
 2f 
 
 64.58 
 
 70 
 
 5 
 
 2rV 
 
 70.13 
 
 75 
 
 51 
 
 2h 
 
 74.56 
 
 80 
 
 5| 
 
 2r 9 « 
 
 79.49 
 
 85 
 
 5M 
 
 m 
 
 84.88 
 
 90 
 
 5fi 
 
 2f 
 
 89.77 
 
 95 
 
 5ft 
 
 2f 
 
 94.59 
 
 100 
 
 5ft 
 
 2| 
 
 99.84 
 
 General 
 Dimensions 
 of Rails. 
 
 6. Before the general manufacture of the Rails is commenced the Manufacturer shall, if 
 required by the Engineer (or by the Purchaser), supply two sets of templates, internal and external, 
 of approved material, for each " B. S." Section of Rail. 
 
 Each template shall be suitably engraved with the Purchaser's name, the number of the 
 " B. S." section (being the nominal weight of the Rail in pounds per lineal yard), the Manufac- 
 turers' name and address, and the date of the Contract. 
 
 These templates shall be submitted to the Engineer (or to the Purchaser) for his approval, 
 and at the commencement of rolling the Engineer will have a competent person present to approve 
 of the section. 
 
 7. Each Section of Rail under this Contract shall be accurately rolled to its respective 
 template. 
 
 8. The whole of the Rails shall be of uniform section throughout, true to templates, per- 
 fectly sound and straight, and free from splits, cracks, burrs and defects of every kind. 
 
 9. A quantity of short lengths will be taken in such lengths and quantities as may be 
 ordered by the Engineer (or by the Purchaser), provided that these short lengths are cut down 
 from longer lengths found to be defective at the ends only, and that the total quantity taken 
 does not exceed 7 J per cent of the contract. 
 
 N. B. — The Committee recommend the adoption of the following, as the normal lengths 
 of Rails, viz. : — 30 feet, 36 feet, 45 feet, or 60 feet. 
 
 10. The Rails shall be the specified length at the temperature of 60° Fahr. No Rail 
 will be accepted which is more than three-sixteenths of an inch (& inch) above or below the 
 length specified, whether for curved or straight line. 
 
 11. When required by the Engineer (or by the Purchaser) rails are to be supplied from 1 to 
 6 inches shorter or longer than the normal specified lengths, and these special lengths are to have 
 about one foot at each end painted with such colors as may be ordered. 
 
 12. Rails shall be supplied for switches and crossings when so ordered, and such Rails shall 
 be of the required lengths and shall be cut from sound long Rails. 
 
 13. The Brand shall be rolled on the web of each Rail to show that the Rail is of British Standard 
 Section and made under the conditions of this Specification; the number of the " B. S." Section 
 (being the nominal weight of the Rail in pounds per yard), the process by which the Rails have 
 been manufactured, the Manufacturer's name, initials, or other recognized mark, and the month 
 and year of manufacture shall be rolled, in letters three-quarters of an inch (f inch) in size, on one 
 
 Templates. 
 
 Rails to Conform 
 to Template. 
 
 Rails to be Free 
 from Defects. 
 
 Length of Rails 
 for Straight 
 Line. 
 
 Permissible 
 Variation 
 in Length. 
 
 Rails of Special 
 Length for 
 Matching in 
 Curved Line. 
 Rails for 
 Switches and 
 Crossings. 
 Branding. 
 
486 
 
 STEEL RAILS 
 
 side of the web of each Rail, e.g., \y B.S. 95, B.A.* 4.04; and the number of the cast or 
 
 blow from which it has been rolled shall be stamped on the end of each Rail in half -inch (§ inch) 
 block figures. 
 
 14. From each cast one rail shall be selected by the representative of the Engineer (or of 
 the Purchaser). From this a piece 5 feet long shall be cut which shall be placed in a horizontal 
 position with the bullhead uppermost upon two iron or steel supports resting on a solid founda- 
 tion and placed so that their centers are 3 feet 6 inches apart, the upper surfaces of the supports 
 being curved to a radius of 3 inches. The test shall comprise two blows delivered midway 
 between the bearings from a falling iron weight of 2240 pounds, the striking face of which shall be 
 rounded to a radius of not more than 5 inches. The heights of the drop for the various sections 
 of Rails shall be as tabulated below. The blows must be sustained without fracture, and the 
 Rail must show a deflection between the limits given below. 
 
 FALLING WEIGHT TEST 
 
 
 
 First Blow 
 
 
 Second Blow 
 
 Number of "B. S." 
 Section and Nom- 
 
 Drop. 
 
 Deflection. 
 
 Drop. 
 
 Deflection. 
 
 i'kiI Uemln nf Kails 
 
 
 
 
 
 per Yard in lbs. 
 
 
 From. To. 
 
 
 From. To. 
 
 Pounds. 
 
 Feet. 
 
 Inches. 
 
 Feet. 
 
 Inches. 
 
 60 
 
 5 
 
 1 1ft 
 
 10 
 
 3 3f 
 
 65 
 
 5 
 
 1 U 
 
 12 
 
 3 3f 
 
 70 
 
 6 
 
 1 1ft 
 
 12 
 
 3 3i 
 
 75 
 
 6 
 
 1 1ft 
 
 12 
 
 3 3J 
 
 80 
 
 6 
 
 1 1ft 
 
 15 
 
 3 4 
 
 85 
 
 6 
 
 1 1ft 
 
 15 
 
 3 4 
 
 90 
 
 7 
 
 1 11 
 
 20 
 
 3 4} 
 
 95 
 
 7 
 
 1 1ft 
 
 20 
 
 3 4J 
 
 100 
 
 7 
 
 1 1ft 
 
 20 
 
 3 41 
 
 Should the length cut from the selected Rail fail to comply with the test specified for its 
 weight, two other Rails from the same cast will be selected and similar lengths cut and tested, and 
 the acceptance or rejection of the cast will be decided by the result of the three tests, so that if 
 two of the Rails selected fail to comply with the test, the entire cast will be rejected. 
 
 15. From each 100 tons of Rails the Manufacturer shall (if required by the representative 
 of the Engineer or of the Purchaser) cut a test piece from any Rail selected as a sample Rail ; such 
 test piece to be stamped to correspond with the sample Rail. It shall then be placed in a testing 
 machine of approved pattern, and shall have an ultimate tensile strength equivalent to not less 
 than 40 tons per square inch, nor more than 48 tons per square inch, with an elongation of not 
 less than 15 per cent upon the Standard Test Pieces C or D (see Fig. 341). Should the test piece 
 fail to fulfil these conditions, the representative of the Engineer (or of the Purchaser) may require 
 the Manufacturer to test two other Rails from the same cast in the same manner, and the ac- 
 ceptance or rejection of the cast shall be decided by the results of the three tests so that if two of 
 the three Rails selected fail to comply with the test the entire cast will be rejected. 
 
 The representative of the Engineer (or of the Purchaser) may then take similar test pieces 
 from a further two rails out of the same 100 tons, and should either fail to comply with the test 
 the whole 100 tons may be rejected. 
 
 Should the Engineer (or the Purchaser) desire to have independent tests made, the Manu- 
 facturer shall provide the necessary test pieces, viz., two for every 200 tons, properly shaped and 
 prepared as described in Fig. 341. 
 
 * The following abbreviations are recommended : 
 S.A. Siemens-Martin Acid. 
 S.B. Siemens-Martin Basic. 
 
 B.A. Bessemer Acid. 
 B.B. Bessemer Basic. 
 
RAIL SPECIFICATIONS 
 
 487 
 
 16. The holes for fishbolts must be drilled through the web from the solid at each end of Holes in RaiLs. 
 the Rails, of the sizes and in the position shown in the British Standard specification for Fish 
 plates for Bull Head Rails (Report No. 47) or on a drawing to be supplied by the Engineer (or 
 the Purchaser). These holes must be clean and square with the web, without burrs on either 
 side, and will be checked with the gauges to be furnished to the Manufacturer by the Engineer 
 (or by the Purchaser). Should any of the holes vary from the correct size or position more than 
 one thirty-second of an inch (^2 inch) the Rails in question will be liable to rejection. 
 
 i u , 
 
 DIA.= .564 IN. 
 
 |i )§ ! 
 
 AREA= V* SQ.IN. 
 
 ! '« 2" GAUGE LENGTH ►» 
 
 u PARALLEL FOR A LENGTH «, 
 
 OF NOT LESS THAN 2 >/4" 
 
 TEST PIECE C. 
 
 
 1 W \ 0IA.-.798 
 
 / W 
 
 ! 1 AREA=HSQ 
 
 1 i 
 
 J 1 
 
 L PARALLEL FOR A LENGTH OF 
 
 NOT LESS THAN 3V«" 
 
 _J 
 
 TEST PIECE D. 
 - Test Pieces C and D, British Standard S 
 
 cifications of Rails. 
 
 The gauge length and the parallel portion are to be as shown, the form of the ends to be as 
 required in order to suit the various methods employed for gripping the test piece. 
 
 17. The Manufacturer shall give to the Engineer (or to the Purchaser), or his representative, Notice of 
 
 at least seven clear days' previous notice, in writing, before the rolling of the first lot of Rails, Rolling to be 
 and at least three clear days' previous notice, in writing, before the rolling of any subsequent lot Given - 
 of Rails, is commenced, in order that arrangements may be made for the presence of the repre- 
 sentative of the Engineer (or of the Purchaser) at the rolling. 
 
 18. The Engineer (or the Purchaser) or his representative shall have access to the works T nS p ect ion 
 of the Manufacturer at all reasonable times. He shall be at liberty to examine the Rails during and Testing. 
 any stage of their manufacture, and to reject any material or finished Rail which does not conform 
 
 to the terms of this specification. 
 
 Before the Rails are put before the representative of the Engineer (or of the Purchaser) 
 for inspection the Manufacturer shall have them examined, and all Rail which he admits to be 
 defective are to be sorted out and placed in a separate stack; the representative of the Engineer 
 (or of the Purchaser) being empowered to refuse to inspect any lot of Rails not put in uniform 
 lengths and sorted. 
 
 19. The Manufacturer shall supply the material required for testing free of charge and shall, 
 at his own cost, furnish and prepare the necessary test pieces, and supply labor and appliances 
 for such testing as .may be carried out at his premises in accordance with this specification. Fail- 
 ing facilities at his own works for making the prescribed tests the Manufacturer shall bear the 
 cost of carrying out the tests elsewhere. 
 
 20. All Rails accepted by the representative of the Engineer (or of the Purchaser) shall be Ac^^ Rails 
 stamped in his presence. 
 
STEEL RAILS 
 
 37. British Standard Specifications of Flat Bottom Railway Rails 
 
 (Report No. 11, Revised July, 1909) 
 
 Issued by The Engineering Standards Committee 
 
 _ Supported by: The Institution of Civil Engineers; The Institution of Mechanical Engineers; The 
 Institution of Naval Architects; The Iron and Steel Institute; The Institution of Electrical Engineers. 
 (Reprinted by permission of the Committee.) 
 
 1. The steel for the Rails shall be of the best quality made by the Bessemer, Siemens- 
 Martin, or other process, as may be approved by the Engineer (or by the Purchaser). 
 
 The Rails, shall show on analysis that in chemical composition they conform to the following 
 limits: 
 
 Carbon from 0.35 to 0.50 per cent. 
 
 Manganese " 0.70 to 1.00 " " 
 
 Silicon not to exceed 0.10 " " 
 
 Phosphorus " " " 0.07 " " 
 
 Sulphur " " " 0.07 " " 
 
 2. The Manufacturer shall make and furnish to the representative of the Engineer (or of 
 the Purchaser) carbon and phosphorus determinations of each cast. 
 
 A complete chemical analysis, representing the average of the other elements contained in 
 the steel, shall be similarly given for each rolling. Such complete analysis shall be made from 
 drillings taken from the Rail or from the tensile test piece or pieces. When the rolling exceeds 
 200 tons, an additional complete analysis shall be made for each 200 tons or part thereof. 
 
 Should the Engineer (or the Purchaser) desire to make independent chemical determinations, 
 the necessary specimens and samples shall be furnished by the Manufacturer. For this purpose 
 not more than two Rails in every 100 tons manufactured shall be selected by the representative 
 of the Engineer (or of the Purchaser) and drillings taken with a drill of 2 inches diameter from the 
 top of the head of the Rail unless otherwise specified by him, and if, upon being subjected to the 
 specified tests, either fail to comply therewith, then all the Rails in the cast of which the test 
 pieces form a part may be rejected. 
 
 The representative of the Engineer (or of the Purchaser) may then take similar samples 
 from a further two rails out of the same 100 tons, and should either fail to comply with the specified 
 analysis the whole 100 tons may be rejected. 
 
 In case of difference between the Engineer (or between the Purchaser) and the Manufac- 
 turer, as to the accuracy of an analysis, either party shall have the right to have samples of the 
 steel analyzed by an independent metallurgist, to be mutually agreed upon The expenses at- 
 tendant upon such independent analysis shall be borne by the party adjudged to be in the wrong. 
 
 3. Each Rail shall be made from an ingot not less than 12 inches square at the smaller 
 end and 14 inches square at the larger end, which must be cogged down into blooms, and have 
 sufficient crop then sheared from each end to ensure soundness. 
 
 All straightening shall be done by pressure and not by hammering. 
 
 4. A rolling margin of | per cent under to \ per cent above the calculated weight will be 
 permitted, but the calculated weight only will be paid for. 
 
 6. Before the general manufacture of the Rails is commenced the Manufacturer shall, if 
 required by the Engineer (or by the Purchaser), supply two sets of templates, internal and external, 
 of approved material, for each " B. S." Section of Rail. 
 
 Each template shall be suitably engraved with the Purchaser's name, the number of the 
 " B. S." section (being the nominal weight of the Rail in pounds per yard), the Manufacturer's 
 name and address, and the date of the Contract. 
 
 These templates shall be submitted to the Engineer (or to the Purchaser) for his approval, 
 and at the commencement of rolling the Engineer will have a competent person present to approve 
 of the section. 
 
RAIL SPECIFICATIONS 
 
 5. TABLE OF GENERAL DIMENSIONS AND WEIGHTS OF "B. S.' ; 
 
 (See Plate XV) 
 
 Number of "B. S." 
 
 
 
 
 Sen hin and Nomi- 
 nal Weight per 
 
 Height of Rail. 
 
 Width of Head. 
 
 Calculated Weight 
 of Rail. 
 
 Yard in Pounds. 
 
 
 
 
 
 Inches. 
 
 Inches. 
 
 Pounds per Yard. 
 
 20 
 
 2i 
 
 
 19.96 
 
 25 
 
 21 
 
 1* 
 
 24.95 
 
 30 
 
 3 
 
 If 
 
 29.98 
 
 35 
 
 3| 
 
 
 35.03 
 
 40 
 
 3| 
 
 1J 
 
 39.98 
 
 45 
 
 3i 
 
 lfs 
 
 45.10 
 
 50 
 
 3fg- 
 
 2tV 
 
 49.94 
 
 55 
 
 4i 
 
 2A 
 
 54.78 
 
 60 
 
 4A 
 
 2i 
 
 60.11 
 
 65 
 
 ^ 
 
 2t% 
 
 64.86 
 
 70 
 
 4f 
 
 2| 
 
 69.77 
 
 75 
 
 4H 
 
 2^ 
 
 74.79 
 
 80 
 
 5 
 
 2| 
 
 79.94 
 
 85 
 
 5^r 
 
 2& 
 
 84.87 
 
 90 
 
 5| 
 
 2f 
 
 89.92 
 
 95 
 
 5t% 
 
 2H 
 
 94.76 
 
 100 
 
 5f 
 
 2| 
 
 99.95 
 
 General 
 Dimensions 
 and Weights 
 of Rails. 
 
 7. Each Section of Rail shall be accurately rolled to its respective template. 
 
 8. The whole of the Rails shall be of uniform section throughout, true to templates, per- 
 fectly sound and straight, and free from splits, cracks, burrs, and defects of every kind. 
 
 9. A quantity of short lengths will be taken in such lengths and quantities as may be ordered 
 by the Engineer (or by the Purchaser), provided that these short lengths are cut down from longer 
 lengths found to be defective at the ends only, and that the total quantity taken does not exceed 
 7J per cent of the Contract. 
 
 10. The Rails shall be the specified length at a temperature of 60° Fahr. No Rail will be 
 accepted which is more than three-sixteenths of an inch (t s s inch) above or below the length 
 specified, whether for straight or curved lines. 
 
 11. When required by the Engineer (or by the Purchaser) Rails are to be supplied from 1 
 to 6 inches shorter or longer than the normal specified lengths, and these special lengths are to 
 have about one foot at each end painted with such colors as may be ordered. 
 
 12. Rails shall be supplied for switches and crossings when so ordered, and such Rails shall 
 be of the required lengths and shall be cut from sound Rails. 
 
 13. The Brand (see sketch) shall be rolled on the web of each Rail to show that the Rail 
 is of British Standard Section and made under the conditions of this Specification; the number 
 of the " B. S." Section (being the nominal weight of the Rail in pounds per yard), the process* by 
 which the Rails have been manufactured, the Manufacturer's name, initials, or other recognized 
 mark, and the month and year of manufacture shall also be rolled, in letters three-quarters of 
 
 an inch (| inch) in size, on one side of the web of each Rail, e.g., W B.S. 95-B.A.* 4.04; 
 
 and the number of the cast from which it has been rolled shall be stamped on the end of each Rail 
 in half-inch (J inch) block figures. 
 
 14. From each cast a piece of Rail (which may be a crop end) shall be selected by the repre- 
 sentative of the Engineer (or of the Purchaser) and stamped with his mark and the number of 
 the cast. From this a piece 5 feet long shall be cut which shall be placed in a horizontal position, 
 with the head uppermost, upon two iron or steel supports resting on a solid foundation, the upper 
 
 Rails to Conform 
 to Templates. 
 Rails to be Free 
 from Defects. 
 
 Length of Rails 
 for Straight 
 Line. 
 
 Permissible 
 Variation in 
 Length. 
 
 Rails of Special 
 Length for 
 Matching in 
 Curved Line. 
 Rails for 
 Switches and 
 Crossings. 
 Branding. 
 
 V 
 
 Impact Test. 
 
 c The following abbreviations are recommended : ■ 
 S.A. Siemens-Martin Acid. 
 S.B. Siemens-Martin Basic. 
 
 B.A. 
 B.B. 
 
 Bessemer Acid. 
 
490 
 
 STEEL RAILS 
 
 surfaces of the supports being curved to a radius of 3 inches. The test shall comprise one blow, 
 delivered midway between the bearings, from a falling iron weight or tup, the striking face of 
 which shall be rounded to a radius of not more than 5 inches. The weight of the tup, the span 
 of the test piece between the centers of the bearings, and the height of the drop for the various 
 sections of Rails shall be as tabulated below. The blow must be sustained without fracture. In 
 addition to the above test the representative of the Engineer (or of the Purchaser) shall select one 
 finished Rail from every 200 offered, and a piece 5 feet in length cut from this Rail shall be similarly 
 tested as specified above. 
 
 Number of "B. S." 
 
 
 Falling Weight Test. 
 
 
 Section and 
 Nominal Weight of 
 
 
 
 
 
 
 
 Rails per Yard 
 in Pounds. 
 
 Weight of Tup. 
 
 Centers of Bearings. 
 
 Drop. 
 
 
 Cwts. 
 
 Feet. 
 
 .Feet. 
 
 20 
 
 5 
 
 3 
 
 8 
 
 25 
 
 5 
 
 3 
 
 9 
 
 30 
 
 10 
 
 3 
 
 10 
 
 35 
 
 10 
 
 3 
 
 12| 
 
 40 
 
 10 
 
 3 
 
 15 
 
 45 
 
 15 
 
 3 
 
 15 
 
 50 
 
 15 
 
 3 
 
 15 
 
 55 
 
 15 
 
 3 
 
 m 
 
 60 
 
 20 
 
 3 
 
 20 
 
 65 
 
 20 
 
 3 
 
 20 
 
 70 
 
 20 
 
 3£ 
 
 20 
 
 75 
 
 20 
 
 3| 
 
 20 
 
 80 
 
 20 
 
 31 
 
 22 
 
 85 
 
 20 
 
 31 
 
 24 
 
 90 
 
 20 
 
 3J 
 
 26 
 
 95 
 
 20 
 
 3| 
 
 28 
 
 100 
 
 20 
 
 3£ 
 
 30 
 
 Should the length cut from the selected Rail fail to comply with the test specified for its 
 weight, two other Rails from the same cast will be selected and similar lengths cut and tested, 
 and the acceptance or rejection of the cast will be decided by the result of the three tests, so that 
 if two of the Rails selected fail to comply with the test the entire cast will be rejected. 
 
 15. From each 100 tons of Rails the Manufacturer shall (if required by the representative 
 of the Engineer or of the Purchaser) cut a test piece from any Rail selected as a sample Rail; such 
 test piece to be stamped to correspond with the sample Rail. It shall then be placed in a testing 
 machine of approved pattern, and shall have an ultimate tensile strength of not less than 40 tons 
 per square inch, nor more than 48 tons per square inch, with an elongation of not less than 15 per 
 cent upon the Standard Test Pieces C or D (see Fig. 341). Should the test piece fail to fulfil these 
 conditions, the representative of the Engineer (or of the Purchaser) may require the Manufacturer 
 to test two other Rails from the same cast in the same manner, and the acceptance or rejection 
 of the cast shall be decided by the result of the three tests, so that if two of the three Rails selected 
 fail to comply with the test the entire cast will be rejected. 
 
 The representative of the Engineer (or of the Purchaser) may then take similar test pieces 
 from a further two Rails out of the same 100 tons, and should either fail to comply with the test 
 the whole 100 tons may be rejected. 
 
 Should the Engineer (or the Purchaser) desire to have independent tests made, the Manu- 
 facturer shall provide the necessary test pieces, viz., two for every 200 tons, properly shaped and 
 prepared as described in Fig. 341. 
 
 16. The holes for fishbolts shall be drilled through the web from the solid at each end of 
 the Rails, of the sizes and in the position shown in the British Standard Specification for Fish 
 Plates for Flat Bottom Rails (Report No. 47), or on a drawing to be supplied by the Engineer 
 (or by the Purchaser). These holes must be clean and square with the web, without burrs on 
 
RAIL SPECIFICATIONS 491 
 
 either side, and will be checked with the gauges to be furnished to the Manufacturer by the 
 Engineer (or by the Purchaser). Should any of the holes vary from the correct size or position 
 more than one thirty-second of an inch (s\ inch) the Rails in question will be liable to rejection. 
 
 17. The Manufacturer shall give to the Engineer (or to the Purchaser), or his representative, Notice of 
 
 at least seven clear days' previous notice, in writing, before the rolling of the first lot of Rails, R? llin S to Be 
 and at least three clear days' previous notice, in writing, before the rolling of any subsequent lot 
 of Rails, is commenced, in order that arrangements may be made for the presence of the repre- 
 sentative of the Engineer (or of the Purchaser) at the rolling. 
 
 18. The Engineer (or the Purchaser), or his representative, shall have free access to the Inspection and 
 works of the Manufacturer at all reasonable times: he shall be at liberty to examine the Rails Testm £- 
 during any stage of their manufacture, and to reject any material or finished Rail which does not 
 
 conform to the terms of this Specification. 
 
 Before the Rails are put before the representative of the Engineer (or of the Purchaser) 
 for inspection, the Manufacturer shall have them examined, and all Rails which he admits to be 
 defective shall be sorted out and placed in a separate stack; the representative of the Engineer 
 (or of the Purchaser) being empowered to refuse to inspect any lot of Rails not put in uniform 
 lengths and sorted. 
 
 19. The Manufacturer shall supply the material required for testing free of charge and shall, Testing 
 at his own cost, furnish and prepare the necessary test pieces, and supply labor and appliances Facilities, 
 for such testing as may be carried out on his premises in accordance with this Specification. Failing 
 facilities at his own works for making the prescribed tests, the Manufacturer shall bear the cost 
 
 of carrying out the tests elsewhere. 
 
 20. All Rails accepted by the representative of the Engineer (or of the Purchaser) shall be Marking of 
 stamped in his presence. Accepted Rails 
 
 38. Specifications for Street Railway Rails 
 
 American Society for Testing Materials, Affiliated with the International Association for Testing 
 Materials. — Standard Specifications for Open-hearth Steel Girder and High Tee Rails. Adopted 
 June 1, 1912. 
 
 I. Manufacture 
 
 1. The steel shall be made by the open-hearth process. The entire process of manufac- Process, 
 ture and testing shall accord with the best current practice. 
 
 2. Bled ingots, and ingots or blooms which show the effects of injurious treatment, shall Bled Ingots, 
 not be used. 
 
 3. A sufficient discard from the top of each ingot shall be made at any stage of the manu- Discard, 
 facture to obtain sound rails. When finished rails show piping, they may be cut to shorter lengths 
 
 until all evidence of this is removed. 
 
 II. Chemical Properties and Tests 
 
 4. The steel shall conform to either of the following requirements as to chemical composi- Chemical 
 tion, as specified in the order : Composition. 
 
 Class A. Class B. 
 
 Carbon, per cent 0.60-0.75 0.70-0.85 
 
 Manganese, per cent 0.60-0.90 0.60-0.90 
 
 Silicon, per cent not over . 20 not over . 20 
 
 Phosphorus, per cent not over 0.04 not over 0.04 
 
 5. To determine whether the material conforms to the requirements specified in Section 4, ^ a ^ e 
 an analysis shall be made by the manufacturer from a test ingot taken during the pouring of 
 
 each melt. Drillings for analysis shall be taken not less than § inch beneath the surface of the 
 test ingot. A copy of this analysis shall be given to the purchaser or his representative. 
 
 6. A Check analysis may be made from time to time by the purchaser from a test ingot or Check 
 drillings therefrom furnished by the manufacturer. Analyses. 
 
STEEL RAILS 
 
 Drop Tests. 
 
 Test 
 Specimens. 
 
 Number 
 of Tests. 
 Retests. 
 
 III. Physical Properties and Tests 
 7. (a) The test specimen shall be tested on a drop-test machine of the type recommended 
 by the American Railway Engineering Association. The specimen shall be placed head upwards 
 on the supports of the machine, and shall not break when tested with one blow in accordance with 
 the following conditions : 
 
 Weight and Height of Rail. 
 
 Rails weighing over 100 lb. per yd. 
 
 and over 7 in. in depth 
 
 Rails weighing 100 lb. or less per 
 
 yd., or 7 in. or less in depth. . . . 
 
 Temperatur 
 
 ul' Specimen; 
 deg. Fahr. 
 
 60-120 
 60-120 
 
 2000 
 2000 
 
 Height of Drop. 
 
 (b) The atmospheric temperature at the time of testing shall be recorded in the test 
 
 report. 
 
 (c) The testing shall proceed concurrently with the operation of the works. 
 
 8. (a) Three rails, each from the top of one of three ingots from each melt, shall be selected 
 
 by the inspector, and a test specimen shall be taken from each of two of these. 
 (b) Drop test specimens shall not be less than 4, nor more than 6 feet in length. 
 
 9. Two drop tests shall be made from each melt. 
 
 10. If the result of the drop test on only one of the two specimens representing the rails in 
 a melt does not conform to the requirements specified in Section 7, a retest on a specimen from 
 the third rail selected shall be made and this shall govern the acceptance or rejection of the rails 
 from that melt. 
 
 IV. Standard Sections, Lengths, and Weights 
 11. (a) The cold templet of the manufacturer shall conform to the specified section as shown 
 in detail on the drawing of the purchaser, and shall at all times be maintained 
 perfect. 
 (6) The section of the rail shall conform as accurately as possible to the templet, and 
 within the following tolerances: 
 
 (1) The height shall not vary more than ^ inch under nor more than -fa inch 
 
 over that specified. 
 
 (2) The over-all width of head and tram shall not vary more than £ inch from 
 
 that specified. Any variation which would affect the gage line more 
 than jV inch will not be allowed. 
 
 (3) The width of base shall not vary more than £ inch under that specified for 
 
 widths less than 6| inches; t\ inch under for a width of 6| inches; and 
 i inch under for a width of 7 inches. 
 
 (4) Any variation which would affect the fit of the splice bars will not be allowed. 
 
 (5) The base of the rail shall be at right angles to the web ; and the convexity 
 
 shall not exceed ¥ z inch. 
 (c) When necessary on account of the type of track construction, and notice to that 
 effect has been given to the manufacturer, special care shall be taken to maintain 
 the proper position of the gage line with respect to the outer edge of the base. 
 12. (a) Unless otherwise specified, the lengths of rails at a temperature of 60° F. shall be 
 60 and 62 feet for those sections in which the weight per yard will permit. 
 
 (b) The lengths shall not vary more than £ inch from those specified. 
 
 (c) Shorter lengths, varying by even feet down to 40 feet, will be accepted to the extent 
 
 of 10 per cent by weight of the entire order. 
 
RAIL SPECIFICATIONS 493 
 
 13. (a) The weight of the rails per yard as specified in the order shall be maintained as Weight. 
 
 nearly as possible after conforming to the requirements specified in Section 11. 
 (6) The total weight of an order shall not vary more than 0.5 per cent from that specified, 
 (c) Payments shall be based on actual weights. 
 
 V. Workmanship and Finish 
 
 14. (a) Rails on the hot beds shall be protected from water or snow, and shall be carefully Straightening. 
 
 manipulated to minimize cold straightening. 
 (b) The distance between the rail supports in the cold-straightening presses shall not 
 be less than 42 inches, except as may be necessary near the ends of the rails. The 
 gag shall have rounded corners to avoid injury to the rails. 
 
 15. (a) Circular holes for joint bolts, bonds, and tie rods shall be drilled to conform to the Drilling and 
 
 drawings and dimensions furnished by the purchaser. Punching. 
 
 (b) In Class A rails the tie-rod holes may be punched. 
 
 16. The ends shall be milled square laterally and vertically, but the base may be undercut Milling. 
 
 sV inch. 
 
 17. (a) Rails shall be smooth on the head, straight in line and surface without any twists, Finish. 
 
 waves, or kinks, particular attention being given to having the ends without kinks 
 or drop. 
 
 (b) All burrs or flow caused by drilling or sawing shall be carefully removed. 
 
 (c) Rails shall be free from gag marks and other injurious defects of cold-straightening. 
 
 VI. Classification of Rails 
 
 18. Rails which are free from injurious defects and flaws of all kinds shall be classed as No. 1 No. i Rails. 
 
 Rails. 
 
 19. (a) Rails which are rough on the head or which by reason of surface or other imper- No. 2 Rails. 
 
 fections are not classed as No. 1 rails, shall be classed as No. 2 rails; providing 
 they do not, in the judgment of the inspector, contain imperfections in such num- 
 ber and of such character as to render them unfit for No. 2 rail uses, and pro- 
 viding they conform to the requirements specified in Section 11. 
 
 (6) Rails which have flaws in the head exceeding \ inch in depth, or in the base exceed- 
 ing § inch in depth, shall not be classed as No. 2 rails. 
 
 (c) No. 2 rails will be accepted to the extent of 10 per cent by weight of the entire order. 
 
 VII. Marking and Loading 
 
 20. (a) The name or brand of the manufacturer, the year and month of manufacture, the Marking. 
 
 letters "O. H.," the weight of the rail, and the section number shall be legibly 
 rolled in raised letters and figures on the web. The melt number shall be legibly 
 stamped on each rail where it will not be covered subsequently by the joint plates. 
 
 (b) Both ends of all short-length No. 1 rails shall be painted green. 
 
 Both ends of all No. 2 rails shall be painted white and shall have two heavy center- 
 punch marks on the web at each end at such a distance from the end that they 
 will not be covered subsequently by the joint plates. 
 
 21. (a) Rails shall be loaded in the presence of the inspector, and shall be handled in such Loading. 
 
 a manner as not to bruise the flanges or cause other injuries. 
 (6) Rails of each class shall be placed together in loading. 
 
 (c) Rails shall be paired as to length before shipment. 
 
 VIII. Inspection 
 
 22. The inspector representing the purchaser shall have free entry, at all times while work Inspection, 
 on the contract of the purchaser is being performed, to all parts of the manufacturer's works 
 
494 STEEL RAILS 
 
 which concern the manufacture of the material ordered. The manufacturer shall afford the 
 inspector, free of cost, all reasonable facilities to satisfy him that the material is being furnished 
 in accordance with these specifications. All tests and inspection shall be made at the place of 
 manufacture prior to shipment, and shall be so conducted as not to interfere unnecessarily with 
 the operation of the works. 
 
 39. Bibliography of Rail Specifications 
 
 Search furnished by the Secretary of the American Society of Civil Engineers, and made 
 in its Library, January 24, 1910, and January 19, 1912, supplemented by the Technology 
 department of the Carnegie Library of Pittsburgh. 
 
 1877 
 " Permanent-way Rolling Stock and Technical Working of Railways," Vol. 1, p. 512, by 
 Ch. Couche, tr. by James N. Shoolbred. Paris, 1877. Dunod, 49 Quai des Augustins. (Contains 
 specifications for chemical composition of steel rails.) 
 
 1879 
 
 " The Chemical Composition and Physical Properties of Steel Rails," by C. B. Dudley. 
 Trans. Am. Inst. Min. Engrs., Vol. 7, p. 172 (1879). (Recommends a formula for the chemical 
 composition of rails for the use of the Pennsylvania Railroad.) 
 
 " Does the Wearing Power of Steel Rails Increase with the Hardness of Steel? " by Charles 
 B. Dudley. Trans. Am. Inst. Min. Engrs , Vol. 7, p. 202 (1879) (four pages). 
 
 " Discussion of Dr. Charles B. Dudley's Papers on Steel Rails." Trans. Am. Inst. Min. 
 Engrs., Vol. 7, p. 357 (1879) 
 
 1880 
 
 " The Wearing Capacity of Steel Rails in Relation to their Chemical Composition and 
 Physical Properties," by Charles B. Dudley. Trans. Am. Inst. Min. Engrs., Vol. 8, p. 321 (1880). 
 (Contains references to specifications for chemical composition.) 
 
 " Specifications for Steel Rails and Track Fastenings." Eng. News, Vol. 20, p. 172 (Sept. 
 1, 1888). (Specifications drawn up and used by Frank Ward & Bro., of Pittsburg, Pa.) 
 
 Same. R. R. Gaz., Vol. 20, p. 587 (Sept. 7, 1888). 
 
 " Steel Rails and Specifications for their Manufacture," by Robert W. Hunt. Trans. Am. 
 Inst. Min. Engrs., Vol. 17, p. 226 (1888). (Contains specifications for Bessemer steel rails.) 
 
 Abstracts of same. R. R. Gaz., Vol. 20, p. 697 (Oct. 26, 1888); Eng. and Min. Jour., 
 Vol. 46, p. 370 (Nov. 3, 1888). 
 
 1893 
 
 " Steel Rails: their Manufacture and Service." Eng. News, Vol. 30, p. 172 (Aug. 31, 1893). 
 (Proposed specifications for steel rails.) 
 
 1895 
 
 " Specifications for Steel Rails of Heavy Sections Manufactured West of the Alleghenies," 
 by Robert W. Hunt. Trans. Am. Inst. Min. Engrs., Vol. 25, p. 653 (1895). (The author states 
 that " the only important features in which the present specifications differ from those of 1888 is 
 in providing for a chemical composition and for drop tests.") 
 
 1897 
 " Brief Note on Rail Specifications," by Robert W. Hunt. Trans. Am. Inst. Min. Engrs., 
 Vol. 27, p. 139 (1897). (One page; report of progress.) 
 
RAIL SPECIFICATIONS 495 
 
 1899 
 " Specifications on Structural Steel and Rails," by W. R. Webster. Journal of the Franklin 
 Institute, Vol. 147, p. 1 (Jan., 1899). (General discussion of the subject.) 
 
 1900 
 
 Proceedings American Railway Engineering and Maintenance of Way Association, Vol. 1, 
 p. 116 (1900). (Refers to specifications for steel rails.) 
 
 " Recent Practice in Rails: an Informal Discussion." Trans. Am. Socy. of Civ. Engrs., 
 Vol. 44, p. 489 (Paper 887, Dec, 1900). (Gives the standard rail specifications of the Louisville 
 & Nashville R. R. Co., Robert W. Hunt's specifications, and specifications of the Western rail 
 mills, and a review of foreign rail specifications.) 
 
 " American Standard Specifications and Methods of Testing Iron and Steel," by Albert 
 Ladd Colby. Journal of the Iron and Steel Institute, Vol. 158, p. 215 (1900). (Gives specifi- 
 cations for steel rails.) 
 
 1901 
 
 " Proposed Standard Specifications for Steel Rails." In Proc. Am. Ry. Eng. and M. of 
 W. Assn., Vol. 2, p. 192 (1901). (Specifications recommended by Committee No. 1 of the Ameri- 
 can Section of the International Association for Testing Materials.) 
 
 " Nature of Metal for Rails, Report (United States)," by P. H. Dudley. International 
 Railway Congress. Proceedings, Sixth Session, 1900, Vol. 1, Question 1, pp. 205, 248. Brussels, 
 1901. P. Weissenbruch, 49 Rue du Poincon. (Gives specifications for steel rails.) 
 
 " Examen des Specifications Normales Americaines Proposees Eprouvettes et Methodes 
 d'Essai du Fer et de l'Acier," by Albert Ladd Colby. In " Communications Presentees devant 
 le Congres International des Methodes d'Essai des Materiaux de Construction tenu a Paris du 
 9 au 16 Juillet, 1900," Vol. 2, Pt. 1, pp. 147, 162. Paris, 1901. Vve. Ch. Dunod, 49 Quai des 
 Grands-Augustins. (Contains a review of foreign specifications for steel rails and proposed 
 American standard specifications.) 
 
 " Specifications for Steel Rails," by P. H. Dudley. R. R. Gaz., Vol. 33, p. 158 (March 8, 
 1901) (one page). 
 
 " Some Suggestions as to Specifications for Steel Rails," by E. F. Kenney. Eng. News, 
 Vol. 46, p. 226 (Oct. 3, 1901). 
 
 1902 
 
 American Society for Testing Materials, Proceedings, Vol. 1, pp. 101, 264 (1899-1902). 
 (Proposed standard specifications for steel rails recommended by American Branch of Committee 
 No. 1, American Section of the International Association for Testing Materials. 
 
 " Review and Text of the American Standard Specifications for Steel, adopted in August, 
 1901," p. 41, by Albert Ladd Colby. Ed. 2. Easton, Pa., 1902. The Chemical Publishing Co. 
 (Contains specifications for steel rails.) 
 
 " Proposed Modifications of the Standard Specifications for Steel Rails, Topical Discussion." 
 Proc. Am. Socy. for Testing Materials, Vol. 2, p. 9, 23 (1902). 
 
 Proceedings of the Am. Ry. Eng. and M. of W. Assn., Vol. 3, p. 201 (1902). (Specifications 
 recommended in 1901 with some amendments.) 
 
 " Specifications for Steel Rails/' by W. R. Webster, Trans. Am. Inst. Min. Engrs., Vol. 31, 
 pp. 449, 967 (1902). (Contains proposed standard specifications recommended May, 1900, by 
 the American Branch of Committee No. 1 of the International Association for Testing Materials.) 
 
 " Steel Rails: Specifications," by Robert Job. Am. Eng. and R. R. Jour., Vol. 76, p. 310 
 (Oct., 1902). (Gives specifications of the Philadelphia & Reading Railway Company for steel rails.) 
 
 1903 
 " The Present Situation as to Specifications for Steel Rails," by William R. Webster. Trans. 
 Am. Inst. Min. Engrs., Vol. 33, p. 164 (1903) (five pages). 
 
496 STEEL RAILS 
 
 "Proposed Modifications in the Specifications for Steel Rails adopted by the American 
 Railway Engineering and Maintenance of Way Association in March, 1903." Proc. Am. Socy. 
 for Testing Materials, Vol. 3, p. 74 (1903). 
 
 1904 
 
 " British Standard Specification and Sections of Bullheaded Railway Rails." Engineering 
 Standards Committee. Report No. 9. Lond., 1904. (10s. 6d. net.) 
 
 " Specifications for Steel Rails of the American Railway Engineering and Maintenance of 
 Way Association, as Amended and Adopted in March, 1904," with Introduction by William R. 
 Webster. Proc. Am. Socy. for Testing Materials, Vol. 4, p. 195 (1904). (Showing main difference 
 in specifications adopted by the two societies.) 
 
 " Standard Specifications for Bessemer Steel Rails." In Proc. Am. Ry. Eng. and M. of 
 W. Assn., Vol. 5, p. 465 (1904). 
 
 " Specifications for Bessemer Steel Rails." Eng. News, Vol. 50, p. 275 (March 24, 1904). 
 (Gives specifications adopted by the American Railway Engineering and Maintenance of Way 
 Association.) 
 
 1905 
 
 " Railroad Construction," p. 243; by Walter Loring Webb, M. Am. Socy. C. E. Ed. 3. 
 N. Y., 1905, John Wiley & Sons, 43 E. 19th St. (Contains proposed standard specifications for steel 
 rails of the American Railway Engineering and Maintenance of Way Association, March, 1902.) 
 
 British Standard Specification and Sections of Flat-bottomed Railway Rails. Engineering 
 Standards Committee. Report No. 11. Lond., 1905. Leslie S. Robertson, Secy., 28 Victoria 
 Street, Westminster, S. W. (10s. 6d. net.) 
 
 " Steel Rails," by William R. Webster. R. R. Gaz., Vol. 38, p. 440 (May 5, 1905). (Gives 
 specifications of the American Railway Engineering and Maintenance of Way Association.) 
 
 1906 
 
 " On Specifications for Steel Rails." Proc. Am. Socy. for Testing Materials, Vol. 6, p. 35 
 (1906). (Gives proposed standard specifications for steel rails.) 
 
 Proc. Am. Ry. Eng. and M. of W. Assn. Vol. 7, p. 553 (1906). (Gives comparison of 
 specifications of American Railway Engineering and Maintenance of Way Association and the 
 American Society of Civil Engineers.) 
 
 " Rails for Lines with Fast Trains." Reports; by P. H. Dudley and Van Bogaert. Inter- 
 national Railway Congress, Seventh Session, 1905, Vol. 1, Question 2, pp. 141, 194. Brussels, 
 1906. P. Weissenbruch. (Contains very brief data on rails specifications.) 
 
 " Specifications for Steel Rails." R. R. Gaz., Vol. 40, p. 280 (March 16, 1906). (Specifi- 
 cations recommended by the American Railway Engineering and Maintenance of Way Association, 
 American Society for Testing Materials, and American Society of Civil Engineers.) 
 
 1907 
 
 " Report of the Special Committee on Rail Sections to the American Society of Civil En- 
 gineers." Proc. Am. Socy. of Civ. Engrs., Vol. 32, p. 52; Vol. 33, p. 290 (1906, 1907). (Contains 
 recommended specifications for Bessemer steel rails.) 
 
 " Manual of Recommended Practice for Railway Engineering and Maintenance of Way," 
 p. 55. Am. Ry. Eng. and M. of W. Assn. Edition of 1907. (Contains specifications for rails.) 
 
 " Proposed Standard Specifications for Steel Rails." Proc. Am. Socy. for Testing Materials, 
 Vol. 7, p. 40 (1907). (Specifications adopted Sept. 1, 1907.) 
 
 " The Steel-rail Discussion, American Society for Testing Materials." Ry. and Eng. 
 Review, Vol. 47, p. 570 (June 29, 1907). 
 
 " Proceedings of the Session of the American Railway Association, October 30, 1907," p. 175, 
 N. Y., 1907. (Report of the Committee on Rail Sections giving specifications for Bessemer 
 steel rails with explanatory notes.) 
 
RAIL SPECIFICATIONS 497 
 
 " Standard Specifications for Steel Rails." Proc. Am. Soc. for Testing Materials, Vol. 7, 
 p. 44 (1907). 
 
 Same. Eng. Record, Vol. 55, p. 774 (June 29, 1907). 
 
 Comparison of American and Foreign Rail Specifications, with a Proposed Standard Speci- 
 fication to Cover American Rails for Export," by Albert Ladd Colby. Trans. Am. Inst. Min. 
 Engrs., Vol. 37, p. 576 (1907). (Contains bibliography.) 
 
 Same. Iron and Coal Trades Review, Vol. 73, p. 357 (July 27, 1906). 
 
 " Rail Sections and Specifications." R. R. Gaz., Vol. 43, p. 250 (Sept. 6, 1907). (Com- 
 pares rail specifications of the American Society of Civil Engineers, American Railway Engineering 
 and Maintenance of Way Association, American Society for Testing Materials.) 
 
 " Some Progress Toward Getting Better Rails." (Editorial.) R. R. Gaz., Vol. 43, p. 577 
 (Nov. 15, 1907). (Brief discussion of rail specifications.) 
 
 " Proposed Standard Rail Sections of the American Railway Association." R. R. Gaz., 
 Vol. 43, p. 627 (Nov. 22, 1907) (illustrated). 
 
 " Rail Specifications." R. R. Gaz., Vol. 43, p. 735 (Dec. 20, 1907). (Contains specifica- 
 tions of the American Railway Association.) 
 
 1908 
 
 " Railway Track and Track Work," p. 78; by E. E. Russell Tratman, Assoc. M. Am. 
 Socy. C. E. Ed. 3. N. Y., 1908. Engineering News Publishing Co., 220 Broadway. $3.50 
 net. (Contains a comparative table of specifications for chemical composition of rails.) 
 
 Proceedings of the American Railway Association, Special Session, February 7, 1908; 
 Regular Session, April 22, 1908, p. 359, N. Y., 1908. (Specifications for Bessemer and open- 
 hearth steel rails, accompanying the Report of the Committee on Standard Rail and Wheel Sec- 
 tions, dated March 23, 1908.) 
 
 " Standard Specifications for Steel Rails." Proc. Am. Soc. for Testing Materials, Vol. 8, 
 p. 44 (1908). (Specifications adopted Aug. 15, 1908.) 
 
 " The Present Status of Rail Specifications." R. R. Age, Vol. 45, p. 76 (Jan. 17, 1908). 
 (A review of the action taken by the American Railway Association.) 
 
 American Society of Civil Engineers, Report of Special Committee on Rail Sections. Eng. 
 News, Vol. 59, p. 105 (Jan. 23, 1908). (Recommended specifications for Bessemer steel rails.) 
 
 " New Steel-rail Specifications of the Pennsylvania Railroad." Eng. News, Vol. 59, p. 426 
 (April 16, 1908). (Gives specifications for chemical composition, process of manufacture, mechani- 
 cal requirements, tests, and inspection.) 
 
 " The Pennsylvania New Rail Sections and Specifications." R. R. Gaz., Vol. 44, p. 539 
 (April 17, 1908). 
 
 " New Rail Sections and Rail Specifications of the American Railway Association." Eng. 
 News, Vol. 59, p. 530 (May 14, 1908). (Specifications for Bessemer and open-hearth steel rails.) 
 
 "American Railway Association's Rail Committee." (Editorial.) Eng. News, Vol. 59, 
 p. 533 (May 14, 1908). (Comments on the rail specifications; one and a half columns.) 
 
 " Steel-rail Breakages; Questions of Design and Specifications," by Harold V. Coes. En- 
 gineering Magazine, Vol. 35, p. 417 (June, 1908). (Gives specifications for the Union and Southern 
 Pacific railways and British standard chemical specifications for steel rails.) 
 
 " Some Features of the Present Steel Rail Question," by Charles B. Dudley, Proc. Am. Soc. 
 for Testing Materials, Vol. 8, p. 19 (1908). (Discusses changed demands on steel rails and pro- 
 posed specifications.) Same. Engineering News, Vol. 60, p. 9. 
 
 1909 
 " Proceedings of the Session of the American Railway Association held in Chicago, Novem- 
 ber 17, 1909," p. 995. N. Y., 1909. W. F. Allen, Secy., 24 Park Place. (American Railway 
 Association specifications for Bessemer and for open-hearth steel rails, adopted as recommended 
 practice April 22, 1908.) 
 
498 STEEL RAILS 
 
 Proceedings Am. Ry. Eng. and M. of W. Assn., Vol. 10, Pt. 1, pp. 369, 374 (1909). (Recom- 
 mended changes in specifications as previously adopted by the Association.) 
 
 " New Rail Section and Specifications, Canadian Pacific Ry." Ry. and Eng. Review, Vol. 
 49, p. 27 (Jan. 9, 1909). (Gives specifications for open-hearth and Bessemer rails.) 
 
 " New Rails for the Canadian Pacific Ry." (Editorial.) Ry. and Eng. Review, Vol. 49, 
 p. 34 (Jan. 9, 1909). (Discusses specifications and rail sections.) 
 
 " New Rail Specifications of the Pennsylvania R. R. System." Eng News, Vol. 61, p. 50 
 (Jan. 14, 1909). (Revision of specifications of Feb. 4, 1908.) 
 
 " Pennsylvania Rail Specifications." R. R. Age Gaz., Vol. 46, p. 101 (Jan. 15, 1909). 
 (Specifications of the Pennsylvania Railroad evised under date of Dec. 10 1908.) 
 
 " The New 85-pound Rail Section of the Canadian Pacific Ry." Eng. News, Vol. 61, 
 p. 272 (March 11, 1909) (illustrated). 
 
 " Recent Developments in Rail Design and a Comparison of Rail Sections." (Editorial.) 
 Eng. News, Vol. 61, p. 276 (March 11, 1909). (Compares rail specifications.) 
 
 " Recent Rail Sections." R. R. A e Gaz., Vol. 46, p. 537 (March 19, 1909) (one page, 
 illustrated). 
 
 " New Rail Orders and Specifications." (Editorial.) R. R. Age Gaz., Vol. 46, p. 535 (March 
 19, 1909). (Discusses rail specifications of various railroads.) 
 
 " Rail Specifications." (Editorial.) R. R. Age Gaz., Vol. 46, p. 925 (April 30, 1909). (Very 
 brief.) 
 
 " Rail Specifications " (letter), by R. Trimble. R. R. Age Gaz., Vol. 46, p. 1018 (May 14, 
 1909). (Brief letter correcting error in the above editorial.) 
 
 " Comparative Rail Specifications." R. R. Age Gaz., Vol. 46, p. 1066 (May 21, 1909). 
 (Compares specifications of the American Railway Association, Steel Manufacturers of America, 
 American Society of Civil Engineers, American Railway Engineering and Maintenance of Way 
 Association, and American Society for Testing Materials, with comments.) 
 
 " Rail Sections and Specifications." (Editorial.) R. R. Age Gaz., Vol. 46, p. 1060 (May 21, 
 1909). 
 
 " Specifications for 90-pound Bessemer and Open-hearth Steel Rails for the Harriman 
 Lines." R. R. Age Gaz., Vol. 47, p. 185 (July 30, 1909). (Specifications to which the Harriman 
 Lines are ordering their 1909 rails. 
 
 " On the Question of Strengthening the Track and the Bridges with a View to Increasing 
 the Speed of Trains Subject II, for Discussion at the Eighth Session of the Railway Congress," 
 by M. L. Byers. Bulletin of the International Railway Congress Association, Vol. 23, p. 908 
 (Sept., 1909). (Gives rail specifications proposed by he American Railway Association and by 
 the Pennsylvania Railroad Committee.) 
 
 " Report of Committee on Rail, American Railway Engineering and Maintenance of Way 
 Association. Bulletin 118 (Dec, 1909). (Specifications for steel rails and review of previous 
 reports.) 
 
 Abstract of same. " Specifications for Steel Rails." Railway and Engineering Review, 
 Vol. 50, p. 118 (Feb. 5, 1910). 
 
 " Standard Specifications for Bessemer Steel Rails." Proceedings American Society for 
 Testing Materials, Vol. 9, p. 62 (1909). (Adopted Aug. 16, 1909.) 
 
 " Standard Specifications for Open-hearth Steel Rails." Proceedings American Society for 
 Testing Materials, Vol. 9, p. 66 (1909). (Adopted Aug. 16, 1909.) 
 
 " La Voie Courante des Chemins de Fer de l'Etat Beige," by Pierre Decamps. Revue Gen- 
 eral des Chemins de Fer et des Tramways, Vol. 32, Pt. 2, p. 267 (Oct., 1909). (Appendix gives 
 rail specifications of the state railroad of Belgium.) 
 
 " Revised Rail Specifications, Pennsylvania Railroad System." Engineering, Vol. 87, 
 p. 218. 
 
 British Standard Sections, No. 47. Engineering Standards Committee (1909). (British 
 standard specifications for bull headed and flat bottom railway rails.) 
 
RAIL SPECIFICATIONS 499 
 
 " Specifications for Standard Open Hearth Steel Rails for A. S. C. E. Sections," Carnegie 
 Steel Co., Jan. 1, 1909. (Two leaflets.) 
 
 " Specifications for Standard Bessemer Steel Rails for A. S. C. E. Sections," Carnegie Steel 
 Co., Jan. 1, 1909. (Two leaflets.) 
 
 " Specifications for Steel Rails." Baltimore and Ohio R. R. Co., No. 163C, Jan. 25, 1909. 
 (Two leaflets.) 
 
 " Specifications for Open Hearth Steel Rails." Proceedings American Street and Inter- 
 urban Railway Engineering Association, Vol. 7, p. 59 (1909). (Includes specifications adopted 
 by the Transit Supply Co., Lorain Steel Co., Pennsylvania Steel Co., and the Manganese Steel 
 Rail Co.) 
 
 1910 
 
 " Permanent Way." R. R. Engr., Vol. 31, p. 18 (Jan., 1910). (Gives Pennsylvania Railroad 
 System specifications for steel rails.) 
 
 " Final Report of Special Committee on Rail Sections." Transactions American Society of 
 Civil Engineers, Vol. 70, p. 456 (Paper 1177, Dec, 1910). (Contains reprint of rail specifica- 
 tions of the American Railway Engineering Association.) 
 
 " The American Railway Association, The American Railway Engineering and Maintenance 
 of Way Association, Specifications for Steel Rails." Proceedings American Railway Engineering 
 and Maintenance of Way Association, Vol. 11, Pt. 1, p. 254 (1910). 
 
 Abstracts of same. "Specifications for Steel Rails." Railway and Engineering Review, 
 Vol. 50, p. 118 (Feb. 5, 1910). "Rail Specifications and Sections," Engineering News, Vol. 63, 
 p. 384 (Mar. 31, 1910). 
 
 " Track Standards and General Rules." Department of Maint. of Way, Metropolitan 
 Street Railway Co. Elec. Ry. Journal, Vol. 35, p. 863. 
 
 " Recent Work of the German Street and Interurban Railway Association." Elec. Ry. 
 Journal, Vol. 35, p. 38. (Considers specifications and standards agreed upon.) 
 
 Hunt (Robert W.) & Co., Engineers. Bureau of Inspection, Tests and Consultation. 
 (Includes "Specifications for Standard Open-hearth Steel Girder and High Tee-Rails," 1910, 
 American Street and Interurban Railway Engineering Association, p. 5; and "Specifications for 
 Standard Open-hearth Steel Girder and High Tee-Rails," Lorain Steel Co., Jan. 1, 1910, p. 14.) 
 
 1911 
 
 " Standard Specifications for Bessemer and Open-hearth Steel Rails." March 21, 1910, 
 United States Steel Products Export Co. (Year-book, American Society for Testing Materials, 
 1911, p. 202.) 
 
 " Rail Sections and Specifications." Elec. Ry. Journal, Vol. 37, p. 8. (Editorial, dis- 
 cussing progress toward uniform specifications in 1910.) 
 
 " Interborough Rails for Tangents and Curves." Elec. Ry. Journal, Vol. 37, p. 82. (Gives 
 recent modifications of specifications of open-hearth steel rails.) 
 
 Same, abstract, Journal of the Iron and Steel Inst., Vol. 84, p. 619. 
 
 " Manufacturers' Standard Specifications for Bessemer Steel Rails," Association of American 
 Steel Manufacturers. Year-book, American Society of Testing Materials, 1911, p. 199. 
 
 " Specifications for Basic Open-hearth Rails," New York Central Lines. (Specifications of 
 Oct. 1, 1909, revised Jan. 11, 1911, to conform to manufacturers' sale per 100 pounds.) 
 
 " Specifications." Report of Committee on Rail. Proceedings American Railway En- 
 gineering and Maintenance of Way Association, Vol. 12 (1911), Pt. 2, p. 12. (Gives short report 
 of progress.) 
 
 Report of Committee A-l. Proceedings American Society for Testing Materials (1911), 
 Vol. XI, p. 48. (Contains reference to international specifications for rails.) 
 
500 STEEL RAILS 
 
 1912 
 
 " Specifications for Carbon Steel Rails." Proceedings American Railway Engineering As- 
 sociation (1912), Vol. 13, p. 565. 
 
 " New Specifications for Steel Rails." Iron Age, Vol. 89, p. 816. (Gives report of rail com- 
 mittee at 1912 meeting of the American Railway Engineering Association and specifications 
 adopted.) 
 
 " Specifications for 85-pound and 100-pound Carbon Steel Rails," 1912, Pennsylvania Rail- 
 road Company. (Two leaflets.) 
 
 " Specifications for Standard Bessemer Steel Tee Rails," 1912 Catalogue, Maryland Steel 
 Company, p. 10. 
 
 " Specifications for Standard Open-hearth Steel Tee Rails," 1912 Catalogue, Maryland 
 Steel Company, p. 12. 
 
 " Specifications for Standard Open-hearth Steel Girders and High Tee Rails," 1912 Cata- 
 logue, Pennsylvania Steel Company, p. 14. 
 
APPENDIX 
 
 REPORTS AND RECORDS 
 
 The forms recommended by the Rail Committee of the American Railway 
 Engineering Association, and contained in the 1911 Manual of the Association, 
 are typical of the best practice, and are shown on Figs. 342 to 359 inclusive 
 and Plate XXXIII. The explanation of the forms as given by the committee 
 is as follows: 
 
 Group I. Reports of Rail Inspection and Shipment at the Mill 
 This set of forms, Figs. 342-344 and Plate XXXIII, is for the use of the 
 railroad company's Inspector at the mills where the rail is rolled, and gives 
 all the information necessary to inform the purchaser that his order has been 
 manufactured in accordance with the specifications and shipped. 
 

 A. B. & C. R. R. 
 
 Report of Chemical and 
 
 Examination 
 
 Co. 
 
 Physica 
 
 ' 1 Sheet No. 
 
 lof Shells' 
 
 
 
 
 For _ 
 Order 
 No. of 
 
 No. _ 
 
 
 
 
 
 
 
 Shrinkage A lowance at Saws inches on 
 
 33-ft. rails 
 
 
 
 Weight of Tup, 2000 lbs. Height of Dr 
 Average Number of Rails per Heat 
 
 
 
 
 
 
 
 Heat No. 
 
 P ™ g ln g ot CrOP 
 
 t 
 
 
 I 
 1 
 
 | 
 
 J 
 
 
 
 §■11 
 
 _. 
 
 Egote 
 
 Top 
 
 e'ciZ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9 
 
 
 
 
 
 
 
 
 
 
 
 
 
 10 
 
 For 
 
 lling in 
 
 vith ty 
 
 ewriter 
 
 column 
 
 si d 
 
 b, ? m 
 
 cd in te 
 
 ths < 
 
 
 
 ., ,|,..|, 
 
 ? nres 11 
 
 7 
 
 4 
 
 5 
 
 5 
 
 5 
 
 S 
 
 5 
 
 5 
 
 5 
 
 5 
 
 5 
 
 ■5 
 
 II 12 
 
 
 
 
 
 
 sheet i 
 
 qmred 
 
 SxlOj i 
 
 chcs. 
 
 
 
 
 13 
 
 
 
 
 
 
 
 
 
 
 
 
 
 14 
 
 
 
 
 
 
 
 
 
 
 
 
 
 15 
 
 
 
 
 
 
 
 
 
 
 
 
 
 16 
 
 
 
 
 
 
 
 
 
 
 
 
 
 17 
 
 
 
 
 
 
 
 
 
 
 
 
 
 18 
 
 
 
 
 
 
 
 
 
 
 
 
 
 19 
 
 
 
 
 
 
 
 
 
 
 
 
 
 20 
 
 
 
 
 
 
 
 
 
 
 
 
 
 21 
 
 
 
 
 
 
 
 
 
 
 
 
 
 22 
 
 
 
 
 
 
 
 
 
 
 
 
 
 23 
 
 
 
 
 
 
 
 
 
 
 
 
 
 24 
 
 
 
 
 
 
 
 
 
 
 
 
 
 25 
 
 
 
 
 
 
 
 
 
 
 
 
 
 26 
 
 
 
 
 
 
 
 
 
 
 
 
 
 27 
 
 
 
 
 
 
 
 
 
 
 
 
 
 28 
 
 
 
 
 
 
 
 
 
 
 
 
 
 29 
 
 
 
 
 
 
 
 
 
 
 
 
 
 30 
 
 
 
 
 
 
 
 
 
 
 
 
 
 31 
 
 
 
 
 
 
 
 
 
 
 
 
 
 32 
 
 Note — Requirements of Standard Specifications are to be sta 
 Instructions 
 One copy of this report to be forwarded to Correct 
 the Chief Engineer M of W 
 
 Approved ■ 
 
 ted on line 1 
 
 Inspector 
 
 
 
 
 
 
 
 
 
 
 \^TZ t , 
 
 Fig. 342. 
 
 M. W. 401. — Report of Chemical and Physical Examination: 
 This blank is filled out from the mill records under the supervision of the 
 Inspector, and gives the chemical contents taken from the ladle analysis and 
 the result of the drop test. (503) 
 
A. B. & C. R. R. Co. 
 
 No._ 
 
 Certificate: of- inspection 
 
 of Process Rails lbs. per yd Section. 
 
 Manufactured by Steel Co. at Works 
 
 For 
 
 Mr 
 
 Chief Engineer M. of W. Date_. 
 
 The following Steel Rails have been inspected and accepted accordir 
 s are certified to be within the limits of the Specifications of the 
 
 ._19-. 
 
 and approved as per details given below. 
 
 All Rails have been inspected and approved for Chemical Analysis, Physical Tests, Section, 
 Weight, Straightening. Drilling, Sawing, Length, Stamping, Finish, Quality. 
 
 All Rails are marked on the web with maker's name, date of manufacture. Heat Number, 
 
 and position occupied in the ingot. Date of Rolling 
 
 No. of Rails Rolled No. of Rails Accepted 
 
 No. of Rails temporarily rejected and cause , -x&-4 
 
 *</ 
 
 __j,€^i._ 
 
 No. of Rails condemned a: 
 
 
 This Certificate covers the run from #" 
 
 Heat No to Heat No both inch 
 
 Calculated Weight. 
 
 Tons Pounds 
 
 Shipper's Scale Weight. 
 
 Amount accepted under this Certificatf 
 
 Total amount of Order 
 
 Balance due on Order 
 
 One copy of this Certificate is to be madi 
 out and forwarded to the Chief Enginee 
 M. of W. of the Railway Company. 
 
 
 Fig. 343. 
 
 M. W. 402. — Certificate of Inspection: 
 
 This is the Inspector's written statement that the material which he has 
 witnessed rolled has been turned out strictly in accordance with the specifications 
 and the order of the railroad company. (504) 
 
2 
 3 
 
 5 
 6 
 7 
 8 
 9 
 10 
 
 14 
 
 8 15 
 16 
 17 
 
 20 
 21 
 22 
 23 
 24 
 25 
 26 
 27 
 
 29 
 
 31 
 32 
 
 of 
 
 
 A. B. & C. R. R. Co. 
 
 No 
 
 Report of Shipment. 
 
 Process Rails lbs. per yd Section 
 
 
 
 Consigne 
 Order N 
 Quality 
 
 d to ... 
 
 . Date of Report 19.. 
 
 ■Jo SheetNo of Sheets. 
 
 Loaded on Cars. 
 
 Numbei of Rails of each Length. 
 
 
 Initial 
 
 No 
 
 
 33 
 
 30 
 
 
 27J 
 
 
 25 
 
 
 RalS 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fo. fill 
 
 
 ypewntc 
 
 
 
 d be spa 
 
 ced in t 
 
 nihs. 
 
 
 *■»■ 
 
 i by thJ 
 
 figures. 
 
 9 
 
 9 
 
 5 
 
 ■5 
 
 5 
 
 ■5 
 
 5 
 
 5 
 
 5 
 
 s 
 
 6 
 
 8 
 
 
 
 
 S,z 
 
 ■„l si, 
 
 et rcqu 
 
 ,,i ,.| 
 
 j inche 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Totals 
 
 
 
 
 
 
 
 
 
 
 
 Total Weight Expressed in Gross Tons and Decimals. 
 
 
 Total Tons of Order Tons previously shipped 
 
 
 Balance due 
 
 
 INSTRUCTIONS _ 
 
 
 
 Onec 
 Ch.ef E 
 the Gen 
 for the I 
 
 raTsuper 
 ivision Si 
 
 of W. 
 ntende 
 
 as 
 
 WO CO 
 
 
 Ap 
 
 
 wi^s^r 
 
 Fig. 344. 
 
 M. W. 403. — Report of Shipment: 
 
 This blank is used for reporting the number and length of rail shipped in 
 each car from the works, and, when properly checked by the Receiving Officer, 
 it furnishes the basis for payment of the bill. 
 
 M. W. 418. — Results of Drop Tests and Surface Inspection of Rails Rolled 
 (Plate XXXIII). (505) 
 
This form is intended for tabulating the results of drop tests and surface 
 inspection of rails rolled. 
 
 Group II. Reports from Division Officers 
 This group, Figs. 345-347, contains all the regular reports which come from 
 the division officers concerning the rails which have been put in service in track. 
 
 A. B. & C. R. R. Co. 
 
 Report of RAIL FAILURES i 
 Date of Report.. 
 
 . . .Branch. 
 
 i Main Tracks 
 
 Rail Sec 
 
 Brand on Rail? ("D" on back) 
 
 Kind of Steel? ("E" on back) 
 
 Heat No. on Rail? ("F"on back) 
 
 Rail No. or Letter? ("F" on back) 
 
 Original Length of Rail? 
 
 Month and Year Rail was Laid 
 
 Location Feet of Mile 
 
 Post 
 
 Which Track? 
 
 Which Rail? 
 
 On Curve or Straight Line ? 
 
 No. of Curve? 
 
 Degree of Curve? 
 
 High or Low Rail, if on Curve? 
 
 Superelevation of Curve at "Break"?.. 
 
 Was Rail "Broken"? 
 
 or "Defective"? 
 
 or Damaged? 
 
 (See "Description of Failur 
 
 Condition of weather? (Wet, dry, - 
 
 Was Rail much c 
 By whom disc 
 
 little 
 
 ?ered.. 
 
 Date 
 
 Was Rail removed?. 
 Date removed ? . . . , 
 Exact gage of Track 
 Was "Break" over 01 
 Was "Break" square 
 Distance between edges of 
 
 
 Condition of Ties each side of "Break ?' ! 
 Kind of" Tie" 
 
 : Plates used? 
 
 i of Line and Surface?. 
 
 Kind 
 
 Was Track properly ballasted? 
 
 Kind of material in roadbed under bal- 
 
 If "Broken," state 
 
 i of break, and describe any flaws found at point of break 
 
 [f "Break" was at 
 
 bolted or insulated 
 
 Were any bolts at jo. 
 
 Was accident or detent 
 
 joint, state kind, number 
 •. 
 
 of holes, 
 
 and 
 
 whether it was full 
 
 
 
 how 
 
 
 
 
 
 
 
 
 Draw on Diagram lines of "Break," or partial fracture, such as long pieces fror 
 side of head and half-moon pieces from base, showing dimensions. Hollows i 
 head should be shown on "End Section." Defects may also be indicated o 
 Diagram. Mark distance from end to "Break." *If "Break" is nearest "Receh 
 ing End," draw pen through words "Leaving End;" if nearest "Leaving End," 
 
 iA*l"«^S>i*t*M K £ji ra EvaI M 
 
 e if known. (See "Description of Failur 
 
 Approved: 
 
 . . Foreman 
 
 and Description of Failur 
 
 M. W. 404. 
 
 Fig. 345. (Face of Form.) 
 
 - Report of Rail Failures in Main Tracks: 
 
This is the basic report of all rail failures and is sent by the Track Foreman 
 to his Supervisor and by him transmitted to the Division Engineer. It contains 
 a classification of rail failures which is used in the tabulations employed in the 
 following blanks. 
 
 INSTRUCTIONS 
 
 The Foreman will send this Report 
 
 discovered, and in the case of a 
 
 out of the track. 
 The Supervisor will forward this Report direct to the Division Engineer. 
 
 The Division Engineer will have copies of this Report made immediately upon 
 
 receipt and send a copy to the Chief Engineer M. of W. 
 The answer to 3 is in raised characters on the web of the rail. 
 The answer to 4 is "Bessemer" (B) ; "Open-Hearth" (O.H.) ; "Nickel" (N.) ; 
 
 "Ferro-titanium (F.T.) ; "Chrome Nickel" (CN.) ; or other method of manu- 
 facture or alloy. 
 The answers to 5 and 6 are stamped into the metal on side of web — figures 
 
 for 5 and a letter for 6. 
 f South 
 
 G. Mile Post No. fro 
 
 IE 
 
 ■nd of Div 
 
 
 DESCRIPTION OF RAIL FAILURES 
 
 When describing Failures of Rails, the following terms should be used. 
 
 Flow or Metal. This term means a "Rolling Out" of the metal o: 
 of the head towards its sides without there being any indication of a I 
 ing down of the head structure, that is, the under side of the head i 
 distorted. 
 
 Crushed Head. This 
 
 mpanied by a crushin 
 
 e >■ 
 
 2 ± 
 
 Fig. 345 (continued). (Back of Form.) 
 

 
 'ON 3"!1 
 
 
 H 
 
 « 
 
 „ 
 
 - 
 
 W 
 
 . 
 
 - 
 
 0= 
 
 0. 
 
 o 
 
 H 
 
 2 
 
 a 
 
 s 
 
 2 
 
 2 
 
 S 
 
 a 
 
 2 
 
 :. 
 
 
 
 1 I 
 
 § 
 
 aoil<3S P in 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3Di[ds ai 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 usj[oja 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 qsM IHflS 
 
 g 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 pe»H WS 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P E3 H 
 
 p 3 qsiu 3 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 [EjaW 
 
 JO *0[j 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ira^jg 
 
 g 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P33EUIEQ 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 II 
 
 1 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 ! i 
 
 1 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 ! 
 J. = 1 
 
 a s " 
 ll'ojli 
 
 piracy 
 
 i|»no K JO Xeq 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I! 
 
 jco A 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 a . 
 
 OS 
 05 
 
 1 
 
 as 
 
 
 |°l 
 
 » 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Jo 3Ajno 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 l 
 
 1 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 a 
 
 sqiSuaq 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 U 
 
 P. 
 
 ■wiiaT 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 l 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ll 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ill 
 
 • 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * | 1 
 
 „ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 S 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 |°J 
 
 » 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P-">A 
 
 JSd 
 
 jqSiaAi 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1| 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 B B: ; | 
 
 M. W. 405. — Superintendent's Monthly Report of Rail Failures in Main 
 
 Tracks : 
 On this blank the Division Engineer informs his Superintendent of the 
 total number of rail failures for the month, tabulated from the Track Foreman's 
 
 (508) 
 
INSTRUCTIONS 
 
 A. The Division Engineer will make out two copies of this report at the end of 
 the month from the Section Foremen's Reports, and send one copy to the 
 Chief Engineer M. of W. and one to the General Superintendent. 
 
 DESCRIPTION OF RAIL FAILURES 
 
 When describing Failures of Rails, the following terms should be used: 
 iROKEN Rail. This term is to be 
 break will come under this head. 
 
 C 2. Flow of Metal. This t< 
 of the head toward, its, 
 ing down of the head s 
 distorted. 
 
 eans a "Rolling Out" of the metal on top 
 .vithout there being any indication of a break- 
 re, that is, the under side of the head is not 
 
 Split Head. This term includes rails split through or near the 
 of the head, or rails with pieces split off the side of the head, 
 term is used it should be further defined by stating whether it 
 accompanied by a seam or hollow head. 
 
 2 A 
 
 EZf 
 
 Fig. 346 (continued). (Back of Form.) 
 
 report, and other officers who are interested, such as the Chief Engineer, Chief 
 Engineer of Maintenance of Way, or General Superintendent, are furnished with 
 copies. In cases where a copy of the Track Foreman's report is sent to the Chief 
 Engineer or Chief Engineer of Maintenance of Way, the monthly report serves 
 as a check on the receipt of all individual rail reports. 
 
 (509) 
 

 A. B. & 
 
 C. R. R. Co. 
 
 Division. 
 
 Main Track. 
 
 December 31, 19 _ . 
 
 Location . 
 
 j 
 
 «. 
 
 * 1 
 
 | 
 
 Length of Feet in Tra:k 
 
 From To 
 
 Laid Previous 
 to 19 
 
 New Steel Laid 
 19 
 
 Steel Laid 
 19 
 
 Remarks. 
 
 M.P. 
 
 + ft. 
 
 M.P 
 
 + ft. 
 
 
 
 
 
 
 
 
 
 I 
 
 
 3 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 5 
 
 4 
 
 5 
 
 3 
 
 12 
 
 4 
 
 5 
 
 8 
 
 7 
 
 7 
 
 16 
 
 
 
 
 
 
 
 
 heet 
 
 required 8 inches 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 No. ft. 601 
 " " 70 
 " " 85 
 " " 100 
 Column 3 to be used for any special rail, s 
 and Re-drilled. 
 
 To be made out and forwarded by the Eng 
 Engineer M. of W., as soon after the close of t 
 
 b.Rail No. Tons 
 
 ich as Re-rolled or Sawed 
 
 Correct: 
 neer M.of W. to the Chief 
 le year as possible. Engineer M.ol W. 
 
 Fig. 347. 
 
 M. W. 406. — Annual Statement of Steel Rails Existing in Main Tracks: 
 This is an annual report sent by the Division Engineer to the Chief Engineer 
 or Chief Engineer of Maintenance of Way, for the permanent record of the com- 
 pany, to show the different kinds of steel in the main tracks at the end of the 
 year. This may be used in conjunction with the rail chart, or take its place 
 altogether, because the rail chart may not be in convenient form for a per- 
 manent record, which may be referred to, after many years, for information 
 concerning the kind of rail in use at a stated period. (510) 
 
* .2 
 
 2D 1 
 
 2 C 
 
 1 i 
 1 3 
 
 ■g.2 
 
 1j 
 
 (511) 
 
Group III. Laboratory Examination of Special Rails 
 
 This group is, at present, represented by the single form shown in Fig. 348. 
 It is used for making check analyses against the mill analyses and for reporting the 
 result of chemical analysis and physical test of special rail or other test pieces which 
 may be sent to the laboratory, from time to time, for examination. Fig. 349 shows 
 standard locations of borings for chemical analyses and also the standard tensile 
 test pieces of the association. 
 
 FOR CHEMICAL ANALYSES. 
 
 To be of maximum 
 
 IF RAIL IS FLANGE WORN, THE 
 BORINGS AND TEST PIECE FROM THE UPPER 
 PART OF HEAD SHALL BE TAKEN FROM THE 
 OPPOSITE CORNER. 
 
 FlHfl— ''i — 
 
 W- 
 
 - Standard Locations of Borings for Chemical Analyses and Standard Tensile 
 Test Pieces. 
 
 Group IV. Compilation of Results for Study 
 
 This group, Figs. 350-354, exhibits the different ways for compiling quan- 
 titative statistics of rail failures. 
 
 M. W. 408 (Fig. 350) is intended for compiling the information relative to 
 rail failures for a period of one year. 
 
 The columns for "specified chemical analysis " are intended for recording 
 the analysis of the particular lot of rail as given in the specification, and is 
 inserted in this blank in order to give an idea as to whether the rail is high or 
 low in carbon, or high or low in phosphorus, etc. 
 
 M. W. 409 (Fig. 351) has been provided on which the results from M. W. 
 408 will be recorded at the end of the year, thus making a continuous 
 record. 
 
 (512) 
 

 s, g i 
 '> i 
 
 |asi 
 Jfgj 
 
 lis"? 
 Ill 
 
 ...n 
 
 - 
 
 « 
 
 n 
 
 » 
 
 „ 
 
 ■ 
 
 - 
 
 CO 
 
 ■ 
 
 2 
 
 ; 
 
 2 
 
 2 
 
 » 
 
 2 
 
 2 
 
 £ 
 
 2 
 
 2 
 
 g 
 
 - 
 
 31 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s| 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 j 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 { 
 
 = 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 | 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 | 
 
 = 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 f 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 «. 
 
 i 
 
 « 
 
 i 
 
 ^ 
 
 z 
 
 u 
 
 1 
 
 « 
 
 1 
 
 ct 
 
 1 
 
 ..'- 
 
 1 
 
 « 
 
 i 
 
 K 
 
 ° 
 
 <a 
 
 | 
 
 <r 
 
 i 
 
 «R 
 
 i 
 
 « 
 
 i 
 
 (K 
 
 i 
 
 «r 
 
 i 
 
 -V 
 
 s 
 
 o 
 
 V, 
 
 ° 
 
 « 
 
 ° 
 
 !.- 
 
 i 
 
 CX 
 
 = 
 
 was" 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i- 
 
 
 11 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 j: 
 
 
 fj 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o|, 1 
 
 If- 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 || 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ..u n 
 
 " 
 
 „ 
 
 " 
 
 " 
 
 ■ 
 
 . 
 
 - 
 
 a 
 
 " 
 
 2 
 
 
 2 
 
 2 
 
 2 
 
 2 
 
 2 
 
 - 
 
 2 
 
 2 
 
 s 
 
 S 
 
 il I 
 
 !1 a 
 
 » put 'a"juo inreiq siq» X< 
 
 ia.2 we „prei irej « 3 ^, 
 
 ' ?..:•"!- ™? SU , M S 2uun' '£ 
 
 i 
 
 " -dnojSqora jo suum|a> aqj dn'jooj ■ 
 'uoipss '}q2i3A\ aqj saiS pire 'pjpq rim lasqs aq; ss jiei |0 spupj A"tmu St ]33qs qoea uo Jajir^ ■ 
 
 •aNvia sihx ni ONrnid hoj SNOixoauia 
 

 33UH 
 
 - 
 
 » 
 
 - 
 
 . * 
 
 • 
 
 - 
 
 CO 
 
 a 
 
 2 
 
 a 
 
 a 
 
 2 
 
 -3 2 2 
 
 a a.a 
 
 
 
 P'H 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 | 
 
 
 
 O H 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 1 
 s 
 
 1 
 
 3 
 
 | 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 5 
 
 I 
 
 T> I 
 
 £ 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 o 
 
 I ! 
 
 j 
 
 1 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 £ 
 
 ,£ " 
 
 3 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 a 
 
 « 5 
 
 s 
 
 3 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 3 
 
 § 
 
 « 1 
 
 I 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 5 
 1 
 
 
 : 
 
 | 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 5 
 
 
 3 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 £ 
 
 3) 1 
 "6 * 
 
 1 
 
 j 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 \ 
 
 >i ° 
 
 s 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 
 E s 
 E " 
 
 J 
 
 j 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 1 
 
 (0 
 
 1 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 i 
 
 
 1 
 
 j 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 7 
 
 
 ! 
 
 7 z 
 
 z ? 
 
 ? ? ? 
 
 z 7 7 
 
 ? ' 
 
 
 7 I 
 
 7 ? 
 
 ? J 
 
 ? f 
 
 ? 
 
 z 
 
 7 7 
 
 ? ? 7 z ? ? 
 
 7p*** * 
 
 b 
 
 
 
 
 - ; 1 
 
 
 I L 'l 
 
 '- i 
 
 ! 2 
 
 I 
 
 - ; 
 
 
 - 
 
 * 
 
 1 
 
 '1 
 
 *| *| 65 | 
 
 *| *| «| 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 s 
 
 
 
 
 
 
 
 i 
 
 , 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 * 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 , 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 
 , 
 
 t' 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 1 
 
 -0 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 •2 
 
 
 
 
 
 
 I 
 
 , 
 
 s 
 
 
 
 
 
 
 
 
 t 
 
 | 
 
 
 
 
 
 
 3 
 
 , 
 
 ° 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 9 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 ~ 
 
 „ 
 
 
 
 
 
 
 
 
 
 
 E 
 
 
 
 
 
 
 i 
 
 , 
 
 
 
 
 
 
 
 
 
 a 
 
 
 
 
 
 
 
 : 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -, 
 
 
 
 
 
 
 
 
 
 J 
 
 Hi 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 8-5 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 1 
 
 
 ssan 
 
 - 
 
 c, 
 
 ™ 
 
 * '° 
 
 ° 
 
 '- 
 
 GO 
 
 o, 
 
 
 
 2 
 
 2 
 
 2 2 2 
 
 a a^l 
 
a 
 
 1 
 
 c j 
 
 l||I 
 
 
 
 
 
 
 *> 
 
 
 
 
 
 
 
 
 
 Sheet No of 
 
 Different Section or Patte 
 
 npanies. 
 
 6h 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 B 
 
 * 
 
 £ 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 jjj 
 
 1 
 
 1 
 
 1 
 B 
 
 B 
 
 1 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 s 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 I 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 ■i 
 
 
 - 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ * j 
 
 (5 o | 
 
 L. 
 
 <t- ,(D 
 
 °it 
 
 (0 
 
 
 
 2u 
 
 1 
 
 I 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 '* 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 ■n 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 = 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 & 
 
 
 
 
 
 
 
 
 
 
 8 
 
 
 -> 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 : 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 J 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 £ s 
 
 <D 
 jD 
 
 E 
 
 D 
 
 z 
 
 
 > 
 
 L 
 
 (0 
 
 a 
 E 
 
 
 
 
 
 § 
 
 r 
 
 , 
 
 
 
 , 
 
 , 
 
 . 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 $ 
 
 Z « £ 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 UJ3MEJ JO 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 ■sqi 'pjEX J3d 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 s 
 
 
 Fig. 352. 
 
 M. W. 410. — Comparative Number of Failures of Steel Rails of Different 
 Section or Pattern, Rolled by Different Steel Companies: 
 
 In order to compare the product of different mills, and also to compare 
 different weights per yard and different sections together, this blank has been 
 provided. It contains the totals taken from M. W. 408 or M. W. 409 as desired. 
 
 (515) 
 

 r a 
 
 
 ^ 
 
 £ 
 
 t 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -, 
 
 
 
 •o 
 
 
 
 
 
 <" 
 
 "H 1 
 °"E 
 
 £2-3 
 
 ill 
 
 "3 -3 & 
 1 §s§ 
 
 sag 
 1 11 
 
 g 8 s 
 
 - %-°L 
 
 2 
 
 2 
 
 i! 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 . 
 
 
 
 
 
 2 
 
 JU 
 
 
 h 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 v 
 
 
 
 
 
 H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 » 
 
 
 
 
 
 
 Q 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 > 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 
 
 , 
 
 
 
 
 
 PQ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 > 
 
 
 § 
 
 , 
 
 
 
 
 
 < 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 > 
 
 
 
 
 
 S 
 
 A 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 i 
 
 ■n 
 
 
 
 
 
 if 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 fa 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .> 
 
 
 
 
 
 
 
 fa 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 § 
 
 1 I 
 
 s * 
 
 I 
 
 1 
 
 Q 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 1 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 v 
 
 - 
 
 
 
 
 a* 
 
 a* 
 
 • c (2 
 
 I** 
 
 fa 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 : 
 
 
 
 
 73 
 
 
 
 
 < 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 , 
 
 i 
 
 
 
 
 a 
 
 * 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■3 
 
 
 
 w, 
 
 
 
 
 
 
 
 
 
 = "" 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 fc 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 v 
 
 "H 
 
 
 
 
 fa 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 >. 
 
 
 b 
 
 
 K 
 
 
 
 
 cH 
 .2 «- 
 
 Q 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 g 
 
 
 
 
 
 
 fa 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 < 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 f 
 
 „ 
 
 
 
 
 
 -' 
 
 1 
 
 A 
 
 1 
 
 
 
 
 
 
 
 
 ! 
 
 
 
 
 
 
 | 
 
 
 
 
 
 
 
 
 
 
 Q. 
 
 s| 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 - 
 
 
 
 
 
 PK 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 "" 
 
 
 
 
 
 H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - ; 
 
 
 
 * 
 
 
 
 
 
 Q 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 pq 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 : 
 
 
 
 " 
 
 
 
 
 
 < 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 v 
 
 
 
 
 
 o S 
 2 £ 
 
 , 
 
 s 
 
 •s 
 i 
 
 s 
 
 S 
 
 - 
 
 £ 
 
 , 
 
 | 
 
 = 
 
 ! 
 
 b 
 
 \ 
 
 I 
 
 1 
 
 ■ : 
 
 i 
 
 = 
 
 is 
 
 : 
 
 H 
 
 
 Fig. 353. 
 
 M. W. 411. — Position in Ingot of Steel Rails which Failed: 
 This is intended to furnish data on the number and character of rail failures 
 according to the original position in the ingot held by the rail in question. 
 
 (516) 
 
(Cover Page for Forms M. W. 408, 409, 410. 411.) 
 
 A. B. & C. R. R. Co. 
 
 Numerical Record and Position in the Ingot of Steel Rails which have 
 Failed in Service. 
 
 M.W.412: 
 
 The information in this group should be bound together in one book; this 
 cover has been provided for convenience and neatness. 
 
 Group V. Progressive Wear of Special Rail under Observation 
 
 In order to keep track of special rail, from time to time, and determine 
 the value of the results being given, it is necessary to have a systematic plan 
 of procedure for examinations and records. This group, Figs. 355-359, is fur- 
 nished for that purpose, and is provided with a cover, as in the case of the 
 previous group. (517) 
 

 
 
 
 
 
 
 1 
 
 
 
 
 Division 
 Rails 
 
 R. F 
 
 
 
 
 
 
 red in 19 
 
 id 
 
 
 
 
 Location D 
 
 of 
 
 Laid in 19 Remo 
 
 Between a 
 
 Office of Chief Engineer M. of 
 Scale 1 in.=l mile Date 
 
 
 
 
 
 
 
 
 
 
 
 
 >< 
 
 i 
 
 t 
 
 6 
 
 | 
 
 I 
 < 
 
 
 « 
 
 
 < 
 
 
 
 
 
 
 » 
 
 
 < 
 
 
 
 
 
 
 « 
 
 
 < 
 
 
 
 
 
 
 « 
 
 
 < 
 
 
 
 
 
 
 » 
 
 
 < 
 
 
 
 
 
 
 « 
 
 
 < 
 
 
 
 
 
 1 
 
 I 
 
 Total sq. in. abraded 
 Ave. sq. in. abraded 
 
 Percentage of Area 
 abraded to total 
 Area of Head 
 
 
 
 
 
 M. W. 413. — Location Diagram: 
 
 This blank is on a scale one inch equals one mile, and is intended for dia- 
 grams showing the location in different places of the same kind of rail under trial. 
 
 (518) 
 

 
 
 
 ! ; 
 
 
 5 
 
 ft! 
 ft! 
 U 
 
 <b 
 
 Location Diagram 
 of 
 
 Rails 
 
 Laid in 19 Removed in 19 
 
 Between and 
 
 Office of Chief Engineer M. of W R. R. 
 
 Scale 2 in.=l mile Date 
 
 
 
 
 
 I 
 
 i 
 
 
 m 
 
 
 < 
 
 
 
 
 « 
 
 
 < 
 
 
 
 ca 
 
 
 < 
 
 
 
 
 «n 
 
 
 < 
 
 
 
 « 
 
 
 < 
 
 
 
 
 * 
 
 
 < 
 
 
 > 
 
 3, 
 
 Total sq. in. abraded 
 Ave. sq. in. abradec 
 
 Percentage of Area 
 abraded to total 
 Area of Head 
 
 
 
 Fig. 356. 
 
 M. W. 414. — Location Diagram: 
 
 This is similar to M. W. 413 except that it is on a scale of two inches equal 
 one mile, and is intended to show the location of a particular portion of the 
 rail given in M. W. 413. It is made on a larger scale, so as to locate the points 
 of measurement. A place is provided on each blank for the summary of the 
 wear or area abraded in percentage of total area of head. (519) 
 
Fig. 357. 
 
 M. W. 415. — Diagram Showing Lines of Wear: 
 
 The measurement of rail section at a specified point is shown on this blank 
 and its position on M. W. 414 is given by the number in the circle of the blank 
 at the top. All statistical information of interest and importance is given on 
 the blank. (520) 
 
A. B. & C. R. R. Co. 
 
 Division 
 
 Record of Comparative Wear of Special Rail 
 
 Date of Report 19 
 
 i 
 
 B 
 
 c 
 
 < en 
 
 < 
 
 
 | = 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 1 
 I 
 ■g 
 
 I 
 I 
 
 1 
 
 o 
 
 2 
 
 11 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 = 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 #1 
 
 
 
 
 
 
 
 
 "» 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 uoday jo ajsp 
 je aoiAjag 
 jo qjguaq 
 
 
 
 
 
 
 
 ; 
 
 10 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P3A01U3-JJ 3JEQ 
 
 
 
 
 
 
 
 
 ; 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 piCT 3JEQ 
 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Chemical Composition 
 
 of Rail 
 
 of Section with 
 
 which Special Rail was 
 
 Compared 
 
 
 
 
 
 
 
 
 s 
 
 «. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9 
 
 > 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 » 
 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 c^ 
 
 
 
 
 
 
 
 J! 
 
 - 
 
 , 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 m 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 0. 
 
 
 
 
 
 
 
 : 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 
 
 
 
 
 
 T 
 
 <• 
 
 ; 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 l 
 
 ! 
 i 
 
 i 
 
 
 
 
 
 
 
 
 s- 
 
 - 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■a 
 
 
 
 
 
 
 
 ■ : 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 B5 
 
 
 
 
 
 
 
 
 * 
 
 ; 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~ 
 
 
 
 
 
 
 
 : 
 
 " 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 JSJMDEjnUBIM 
 
 
 
 
 
 
 
 £ 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 uonosg jo sdi<x 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 pjE.< aad iqSiaM 
 
 
 
 
 
 
 
 £ 
 
 «, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 piBlsuoj,joo N 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 
 
 
 
 
 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 M. W. 416. — Record of Comparative Wear of Special Rail: 
 This blank is intended for compiling the information given in the previous 
 ones, so as to give a general summary of the results. (521) 
 
(Cover Page for Forms M.W. 413, 414, 415, 416.) 
 
 A. B. & C. R. R. Co. 
 
 RAIL SECTIONS 
 
 Showing Progressive Wear 
 
 M. W. 417: 
 
 The information in this group should be bound together in one book; this 
 cover has been provided for convenience and neatness. 
 
 In some cases it may prove desirable to use charts or diagrams which illus- 
 trate graphically the information found in the records. The diagrams of rail 
 failures found in the Proceedings of the American Railway Engineering Asso- 
 ciation are examples of this. Fig. 360 shows a method of recording the failures 
 of different groups of rails during a period of years, and shows as well the dis- 
 tribution of the failures during each year. It will be noted from the figure that 
 the rails failed, with a few exceptions, during the winter and spring months. 
 A modification of this diagram can be obtained by making the record cumu- 
 lative, which affords a ready comparison of the behavior of different rollings 
 after having been in service any considerable length of time. 
 
 (522) 
 
00999 
 
 
Fig. 361 presents a further example and shows a diagram of rail failures on 
 the Harriman Lines. 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 § 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 7 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 $ 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 $ 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 § ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * L 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 r 
 
 <s 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 f^ 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 , 
 
 
 
 s 
 
 
 
 
 
 
 
 
 
 MSiB*- 
 
 
 
 -Z£/± 
 
 — -\" 
 
 
 
 
 
 
 
 -, .-;~^ 
 
 
 i. 
 
 
 
 
 
 
 
 
 "V** 
 
 
 
 
 
 
 a S 
 
 
 — - 
 
 
 -£ 
 
 — j 
 
 
 
 
 
 
 for all lines for the year 1907 with respect to the uai-,oas months. EXAMPLE: Dio- 
 
 miles of 90 /6 section in the track: 4.7 broken 30 /& rails per lOOmileToTso lb roil 
 in the track: and Zl broken 7S/b rai/s per lOOmi/es of 75 /b roi/ in track. 
 
 CONCLUSIONS. 4 The diaqram would seem to Indicate, that for alf weights of 
 roil, and more especial/u for the 90 lb. secfi on, the laraer m umber of breoks took 
 place during the co/der months of the ueor. 
 
 - Diagram of Rail Failures, Harriman Lines. (Am. Ry. Eng. Assn.) 
 
 (524) 
 
Total 7.89 sq. in. 100% Ratio Total Periphery to Total Area 2.78 Total ! 
 
 Moment of Inertia, 27.25 
 Section Modulus, Head, 10.30 
 Section Modulus, Base, 11.97 
 
 Plate I. — Standard Rail Sections of the American Society of Civil Engineers (adopted 1893). 
 
 . 100% Ratio Total Periphery to T 
 Moment of Inertia, 43.8 
 Section Modulus, Head, 14.44 
 Section Modulus, Base, 16.11 
 
Plate II.— Rail Sections used before the Adoption of the A. S. C. E. Standard Sections 
 
Standard Wheel Sections showing Coning 
 of Wheel. 
 


 
 
 — .2tf*- * 
 
 
 
 •J 
 
 [ ;i IOC lb. 
 
 
 ft". 
 
 
 J 
 
 
 
 L T 
 
 
 
 
 
 o 
 
 7 -j. - 
 
 "sa.it 
 
 i 
 
 ^ 
 
 
 L r ,;" 
 
 
 Area of Head 3.05 sq. in. 38.8% 
 
 ' Web 1.65 ' 
 ' Base 3.16 " 
 
 21.0% 
 40.2% 
 
 Ratio Periphery Head to Area of Head 1.93 
 
 Web " " " Web 3.57 
 
 " " Base " " " Base 2.52 
 
 Area of Head 3.64 sq. in. 36.9% 
 " " Web 2.29 " " 23.4% 
 " " Base 3.91 " " 39.7% 
 
 Ratio Periphery Head to Area of Head 
 
 Web " " " Web 
 
 " " Base " " " Base 
 
 l. 100.0% Ratio Total Periphery to Total Area 
 
 Moment of Inertia, 28.80 
 Section Modulus, Head, 10.24 
 Section Modulus, Base, 12.46 
 
 9.84 sq. in. 100.0% Ratio Total Periphery to Total Area 
 
 Moment of Inertia, 48.94 
 Section Modulus, Head, 15.04 
 Section Modulus, Base, 17.78 
 
 3.21 
 3.29 
 
 2^92 
 
 Plate VII. — Proposed Standard Rail Section of the American Railway Association, Series " A " (recommended 1907) . 
 
Area of Head 3.07 sq. in. 38.8% 
 " " Web 1.54 " " 19.5% 
 « " Base 3.30 " " 41.7% 
 
 Ratio Periphery Head to Area of Head 1.79 
 
 Web " " " Web 3.57 
 
 " " Base " " " Base 2.72 
 
 7.91 sq. in. 100.0% Ratio Total Periphery to Total Area 2.53 
 
 Moment of Inertia, 25.1 
 Section Modulus, Head, 9.38 
 Section Modulus, Base, 11.08 
 
 Area of Head 3.95 sq. in. 40.2% Ratio Periphery Head to Area of Head 1.64 
 
 " " Web 1.89 " " 19.2% " " Web " " " Web 3.60 
 
 " " Base 4.01 " " 40.6% " " Base " " " Base 2.49 
 
 Total 9£5 sq. in. 100.0% Ratio Total Periphery to Total Area 2.37 
 
 Moment of Inertia, 41.3 
 Section Modulus, Head, 13.70 
 Section Modulus, Base, 15.74 
 
 Plate VIII. — Proposed Standard Rail Section of the American Railway Association, Series " B " (recommended 1907). 
 
Ratio Periphery of Head to Area of Head 1.73 
 " " Web " " " Web 3.81 
 
 8.47 sq. in. 100% Ratio Total Periphery to Total Area 2.48 
 
 Moment of Inertia, 29.1 
 Section Modulus, Head, 10.77 
 Section Modulus, Base, 12.02 
 
 Area of Head 4.09 sq. in. 
 " " Web 1.85 " " 
 " " Base 4.03 " " 
 
 41% 
 19% 
 40% 
 
 100 lbs. per yd. 
 
 Ratio Periphery of Head t 
 " Web ' 
 
 i Head 1.59 
 Web 3.58 
 
 9.97 sq. in. 100% Ratio Total Periphery to 1 
 
 Moment of Inertia, 41.9 
 Section Modulus, Head, 13.71 
 Section Modulus, Base, 15.91 
 
 Plate IX. — Standard " P. S." Rail Section of the Pennsylvania Railroad System (adopted 1907). 
 
96.7 lbs. per yd. 
 i, Lyon and Mediterranean Railway 
 
 91 lbs. per yd. 
 French Eastern Railway. 
 
 92.7 lba. per yd. 
 Egyptian State Railways. 
 
 Plate X. — Rail Sections of the Vignole Type. 
 
87.9 lbs. per yd. 
 Bavarian- Wiirttemberg Railways. 
 
 91 lbs. per yd. 
 Prussian-Hessian Railways. 
 
 Plate XI. — Rail Sections used on German Railways. 
 
 92.3 lbs. per yd. 
 Saxony Railways. 
 
-ACTUAL WEIGHT jpp jp3 LBS. PER YARD. - 
 
 LENGtH 
 
 3$ FEET. 
 
 DETAILS OF STEEL FISH BOLT & STEEL LOCK NUT. 
 
 -WEIGHT 36 LBS. PER PAIR - 
 
 -ELEVATION OF FISHPLATE- 
 
 Plate XII. — Midland RaUway, Permanent Way, 1907. 100 lbs. Steel Rail, Steel Fishplates and Bolts, half size. 
 
l„,}„ 
 
 
 B C 
 
 1 il i 
 
 i i i (] f] ' 
 
 I il 
 
 i] ii ij n i| rj i] i] n n n 
 
 I | J o □ □ □ □ □ d~H d d U U\H 
 
 ! 
 
 EE 
 
 Q| Q_ ® 
 
 w 
 
 IE 
 
 iM 
 
 
 W«i|l>t of 1 pair ■TIUlWii -311b. 
 ' 1 
 
 
 l 
 
 
 
 
 | <B <3> 
 
 
 f 
 
 
 
 lie joint chairs are 91 ins. wide, and mil* 54 
 
 @ 
 
 te chairs are 71 ins. wide, and weigh 46 
 
60 lbs. per yd. 
 \. S." Section No. 60. 
 
 80 lbs. per yd. 
 "B. S." Section No. 80. 
 
 Plate XIV. —British Standard Bull Head Railway Rails. 
 
Pi^te XV. — British Standard Flat Bottom Railway Rails. 
 
Tram Girder Rails, Pennsylvania Steel Company. 
 
 85 lbs. per yd. 
 
 100 lbs. per yd. 
 
 lbs. per yd. 
 
 Standard High Tee Rails of the American Electric Railway Engineering Association. 
 
 Plate X\I. — Rail Sections for Street Railways, Tram Girder Rails and High Tee Rails. 
 
7-in. Groove Rail, 106 lbs. per yd. 
 
 9-in. Groove Rail, 120 lbs. per yd. 
 
 7-in. Guard Rail, 122 lbs. per yd. 
 
 9-in. Guard Rail, 141 lbs. per yd. 
 
 View showing amount that may be 
 worn from head of Groove Rail before 
 standard wheel flange will come in 
 contact with lip of rail. 
 
 Plate XVII. — Rail Sections for Street 
 Railways: Standard Girder Sec- 
 tions of the American Electric 
 Railway Engineering Association. 
 For use in paved streets with heavy 
 vehicular traffic. 
 
 24-25J 
 
 
Plate XVIII. — British Standard Tramway Rails. 
 
Crossing Frogs in Poor 
 
 DATUM LINE 
 
 DATUM LINE 
 
 Applir.ilioii of Brakes, 
 
 Straight, Level Track. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 DATUM I 
 
 JNE 
 
 
 = — 
 
 
 
 
 
 
 
 
 
 DATUM LINE 
 
 24 
 
 30 
 
 [Plate XIX. — Deflection of Drh 
 
 15 18 21 
 
 TIME IN SECONDS 
 
 Note. — Static Load Deflection 0.5 in. below Datum Line 
 'heel Spring, Consolidation Engine No. 1064, Boston and Maine Railroad. (Coes and Howard.) 
 
2TT- 
 
 Atlantic Type Passenger Locomotive. 
 hooded Weights. 
 
 On leading truck 42,500 pounds 
 
 On driving wheels 105,000 pounds 
 
 < )n liiiilni!- I ruck 42,500 pounds 
 
 Tot ;il engine 190,000 pounds 
 
 Tender 130,000 pounds 
 
 Prairie Type Locomotive 
 Loaded Weights. 
 
 On leading truck 
 
 On driving wheels 1 
 
 10 pounds 
 K) pounds 
 
 Pacific Type Passenger Locomotive. 
 Loaded Weights. 
 
 On leading truck 36,900 pounds 
 
 On driving wheels 147,800 pounds 
 
 On trailing truck 40,300 pounds 
 
 Tola! engine 225,000 pounds 
 
 Tender 162,800 pounds 
 
 Plate XX. — Pass 
 
 can Locomotive Co.) 
 
Loaded Weights. 
 On driving wheels 445,000 pounds 
 
 Mogul Type Freight Locomotive. 
 Loaded Weights. 
 
 On driving wheels 1511.000 pounds 
 
 On leading truck 28,000 pounds 
 
 Total engine 187,000 pounds 
 
 Tender 140,000 pounds 
 
 Plate XXI. — Freight Locomotive Diagrams. (From 
 
dii 
 
 "V" 
 
 f 
 
 DIP 
 
 """V 
 
 b^ 
 
 
 
 p LJ 
 
 p ,, 
 
 
 
 r - T: 
 
 
 n 
 
 n 
 
 n 
 
 n 
 
 __, , _is^, 
 
 07 
 
 Di 
 
 n 
 n 
 
 p 
 
 3 
 
 fl] 
 
 = 
 
 MoKee Single-flange Plate. 
 
 Plate XXII. — Examples of American Tie Plates. 
 
 Pittsburgh and Lake Erie Railroad Tie Plate. 
 
V 
 
 :_ 
 
 __ 
 
 Plate XXIII. — Rail Diagram of Love. 0.46 original size. 
 Note.— In the curve of bending moments, the maximum bending moments under the wheels arc 
 determined by rombmin:: ihc nonnil moment i.lia^nmis, -luiw-n in dotted hnes, with the moments a 
 the adjacent ties. 
 
Plate XXIV.— Examples of American Rail Joi 
 
/OO /25 /SO 
 
 Axle Spacing' Inches 
 
 175 
 
 ZOO 50 
 
 
 
 
 
 1^50 
 
 
 
 
 ^^ 
 
 v 
 
 
 
 
 
 
 Plate XXVI.-Dynamic Wheel Loads for Various Rails and Axle Spacing. 
 
 
 
 
 
 
 — £^to 
 
 ^-^ 
 
 
 ^-- 
 
 
 --- 
 
 
 
 
 — ^JS^ 
 
 I I I I I I . 
 
 n ? S c /0 ° , , ^ 5 50 75 r 100 /25 
 
 ftxle Spacing Inches /jx/e Spacing Inches 
 
 Class B Track. Class C Track. 
 
500.000 
 
 
 
 J =50 
 
 
 
 
 
 
 
 
 T=4-o 
 
 
 
 
 1=30 
 
 
 A 
 
 /^' 
 
 I=ZO__ 
 
 
 
 
 f 
 
 
 
 
 
 
 
 
 
 
 
 
 
 /^30 
 
 
 ^^ 
 
 
 ^ 
 
 I*ZO 
 
 
 
 
 BENDING MOMENTS IN DIFFERENT RAILS WHICH 
 WILL CAUSE AN EXTREME FIBER .STRESS OF 
 20,000 LBS. PEE SQ. IN. IN THE BASE OF THE 
 
 75 100 125 /50 
 
 Axle 5pacmg Inches 
 
 /75 
 
 ZOO 50 
 
 it Weights of Rail Corresj 
 
 75 100 Its 
 
 Axle Spacing Inches 
 
 ■ 
 
 "SZZSSS? 
 
 *s$£ixsr- 
 
 40 
 35 
 
 8 
 
 1S.0 
 13.5 
 
 27C 
 24C 
 
 000 
 000 
 
 000 
 000 
 
 000 
 
 
 
 
 
 
 I' 50 
 
 
 /-- 40 
 
 
 I* 30 
 
 
 /= Zo 
 
 50 a 7 , 5 c w0 , '25 
 
 Mxie j pacing inches 
 
v<» 
 
 eg 80 
 
 
 ft. 
 
 4 
 
 n 
 
 E 
 
 30,000 40,000 50,000 
 Static Axle load lbs. 
 
 
 <3L 
 
 la 
 
 
 
 1 
 
 h 
 
 % 
 
 A 
 
 
 
 i 
 
 1 
 
 // 
 
 
 
 
 i 
 
 A 
 
 
 
 
 
 1 
 
 /I 
 
 
 
 
 
 t 
 
 
 
 
 
 /( 
 
 
 
 
 
 30,000 40,000 50,000 6QOO0 
 
 Static Axle load Ib5. 
 
 ^3 
 
 01 
 
 -ffi. 
 
 ..JOS:^ 
 
 mi 
 
 SL 
 
 i, I OonsolulMion Lncomotive-". 
 
 ^P^P^T 
 
 
 
 
 
 
 i 
 
 V 
 
 
 
 
 
 
 i 
 
 
 
 p f 
 
 rf 
 
 
 / 
 
 i i 
 
 
 
 / 
 
 
 
 / 
 
 1 
 
 i 
 
 
 
 
 
 
 // 
 
 ' 
 
 
 
 // 
 
 1 1 
 
 
 // 
 
 
 
 20,000 30,000 40,000 50,000 
 
 •Static Axle load lbs. 
 
 Era. B.— Ten Wheel Loci 
 
 CD QCDojn 
 
 ^ 
 
 ^ 
 
 
 ^70 
 
 
 
 
 
 
 7 
 
 
 
 
 
 4 
 
 V 
 
 
 1 
 / 
 
 
 
 // 
 
 
 
 k\ 
 
 
 
 // 
 
 
 
 1 
 i 
 
 
 
 '/ 
 
 
 i 
 1 
 
 
 
 / 
 
 / 
 
 
 £Q0O0 JQGO0 4Q000 
 Static Axle load Ib5. 
 
 Era. D. — Freight Cars and Six Wheel Passenger Cars 
 
 F 
 
 CD CO 
 
 
 ^q 
 
 P 
 
 CD CD CO 
 
 i Steam Roads. 
 
 h/00 
 
 
 Class A Track. - 
 Class 13 Track. - 
 Class C Track. - 
 
 — rf— 
 
 Zi 
 
 >£ 
 
 ^Q000 JQOOO 40,000 
 
 Static Axle load lbs. 
 
 9 o 
 
 on Electric Roads. 
 
Plate XXLX. — Plan of Gary Steel Plant. (Harbord and Hall.) 
 
Plate XXX. — Reversing Cogging Mill. (Harbord and Hi 
 
Platk XXXI. —American Three-high Mill, with 36-inch Rous. (Harbord and Hi 
 
m XXXII. (Puppe.) — Power required to Roll Rails about 35.S kg. per meter. 
 For further data see Table XCVI. 
 
 Output of Motors. 
 
 Speed curve. 
 

 
 r- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 RESULTS OF 
 
 A. B. & a R. R. Co. 
 
 DROP TESTS, JUn~D SUREACE mSPECTIOH OF RAILS POL 
 
 i»i forme Jnalysis of Heals fe/ecfej 
 
 LED DXJRmq %fcSrt^ffi%£U/&!Z 
 
 
 
 li 
 
 z„ 
 
 1 
 
 I 
 
 Name ofCompany 
 
 Pail 
 
 it 
 
 ■'■>.■:. 
 
 54 
 
 No. of 'Tons Received 
 
 of each Classification. 
 
 /ejected. 
 
 I 
 
 I 
 i 
 
 1 
 
 
 1 
 
 No. of Pails Accepted No. of Pails 
 
 1 Ts 8\. 
 
 No. of Pa 
 
 I :,';,. 
 
 ■ ■ 
 III? 
 
 Hi 
 
 in 
 
 lift 
 
 Average Analysis j Average Analysis 
 of Heats Accepted j of Heats Rejected 
 
 'o:L'!%td 
 
 ll 
 
 1, 
 
 ta&ESS 
 
 
 
 f 
 
 s 
 
 No.l 
 
 Specials 
 
 No. 2 
 
 is! 
 
 1 
 
 '.: 
 
 No.l 
 
 Specials 
 
 No. 2 
 
 11 
 
 J 
 
 $1 
 
 1! 
 
 £* 
 
 P. 
 
 
 1 
 
 
 i\ 
 
 
 Tons 
 
 
 A 
 
 B. 
 
 c. 
 
 
 
 * 
 
 I 
 
 it 
 
 Pails 
 
 A 
 
 P 
 
 r 
 
 p 
 
 1 
 
 J J 
 
 1 
 
 1 '' * 
 
 M it 
 
 |M 
 
 1 
 
 \ 
 
 I 
 
 1 
 
 J_ 
 
 1 
 
 * 
 $ 
 
 1 
 
 
 | 
 
 No. 
 
 \.i. 
 
 No 
 
 
 No. 
 
 
 No. 
 
 
 
 sk^^ 
 
 Ma.ng. 
 
 PhOS 
 
 Sit 
 
 Sulpha 
 
 
 m> 
 
 Phos 
 
 Si! 
 
 Solplti 
 
 
 
 
 
 / 
 
 
 
 
 
 
 ° 
 
 
 
 
 " 
 
 " 
 
 
 
 
 
 * 
 
 
 ,. 
 
 ' J 
 
 " 
 
 
 " 
 
 
 
 
 " 
 
 " 
 
 " " 
 
 
 " 
 
 
 
 " 
 
 
 
 " 
 
 " 
 
 
 " 
 
 
 
 
 " 
 
 " 
 
 - 
 
 
 •• 
 
 
 "1 » 
 
 ^r 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 | 
 
 I 
 
 H— 
 
 
 2 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — t- 
 
 
 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 r 
 
 -j— 
 
 
 — 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 f— 
 
 
 
 
 
 e 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 
 
 
 „ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — 
 
 
 
 9 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ — 
 
 
 
 — 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 • r~ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 12 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 13 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 
 « 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r 
 
 
 
 
 
 
 is 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i ■ 
 
 
 m 
 
 
 
 Grand Total 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 These two columns 
 ire for office use 
 
 'everal rollings and 
 shipments, and are 
 'o be removed on line 
 1'frem the finished 
 
 <* 
 
 
 NQTE - DROP TESTS TESTS AND INSPECTION. 
 
 If Specials are not accepted, columns 8,9.10,11 and Drop Tests shall be made on pieces of rail rolled from the top of (a) Two pieces shall be tested from each heat of steel. If of Hie same heat. If two out of three of these second test pieces 
 Z3.Z4-.Z5.26 should be omitted. the ingot, not less than four (4)ft. and not more than sixfelft. bng.from either of these test pieces breaks.a third piece shall be 'tested, break, the remainder of the rails of theheat will also be rejected 
 each heat of steel. These test pieces shall be cut from the rail bar If two of the test pieces break without shorting physical defect Iftm out of three of these second test pieces do not break the 
 next to either end of the top rail, as selected by the inspector all rails of the heat mil be rejected absolute//. If two of the test remainder of the rails of the heat mil be accepted provided they 
 
 pieces do not break, all rails of the heat will be accepted as No 1 conform to the other requirements of these specifications as No 1 
 The test pieces shall be placed head upward on solid supports, five or No 2 classification, according as the deflection is less or or No. 2 classification, according as the deflection is less or more 
 (5) in top radius, three (3.) ft between centers, and subjected to more, respectively, than the prescribed limit. respectively, than the prescribed limit. • 
 impact tests, the tup failing free from the following heights: 
 
 70 lb. rail. left. &) If, however, any test piece broken under testy.shom (d) If any test piece, teste;does not break' but when nicked and 
 
 60,85 and 90 lb rail. 18ft. physical defect, the top rail from each ingot of that heat tested to destruction shows interior defect the fop rails from 
 
 lOOlb.rail. 20ft. shall be rejected. each ingot of that heat shall be rejected. ' 
 
 The test pieces which do not break under the first drop shall 
 
 be nicked and tested to destruction. Cc) Additional tests shall then be made of test pieces 
 
 selected by the Inspector from the top end of any second rails 
 
 
INDEX 
 
 An, Ar 2 , Ar 3 points 426 
 
 Aberdare rails, effect of cold on 285 
 
 Adriatic Railway tie 102, 104 
 
 Affleck tie 104 
 
 Algoma Steel Company: 
 
 rail mill 437, 438, 444 
 
 shrinkage allowed at , 444 
 
 teeming practice 399 
 
 Alloy steel (see Special steels). 
 
 Alsace-Lorraine Railways, data on track. . . . 218 
 
 Aluminum : 
 
 effect of, in casting steel 401, 405 
 
 in iron ore 364 
 
 America (see also American) : 
 
 earl}' steel rails used in 2 
 
 length of rails used in 267 
 
 production of rails in 382 
 
 steel manufacturers of (see Association of 
 American Steel Manufacturers). 
 
 use of English rails in 2, 4 
 
 American (see also America and United 
 States) : 
 
 engines, examples of 32, 33, 34, 72 
 
 joints, examples of 264 
 
 rail mills, practice at 399, 438, 444 
 
 specifications for rails 463 
 
 speed of railway trains 21, 23, 27, 28 
 
 Steel Manufacturers, Association of (see 
 Association of American Steel Manu- 
 facturers), 
 steel rails: 
 
 examples of girder rails 19 
 
 T-rails, early 6 
 
 T-rails, present 10, 460 
 
 prices of 325 
 
 tie plates, examples of 122 
 
 type engine (see Eight-wheel type engine) . 
 American cypress, resistance to pulling 
 
 of spike 140 
 
 American Electric Railway Engineering 
 
 Association, rail sections 19 
 
 American Forest Congress, ties 106 
 
 American Institute of Mining Engineers: 
 
 blowholes 404 
 
 effect of recarbonizing 391 
 
 Gayley dry blast 360 
 
 American Institute of Mining Engineers: 
 
 Lake Superior ores 351 
 
 piping in ingots 401, 409 
 
 relation between chemical composition and 
 
 strength of rails, Dudley on 326 
 
 roasting iron ores 344 
 
 Sauveur on rail structure 429 
 
 American Locomotive Company: 
 
 examples of modern locomotives . 32, 33, 34, 72 
 
 excess pressure of driving wheels 35 
 
 American McKenna Process Company, 
 
 rerolling rails 459 
 
 American Railway Association: 
 
 committee on standard rail and wheel sec- 
 tions, report of 14 
 
 discard, investigation of 417 
 
 lettering rails from different parts of the 
 
 ingot 416 
 
 rail sections, principles governing design 
 
 of 16 
 
 proposed by 14 
 
 series A, description of 15 
 
 failures of 10 
 
 B, description of 15 
 
 test to determine piping 417 
 
 American Railway Engineering and Main- 
 tenance of Way Association (see 
 American Railway Engineering As- 
 sociation). 
 American Railway Engineering Association: 
 
 classification of defective rails 10 
 
 comparison of rail failures of different 
 
 sections 10 
 
 distribution of pressure to subgrade 185 
 
 drop testing machine, proposed by 290 
 
 tests 293 
 
 flow of rail head under wheel loads 205 
 
 grain size in head of rail 432 
 
 impact tests 64 
 
 report on flat spots on car wheels 56, 61 
 
 manufacture of rails, 
 
 363, 374, 378, 436 
 
 metal and composite ties 90 
 
 scleroscope 298 
 
 reports and records proposed by 10, 501 
 
 screw spikes 142 
 
American Railway Engineering Association: 
 
 size and spacing of ties 121 
 
 special steels 333, 339 
 
 specifications for rails 463 
 
 standard locations of borings for chemical 
 
 analysis and tensile test pieces 512 
 
 statistics of defective rails 10 
 
 strength of head and web of rail 252 
 
 rail steel 304 
 
 stresses in the rail 210, 218 
 
 tests on flat spots in car wheels 61 
 
 joints 264 
 
 strength of rail steel 304 
 
 strength of wood 169 
 
 tie plates 133, 170 
 
 tree plantations for ties Ill 
 
 unit stresses of different woods 168 
 
 American Railway Master Mechanics' 
 Association: 
 
 excess pressure of driving wheels 35 
 
 tire wear 197 
 
 American Society for Testing Materials : 
 
 bibliography of impact tests 288 
 
 cupro-nickel steel 332 
 
 drop tests 288 
 
 ductility in rail steel 286, 287 
 
 experiments on repeated stress 278, 280 
 
 finishing temperature 434 
 
 hardness tests 298 
 
 Howe, on welding blowholes 404 
 
 ferrite grains 427 
 
 impact tests 288 
 
 influence of titanium on segregation 405 
 
 manganese sulphide in steel 390 
 
 manufacture of car wheels 57 
 
 specifications for girder rails 491 
 
 T-rails 463,491 
 
 stremmatograph experiments 236 
 
 Talbot, rolling practice in England 434 
 
 tests on nickel steel 332, 334 
 
 American Society of Civil Engineers: 
 
 committee on rail sections 6 
 
 finishing temperature in rails 7 
 
 rail section 6 
 
 difficulty in rolling heavier rails. . . 7, 14, 435 
 
 failures of 10 
 
 necessity for rolling at high tem- 
 peratures 435 
 
 strains produced in, by straightening. . . 445 
 
 report on early steel rails 4 
 
 specifications for rails, bibliography of . . . . 494 
 American Society of Mechanical Engineers: 
 
 Bessemer process 366 
 
 electric locomotives 74 
 
 pressure of locomotive drivers 32, 62 
 
 wire tests of Professor Goss 62 
 
 American Wood Preservers' Association, 
 
 report on timber supply 113 
 
 Amsler-Laffon machine for testing hardness.. 301 
 Analysis : 
 
 converter metal 374 
 
 copper alloys used by Ball and Wingham. 331 
 
 early English rails 326 
 
 ferromanganese 374, 380 
 
 ferrosilicon 380 
 
 ferrotitanium 405 
 
 iron ores 363, 364 
 
 location of standard borings for (American 
 
 Railway Engineerirg Association) . . . 512 
 
 mixer metal 380 
 
 specification for (see Specifications). 
 
 steel, Bessemer 11, 253, 310 
 
 electric 385 
 
 manganese 333 
 
 nickel 333 
 
 open hearth 310 
 
 titanium 340, 406 
 
 Angle bars (see Joints). 
 Angle: 
 
 fishing (see Joints). 
 
 of friction of soils 314 
 
 of incision in rolling 457 
 
 of repose of various soils 314 
 
 Ann Arbor Railroad, concrete ties on 105 
 
 Annealed rails 208 
 
 Annealing, effect of, on special steels . . . 336, 337 
 
 Anthracite coal in blast furnace 361 
 
 Archduke Albrecht Steel Works at Carl- 
 
 shtitte 3 
 
 Arnold, J. O.: 
 
 causes of rupture in steel 271 
 
 influence of bismuth on copper 271 
 
 Arsenic, effect of, in steel 330 
 
 Articulated compound engine (see Articu- 
 lated engine). 
 Articulated engine: 
 
 classification of 29 
 
 dimensions of 34, 72 
 
 typical dynamic wheel loads 73 
 
 weights of 34,72 
 
 Ash, physical properties of 164 
 
 Association of American Steel Manufacturers : 
 
 chemical specifications for rails 342 
 
 drop testing machine 290 
 
 specifications for rails 342, 463 
 
 Association of Engineering Societies, Howard, 
 
 on rail failures 203 
 
 Aston, on cupro-nickel steel 332 
 
 Atchison, Topeka and Santa Fe Railway: 
 
 effect of subgrade on track 313 
 
 rail section 461 
 
 screw spikes on 141 
 
527 
 
 Atchison, Topeka and Santa Fe Railway: 
 
 speed of trains 25, 26, 27, 28 
 
 Atlanta, Birmingham and Atlantic Railroad, 
 
 size and spacing of ties 121 
 
 Atlantic City Railway, speed of trains, 
 
 23, 25, 26, 28 
 
 Atlantic Coast Line Railroad: 
 
 size and spacing of ties 121 
 
 speed of trains 24 
 
 Atlantic type engine: 
 
 allowable axle loads 322 
 
 classification of 29 
 
 description of 21 
 
 dimensions of 32, 72 
 
 effect of excess balance and angularity of 
 
 main rod 40, 70 
 
 pressure rounding curve 259 
 
 rail stresses caused by 212 
 
 speeds of 21, 24, 25, 26, 70 
 
 strength of track required for 322 
 
 typical dynamic wheel loads 72 
 
 weight of rail required for 322 
 
 weights of 32, 72 
 
 Atlas, Barrow and Dowlais, steel rails made 
 
 by 2 
 
 Austenite 427 
 
 Australia, length of rails used in 267 
 
 Austria, early steel rails 3 
 
 Austrian State Railways, hardness tests 
 
 on 302 
 
 Axle loads : 
 
 cars 85, 89 
 
 dynamic (see Dynamic). 
 
 effect of size of wheel on . 202 
 
 given in modern bridge specifications 211 
 
 increase in 15, 30 
 
 locomotives, electric 74, 79, 80 
 
 steam 31, 32, 33, 34, 72 
 
 maximum allowable, on rail 319 
 
 used on Paris, Lyons and Mediterranean 
 
 Railway 19 
 
 Axle spacing: 
 
 effect of, on load 247 
 
 s of rail for different. ..'.... 319 
 
 Back driver (see Driving wheels). 
 
 Baggage car 84 
 
 Bairstow and Stanton, experiments on re- 
 peated stress 281 
 
 Baker, Benjamin, experiments on repeated 
 
 stress 282 
 
 Bald cypress, physical properties of . . . . 164, 168 
 Baldwin Locomotive Works, progress in loco- 
 motive building 30 
 
 Ball, influence of copper on steel 331 
 
 Ball pressure tests 300 
 
 Ballast: 
 
 bearing power of 179 
 
 effect of dynamic load on 189 
 
 Cuenot's experiments on 172 
 
 depression of tie in, U. S. Government 
 
 experiments 219, 222, 233 
 
 depression of track in 172, 176, 189 
 
 depth required for different classes of 
 
 track 317 
 
 distribution of load to subgrade 180, 185 
 
 experiments on, in Germany 180 
 
 frozen, effect of, on bending moment of 
 
 rail 229,288 
 
 frozen, tests on, by U. S. Government 225 
 
 influence of kind of, on stresses in rail .... 229 
 
 Pennsylvania Railroad tests 184 
 
 weights of 316 
 
 Baltic fir (see also Fir), force required to pull 
 
 spike from 140 
 
 Baltimore and Ohio Railroad: 
 
 articulated compound engine 34 
 
 chrome-nickel rails on 333 
 
 electric locomotives 79 
 
 impact tests on bridges 64 
 
 ore docks 357 
 
 rail section 461 
 
 size and spacing of ties 121 
 
 speed of trains 24, 28 
 
 titanium rails on 339 
 
 use of scleroscope on 298 
 
 Bangor and Aroostook Railroad, size and 
 
 spacing of ties 121 
 
 Barnes, rail pressures of locomotive driving 
 
 wheels 32 
 
 Base rail: 
 
 broken (see Broken base). 
 
 indirect pressure used in rolling. . . . 457 
 
 principles governing design of 16 
 
 strength of steel from 306 
 
 wheel: 
 
 of cars 85, 89 
 
 of electric locomotives 79, 80 
 
 of steam locomotives 32, 33, 34, 72 
 
 Basquin, O. H., endurance tests 280 
 
 Bavarian State Railways: 
 
 arrangement of joints on 144 
 
 data on track 218 
 
 spikes used on 144, 145 
 
 tie plate 128, 132 
 
 weight of rail on 144 
 
 Bearing power (see Supporting power). 
 
 Beck, effect of cold on rails 285 
 
 Beech: 
 
 force required to pull spike from 140, 152 
 
 ties, amount purchased in the United 
 
 States 156 
 
Beech ties, cost of 156 
 
 on French Eastern Railway .... 117 
 Belgian State Railways, rail and tie plate, 125, 126 
 Bell, Sir I. Lowthian: 
 
 cost of transporting ore 349 
 
 depression of rails at different speeds 190 
 
 early iron rails 1 
 
 Bending moment: 
 
 maximum for different loadings of rail. . . . 320 
 
 rail, calculation of 239 
 
 determined by Love 241 
 
 effect of joint on 262 
 
 unequal tie pressure on . . 239, 288 
 
 proposed solutions of 210 
 
 stremmatograph tests of 236 
 
 U. S. government experiments 218 
 
 tie 171, 177, 179 
 
 tie plate 122, 170 
 
 Bending strength of wood: 
 
 effect of moisture on 169 
 
 steaming on 158 
 
 treatment on 158 
 
 in natural state 158, 164, 166, 168 
 
 Bending tests on worn rails 205 
 
 Benjamin, Professor C. H., flat spots on 
 
 car wheels 60 
 
 Berlin-Zossen line, speed of electric loco- 
 motives 27 
 
 Bessemer, Sir Henry, origin of the Bessemer 
 
 Bessemer: 
 
 comparison of, with open hearth process. . 380 
 
 description of plant 369 
 
 early rails, in America 3 
 
 invention of process 366 
 
 process, description of 370, 374 
 
 economies possible in 383 
 
 steel, analysis of 11, 253, 310 
 
 segregation in 405 
 
 strength of 309, 310 
 
 rails, amount in tracks of American 
 
 railroads 11 
 
 branding 447 
 
 ductility of 287 
 
 failures of 10 
 
 hardness determined by sclero- 
 
 scope 299 
 
 prices of 325 
 
 production of 382 
 
 rolling tests on, at Watertown 
 
 Arsenal 420 
 
 specifications for (see Specifi- 
 cations) . 
 
 Steel Works at Troy, early steel rails 4 
 
 Bessemer and Lake Erie Railroad, steel 
 
 ties on 90 
 
 Bethlehem Steel Company: 
 
 chrome nickel steel rails 333 
 
 rail mill 438, 444 
 
 shrinkage allowed at 444 
 
 teeming practice 399 
 
 use of iron with high copper 332 
 
 Bibliography : 
 
 of chemical composition 343 
 
 of impact tests 288 
 
 of joints 268 
 
 of literature on steel manufacture (see 
 Preface). 
 
 of piping and segregation 418 
 
 of rail specifications 494 
 
 Birch ties: 
 
 amount purchased in the United States. . . 156 
 
 cost of 156 
 
 Birkinbine, John, roasting iron ores 344 
 
 Bitternut hickory, physical properties of . . . . 164 
 
 Bituminous coal in blast furnace 361 
 
 Black locust, planting, for tie timber 109 
 
 Bland, J. C, stresses in the rail 211 
 
 Blast furnace: 
 
 description of 357 
 
 dry blast 360 
 
 fuel used in 361 
 
 location of plant , 345 
 
 operation of, at Gary 364 
 
 Maryland Steel Company 363 
 
 performance of, with dry blast 361 
 
 stoves 359 
 
 typical examples of 357 
 
 Bled ingots, specifications for 473 
 
 Blessing, concrete tie 104 
 
 Bloom: 
 
 elimination of defects in, early mill 
 
 practice 434 
 
 entering rolls 421 
 
 size of 438 
 
 Blooming: 
 
 passes, reduction in 437, 438 
 
 practice at American mills 438 
 
 Blowholes : 
 
 Brinell's experiments on 402 
 
 caused by iron oxide 389 
 
 description of 401 
 
 effect of fluid compression on 416 
 
 in titanium steel 406 
 
 means for preventing 401 
 
 Blum: 
 
 report on speed of trains as affecting track . 28 
 
 stresses in the rail 217 
 
 Bolt, track (see also Joints) : 
 holes, arrangement of (see Joints), 
 specifications for (. 
 drilling). 
 
529 
 
 Bolt, track, pull required to tighten 260 
 
 strength of 260 
 
 used on English bull head rail 19 
 
 Bonzano joint 264 
 
 Borings, location of standard, for chemical 
 analysis, American Railway Engineer- 
 ing Association 512 
 
 Boston and Albany Railroad: 
 
 angle bar tests 260 
 
 early steel rails on 2 
 
 track experiments on, by U. S. Govern- 
 ment 218, 225 
 
 Boston and Lowell Railroad, granite ties 
 
 on 90 
 
 Boston and Maine Railroad: 
 
 rail section 10 
 
 failures 10 
 
 size and spacing of ties 121 
 
 tests on locomotive springs 52 
 
 Boston and Providence Railway, early steel 
 
 rails on 2 
 
 Box car 81 
 
 Brakes, effect of application of, on locomo- 
 tive springs 52 
 
 Branding : 
 
 practice at different mills 447 
 
 specifications for (see also Specification in 
 
 question) 477 
 
 Breuil, P., impact tests 292 
 
 Bridge: 
 
 effect of velocity of load on 69 
 
 impact on, caused by moving train 64 
 
 specifications, impact allowed in 210 
 
 limiting weights on axles. . . 211 
 Brinell: 
 
 ball pressure tests 300 
 
 experiments on blowholes 402 
 
 hardness test 300 
 
 British Standards Committee: 
 
 bull head rails 19 
 
 flat bottom rails 19 
 
 rail sections 19 
 
 specifications for bull head railway rails . . 484 
 
 flat bottom railway rails . 488 
 
 tramway rails 19 
 
 Broken base: 
 
 classification of, American Railway Engi- 
 neering Association 10 
 
 due to defects in casting 390 
 
 photographs of typical failures 11 
 
 rail failures, six months ending April 30, 
 
 1909 10 
 
 Broken rails: 
 
 caused by defective equipment 57 
 
 classification of, by American Railway 
 
 Engineering Association 10 
 
 Broken rails: 
 
 failures, six months ending April 30, 
 
 1909 10 
 
 photographs of typical failures 11 
 
 Broken stone: 
 
 ballast (see Ballast). 
 
 weight of 316 
 
 Brown hoist machine for unloading ore 357 
 
 Brown, John, and Company: 
 
 chemical composition of rails rolled by.. . . 326 
 
 early steel rails 4 
 
 rails rolled for Ashbel Welsh 5 
 
 Brown, J. P., on use of catalpa for ties 109 
 
 Brunson concrete tie 104 
 
 "B. S." rail sections (see British Standards 
 Committee). 
 
 Buffalo, Rochester and Pittsburgh Railway, 
 
 size and spacing of ties 121 
 
 Buhrer: 
 
 composition tie 96 
 
 concrete tie 97, 105 
 
 steel tie 96 
 
 Bull head rail (see also British Standards 
 Committee) : 
 
 specifications for 484 
 
 types of 19 
 
 Burbank tie 104 
 
 Burden of blast furnace 363, 364 
 
 Bureau of Forestry (see U. S. Forest Service). 
 
 Bureau of Standards, magnetic testing of 
 
 rails 303 
 
 Burgess, on cupro-nickel steel 332 
 
 Burlington (see Chicago, Burlington and 
 Quincy Railroad). 
 
 Calcination of iron ore 344 
 
 Caledonian, Lancashire and Yorkshire Rail- 
 way, early steel rails on 3 
 
 Caledonian Railway: 
 
 rail fastenings on 147 
 
 speed of trains 28 
 
 Camber, amount of, in rails 444, 477, 482 
 
 Cambering machine 444 
 
 Cambering, specifications for (see Specifi- 
 cations). 
 Cambria Steel Company: 
 
 early steel rails 3 
 
 handling ore at 363 
 
 mixer used at 365 
 
 oil tempered joints 266 
 
 ores formerly used at 344 
 
 rail mill 438,444 
 
 shrinkage allowed at 444 
 
 teeming practice 399 
 
 tests to detect pipes 469 
 
 titanium steel rails 340 
 
530 
 
 Camden and Amboy Railroad, first use of 
 
 T-rail 6 
 
 Camden and Atlantic Railroad, speed of 
 
 trains 25 
 
 Campbell, H. H.: 
 
 experiments on nickel steel 335 
 
 rolling 430 
 
 Campbell tie 105 
 
 Canada, length of rails used in 267 
 
 Canadian Pacific Railway: 
 
 indentation test 300 
 
 rail section 461 
 
 speed of trains 24 
 
 ten-wheel type engine 33 
 
 Canadian Society of Civil Engineers: 
 
 Dutcher on indentation test 300 
 
 long ties used on Muskeg swamp 188 
 
 Capillary attraction, movement of water by, 
 
 in subgrade 314 
 
 Carbon : 
 
 content in rails: 
 
 increase in following Dudley's experi- 
 ments 328 
 
 of Bessemer steel 11, 253, 310 
 
 of open hearth steel 310 
 
 of special steel (see Special steels) . 
 specifications for (see also Specifications), 
 
 328, 342, 466 
 
 effect of, in steel 329 
 
 on ductility of rail steel 287 
 
 iron diagram 427 
 
 modification of, for low phosphorus 466 
 
 segregation of 409 
 
 Carbon dioxide, evolution of, in cooling 
 
 steel 401 
 
 Carbon-iron diagram 427 
 
 Carbon monoxide, evolution of, in cool- 
 ing steel 401 
 
 Carbonic anhydride, compression of the ingot 
 
 by means of 415 
 
 Carnegie Library of Pittsburgh, bibliography 
 
 of rail specifications 494 
 
 Carnegie Steel Company: 
 
 blast furnaces 361 
 
 chemical composition of rails 328 
 
 dry blast at 361 
 
 nickel steel rails 333 
 
 rail mill 438, 444 
 
 shrinkage allowed at 444 
 
 steel ties in tracks at Duquesne Plant 94 
 
 teeming practice 399 
 
 titanium steel rails 340 
 
 Carnegie steel tie 90 
 
 Cars: 
 
 allowable wheel loads of 322 
 
 dynamic augment of wheel pressure. ... 85, 89 
 
 Cars: 
 
 strength of track required for 322 
 
 weight of rail required for 322 
 
 weights of 85, 89 
 
 Cartesian coordinates, stress-repetition 
 
 curves drawn to 281 
 
 Cast iron: 
 
 manufacture of 357 
 
 wheel, effect of, on rail 195 
 
 Casting: 
 
 effect on steel 390, 395 
 
 fluid compression during 410, 415 
 
 the ingot 395 
 
 Casting ladle: 
 
 effect of agitation of steel in 390 
 
 example of 389 
 
 Catalpa : 
 
 force required to pull spike from 139 
 
 planting for tie timber 109 
 
 Cedar: 
 
 physical properties of 164, 168 
 
 ties, amount purchased in the United 
 
 States 154, 156 
 
 cost of 154, 156 
 
 Cedar elm, physical properties of 164 
 
 Cell (see Grain). 
 
 Cementite 427 
 
 Census, United States: 
 
 distribution of lumber products in the 
 
 United States 108 
 
 ties purchased in the United States 156 
 
 Center of gravity in locomotives 74, 259 
 
 Central of Georgia Railway, size and spacing 
 
 of ties 121 
 
 Central Railroad of New Jersey: 
 
 size and spacing of ties 121 
 
 special steel rails on 333 
 
 speed of trains 24, 28 
 
 Centrifugal force: 
 
 of counterbalance 38 
 
 of locomotive when rounding curve 259 
 
 of wheel on irregular track 45 
 
 Chair used with bull head rail 19 
 
 Chanute, Octave: 
 
 investigation of proper design of rail 6 
 
 report on steel rails 5 
 
 Chapin ore 364 
 
 Charcoal in blast furnace 361 
 
 Charge of blast furnace 363, 364 
 
 Charpy, G., impact tests 292 
 
 Chemical composition: 
 
 bibliography of recent literature on 343 
 
 comparison of early and recent, in rails. . . 328 
 
 Dudley's formula 327 
 
 effect on physical properties of steel (see 
 Element in question) . 
 
Chemical composition : 
 
 effect on rail breakages in coll weather . . . 286 
 form for reporting, American Railway En- 
 gineering Association 503 
 
 increase in hardening constituents in rails . 328 
 location of standard borings for analysis, 
 American Railway Engineering 
 
 Association 512 
 
 of Bessemer steel 11, 253, 310 
 
 of different steels (see Analysis). 
 
 of early rails 326 
 
 of joints 234, 266 
 
 of open hearth steel 310 
 
 of special steels (see Special steels), 
 specifications for, in rails (see also Specifi- 
 cation in question) 328, 342, 465 
 
 Chemins de Fer d'Orleans (see Orleans Rail- 
 way) . 
 Chemins de Fer de l'Est (see French Eastern 
 
 Railway). 
 Chemins de Fer de l'Etat (see State Railways 
 
 of France). 
 Chemins de Fer de l'Ouest (see Western Rail- 
 way of France). 
 Chemins de Fer de Paris a Lyon et a la Medi- 
 terranee (see Paris, Lyons and Medi- 
 terranean Railway). 
 Chemins de Fer du Midi (see Southern Rail- 
 way of France) . 
 Chemins de Fer du Nord (see Northern Rail- 
 way of France). 
 
 Chenoweth concrete tie 104 
 
 Chesapeake and Ohio Railway articulated 
 
 compound engine 34 
 
 Chestnut : 
 
 force required to pull spike from 139 
 
 ties, amount purchased in the United 
 
 States 154, 156 
 
 cost of 154, 156 
 
 Chicago and Alton Railroad: 
 
 concrete ties on 99, 105 
 
 early steel rails on 2 
 
 Chicago and Eastern Illinois Railroad, size 
 
 and spacing of ties 121 
 
 Chicago and North Western Railway: 
 
 concrete ties on 105 
 
 early steel rails on 2, 4 
 
 new rail section 462 
 
 size and spacing of ties 121 
 
 speed of trains 24, 26, 28 
 
 ten-wheel type engine 33 
 
 Chicago, Burlington and Quincy Railroad : 
 
 concrete ties on 104 
 
 plantations for tie timber Ill 
 
 prairie type engine 33 
 
 rail section 10, 461 
 
 Chicago, Burlington and Quincy Railroad : 
 
 rail section failures 10 
 
 size and spacing of ties 121 
 
 speed of trains 24, 25, 26, 27 
 
 track experiments 219 
 
 Chicago Great Western Railroad, size and 
 
 spacing of ties 121 
 
 Chicago, Indiana and Southern Railroad : 
 
 consolidation type engine 34 
 
 size and spacing of ties 121 
 
 Chicago, Indianapolis and Louisville Rail- 
 way, size and spacing of ties 121 
 
 Chicago Junction Railway, concrete ties on. . 104 
 
 Chicago, Milwaukee and Puget Sound Rail- 
 way, size and spacing of ties 121 
 
 Chicago, Milwaukee and St. Paul Railway: 
 
 excess pressure of driving wheels 36 
 
 fiat spot in engine wheel 56 
 
 Pacific type engine 32 
 
 size and spacing of ties 121 
 
 speed of trains 25 
 
 Chicago, Rock Island and Pacific Railway: 
 
 Atlantic type engine 32 
 
 early steel rails on 2 
 
 size and spacing of ties 121 
 
 Chicago, St. Paul, Minneapolis and Omaha 
 
 Railway, size and spacing of ties. . . . 121 
 
 Chrome-nickel steel rails: 
 
 production of 341 
 
 use of 333 
 
 Chromium: 
 
 effect of, in steel 334 
 
 in combination with 
 
 nickel 335 
 
 steel rails, production of 341 
 
 Cincinnati, Hamilton and Dayton Railway, 
 
 size and spacing of ties 121 
 
 Cinder ballast (see Ballast). 
 
 City cars 88 
 
 Clamer, copper and nickel in steel 332 
 
 Clary tie plate 122 
 
 Classification: 
 
 of locomotives 29 
 
 of track 317 
 
 Whyte's system 29 
 
 Clausen, L. R., example of flat wheels 56 
 
 Clay: 
 
 angle of friction of 314 
 
 bearing power of 187 
 
 Cleveland, Cincinnati, Chicago and St. 
 Louis Railway: 
 
 plantations of tie timber 112 
 
 size and spacing of ties 121 
 
 Cleveland, Painesville, Ashtabula Railway, 
 
 car used on 88 
 
532 
 
 Coal: 
 
 cars (see also Cars), effect of heavy, on 
 track 82, 
 
 used in blast furnace 
 
 Coefficient : 
 
 of ballast 
 
 of friction of soils 
 
 of slip in driving wheels 198, 
 
 of yielding in ballast on German railways. 
 Coes and Howard, experiments on rocking 
 
 of engine 
 
 Cogging mill 437, 
 
 Cogging rolls used in Puppe's tests 448, 
 
 Coke car 
 
 Coke used in blast furnace 361, 363, 
 
 Colby, A. L., experiments on nickel steel. . . . 
 Cold: 
 
 defective tires due to 
 
 effect of, on strength of rail 
 
 rolling of head of rail caused by wheel 
 
 rolling of rails i 
 
 shortness in steel I 
 
 straightening press < 
 
 straightening rails 4 
 
 work, effect of, on structure of steel . 424, i 
 
 Collet trenail 148, ] 
 
 Colorado and Southern Railway, size and 
 
 spacing of ties ] 
 
 Colorado Fuel and Iron Company: 
 
 rail mill 438,' 
 
 shrinkage allowed at l . 
 
 Columbia type engine, classification of 
 
 Composite ties used in Cuenot's experi- 
 ments ] 
 
 Compression : 
 
 fluid, of ingot 410, < 
 
 modulus of, at point of contact of wheel 
 
 and rail 194, ] 
 
 effect of bearing surface on.. . 1 
 strength of special steels (see Steel in 
 question) . 
 
 strength of steel ] 
 
 effect of chemical composition on (see 
 Element in question). 
 
 strength of wood 158, 165, 167, 168, 1 
 
 effect of treatment on ] 
 
 Concrete ties: 
 
 (see also Tie in question) 97, 1 
 
 service tests on ] 
 
 Cone pressure tests i 
 
 Conical tires 
 
 Connecticut River Railroad, early steel rails 
 
 on 
 
 Consolidation type engine: 
 
 allowable axle loads c 
 
 Consolidation type engine: 
 
 classification of 29 
 
 dimensions of 34, 72 
 
 effect of excess pressure and angularity of 
 
 main rod 44, 70 
 
 pressure rounding curve 259 
 
 rail stresses caused by 228 
 
 speeds of 70 
 
 strength of track required for 322 
 
 tests on driving wheel springs 52 
 
 typical dynamic wheel loads 73 
 
 U. S. Government experiments with 228 
 
 weight of rail required for 322 
 
 weights of 34, 72 
 
 Continuous : 
 
 girder, rail as 189, 240, 247 
 
 joint 264 
 
 process of making steel 375 
 
 rails 267 
 
 record of rail failures, form for reporting 
 
 graphically 523 
 
 Conversion of iron into steel 366 
 
 Converter (see Bessemer). 
 
 Converter metal, analysis of 374 
 
 Cooling curve of different substances. . . 425, 426 
 
 Coombs, R. D., on concrete ties 104 
 
 Cooper, Hewitt and Company, first open 
 
 hearth furnace in America 375 
 
 Copper: 
 
 cooling curve of 425 
 
 effect of, in steel 331 
 
 granular structure 270 
 
 Cornwall irons, high copper in rails made 
 
 from 332 
 
 Corrugations of rails 209 
 
 Cort, early methods of making steel 366 
 
 Cost: 
 of Bessemer compared to open hearth 
 
 process 383 
 
 of forest land . . . . ' 114 
 
 of plantations for tie timber Ill 
 
 of rails, 1855 to 1910 325 
 
 ferrotitanium 341 
 
 rerolling 459 
 
 of ties, annual charge of 115 
 
 treated 115 
 
 in the United States 154, 156 
 
 of track of German railways 218 
 
 with screw spikes 142 
 
 of transporting ore 349, 352 
 
 Coiiard, depression of ties in ballast.. . . 172, 190 
 
 Counterbalance pressure : 
 
 absence of, in electric locomotives 78 
 
 amount of, for different types of 
 
 engines 35, 70 
 
 calculation of 35 
 
533 
 
 Counterbalance pressure: 
 
 effect of inertia of track on stresses pro- 
 duced by 69 
 
 on bridges 68 
 
 tire wear 202 
 
 Professor Goss' experiments on 62 
 
 Cow oak, physical properties of 164 
 
 Crandall, Professor, experiments on steel 
 
 rollers, on steel plates 194 
 
 Creeping of rails 153 
 
 Creosote treatment (see Treated Ties). 
 
 Crescent breaks in flange: 
 
 effect of casting on 390 
 
 examples of 11 
 
 Creusot rails, effect of cold on 285 
 
 Creusot, steel works, early steel rails 3 
 
 Critical point, effect of rolling below 427 
 
 Crop (see Discard). 
 
 Cross ties (see Ties). 
 
 Crushed head: 
 
 classification of, American Railway En- 
 gineering Association 10 
 
 effect of casting on 390 
 
 investigation of 391 
 
 photographs of typical failures 11 
 
 rail failures, six months ending April 30, 
 
 1909 10 
 
 Crushing strength (see Compression). 
 
 Crystal (see Grain). 
 
 Cuban pine, physical properties of 164 
 
 Cuenot, G.: 
 
 advance wave of rail 223 
 
 experiments on ties 172 
 
 profile of rail 46 
 
 Cupola for melting pig iron 366 
 
 Cupro-nickel steel 332 
 
 Curve: 
 
 comparison of rail failures on, with tangent. 10 
 
 elastic, of rail 241, 242 
 
 of tie 172, 176, 177 
 
 horizontal pressure exerted by engine when 
 
 rounding 259 
 
 relation of coning of wheel to 7 
 
 Cushing, W. C: 
 
 design of screw spike proposed by 147 
 
 discard of ingot 417 
 
 discussion of screw spikes 142 
 
 on ties (translation of M. Cuenot 's 
 
 experiments) 172 
 
 Cylinder: 
 
 compression modulus of surface of 193 
 
 size of, on modern steam locomotives, 
 
 40, 41, 42, 43, 44 
 
 Cylindrical tires 6 
 
 Cypress : 
 
 force required to pull spike from 140 
 
 Cypress: 
 
 physical properties of 164, 168 
 
 ties, amount purchased in the United 
 
 States 154, 156 
 
 cost of 154, 156 
 
 De Paris a Lyon et a la Mediterranee (see 
 Paris, Lyons and Mediterranean Rail- 
 way). 
 Decapod engine: 
 
 classification of 29 
 
 example of 31 
 
 Defective equipment: 
 
 broken rails caused by 57 
 
 due to excessive loads . . • 85 
 
 small diameter of wheel 202 
 
 effect of, on track 57 
 
 examples of long flat spots in wheels ... 56, 57 
 lack of roundness in chilled car wheels. ... 58 
 Defective rails: 
 
 classification of, by American Railway 
 
 Engineering Association 10 
 
 for six months ending April 30, 1909 10 
 
 forms for reports and records of, American 
 
 Railway Engineering Association 506, 512 
 forms for reports and records of, chart for . 523 
 forms for reports and records of, used on 
 
 Harriman Lines 524 
 
 on American railroads 10 
 
 photographs of typical failures 11 
 
 Defective wheels (see Defective equipment). 
 Deflection: 
 
 in drop test, specifications for 470, 486 
 
 of driving wheel springs 48, 49, 53 
 
 of head of rail under eccentric load . . 255, 257 
 of rail, comparison of worn and un- 
 worn 204, 205 
 
 in drop test 292, 294, 470, 486 
 
 in track, amount of 190, 233 
 
 in track, effect of, on stress in 
 
 rail 243 
 
 Deflectometer used by Turneaure in impact 
 
 tests 64 
 
 Delaware and Hudson Company: 
 
 articulated compound engine 34 
 
 consolidation type engine 34 
 
 80-lb. rails used by, in 1893 7 
 
 management of timber lands 113 
 
 size and spacing of ties 121 
 
 Delaware, Lackawanna and Western Rail- 
 road : 
 
 Mogul type engine 34 
 
 plantations of tie timber Ill 
 
 size and spacing of ties 121 
 
 speed of trains 25 
 
 ten-wheel type engine 33 
 
534 
 
 De L'Est (see French Eastern Railway). 
 
 De L'Etat (see State Railways of France). 
 
 De L'Ouest (see Western Railway of France). 
 
 Denver and Rio Grande Railroad, size and 
 
 spacing of ties 121 
 
 Denver, North Western and Pacific Railway, 
 
 ten-wheel type engine 33 
 
 Depression (see also Deflection) : 
 
 of tie in ballast, amount of 172, 176, 189 
 
 effect of, on stress in rail. 243 
 of track, U. S. Government, experiments 
 
 on 218 
 
 Design : 
 
 of rail 6, 17, 458 
 
 of rolls 457 
 
 of track 323 
 
 Detroit River Tunnel Company's loco- 
 motive 75, 80 
 
 Dickerson, S. K., tests on chilled car 
 
 wheels 57 
 
 Dining car 84 
 
 Direct pressure in rolling 456 
 
 Direct process of making steel 365 
 
 Discard : 
 
 amount of, necessary 417 
 
 report of American Railway Association 
 
 on 416 
 
 required on compressed ingots 415 
 
 specifications for (see also Specification 
 
 in question) 473 
 
 Docks, ore 348, 350, 357 
 
 Dolomite used in blast furnace 363 
 
 Dominion Iron and Steel Company: 
 
 rail mill 438, 444 
 
 shrinkage allowed at 444 
 
 D'Orleans (see Orleans Railway). 
 
 Double-headed rail: 
 
 examples of 19 
 
 specifications for 484 
 
 Douglas fir (see also Fir), physical proper- 
 ties of 166, 168 
 
 Douglas spruce, physical properties of 164 
 
 Dowel: 
 
 increased resistance due to use of, in ties. 152 
 use of, with screw spike 142, 150 
 
 Drainage : 
 
 effect of water on gravel ballast 187 
 
 necessity of, for subgrade 314 
 
 Draw bar: 
 
 constant pull on, of electric locomo- 
 tives 78 
 
 effect of, on pressure of drivers 71 
 
 pull in Mallet locomotive 22 
 
 Drill test for hardness 303 
 
 Drilling, specifications for, in rails (see also 
 
 Specification in question) 477 
 
 Driving wheels (see also Axle and Wheel) : 
 
 coefficient of slip of 198, 199 
 
 effect of position of, on allowable load .... 322 
 increase in pressure on rail due to excess 
 
 balance and angularity of main rod. . 35, 70 
 increase in pressure on rail due to irregu- 
 larities in the track 45, 71 
 
 increase in pressure on rail due to rocking 
 
 of engine 54, 71 
 
 increase in weight on 15, 30 
 
 springs (see Springs). 
 
 weights on, electric locomotives 79, 80 
 
 steam locomotives . . 31, 32, 33, 34 
 
 wheel base, electric locomotives 79, 80 
 
 steam locomotives . . 32, 33, 34, 72 
 Drop test: 
 
 bibliography of 288 
 
 comparison of deflections obtained with 
 
 different machines 291 
 
 deflections obtained in 291, 294 
 
 description of recent machines 289 
 
 energy dissipated in 293 
 
 form for reporting results of, American 
 
 Railway Engineering Association .... 505 
 
 losses in 293 
 
 machine, standard 290 
 
 measurement of ductility in 287 
 
 specifications for 290 
 
 specifications for (see also Specification 
 
 in question) 470 
 
 theoretical considerations of 293 
 
 Dry blast (see Gay ley). 
 Ductility: 
 
 in rail, specifications for 470, 480 
 
 steel 287 
 
 in special steels (see Steel in question). 
 of steel, effect of chemical composition 
 on (see Element in question). 
 
 effect of cold on 286 
 
 effect of titanium on 287 
 
 Dudley, C. B.: 
 
 proposed formula for chemical composi- 
 tion of rails 327 
 
 tests on relation between chemical com- 
 position and wearing of rails 326 
 
 Dudley, P. H.: 
 
 casting steel 388 
 
 depression of track in ballast 190, 213 
 
 design of new section with large fillet 462 
 
 ductility in rail steel 286 
 
 dynamic augment to wheel load 212 
 
 effect of cold on rails 287 
 
 draw bar pull on wheel pressure . . 71 
 
 friction in splice bars 261 
 
 grains per square inch in rail steel 424 
 
 joint 265 
 
Dudley, P. H.: 
 
 lettering rails from ingot 416 
 
 reheating furnace, effect of, on ingot 400 
 
 section of rail 10, 462 
 
 failures 10 
 
 stremmatograph tests 212, 236 
 
 tonnage service of wheels and rails 202 
 
 wear of rails 328 
 
 Duguet, Captain, impact tests 292 
 
 Duluth and Iron Range Railroad, ore 
 
 dock 348 
 
 Duluth, Mesabi and Northern Railway, ore 
 
 docks 351 
 
 Duluth, South Shore and Atlantic Railway, 
 
 size and spacing of ties 121 
 
 Dumas, report on nickel steel 336 
 
 Du Midi (see Southern Railway of France). 
 
 Dummy pass 443, 444 
 
 Du Nord (see Northern Railway of France). 
 
 Duplex process for making steel 388 
 
 Duquesne joint 264 
 
 Dutcher, indentation test 300 
 
 Dynamic : 
 augment of wheel load: 
 
 amount of, for cars 85, 89 
 
 for electric locomotives .... 78 
 for steam locomotives. ... 71, 72 
 
 assumed by Bland 211 
 
 Dudley 212 
 
 causes of 32 
 
 due to excess pressure of counter-bal- 
 ance and angularity of main rod 35 
 
 flat spots in wheels 54 
 
 impact 68 
 
 irregularities in track 45 
 
 rocking of engine 54, 71 
 
 velocity of load 70 
 
 load, effect of, on driving wheel springs ... 54 
 
 track 189 
 
 tests on ties 175, 190 
 
 typical, load diagrams for cars 85, 89 
 
 electric engines. 78 
 steam engines. 72, 73 
 
 wheel load allowable for 1004b. rail 247 
 
 wheel loads for different weights of rail 
 
 and axle spacing 319 
 
 E-60 loading 211 
 
 Earth: 
 
 angle of friction of 314 
 
 bearing power of 313 
 
 East Coast Railway of England, speed of 
 
 trains 25, 28 
 
 Edgar Thomson Works: 
 
 early use of mixer at 365 
 
 fluid compression of the ingot 415 
 
 Egyptian State Railways, rail section 19 
 
 Eight- wheel type engine: 
 
 classification of 29 
 
 coupled, classification of 29 
 
 description of 21 
 
 effect of excess balance and angularity of 
 
 main rod 35 
 
 rail stresses caused by 219, 235, 236 
 
 stremmatograph tests with 236 
 
 U. S. Government experiments with.. 219, 228 
 
 wear of tires 199 
 
 El Cuero ore used at Maryland Steel 
 
 Company 363 
 
 El Paso and Southwestern System, size and 
 
 spacing of ties 121 
 
 Elastic curve: 
 
 calculation of, for 100-lb. rail 242 
 
 of rail, calculation of, by Love 240 
 
 of ties 172, 176, 177 
 
 Elastic limit: 
 
 definition of 310 
 
 effect of repeated loads on 311 
 
 effect on structure of metals of straining 
 
 beyond 273 
 
 necessity for high, in rail head 208 
 
 of rail at point of contact with wheel . . 194, 195 
 
 of rail steel 306 
 
 relation of breaking strength to, under 
 
 repeated stress 273, 278 
 
 relation of, to working load 312 
 
 of special steels (see Steel in question). 
 Elasticity, modulus of: 
 
 of steel 225, 241 
 
 of wood 158, 166, 168 
 
 Electric: 
 
 cars 87, 88, 322 
 
 furnace: 
 
 description of 383 
 
 steel from, analysis of 385 
 
 production of 341 
 
 strength of 386 
 
 locomotives: 
 
 comparison of, with steam 74 
 
 dynamic augment to wheel load (see 
 Dynamic). 
 
 general characteristics of 79, 80 
 
 pressure of, rounding curve 257 
 
 speeds of 27, 29, 79, 80 
 
 typical dynamic load diagrams 78 
 
 magnet for loading rails 446 
 
 motor for rolling mills: 
 
 advantages of 459 
 
 use of, at Gary 437 
 
 railway : 
 
 cars 87, 88, 322 
 
 corrugations in rails 209 
 
536 
 
 Electric : 
 railway: 
 
 cross-ties purchased by 154, 156 
 
 rails, specifications for 491 
 
 use of manganese in 339 
 
 used on 19 
 
 roaring rails 209 
 
 weight of rail for various axle loads .... 322 
 
 welded joints on 267 
 
 steel rails, production of 341 
 
 Elgin, Joliet and Eastern Railway, concrete 
 
 ties on 105 
 
 Elm, physical properties of 164 
 
 Elm ties: 
 
 amount purchased in the United States.. . 156 
 
 cost of 156 
 
 Elongation: 
 
 effect of chemical composition on (see 
 Element in question). 
 
 effect of size of grain on 424 
 
 in electric steel 386 
 
 in rail, specifications for 470, 480 
 
 in rail steel 306 
 
 determined in drop testing machine .... 287 
 
 in rolling, Puppe's tests 456 
 
 in special steels (see Steel in question). 
 
 under repeated stress 280 
 
 Elsass-Lothringen State Railways: 
 
 arrangement of joints on 144 
 
 rail and tie plate 130 
 
 spikes used on 144, 145 
 
 weight of rail on 144 
 
 Empire State Express, speed of 23 
 
 Engineering Standards Committee (London) : 
 
 bull head rail 19 
 
 flat bottom rail 19 
 
 length of rails 268, 485, 489 
 
 rail sections 19 
 
 specifications for bull head railway rails. 484 
 
 flat bottom railway rails . 488 
 
 tramway rail 19 
 
 Engineers' Club of St. Louis, Johnson on 
 
 compression moduli 193 
 
 Engines (see Locomotive). 
 England (see English). 
 English: 
 
 chemical composition of early steel rails . . 326 
 
 early steel rails 2, 326 
 
 iron rails, life of 2 
 
 prices of 325 
 
 length of rails 268 
 
 practice in rolling rails 434 
 
 rail fastenings 19, 147 
 
 rail mills, hot straightening at 446 
 
 reheating furnace 398 
 
 screw spikes 19, 146, 147 
 
 English : 
 
 specifications for rails 484, 488 
 
 speeds of railway trains 24, 25, 28 
 
 steel rails, prices of 325 
 
 used in America 2, 4 
 
 three-high rail mill 437 
 
 trenails used on, railways 19, 146, 147 
 
 types of rails 19 
 
 weights of rails used on, railways 19 
 
 Erakoff, effect of cold on rails 285 
 
 Erie Railroad: 
 
 chrome-nickel rails on 333 
 
 early steel rails on 4 
 
 size and spacing of ties 121 
 
 speed of trains 25, 26 
 
 Europe: 
 
 rail fastenings used in 144 
 
 rails used in 19, 125 
 
 types of tie plates 125 
 
 Eutectoid steel 427 
 
 Ewing, J. A., experiments on repeated stress 273 
 Excess balance (see Counterbalance pressure). 
 
 Exhausted metal in head of rail 204 
 
 Extensometer used in Turneaure's impact 
 
 tests 64 
 
 Extreme fiber stress: 
 
 allowable in rail 312 
 
 bending moments corresponding to 248, 320 
 
 in 100-lb. rail, calculation of 239 
 
 in rail, determined by Bland 211 
 
 Freeman 211 
 
 Government tests. . 218 
 
 Selby 210 
 
 on German railways." 218 
 
 shown by stremmatograph. . . . 212, 236 
 in tie 171, 179 
 
 Factor of safety : 
 
 amount of, in different structures 311 
 
 in rails 312 
 
 Failures: 
 
 rail (see Defective rails). 
 
 wheel (see Defective equipment). 
 
 Farlington forest 110 
 
 Fastenings, rail (see Joints). 
 Fatigue of metal: 
 
 in head of rail 204, 284 
 
 under repeated stress 270 
 
 Faustmann's formula for productivity of 
 
 woodlands 114 
 
 Fay, Henry, manganese sulphide in steel. . . 390 
 Felt: 
 
 tie plate on French Eastern Railway 132 
 
 London and North Western 
 Railway 19 
 
Felton, S. M., increase in loads on engine 
 
 drivers 30 
 
 Feodossieff, effect of cold on rails 285 
 
 Ferrite 427 
 
 Ferromanganese (see also Recarbonizing) : 
 
 amount added to recarbonize 374, 380 
 
 analysis of 374, 380 
 
 manganese in steel due to use of 330 
 
 Ferrosilicon 380 
 
 Ferrotitanium (see Titanium). 
 
 Ferrule, oak, for spike 19 
 
 F6ry pyrometer 434 
 
 Fiber stress (see Extreme fiber stress) . 
 Finishing : 
 
 pass 437, 438 
 
 specifications for, in rails (see also Specifi- 
 cation in question) 477 
 
 Finishing temperature : 
 
 effect of, in rolling 430 
 
 in A.S.C.E. rails 7, 435 
 
 in English rails 435 
 
 specifications for (see also Specification 
 
 in question) 475 
 
 Fir: 
 
 force required to pull spike from 140 
 
 physical properties of 166, 168 
 
 ties, amount purchased in the United 
 
 States 154, 156 
 
 cost of 154, 156 
 
 Fish bolt (see Bolt). 
 Fishing angles of rail: 
 
 determined by Chanute 6 
 
 principles of design of 6, 17 
 
 Fishplate (see Joints). 
 
 Fissures, transverse, in head of rail 203 
 
 Flange, rail (see Base, rail). 
 Flat bottom railway rails: 
 
 specifications for 488 
 
 types of 19 
 
 Flat car 82 
 
 Flat spots in wheels : 
 American Railway Engineering Asso- 
 ciation, experiments with 61 
 
 broken rails caused by 57 
 
 effect of inertia of track on stresses pro- 
 duced by 68 
 
 example of, on the Chicago, Milwaukee 
 
 and St. Paul Railway 56 
 
 excess pressure caused by 54 
 
 M. C. B., rule for length of 56 
 
 Professor Benjamin's apparatus for testing 60 
 Professor Hancock's mathematical in- 
 vestigation of 55 
 
 Florida East Coast Railway: 
 
 concrete ties on 105 
 
 size and spacing of ties 121 
 
 Flow of metal in head of rail: 
 
 classification of, by American Railway 
 
 Engineering Association 10 
 
 effect of, on bending properties of rail .... 204 
 
 wheel load on 205 
 
 photographs of typical failures 11 
 
 rail failures, six months ending April 
 
 30, 1909 10 
 
 Flux used in blast furnace 361 
 
 Flywheel of rolling mill, energy in 459 
 
 Foppl, experiments on repeated stress 282 
 
 Force, H. J., effect of copper on steel 332 
 
 Forest: 
 
 original, of the United States 106 
 
 wasteful cutting of, for ties 116 
 
 Forest Service (see U. S. Forest Service). 
 Forestry: 
 
 application of methods to growing tie 
 
 timber 114 
 
 Faustmann's formula 114 
 
 original forests in the United States 106 
 
 plantations for ties 109 
 
 Forney coupled engine, classification of. . . . 29 
 
 Forney, M. N., report on steel rails 5 
 
 Forsyth, Robert: 
 
 phosphorus in rail steel 465 
 
 piping of ingots 399 
 
 transferring ladle 390 
 
 Four-wheel engine: 
 
 classification of 29 
 
 coupled, classification of 29 
 
 Fowler, G. L.: 
 
 effect of repeated stress on rails 278 
 
 experiments on contact between wheel 
 
 and rail 195 
 
 France (see French). 
 Freeman, F. B.: 
 
 reaction of tie in ballast 191 
 
 stresses in the rail 211 
 
 Freight car (see Cars). 
 
 Freight locomotives (see Locomotives). 
 
 French: 
 
 early screw spikes 141 
 
 steel rails 3 
 
 rail fastenings 143, 144 
 
 screw spikes 143, 144 
 
 speeds of railway trains 24 
 
 tie plates 132 
 
 tie plug 149 
 
 types of rails 18 
 
 weights of rail used on, railways 18, 144 
 
 French Eastern Railway: 
 
 arrangement of joints on 144 
 
 early screw spikes used on 141 
 
 half-round ties on 117 
 
 screw spikes on 143, 144 
 
538 
 
 French Eastern Railway: 
 
 section of rail 18 
 
 tie plate 132 
 
 tie plug 149 
 
 weight of rail on 18 
 
 French Government, fluid compression of 
 
 ingots required by 415 
 
 Friction, coefficient of, for earth and gravel . 314 
 in joints: 
 
 effect of, on rail stresses 262 
 
 tests on 260 
 
 of soils, angles of 314 
 
 Front driver (see Driving wheels). 
 Frost (see also Cold): 
 
 effect of, on depression of track 233 
 
 on rail breakages 288 
 
 Ft. Worth and Denver City Railway, size 
 
 and spacing of ties 121 
 
 Fuel used in blast furnace 361 
 
 Furnace: 
 blast (see Blast furnace). 
 
 cupola 365 
 
 electric (see Electric). 
 
 open hearth (see Open hearth). 
 
 Gagging, effect of, on rail 445 
 
 Gagging press 445 
 
 Galvanized screw spike 19 
 
 Galveston, Harrisburg and San Antonio 
 Railway: 
 
 concrete ties on 105 
 
 size and spacing of ties 121 
 
 plantations of tie timber Ill 
 
 Gary: 
 
 blast furnaces at 358, 361 
 
 general arrangement of plant 361 
 
 open hearth furnaces at 378 
 
 rolling mill practice 437, 438, 444 
 
 shrinkage allowed at 444 
 
 soaking pits at 397 
 
 strength of steel from 306 
 
 teeming practice 399 
 
 Gas electric car 86 
 
 Gas in molten steel 366, 401 
 
 Gay ley dry blast : 
 
 at Bessemer converter 389 
 
 description of, for blast furnace 360 
 
 General electric gas electric car 86 
 
 Georgia Railroad, size and spacing of ties . . . 121 
 German : 
 
 electric steel 385 
 
 experiments on ballast 180 
 
 rolling mills 447 
 
 rail fastenings 144, 145 
 
 rails 19 
 
 German: 
 
 screw spikes 144, 145 
 
 speeds of electric locomotives 27, 29 
 
 tests on tie plates 133 
 
 tie plates 125 
 
 track, data on 218 
 
 weights of rail used on, railways 19, 144 
 
 Germany (see German). 
 
 Gibbs, George, electric locomotives 74 
 
 Girder rail (see Street railway). 
 
 Goldie tie plate 122 
 
 Gondola car 82 
 
 Goss, Professor, tests on counterbalance 
 
 pressure 62 
 
 Grab bucket for unloading ore 357 
 
 Grade, effect of changes in, on wheel pres- 
 sure 45 
 
 Grading ore 344 
 
 Grain: 
 
 changes in, under repeated stress 273 
 
 effect of rolling on 427 
 
 on strength of steel 272, 424 
 
 temperature on 427 
 
 number in rail steel 424 
 
 size of, in head of rail 392, 394, 424, 431 
 
 structure of different metals 270 
 
 Grand Duchy of Baden State Railways, early 
 
 screw spike used on 140 
 
 Grand Rapids and Indiana Railway, size 
 
 and spacing of ties 121 
 
 Grand Trunk Railway: 
 
 effect of cold on steel rails 4 
 
 electric locomotives 80 
 
 size and spacing of ties 121 
 
 Granite ties 90 
 
 Granular structure (see Grain). 
 
 Gravel : 
 
 angle of friction of 314 
 
 ballast (see Ballast). 
 
 weight of 316 
 
 Gravity: 
 
 center of, in steam and electric locomo- 
 tives 74 
 
 effect of, on wheel on irregular track 47 
 
 specific (see Specific gravity) . 
 
 Great Britain (see English). 
 
 Great Central Railway (of England), speed 
 of trains 28 
 
 Great Eastern Railway (of England), rail 
 fastenings on 147 
 
 Great Lakes (see Transportation). 
 
 Great Northern Railway: 
 
 accident on, at Sharon, N. D 325 
 
 electric locomotives 80 
 
 ore dock 350 
 
 size and spacing of ties 121 
 
539 
 
 Great Northern Railway (of England) : 
 
 rail fastenings on 147 
 
 speed of trains 24, 28 
 
 Green ash, physical properties of 164 
 
 Grooved rail: 
 
 examples of 19 
 
 specifications for 491 
 
 Guerhard, effect of cold on rails 285 
 
 Gum: 
 
 physical properties of 164, 170 
 
 plantations of, for tie timber 112 
 
 ties, amount purchased in the United 
 
 States 156 
 
 cost of 156 
 
 Hadfield: 
 
 experiments on nickel steel 334 
 
 manganese steel 336 
 
 Half-round tie 116 
 
 Hancock, Professor, flat spots in wheels .... 55 
 
 Hansen steel tie 94 
 
 Harbord, influence of arsenic on steel 331 
 
 Hardness tests: 
 
 Brinell's ball test 300 
 
 cone pressure test 301 
 
 indentation test 300 
 
 Keep drill test 303 
 
 sclerometer, of Turner 302 
 
 scleroscope 298 
 
 Hardy catalpa: 
 
 force required to pull spike from 139 
 
 use of, for tie plantations 109, 112 
 
 Harmet process for compression of ingot 411, 415 
 
 Harrell tie 104 
 
 Harriman Lines: 
 
 chart of rail failures 524 
 
 specifications for rails 463 
 
 Harrison, T. E., iron rails, life of 1 
 
 Hatt, W. K.: 
 
 impact tests 292 
 
 tests on strength of timber 157, 169 
 
 tie plates 169 
 
 Hawaiian Ohia ties: 
 
 amount purchased in the United States.. . 156 
 
 cost of 156 
 
 Head: 
 
 bull, rail, examples of 19 
 
 specifications for 484 
 
 double, rail (see Bull head rail), 
 rail: 
 
 crushed (see Crushed head). 
 
 effect of casting on 390 
 
 flat spot in wheel on 56 
 
 fatigue of metal in 203, 284 
 
 flow of metal in (see Flow of metal) . 
 
 principles governing design 16 
 
 392, 394, 424, 431 
 
 Head rail: 
 
 size of grain in 
 
 split (see Split head), 
 strength of, experiments on, at Mary- 
 land Steel Company . . . 
 
 steel from 
 
 stress at point of contact with wheel . 
 
 thermal cracks in 
 
 transverse fissures in 
 
 unsound metal in 
 
 205 
 203 
 
 sweep 
 
 Heath, Josiah, early experiments with open 
 
 hearth furnace 
 
 Heating curve of steel 
 
 Hecla Belt Line, concrete ties on 
 
 Hemlock : 
 
 physical properties of 166, 
 
 ties, amount purchased in the United 
 
 States 154, 
 
 cost of 154, 
 
 Hennebique, concrete tie 
 
 Heroult electric furnace 
 
 Hickey tie 
 
 Hickory, physical properties of 
 
 High T-rails: 
 
 examples of 
 
 specifications for 
 
 Hill fastening on Carnegie steel tie 
 
 Hiroi, I., web stresses 
 
 Hocking Valley Railroad, size and spac- 
 ing of ties 
 
 Hoffman, elastic curve of tie 
 
 Holley, A. L.: 
 
 bottom casting used by 
 
 design of Bessemer plant 
 
 Honigsberg, O., measurement of forces be- 
 tween wheel and rail 
 
 Hook plates: 
 
 examples of 127, 128, 129, 130, 
 
 German experiments with 
 
 Hook spike (see Spikes). 
 
 Horizontal pressure of wheel on rail 
 
 Hot beds 
 
 shortness in steel 
 
 work, effect of, on steel 
 
 Howard, Coes and, experiments on rocking 
 
 of engine 
 
 Howard, James E. : 
 
 examination of rolling at different stages . . 
 
 experiments on repeated stress 
 
 flow of metal in head of rail 
 
 report on Great Northern wreck 
 
 Lehigh Valley wreck 
 
 Howe, H. M.: 
 
 blowholes 
 
 259 
 445 
 445 
 330 
 428 
 
 420 
 
 278 
 
 325 
 203 
 
540 
 
 Howe, H. M.: 
 
 effect of copper on steel 332 
 
 nickel on steel 335 
 
 rolling on structure of steel 427 
 
 phosphorus in rail steel 465 
 
 piping and segregation 401, 404, 409 
 
 network and ferrite grains in steel 427 
 
 on nickel steel 335 
 
 relation between carbon content and 
 
 strength of steel 329 
 
 Howe, M. A., bearing power of earth 313 
 
 Howorth, Captain, effect of slag in steel. . . . 391 
 
 Hulett ore unloader 354 
 
 Humfrey, J. C. W., experiments on repeated 
 
 stress 274 
 
 Hundred per cent joint 264 
 
 Hunnewell plantation 109 
 
 Hunt, R. W.: 
 
 effect of copper on steel 332 
 
 manufacture of early rails 324 
 
 piping of ingots 399 
 
 Hunt, R. W., and Company: 
 
 American rolling mill practice 438, 444 
 
 branding of rails 447 
 
 method of inspection at mills 464 
 
 shrinkage allowed at American rail mills . . 444 
 
 teeming practice at American rail mills. . . 399 
 
 Hydrogen, evolution of, in cooling steel 401 
 
 Hyper-eutectoid steel 427 
 
 Hypo-eutectoid steel 427 
 
 Illingworth's process for compression of 
 
 ingot 411,415 
 
 Illinois Steel Company (see Gary; South Works). 
 Impact : 
 
 bibliography of 288 
 
 discussion of, as applied to tests 293 
 
 effect of, on strength of pine ties 158 
 
 track 68 
 
 in bridge specifications 210 
 
 increase in wheel load due to 68 
 
 tests by Professor Goss on engine drivers 62 
 drop (see Drop test). 
 
 on bridges 64 
 
 Incision, angle of, in rolling 457 
 
 Indentation test 300 
 
 India, length of rails used in 267 
 
 Indiana Engineering Society, flat spots on 
 
 car wheels 55 
 
 Indiana Railroad Commission (see Railroad 
 
 Commission of Indiana). 
 Indiana Steel Company (see Gary). 
 
 Indirect pressure in rolling 456 
 
 Inertia: 
 
 moment of (see Moment of inertia). 
 
 of roadbed, effect of, on bearing power 189, 318 
 
 Inertia: 
 
 of track, effect of, on impact of wheel. ... 68 
 
 rail stresses 318 
 
 Ingot : 
 
 bled, specifications for 473 
 
 blowholes in 401 
 
 casting 395 
 
 discard from 416 
 
 fluid compression of 410, 415 
 
 form for reporting rails which failed from 
 different parts of, American Railway 
 
 Engineering Association 516 
 
 lettering rails from 416 
 
 piping of 399 
 
 reduction necessary for different parts of. . 420 
 
 segregation of 404 
 
 size of, at American mills 438 
 
 stripper 397 
 
 Inspection : 
 
 form for, American Railway Engineering 
 
 Association 504 
 
 specifications for, in rails (see also Specifi- 
 cation in question) 464 
 
 tendency toward greater, at mill 464 
 
 Institution of Civil Engineers: 
 
 Bell, on deflection tff rails in track 190 
 
 Kirkaldy, on wear of rails 205 
 
 Sandberg, manufacture and wear of rails . . 1 
 
 Williams, maintenance of permanent way. 1 
 
 Intergranular weakness in steel 272 
 
 International Association for Testing 
 Materials : 
 
 drop testing machine 291 
 
 hardness tests 300 
 
 impact tests 292 
 
 slag in steel 393 
 
 thermoelectric measurements of stress .... 311 
 International Railway Congress: 
 
 Dudley, on tonnage of rails and wheels. . . . 202 
 report on contact area between wheel and 
 
 rail 195 
 
 effect of speed on the track 28 
 
 electric traction 74 
 
 joints 266 
 
 length of rail 266 
 
 use of screw spikes 150 
 
 Interstate Commerce Commission, report 
 
 on Lehigh Valley Railroad wreck 203 
 
 Interurban cars 87, 88 
 
 Interurban railways (see Electric railway). 
 
 Interurban Rapid Transit, concrete ties on.. 104 
 Ireland (see English). 
 Iron: 
 
 carbon diagram 427 
 
 cast, manufacture of 357 
 
 content in ore 344, 363, 36^ 
 
541 
 
 Iron: 
 
 conversion of, into steel 366 
 
 cooling curve of 426 
 
 cupola 365 
 
 effect of repeated stress on 274, 277 
 
 extraction of, from its ore 344 
 
 ore (see Ore). 
 
 oxide, blowholes caused by 389 
 
 pig, manufacture of 357 
 
 rails, life of, in American railroads 4 
 
 United Kingdom 2 
 
 on North Eastern Railway 1, 2 
 
 price of 2, 325 
 
 wear of 1 
 
 structure of 270 
 
 Zores 173 
 
 Iron and Steel Institute: 
 
 effect of slag in rail steel 391 
 
 experiments on influence of arsenic on 
 
 steel 331 
 
 on influence of copper on 
 
 steel 331 
 
 on repeated stress 276 
 
 Iron rails (see Iron). 
 
 Jaggar test for hardness 303 
 
 Jeans, J. S.: 
 
 early steel rails 2 
 
 origin of pneumatic process of making steel 367 
 
 Job, Robert: 
 
 cells per square inch in rail steel 424 
 
 unsoundness of head of rail 391 
 
 John Brown and Company (see Brown, 
 John, and Company). 
 
 Johnson, L. E., ties, supply of 106 
 
 Johnson, Professor: 
 
 compression modulus as affected by sur- 
 face of contact 193 
 
 experiments on contact between wheel 
 
 and rail 194 
 
 Johnson, Thomas H. : 
 
 distribution of pressure through ballast . . 185 
 
 drop testing machine, tests on 291 
 
 Joints (see also Joint in question) : 
 
 American Railway Engineering Associa- 
 tion tests on 264 
 
 bibliography of recent literature on 268 
 
 chemical composition 264, 266 
 
 economic distribution of metal in 263 
 
 effect of, on rail stresses 259 
 
 fishing angle, determined by Chanute .... 6 
 
 for bull head rail 19 
 
 friction of 260 
 
 on American railways 264 
 
 on English and Scotch railway.-, arrange- 
 ment of 19 
 
 Joints : 
 
 on French railways, arrangement of 144 
 
 on German railways, arrangement of 144 
 
 size of 218 
 
 oil tempered 266 
 
 shear in 262 
 
 Watertown Arsenal tests on 260, 264 
 
 welded 267 
 
 Jones, Capt. Wm. R., development of mixer 
 
 by 365 
 
 Jones and Laughlin Steel Company: 
 
 duplex process employed at 388 
 
 tilting open hearth furnaces at 377 
 
 Jouraff sky, effect of cold on rails 285 
 
 Kansas City, Mexico and Orient Railway: 
 
 Mogul type engine 34 
 
 size and spacing of ties 121 
 
 Keep test for hardness 303 
 
 Kellogg, R. S., timber supply of the Unite 1 
 
 States 106 
 
 Kelly, William, pneumatic process of mak- 
 ing steel 367 
 
 Kennedy-Morrison process 435 
 
 Kennedy stove 360 
 
 Key, oak for double-headed rail 19 
 
 Kimbal tie 99, 105 
 
 Kingdom of Saxony State Railways: 
 
 arrangement of joints on 144 
 
 data on track 218 
 
 rail and tie plate 129 
 
 spikes used on 144, 145 
 
 weight of rail on 19, 144 
 
 Kingdom of Wiirttemberg State Railways: 
 
 arrangement of joints on 144 
 
 rail and tie plate 127 
 
 screw spikes on 144, 145 
 
 weight of rail on 19 
 
 Kirkaldy, wear of rails 205 
 
 Kneedler concrete tie 104 
 
 Krupp mills, fluid compression of ingot 415 
 
 Lackawanna Steel Company: 
 
 rail mill 438,444 
 
 shrinkage allowed at 444 
 
 strength of steel from 309 
 
 teeming practice 399 
 
 titanium steel rails ' 340 
 
 Ladle, casting 389 
 
 Lake Erie and Western Railroad: 
 
 concrete ties on 105 
 
 size and spacing of ties 121 
 
 Lake Shore and Michigan Southern Railway: 
 composition and metal ties on.. . . 96, 104. 105 
 
 concrete ties on 104, 105 
 
 prairie type engine 33 
 
542 
 
 Lake Shore and Michigan Southern Railway : 
 
 size and spacing of ties 121 
 
 speed of trains 24, 25, 26, 27, 28 
 
 tests on chilled car wheels 57 
 
 Lakeside and Marblehead Railroad, concrete 
 
 ties on 105 
 
 Lake States, lumber production of 108 
 
 Lake Superior: 
 
 ore industry 349 
 
 ores, iron content 344 
 
 Lakhovsky screw spike 148 
 
 Lancashire and Yorkshire Railway, rail 
 
 fastenings on 146, 147 
 
 Lanza, Gaetano: 
 
 driving wheel spring tests 49 
 
 effect of suddenly applied load 54 
 
 Larch: 
 
 force required to pull spike from 140 
 
 physical properties of 140, 166 
 
 Leading truck: 
 
 allowable weights on 321 
 
 on freight and passenger engines 32, 33, 34, 72 
 
 Lebanon iron used at Maryland Steel 
 
 Company 374 
 
 Lebasteur, impact tests 293 
 
 Lehigh Valley Railroad: 
 
 size and spacing of ties 121 
 
 speeds of trains on 25 
 
 wreck caused by broken rail 203 
 
 Length of rails: 
 
 report on, International Railway Con- 
 gress 266 
 
 specifications for (see also Specification in 
 
 question) 474 
 
 Lettering rails from different parts of ingot . . 416 
 
 Limestone used in blast furnace 365 
 
 Loading: 
 
 axle (see Axle loads). 
 
 dynamic, for different types of cars .... 85, 89 
 for different types of electric 
 
 engines 78 
 
 for different types of steam 
 
 engines 72, 73 
 
 E-60 211 
 
 maximum axle, allowable on rail 319 
 
 rail, for different classes of track 319 
 
 specifications for, rail 478 
 
 weights of rail for different conditions 
 
 of 322 
 
 Loblolly pine (see also Pine) : 
 
 forse required to pull spike from 139 
 
 physical properties of 158, 164, 166, 170 
 
 tie, decay in spike hole 139 
 
 wear of, under tie plate 123 
 
 Locomotives: 
 
 allowable axle loads of 322 
 
 Locomotives: 
 
 axle loads of (see Axle loads). 
 
 classification of 9 
 
 counterbalance pressure (see Counter- 
 balance). 
 
 decapod 29, 31 
 
 development in, at Baldwin Locomotive 
 
 Works 30 
 
 on Pennsylvania Rail- 
 road 30 
 
 draw bar pull, effect on track 71 
 
 driving wheels (see Driving wheels) . 
 dynamic augment of wheel pressure (see 
 Dynamic) . 
 
 effect of rocking of, on track 49, 71 
 
 on track of badly balanced 39 
 
 electric (see Electric locomotives), 
 freight, development of, on Pennsylvania 
 
 Railroad 30 
 
 examples of 33, 34, 72 
 
 passenger, development of, on Pennsyl- 
 vania Railroad 30 
 
 examples of 32, 33, 72 
 
 Pennsylvania Railroad tests on steam and 
 
 electric 74, 259 
 
 rail stresses caused by (see Stresses). 
 
 speed of 21, 23, 29 
 
 springs (see Springs). 
 
 steam, best speeds of 27 
 
 dimensions of 32, 33, 34, 72 
 
 weights of different types, 
 
 29, 32, 33, 34, 72 
 
 strength of track required for 322 
 
 tires (see Tires). 
 
 types of (see also Type in question) : 
 
 American 21 
 
 articulated 34, 72 
 
 Atlantic 21, 29, 32 
 
 consolidation 29, 34 
 
 eight-wheel 21, 29 
 
 Mallet 21 
 
 Mikado 31 
 
 Mogul 34,72 
 
 Pacific 21,29,32 
 
 prairie 33, 72 
 
 ten-wheel 33, 72 
 
 typical dynamic wheel loads 72, 73 
 
 weight of rail required for 322 
 
 weights of 29, 32, 33, 34, 72 
 
 wheels (see Wheel) . 
 Locust, plantations of, for tie timber. . . 109, 111 
 Lodgepole pine ties : 
 amount purchased in the United States 154, 156 
 
 cost of 154, 156 
 
 London and North Western Railway: 
 
 early steel rails on 3 
 
543 
 
 London and North Western Railway: 
 
 permanent way 19 
 
 rail fastenings on 19, 147 
 
 screw spikes on 19, 146, 147 
 
 section of rail 19 
 
 speed of trains 24, 26, 28 
 
 London, Brighton and South Coast Railway, 
 
 early steel rails on 3 
 
 Long Island Railroad, size and spacing of 
 
 ties 121 
 
 Longleaf pine (see also Pine), physical 
 
 properties of.. 164, 166, 167, 168, 169, 170 
 
 Losses in drop testing machine 293 
 
 Louisville and Nashville Railroad : 
 
 plantations for tie timber Ill 
 
 size and spacing of ties 121 
 
 Love, C. E.: 
 
 analysis of Government track experi- 
 ments 240 
 
 relation between depression and pressure 
 
 on tie 192 
 
 Ludwick, hardness tests 300 
 
 MacPherson, D., long ties used on muskeg 
 
 swamp 188 
 
 Magnetic crane for loading rails 446 
 
 testing of rails 303 
 
 Main driver (sec Driving wheels). 
 Main rod: 
 
 angularity of, effect of inertia of track on 
 
 'stresses produced by 69 
 
 effect of angularity of, calculation of 35 
 
 pressure on rail caused by angularity of . . 35 
 Maine Central Railroad, size and spacing 
 
 of ties 121 
 
 Mallet type engine (see also Articulated 
 engine) : 
 
 description of 21 
 
 dimensions of • • 34, 72 
 
 draw bar pull of 22 
 
 speeds of 22 
 
 weights of 34, 72 
 
 Maltitz, E. von: 
 
 blowholes 389, 391, 404 
 
 effect of recarbonizing 391 
 
 Manganese: 
 
 content in rails: 
 
 of Bessemer steel 11, 253, 310 
 
 of open hearth steel 310 
 
 specifications for (see also Specification 
 
 in question) 328, 342, 466 
 
 effect of, in casting 402 
 
 steel 330 
 
 on blowholes 402 
 
 ferro 330, 374, 380 
 
 in iron ore 363, 364 
 
 Manganese: 
 
 segregation of 409 
 
 specifications for, in rails (see also Specifi- 
 cations in question) 328, 342, 466 
 
 steel 333, 336, 341 
 
 rails, production of 341 
 
 use of, on steam railways 333 
 
 street railways 339 
 
 sulphide in steel 390 
 
 Manufacture: 
 
 bibliography of literature on steel (see 
 Preface). 
 
 casting the ingot 395 
 
 conversion of iron into steel 366 
 
 difficulty in heavy A.S.C.E. rail 7, 14 
 
 early process of rail 324 
 
 extraction of the iron from its ore 344 
 
 influence of detail of 324 
 
 of car wheels 57 
 
 of iron rails 1,2 
 
 rolling 420 
 
 specifications for (see also Specifica- 
 tions) 473 
 
 Manufacturers (see Association of American 
 Steel Manufacturers): 
 Steel, of America (see Association of Ameri- 
 can Steel Manufacturers). 
 Maple ties: 
 
 amount purchased in the United States ... 156 
 
 cost of 159 
 
 Marston, A., experiments on steel rollers on 
 
 steel plates 194 
 
 Martens, Professor, impact tests 292 
 
 Martin, S. S., drop tests 288 
 
 Maryland Steel Company: 
 
 Bessemer converters at 370, 374 
 
 blast furnaces at 363 
 
 cold straightening press at 446 
 
 duplex process employed at 388 
 
 experiments on rolling rails 430 
 
 iron cupola at 365 
 
 Mayari ore used at 335 
 
 rolhng mill practice 438, 444 
 
 shrinkage allowed at 444 
 
 teeming practice 399 
 
 tests on strength of rail head 252 
 
 titanium rail steel 340 
 
 Massachusetts Institute of Technology: 
 Coes and Howard's thesis on driving wheel 
 
 springs 49 
 
 congress of technology 30 
 
 Master Car Builders' Association: 
 
 limiting length of flat spot in car wheels . . 56 
 
 standard tire 7 
 
 wheel in relation to design of 
 
 rail 6, 17 
 
544 
 
 Master Mechanics Association (see American 
 Railway Master Mechanics Associa- 
 tion). 
 
 Mayari ore 335 
 
 McGill University, indentation tests 300 
 
 McKee tie plate: 
 
 examples of 122 
 
 tests on 122 
 
 McKeen motor car 86 
 
 McKenna process for rerolling rails 459 
 
 Mechanical work, influence of, on steel 427 
 
 Mellor, J. W.: 
 
 changes during cooling of metals 425 
 
 crystalline structure of metals 270 
 
 Menard, friction of joints 261 
 
 Mesquite ties: 
 
 amount purchased in the United States ... 156 
 
 cost of 156 
 
 Metal ties (see also Tie in question) 90 
 
 Metcalf, William, phosphorus in rail steel . . . 465 
 
 Mexican Railway steel tie 96 
 
 Michel, Jules: 
 
 holding power of spikes 140 
 
 influence of form of thread on holding 
 
 power of screw spikes 148 
 
 Michigan Central Railroad: 
 
 chemical composition of rails 328 
 
 concrete ties on 105 
 
 consolidation type engine 34 
 
 continuous rail on 268 
 
 80-lb. rails used on, in 1893 7 
 
 electric locomotives in Detroit River 
 
 tunnel 80 
 
 plantations for tie timber 112 
 
 size and spacing of ties 121 
 
 speeds of trains on 24, 26 
 
 Middle driver (see Driving wheels). 
 
 Midland Railway: 
 
 permanent way 19 
 
 rail fastenings on 19, 146, 147 
 
 section of rail 19 
 
 speed of trains 28 
 
 Mikado type engine: 
 
 classification of 29 
 
 description of 31 
 
 Mill: 
 
 American, practice 399, 438, 444 
 
 blooming 436, 438 
 
 reversing cogging 437 
 
 rolling (see Rolling). 
 
 three-high 437 
 
 Minneapolis and St. Louis Railroad, size and 
 
 spacing of ties 121 
 
 Minnesota, iron ore mines in 346, 347 
 
 Missouri and North Arkansas Railroad, size 
 
 and spacing of ties 121 
 
 Missouri, Kansas and Texas Railway System, 
 
 size and spacing of ties 121 
 
 Missouri, Pacific Railway Systems, size and 
 
 spacing of ties 121 
 
 Mixer: 
 
 development of 365 
 
 metal used in open hearth heat, analysis 
 
 of 380 
 
 Mobile and Ohio Railroad, size and spacing 
 
 of ties 121 
 
 Mockernut hickory, physical properties of. . . 164 
 
 Modulus: 
 
 of compression (see Compression). 
 
 of elasticity of steel 225, 241 
 
 wood 158, 166, 168 
 
 of rupture (see also Ultimate strength). 
 
 of steel 306 
 
 wood , . 158, 164, 166, 168 
 
 section, of different rails (see also Rail in 
 
 question) 319 
 
 Mogul type engine: 
 
 allowable axle loads 322 
 
 classification of 29 
 
 dimensions of 34, 72 
 
 effect of excess balance and angularity of 
 
 main rod 43, 70 
 
 rail stresses caused by 222, 237 
 
 speeds of 70 
 
 strength of track required for 322 
 
 typical dynamic wheel loads 73 
 
 U. S. Government experiments with 222 
 
 weight of rail required for 322 
 
 weights of 34, 72 
 
 Moisture (see Water). 
 
 Moment, bending (see Bending moment). 
 
 Moment of inertia: 
 load diagram for rails having different . . . 319 
 of different rails (see also Rail in question). 319 
 of rails and joints on German railways. . . . 218 
 
 Motor cars 85, 86 
 
 Munshet, improvement in making steel 367 
 
 Nail spike (see Spikes). 
 
 National Conservation Commission, report 
 
 on ores 345 
 
 New England States, lumber production 
 
 of 
 
 108 
 
 New Jersey Steel and Iron Company, first 
 
 open hearth furnace in America 375 
 
 New York Central and Hudson River Rail- 
 road: 
 
 concrete ties on . . 104 
 
 80-lb. rails used on, in 1893 7 
 
 electric locomotives 79 
 
 Pacific type engine 32 
 
 rail section 10 
 
New York Central and Hudson River Rail- 
 road : 
 
 rail failures 10 
 
 size and spacing of ties 121 
 
 speed of trains 23, 24, 25, 26, 28 
 
 stremmatograph tests 237 
 
 ten- wheel type engine 33 
 
 wear of rails on 328 
 
 New York Central Lines: 
 
 specifications for rail 478 
 
 speed of trains 23 
 
 titanium steel rails 341 
 
 New York, Chicago and St. Louis Railroad, 
 
 size and spacing of ties 121 
 
 New York, New Haven and Hartford Rail- 
 road: 
 
 electric locomotives 74, 80 
 
 size and spacing of ties 121 
 
 speed of trains 28 
 
 New York, Ontario and Western Railway, 
 
 size and spacing of ties 121 
 
 New York Railroad Club, annealed rails .... 208 
 Nickel, chrome, steel rails: 
 
 production of 341 
 
 use of 333 
 
 Nickel, effect of, in steel 334 
 
 Nickel Plate, concrete ties on 105 
 
 Nickel steel rails : 
 
 production of 341 
 
 use of 333 
 
 Nicolaieff, ore used at Maryland Steel 
 
 Company 363 
 
 Nicolas Railway (of Russia), early steel 
 
 rails on 3 
 
 Nicolia, effect of cold on rails 285 
 
 Nisbet, productivity of woodlands 114 
 
 Nitrogen: 
 
 effect of, in steel 287, 341 
 
 on ductility of rail steel 287 
 
 titanium on, in steel 287, 341 
 
 evolution of, in cooling steel 401 
 
 Norfolk and Western Railway: 
 
 chemical composition of rails 328 
 
 manganese rails on 333 
 
 plantations for tie timber 112 
 
 size and spacing of ties 121 
 
 Norfolk Southern Railroad, size and spacing 
 
 of ties 121 
 
 North Austrian Railways, experiments on 
 
 rails rolled for 450 
 
 North British Railway, rail fastenings on, 146, 147 
 
 North Chicago Rolling Mill, early steel rails 3 
 North Eastern Railway (of England) : 
 
 iron rails on 1 
 
 rail fastenings on 147 
 
 speed of trains 28 
 
 Northern Pacific Railway: 
 
 management of timber lands on 113 
 
 size and spacing of ties 121 
 
 Northern Railway (of Austria), early steel 
 
 rails on 3 
 
 Northern Railway (of France) : 
 
 arrangement of joints on 144 
 
 early screw spikes used on 141 
 
 screw spikes on 143, 144 
 
 speed of trains 24 
 
 Northern Railway (of Spain), concrete ties 
 
 on 104, 105 
 
 North Western Pacific Railroad Company, 
 
 size and spacing of ties 121 
 
 Norway pine, physical properties of . . . . 166, 168 
 
 Nutmeg hickory, physical properties of 164 
 
 Oak: 
 
 ferrule for spike 19 
 
 force required to pull spike from 139, 140 
 
 key, for double-headed rail 19 
 
 permissible load under tie plate 171 
 
 physical properties of 164, 168, 170 
 
 plantations of, for tie timber Ill 
 
 ties, amount purchased in the United 
 
 States 154, 156 
 
 annual charge of 115 
 
 cost of 154, 156 
 
 on French Eastern Railway 117 
 
 Oklahoma Railway Company, car used on. . . 88 
 Old Colony and Newport Railway, early steel 
 
 rails on 2 
 
 Open hearth: 
 
 comparison of, with Bessemer process. . . . 380 
 
 description of process 375, 378 
 
 early experiments with 375 
 
 steel, analysis of 310 
 
 rails, branding 447 
 
 ductility of 286 
 
 hardness determined by sclero- 
 
 scope 299 
 
 production of 382 
 
 specifications for, 342, 463, 478, 491 
 
 strength of 306 
 
 Talbot continuous process 375 
 
 tilting furnace 375 
 
 Ore: 
 
 analyses of 363, 364 
 
 docks 351, 357 
 
 extraction of iron from 344 
 
 handling of 351, 357 
 
 iron content in 344, 363, 364 
 
 Lake Superior mining 351 
 
 Mayari 335 
 
 roasting 344 
 
 supply of, low phosphorus 374, 381 
 
Ore: 
 
 transportation of , 349 
 
 unloader 352 
 
 used in blast furnace 360, 363, 364 
 
 open-hearth furnace 378, 380 
 
 vessels 349 
 
 Oregon Electric Railway Company, cars 
 
 used on 87 
 
 Oregon Railroad and Navigation Company, 
 
 ten- wheel type engine 33 
 
 Oregon Short Line Railroad, Atlantic type 
 
 engine 32 
 
 Orleans Railway: 
 
 arrangement of joints on 144 
 
 screw spikes on 144 
 
 Osmond, transition points in steel 426 
 
 Overcup oak, physical properties of 164 
 
 Oxygen, effect of, in steel 401, 404 
 
 Pacific States, lumber production of 108 
 
 Pacific type engine: 
 
 allowable axle loads 322 
 
 classification of 29 
 
 description of 21 
 
 dimensions of 32, 72 
 
 effect flf excess balance and angularity of 
 
 main rod 41, 70 
 
 pressure rounding curve 259 
 
 rail stresses caused by 215 
 
 speeds of 21, 26, 70 
 
 strength of track required for 322 
 
 typical dynamic wheel loads 72 
 
 weight of rail required for 322 
 
 weights of 31, 32, 72 
 
 Paris, Lyons and Mediterranean Railway: 
 
 arrangement of joints 144 
 
 axle loads used on 19 
 
 early steel rails on 3 
 
 experiments on ties 172 
 
 screw spikes 143, 144 
 
 section of rail 19 
 
 m, South Works 442 
 
 r cars (see Cars). 
 locomotives (see Locomotives). 
 
 Passes: 
 
 cogging, Puppe's tests 448, 450 
 
 number in rolling rail 438 
 
 Pay-as-you-enter cars 88 
 
 Pearlite 427 
 
 Pecan hickory, physical properties of 164 
 
 Pennsylvania Lines: 
 
 concrete ties on 97, 102, 104, 105 
 
 continuous rail on 267 
 
 nickel steel rails on 333 
 
 Pacific type engine 32 
 
 plantations for tie timber 112 
 
 Pennsylvania Lines: 
 
 prairie type engine 33 
 
 size and spacing of ties 121 
 
 special steel rail on 333 
 
 speed of trains 25, 26 
 
 steel ties on 94 
 
 Pennsylvania Lines West of Pittsburgh (see 
 Pennsylvania Lines). 
 
 Pennsylvania Railroad: 
 
 chemical composition of early rails 328 
 
 rails 328, 342 
 
 committee on rail section 18 
 
 continuous rail 267 
 
 development in locomotives since 1850 ... 31 
 
 early experiments on wear of rails 326 
 
 steel rails used on 4 
 
 electric locomotive 74, 76, 79 
 
 horizontal pressure on rail exerted by 
 
 engine wheel 259 
 
 joint 264 
 
 plantations for tie timber Ill 
 
 prairie type engine -. . 33 
 
 rail section 10, 18 
 
 deflection in drop test 291 
 
 failures of n 
 
 "P. S." section (see Pennsylvania Stand- 
 ard section) . 
 
 size and spacing of ties 121 
 
 specifications for rail 342, 463 
 
 speeds of trains 24, 26, 27, 28 
 
 steam and electric locomotives, tests on 74, 257 
 
 steel ties on 96 
 
 tests on ballast 184 
 
 track experiments on, by TJ. S. Govern- 
 ment 225 
 
 weight of sleeping cars 82 
 
 Pennsylvania Standard ("P. S.") section: 
 
 deflection in drop test 291 
 
 design of 18, 460 
 
 failures of 10 
 
 Pennsylvania Steel Company: 
 
 bottom casting at 410 
 
 early steel rails 4 
 
 machine for testing rail wear 303 
 
 rail mill 438, 444 
 
 shrinkage allowed at 444 
 
 use of iron with high copper 332 
 
 Percival concrete tie 101, 105 
 
 Pere Marquette Railroad, concrete ties on, 
 
 104, 105 
 
 Permanent Way: 
 
 of English railways 19 
 
 maintenance of 1 
 
 "Pewter" rails 12 
 
 Philadelphia and Reading Railway: 
 
 90-lb. rails used on, in 1893 7 
 
547 
 
 Philadelphia and Reading Railway: 
 
 performance of fine structure rails on 424 
 
 size and spacing of ties 121 
 
 speed of trains 24, 28 
 
 tests on annealed rails 208 
 
 Philadelphia Rapid Transit, concrete ties 
 
 on 104 
 
 Philadelphia, Wilmington and Baltimore 
 
 Railway, early steel rails on 2, 4 
 
 Phosphorus : 
 
 content in rails of Bessemer steel.. 11, 253, 310 
 
 open hearth steel 310 
 
 effect of, in steel 329 
 
 on ductility of rail steel 286 
 
 in iron ore . . 
 
 segregation of 408 
 
 specifications for, in rails (see also Speci- 
 fications) 328, 342, 466 
 
 supply of ores low in 374, 381 
 
 Physical properties: 
 
 of rails, form for reporting, American 
 
 Railway Engineering Association. . . . 503 
 of special steels (see Special steels). 
 of steel (see Steel). 
 
 of wood, treated 158 
 
 untreated 158, 164, 166, 168 
 
 Pig iron, manufacture of 357 
 
 Pignut hickory, physical properties of 164 
 
 Pine: 
 
 effect of moisture on strength of 169 
 
 force required to pull spike from 139, 152 
 
 permissible load under tie plate 171 
 
 physical properties of, 
 
 158, 164, 166, 167, 168, 169, 170 
 ties: 
 
 amount purchased in the United 
 
 States 154, 156 
 
 annual charge of 115 
 
 cost of 154, 156 
 
 effect of preservative treatment on 
 
 strength of 158 
 
 tests on treated 158 
 
 wear of, under tie plate 123 
 
 Pipes in ingots: 
 
 bibliography of recent literature on 418 
 
 cause of 399 
 
 effect of fluid compression on 410, 415 
 
 rapid charging into heating fur- 
 nace 400 
 
 titanium on 405 
 
 Howe's experiments on 409, 411 
 
 tests to detect 464, 469 
 
 Pittsburgh and Lake Erie Railroad : 
 
 concrete ties on 101, 105 
 
 tie plate 122 
 
 Pittsburgh, Cincinnati, Chicago and St. 
 Louis Railway (see also Pennsylvania 
 Lines) : 
 chemical composition of early steel rails 
 
 on 326 
 
 speed of trains 24 
 
 Pittsburgh, Fort Wayne and Chicago Rail- 
 way (see also Pennsylvania Lines), 
 
 speed of trains 24 
 
 Pittsburgh Railway Club, Fowler's experi- 
 ments on contact of wheels with 
 
 rails 195 
 
 Plantations for growth of tie timber 109, 111 
 
 Pneumatic method of making steel (see 
 Bessemer) . 
 
 Pole tie 119 
 
 Poplar: 
 
 force required to pull spike from 140 
 
 plantations of, for tie timber Ill 
 
 tie plates of, on French Eastern Railway . . 132 
 
 Post oak, physical properties of 164 
 
 Poutzen, friction of joints 261 
 
 Prairie type engine: 
 
 classification of 29 
 
 dimensions of 33, 72 
 
 speeds of 70 
 
 typical dynamic wheel loads 72 
 
 weights of 33, 72 
 
 Prepayment cars 88 
 
 Preservation, tie (see Treated ties). 
 Pressure: 
 
 allowable on subgrade, amount of. 187, 189, 317 
 
 calculation of 313 
 
 under tie plate 171 
 
 casting ingots under 410, 415 
 
 direct, in rolling 456 
 
 distribution of, to subgrade 180, 185 
 
 effect of steam, on treated ties 158, 162 
 
 indirect, in rolling 456 
 
 of wheel on rail: 
 
 at point of contact 193 
 
 caused by drawbar pull 71 
 
 flat spot on wheel 54 
 
 excess pressure of counter- 
 balance and angularity of 
 
 main rod 35, 70 
 
 impact 68 
 
 irregularities in the track. . 45, 71 
 
 rocking of engine 49, 71 
 
 velocity of load 70 
 
 weight of tender 71 
 
 dynamic, of cars 85 
 
 electric engines 78 
 
 steam engines 71, 72 
 
 effect of change in grade on 45 
 
 inertia of track on 69 
 
548 
 
 Pressure: 
 
 horizontal component of, on curves .... 259 
 
 static, cars 85 
 
 electric locomotives 77, 79, 80 
 
 steam locomotives 29 
 
 working, of steam locomotives, 
 
 37, 40, 41, 42, 43, 44 
 Production: 
 
 lumber, in the United States 108 
 
 rail, in the United States 382 
 
 ties, in the United States 154, 156 
 
 Profile of rail 46, 47 
 
 Prussian Government Railroads: 
 
 experiments on tie plates 133 
 
 screw spikes used on 144, 145 
 
 tests on strength of rails 302 
 
 Prussian Hessian Railways, data on track. . . 218 
 Prussian Railway Department, hardness 
 
 tests 302 
 
 Prussian State Railways: 
 
 arrangement of joints on 144 
 
 rail and tie plate 131 
 
 speed of trains 27, 29 
 
 spikes used on 144, 145 
 
 ties with screw dowels 151 
 
 weight of rail on 131, 144 
 
 "P. S." rail section (see Pennsylvania 
 Standard section). 
 
 Puppe, experiments on rail rolling 447 
 
 Purdue University, tests on tie plates 169 
 
 Pyrometers 434, 476 
 
 Queen and Crescent, chemical composition 
 
 of rails 328 
 
 Rail: 
 
 advance wave determinations of 221 
 
 annealed 208 
 
 base (see Base rail). 
 
 bearing of, on tie 122 
 
 bending, affected by flow of metal in 
 
 head 204 
 
 bending moment (see Bending moment). 
 
 branding 446 
 
 broken (see Broken rails) . 
 
 base (see Broken base). 
 
 by defective equipment 57 
 
 bull head 19,484 
 
 camber in 444, 477, 482 
 
 chemical composition (see Chemical com- 
 position). 
 
 continuous 267 
 
 corrugations in . . 209 
 
 creeping of 153 
 
 crushed head (see Crushed head). 
 
 Rail: 
 
 defective (see Defective rails). 
 
 design of (see also Section), investigation 
 
 of, by Ashbel Welsh 5 
 
 design of (see also Section), investigation 
 
 of, by Chanute 6 
 
 design of (see also Section), principles 
 
 governing 16 
 
 double head 19, 484 
 
 ductility of steel (see Ductility). 
 
 dynamic load for different classes of track . 319 
 
 effect of cold on 284 
 
 repeated stress on 284 
 
 elastic curve of 241, 242 
 
 electric steel 383 
 
 exhausted metal in 204, 284 
 
 failures, chart of, Harriman Lines 524 
 
 reported by American Railway 
 
 Engineering Association .... 10 
 
 typical, photographs of 11 
 
 fastening of, to tie 138 
 
 fastenings (see Joints), 
 flange (see Base, rail). 
 
 flat bottom 19, 488 
 
 flow of metal in head of (see Flow of 
 
 metal). 
 
 gagging, or cold straightening 446 
 
 girder (see Street railway). 
 
 grooved (see Street railway). 
 
 hardness tests on (see Hardness tests). 
 
 head (see Head, rail). 
 
 horizontal pressure of wheel on curves. . . . 259 
 
 iron 1,324 
 
 joint (see Joints). 
 
 length of 267, 474, 485, 489, 492 
 
 lettering from different parts of ingot . . . 416 
 
 load for different classes of track 317, 322 
 
 manufacture (see Manufacture), 
 manufacturers (see Association of Ameri- 
 can Steel Manufacturers). 
 mill (see Rolling), 
 moment of inertia of standard (see also 
 
 Rail in question) 319 
 
 performance of fine structure 424 
 
 heavy sections 7, 10 
 
 prices of 325 
 
 production 382 
 
 profile of 46, 47 
 
 records (see Reports and records), 
 reports (see Reports and records). 
 
 rerolled 459 
 
 road (see Road in question). 
 
 roaring 209 
 
 rolling (see Rolling). 
 
 section modulus of standard (see also Rail 
 
 in question) 319 
 
Rail: 
 
 sections (see Section), 
 shear in (see Shear), 
 special (see Special rails), 
 specifications (see Specifications). 
 
 split head 10, 390 
 
 web 10 
 
 stamping 446 
 
 steel, first made in America 3 
 
 used in America 2 
 
 foreign countries ....... 3 
 
 John Brown and Company .... 4, 5, 326 
 
 special (see Special rails). 
 
 straightening 446 
 
 strains produced in, by straightening 445 
 
 street railway (see Street railway). 
 
 strength of 270, 310 
 
 stresses in (see Stresses) . 
 
 supports of 90 
 
 T- (see T-rail). 
 
 temperature, effect of on, cold 284 
 
 finishing (see Fin- 
 ishing) . 
 tests of, hardness (see also Hardness tests) . 298 
 impact (see Drop test). 
 
 tests on, by U. S. Government 218 
 
 stremmatograph 71, 212, 236 
 
 tramway (see Street railway). 
 
 transverse fissures in head 203 
 
 way (see Road in question), 
 wear (see Wear), 
 web (see Web, rail). 
 
 weights of, for different loading 310, 322 
 
 on English railways 19 
 
 French railways. ... 18, 19, 144 
 German railways . . 19, 125, 144 
 specifications for (see also Speci- 
 fication in question). 
 Railroad (see Road in question). 
 Railroad Commission of Indiana: 
 
 Buffington, on quality of rail 325 
 
 Cushing, on discard of ingot 417 
 
 Dudley, on wear of rails 328 
 
 Railway (see Road in question). 
 
 Rainfall on ballast in Pennsylvania Railroad 
 
 tests 184 
 
 Rasch, thermoelectric measurements of stress 311 
 Rear driver (see Driving wheels). 
 
 Recalescence 425 
 
 Recarbonizer (see also Ferromanganese) .... 366 
 Recarbonizing: 
 
 effect of, on steel 389 
 
 in Bessemer converter 374 
 
 in casting ladle 391 
 
 in electric furnace 385 
 
 in open-hearth furnace 380 
 
 Reciprocating machine, tests on rail head 
 
 with 256 
 
 parts on locomotives (see also 
 
 Counterbalance) 32, 35 
 
 Records (see Reports and records). 
 
 Red cedar, physical properties of 168 
 
 Red gum : 
 
 crushing strength of 170 
 
 plantations of, for tie timber 112 
 
 Red oak (see also Oak): 
 
 force required to pull spike from 139 
 
 physical properties of 139, 164, 170 
 
 plantations of, for tie timber Ill 
 
 tie, annual charge of 115 
 
 Red pine, physical properties of 164 
 
 Red shortness in steel 330, 332 
 
 Red spruce, physical properties of 166 
 
 Reduction necessary from ingot to rail 420 
 
 Reduction of area: 
 
 effect of chemical composition on (see 
 Element in question). 
 
 size of grain on 424 
 
 in rolling, Puppe's tests 448 
 
 of electric steel 386 
 
 of rail steel 306 
 
 Redwood: 
 
 physical properties of 166, 168 
 
 ties, amount purchased in the United 
 
 States 154, 156 
 
 cost of 154, 156 
 
 Reheating furnace 396 
 
 Repeated stress (see Stresses). 
 Reports and records: 
 by American Railway Engineering As- 
 sociation : 
 
 compilation of results for study 512 
 
 from Division officers 506 
 
 inspection and shipment at mill 501 
 
 laboratory examination of special rails.. 511 
 
 progressive wear of special rail 517 
 
 chart of rail failures, Harriman Lines 524 
 
 form for continuous record shown 
 
 graphically 523 
 
 Republic Iron and Steel Company, blast 
 
 furnace 355, 356 
 
 Rerolled rails 459 
 
 Resal, friction of joints 261 
 
 Reversing cogging mill 437 
 
 Reynolds and Smith, experiments on re- 
 peated stress 278, 282 
 
 Rhymney Railway (of England), early steel 
 
 rails on 3 
 
 Richards, R. H.: 
 
 influence of copper on steel 332 
 
 roasting ores 344 
 
 Riegler tie 102 
 
Riese, elastic curve of tie 172 
 
 Roadbed (see Subgrade). 
 
 Roaring rails 209 
 
 Roasting ore 344 
 
 Rock (see Stone). 
 
 Rock Island (see Chicago, Rock Island and 
 Pacific Railway). 
 
 Rocking of engine: 
 
 Coes and Howard's experiments on 49 
 
 effect of inertia of track on stresses pro- 
 duced by 69 
 
 on wheel pressure 49, 71 
 
 Roechling Rodenhauser electric furnace .... 385 
 
 Rogers, F., experiments on repeated stress. . 282 
 
 Rolling: 
 
 American practice 438 
 
 direct .pressure in 456 
 
 early rails 324, 434 
 
 effect of, on grain size 424, 427 
 
 experiments at Maryland Steel Company . 430 
 Watertown Arsenal by 
 
 Howard 420 
 
 by Puppe 447 
 
 heavy A.S.C.E. sections 7, 435 
 
 indirect pressure in 456 
 
 Kennedy-Morrison process 435 
 
 mill at Algoma Steel Company 437, 438 
 
 Gary 437,438 
 
 South Works 438, 443, 444 
 
 practice at American 438, 444 
 
 power required for 458 
 
 reduction at each pass 436, 438 
 
 structural changes during 420 
 
 tests on head of rail 205, 256 
 
 work done in, Puppe's tests 448 
 
 worn rails, McKenna process 459 
 
 Rolls, design of 457 
 
 Roozeboom, H. W. B., carbon-iron diagram . 427 
 
 Rosenhain, W. : 
 
 experiments on repeated stress 273 
 
 slag enclosures in steel 393 
 
 Roughing pass 437, 438 
 
 Rudeloff, experiments on nickel steel 334 
 
 Russia: 
 
 early steel rails 3 
 
 effect of cold on rails in 285 
 
 Rutland Railroad, size and spacing of ties. . . 121 
 
 San Antonio and Arkansas Pass Railway: 
 
 plantations of tie timber 112 
 
 size and spacing of ties 121 
 
 Sand: 
 
 angle of friction of 314 
 
 weight of 316 
 
 Sandberg, C. P. : 
 
 drop testing machine 288 
 
 Sandberg, C. P.: 
 
 experiments on effect of cold on rails . . . 284 
 
 manufacture and wear of rails 1 
 
 Santa Fe (see Atchison, Topeka and Santa 
 
 Fe Railway). 
 Santa Fe, Prescott and Phcenix Railway, 
 
 size and spacing of ties 121 
 
 Sarada tie 102, 104 
 
 Sarnia tunnel, electric locomotive 80 
 
 Sauveur: 
 
 rail structure 429 
 
 relation between size of grain and physi- 
 cal properties of steel 424 
 
 Saw: 
 
 hot. 
 rail. 
 
 445 
 
 445 
 
 runs at American mills 445 
 
 Saxony State Railways (see Kingdom of 
 
 Saxony State Railways). 
 Schubert : 
 
 depression of tie in ballast 190 
 
 experiments on ballast 180 
 
 Schwald, elastic curve of tie 172 
 
 Sclerometer 302 
 
 302 
 
 Scratch test for hardness ' 
 
 Screw spikes (see Spikes). 
 
 Seaboard Air Line Railway, size and spacing 
 
 of ties 121 
 
 Seasoning, effect of, on strength of wood .... 168 
 
 Seating capacity of electric cars 87, 88 
 
 Section, rail (see also Section in question) : 
 
 American 10, 14, 18, 19, 460 
 
 American Electric Railway Association 19 
 
 American Railway Association 14 
 
 American Society of Civil Engineers 6 
 
 British standard 19 
 
 bullhead 19 
 
 comparison of failures of different 
 
 sections 10 
 
 designed by Ashbel Welsh 5 
 
 Chanute 6 
 
 Dudley 11 
 
 early 1 
 
 English 19 
 
 flat bottom 19 
 
 French 19 
 
 German 19, 125 
 
 girder 19 
 
 grooved 19 
 
 high tee 19 
 
 New York Central and Hudson River 
 
 Railroad 11 
 
 Pennsylvania Railroad 11 
 
 Pennsylvania standard 18 
 
 14, 460 
 
551 
 
 Section, rail: 
 
 principles governing design of 16 
 
 specifications for (see also Specification in 
 
 question) 476 
 
 street railway (see Street railway). 
 
 tramway 19 
 
 Vignole 18 
 
 Section moduli of different rails (see also Rail 
 
 in question) 319 
 
 Seeley concrete tie 104 
 
 Segregation: 
 
 bibliography of recent literature on 418 
 
 cause of 404 
 
 effect of fluid compression on 410, 416 
 
 titanium on 406 
 
 Howe's experiments on 409, 411 
 
 Selby, O. E., stresses in the rail 210 
 
 Series "A," American Railway Association. . 14 
 
 "B," American Railway Association. . 14 
 
 Shagbark hickory, physical properties of ... . 164 
 Shear: 
 
 calculation of, for 100-lb. rail 239 
 
 distribution of, in 100-lb. rail 250 
 
 in rail and joint 262 
 
 calculation of 239 
 
 Love's diagram of 241 
 
 intensity of, at any point 250 
 
 specifications for (see also Specification in 
 
 question) 473 
 
 Ship, used on Great Lakes for transporting 
 
 iron ore 349 
 
 Shipment: 
 
 form for reporting, American Railway 
 
 Engineering Association 505 
 
 of iron ore (see Transportation). 
 Shock (see Impact). 
 Shortleaf pine (see also Pine) : 
 
 physical properties of 164, 166, 168, 170 
 
 tie cut through spike holes 138 
 
 Shrinkage: 
 
 allowed at American mills 444 
 
 English mills 435 
 
 specifications for (see also Specification 
 
 in question) 475 
 
 Shwedler, elastic curve of tie 172 
 
 early open hearth furnace 375 
 
 furnace 375 
 
 reheating furnace 437 
 
 Sierra Morena ore, used at Maryland Steel 
 
 Company 363 
 
 Silica, in iron ore 363, 364 
 
 Silicon: 
 
 content in rails of Bessemer steel. 11, 253, 310 
 
 open hearth steel 310 
 
 effect of, in casting 401, 402 
 
 Silicon, effect of, in steel 329 
 
 on blowholes 402 
 
 segregation of 409 
 
 specifications for, in rails (see also Specifi- 
 cation in question) 328, 342, 466 
 
 Six-wheel engine: 
 
 classification of 29 
 
 coupled, classification of 29 
 
 Slade, F. J., first open hearth furnace in 
 
 America 375 
 
 Slag: 
 
 in steel 391 
 
 unsound metal caused by 391 
 
 Sleeper (see Ties). 
 
 Sleeping car 84 
 
 Slip: 
 
 bands caused by repeating stress 274 
 
 coefficient of, locomotive drivers 198, 199 
 
 means of determining, in 
 
 earth 313 
 
 Smelting, iron ore 344, 357 
 
 Smith, H. E., tests on chilled car wheels. . . 57 
 Smith and Reynolds, experiments on re- 
 peated stress 278, 282 
 
 Snyder steel tie 96 
 
 Soaking pits 396 
 
 Soils, supporting power 313 
 
 Solid solution of iron 427 
 
 South Africa, length of rails used in 267 
 
 South America, length of rails used in 267 
 
 South Works, Illinois Steel Company: 
 
 electric furnace at 383 
 
 experiments with new section of rail 462 
 
 Forsyth's transferring ladle 390 
 
 pass diagram 442 
 
 rail mill 438,442 
 
 shrinkage allowed at 444 
 
 teeming practice 399 
 
 Southern Pacific Company: 
 
 Mogul type engine 34 
 
 plantations of tie timber 112 
 
 size and spacing of ties 121 
 
 speed of trains 26, 28 
 
 timber lands of 113 
 
 Southern pine (see Pine). 
 Southern Railway: 
 
 Atlantic type engine 32 
 
 chemical composition of early rails on ... . 326 
 
 plantations for tie timber 112 
 
 size and spacing of ties 121 
 
 Southern Railway (of France): 
 
 arrangement of joints on 144 
 
 screw spikes on 144 
 
 Southern States: 
 
 lumber production of 108 
 
 ores in 381 
 
552 
 
 Spangenberg's experiments on repeated 
 
 stress 277 
 
 Spanish oak, physical properties of 164 
 
 Special rails (see also Special steels) : 
 
 form for reporting laboratory examina- 
 tion of 511 
 
 form for reporting wear of 517 
 
 Special steels: 
 
 chrome nickel 333, 341 
 
 chromium 333, 341 
 
 cupro-nickel 332 
 
 electric 341, 383 
 
 manganese 333, 336, 341 
 
 nickel 333, 334, 341 
 
 production of 341 
 
 titanium (see Titanium). 
 Specific gravity: 
 
 of manganese sulphide 390 
 
 of steel 390 
 
 of wood 158, 163 
 
 Specifications: 
 
 axle loads given in bridge 211 
 
 for drop testing machine 290 
 
 Specifications, rail: 
 
 American, comparison of: 
 
 bled ingots 473 
 
 branding 477 
 
 chemical composition 342, 465 
 
 discard 473 
 
 drilling 477 
 
 drop test 470 
 
 finishing 477 
 
 finishing temperature 475 
 
 inspection 464 
 
 length 474 
 
 loading 478 
 
 Nos. 1 and 2 rails 472 
 
 physical requirements 467 
 
 process of manufacture 473 
 
 quality of manufacture 473 
 
 section 476 
 
 473 
 
 475 
 
 weight 476 
 
 bibliography of 494 
 
 British standard bull head railway rails.. . 484 
 flat bottom railway rails. . 488 
 chemical composition (see also Specifica- 
 tion in question) 342, 465 
 
 for street railways, American Society for 
 
 Testing Materials 491 
 
 of American Railway Engineering Asso- 
 ciation 463 
 
 of American Society for Testing Materials. 491 
 of Association of American Steel Manu- 
 facturers 342, 463 
 
 Specifications, rail: 
 
 of Harriman Lines 463 
 
 of New York Central Lines 478 
 
 of Pennsylvania Railroad System 342, 463 
 
 Speed: 
 
 effect on bridges due to velocity of load . . 69 
 
 counterbalance pressure 35 
 
 depression of tie 190 
 
 track 29, 189, 191 
 
 due to velocity of load 70 
 
 Prussian State Railways 29 
 
 fast runs in last three decades 23 
 
 in rolling mills, Puppe's tests 458 
 
 of electric locomotives 27, 29, 79, 80 
 
 of modern trains 21, 23 
 
 Spiegle (see also Ferromanganese) 374 
 
 Spiegle-eisen (see also Ferromanganese) 374 
 
 Spikes: 
 
 common or nail 140 
 
 used on English railways. . . 146, 147 
 
 German railways . . . 144, 145 
 
 comparative cost of screw and common. . . 142 
 
 effect of treatment on holding force of . . . . 159 
 
 holding power of 139 
 
 hook 140 
 
 screw 140, 142 
 
 cost of equipping track with 142 
 
 examples of English 19, 146 
 
 French.. 143,144 
 
 German 144, 145 
 
 influence of design of thread on holding 
 
 power 148 
 
 machine for preparing ties for 142 
 
 on the Atchison, Topeka and Santa Fe 
 
 Railway 141 
 
 use of dowel with 142, 150 
 
 Splice bars (see Joints). 
 Split head: 
 
 classification of, American Railway En- 
 gineering Association 10 
 
 effect of casting on 390 
 
 photographs of typical failures 11 
 
 rail failures, six months ending April 30, 
 
 1909 10 
 
 Split web : 
 
 classification of, American Railway En- 
 gineering Association 10 
 
 photographs of typical failures 11 
 
 rail failures, six months ending April 30, 
 
 1909 10 
 
 Spokane and Inland Railroad, electric 
 
 locomotives 80 
 
 Springs, locomotive driving: 
 
 Coes and Howard's experiments on 49 
 
 dimensions of 48 
 
 effect of inertia of track on 69 
 
553 
 
 Springs, locomotive driving: 
 
 effect of suddenly applied load 54 
 
 rocking of engine on 49 
 
 weight borne by, in electric locomotives ... 74 
 Spruce : 
 
 physical properties of 164, 166, 168 
 
 ties, amount purchased in the United 
 
 States 156 
 
 cost of 156 
 
 Spruce pine, physical properties of 164 
 
 Spuyten Duyvil Rolling Mill Company, 
 
 early steel rails 4 
 
 St. Etienne Works, Harmet process at 415 
 
 St. John, I. M., report on steel rails 5 
 
 St. Louis and San Francisco Railroad: 
 
 articulated compound engine 34 
 
 plantations of tie timber 112 
 
 size and spacing of ties 121 
 
 ten-wheel type engine 33 
 
 St. Louis, Brownsville and Mexico Railway, 
 
 size and spacing of ties 121 
 
 St. Louis Southwestern Railway, size and 
 
 spacing of ties 121 
 
 Stamping rails 446 
 
 Standard drop testing machine 290 
 
 Stanton and Bairstow, experiments on re- 
 peated stress 281 
 
 Stassano electric furnace 385 
 
 State Railways of France (see also Road in 
 question) . 
 
 concrete ties on 105 
 
 State Railways of Germany (see also Road 
 in question). 
 
 concrete ties on 105 
 
 rail fastenings on 144 
 
 rails and tie plates on 125 
 
 Static axle loads (see Axle loads). 
 
 Stead, granular structure of metals 271 
 
 Steam: 
 
 effect of, on ties 158, 162 
 
 engines (see Locomotives), 
 
 loss in reversing, engines for rolling 
 
 mills 459 
 
 shovels, used in mining iron ore 351 
 
 use for compression of ingot 415 
 
 Steaming, effect of, on strength of wood 158, 162 
 Steel: 
 
 r (see Bessemer). 
 
 which take place during 
 
 cooling 426 
 
 compression strength (see Compression). 
 
 cooling curve of 426 
 
 ductility of (see Ductility). 
 effect of chemical composition on physi- 
 cal properties (see Element in ques- 
 tion). 
 
 Steel: 
 
 effect of mechanical work on strength 
 
 of 424, 427 
 
 repeated stress on strength 
 
 of 276 
 
 size of grain on strength of 424 
 
 temperature on strength of 284 
 
 electric (see Electric). 
 
 eutectoid 427 
 
 granular structure of (see Grain) . 
 
 heating curve 426 
 
 hyper eutectoid 427 
 
 hypo eutectoid 427 
 
 manufacture of (see Manufacture). 
 open hearth (see Open hearth), 
 rails (see Rail), 
 special (see Special steels), 
 strength of (see Strength). 
 tensile strength (see Tension) . 
 
 ties (see also Tie in question) 90 
 
 Steel Manufacturers of America (see Asso- 
 ciation of American Steel Manufac- 
 turers). 
 Stetson, E. E.: 
 
 flat spots in wheels 56 
 
 horizontal pressure on the rail 257 
 
 Stevens, Robert L., inventor of T-rail 6 
 
 Stock car 83 
 
 Stone: 
 
 ballast (see Ballast). 
 
 used in blast furnace 363, 364 
 
 weight of 316 
 
 Stoughton, B., piping and segregation 411 
 
 Stoves for blast furnace 359 
 
 Straightening press 446 
 
 rails 446 
 
 Strain, influence of, on strength of rail 
 
 steel 270 
 
 Street railway: 
 
 cars used on 87, 88 
 
 rails, corrugations in 209 
 
 sections of 19 
 
 specifications for 491 
 
 use of manganese in 339 
 
 roaring rails 209 
 
 ties purchased by 154 
 
 welded joints on 267 
 
 Stremmatograph tests 71, 212, 236 
 
 Strength: 
 
 of electric steel 386 
 
 rail, as shown by drop test 293 
 
 calculation of 239 
 
 tests of 288 
 
 steel compression (see also Compression). 
 
 effect of cold on 284 
 
 influence of stress and strain on 270 
 
554 
 
 Strength: 
 
 steel in rail 306 
 
 in rail head 193, 205 
 
 special (see Steel in question). 
 
 tension (see Tension), 
 ties (see Ties). 
 
 track 317 
 
 wood 158, 164, 166, 168 
 
 bending in rail (see Bending moment). 
 
 calculation of rail 239 
 
 effect of low temperature on rail 262, 288 
 
 repeated, Ewing and Humfrey's 
 
 experiments 274 
 
 Ewing and Rosenhain's 
 
 experiments 273 
 
 Howard's experiments 278 
 Spangenberg's experi- 
 ments 277 
 
 Watertown Arsenal 
 
 experiments 278 
 
 Wohler's experiments. 277 
 extreme fiber (see Extreme fiber stress). 
 
 in German rails 218 
 
 influence of, on strength of rail steel 270 
 
 lines of principal 251 
 
 proposed solutions of rail 210 
 
 rail 193 
 
 at point of contact with wheel 193 
 
 calculation of, by Love 240 
 
 effect of cold on 288 
 
 friction of joint on 262 
 
 inertia of track on 68 
 
 joint on 259 
 
 position of wheel load on 230 
 
 influence of kind of ballast on 229 
 
 produced by rolling 435 
 
 shearing (see Shear). 
 
 stremmatograph, determined by 212, 236 
 
 tests to determine rail, by Dudley. ... 212, 236 
 by U. S. Govern- 
 ment 218 
 
 stremmatograph 212, 236 
 
 tie 171, 177 
 
 on German railways 218 
 
 U. S. Government tests on rail 218 
 
 web, discussion of, by Hiroi 247 
 
 working rail 312 
 
 wood, compression under rail .... 171 
 
 in cross bending 168, 171 
 
 Stripping ingots 397 
 
 required at iron ore mines 351 
 
 Stubbs, F., influence of copper on steel .... 331 
 Styffe: 
 
 drop testing machine 288 
 
 experiments on effect of cold on rails. . . . 284 
 
 Subgrade: 
 
 bearing power of 187, 317 
 
 depression of, U. S. Government experi- 
 ments on 224 
 
 distribution of load to 180, 185 
 
 effect of heavy traffic on supporting power 
 
 of. 
 
 318 
 
 313 
 
 on rail breakages 
 
 long ties used on weak 188 
 
 supporting power required for different 
 
 classes of track 317 
 
 Suddenly applied load (see Dynamic). 
 
 Sugar gum, plantations of, for tie timber. . 112 
 
 Sulphur: 
 
 content in different iron ores 363, 364 
 
 rails of Bessemer steel. 11, 253, 310 
 
 open hearth steel 310 
 
 effect of, in steel 330, 390 
 
 removal of, from iron ore by roasting 344 
 
 segregation of 407 
 
 specifications for, in rails (see also Specifi- 
 cations) 328, 467 
 
 Support of the rail 90 
 
 Supporting power of ballast 180 
 
 of soils 313 
 
 of the subgrade 180, 313 
 
 of the subgrade required 
 for different classes of 
 
 track 317 
 
 of the track 188, 191, 317 
 
 Swedish Government Railroads, reinforced 
 
 joint 264 
 
 Swedish iron: 
 
 effect of repeated stress on 274 
 
 structure of 270 
 
 Sweet gum, physical properties of 164 
 
 T-rail (see also Rail) : 
 
 examples of American 6, 10, 14, 18, 460 
 
 English 19 
 
 European 18, 19 
 
 for street railways 19 
 
 high, examples of 19 
 
 specifications for 491 
 
 inventor of 6 
 
 Talbot, A. N., experiments on distribution of 
 
 pressure through gravel 187 
 
 Talbot, Benjamin: 
 
 advantages of open hearth furnace 382 
 
 comparison of Bessemer and open hearth 
 
 processes 380 
 
 continuous process of making steel 375 
 
 English practice of rolling rails 434 
 
 segregation, effect of aluminum on 405 
 
 Tamarack: 
 physical properties of 166, 168 
 
555 
 
 Tamarack: . 
 
 ties, amount purchased in the United 
 
 States 154, 156 
 
 cost of 154, 156 
 
 Tamping, effect of, on elastic curve of tie. . . 177 
 Tangent, comparison of rail failures on, with 
 
 curve 10 
 
 Technical Conventions of the Union (Ger- 
 man), design of rails 19 
 
 Tee rail (see T-rail). 
 Teeming (see Casting). 
 Temperature: 
 
 critical, in rolling 426 
 
 effect of, in casting 410, 413 
 
 rolling 427, 430, 434 
 
 on conversion of steel 366 
 
 strength of rails 284 
 
 finishing (see Finishing). 
 
 means for measuring, in rails 434 
 
 Ten-wheel type engine: 
 
 allowable axle loads 322 
 
 classification of 29 
 
 coupled, classification of 29 
 
 dimensions of 33, 72 
 
 effect of excess balance and angularity of 
 
 main rod 36,42,70 
 
 speeds of 24, 25, 26, 70 
 
 strength of track required for 322 
 
 typical dynamic wheel loads 73 
 
 wear of tires 198 
 
 weight of rail required for 322 
 
 weights of 33, 72 
 
 Tenacity (see Tension). 
 Tender: 
 
 effect of, on pressure of drivers 71 
 
 weights of 32, 33, 34 
 
 Tennessee Central Railroad Company, size 
 
 and spacing of ties 121 
 
 Tennessee Coal, Iron and Railroad Com- 
 pany: 
 
 rail mill 438, 444 
 
 shrinkage allowed at 444 
 
 teeming practice 399 
 
 Tension: 
 
 stress in rails on American railways (see 
 Stresses) . 
 
 German railways 218 
 
 in ties on German railways 218 
 
 test pieces, standard, American Railway 
 Engineering Asso- 
 ciation 512 
 
 Engineering Stand- 
 ards Committee . 487 
 Tension strength of iron-carbon alloys. . . . 329 
 Tension strength of special steels (see Steel in 
 question) . 
 
 Tension strength of steel 306 
 
 effect of size of grain on 424 
 
 influence of chemical composition on (see 
 Element in question). 
 
 Texan oak, physical properties of 164 
 
 Thermal cracks in head of rail 205 
 
 Thermoelectric measurements of stress 311 
 
 Thiollier helical lining 148, 150 
 
 Three-high mill 437 
 
 Thurston, R. H.: 
 
 arrangement of Bessemer plant 369 
 
 location of blast furnace 345 
 
 strength of materials 311 
 
 Tie plates: 
 
 allowable load under 170, 171 
 
 American 122 
 
 European 125 
 
 experiments in Germany 133 
 
 felt, on L. & N. W. R 19 
 
 function of 122, 133 
 
 tests at Purdue University 169 
 
 by American Railway Engineering 
 
 Association 169 
 
 on McKee 122 
 
 wooden 132 
 
 Tiemann, H. D., testing by impact 293 
 
 Ties: 
 
 allowable load on, as determined by bear- 
 ing on subgrade 188 
 
 allowable load on, as determined by bear- 
 ing strength under rail 171 
 
 allowable load on, as determined by ex- 
 treme fiber stress in bending 179 
 
 allowable load on, as determined by safe 
 
 bearing on ballast 180 
 
 amount used annually in the United 
 
 States 154, 156 
 
 annual charge of 115 
 
 bearing of rail on 122, 171 
 
 effect of dynamic load on 189 
 
 bending moment in 171, 177, 179 
 
 effect of dynamic load on 189 
 composite, used in Cuenot's experiments.. 173 
 
 composition (see also Tie in question) 96 
 
 concrete (see also Tie in question) 97 
 
 service tests on 104 
 
 cost of 154, 156 
 
 Cuenot's experiments on 172 
 
 depression of, in ballast 172, 176, 189 
 
 on German railroads 218 
 
 U. S. Government experiments . 219, 222, 233 
 
 distribution of load in 118 
 
 effect of dynamic load on 189 
 
 elastic curve of 172, 176, 177 
 
 granite 90 
 
 half-round 116 
 
Ties: 
 
 holding force of spikes in 138 
 
 kind of wood required for different classes 
 
 of track 188,317 
 
 kinds of wood used for 154, 156 
 
 life of, in track 115 
 
 metal (see also Tie in question) 90 
 
 permissible load under tie plate 171 
 
 pole 119 
 
 preservation of (see Treated ties). 
 
 | reaction of, in track 191 
 
 i size and spacing of, for different classes of 
 
 track 317 
 
 size of, on American railways 121 
 
 German railways 218 
 
 L. &N. W. R 19 
 
 spacing of, on American railways 121 
 
 German railways 218 
 
 L. &N. W. R 19 
 
 steel (see also Tie in question) 90 
 
 strength of 153 
 
 wood 158, 170 
 
 stresses in (see Stresses). 
 
 supply of 106 
 
 supporting power of 188, 191 
 
 treated (see Treated ties). 
 
 twelve foot, on muskeg swamp 188 
 
 U. S. Government experiments on depres- 
 sion of, in ballast 190 
 
 wood, future supply of 106 
 
 wooden plug for 149 
 
 Tilting open hearth furnace , 375 
 
 Timber: 
 
 supply of, for ties 106 
 
 in the United States 108 
 
 wasteful cutting of, for ties 116 
 
 Tirefond (see also Spikes, screw) 142 
 
 Tires: 
 
 cylindrical and conical, comparison of ... . 6 
 defective (see Defective equipment). 
 
 M. C. B. standard 7 
 
 wear of : 196 
 
 Titanium : 
 
 influence of, on segregation , 405 
 
 ferro, analysis of 405 
 
 effect of, in casting steel 341, 405 
 
 Titanium steel: 
 
 analysis of 340, 406 
 
 behavior of, at low temperatures 286 
 
 branding 447 
 
 cost of 341 
 
 ductility of 286 
 
 hardness of, as determined by the sclero- 
 
 scope 300 
 
 rails, production of 341 
 
 strength of 287, 340 
 
 Toledo Terminal Railway, concrete ties on . . 104 
 
 Torrey continuous rail 268 
 
 Track: 
 
 automatic inspector machine 45 
 
 bolt (see Bolt). 
 
 classification of , 317 
 
 depression of, U. S. Government experi- 
 ments on 219, 222, 233 
 
 effect of dynamic load on 189, 190 
 
 inertia of, on rail stresses 318 
 
 irregularities in, on wheel pres- 
 sure 45 
 
 speed on 29, 189, 190 
 
 experiments by U. S. Government 218 
 
 loading for different classes of 322 
 
 principles governing design of 313 
 
 stremmatograph experiments 236 
 
 strength of 310 
 
 U. S. Government experiments 218 
 
 Trackman's surface, depression of rail below 190 
 
 Tractive force, effect of, on pressure of 
 
 drivers 71 
 
 Trailing truck (see Truck). 
 
 Train: 
 
 effect of, on pressure of drivers 71 
 
 speeds of 21, 23 
 
 weights of 21, 23 
 
 Tram rails, examples of 19 
 
 Tramway rails, British standard 19 
 
 Transition points in cooling steel 426 
 
 Transportation: 
 
 cost of ore 345, 349 
 
 docks used in ore 351, 357 
 
 effect of, on location of blast furnace plant . 345 
 
 of iron ore on Great Lakes 349 
 
 vessels used in ore 349 
 
 Transverse fissures in head of rail 203 
 
 Trap rock, tests on ballast of 184 
 
 Tread (see Tires). 
 
 Treated ties: 
 
 amount treated in the United States 157 
 
 cost of 115 
 
 strength of 158 
 
 U. S. Forest Service, tests on 158 
 
 Tree plantations 109 
 
 Trenail: 
 
 Collet 148, 150 
 
 used on English railways 146, 147 
 
 Truck: 
 
 allowable weights on 320 
 
 leading, on freight and passenger en- 
 gines 32, 33, 34 
 
 trailing, on passenger engines 32, 33 
 
 weight of car 82, 89 
 
 on electric engine 79, 80 
 
 steam engine 32, 33, 34, 72 
 
Trzynietz Iron Works (Austria), hardness 
 tests at 
 
 Tucket, influence of arsenic on steel 
 
 Tup (see Drop test machine). 
 
 Turneaure, Professor, impact tests on 
 bridges 
 
 Turner: 
 
 sclerometer 
 
 thermoelectric measurements of stress. 284, 
 
 Twelve-wheel engine, classification of 
 
 29 
 
 Ulster and Delaware Railroad tie 104 
 
 Ultimate strength: 
 
 of steel 306,312 
 
 effect of repeated stress on (see 
 
 of wood 158, 164, 166, 168 
 
 Unciti, concrete tie 105 
 
 Union Pacific Railroad: 
 
 consolidation type engine 34 
 
 plantations for tie timber 112 
 
 size and spacing of ties 121 
 
 speed of trains 25, 26, 28 
 
 Union Railroad, steel ties on 90 
 
 Union Steel Works: 
 
 experiments on piping in ingots 399 
 
 Forsyth transferring ladle 390 
 
 United States (see also American) : 
 forest service (see U. S. Forest Service). 
 
 forests in 107 
 
 government tests on rail 218 
 
 Love on 240 
 
 ties used in, number, kind and value. 154, 156 
 
 timber, supply of, in 108 
 
 United States Census: 
 
 distribution of lumber product in the 
 
 United States 108 
 
 ties purchased in the United States 156 
 
 United States Ordnance Department, hard- 
 ness tests 
 
 United States Steel Corporation (see Plant 
 in question). 
 
 Universal metallic tie 94 
 
 Unwin, Professor, remarks on factor of safety 312 
 
 U. S. Forest Service: 
 
 half-round tie, proposed by 116 
 
 holding force of railroad spikes 139, 152 
 
 strength of different woods 158, 164, 166 
 
 tests on treated ties 158 
 
 the hardy catalpa 110 
 
 ties purchased in the United States 154 
 
 timber supply of the United States 108 
 
 303 
 
 Vandalia Railroad: 
 Mogul type engine . 
 Pacific type engine . 
 
 Vandalia Railroad: 
 
 speed of trains 25 
 
 Vaughan, H. H., flat spots on car wheels . . . 55 
 
 Velocity of load, effect of 69 
 
 Vessels for transporting ore 349 
 
 Vignole rail 18 
 
 Voiron St. Beron Railway (of France), con- 
 crete tie 105 
 
 Von Maltitz, E.: 
 
 blowholes 389, 391, 404 
 
 effect of recarbonizing 391 
 
 Von Schrenk, H.: 
 
 cross-tie forms 116 
 
 use of wooden tie plate 132, 133 
 
 Wabash Railroad: 
 
 Atlantic type engine 32 
 
 speed of trains 25 
 
 Walker, W. R., agitation of steel in casting. . 390 
 
 Walnut, plantations of, for tie timber Ill 
 
 Wanner pyrometer 434 
 
 Washington, Baltimore, Annapolis Railway, 
 
 car used on 87 
 
 Water: 
 
 contained in iron ore 360, 363 
 
 cooling curve of 425 
 
 effect of, on gravel ballast 187 
 
 strength of wood 168 
 
 subgrade 314 
 
 removal of, by dry blast of Gay ley 360 
 
 from iron ore by roasting .... 344 
 used in ballast tests, Pennsylvania 
 
 Railroad 184 
 
 Water hickory, physical properties of 164 
 
 Waterhouse, G. B.: 
 
 examination of strength of rail steel 309 
 
 titanium steel 340, 406 
 
 Water oak, physical properties of 164 
 
 Watertown Arsenal: 
 
 examination of steel at different stages 
 
 of rolling 420 
 
 experiments on repeated stress 278 
 
 tests on joints 260, 264 
 
 steel at different temperatures. . . 286 
 Wear of rails: 
 
 Dudley's investigation of 326 
 
 form for reporting 517 
 
 iron 1, 2 
 
 Kirkaldy on 205 
 
 machine for testing 304 
 
 of similar chemical composition 327 
 
 Wear of tires 196 
 
 Web rail: 
 
 principles governing design of 16 
 
 strength of 247, 252 
 
 steel from 307 
 
558 
 
 Web rail: 
 stresses (s< 
 
 Weber, depression of tie in ballast 189 
 
 Weber joint 264 
 
 Wedding, experiments on nickel steel 334 
 
 Weight: 
 
 ballast 316 
 
 cars 85, 89 
 
 locomotives, electric 74, 79, 80 
 
 increase in 15, 30 
 
 modern steam 29 
 
 of rail for different conditions of loading 310, 322 
 specifications for (see also Specification 
 
 in question) 476 
 
 wood 158, 163 
 
 Welded joints 267 
 
 Wellington, A. M., on soft rails 328 
 
 Wellman tilting furnace 377 
 
 Welsh, Ashbel: 
 
 rail section 5 
 
 report on steel rails 5 
 
 West Coast Railway of England, speed of 
 
 trains 25, 28 
 
 West Jersey and Seashore Railway : 
 
 experiments on horizontal component of 
 
 wheel pressure 259 
 
 speed of trains 24, 28 
 
 West, T. D., manufacture of ear wheels 57 
 
 Western hemlock, physical properties of 166, 168 
 
 Western larch, physical properties of 166 
 
 Western Railway Club, flat spots on car 
 
 wheels 60 
 
 Western Railway of France : 
 
 screw spikes on 143, 144 
 
 Westinghouse, George, on electric loco- 
 motives 74 
 
 Wheel: 
 
 defective (see Defective equipment). 
 
 dynamic load of (see Dynamic). 
 
 effect of position of, on stresses in rail .... 230 
 
 flat spot in 54 
 
 horizontal pressure of, on curves 259 
 
 impact of 62 
 
 load (see also Axle loads) : dynamic augment 
 
 of 69 
 
 effect of spacing 
 
 of wheels on. . 247 
 maximum, allow- 
 able on rail .. . 319 
 main, increase in pressure due to angu- 
 larity of main rod 35 
 
 manufacture 57 
 
 path of, on irregular track 47 
 
 pressure of, on rail (see also Pressure) .... 21 
 spacing given in modern bridge specifi- 
 cations 211 
 
 Wheel. 
 
 spacing of cars 85, 89 
 
 locomotives 32, 33, 34, 72 
 
 stresses at point of contact with rail 193 
 
 tests on chilled car 57 
 
 tire wear 196 
 
 tread (see Tires). 
 Wheel base: 
 
 cars 85, 89 
 
 electric locomotives 79, 80 
 
 steam locomotives 32, 33, 34 
 
 Wheeling and Lake Erie Railroad, size and 
 
 spacing of ties 121 
 
 White ash, physical properties of 164 
 
 cedar, physical properties of 164 
 
 elm, physical properties of 164 
 
 oak (see also Oak) : 
 
 force required to pull spike from . . 139 
 
 physical properties of 164, 168 
 
 tie, annual charge of 115 
 
 pine, physical properties of 164, 168 
 
 spruce, physical properties of 166 
 
 Whitwell stove 359 
 
 Whitworth's process for compression of 
 
 ingot 411,415 
 
 Whyte's classification of locomotives 29 
 
 Wickhorst, M. H.: 
 
 blast furnace practice 363 
 
 comments on Howard's rolling tests 420 
 
 description of process of making Bessemer 
 
 steel 374 
 
 description of process of making open 
 
 hearth steel 378 
 
 energy dissipated in drop test 293 
 
 flow of rail head under wheel loads 205 
 
 grain in head of rail 432 
 
 strength of rail head and web 252 
 
 steel 304 
 
 Wille, increase in axle loads 15 
 
 Williams' process for compression of ingot 410, 415 
 
 Williams, R. Price, iron rails 1, 2 
 
 Willow oak, physical properties of 164 
 
 Winckler: 
 
 depression of tie in ballast 189 
 
 elastic curve of tie 172 
 
 Wingham, influence of copper on steel 331 
 
 Winslow and Griswold, Bessemer Steel 
 
 Works at Troy 4 
 
 Wire tests of Professor Goss 63 
 
 Wohler's experiments on repeated stress. . . . 277 
 Wolhaupter: 
 
 joint 264 
 
 tie plate 122 
 
 Wood: 
 
 ferrule for spike 19 
 
 future supply of 106 
 
Wood: 
 
 key for double-headed rail 19 
 
 preservation of (see Treated ties). 
 
 specific gravity of 163 
 
 strength of 158, 164, 166, 168 
 
 effect of moisture on 168 
 
 steaming on 158 
 
 treatment on 158 
 
 test of, by American Railway Engineer- 
 ing Association 169 
 
 tie (see Ties). 
 
 plates 132 
 
 unit stresses recommended by American 
 
 Railway Engineering Association .... 168 
 
 used for ties, kinds of 154, 156 
 
 weight of 158, 163 
 
 Working load, relation of, to ultimate 
 
 strength 311 
 
 pressure, steam locomotives, 
 
 37, 40, 41, 42, 43, 44 
 
 Wiirttemberg State Railway (see Kingdom 
 
 of Wiirttemberg State Railways). 
 Wyandotte Rolling Mill, early steel rails ... 3 
 
 Yellow locust, plantations of, for tie timber . . Ill 
 
 oak, physical properties of 164 
 
 Yield point (see Elastic limit). 
 Yielding: 
 
 coefficient of, in ballast on German rail- 
 ways 218 
 
 of tie in ballast, amount of 172, 176, 189 
 
 effect of, on stress in rail 243 
 
 Zimmerman, H.: 
 
 depression of tie in ballast 189 
 
 effect of dynamic loading 69 
 
 elastic curve of tie 172 
 
 stresses in the rail 217 
 
 Zinc chloride treatment (see Treated ties) . 
 
 Zores iron 173 
 
" N ° V V , 
 
 " 
 
 
 X 
 
 v -. * 
 
 
 l, ^ C* 
 
 ^v *>* 
 
 .y %. *< 
 
 
 % *** 
 
 
 <> ^ 
 
 
 
 
 
 ^ ^ 
 
 
 
 ^ .^ v 
 
 
 v° °-<, 
 
 ■^ ^ 
 
 ,^%. 
 
 ■'•/>, ^\ N " ;* ^ 
 
 •^ ^ 
 ^ 
 
 
 
 „ > v> 
 
 
 
 '^.oO 
 
 ■4. -/* 
 
 
. 
 
 \° °* 
 
 
 
 ,0 O 
 
 
 V < 
 
 
 •^ '^ 
 
 r»V s' ,, 'r. V 
 
 ^ ^. 
 
 
 
 '-., 
 
 *• -f 1 
 
 \^ ^ , 
 
 <$*%, 
 
LIBRARY OF CONGRESS ftl 
 
 iiiiiiiuiiiiii 
 
 021 218 399 4