a a paren\ waiter ._e te aa me Cornell University Library OF THE College of Hrchitecture — FO cts cecsieponmener epee REO 1076 ornell University Libra Tests of bond between concrete and steel a7 UNIVERSITY OF ILLINOIS: “BULLETIN: = “ Issuep Weer Ly ee Vol. XI pig: toe OE, DECEMBER, 8, 1913 eS No. ‘3 ‘ hel : gnntered as second lags matter Dec. 4, 1912, at the Poat Office at. Urbana, In, under the Act of aug 24, 1912, oe. TESTS OF BOND. BETWEEN o CONCRETE AND STEEL | “ ‘sf ' i ee 4h BY | _DUFF A. ABRAMS. . ‘ BULLETIN a ENGINEERING EXPERIMENT STATION _ PusLisEp BY THE Unrvarstry oF Intinots, Unsaxa “gt,” ¢ , 0 ee ' a Price; ONE DOLLaR’ 73 ( ds i, : EUROPEAN AGENT nae { QHAPMAN anp Hatt, Lrp;, LONDON es % ‘ ‘HE “Engineering ‘Experiment Station 1 was established by act: of the ‘Board of Trustees,. ‘December 8,;1908. - “It is the purpose’ of ' the Station. to ‘carry on investigations. along various lines of ‘ engineer- oy ing anil to study problems of. importance to proféssional engineers | and ' to the ‘manufacturing, . railway, maining, ‘constructional, “and ‘industrial Ps interests of the State: “Sy “The control of the. ‘Rugineeziig Experiment. Station i is. vested 4 in the heads of the.several ‘departmients ‘of the College of. Engineering. These constitute, the ‘Station Staff and, with the ‘Director, determine the: ‘char-_ acter of the investigations. to. be undertaken, | The’ work. is carried on, : under the. ‘supervision of the. Staff, - sometimes. by” esearch fellows as graduate work, ‘sometimes, by members of the instructional: staff of. the’. ‘College of Engineering, ‘but more. frequently by, investigators belonging ; ‘to the Station c corps, - __., The ‘Tesults ‘of these - favestightions ate published: in. the’ fo n of | bulletins, which, Tecord mostly, the experiments of the Station’s own’ staff , of investigators. There will also be issued from time to time in the form of. circulars, conipilations giving the results of the experiments of | engineers, ‘industrial ; orks, technical institutions, and. governmental testing departments. a The: volume and. number af the top. of the. title page of. iis: cover “are. merely arbitrary numbers and. refer to the general publications of the. University of Illinois; either above the tiile or below the seal és given the number. of Engineering | Experiment. Station bulletin’ or eirewlar, “which should be used in referring to these publications. . . For. ‘copies: of bulletins, circulars, or other information address the Baginering Sag Sistion Urbana, Tlinojs. oie UNIVERSITY OF ILLINOIS ENGINEERING EXPERIMENT STATION Buuwetin No. 71 . DrcEemBeER, 1913 TESTS OF BOND BETWEEN CONCRETE AND STEEL By Durr A. ApraAms, ASSOCIATE IN THEORETICAL AND APPLIED MECHANICS CONTENTS I. Inrropucrion Page Te: “Prelimitiaty® 3 cu avi newt ower ech ieee oeawawie nes 7 2... Scope of Bulletin: i eicavs acvivae ise gvvgees cia dens ened it 3, Acknowledgment ic cocina cau ot eats Pemee ee eg gay 9 II. Materrats, Test Preces AnD MretHops or TESTING 4, Concrete Materials ...... 0... c cece cece cee eee eee eens 9 5. \CONGTELG xs8 ours eave e eens Saceetigey ee Mivewie wees 11 6.. Reinforcing Steel ..isessese.sesaaw ae deans wewews een eeen 13 %, “Pullsout Specimens: 62... es. cee ee ne eee ee been te meen 13 8. Reinforced Concrete Beams..... 00... 0.0 cece eee ccc e eee eee 15 9. Auxiliary Test Specimens ........ 0. ee cece eee ee eee 1? 10. Storage of Test Specimens. ......... 0... cece eee eee 17 11. Method of Making Pull-out Tests.............. 00.0. eee eee 17 12. Method of Making Beam Tests.............0.. eee e eee 19 3.’ “Anise ary: U eStS act evet eeu koteas Wiel ea aoe Aes Bled pacts: 20 III. ExperrMentaL Data anp Discussion 14. Preliminary Discussion of the Nature of Bond Resistance.... 20 15. Strength of Concrete ..sc.5 erevsyeera vesiasewinsee ee ene 22 A. Puut-out TEsts 16. Classification of Pull-out Tests.............0 0. ee eee eee 25 17. Stresses and Deformations in a Pull-out Specimen........... 26 18. Phenomena of Pull-out Tests......... 00... ce eee cece eens 28 19. 20. 25. 26. 27. 28. 29. 30. él. 33. 34, 35. 36. 3%. 38. 39. 40. 41, 42. 43. 44, 45, 47. 48. 49. 22. 23. 24, ILLINOIS ENGINEERING EXPERIMENT STATION a. Effect of Variations in the Dimensions of Pull-out Specimens age Effect of Size of Plain Bars; Embedment Variable.......... 31 Effect of Size of Corrugated Bars; Embedment Variable...... 35 Effect of Size of Plain Bars; Embedment 8 in.............. 37 Effect of Length of Embedment; 114-in. Plain Rounds....... 37 Effect of Length of Embedment; 114-in. Corrugated Rounds.. 42 Effect of Variation in Diameter of the Concrete Block........ 45 b. Effect of Shape of Section and Condition of Surface of Bar Bond Resistance with Rusted Bars...........-. cece ences 4” Bond Resistance with Flat Bars.............. 000 c cece eee 49 Bond Resistance with T-Bars........ 0... ccc cee ce cece neues 49 Bond Resistance with Polished Round Bars................. 50 Tests with Tapered Bars............... cece eee e eee teas 51 ce. Effect of Condition of Storage Preliminary: so. savoninsee ean an eee gd amas enaaasavaccouee ys 53 Batches:3: and 4. ccc cag.ce eee g eo eave gah s eee saweeeNe cans 53 Batch 3B) sie avacaimavwaiare a eal atta eile Weta betnicacacarbaeatanaiiataca eee whee igepaary 57 BALCH Mis. a si davataalenndlicoactlaretcnate tudes buat eee ebm 58 Batch BO sass cp ete. ay anna ny en mie ace eects ee Bact econ: dy dseviata va aGhsoedionid aris 59 Discussion of Effect of Storage............ 00... cece eee 59 d. Bond Tests with Deformed Bars PTElUMI MARY hs Weiss ere tvnaag doo Sea ia eps emer s As Oia dee 60 Phenomena of Pull-out Tests with Deformed Bars........... 60 Basis of Comparison of the Bond Resistance of Deformed Bars. 65 Discussion of Deformed Bar Tests...............0.c0cceeee 66 PEPER ded Bare x yicsnsaist goed aceasta stars eae tena dod fecguenadng a assis 72 Twisted Square Bars s..6c6 vices ce een eew eves vives een 73 Elements of the Bond Resistance of Deformed Bars.......... 16 e. Effect of Age and Miz Preliminary? faeeiviaisie aye seein ie ssiskae inerosdens dade begin dosieeea aie 81 Effect of Age on Bond Resistance .............-00000000... 84 Effect of Mix on Bond Resistance......................... 90 Mixes of the Form 1-2-m........ 0.60. e cece ccc eee ene 90 Mixes of the Form 1-n-2n......... cc cece cece cece eee 90 Relation of Bond Resistance to the Compressive Strength of _ Conerete aTiagisa tee alc Rietia ods cas pecan ued see aaa a 94 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 3 f. Effect of Anchoring Ends of Bars Page Preliminag? oxaie nip nsii esas asada saueeetMwetawuses 97 Bars Anchored with Nuts and Washer.............. 00 2.00. 97 Bars Anchored by Means of Hooks and Bends............... 99 Discussion of Tests with Anchored Bars.............-...005 100 g. Miscellaneous Tests Prelim <4.910ce¢igiivassee ty eee li newe ely er emiees 101 Repeated Loads on Pull-out Specimens.................04. 101 Effect of Distribution of Load on the Face of the Concrete Block? suis sannaieeinignd ape pa wld arse aine ee catered 104 Effect of Position of Bar During Molding.................. 104 Effect of Method of Applying Load in Pull-out Tests......... 105 Autogenous Healing in Concrete.......... 0... 0c cece eee 107 Effect of Loads Reapplied after Failure of Bond............ 108 Bond Resistance of Concrete which Set under Pressure....... 115 Compression Tests of Concrete which Set under Pressure..... 117 Effect of Loads Reapplied after Failure in Compression..... 121 Effect of Reinforcing the Pull-out Specimen against Bursting.122 B. REINFORCED CoNCRETE BEAM TESTS Preliminary: a: ns execs vse hae eS Oe ae ee RG 124 Measurements of Slip of Bar........... cee eee c eee eee 124 Computation of Stresses im a Reinforced Concrete Beam...... 125 Phenomena of Beam Tests....... 00... cece cece cece eee 126 a. 1909 Beam Tests Outline of Series ...... 0.0... cece cece eens pace. 130 Bond with Plain Round Bars.......... 0.0. c cece ence ee eeee 131 Bond with Deformed Bars ......... 0.0 ce cece cence eee eens 131 b. 1911 Beam Tests Olitline of Series aca wine hae tiee ee ew Rilds Hare 133 Basis of Comparison of Bond Resistance in Beams and in Pull- Olt SROBES hs dusicg aie iowa eid aime taiceserens ate elindha oe dirael Sees Gita 133 Bond with Plain Round Bars.............. cesses eee eeeee 140 Effect of Vertical Stirrups on Bond Resistance.............. 142 Effect of Depth of Concrete Below the Reinforcement........ 143 Effect of Span Length......-.-. ssc eese eee eee eee e eens 144 Effect of Length of Overhang of Ends of Beams............. 145 Bond with Corrugated Bars......... 0. esse eee eee eee eee 147 Bond Resistance in 1911 Beam and Pull-out Tests.......... 149 100. 101. ILLINOIS ENGINEERING EXPERIMENT STATION c. 1912 Beam Tests Outline. of Series: . accesses sie eaee ie vise eas eee dee es Bond with Plain Round Bars.......... 0.0.00 cee cece eee Bond with Plain Square Bars.......... 000s eee cee eee e eee Bond with Twisted Square Bars......... 0... cece eee eeeee Bond with Corrugated Bars.... 0.0.0... ccc eee ee eens Effect of Span Length... 0.0... ccc eee eee cette Effect of Position of Loads on Beams..............0..0 000 Effect of Auxiliary Reinforcement in Ends of Beam.......... Effect of Variation in Number and Size of Reinforcing Bars... Effect of Repeated and Continued Loads on Beams........... Pull-out Tests with 1912 Beam Series.................000. Bond Resistance in 1912 Beam and Pull-out Tests........... Variation in Results of Cube and Pull-out Tests............. d. Discussion of Reinforced Concrete Beam Tests Slip of Reinforcing Bar at Internal Points in Beams......... Distribution of Bond and Tensile Stress along Bar.......... Relation of Slip of Bar to Diagonal Tension Cracks........ Influence of Slip of Bar upon Beam Deflection............ Critical Bond Stress in Reinforced Concrete Beams.......... Working Stresses for Bond............ 0... c cece eee ee ee eee IV. Summary SUMMATY gd su nowy Sua airbus veda aaa ne Concluding Remarks ............. 0.00. eee eee ee eee ee H 11. Se IS ISEB ke eR HO 12, SLPMne Ree ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL LIST OF TABLES Page Briqneine Tests af Cemevisiacs s euwaws 4 6 eo50 0a nea a 4 Bawes ON YRRS 4 Mew a eRe s 10 Mechanical Analysis of Sand............eeeeee ade thugid daha smanerSex) guateviede-d Seaaees fave shea s 12 Mechanical Analysis of Stotteeis y cesses sce gues senncecaxy 12 Record of Batches of Concrete Used i in 1909 Pull- ii Tests. 23 Compressive and Flexural Strength of Concrete............ 24 Classification of Pull-ant: Testévsas ycxceae os eon oye eden d be eE 4d PEA 6 LOG Riko 26 Effect of Size of Bar; Embedment Variable............. esc e cece eee e eee e eens 31 Effect of Size of Plain Round Bars; Embedment 8 in............cecee eee eeees 36 Effect of Length of Embedment..........2eeeeceeeeees ee eee ee 89 Effect of Variation in Diameter of Concrete Block.......0ecseeee eset er eeeeeee 46 Effect of Shape of Section and Condition of Surface of Bar...... cece cece anes * 48 Effect of Condition of Storage........ See oes Be dgureeides dawned sents ganas 54 Notes on Deformed Bars Used in 1909 Pull-out Tests........... 000s e ese e cues 62 Pull-out Tests with Deformed Bars (1909)........ccecceceeeeeeee eee eneeetnes 68 Bearing Stresses Developed by Deformed Bars........cecceeeeeeeneeeeeenteees 79 Effect of Age and Mix........... cece ee cenee BSsy sanauac: Ses hantnsetsaudvene gue al cuadeyaes Somaaand 82 Comparative Values for Bond on Plain Round Bars........... cece eee eee e eee 89 Effect of Varying Quantity of Cement and Mortar.............. cece eee ce eens 92 Effect. of Anchoring: Ends: Of ‘Bars sca nce e.c1tsd saa eee SR eg SS 96 Miscellaneous Pullout: Tests. «4 scdis.e's ssceriiene overeat es asanain we 4 aegntee W graces 4 gael ioe - 108 Effect of Loads Reapplied after Failure of Bond......... 0. ccc eee cere eee eaee 110 Bond Resistance of Concrete which Set under Pressure........-. sess eeeeeeeeees 116 Compression Tests of Concrete which Set under Pressure........ 0... ceeeeeeeeee 119 Effect of Reinforcing the Concrete Block against Splitting...............-.ee uae 123 Data of Reinforced Concrete Beams—1909 SerieS....... eee e cece e cere e ene eenee 130 Tests of Reinforced Concrete Beams—1909 Series.........cee eee eec cee eeeerenes 131 Data of Reinforced Concrete Beams—1911 Series........ esse sees eee e eee e eens 134, 136 Tests of Reinforced Concrete Beaims—1911 Series. ....--..0ceseeeeeeceeeeneee 135, 137 Comparison of Bond Resistance in Beams and Pull-out Tests—1911 Series....... - 188 Summary of Bond Stresses in Beam and Pull-out Tests—1911 Series............. 141 Data of Reinforced Concrete Beams—1912 Scries........ cece cetera 152, 154 Tests of Reinforced Concrete Beams—1912 Series.......... ee ee ee 153, 155 Comparison of Bond Resistance in Beams and Pull-out Tests—1912 Series....... 156 | Summary of Bond Stresses in Beam and Pull-out Tests—1912 Series............ 160 Bond Resistance of 6-ft. Beams Reinforced with Plain Round Bars........... +. 162 Bond Resistance with Twisted Square, Square, and Round Bars 163 Variation in the Compressive Strength of 6-in. Concrete Cubes 176 Variation of Values of Bond Resistance in Pull-out Tests............ 176 Applied Load at First Slip of Bars at Intermediate Points in Reinforced Concrete Beams oe eee ce cece ee err eteerereee Ba fey Sis a dgsetatig & ate righ c Wore Asta GENER Se wide SMa ora 180 Distribution of Bond ‘Sieess in Reinforced Concrete Beams........+seeeeereeeee 198 LIST OF FIGURES Types of Pull-out Speciments UUsedian the Wests. eis iadas homies Oars ened 14 Typical Forms of Reinforced Concrete Beams Used in the Tests......-....+.+-- 16 Pull-out Specimen in Machine Ready for Test.......seeeestere cere e eee enennes 18 Diagram Showing Principal Stresses in a Pull-out Specimen...............0000. 27 Load-Slip Curves after Eliminating Size of Bar and Length of Embedment...... 29 Load-Slip Curves from Pull-out Tests with Plain Round Bars.....-....+-..++.++5 30 Load-Slip Curves for Plain Round Bars of Different Size......... ie fa falitve asa; eae 32 Bond Resistance of Plain Round Bars; Embedment Variable.........+++..--ee++ 33 Bond Resistance of Corrugated Bars; Embedment Variable....-...+....-seeeeeee 36 Bond Resistance of Plain Round Bars of Different Size; Embedment 8 TG 43 54 38 Load-Slip Curves for 1%4-in. Plain Round Bars; Embedment Variable............ 40 Bond Resistance of 14-in. Plain Round Bars; Embedment Variable.......... be 41 5 ILLINOIS ENGINEERING EXPERIMENT STATION Page 18. Load-Slip Curves for 1%-in. Corrugated Round Bars; Embedment Variable..... 43 14. Bond Resistance of 1% in. Corrugated Round Bars; Embedment Variable........ 44 15. Load-Slip Curves for 34-in. Plain Round Bars in Cylinders of Different Diameters 46 16. Bond Resistance of 34-in. Plain Round Bars in Cylinders of Different Diameters 47 17. Load-Slip Curves for Plain Round, Flat and T-Bars..........0eeeeeeee Ave Riek Ns 49 18. Bond Resistance with Polished Bars of Wedging Taper..........s.eeseeceeeeee 52 19. Load-Slip Curves for Different Conditions of Storage........sceseeeereereneeee 56 20. Maximum Bond Resistance for Specimens Stored in Air and in Water.......... 58 21. Deformed Bars Used in Pull-out Tests.............005 bi ovslepeidsncenbatece cuomrauts akedeued 61 22. Composite Load-Slip Curve from Pull-out Tests with Deformed Bars............ 64 23. Load-Slip Curves from Pull-out Tests with Deformed Bars......-...0ecseeeeeaee 70 24. Load-Slip Curves from Pull-out Tests with Deformed Bars (Enlargement of a Portion: :Of Figs. (28 iis. sicsctsvace Guar wtse Bay oik-2 sieht Sg eho Meet ah We Sarees Sting 71 25. Relative Values of Bond Resistance of Deformed Bars....... aga Rl aedte eR ENS 8 Bee 72 26, Load-Slip Curves for Twisted Square Bars.......- se sce ec eee e cert eet neeeneee 75 27, Effect of Age and Mix on Bond Resistance of Plain Round Bars..........-++065 84 28. Load-Slip Curves for Plain Round Bars Embedded in 1-2-4 Concrete,........... 85 29. Effect of Age and Mix on Bond Resistance of Corrugated Bars.........06000 2005 87 30. Load-Slip Curves for Corrugated Bars Embedded in 1-2-4 Concrete.............. 88 31. Load-Slip Curves for Plain Round Bars............. oghs s stesso raiciee eich ee eos 91 32. Bond Resistance for Different Mixes......... see ccc cee ccc cer cece eeet ne enenens 93 33. Relation of Bond Stress at End Slip of 0.0005 in. to the Compressive Strength of 6-in. Concrete Cubes ....... da ies SNeacolats ee weidvae Cette soe Pharl p< Sits AAEKE & BMTOiga HEM ANg Se aleaee 94 34, Relation of Maximum Bond Resistance to the Compressive Strength of 6-in. Concrete Cubes........... Syiie BaMrele ee wreie wR BGA see Asad se deed 95 86. Effect of Age on the Compressive Strength of Concrete..........seeec scence aee 96 36. Forms of Pull-out Specimens with Anchored Bars.............:eececccecsuees 97 87. Load-Slip Curves for Bars Anchored with Nuts and Washers...........---.0e00% 98 88. Load-Slip Curves for Repeated Loads on Pull-out Specimens................... 102 89. Load-Slip Curves for Repeated Loads on Pull-out Specimens................... 102 40. Load-Slip Curves for Miscellaneous Pull-out Specimens.....................000, 104 41, Load-Slip Curves for Specimens of the Form Shown............-:.sceeeeeeeee 106 42. Load-Slip Curves for Specimens of the Form Shown............eccesceeeeeees 107 43. Load-Slip Curves for Loads Reapplied after Failure of Bond... 112 44, Load-Slip Curves for Loads Reapplied after Failure of Bond 113 45. Load-Slip Curves for Miscellaneous Pull-out Specimens ................0.000005 117 46. Stress-Deformation Curves for 8 by 16-in. Concrete Cylinders.................., 120 47. Effect of Pressure During Setting on the Properties of Concrete................ 122 48. Effect of Length of Overhang of Ends of Beams....... priggact: Siauesy io baadas eke, acta earners 146 49. Comparison of Bond Stresses in Beams and Pull-out Specimens—1911 Series..... 148 50. Effect of Position of Loads on Reinforced Concrete Beams....................., 165 51. Load-Slip and Load-Deflection Curves for a Beam under Repeated Loads......... 169 52. Slip of Bar in Beams under Constant Loads...........cs cece eee e cece ee aveene 170 58. Slip of Bar and Center Deflection of Beam No. 1055.8............2ccceeeeceeue . 171 54, Load-Slip Curves for Beam No. 1055.8........ cece cece eects ete eeeeaes 172 55. Load-Slip Curves for Pull-out Tests from 1912 Beam Series...................., 174 56. Reinforced Concrete Beams in Machine Ready for Test.............ce...eceaee 183 BYS8. Reintorted Colitréte Beattie alter Teéstves < reais iiccd osanwndeeaad cence pee 184 to 190 64. Diagrams from Test of Beam No. 1055.6. One 1-in. Plain Round Bar; Span 10 ft. 194 65. Distribution of Tensile Stress in Bar in Beam No. 1055.38. One 1-in. Plain Round Bar; Spar 10 ft.... 2... sees cece nee eect eens cern ene eae cers rettacenasseses 197 66. Diagrams from Test of Beam No. 1049.3. One 1%-in. Corrugated Round Bar; Span: Ol ft sae votes sos. alba ees FG e Male Paw vee paweleaars aw eaulee Sauce 198 67. Diagrams from Test of Beam No. 1049-3. One 1%-in, Cortugated Round Bar Span 10) Ftc wick vas ohavenee tema one Sain Maratea wine aahS.S OMe & Sealat ace di ebiSia 199 68. Idealized Diagram Showing the Distribution of Tensile and Bond Stresses in a Simple Reinforced Concrete Beam........ Bie RHE es ccc set MOGs Beatie ea, 203 69-76. Load-Slip Curves for Bars in Reinforced Concrete Beams................ 221 to 228 77-86. Load-Deflection and End-Slip Curves for Reinforced Concrete Beams......229 to 238 TESTS OF BOND BETWEEN CONCRETE AND STEEL I. Invropuction. 1. Preluminary.—The usefulness of reinforced concrete as a struc- tural material depends on the strength and permanency of the bond between the concrete and the reinforcing metal, and for this reason bond resistance has received much attention from engineers and experimenters. It is said that Thaddeus Hyatt made tests to determine the bond be- tween concrete and iron bars as early as 1876. During the past decade numerous bond tests have been reported. These tests have been charac- terized by a lack of uniformity in the form of the test specimen and in the methods of conducting the tests, as well as by the wide variations in the values reported for bond resistance. In nearly all the tests thus far published values of maximum bond resistance only have been given. These test results and the discussions called forth by them have furnished the basis for a great variety of opinions as to the value of bond resistance. Many explanations of the source and nature of bond resistance have been given. Various methods have been advocated for increasing bond resistance and numerous devices have been employed for this purpose. Present American practice is fairly standardized as to the bond stresses to be used in designing, but a rational basis for the stresses used is lacking and there is a great diversity of practice in the methods of calculating these stresses. There are many phases of bond action which are not now understood. It is evident that the distribution of bond stress in reinforced concrete members under load and the nature and value of bond resistance under given conditions may well be the subject of experimental investigation. 2. Scope of Bulletin.—The tests reported in this bulletin were un- dertaken with a view to securing additional information on the nature of the bond resistance of reinforcing bars in concrete, to determining val- ues of bond resistance for a wide range of conditions, and to studying bond 8 ILLINOIS ENGINEERING EXPERIMENT STATION action in specimens of different forms. Tests were made on pull-out specimens and on reinforced concrete beams. In both forms of speci- men attention was given to obtaining accurate measurement of the slip of bar through the concrete as the loading progressed. In many of the beam tests the slip of bar at various points along its length was meas- ured for different loads. In the discussion of bond resistance the load- slip-of-bar relation has been utilized to a considerable extent. These measurements are useful in indicating the distribution of bond stress. They are particularly significant in the beam tests. In a few of the beam tests the distribution of bond stress was studied by measuring the changes in the stress in the longitudinal reinforcement throughout the length. The values found for bond resistance and the relative bond resistance found in beam tests and pull-out specimens are also interest- ing features of the investigation. The pull-out tests consisted in applying load to a short reinforcing bar embedded in a block of concrete. The concrete block was generally 8 in. in diameter and 8 in. long, with the bar embedded axially. In certain groups of tests these dimensions were varied. The size of bar used varied between 14 in. and 144 in. The pull-out tests covered a wide range and included effect of dimensions of specimen, effect of form of bar, effect of conditions of storage, effect of age and mix, using both plain and deformed bars, effect of different methods of loading, bond resistance of concrete setting under pressure, effect of reapplied loads, comparison with the bond resistance of reinforced concrete beams, etc. The deformed bars used included most of the forms in use at the time the work was begun, but it should be noted that the tests with de- formed bars were intended to bring out the action of the deformed bar as contrasted with the plain bar and not to determine the value of particular forms of bars. A special effort was made to determine the behavior of beams sub- jected to high bond stresses. The beams tested were 8 by 12 in. in section with an effective depth of 10 in. The span length was gener- ally 6 ft.; a few beams were tested with span lengths of 5 to 10 ft. All beams were tested with two symmetrical loads, generally at the one- third points of the span. With the exception of six tests, the longi- tudinal reinforcement consisted of a single bar of large diameter placed horizontally throughout the length of the beam. Both plain and de- formed bars were used. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 9 3. Acknowledgment.—The tests reported herein were made in the Laboratory of Applied Mechanics of the University of Illinois and formed a part of the investigations of reinforced concrete and other struc- tural materials which are being conducted by the Illinois Engineering Experiment Station. These tests cover the experiments which were designed with special reference to a study of bond between concrete and steel during the period 1909-1912. The work was done under the direc- tion of A. N. Talbot, Professor in Charge of the Department of Theo- retical and Applied Mechanics. The writer is indebted to Professor Talbot for many helpful suggestions in planning the tests and in inter- preting the data. In the 1912 beam series the work of testing was done under the writer’s supervision by Messrs. W. W. Manspeaker and A. W. Wand, senior civil engineering students of the class of 1912. These tests fur- nished the subject-matter of their baccalaureate thesis, where a very creditable report on the tests was presented. These men exercised great care in carrying out the routine of the work and they are to be com- mended for the way in which they met the demands of an unusually arduous program of tests. The other tests were made by the writer with the assistance of various members of the Laboratory staff. II. Marerrars, Test Pirces anD Metuops oF TESTING. 4. Concrete Materials—The materials were the same as used for other concrete and reinforced concrete specimens made and tested by the Engineering Experiment Station during the period 1909 to 1912. The quality of the materials may be considered as representative of that used in first-class concrete work in the central states. Cement. Most of the test specimens were made with Universal portland cement, which was furnished by the manufacturers. Chicago AA portland cement was used in the 1909 series of beam tests and in Batch No. 4 of the 1909 pull-out tests. Lehigh cement was used in part of the 1911 beam series. The Chicago AA and Lehigh cements were purchased from a local dealer. The results of briquette tests of these cements are given in Table 1. Samples were taken at intervals throughout the season. Each value given is the average of five briquette tests. Vicat needle tests on the three samples of the 1909 lot of Uni- versal cement showed initial set to occur at 1 hr. 45 m., and final set 10 ILLINOIS ENGINEERING EXPERIMENT STATION at 3 hr. 45 m. after mixing. Sieve analysis showed 96.5% passing a No. 100 sieve and 81.9% passing a No. 200 sieve. The 1912 lot of Universal cement (7 samples) gave the following average values: initial set, 3 hr. 5 m., final set, 6 hr. 32 m.; 97.2% passing a No. 100 sieve and 81.8% passing a No. 200 sieve. All cement tests recorded in this bulletin were made by Mr. B. L. Bowling, Assistant in Charge of Cement Laboratory, University of Illinois. TABLE 1. BRIQUETTE TESTS or CEMENTS. Each value is the average of five tests. Unless otherwise noted, standard Ottawa sand was used in the 1-3 briquettes. The results are expressed in pounds per square inch. Neat 1-3 Neat 1-3 Sample No. 7 days 28 days 7 days 28 days 7 days 28 days 7 days 28 days Universal Cement (1909) Chicago AA Cement. 1 607 732 197 281 742 783 205 270 2 595 772 179 280 716 807 232 306 288* 331° 3 417 853 160 278 725 768 176 254 Average 606 786 179 280 728 786 204 277 Universal Cement (1911) Lehigh Cement. 1 589 674 198 278 719 805 pase aa3° 248 329 2 684 709 227 283 3 653 731 240 319 4 662 696 214 282 Average || 647 702 220 290 Universal Cement (1912) 1 585 685 239 315 2 577 694 225 297 3 691 715 242 306 4 617 792 231 326 5 588 672 246 333 6 612 758 253 323 7 698 884 287 372 Average 624 743 246 325 * Made with sand used in making concrete; not included in average. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 11 Sand. The sand came from a deposit of glacial drift near the Wabash River at Attica, Indiana. Nearly all the clay or loam had been removed by washing. Fineness tests of samples from the three lots of sand are given in Table 2. A single set of briquette tests made from a sample of the 1909 lot of this sand, using Chicago AA cement, gave values about 10% higher than briquettes made from the same sample of cement using standard Ottawa sand. A set of briquettes made from the 1911 sand using Universal cement gave.values about 25% higher than the standard sand. The values from these tests are included in Table 1. The sand was well graded, but it will be noted that the 1912 sand was somewhat coarser than the other lots. Stone. The crushed limestone came from Kankakee, Illinois. It had been screened through a 1-in. screen and over a 14-in. screen. Mechanical analyses of a number of samples of this stone are given in Table 3. It contained about 48% voids. 5. Concrete—The concrete was proportioned by loose volume. The material was weighed also in order to obtain an independent check on the proportions of each batch. The ratios of the weights of the materials used in most of the specimens are given in Tables 4, 25, 27 and 31. In making the 1909 specimens the cement was measured in buckets, as was done for the other materials. This resulted in con- siderable variation in the batches. The practice of considering 95 |b. of cement equal to 1 cu. ft. was adopted for the 1911 and 1912 tests. This is in reality a weight proportioning for the cement and is more rational than the older method. The work of mixing and placing the concrete was done by men - of considerable experience in concrete work. The foreman has been employed in the laboratory since 1905, but has spent five or six months each year on contract work in concrete. All of the concrete except that used in the last two-thirds of the 1912 beam series was mixed by hand. The hand-mixing was done directly on the floor of the concrete labora- tory. The cement and sand were first mixed dry; the stone, which had previously been thoroughly moistened, was added and the batch then turned until it presented a uniform appearance. The first operation usually required five or six turnings, and the last two or three. Water was added and the material then turned until thoroughly mixed. In December, 1911, a 9-cu. ft., motor-driven, batch-mixer, made by the Marsh-Capron Mfg. Co., Chicago, Illinois, was installed. The second and thiftd beams in each set of three in the 1912 series and the corresponding auxiliary specimens, as well as some later miscellaneous 12 ILLINOIS ENGINEERING EXPERIMENT STATION pull-out specimens, were made of machine-mixed concrete. With the machine running continuously, the stone and sand were placed in the mixer and about one-half the required amount of water admitted. The cement was then added and the remainder of the water admitted at the same time. The amount of water used in each batch was measured and recorded. The drum of the mixer operated at about 22 revolu- tions per minute. Each batch was mixed for about 5 minutes after adding the cement. When the mixing was complete, the batch was dis- TABLE 2. MECHANICAL ANALYSIS OF SAND. ass Separation Per cent Passing Each Sieve inches 1909 1911 1912 3 0.28 99.7 99.8 100.0 5 174 95.0 967 88.0 10 091 77.5 74.4 54.3 12 067 70.3 67.6 47.5 16 Bids 62.8 61.0 41.7 18 043 51.1 50.2 32.9 30 027 28.2 30.3 21.2 40 019 16.9 17.0 13.3 50 -013 5.8 7.0 5.1 74 009 3.1 3.3 2.7 150 aeease 0.8 0.8 1.0 TABLE 3. Mecuanicat ANALYsis or STONE. Per cent Passing Each Sieve aed ae od Opening inches 22 Samples 8 Samples 5 Samples lin. 100 100 100 34 in. 87 95 95 1 in. 46 57 67 Mem pes 29 32 46 No. 3 0.280 14 18 26 No. 5 174 2 3 8 No. 10 091 1 2 3 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 13 charged onto the concrete floor, and was later removed to the forms directly with shovels or by means of a wheelbarrow. The concrete was mixed rather wet so that very little ramming was necessary after placin: it in the forms. 6. Reinforcing Steel—The plain round bars of 34-in. diameter and larger sizes used in the 1909 and 1911 tests were of high-carbon steel. The smaller sizes of plain round bars used in a few of the pull- out specimens and all the plain bars used in the 1912 series were of mild steel. The corrugated bars and most of the other types of de- formed bars used were of high-carbon steel. The Thacher bars were of mild steel. See Table 13 and Fig. 21 for details of deformed bars. Additional notes concerning the character and preparation of the steel are given with the discussion of the various groups of tests. Tensile tests on the steel are not given in this bulletin. In only a few of the tests was the steel stressed to the yield point, and these are noted in the tables. 7. Pull-out Specimens.—The specimens for pull-out tests consisted of a cylindrical block of concrete with a steel bar embedded axially. The blocks were generally 8 in. in diameter with embedment of 8 in. In some of the tests both the diameter of the block and length of em- bedment varied from these figures. The pull-out specimens were usually cast in a vertical position with the bar projecting about 16 in. below and 14 in. above the finished block. The forms, consisting of gal- vanized sheet-steel, were set up on a platform made of two 8-in. or 12-in. I-beams placed with their flanges about 2 in. apart. A planed cast-iron base plate made the bottom of the form for the concrete cylinder. A central hole in the base plate allowed the rod to pass through. These plates were removed when the specimens were taken from the forms. The specimens were tested in the same position, the load being applied to the longer end of the bar. The usual form of pull-out specimen is shown in Fig. 1 (a). Other pull-out specimens of unusual form are shown in Figs. 36, 41 and 42. Nearly all of the pull-out specimens with deformed bars in the 1909 series and certain other groups of specimens were reinforced against bursting by means of six or seven turns of 14-in. wire in the form of a spiral. This spiral was set inside the form before placing the con- crete and was near the outer surface of the concrete block; see Fig. 1 (b). In general, the spiral reinforcement was not used in the pull-out specimens made with the reinforced concrete beams. 14 ILLINOIS ENGINEERING EXPERIMENT STATION Generally, the test pieces were made in sets of five; but for a few of the tests the specimens were made in sets of two to ten. The pull- out specimens made as companion pieces to the 1911 and 1912 series of beams were made in sets of three. When practicable, the individual specimens of a set in the 1909 pull-out series were made from different batches, thus minimizing accidental effects in a given set. Frequent duplicate sets were made in order that comparison might be made 12”: (9) Fic. 1. Types or Putt-our Specimens Usep IN THE TEsTs. between concretes from different batches. The numbers of the batches from which the pull-out specimens of the 1909 series were made are given with the tables of test data. In the group of tests on deformed bars (Table 14) one specimen of each set was made from each of five batches which were mixed at intervals of about two weeks. In the series on “Effect of Age and. Mix,” (Table 16) three specimens with plain round bars and two with corrugated bars for each age were made from one batch, and the re- maining specimens from a second batch. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 15 8. Reinforced Concrete Beams.—The 1909 series of tests included 11 reinforced concrete beams; the 1911 series, 36 beams, and the 1912 series, 63 beams. All of the beams (with the exceptions mentioned in Table 27) were 8 by 12 in. in section. The length was generally 614 ft. A few of the beams in the 1911 and 1912 series were made in lengths of 514, 74, 814 and 10% ft. The arrangement of the rein- forcement for typical forms of beams is indicated in Fig. 2. The ends of the bars were squared, and extended flush with the ends of the beam. In all but two sets of beams the longitudinal reinforcement consisted of a single straight bar placed in the middle of the width of the beam, with its center 10 in. below the top. One set of three beams in the 1912 series was reinforced with 3 34-in. rounds and one set with 4 56-in. rounds. All of the beams, except a part of the 1911 series, were reinforced with vertical stirrups of plain round bars in sizes varying from %4 to 5g in. The stirrups engaged the longitudinal reinforcement and ex- tended to the top surface of the beams; they were placed 4 in. apart throughout the outer thirds of the beam in the 1909 series and 6 in. apart in other beams in which stirrups were used. In the beams in which the longitudinal reinforcement consisted of a single bar, the stirrups were V-shaped; in the beams in which three or four smaller rods were used for longitudinal reinforcement, the stirrups were U-shaped; in the 1911 and 1912 beams the ends of the stirrups were curved inward. The beams were made in wooden forms directly on the concrete floor of the laboratory, a sheet of building paper having been placed under the form. Generally enough concrete was mixed at one time to make two beams and the corresponding auxiliary specimens. Enough concrete was placed in the form at first to fill it a little above the point where the center of the reinforcing bar should be. After placing the reinforcing steel in position, the form was filled in layers of about 3 in. depth. From the first few batches of machine-mixed concrete only one beam was made; later it became the practice to discharge two batches together on the cement floor of the laboratory before commencing the making of the beams. This supplied enough concrete for two or three beams and the corresponding auxiliary specimens. The 1909 beams were considered a preliminary group. The beams of the 1911 and 1912 series were generally made and tested in sets of three. is ? 16 ILLINOIS ENGINEERING EXPERIMENT STATION 2-07--- port ——— 6*0* (a) Typical Beam; 1/909 Series. 1o" (6) Typical Geam, 1911 Serles —-——— 6*o* (¢) Beam with Ends Overhanging Supports 9in, 1H) Series. —— T 1 } t ! { 1o" = 4 \ (a)Group 6; 1/91! Series. to4 fs bee pee ees a GO Sea (e) 64 Beams, Varkable Spacing of Loads; 1912 Serres. (4) Beams loaded at 4 moi Veriable Span, 12 Series, (9) Group 16 , 1912 Series. SS a SS (h) Group 28 ; 1912 Series. Fic. 2, Typrca, Forms or Retnrorcep Concrete Beams Usep IN THE TESTs, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 17 9. Auailiary Test Specimens—From each batch of concrete three or more 6-in. cubes were made for compression tests. With several groups of the 1909 pull-out specimens in which the age at test or con- dition of storage were varied, sets of cubes were made for each varia- tion. The 6-in. cubes were made in metal forms. Plain concrete beams 6 by 8 in. in section and 42 in. long, for flexure tests, were made from each batch of the 1909 pull-out tests. Pull-out specimens were made from the same material as was used in many of the reinforced concrete beams of the 1911 and 1912 series. 10. Storage of Concrete Specimens.—The 1909 pull-out speci- mens and the corresponding 6-in. cubes were stored in damp sand unless otherwise noted. The plain concrete flexure beams were stored in open air. These forms were removed after four days except for the specimens tested at age of two days. The specimens were stored in a damp room where the temperature range was about 65° to 75° F. All of the reinforced concrete beams were stored in the open air of a room in which the temperature varied from 50° to 75° F. They were wet daily with water from a hose until about two months old. The pull-out specimens made with the 1911 and 1912 series of beams were stored under the same conditions as the beams. The 6-in. cubes made with the beams were stored in damp sand. The forms of the beams and their auxiliary specimens remained in place seven days. The variations in the conditions of storage of the pull-out speci- mens made at different times should be borne in mind when comparing the bond stresses developed in the different series of tests. The 1909 pull-out specimens were made between Jan. 1 and May 5; the 1909 beams between Feb. 1 and April 15; the 1911 series of beams between Jan. 1 and April 15; and the 1912 series of beams between Nov. 1, 1911, and Jan. 22, 1912. Other smaller groups included under “Miscellaneous Tests” were made during the season 1911-1912. The 1909 pull-out tests were generally made at age of about 60 days. In a few of the groups of pull-out tests the age ranged from 2 days to about 314 years. The 1909 beams were tested at about 100 days; the 1911 beams and pull-out tests at about 8 months; the 1912 beams and companion specimens at about 60 days. 11. Method of Making Pull-out Tests—In testing, the pull-out specimen was placed above the weighing head of the testing machine as shown in Fig. 8. The lower end of the embedded bar was engaged by the grips of the pulling head of the testing machine, the load being 18 ILLINOIS ENGINEERING EXPERIMENT STATION transmitted from the concrete cylinder which rested on a planed cast- iron base plate, through a rubber cushion to a spherical bearing block through which it passed to the weighing head of the machine. The rubber cushion served to reduce the rate of application of the load during the earlier stages of the tests and to minimize the effect of shocks arising from the slipping of the grips or vibration of the testing Concrete Block: Cast iron Base Plate ‘Rubber Cushion t Cast iron Pate - Spherical Bearing Block-—~|_” Top View of Yoke iia. ——, | Moving Head Fic. 3. Putt-out SPECIMEN IN Macuine Reavy ror Tsst. machine. The spherical bearing block allowed the bar to take a vertical position and tended to prevent bending action due to the bar being non-central in the machine or not parallel to the axis of the cylinder. A 100,000-lb. Riehle testing machine was used for all the pull-out tests. In these tests the moving head of the testing machine moved at the rate of 0.05 in. per minute. In the pull-out tests the amount of movement of the free end of the embedded bar was measured by means of an Ames gage in contact with the upper or free end of the bar. The instrument was mounted ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 19 on a yoke which was attached near the top of the concrete cylinder, as shown in Fig. 3. This instrument is self-indicating and requires no manipulation during the test. It is graduated directly to 0.001 in., and fractional parts of a division may readily be estimated. In order to test the stability of the yoke and to determine whether the top face of the cylinder remains a plane section during the test, four additional Ames gages were attached to the yoke at different points along the diam- eter of the cylinder in a number of tests. The yoke remained perfectly stable. The concrete 14 in. from the bar showed a measurable depression, about 0.0002 in. at loads near the maximum, but no depression could be measured at points further than 1 in. from the edge of the bar. The load was applied continuously, except in a few of the tests in which the load was released and reapplied after the beginning of slip at the free end of the bar. As the test progressed the loads on the test- ing machine corresponding to an end slip of 0.0005 in., 0.001, 0.002, 0.005, 0.01, 0.02, 0.05, and 0.10 in., were noted. A slip of 0.0005 in. is about the smallest amount that should be used in making comparison, although smaller amounts can readily be measured. Two men were required to conduct a test; one man operated the testing machine and observed the loads, while the other observed the amount of movement of the bar and kept the test notes. 12. Method of Making Beam Tests.—The reinforced concrete beams were loaded in a 200 000-lb. Olsen testing machine. The beams were tested under two equal loads applied generally at the one-third points of the span length; exceptions to this method of loading are noted in the tables. In general the ends of the beams overhung the supports 3 in.; in some of the 1911 tests the overhang was 9 and 15 in., respectively. The supports consisted of roller or rocker bearings. Rollers were used also to transmit the load to the top of the beams. Fig. 56 shows a beam set up in the testing machine ready for loading. Center deflections and the movement of the ends of the reinforcing bars were measured by means of Ames gages. For measuring deflec- tions the Ames gage was attached at the middle of a wooden bar which was carried by metal points at the mid-depth of one face of the beam at points directly over the supports. A small metal bracket attached to the beam at mid-span, transmitted the movement of the beam to the gage. For measuring the movement of the ends of the reinforcing bar, the gages were carried by yokes in such a way that the plunger had a direct bearing against the end of the bar. 20 ILLINOIS ENGINEERING EXPERIMENT STATION In many beams of the 1911 and 1912 series observations were made on the amount of movement of the reinforcing bar with respect to the adjacent concrete at several points along the length of the beam as described in Art. 66. The method of attaching the instruments to the beam is shown in Fig. 56. The load was applied in increments of 1000 or 2000 lb. In all but five of the beam tests the load was increased progressively to failure. 13. Auailiary Tests—The 6-in. cubes were tested for compressive strength only. About one day before testing, the faces to be loaded were bedded in a thin layer of plaster of paris. At least one set of three cubes made with each batch of the 1909 pull-out specimens was tested at age of 60 days. The 6 by 8 in. plain concrete beams made with the 1909 pull-out specimens were loaded at the one-third points of a 3-ft. span with the 8-in. dimension vertical. The age at test was generally 60 days. III. ExprrrMEentaL Data anp DISCUSSION. 14. Preliminary Discussion of the Nature of Bond Resistance.—- In this preliminary discussion it will be necessary to anticipate certain conclusions which appear in the following pages. The tests therein reported indicate that if a bar embedded in concrete is subjected to a tensile stress sufficient to overcome the bond resistance and withdraw the bar, certain definite relations exist between the amount of movement of the bar and the bond stresses developed. In the case of plain bars of ordinary mill surface there is no appreciable movement of the bar until a bond stress about 60% of the maximum bond resistance has been developed. If the bar is further stressed until the slip amounts to, say, 0.1 in. it will be seen that the bond-slip relations have undergone numerous changes. After slipping begins, the bond stress increases with further movement of the bar, very rapidly at first, then more slowly until the maximum bond resistance is reached. After the maxi- mum bond resistance is reached, the bond stress drops off with further slip and at a slip of, say, 0.1 in. amounts to about 50 to 60% of the maximum. For purposes of this discussion bond resistance may be divided into two elements which we shall designate as (1) adhesive resistance and (2) sliding resistance. The term adhesive resistance is used to designate the bond resistance developed before movement of bar with respect to the adjacent concrete begins. Sliding resistance is applied to ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 21 the resistance developed as a result of the movement of the bar. Ad- hesive resistance may be considered as due to tangential adhesion be- tween the concrete and steel and to static friction. The origin of tangential adhesion is not well understood, but its presence is a matter of universal experience with materials of the nature of mortar and concrete. It is recognized that tangential adhesion does not account for all the resistance developed before slipping begins, but the difference may be attributed to static friction developed by the pressure or grip of the concrete on the bar. As soon as the adhesive resistance is over- come, that is, as soon as the bond stress exceeds the sum of the tan- gential adhesion plus the static friction, there is a movement of the bar with respect to the adjacent concrete. Although we cannot ac- curately divide this adhesive resistance into its component elements, the sum of these elements or the total adhesive resistance can be determined, and it is found that under certain conditions it bears a definite ratio to the ultimate bond resistance. The bond resistance which is developed after slip of bar begins may be said to be due entirely to sliding resistance. Friction between bodies in contact arises primarily from roughnesses of their surfaces; its value is expressed as the product of the coefficient of friction into the normal pressure which exists at the surfaces of contact. The static friction mentioned above is due to the same cause. The roughness of surface in the case of a bar embedded in concrete is due to inequalities in the surfaces of contact which arise from irregularities of section and align- ment of the bar and the corresponding conformation of the concrete. The coefficient of friction between concrete and steel for the conditions present in these tests has not been determined. The normal pressure at the surface of the bar may be due to the following causes: (a) initial stresses generated by shrinkage of the concrete during setting and_ hardening; (b) wedging of the bar in the concrete following a move- ment from its original position. This wedging is caused by the inequal- ities in the bar mentioned above and in itself will give an added resist- ance. Apparently it is augmented under certain conditions by concrete adhering to the bar. In the above discussion the various components of bond resistance have been mentioned in the order in which they become effective in re- sisting bond stress. It is evident from these considerations that what has been termed adhesive resistance, that is, the amount of bond resist- ance which may be utilized before slipping begins, is by far the more 22 ILLINOIS ENGINEERING EXPERIMENT STATION significant element. While frictional resistance is of importance, re- liance should not be placed on this element of bond resistance. It was stated above that we cannot accurately divide adhesive re- sistance into its component elements; however, certain tests do give some indication of the values of these components. A large number of specimens with polished round bars embedded 8 in. in 1-2-4 concrete tested at age of 60 days gave a maximum bond resistance of about 160 Ib. per sq. in. In these tests frictional resistance (both static and slid- ing) was reduced to a minimum and bond resistance after slipping began was almost nil. We may conclude then that 160 lb. per sq. in. represents about what may be expected for the tangential adhesion between steel and concrete of this quality. It seems probable that the same value may be used for the tangential adhesion between any clean steel surface and concrete of the quality used. Round bars with ordi- nary mill surface tested under the same conditions as the polished bars show slip to begin at, say, 260 lb. per sq. in. with a maximum bond resistance of about 440 lb. per sq. in., developed after a slip of about 0.01 in. had occurred. The difference of about 100 lb. per sq. in. between the bond resistance developed by the polished bars and by the ordinary bars when slipping “begins (corresponding also to the maximum for the polished bars) may be said to represent the value of static friction for the bars of ordinary mill surface above that’ of polished bars embedded in the same concrete. The additional bond resistance developed by the ordinary bars after slipping begins is due to frictional resistance. The above observations refer primarily to the general case of bond between concrete and plain bars; for deformed bars, certain obvious modifications in these statements would be necessary. It will be seen later that the conditions under which the specimen is molded and tested have an important bearing on the bond resistance developed in any case. In reinforced concrete beams, where the reinforcing bars are con- sidered to take the main tensile stresses, the phenomena of bond action are complicated by the stiffness of the adjoining concrete in resisting stretching concurrently with the steel. This results in anti-stretch slip, a term which is discussed at length in Art. 68. The presence of this stress makes it desirable to distinguish between the phenomenon of anti-stretch slip and the slip produced by ordinary beam action. 15. Strength of Concrete-—The compressive and flexural strength of the various batches of concrete used in the 1909 series of pull-out specimens are given in Table 4. For convenience of reference a sum- ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 23 TABLE 4. Recorp or Batcuss or Concrete Usep IN THE 1909 Puut-out Tzsts. The average compressive strength of 69 6-in. cubes from 19 batches of 1-2-4 con- crete is 2150 Ib. per sq. in.; the average modulus of rupture of 18 6 by 8-in. plain concrete beams loaded at the ¥ points of a 36-in. span, is 296 Ib. persq.in. The aver- age values with the mean variations for the cube and plain beam tests may be ex- pressed as 2150-58 and 296-16 lb. per sq. in. respectively. The tests given in this table were made at age of about 60 days. Stresses are given in pounds per square inch. Number of Specimens i i: Date Mixture ; Compreasive | Modulus of Batch Made from Batch by Mixture | Strength | Rupture of No. (1909) ; Loose by of 6-in. | 6x8x 36-in. Pull 6-in. | Flexure} Volume Weight Cubes Plain Beams ull-out | Cubes | Beams 1 Jan. 1 47 27 1 1-24 1-2.4-4.3 1815 376 2 Jan. 7 48 18 1 1-144-3 1-1.8-3.3 3060 359 3 Jan, 12 25 15 1 1-2-4 1-2.5-4.0 2898 550 4° Jan. 13 40 15 1 1-2-4 1-2.4-5.5 1913° 338° 5 Jan. 18 39 39 1 1-3-6 1-3.6-6.3 1908 282 6 Jan, 25 5 6 1 1-1-2 1-1,5-2.0 4061* 374 7 Jan, 22 5 6 1 1-4-8 1-5.0-8.8 1155* 151 8 Jan, 22 5 6 1 1-2-2 1-2.1-1.9 2696* 332 9 Jan. 22 5 6 1 1-5-10 1-6.3-10.2 533* 62 10 Jan. 27 43 6 1 1-2-4 1-2.4-4.2 2180* 255 11 Feb. 1 24 3 1 1-2-4 1-2.44.3 1840 375 12 Feb. 9 47 27 1 1-2-4 1-2.2-3.8 2637 303 13 Feb. 15 25 18 1 1-144-3 1-1.9-3.3 2688 385 14 Feb. 15 5 3 1 1-214-5 1-3.0-5.1 1555 223 15 Feb. 15 5 3 1 1-2-3 1-1.8-2.4 1733 302 16 Feb. 25 5 3 1 1-2-5 1-2.1-5.0 1685 122 17 Feb. 19 50 6 1 1-2-4 1-2.3-4.3 1683* 240 18 Feb. 25 5 3 1 1-2-1 1-2,1-1.1 2870 336 19 Feb. 25 30 21 1 1-3-6 1-3.4-6.2 1242 183 20 March 1 42 18 1 1-1-2 1-1.2-2.1 3000 367 21 March 6 45 21 1 1-4-8 14.6-8.3 1298 167 22 March 10 23 3 1 1-2-4 1-2 .3-4.2 2043 213 23 March 11 33 18 1 1-1-2 1-1.2-2.0 3488 335 24 March 15 51 6 1 1-2-4 1-2.0-3.5 1468* 151 25 March 19 25 3 1 1-2-4 1-2.4-4.3 1950* 165 26 March 20 30 18 1 14-8 14.8-8.5 800 27 March 23 57 6 1 1-2-4 1-2.4-4.3 1675 247 28 March 30 5 3 1 1-2-6 1-2.4-6.6 1108 113 29 March 30 10 3 1 1-2-0 1-4.2-0. Syne 357 30 March 27 53 6 1 1-2-4 1-2.4-4,2 1715 172 1 March 31 38 3 1 1-2-4 1-2.3-4.1 2300 332 32 March 31 5 3 1 1-2-8 1-2.4-8.0 1240 117 33 April 8 40 24 1 1-114-3 1-1.6-3.1 itt 371 34 April 19 42 3 1 1-2-4 1-2.5-4.5 1950 202 35 April 23 45 9 1 1-2-4 1-3.1-5.9 3040 427 ‘il 16 26 3 1 1-2-4 1-2.4-4.5 1770 238 i Aba 24 48 3 1 1-2-4 1-2.:4-4.2 2460 357 38 May 5 7 3 1 1-3-6 1-3.7-6.5 967 257 39 April 29 60 21 1 1-2-4 1-2.2-4.1 2190 379 40 May 3 55 36 wk 1-2-4 1-2.4-4.3 2560 Total 1198 444 39 ° Batch No. 4 was made from Chicago AA cement; all others from Universal cement. * Average of 6 tests; all other values for cube strength are the averages of 3 tests. 24 ILLINOIS ENGINEERING EXPERIMENT STATION mary of the strength of concrete from the most important groups of tests is given in Table 5. This table includes all the tests on 6-in. cubes which were stored in damp sand and tested at age of about 60 days. The average compressive strength of 69 6-in. cubes from 19 batches of 1-2-4 concrete in Table 4 is 2150 lb. per sq. in.; the average flexural strength of the same concrete, 296 lb. per sq. in. Values for the compressive strength of the concrete used in each group of specimens are given in the tables of test data. Other data of compressive strength for different mixes and ages are shown in Tables 16 and 18 and in Fig. 35. Values for different conditions of storage are given in Table 12. TABLE 5. CoMPRESSIVE AND FLEXURAL STRENGTH OF CONCRETE. All cubes given in this table were stored in damp sand. The plain beams were stored in the open air. Cube tests for other ages and conditions of storage are given in Tables 12 and 16. Stresses are given in pounds per square inch. 6 x 8 x 36-in. 6-in. Cubes Plain Beams se Method | Number | 4zeat Mix Cement of of Test Numker | Com- Mixing | Batches days Number Modulus of Tests Tessive of Tests of Rupture Btrength 1909 1-1-2 Universal Hand 3 61 3 362 12 3653 1-134-3 Universal Hand 3 62 3 338 6 2874 1-2-4 Universal Hand 19* 64 18 296 69 2150 1-214-5 Universal Hand 1 61 1 223 3 1555 1-3-6 Universal Hand 3 62 3 241 9 1372 14-8 Universal Hand 3 60 2 159 12 1102 1-5-10 Universal Hand 1 60 1 62 6 533 1-2-0 Universal Hand 1 72 1 357 a 1-2-1 Universal Hand 1 60 1 336 3 2870 1-2-2 Universal Hand 1 60 1 332 6 2696 1-2-3 Universal Han:) 1 61 1 302 3 1733 1-2-5 Universal Hand 1 61 1 122 3 1685 1-2-6 Universal Hand 1 65 1 113 3 1108 1-2-8 Universal Hand 1 73 1 117 3 1240 1-2-4 Chicago AA Hand 6t 100 18 1700 1911 1-2-4 Universal Hand 11 240 ie gies 1-2-4 Lehigh Hand 6 240 se 6 ia 3087 1912 1-2-4 Universal Hand 12 63 sa Ra 1-2-4 Universal Machine 31 63 “Ata ae a ee 1-2-4 Universal Machine 4 240 bl 12 3774 * Includes one batch (No. 4) made of Chicago AA cement. { From the concrete used in the 1909 beams. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 23 The compressive strength of 6-in. cubes from the 1912 1-2-4 con- crete tested at average age of 63 days was as follows: 36 cubes of hand- mixed concrete, 2200 lb. per sq. in.; 93 cubes of machine-mixed con- crete, 2800 lb. per sq. in. The machine-mixed cubes gave about 30% higher strength than the hand-mixed. However, it will be seen later that all of the tests do not bear out this conclusion; in some of the beam tests the hand-mixed concrete gave the higher values for bond and vertical shearing stress. In making some of the larger batches of concrete for the 1909 pull- out tests a period of from three to four hours elapsed between the making of the first and last specimens. From five batches of this kind’a set of three 6-in. cubes was made ag soon as the mixing was completed and a second set after about 50 pull-out specimens had been finished. The portion of the batch remaining on the mixing floor was turned two or three times during the interval between making the two sets of cubes. The average compressive strength of five sets of cubes made as soon as the mixing was complete, was 1756 lb. per sq. in. The average for cubes made 31% hours after mixing, was 1733 lb. per sq. in., a difference of about 1%. It will be remembered that the Vicat needle test showed this cement to reach final set at 334 hr. A. PuLt-out TEsTs. 16. Classification of Pull-out Tests—A total of 1500 pull-out tests are reported in this bulletin. The 1909 series included about 1000 pull- out specimens. Forty-two pull-out specimens were made as companion pieces to the 1911 series of reinforced concrete beams, and 180 with the 1912 series of beams. The remainder of the pull-out tests are grouped under “Miscellaneous Tests,” Art. 54 to 64. The pull-out tests will be discussed under eight different headings as shown in Table 6, which indicates the number of tests included under each sub-division and the tables and figures in which details of the tests may be found. An inspection of the table will show that these divisions are not exclusive, as such elements as effect of age, proportions of con- crete, condition of storage, etc., are found to be important variables in more than one of the groups. A few tests are included under two sub- divisions, but this is indicated in the tables of test data. However, each sub-division may be considered as a more or less complete series of tests, 26 ILLINOIS ENGINEERING EXPERIMENT STATION and they were so considered in designing and making the specimens. In making comparisons between different groups, the conditions of storage and the quality of the concrete as shown by the cube tests, should be taken into consideration. TABLE 6. CLASSIFICATION OF PuLL-ouT TsstTs. Test Data in Number Item of Tables Figures Tests No. No. a. Effect of Variations in the Dimensions of Pull-out Specimens . 163 7, 8, 9, 10 5 to 16 b. Effect of Shape of Section and Condition of Surface of Bar... . 90 11 17,18 c. Effect of Condition of Storage..............0seeeee eee eee 189 1 19, 20 d. Bond Tests with Deformed Barg............s0seeeeeeeeee 116 13, 14, 15 21 to 26 e. Effect of Age and Mix............. cece c eee e cece eee ee 265 16, 17,18 27 to 35 f. Effect of Anchoring End of Bar.............----.-.0-0 00+ 97 19 36, 37 g. Miscellaneous Tests...............cc eee cece cece eee eee 337 20, 21, 22, 24 38 to 45 Companion Tests to Reinforced Concrete Beams........... 222 28, 29, at, tng 33 49, 55 and « 17. Stresses and Deformations in a Pull-out Specimen.—As load is applied to the bar in a pull-out specimen of the form used in these tests, the tensile stress in the bar is gradually taken off along the em- bedded length by bond between the bar and the surrounding concrete. Thus the total tension in the bar, the total compression over the lower face of the concrete block and the total bond stress between the concrete and steel are equal. The principal stresses existing in a pull-out speci- men of the form generally used are indicated in Fig. 4. If we con- sider the bond stress to be uniformly distributed along the length of the bar, the bond unit stress is: P Oh where P is the total load on the bar, m is the perimeter of the bar, and h is the length of embedment. It is evident that owing to the elonga- tion in the steel due to tensile stress and the shortening in the concrete block due to compressive stress, acting in opposite directions, the greatest bond stress during the early stages of the test and consequently the first slip between concrete and steel must occur at the point where the bar enters the block, and that slip will always be a little greater here than at points of greater embedment. Experimental verification of this statement will be found in the tests. A numerical example will assist ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 27 in fixing ideas. Consider a 1-in. round bar embedded 8 in. in an 8-in. cylinder. Tests show that a specimen of these dimensions, of 1-2-4 con- crete about 60 days old, will withstand a bond stress of about 260 Ib. per sq. in. before slip begins. At this stress the total compression in the concrete and tension in the bar are each 6500 Ib. The stresses in the concrete and in the steel are 132 and 8300 Ib. per sq. in., respect- ively. If we assume that the deformations in the concrete and the steel are proportional to the stress and that the bond stress is uniformly dis- tributed along the length of the bar we shall have 0.0002 in. and 0.0011 in. for the total deformation in the concrete and steel, using 2 500 000 and 30000 000 Ib. per sq. in., respectively, as the moduli of elasticity | | Early Stages After Slipping _ of Test becomes General Distribution of Bond Stress e Fic. 4. Dracram SHOWING PRINCIPAL STRESSES IN A PULL-oUT SPECIMEN. of the materials. This gives a total slip at the lower end of the block of 0.0013 in. when slip at the free end first becomes perceptible. These considerations show that for a 1-in. bar the relative movement at the bottom due to steel deformation is over five times as great as that due to concrete deformation, and indicate that as far as the bond resistance is concerned the compressive stress developed in the concrete block is of minor importance as compared with the steel stress. For smaller bars the influence of the concrete stress is negligible. As an extreme case consider a 114-in. plain bar embedded 24 in. as in the tests in Table 9. Slipping at the free end of the bar began at an average bond stress of 278 lb. per sq. in. The average compressive stress over the lower face of the concrete block at this stage of the test was 540 lb. per sq. in. and the tensile stress in the bar, 21 300 lb. per sq. in. If the bond stress be considered uniformly distributed along 28 ILLINOIS ENGINEERING EXPERIMENT STATION the length of the bar, this gives a slip of 0.0112 in. at the bottom when slip at the top of the block begins. It will be found that for specimens of the form generally used in these tests, the bond resistance of plain bars increased as slip progressed and reached a maximum when the bar had slipped about 0.01 in. The tests on 114-in. plain bars in Table 9 and Fig. 11 bear out the conclusion that was reached from these com- putations, that with 24-in. embedment, the maximum bond resistance of the bar as a whole does not differ much from that corresponding to first slip at the free end. In addition to the longitudinal stresses set up in the concrete and steel, stresses are developed normal to the surface of the bar. During the later stages of the test these stresses become considerable and may be sufficient to split the concrete surrounding the bar. This is espe- cially evident with deformed bars, though the splitting action was found with plain bars. 18. Phenomena of Pull-owt Tests——Fig. 6 summarizes load-slip curves for plain round bars for a variety of conditions of age, mix, size of bar, length of embedment and storage. Each curve is a composite of from 5 to 10 tests. For ease of making comparisons all bond stresses have been plotted as a percentage of the maximum bond resistance. Only the portions of the curves preceding the maximum are plotted here; all of these tests are discussed in detail in the following pages. These curves are quite similar, considering the wide variations of conditions present. Attention should be called to the fact that in this figure the values of slip of bar were measured at the free end. Generally, slip at the free end began at a load between 60% and 80% of the maximum. In nearly all cases the maximum load occurred at an end slip of about 0.01 in. It seems that a movement of 0.01 in. between the concrete and steel was sufficient to destroy the irregularities which were most effective in increasing bond resistance after the adhesion was broken. The curves in Fig. 5 represent the successive condition at each point along the length of the bar. The basis of these curves is given in Arts. 19, 22 and 23. The solid curve may be considered as typical of the load-slip relation found in pull-out tests with plain bars. This curve represents what may be considered as the bond-stress-slip history of each unit of the embedded length of a plain bar. It exhibits the fol- lowing characteristics : (1) There is no measurable slip of bar until a bond stress of about 260 lb. per sq. in. has been developed. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 29 (2) Slipping begins at about 60% of the maximum bond resist- ance. (3) When the bar has slipped 0.001 in. at any point the bond stress there is about 75% of the maximum bond resistance. (4) When the slip at any point reaches 0.005 in. the bond stress there is 95% of the maximum. 700 |--4 ds 4 4 gout 600 ot@a— rue a 5 at i § 500 “ : “ Z | \ 00} 4 ~ ' 9 / a Lip 1 x t Chir § t+ yO Ss Q © /g-in Corrugated founds, Embedment Variable. 200 + /3-17, Plain Fourids, Ernbeamert Variable. ° Plain founds, Diarnerer » and Ermbeamert Variable. 100 { 1-2-4 COCTere, Age about Crrorths, | 0 Pees eif o Of 2 03 04 OS Sip of Gar- inches Fic. 5. Loap-stip Curves AFTER ELIMINATING S1zE oF BAR AND LENGTH oF EMBEDMENT. (5) The maximum bond resistance occurs at a slip of about 0.01 in., and the bond resistance decreases with further movement of the bar. (6) When the bar has slipped twice the amount which was meas- ured at the maximum bond resistance the bond stress has decreased only 10%. (7) When the bar has moved 0.05 in., five times the slip at the maximum bond resistance, the bond stress is still about 70% of the maximum. The curve shows that the term “running friction” loses its signifi- cance in view of the data of these tests, since, properly speaking, we are dealing with “running friction” throughout the test after move- ment begins. 30 ILLINOIS ENGINEERING &XPERIMENT STATION It should be borne in mind that these deductions apply only to test specimens of the form used, under a progressively applied load that does not exceed the yield point strength of the bar. It will be seen later that the continuation of a constant load after slipping has begun, and other variations materially modify the load-slip relation. The broken line in Fig. 5, from a series of tests on corrugated round bars, indicates the corresponding load-slip relations for this type of bar. No 100 90 60 70 fF 60 01 %-in, Plain Rounds, 50 1-2-4 Corrcrere, Gin, 02 of Storage /-Ir, Plait Rounds, Linbeadment Bin, Age 60 adays ° al 2 03 oO a 02 103 4 Bool 14-in. Plain Rounds 1-2-4 Coricrete Age Percent of Maximum Bond Resistance $328 8 § € B wa eB 8 Plain founds, db 8 OM. Age akeur 60aays. w 5 06 07 OES 03 04 OS Slip of Bar-inches Fic. 6. Loap-stip Curves From Puxt-our Tests with PLarn Rounp Bans. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 31 a—Hffect of Variations in the Dimensions of Pull-out Specimens. 19. Effect of Size of Plain Bars; Embedment Variable.—In this group of tests the diameter of the concrete cylinders for all the speci- mens was 8 in., but the size of bar and length of embedment varied as shown in Table 7, between the limits 14-in. plain round bar embedded 3 in. and 114-in. round bar embedded 16 in. It will be seen that these dimensions are such as to give a nearly constant ratio (about 50) be- TABLE 7. Errsct or SizE or Bar; EMBEDMENT VARIABLE. Diameter of concrete cylinders 8 in. in all cases. 1-2-4 hand-mixed concrete from Batches 11, 22, 25, 31, and 36. (See Table 4.) The average compressive strength of 18 6-in. cubes from this concrete, tested at age of about 60 days, was 1975 lb. per sq. in. Stresses are given in pounds per square inch. Length of Embed- Bond Stress at : Number Age at ; Maximum Size of = of Test ae ae Bond Ber inches diameters Tests days 0008 in. 001 in, Resistance Plain Round Bars. Yin. 3 12.0 4 71 376 389 476 ¥ in. 434 12.6 4 73 335 371 423 Yin. 6 12.0 5 71 263 316 408 & in.? 8 12.8 5° 72 266 295 405 34 in. 94 12.6 5 70 287 311 386 1 in. 1234 12.8 6 70 305 327 392 1 in.* 16 12.8 9* 75 275 298 359 AVGRR Gos cass |e 2 years Peewee we ed swoon eeu 72 301 330 407 Corrugated Bars. Xf in. aq. 34 17.5 5 71 282 335 739T 4 in. aq. 443 15.0 4 71 324 356 in. sq. 8 16.0 5 71 341 371 702t 34 in. 8q. 12 16.0 5 71 339 367 730T 1} in. rd.* 24 19.2 9* 77 268 301 688+ Average consis: |: «sesso nteses [4 ae pests 72 311 346 T15t © The same tests are included in Table 8. * Includes 4 testa made with group in Table 9. are: 4 eels Blocka were reinforced against bursting by means of 14-in. spirals. The maximum bond stress given in the tables for corrugated bars is the average stress developed at an end slip of 0.1 in. 32 ILLINOIS ENGINEERING EXPERIMENT STATION tween the embedded area and the cross sectional area of the bar and corresponds to an embedment of about 12 diameters for each of the bars. It was felt that this was the proper basis for a series of tests to show the relation between the bond resistance and the size of bar. The load-slip curves are plotted in the upper portion of Fig. 7%. In Fig. 8 the loads have been plotted for slips at the free end of the bar of 0.0005, 600 Embedment ia 400 1-2-4 Corcrefe, 50 Found Bars, Age J? days. 100 Oo Ol 02 OF 04 OS 06 O07 OS O2 40 Slip of Bar sriches Fic. 7. Loap-strp Curves ror Pain Rounp Bars or Dirrerent Size, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 33 0.001, 0.002, 0.005 in. and for the maximum bond resistance, correspond- ing to a slip of about 0.01 in.; the corresponding points are connected by straight lines. A slip of 0.0005 in. may be considered as the begin- ning of slip. In general the smaller bars gave a bond resistance slightly higher than the bars of larger size, but the relation is not well defined at all stages of the tests. The maximum bond resistance decreased as the diameter of the bar increased. The 36-in. bars gave values at the maxi- mum load about 15% higher than the 144-in. bars. The 14-in. bars embedded 3 in. were somewhat erratic in their behavior. Some of the €00 400 iN i J00 q 3 k 5s yy 200 Embedment Variable, 4-2-4 Concrere. 100 Plain Round Bars, Age jdays, - hte / Qianerer a Bar. Entedtnen-n 3 446 8 % lel 46 Lntbedmet-lianm 12 126 12 3 (26 18 128 Fic. 8.. Bonp ResIstANCE OF PLAIN RounpD Bars; EMBEDMENT VARIABLE. concrete blocks split at the maximum load. The cause for this is some- what problematical, but it probably was due to slight irregularities in the contact between the concrete block and the bearing plate and to lack of stiffness in the blocks which prevented them from distributing the compressive stress over their entire base. With a constant ratio between the average bond stress throughout the length of the bar and the tensile stress in the steel, that is, with the bars embedded a constant multiple of the diameter, there are several variables which may affect the bond stresses developed. Among these may be mentioned: 34 ILLINOIS ENGINEERING EXPERIMENT STATION (1) Variation in the compressive stress in the concrete block ; (2) Contraction in the section of the bar due to the tensile stress ; (3) Variations in the section and alignment of bars of different sizes. (1) The compressive stress in the concrete blocks in these tests varies directly with the section of the bar, if the bond unit stress is the same. The maximum compressive stress at the bottom of the concrete blocks in the tests on the 14-in. bars embedded 6 in. and on the 144-in. bars embedded 16 in. are 76 and 622 Ib. per sq. in., respectively, using 400 lb. per sq. in. bond stress in both cases. It is evident from these tests and the tests discussed in Art. 21 and 22 that a wide variation in the amount of compressive stress in the concrete block has very little influence on the bond stresses developed in the tests. (2) The total contraction of the section of the bar due to tensile stress will be proportional to the diameter of the bar for the same unit stress in the steel. For the 114-in. bars used in this group of tests the diameter of the bar is shortened as much as 0.0004 in. when it is carry- ing the highest stress. A portion of this shortening was counteracted by the corresponding expansion in the concrete, but since the distribu- tion of the concrete stresses is not known and the amount of this expan- sion uncertain, the relation of these deformations cannot be determined. It seems that the contraction in the section of the bar may have a slight influence in reducing the bond resistance of the larger bars for loads near the maximum. This influence would not be important in the group of tests discussed in Art. 22, since the steel stresses were not high. (3) The bars used presented surfaces which apparently were sim- ilar in all respects. It has frequently been observed, however, that small rolled bars are more irregular in section and alignment than larger bars. This fact may partially account for the higher stresses developed by the smaller bars in the later stages of the tests. The lower bond stress developed by the smaller bars after an end slip of 0.01 in. has occurred lends color to this opinion. These tests indicate that small bars give a somewhat higher bond resistance than larger bars. Earlier discussions of this subject have been based on the maximum bond stresses developed by bars of different sizes embedded equally without regard to the load-slip relations present. A group of tests of this kind is discussed in the following article. In Fig. 8 dotted lines have been drawn which indicate the gen- eral trend of the values of bond resistance in these tests for the various ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 35 amounts of slip. In constructing these lines least weight has been given to the values for 14 and 3¢-in. bars, since the load-slip curves in Fig. 7 show these tests to be abnormal. The points at which these lines inter- sect the vertical axis may be said to represent the bond stresses for a bar of infinitesimal diameter embedded an infinitesimal length; the em- bedded length is still equal to about 12 diameters. These values and values obtained in a similar way from Fig. 12 have been used to plot the load-slip curve shown by the solid line in Fig. 5. It will be seen that the values given in the figure for the two series of tests are nearly identical. This curve may be said to show the successive variations in bond resistance at each point along the length of the bar during the progress of the test after eliminating the size of bar and length of em- bedment. This consideration shows slipping to have begun at a some- what lower load than was indicated by the measurements of slip at the free end of the bar and that the maximum bond resistance from point to point was a little higher than that found for the bar as a whole. This is as might have been expected, since, up to the load at which slip reached a considerable amount and became general, each point along the length of the embedded bar was in a different stage of its load-slip history. After the maximum load the curve for infinitesimal embedment follows the general course of the other curves, which indicates that after the adhesion was entirely broken and slip became general, the frictional resistance was about the same for all sizes of bars of the kind used in these tests. It will be seen in the discussion of Fig. 12 in Art. 22 that a curve of almost exactly the same form is found from a consideration of a group of tests on 114-in. plain rounds in which the length of em- bedment varied. 20. Effect of Size of Corrugated Bars; Embedment Variable.— The group of tests on corrugated bars given in Table 7 may be con- sidered as forming a series parallel to the group of tests on plain rounds discussed in Art. 19; they were made from the same concrete. The specimens varied from a 14-in. corrugated square bar embedded 31 in. to a 114-in. corrugated round bar embedded 24 in. As in the group of tests on plain rounds these dimensions were such as to give approxi- mately a constant ratio (about 48) of bond area to the cross-sectional area of the bars, corresponding to an embedment of about 16 short diameters of the bars (21.4 diameters in the case of 114-in. corrugated round bars). The bond resistances for various amounts of end slip are plotted in Fig. 9. The values for 14-in. bar embedded 4% in. have \ 36 ILLINOIS ENGINEERING EXPERIMENT STATION 600 700 § § Bond Stress -lb. per 5p./7. g si Embeament Variable. 4-2-4 Corcrefe. Corrugated Bars. Blocks reintorced 100 against bursting. Age 72? o Size of Bar -in £59559 259. 23g. lg round Linbeament-in. Fg 43 a 42 aA e diam 140 (25 16 46 ala Fic 9. Bonp RESISTANCE OF CoRRUGATED BARS; EMBEDMENT VARIABLE. TABLE 8. Errect or S1zE OF PLAIN Rounp Bars; EMBEDMENT 8 IN. Made from the same concrete as the tests in Table 7. Stresses are given in pounds per square inch. 7 . Length of Bond Stress Size of Embedment Number Age at at End Slip of Maximum Bar ee A ee es es ol Test Bond ‘ : Tests days i Resistance inches diameters .0005 in. .001 in. Win. 8 16.0 4 72 323 339 381 5% in. 8 12.8 5* 72 266 295 405 34 in, 8 10.7 bt 81 275 303 387 1 in. 8 8.0 5 64 247 281 385 1 in. 8 6.4 12} 74 269 296 397 * The same tests as given in Table 7. : meh the similarity between these values and those for 9}4-in. embedment in Table 7; compare also tests in Table 10. } The same tests as given in Table 9. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 37 been omitted from the table and diagram after a slip of 0.001 in, since the data for some of the tests are not complete; the position of the two points in the figure indicate that the remaining points prob- ably would have occupied their proper places in the diagram. In general, slipping began at a bond unit-stress only a little greater than for the plain rounds. For end slips of 0.0005 and 0.001 in. the average bond stresses are 301 and 330 lb. per sq. in. for plain rounds and 311 and 336 Jb. per sq. in. for the corrugated bars; at this stage of the tests the stresses developed by the plain rounds were about 97% of those given by the corrugated bars. The highest bond stresses reported (end slip of 0.1 in.) are not materially different for the bars of the sizes used in these tests, and average about 715 Ib. per sq. in. The some- what lower values given by the 114-in. corrugated round bars are prob- ably due to the design of this bar as compared with the square bars; the projections on this bar present a smaller area to take the bearing stress which replaces the bond resistance after the failure of the ad- hesion than the square bars of type B which were used in the other tests in this group. 21. Effect of Size of Plain Round Bar; Embedment 8 in.—In this group of tests the diameter of the concrete cylinders was 8 in. and the length of embedment 8 in., while the diameters of the plain round bars used varied from 14, to 114 in. The results are summarized in Table 8. The load-slip curves are plotted in the lower portion of Fig. 7. The loads for various amounts of slip of the free ends of the different bars have been plotted in Fig. 10. The broken lines drawn in Fig. 10 show the trend of the values for the different amounts of slip and for the maximum loads. The values for end slip of 0.0005 in. may be taken as the beginning of slip. It is seen that in the earlier stages of the tests, the smaller bars develop the higher bond stresses; the load at beginning of slip of the free end of the bars varies from 323 lb. per sq. in. for the 14-in. rounds to about 253 lb. per sq. in. for the 1 and 114-in. bars. The maximum bond resistance is nearly constant and averages 391 lb. per sq. in. 22. Effect of Length of Embedment; 114-in. Plain Rounds.—In this group of tests the length of embedment varied from 4 to 24 in., corresponding to 3.2 to 19.2 diameters. One and one-fourth-inch bars were used in order to secure a wide range of embedded lengths without’ overstressing any of the bars. A summary of the tests is given in 38 ILLINOIS ENGINEERING EXPERIMENT STATION Table 9. The load-slip curves have been plotted in Fig. 11. The bond stresses corresponding to various amounts of end slip up to and including the maximum load are shown for the different lengths of embedment in the upper portion of Fig. 12. In the lower part of the figure the maximum resistance and the stresses after the maximum for slips of 0.02 in., 0.05 in. and 0.10 in. are plotted. The load-slip curve for 4-in. embedment drops off sharply after the maximum load, as a result of the splitting of some of the concrete blocks. Fig. 12 shows that the average bond stress corresponding to 600 Bond Stress -Lb per SQM, ge 8 8 o Dias, of Gar-lra / lmbeanent-diam 16 123 127 o 64 Fic 10. Bonp Resistance or PLain Rounp Bars orf DIFFERENT SIZE; EMBEDMENT 8 IN. beginning of slip at the free end of the bar varies but little with increased embedment. At a slip of about 0.001 in. the average bond stress is nearly the same for all lengths of embedment included in these tests and amounts to about 300 Ib. per sq. in. It is evident that when a slip of 0.001 in. has occurred at the free end of the bar, slip- ' ping has become general throughout the length of the bar, but it will be seen from a consideration of the conditions present and from the load-slip curves that the amount of slip represents quite different stages of the test in the specimens of different lengths of embedment. For the 4 or 6-in. embedment the difference between the amount of slip at the two ends of the block for any given load is not great, and ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 39 TABLE 9. Errect or Lenctu or EMBEDMENT. 1-2-4 concrete from Batches 10, 17, 24, 27 and 30. Diameter of concrete cylinder 8 in. in all tests. These specimens were made from the same concrete as the specimens with deformed bars in Table 14. The average compressive strength of 24 6-in. cubes from same concrete tested at about 60 days was 1760 Ib. per sq. in. Stresses are given in pounds per square inch. Length of Embedment Bond Stress at Number Age at End Slip of Maximum . . of Test — Bond inches diameters Tests days 0005 in, 001 in. Resistance 114-in. Plain Rounds. 4 3.2 5 74 265 314 375 6 4.8 5 74 243 391 420 8 6.4 12* 74 269 296 397 12 9.6 5 75 284 312 390 16 12.8 ot 75 275 298 359 24 19.2 5 76 278 300 328 1)%-in. Corrugated Rounds’. 4 3.6 5 74 228 272 830t 8 7.1 5 86 250 286 775t 16 14.2 5 76 281 306 846t 24 21.4 gt 77 268 301 688t * Includes five tests made with the group in Table 8 and two tests from Batch 37. + Includes five tests made with the group in Table 7. © Blocks were reinforced against bursting by means of a 14-in. spiral. t Bond stress corresponding to end slip of 0.1 in. 40 ILLINOIS ENGINEERING EXPERIMENT STATION we may expect that the load-slip relations given by these specimens (barring the premature splitting of the 4-in. blocks) are not much different from those for a unit of area at any point along the em- bedded length of the bar, while for the 24-in. embedment the load- slip relation at the free end of the block must be quite different from that at the other end. If the straight lines which represent the general trend of the values of bond resistance for the amounts of slip shown in Fig. 12, be extended to the left to intersect the vertical axis of coordinates, 600 500 & 400 & Ss 4 s 8 8 300 g 5 § 200 ag 1-2-4 Concrete /00 lz-n. Plain founds Qo oO Ol 02 .O2 .O7 O83 :O9 10 Slip of Bar-Inches Fic. 11. Loap-siip Curves For 114-1n. PLain Rounp Bars; EMBEDMENT VARIABLE, the points of intersection may be said to represent the bond resistance per unit of area for a very short length of the embedded portion of the bar. These points have been plotted as shown by the dotted curve in Fig. 11. The load-slip curve which they form differs from the others in the same group in that slip begins at a somewhat lower proportional load and the bond resistance reaches a higher maximum value than indicated by the other curves, and the curve drops off a little more rapidly after the maximum. This curve may be said to represent the conditions for an infinitesimal embedment of a bar of this kind, and indicates what happens at each point along the em- bedded length of a bar during the progress of the pull-out tests in ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 41 500 400 200 Before Maximum Load 8 Ss 1-2-4 Concrete Age 75 days Bond Stress -Lb.per 59./n. 9 8 After Maximum Load 100 4 6 a /2 16 4 ” Gam 32 4G 6A 96 28 192 Fic. 12. Bonp RESISTANCE OF 14-IN. PLatn Rounp Bars; EMBEDMENT VARIABLE. 42 ILLINOIS ENGINEERING EXPERIMENT STATION the same way as indicated in Art. 19 for another group of tests. The values derived in this way are as dependable as those given for other lengths of embedment, since they are based on the mean stresses from the entire group of tests. Data from these tests were used in plotting the solid curve in Fig. 5. The points in Fig. 5 for the plain round bars were derived from these tests and from the tests dis- cussed in Art. 19. Only one curve has been drawn for the plain bars, although the values from the two series of tests are indicated by distinctive symbols. The similarity of the values obtained in this way is noteworthy, when we consider that these specimens were made from bars of different size and from different batches of concrete. The uni- formity of these values gives considerable confidence in the tests and in the conclusions based on them. These values are significant in arriving at a proper conception of bond action for bars of this kind. The maximum bond stress decreased as the length of embedment increased and varied quite uniformly from 410 lb. per sq. in. for 6 in. embedment to 320 Ib. per sq. in. for 24 in. embedment. The lines in the lower division of Fig. 12 converge at a point about 290 lb. per sq. in. bond stress and 30 in. embedment. ‘This indicates that with a longer embedment than 24 diameters the maximum average bond resistance would be less than that which gives first slip with shorter embedments. When a slip of 0.1 in. is reached, approximate equality of bond stress for all lengths of embedment is again found at about 235 Ib. per sq. in. 23. Effect of Length of Embedment; 114-in. Corrugated Rounds; Blocks Reinforced against Bursting—The tests with corrugated bars embedded in reinforced blocks summarized in Table 9, may be con- sidered as parallel to the group of tests on plain rounds made from the same concrete and included in the same table. In this group the concrete cylinders were all 8 in. in diameter, but the length of em- bedment varied from 4 to 24 in. The concrete blocks were reinforced against bursting by means of a 14-in. wire in the form of a spiral as shown in Fig. 1 (b). The load-slip curves are given for the four lengths of embedment in Fig. 18. Each curve is the composite of all the tests in a set. According to the practice in all tests on deformed bars described in this bulletin, the highest bond stress which has been reported was that developed at an end slip of 0.1 in., although in most of the tests the load continued to rise beyond that point. In Fig. 14 the bond stress corresponding to various amounts of end ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 43 slip of the bar have been plotted against the length of embedment. The general direction of each of the zig-zag lines is indicated by the short dotted lines at the left margin of the figure. The points of intersection of these lines with the vertical axis of co-ordinates may be taken to represent the action of a similar bar of infinitesimal em- bedment, based on the general trend of the values from these tests. The 3900 s § 8 8 Bond Stress -Lb per 53. /n. § /-2-4 Concrete, 1 -In, Corrugated Rounds, Blocks reinforced against bursting: Age 7 aays. 100 oO oO Ol 02 ‘07 aos we do So of Bar-Incnes Fic. 13. Loap-strp Curves For 1¥%-1n. CorruGATED Rounp Bars; EMBEDMENT VARIABLE. load-slip curve corresponding to these points is shown by the dotted line in Fig. 13. As may be expected, it follows closely the curve for the 4-in. embedded length. If the lines in Fig. 14 are projected to the right it will be found that they intersect at approximately the same point. The interpreta- tion of this feature of the tests is probably the same as that suggested in the preceding article for plain round bars. For an embedment of more than 56 in. (50 diameters) the excessive deformation developed 44 ILLINOIS ENGINEERING EXPERIMENT STATION in the concrete and steel would so modify the distribution of bond stresses that an average stress greater than, say, 320 lb. per sq. in. could not be developed in a pull-out test with a bar of this kind. In these tests slip began at a lower unit stress and the maximum stress was higher in the specimens of short embedment than in those 900 700 8 8 8 Sond Stress -L per 5o.1ta /-2-4 Corncrefe Age 7édays oO Linbeamerl-in, 4 a le 16 20 24 & “ Diam 36 Ww M2 alt Fic, 14, Bonp RESISTANCE oF 1)£-1N. CorrucATED Rounp Bars; EMBEDMENT VARIABLE. of longer embedment. The bond stress at a slip of about 0.002 in. was about the same (320 Jb. per sq. in.) for all lengths of embedment in- cluded in the tests. There is a striking similarity between the bond stresses found for the beginning of slip of the 114-in. corrugated round bars and those for the 144-in. plain rounds in the preceding article. Up to a slip of about 0.002 in. the plain rounds give slightly higher values at each ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 45 stage of the tests. It is notable also that the line of equal bond stresses for all lengths of embedment for each of these groups came at a bond stress of about 300 lb. per sq. in. and at approximately the same amount of slip. The load-slip curve for the 114-in. corrugated rounds correspond- ing to an infinitesimal embedment has been plotted in Fig. 5. The interpretation of this curve is much the same as given above for the curve obtained in a similar manner for plain bars and included in the same figure. A comparison of the values for an infinitesimal embedment for this group of tests on 144-in. corrugated rounds, with the similar curve from the two groups of plain bars discussed in Art. 19 and 22, is of interest. The plain rounds gave higher bond resistance than the corrugated bars for all amounts of end slip up to that corresponding to nearly the maximum resistance of the plain bars; after this point the corrugated bar steadily gained in bond re- sistance while the bond resistance of the plain bar decreased in the typical manner, after passing a slip of 0.01 in. Pull-out tests with corrugated bars embedded in concrete blocks without spiral reinforcement are discussed in Art. 64. 24. Effect of Variation in Diameter of the Concrete Block.— Table 10 summarizes a group of pull-out tests using %4-in. plain round bars in which the diameters of the concrete cylinders varied from 3 to 12 in. The embedment was 8 in. for all tests except the specimens with 8-in. cylinders which were embedded 944 in. The load-slip curves for these tests are plotted in Fig. 15. In Fig. 16 the values of bond unit-stress for various amounts of slip and for maxi- mum loads have been plotted. The values for the 3-in. cylinders do not show the usual increase of bond resistance after beginning of end slip. This is probably due to the high compressive stress developed in the con- crete; at the maximum load this stress was about 930 lb. per. sq. in. In the 3-in. cylinders slip began at approximately 78% of the maxi- mum load. If we disregard the 3-in. cylinder tests, the bond stresses showed a. decided falling-off for all amounts of slip as the diameter of the cylinder increased. The maximum loads varied quite uniformly from 420 Ib. per sq. in. for the 4 and 6-in. cylinders to 358 Ib. per sq. in. for the 12-in. cylinders. The values for the 12-in. cylinders were about 84% of the average values for the corresponding amount of slip in the 4 and 6-in. cylinders. The difference may be due to variation in relative shrinkage in longitudinal and radial directions. 46 ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 10. Errect OF VARIATION IN DIAMETER OF CONCRETE BLOCK. All bars 34-in. plain rounds. Made from same batches as tests in Table 7. Stresses are given in pounds per square inch. Prssiebes ob oe of Age at Bond Stress at End Maximum Cylinder J ent Nezaber of Test Slip of Hoot - inches meats aoe 0005 in.) 001 in. 3 8 5 71 ‘ 253 | 288 326 4 8 5 71 283 319 413 6 8* 5 71 297 333 426° 8 914° 5 70 287 311 386° 12 8 5 72 245 272 358 * Compare the specimens with 3{-in. bars in Table 8. ° The same tests are given in Table 7. g 8 g Bond Stress -Lh per 5g./p. 100 Age about jo aays Qo Or .02 .O3 Od OF 06 O7 a O09 +0 Sip of Bar-lnches Fic. 15. Loap-s_ip CurvES For 34-1N. PLarn Rounp Bars in CYLINDERS OF DIFFERENT DIAMETERS. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 47 600 500 § § Bond ress -Lb. per 5. In. g ¥-in. Plain Fours, /-2-4 Concrefe, Age about 7odays. o 3 6 9 l2 1s Diameter of Concrefe Olinder-lches Fic. 16. Bonp REsIsTANCE oF 34-IN. Plain Rounp Bars IN CYLINDERS OF DIFFERENT DIAMETERS. b. Effect of Shape of Section and Condition of Surface of Bar. 25. Bond Resistance with Rusted Bars—Fig. 17 gives the load- slip curves for the specimens in Table 11 which have ordinary mill surfaces and rusted surfaces. For the bars having ordinary mill sur- faces, end slip began at 267 lb. per sq. in.; the maximum bond re- sistance was 380 lb. per sq. in. and corresponded to an end slip of 0.01 in. The rusted bars had a heavy coat of firm rust caused by allowing them to remain below the surface in a pile of damp sand for five weeks previous to making the specimens. The tests on rusted bars gave a bond resistance higher than that developed with bars having ordinary mill surfaces. End slip began at 302 lb. per sq. in.; 13% higher than for round bars with ordinary mill surfaces. The maximum bond resistance of the rusted bars was 440 lb. per sq. in.; 16% higher than for ordinary rounds. The load-slip curve shows the maximum bond resistance to come at a somewhat greater slip than in the round bars of ordinary surface; this result is a natural consequence of the rougher surface, which is responsible also for the higher bond resistance devel- oped by the rusted bars. 48 ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 11. ErrectT OF SHAPE OF SECTION AND CONDITION OF SURFACE OF Bar. Embedment 8 in. 1-2-4 hand-mixed concrete from Batches 10, 11, 17, 22, 24, 25, 27, 30, 31 and 36. The average compressive strength of 42 6-in. cubes made from 10 different batches of concrete, tested at about 60 days, was 1850 lb. per sq. in. All stresses are given in pounds per square inch. Bond Stress at : Number H Maximum Size and Kind of Bar ol epee HANDY 8 Bond Tests Resistance daya 0005 in, | 001 in. Plain Round Bars. 1-in., ordinary mill surface............... 5 69 267 302 380 Trims, PUB bed Fa jeciankag days areetaratoncena gets. 5 69 302 331 440 Flat Bars. Axess sicsitiecanand va earned 255 6 69 359 395 459 PAGAN ai sicigmiist age Lxslstarasigninlgen ott 4 81 239 263 293 T-Bars. Ixlx}-in. (area 0.27 5g. in.)............. 5 64 282 295 310 134x1}4x yin. (area 0.48 aq. in.)......... 5 71 227 282 305 2x2x}4-in. (area 1.07 sq. in.)............. 5 64 202 221 242 Polished Round Bars. 1-in., polished........... 5 69 149 san 152 34-in. tool steel, polished : 5 69 137 146 160 %%-in. tool steel, ordinary surface......... 6 71 170 192 255 Bars of Wedging Taper*. Polished 1-in. Rounds. Tapered 0.025 in. per ft................. 4 66 171 250t Tapered 0.07 in. pertt................. 5 69 162 482T Tapered 0.10 0. per ft-.sos 10s cers ceaas 5 69 173 547 Tapered 0.20 in. per ft.................. 5 69 163 633t Average....... iehccetesscpuenaeen san nl oni ann 68 167 Bars of Non-Wedging Taper. Polished 1-in. Rounds Tapered 0.025 in. per ft .. 5 69 164 170 Tapered 0.07 in. per ft 5 69 201 222 Tapered 0.10 in. per ft 5 69 204 210 Tapered 0.20 in. per ft.. 5 69 150 155 Average sccsiacivinns 5 os exsaceoine ay We bya da kaijin 69 180% ll isievesraevenraavs 189 * The concrete blocks were reinforced Fig. 18 for sketch of specimen. ¢ Corresponding to an end slip of 0.1 in. against splitting by means of a 14-in. wire in the form of a spiral, See ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 49 26. Bond Resistance with Flat Bars—The flat bars presented about the same character of surface as the plain rounds. Through an error in making the test pieces, 6 specimens were made with 1 by ¥-in. bars and 4 with 2 by 14-in. bars. Direct comparison of the results of the tests is not as convincing as it might have been with an equal number of specimens of each size. The load-slip curves are plotted in Fig. 17. The 1 by 14-in. bars show a higher bond resist- ance and the 2 by 14-in. bars a lower bond resistance than the corres- ponding plain rounds. 500 s 8 Bond Stress -Lb per 59./n. /-2-4 Concrete LEribedmert ir. Age about 7Odays Nurmber of rests showr thug, (5) 8 Qo WO 20 O70 240 BO 060 070 .O3O .O90 oe Sip of Bar inches Fic. 17. Loap-stip Curves For Pitan Rounp, Firat anp T-Bars. 27. Bond Resistance with T-bars.—Pull-out tests were made on three sizes of T-bars, 1, 114 and 2-in., as shown in Table 11. The bond-slip curves in Fig. 17 show that the bond resistance for the T-bars reaches a maximum at a smaller amount of slip than for the rounds or flats; this is especially true of the larger sizes. The curves for the 144-in. and 2-in. T-bars drop off more sharply than usual immediately after the maximum. The smaller T-bars develop higher bond stresses than the larger ones, but all the values fall below those for the plain rounds. On the other hand the bond resistance per lineal inch of bar is twice as high for the 2-in. T-bars as for the 1-in. plain rounds, and the bond resistance per lineal inch for the 1-in. T-bar is about the same as for the 1-in. round. 50 ILLINOIS ENGINEERING EXPERIMENT STATION 28. Bond Resistance with Polished Round Bars.—A comparison of the bond resistance with polished bars and that with plain bars of the ordinary mill rolling may be expected to show how much of the bond is due to adhesive resistance and how much is due to what has been termed sliding resistance, resulting from irregularities of the surface of ordinary mill steel, and to inequalities in the section or align- ment of the bars usually furnished for reinforced concrete construction. In other words, to what degree does the plain bar partake of the nature of a deformed bar? ‘Two sets of tests were made on polished round bars; see Table 11. The 1-in. polished bars were round cold-rolled steel; they were polished by turning rapidly in a lathe and applying fine emery cloth. The 34-in. tool steel was used as it came from the rolls; these bars had the smooth surface and uniform diameter which are usually found in bars of this kind. It was characteristic of the tests of specimens with polished bars that as soon as an end slip amounting to between 0.0005 and 0.001 in. was developed, the adhesion was suddenly destroyed and the bar pulled out so rapidly that it was generally impossible to weigh the decreasing load on the testing machine. However, readings were obtained in one test, and the load-slip relation has been plotted in the lowest curve of Fig. 18. This curve differs from those of plain round bars in that the maximum load comes at the very beginning of slip and the curve drops off quite rapidly as soon as slip becomes appreciable. The mean values of bond resistance from five tests each on 34 and 1-in. polished bars are 143 Ib. per sq. in. at slip of 0.0005 in. and 156 Ib. per sq. in. at the maximum. The corresponding values for the 1-in. bars of ordinary surface are 267 and 380 Ib. per sq. in. If special weight is given to the bond stresses developed for a small amount of slip, it may be said that the adhesive resistance of clean steel to concrete of this quality is about 150 lb. per sq. in., which amounts to about 55% of the bond resistance of bars of ordinary mill surface at a small amount of slip. The results of the pull-out tests with tool steel bars having the original surface take a mean position between the values for the polished bars and ordinary plain round bars. Tests reported in Bulletin No. 1 (also in Table 25, Bulletin No. 8), University of Illinois Engineering Experiment Station, gave 136 and 147 lb. per sq. in. as the maximum bond resistance of 1-in. cold rolled bars, and 34-in. round tool steel bars, respectively, embedded 6 in. in 1-3-6 concrete, and tested at 60 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 51 days. The pull-out tests on specimens with cold rolled round bars in which the concrete set under pressures of 100 lb. per sq. in. gave values about 10% higher than those in concrete setting without pressure (see Table 22.) The tests on the effect of loads reapplied after failure of bond on specimens in which smooth bars were used are discussed in Art. 60. Double pull-out specimens with cold rolled rounds are dis- cussed in Art. 61. 29. Tests with Tapered Bars—The tests on tapered bars were designed to throw light on the nature of bond resistance. Tapered bars of two distinct types were used. These types will be referred to as bars with wedging taper and bars with non-wedging taper. The form of the wedging-taper bar is shown in Fig. 18. Four different degrees of taper were used for each type. The taper varied from 0.025 to 0.2 in. per foot for each of the two forms. These bars were polished after machining so as to reduce them to a surface condition as nearly uniform as pos- sible. The specimens with bars with wedging taper were reinforced against bursting by means of 6 or 7 turns of 14-in. wire in the form of a spiral. The results of the tests are given in Table 11. The bars with non- wedging taper developed an end slip of less than 0.001 in. and an average maximum bond resistance of 189 lb. per sq. in. before the adhesion was destroyed. Almost immediately the bond resistance fell to nothing. In only a few of the tests did the bar slip as much as 0.001 in. before the maximum resistance was reached. There was no apparent difference in the amount of slip developed before failure in the bars of the different degrees of taper, and there was no systematic relation between bond resistance and amount of taper. These bars gave approx- imately the same maximum bond resistance as the straight polished bars. The tests on bars with wedging taper gave average loads for all] the tests at the beginning of slip nearly the same as the average for the bars with non-wedging taper for the same amount of slip. The phenomena of these tests were quite different from those of other forms of polished bars or bars with mill surface, and exhibited some unusual features. The uniformity of the loads at first end slip is noteworthy. The load-slip curves for the bars with wedging taper are given in Fig. 18. Each curve is a composite of all the tests in a set. For comparison the load-slip curves for a single test on a straight polished bar and for the set of plain rounds with mill surface have been plotted on the same diagram. 52 ILLINOIS ENGINEERING EXPERIMENT STATION It should be noted that the concrete blocks in this group of tests were reinforced against bursting and this allowed very high bond stresses to be developed. Up to the time that slip began, the bond stress de- veloped was much the same as in the tests with the cylindrical polished bars; and it was not materially different from that found in the tests with the bars with non-wedging taper. In all the tests except those with a taper of 0.2 in. per ft., after a slip of about 0.005 in., corre- sponding to a bond stress of about 205 lb. per sq. in., the load dropped off as the bar was being withdrawn. With continued slip (the amount 600 § Test Specimen with tapered bar g Band Srress -LL per 5g.l, g /-@-4 Concrete, 100 Lrbedimert Clr, Age CadhayS Qo O Of OC? OF Of OS 06 OT O83 09 10 ee Mt 46 48 0 Sip of Bar -lriches Fic. 18. Bonp REsisTANCE witH PotisHep Bars oF WEDGING TAPER. depending on the degree of taper), the load finally began to rise. Dur- ing the first stage of the test it appears that the load was carried prin- cipally by adhesive resistance. After the breaking of the adhesion the load on the bar was taken by (a) frictional resistance developed by the compression and (b) the longitudinal component of the compression. The latter was small except with the largest tapers. That the break in the curves is a result of failure in adhesion is shown by the similarity of the stresses up to a slip of about 0.005 in., and by the fact that this bond stress is only a little higher than that found in the tests of pol- ished straight bars and polished bars with non-wedging tapers. These load-slip curves indicate that after adhesion was destroyed the bar ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 53 slipped an amount which gave only a very small increase in diameter of section—less than 0.0005 in.—before the bond resistance again reached the maximum amount taken before adhesion was destroyed. As slip continued, the amount of increase in diameter at a given section of the concrete block, which corresponds to an additional bond resistance of 100 lb. per sq. in., was about 0.0005 in., being nearly the same for the several tapers. This change of diameter accompanied the compressive stress set up in the concrete. If the coefficient of friction between con- crete and steel be taken at 0.25, the calculated normal compressive unit- stress between the concrete and steel will range from two to four times the bond unit-stress; the former figure being for the largest taper. It will be noted that for the larger tapers at the higher loads the load-slip curves change character and round off toward uniform bond resistance. It seems evident that this condition is due to the very high normal compressive stresses in the concrete. The most notable feature of these tests is the sharp line of demarka- tion between the effect of adhesion and the effect of wedging action. The similarity of the behavior of the bars with wedging taper and the twisted square bars is discussed in Art. 41. c. Effect of Condition of Storage. 30. Preliminary.—lIt is a not uncommon belief that the permanency of submerged concrete work reinforced with plain bars may be seriously impaired by the ultimate failure of the bond between the concrete and teel. With a view to obtaining information on the effect of a variety of conditions of storage the tests summarized in Table 12 were made. The tests were generally made at age of about 60 days. In two of the groups the age at test varied from 7 days to about 3 years. All speci- mens, except those placed out-doors, were stored in the Hydraulic Laboratory. Thus the air-stored specimens were in a warm, damp atmosphere. The water-stored specimens were generally tested about 1 to 6 hours after removal from the water. Table 12 also contains the results of the compression tests of 6-in. cubes which were stored under the same conditions as the corresponding pull-out specimens. 31. Batches 3 and 4.—From batches 3 and 4 two nearly parallel groups of tests were made using different cements. Five specimens were made for each condition of storage and all tests were made at about 60 days. These specimens were made on January 12 and 13. Twenty-four hours after making, the forms were removed and the speci- ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 12. EFFEcT oF CONDITION OF STORAGE. 1-2-4 hand-mixed concrete. Embedment 8 in. Specimens were stored indoors unless otherwise noted. Stresses are given in pounds per square inch. Compres- Bond Stress ; sive Size of A at End Maxi- Strength Round t Pe q Number Slip of mum of 6-in. Bar va ee EY ta Storage one Cubes, A ays esist- eens 0.0005 | 0.001 | ance | Avaraee inches | inches Tests Batch 3, Universal Cement. ye 62 5 Damp Aan ngs. ssrscercsacan yews Rapaciesuint e402 556 620 702 2898 4 60 5 Air in laboratory .............00..000000s 313 375 498 1937 4 62 5 Water in laboratory..............-....4-. 493 562 670 3253 A 62 5 Water outdoors............ 0... c eee eee 413 421 523 2558 % 63 5 Outdoors in air................ 0. cece ee 354 389 454 2022 Batch 4*, Chicago AA Cement. a4 61 5 Damip sand sé ciccacinssc eras seeaseeink sae 387 421 538 1913 4 50 5 Air in laboratory ...............00..0000- 246 278 398 1295 a4 62 5 Water in laboratory... .............0.-006 359 405 522 1993 a% 61 5 Water outdoors..............--6-2.0000e 415 447 539 1677 34 62 5 | Outdoorsin air, ............ cece eee 335 366 489 1642 % 62 5 Made outdoors in freezing weather........ 64 67 80 diag Batch 35, Universal Cement. 1 7 3 200 224 301 1 7 3 239 273 392 1 41 3 283 314 491 1 41 3 465 514 607 1 41 3 429 482 606 1 60 3 221 254 416 sce 1 60 3 415 492 601 esa 1 60 3 475 485 601 3040 1 14 mo. 3 332 389 562 1 14 mo. 3 635 731 817 1 26 mo. 8 445 513 668 1 26 mo. 3 624 772 950 * It should be noted that Batch 4 was a much leaner mix than Batch 3 weight are 12.7 and 15.4, respectively. The tests in Table 1 show that these tw strength. ; the percentages of cement by © cements were of about equal ABRAMS—TESTS OF 1-2-4 hand-mixed concrete. TABLE 12—ContInvuEp. EFFECT OF CONDITION OF STORAGE. Embedment 8 in. Specimens were stored indoors unless otherwise noted. Stresses are"given in pounds per square inch. BOND BETWEEN CONCRETE AND STEEL 55 Bond Stress Mempiese Size off| Age aie me Maxi- Strength Round | g¢ Test | Number mo of 6-in. Bar of Storage Bond Cubes inches days Tests eis: Average 0.0005 | 0.001 | tance ge inches | inches Tests Batch 41, Universal Cement. 1 7 2 SAMr ci. sScecsdstirs Yar awlapestoaaoog areca taeda 243 279 381 Ssaivis 1 7 2 Water era sccecas sxsronartngstnencia-ai a sepensuisnt 281 343 450 aetisie 1 7 5 Water 4 days; air 3 days........:. 253 285 380 seat 1 31 2 pane sand 456 528 694 2550 1 31 2 |) AIP SS csscsecers is 309 426 632 2190 1 31 2 Water 383 437 640 2203 1 31 5 Water 14 days; air 17 days........ 417 465 562 2484 1 60 2 Damp sand............. esses eee 395 440 608 2560 1 60 2 i ja sahiaiets SG tweens eae Sea a 381 439 586 2650 1 60 DW Wikktet ic ncccs coannccnte nackte otal 506 600 732 2420 1 60 5 Water 30 days; air 30 days........ 532 598 691 3160 1 14 mo. 2 Damp BBN pn alde sien caramomie da te 821 967 1060 1 14 mo. 2 AIP scans sees Sao wate aes aang 576 668 806 1 14 mo. 5 Water 3 mo.; air 11 mo........... 669 739 848 1 26 mo. 2 Albis hey einsacnnsee ae aaunietag NTRS 531 656 815 1 26 mo. 2 Waterss satndccnsbitsuieiet oe sarees 676 836 984 1 37 mo. 4 WHO i vacuggenrommaci mua’ 750 780 936 Batch 39+, Universal Cement. 84 61 4 Water 50 days; air 11 days........ 495 532 671 2360 14% 61 4 Water 50 days; air 11 days........ _ 430 477 634 + oe8 4 61 4 Water 55 days; air 6 daya........ 476 517 641 2560 14% 61 4 Water 55 days; air 6days........ 431 483 572 eabes 74 61 4 Water 58 days; air 3 days........ 504 554 709 3007 1% 61 4 Water 58 days; air 3days........ 437 498 621 juste % 61 4 Water 61 days..............-.04. 387 559 658 3020 1% 61 4 Water 61 Cay. cccnwaas adios ses 2 318 449 471 a shat % 70 4 Water 60 days; air 10 days........ 586 648 789 3223 1% 70 4 Water 60 days; air 10 days........ 392 451 594 eave 4 79 4 Water 60 days; air 19 days.... .... 561 673 766 2713 1% 79 4 Water 60 days; air 19 days........ 460 536 619 wie t The compressive strength of 6-in. cubes from Batch 39 which were stored in damp sand and tested at abou 60 days was 2190 Ib. per sq. in. 56 ILLINOIS ENGINEERING EXPERIMENT STATION mens stored as indicated in the table. The specimens stored out-doors in air were placed near the north wall of the building, where they were exposed to the weather during the greater part of an unusually severe winter. The specimens marked “made outdoors in freezing weather” were made at about noon with the thermometer at 15° F. The concrete was mixed indoors and carried to the forms in buckets. The specimens were left outdoors till the time of test. The concrete probably froze and thawed several times before finally hardening. When the specimens were broken up after testing, evidence of freezing could be seen in innumerable fossil-like crystal forms which were distributed throughout 600 500 g 8 Band Stress -Lh perSo.tn. 8 of Storage Batch 4 0 0 Oo 0? OF Od -O5 06 .07 08 .09 /0 Slip of Bar-Inches Fic. 19. Loap-stip Curves For DIFFERENT CoNDITIONS OF STORAGE. the concrete. Fig. 19 shows the load-slip curves for the tests from Batch 4; each curve is a composite of the curves for the five tests in a set. An examination of the values for maximum bond resistance given in the table shows that for both groups the damp sand storage gave the highest resistance, with water storage a close second. For the cube tests, these relations are reversed. There is little difference between the values for water storage indoors and outdoors—an average of 11% in favor of the indoor storage. In the case of air storage the outdoor specimens (average of batches 3 and 4) were 5% stronger than those stored in- doors. The average of all water-stored specimens is 23% greater than for the corresponding air-stored specimens. For the cube tests the values are in the same order; the corresponding percentages are 21, 14 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 57 and 42, respectively. The similarity of the results for indoor and out- door storage is surprising when the difference in temperatures is con- sidered. The mean outdoor temperature during the season of storage was 31° F., while the temperature indoors probably seldom fell below 70° F. It seems probable that the loss of water due to evaporation from the specimens indoors in air had more effect in reducing the con- crete strength than the low temperature outdoors. The excess of 11% for the water-stored specimens indoors over those stored in water out- doors may be considered to represent the difference in strength due to the higher indoor temperature. The specimens made outdoors in freezing weather were almost de- void of bond strength. The cubes for this set were accidentally destroyed before the time of test. The curves in Fig. 19 show considerable similarity in the bond- slip relations of the specimens stored differently. This similarity is maintained until the tests were discontinued at a slip of 0.1 in. 32. Batch 85.—In this group a comparison of sand, water and air storage was made for ages varying from 7 days to 26 months. The maximum bond resistances for the different ages for the air-stored and water-stored specimens are plotted in Fig. 20. The values for damp sand and water storage are nearly identical for both ages at which tests were made—41 and 60 days. The bond resistance for water storage is greater than for air storage for all ages at which tests were made; the maximum bond resistance for the water-stored specimens is greater by 30% and 23% at 7 and 41 days, and 77% and 42% at 14 mo. and 26 mo. The 60-day tests show an increase over the 7-day tests of 38% for the air-stored and 58% for the water-stored specimens. The 26-mo. tests show an increase over the 60-day tests of 50% for air storage and 64% for water storage. The specimens tested at 60 days show a slightly lower bond resistance than the specimens tested at 41 days; this, how- ever, is probably an accidental variation. The bond resistance at beginning of end slip shows about the same relation as at the maximum loads. The high bond resistance developed in the older specimens is noteworthy. For the water-stored specimens tested at 26 mo. the bond stress at beginning of slip of the free end of the bar was 624 lb. per sq. in., and the maximum bond resistance was 950 lb. per sq. in. The bond stresses developed by the specimens from Batch 35 are somewhat lower than those from Batch ,3 for the same age 58 ILLINOIS ENGINEERING EXPERIMENT STATION and storage, although the 6-in. cube tests indicate that the concrete was of about the same quality. 33. Batch 41.—The tests from Batch 41 were in part duplicates of those in Batch 35. Four sets of specimens were placed in water during the first half of their storage period and in air for the remainder of the time. In this group the water-stored specimens gave a higher bond resistance at all stages of the tests than the air-stored. At 31 days the sand-stored specimens were stronger than the water-stored, while at 60 days the reverse was true. The highest values for bond on plain bars from the tests reported in this bulletin were found in the 1000 900 eal Le) La | 600 > x 6s soe | § 700 wal? & 7 aa § 600 / = ; | i 500 | N ir 3108 Le 8 ao ee § & § 300 Effect of Storage Batch 35 200 100 o oO e 4 6 é 10 le 18 16 18 20 @P 2h 26 & Age al Test -Monirs Fic. 20. Maximum Bonn RESISTANCE For SPECIMENS STORED IN AIR AND IN WATER. tests of the sand-stored specimens in this series, which were tested at 14 mo.; the maximum bond stress of 1060 lb. per sq. in. from these tests corresponds to a steel stress of 31400 lb. per sq. in. in the 1-in. bar which was embedded only 8 in. The specimens stored 4 days in water and 3 in air gave about the same values as those stored in air for 7 days. Those stored 14 days in water and 17 days in air gave a somewhat lower maximum bond resist- ance than those stored 31 days in air. The specimens stored 30 days in ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 59 water and 30 days in air gave values between the water and air stored specimens tested at 60 days. The specimens stored 3 mo. in water and 11 mo. in air gave values between the sand-stored and the air-stored specimens tested at 14 mo. In general these tests show a progressive increase of bond strength with age for all conditions of storage. The values at the older ages for all conditions of storage are high. 34. Batch 39W—In order to study the effect of storing specimens for a period in water followed by varying periods of air storage tests were made as given under Batch 39 in Table 12. The specimens stored in water 61 days were tested immediately upon removal from the water and were thoroughly wet when the load was applicd. These tests indicate that the drying-out of the water-stored specimens before testing has an influence in increasing the bond resistance, although the evidence is not entirely conclusive. 35. Discussion of Effect of Storage—The tests in Table 12 show about the same bond resistance for damp-sand storage as for water storage for ages up to 60 days; and the high stresses developed by the specimens in Batch 41 stored in damp sand for 14 mo. indicate that this relation may be expected to hold indefinitely unless affected by other agencies. The water-stored specimens gave from 10% to 45% higher bond resistance than the corresponding air-stored specimens, an average of 26% for 13 parallel sets of tests, based on the maximum bond stresses given in Table 12. This difference seems to increase with age; three parallel sets of tests made after 1 year show an average increase in bond resistance of 37% in favor of the water-stored concrete. The compression tests on 6-in. cubes show an average excess of strength of 25% for the water-stored specimens over the air-stored specimens. Since .sufficient water was used in mixing the concrete, it is probable that evaporation of water from the concrete was the cause of the lower bond resistance in the air-stored specimens. The presence of water not only does not injure the bond between the concrete and steel for ages up to 3 years, but it is an important factor in producing conditions which result in high bond resistance. Conditions of stress in concrete under load may modify these results, but it seems likely that the relative bond resistances in service will not be materially different from those found in the above tests with concrete of the same quality and with similar exposure. 60 ILLINOIS ENGINEERING EXPERIMENT STATION d. Bond Tests with Deformed Bars. 36. Preliminary.—The term “deformed bar” is applied to a form of bar with projections on the surface or other frequent irregularities of section and designed and rolled specially for concrete reinforcement. These irregularities of section or projections are expected to increase the bond resistance of the bar. Some hundreds of patents have been issued on bars of this kind, but only a few forms have been used to any extent. Present building codes and engineering specifications gen- erally permit the use of higher bond stresses in the design of structures in which deformed bars are to be used than are permitted in the case of plain bars. In the 1909 series, pull-out tests were made with seven commercial types of deformed bars, which number included most of the types of deformed bar in use at the time the tests were begun. Fig. 21 shows the forms of the bars used, and Table 13 gives the characteristics of these bars. For comparison, tests on specimens with plain round, twisted square, and threaded round bars were included. The size of the bars varied from 1% to 11@ in. The concrete cylinders were 8 in. in diameter with 8-in. embedment. The concrete was 1-2-4, hand-mixed. All specimens were stored in damp sand and tested at the age of about 2 months. In general, the tests were made in sets of five, one specimen of each kind having been made from each of five batches of concrete. A period of about 2 months elapsed between the making of the first and the last specimens. In all the specimens except those with plain round bars the concrete cylinders were reinforced against splitting by means of a wire spiral, since the load-slip relation was desired rather than values for ultimate strength. A summary of the average values from the tests is given in Table 14. The load-slip curves are plotted in Fig. 23 and 24. Deformed bars were used in several other groups of pull-out tests which are not included in the present discussion. Tests with deformed bars in which the concrete blocks were not reinforced against splitting are discussed in Art. 64. 37. Phenomena of Pull-out Tests with Deformed Bars—In the discussion of bond on plain bars it was seen that the adhesion between the concrete and steel was broken by a very small movement of the bar and as movement continued the bond was gradually taken by sliding resistance, which resulted from inequalities of the surface of the bar and from irregularities of its section and alignment. The projections on a ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 61 V) vv itt Hf i , dn00 Ng Oyaaeg ‘ ai AA adadadi FOUND BAR WITH STANDARD THREADS Fic. 21. Derormep Bars Usep 1n Putt-out TEsts. 62 ILLINOIS ENGINEERING EXPERIMENT STATION deformed bar give an exaggerated case of inequality of surface and irregularity of section. The tests described below indicate that the pro- jections do not assist in resisting a force tending to withdraw the bar until a slip of bar has occurred approximating that corresponding to the maximum sliding resistance of plain bars. The sliding resistance of plain bars reaches its maximum yalue at a slip of about 0.01 in. It TABLE 13. Notes oN DEFORMED Bars UseEp IN 1909 PuLt-out Tzsts. The forms of these bars are shown in Fig. 21. Perim- Bar Equivalent Section eter,* Remarks inches Y-in. twisted lug bar° Hexagonal in section,} 1.87 Lugs 2.4 in. apart; 2.7 twists per ft. -54-in. twisted lug bar® equivalent to square} 2.32 | Lugs 2.4 in. apart; 2.0 twists per ft. 34-in. twisted lug bar°® bar of same nominal] 2.78 Lugs 2.4 in. apart; 1.7 twists per ft. 1 -in. twisted lug bar° size. 3.78 Lugs 2.4 in. apart; 1.0 twists per ft. 14-in. cup bar Round bar, equivalent} 1.77 Cups 1__ in. long. 34-in. cup bar to square bar of same] 2.65 Cups 1} in. long. 1 -in. cup bar nominal size. 3.56 Cups 124 in. long. 44-in. Corrugated sq. Type A] Square bar having the] 1.32 Corrugations .06 in. high 0.5 in. apart on all faces. 34-in. Corrugated sq. Type A} _ perimeter shown. 2.50 Corrugations .08 in. high 0.7 in. apart on all faces. 1 -in. Corrugated sq. Type A 3.56 Corrugations .09 in. high 0.9 in. apart on all taces. 14-in. Corrugated sq. Type B} Square bar of same size] 2.00 Corrugations .06 in. high 34 in. apart on all faces. 84-in. Corrugated sq. Type B| Square bar of same size| 3.00 Corrugations .09 in. high 114 in. apart on all faces. 1 -in. Corrugated sq. Type B| Square bar of same size| 4.00 Corrugations .11 in. high 134 in. apart on all faces. fe-in. Corrugated round] Round bar 1.77 Circumferential corrugations 0.05 in. high, . e -in. apart. 1%-in. Corrugated round] Round bar 3.34 Circumferential corrugations 0.09 in. high, Type C 1.6 in. apart. 34-n Thacher 34-in. round bar 2.35 | Plain round bar flattened at intervals. 4-in. sq. twisted ¥ in. square 2.00 Cold twisted; 2 twists per lineal foot. 1 -in. sq. twisted 1 in. square 4.00 Cold twisted; 1 twist per lineal foot. 1-in. round bar, threaded Round, diameter 0.89in.| 2.80 Standard V-shaped threads, 8 per inch. ° The straight lug bars were like those given in the table, except for the twisting. * The perimetera given are those of bars of the same weight and having the form of section indicated in the second column. will be shown in Art. 51, in the discussion of tests on bars anchored by means of nuts and washers, that with the large bearing area present a distinct movement of the bar was necessary to bring this anchorage into action. This makes it clear that the adhesive resistance must be de- troyed and sliding resistance largely overcome and that the concrete ahead of the projections must undergo an appreciable compression before the projections become effective. The action of anchored and deformed ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 63 bars during the first stage of the tests does not differ much from that of plain bars. This conclusion might have been reached from our knowledge of the behavior of elastic bodies, but it is one which gen- erally has been overlooked in discussions of the action of deformed bars. After the adhesion is overcome and the sliding resistance of the smoother portions of the bar reaches its maximum value, the projec- tions become effective and we may recognize a third stage in a pull-out test of a deformed bar in which the bond stress arises principally from the bearing stress between the projections and the concrete ahead. As slip continues, a larger and larger portion of the stress is taken in direct bearing. This bearing stress is opposed by a shearing stress over an area of concrete enveloping the projections. The exact stage of the test at which the projections become effective in taking stress and the amount of stress which finally may be taken in this way, depend, of course, on the design of the bar. The influence of the form of the bar and the secondary stresses developed in the concrete by deformed bars of different types are discussed in Art. 42. Due to the elongation of the bar and to the yielding of the con- crete ahead of the projections it is evident that the three stages men- tioned above (adhesion, sliding resistance and bearing resistance) may be co-existent over a comparatively short embedded length of a deformed bar. In other words, the transition from the condition in which prac- tically all the bond stress is taken by adhesion, only a little by sliding resistance and none by bearing, to that in which none is taken by adhesion, only a little by sliding resistance and practically all by bear- ing, is so gradual that we may not expect to find a sharp line of demarka- tion between them, but that the load-slip relation for deformed bars will be a continuous curve, as long as the concrete is intact. Fig. 22 gives a typical load-slip curve for deformed bars. This curve is a composite from 55 tests, including all the tests given in Table 14 on bars of 34 in. and larger sizes, except the plain round, twisted square and threaded bars. The twisted square and threaded bars are not considered in the present discussion of deformed bars. The bond stresses for various amounts of end slip are given as percentages of the average bond stress which was developed at an end slip of 0.1 in. Since the value used for the highest bond resistance considered is purely arbitrary, it will be seen that the percentages given in Fig. 22 are only relative. However, the curve plotted in this form is useful in showing the relation of bond resistance to end slip of bar. It would have mae little difference in the relative position of the 64 ILLINOIS ENGINEERING EXPERIMENT STATION curve if the average bond resistance corresponding to an end slip of, say, 0.05 in. or 0.2 in. had been used as 100%. End slip of bar began at about one-third the stress corresponding to an end slip of 0.1 in. After slipping began, the increase in bond resistance was nearly pro- portional to the amount of slip up to an end slip of about 0.01 in. After an end slip of 0.05 in. the bond resistance increased very slowly with further withdrawal of the bar. It should be borne in mind that 100 La rs & 40 7) a Ss Ss J S 70 7 $s A B& 60 s y 3 [ a 50 wH ye & 4o f © 30 oe Composite Logd-S/ip Curve 8 tor Deformed Bars; Confeaiting & 20 ol! the Tests In Kathe 14. with = ——) 8 Bars Z-in, and larger, excepr & 0 twisted Square , Plain ound, and Threaded Bars. 7 eel O Of 02 OF Of OF 06 O07 O08 09 O} Slip of Bar -/nches. Fic. 22. Composite Loap-stip Curve From PuLL-ouT TESTS WITH DeForMED Bars. spiral reinforcement was used to prevent splitting of the concrete blocks. It will be seen later that the presence of this reinforcement had a marked influence in increasing the bond resistance of the deformed bars even at small amounts of end slip. A reference to Fig. 23 and 24, where the individual load-slip curves. for these tests are plotted, will show that the form of the curves for deformed bars and their position on the diagram are not materially different from the curves for the plain bars up to an end slip of about 0.01 in., corresponding to the maximum bond resistance of the plain bars. This makes it apparent that up to this point the bond resistance of deformed bars is principally due to the same causes as in the plain bar tests, and that the projections have not appreciably come into action at this stage of the tests. All further increase in bond resistance is due ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 65 entirely to the presence of the projections. The form of the load-slip curve after slip became general varied widely for bars of different types. The form assumed by any curve probably depends on the shape, size and spacing of the projections. Some types of deformed bar developed very high bond stresses and it is apparent that the bearing stress between the concrete and the adjacent projections was abnormally high. Some notion of the impor- tance of these stresses and their bearing on the proper design of a deformed bar which will best resist slipping through the concrete is given in Art. 42. 38. Basis of Comparison of Bond Resistance of Deformed Bars.— In computing the unit stresses given in Table 14 and for other tests with deformed bars, the perimeter used was that of a bar having the same sectional shape and the same weight per unit of length as the deformed bar under consideration. The perimeters given in Table 13 for these bars were determined from the weights of the bars in this way. In order to secure a uniform basis for comparison the bend stresses corresponding to given amounts of end slip will be used. Most of the deformed bers showed an increasing bond resistance even after a large amount of slip had occurred. The highest bond resistance reported for deformed bars was that corresponding to an end slip of 0.1 in, unless the maximum load came at a smaller amount of slip. A slip of 0.1 in. is very much larger than would be available or permissible in reinforced concrete construction. It will be seen in the discussion of bond failures in beams reinforced with either plain or deformed bars that the amount of slip at the free end of the bar which may be developed before the maximum bond resistance in the beam is reached is much less than in the pull-out tests. It is evident that if we are to secure a proper criterion of the bond value of a deformed bar from the pull-out tests we must use a load which corresponds to a slip very much smaller than that at the maximum bond stress as defined above. Certain beam tests indicate that if the load which produces a slip of about 0.001 in. at the free end of the bar is repeated or maintained constant for a considerable period the beam will ultimately fail in bond at this load. This conclusion is based on tests of beams in which the longitudinal steel was not highly stressed—see Art. 90. In the case of plain round bars the pull-out tests show that the bond resistance at an end slip of 0.001 in. is about 75% of the 66 ILLINOIS ENGINEERING EXPERIMENT STATION maximum. Professor J. L. Van Ornum, Trans. Am. Soc. C. E., Vol. LVIII, 1907, p. 294, and Professor M. O. Withey, Bulletin No. 321, University of Wisconsin, have found that with plain bars a sufficient number of repetitions of a load 60 to 80% of what would be the maxi- mum bond resistance under a load applied progressively to failure will cause bond failure in pull-out tests or in beam tests. It will be seen also in the discussion of the beam tests that begin- ning of slip at the free end of the bar (slips amounting to 0.0002 in. or less) is almost simultaneous with the appearance of cracks in the outer region of the beam. This is an important reason for using for com- parison bond stresses corresponding to small amounts of end slip. These considerations show that it is the bar which longest resists beginning of slip that should be rated highest in comparison. In the following discussion of deformed bars, the bond resistance at a slip of 0.001 in. will be used as the principal basis of comparison of the bond resistance of bars of different forms. In the summary of the tests, Table 14, the relation of the bond resistance at an end slip of 0.001 in. has been given for each set of tests in terms of the resistance of plain round bars at the same slip. Similar ratios are given for the bond stresses at an end slip of 0.01 in. Mean values for 1 and 114-in. plain bars have been used as a basis of comparison in determining the ratios given in the table. The high values found in the pull-out tests are much larger than would be obtained without the restraint of the spiral reinforcement. While such values may be available in the case of a bar simply anchored, it is evident that they are not available in beam or similar construction. 39. Discussion of Deformed Bar Tests—The plain round bars used in this group of tests were of high-carbon steel with ordinary mill surface. The values of bond resistance found for the plain rounds are about the same as found in other groups of tests on plain bars. The average values for the two sizes of bar are 306 lb. per sq. in. for an end slip of 0.001 in. and 405 Ib. per sq. in. for the maximum bond resistance. The blocks with plain round bars were not reinforced against bursting. Using the above values from the plain bar tests as a basis, we may make some interesting comparisons. At an end slip of 0.001 in. the 12 sets of deformed bars of 34-in. and larger sizes in Table 14 (the 55 tests which were used in plotting the curve in Fig. 22) developed an average bond resistance of 318 lb. per sq. in., or an average of 4% more than the plain rounds at the same end slip. At this stage of the tests, two ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 67 sets of deformed bars gave practically the same bond resistance, five sets gave lower values, and five sets higher values than the plain bars. At an end slip of 0.01 in., corresponding to the maximum bond resist- ance of plain bars, the average bond resistance of the 12 sets of deformed bars was 445 Ib. per sq. in., or 10% higher than the plain bars. At this stage, two sets gave about the same resistance, two sets gave lower values and eight sets gave higher values than the plain bars. The highest value (2 tests) was about 45% in excess of the plain bars; the average of the three next highest sets was about 23% in excess of the plain bars. In the following discussion, little will be said about the tests with deformed bars smaller than 34 in., but the values may be seen in Table 14. The tests with threaded round and twisted square bars are dis- cussed in Art. 40 and 41. Tests were made on three types of corrugated bars—types A, B, and C. Types A and B are square bars, type C is designated as round but has an oval contour. These bars were of high-carbon steel. At a slip of 0.001 in. with bars of type A (formerly known as corru- gated square, old style) the 34-in. size gave 89% of the bond resistance of plain bars. The 1-in. bars gave 123%. The individual tests for the 34-in. bars show a wide variation; the values ranged from 183 to 390 lb. per sq. in. at a slip of 0.001 in., with an average of 273 lb. per sq. in. The load-slip curves follow closely the composite curve for deformed bars given in Fig. 22. Type B (formerly known as corrugated square, new style) gave a percentage of 101 for both the 34 and the 1-in. bars; the value for bond stress at slip of 0.001 in. averaged 308 lb. per sq. in. for each of these sizes. Type C (commonly known as corrugated round) gave a bond resistance of 256 lb. per sq. in., for the 114-in. bar at an end slip of 0.001 in., a value below all the other deformed bars. However, in other groups of pull-out tests and in the beam tests in which this bar has been used the comparison is more favorable. In the tests on effect of length of embedment, Table 9, the 114-in. corrugated rounds give a bond resistance of 291 lb. per sq. in. at a slip of 0.001 in., as the average of four lengths of embedment, ranging from 4 to 24 in., while the 114-in. plain rounds for the same lengths of embedment gave 302 Ib. per sq. in. The bars of type A give the highest bond resistance both at beginning of slip and at maximum load found in the tests on corrugated bars. It is shown in Art. 42 that these differences may readily be accounted for by a consideration of the details of the design of the bars. The relative values for the larger sizes of corrugated bars may be seen by reference to Figs. 23, 24 and 25. ILLINOIS ENGINEERING EXPERIMENT STATION 68 “By [9'0 Jo dys pus ue ye pus “ut 790'9 Jo dys pus us 48 spenor med “UI-J{] pus | Jo soumstsar puog aBerOAN aM Jo 4090 red UT possardxa seq pouToyap Jo soueysIsa1 puog ay “dis Jo FUMOUTE Ja[[BUIS B 4B PaTOvE SBA LINUNTXBUL 8 ssa[UN “TT [°Q Jo dijs pus Us 04 spucdselz00 perepismod asa148 puog YeayFIY OY} Seq PotTOFEp OY} 10] y 291 002 ShL pe. SL 629 849 989 319 oes 9L OR | Orta nee punos papeazqy “UI-] $8 88 SOF sce one che Ize 062 892 63 LL Fie | |e teerseer ‘T 1 Tarenbs pa4siai9 “UI-] 26 Tt 4 yee 188 Ue 298 ose Tee PEE 9b > ||reeeas SPREE erenbs paystasg “UI-34 201 “16 612 9g OLF slp ole 608 61 8h OL Gi | PSE ees eq seqoegL, “W-% 28 +8 528 429 iy pee 662 12 992 98% ¥L g “1 1***(9 ed4q) punos paqesnss09 “a-§4 7 Ost. S21 8F01 918 629 92g 6rF 868 LL 9e8 92 $ “""*(Q ad44) punoz pozesnz09 “ui-3t eer TOT 16L 1574 06¢ 164 827 ggg 808 18% 89 g **(q ed49) arenbs poyednzoo “ur-] FIL 101 OTL 269 6s 29oF 10 He 808 612 69 $ (q ed4q) orenbs poze8nss09 “uI-Fg Ov Tel ¥18 FOL 119 89¢ S64 oer T0F Ole el ¥ *(G ed44) arenbe pozesnss0o “u1-5¢ SFT S21 9z0t ore eZ 88g 66h 1% LL yee &L z nena (y ed44) oxenbs pozeZnt09 “a-] SIT 68 198 028 119 697 198 org £12 oF 0 9 “r+'*(y adAq) oxenbs pozesn109 “w1-% T9t oe £06 ae 261 $99 ore oF 868 ose 9L g (y 0d44) exenbs pozesnisoo “mi-34 66 +6 FOL ggg oH 807 743 9z8 882 193 L g ***(p2qets4) Bny “uty 16 ¥8 998 589 SLF 69¢ 8Te 813 892 682 TL $ “*(paqsray) Bny “ar-% er 96 THOT oS. ggg eg Sty oes 16z £92 29 & 1777 1(paqsus9) Bny -U1-94 TAT ost Bear ZL01 | €88 69 98g OL 66 998 TZ ¢ **(paqstang) Bny “wI-$4 I2t Ian +9 z9¢ £09 06F cP 80% Le vee &L |S RAR CIR tae (qqsrexy8) Bny “ur-T Il 80T 092 149 Tag ao STF gge ee 908 69 $ “ott Qyseays) Bny “ur-¥g 2st SFI PLL oll 0L 989 86 Ig oop 90F 9 Gel pry rreae ee (aysteHs) any “UT-$¢ P21 901 F801 916 oe9 50S 9% cog See 062 €L ¢ a TT sil 9FIT 986 £89 89F SIF 198 LE 1% OL $ **dno “UI-¥% ver 6eT 168 828 189 oes 164 Soy Sey 908 9 g dno “ut-§{ oot oot SIF 118 06g SIF 10 ese re 222 OL 6 *“punos urerd “Ul-24] ge £08 ose ose 298 828 108 892 &L S punog ured “ut-T +Pareprewoy | = ¢0" 20° To" $00" 200" Too" | $000" dep $100 | FU! 1000 S824 450, 8389], jo dig pug | jo dyg pug puog (seqout) i jo Jeg Jo pury pue azig 48 $90 Jeg | 4e 4ua0 Jeg qsoqaty jo dyg pug 48 ssezjg puog 4B ady Joqunyy “ar g quompequg, (6061) SHVG GaWUOTTC] HLIM SLISH], LNO-1T1ng Fl WTAVL -ae10j8 puws due ‘qout exenbs sad spunod wl waald aie sasseyg ‘at “bs sod ‘q] OGL sv ‘SBP CQ Jo OF 4B Saqnd “UI-g 27 JO IZUAI8 OAIssaIdmIOD BZBI9Ae ONY, ‘jeuds © Jo UOJ OT} Ul OIA “OI-}¢ Jo suiny 4 JO g JO suveM Aq Zuremmq yeuIEse padloyulel alom sIBq PUNO Ulefd yyIM eso} Ydeoxe sueuMoads [TV “OE PUG LZ ‘FZ “LT ‘OT Saye WoIy JoINN0N pexTUI-pueY F-Z-T ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 69 The general characteristics of the cup bar are somewhat similar to the corrugated rounds. These bars are round in section and have a sectional area equal to that of a square bar of the same nominal size. These bars were of high-carbon steel. At an end slip of 0.001 in. the 34-in. cup bars show an efficiency of 118% and the 1-in. bars 106%, as compared with the plain rounds. The bond resistance developed by the cup bar at an end slip of 0.1 in. was the highest developed by any of the larger sizes of deformed bars; it averaged about 1100 lb. per sq. in. for the larger bars. However, it should be remembered that the con- crete blocks were reinforced against bursting. Tests were made on two different types of lug bar—straight and twisted. They were of high-carbon steel. The amount of twist and the spacing of the lugs are shown in Table 13. The 34-in. straight lug bars show an efficiency of 108% and the 1-in. bars 121% as compared with the plain round bars at an end slip of 0.001 in. The %-in. twisted lug bars show an efficiency of 95%, the 34-in. bars 84%, and the 1-in. bars 94%. Considering only the 34 and 1-in. bars, the values of bond resistance are from 10% to 30% higher for the straight bars than for the twisted bars. This is true for all stages of the tests except at the highest loads considered. It will be seen in the discussion of tests on square twisted bars (Art. 41) that this is a characteristic phenomenon of pull-out tests with twisted bars. It is noteworthy that at an end slip of 0.001 in. the straight lug bars had a bond resistance about 15% greater and the twisted lug bars about 10% less than plain rounds. The Thacher bar consists of a round mild steel bar which is flat- tened at intervals. Only 34-in. bars of this type were tested. This bar gave a bond resistance of 279 lb. per sq. in. at a slip of 0.001 in., 91% of the bond resistance of plain round bars at the same slip. The average - load-slip curve for this bar follows almost exactly the composite curve for the larger sizes of deformed bars as given in Fig. 22. In Fig. 25 the significant stresses from the pull-out tests on de- formed bars have been plotted in a way that shows their relative values. The heavy solid line shows to scale the bond stress at an end slip of 0.001 in., the right end of the broken line indicates the bond stress at an end slip of 0.01 in., and the open line corresponds to the bond stress at an end slip of 0.1 in., which is the highest bond stress which: has been considered in the discussion of the deformed-bar tests. The bars have been arranged in the figure in the order of the stress developed at an end slip of 0.001 in. For sake of comparison the tests on threaded bars, 70 ILLINOIS ENGINEERING EXPERIMENT STATION 4200 1100 1000 Bond Stress- Lh per 5g. ltr. 8 8 8 8 300 £00 Shown ro en- larged scale In following sgure 100 o Ol 02 /-2-d Carrere, Linbedinert in, Average age lédays, Bars stnaller than %q-in, anitted ‘ 06 07 0B Sp of Bar - Inches 09 0 Fic, 23. Loap-sLip Curves rrom PuLt-out Tests with Drrormep Bars. Bond Stress -L4 per 5G. lr ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 71 JOO | Y r | Lin, “gf erprged » get / , y wa uJ YO J 450 g PAA % 8 250 200 o OOP 004 006 .OOF O10 Oe? Old O16 O17 0 Nip of Bar-liches Fic. 24. Loap-stip Curves FroM Putt-out Tests with DerorMep Bars. (ENLARGEMENT OF A PorTIoNn oF Fie. 23.) 72 ILLINOIS ENGINEERING EXPERIMENT STATION twisted square and plain rounds made from the same batches have been included. The average values for the 1 and 114-in. plain rounds have been used. With this arrangement it is of interest to note that the plain bars occupy about a mean position, with seven commercial bars above and six below. oO 300 400 600 700 800 900 1000 100 [£00 Round |hreaded) -in, Con Square (A} “in, Lug (straight } Zin. Cup &-in Lug (rraig} /-in. Cup /-lta Coe Square (8) $11. Cor Square (8) and! ligin Pith Round “in, Lug [twisted] %-ltn Thacher Zi. Cor. Square (A) ln, Twisted Square Zin, Lug (twisted) Car Fourd 0 100 200 300 400 $500 600 700 800 900 1000 11001200 Bond Stress -Lh per 5e In. Fic. 25. RELATIVE VALuges oF Bonp RESISTANCE OF DEFORMED Bars. The heavy solid lines show to scale the bond stresses developed in the pull-out tests at an end slip of 0.001 in.; the right ends of the broken lines indicate the bond stresses at an end slip of 0.01 in.; the open lines indicate the bond stresses at an end slip of 0.10 in. 40. Threaded Bars.—In order to study the bond resistance of bars in which the projections were more numerous than in any of the avail- able commercial bars, tests were made on 1-in. round bars on which a standard thread had been cut for the entire embedded length. The relation of this bar to the other deformed bars used is shown in Table 13. The results of 10 tests are given in Table 14 and the load-slip curves ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 73 in Fig. 23. This bar offers a much higher resistance to beginning of slip than any other form of bar used in these tests. The bond stress at an end slip of 0.001 in. varied from 500 to 742 lb. per sq. in.; average 612 lb. per sq. in. It will be seen that the lowest value for this bar is nearly as great as the highest for any of the commercial deformed bars of about the same size. This bar gave twice as much bond resistance as the plain round bars at a slip of 0.001 in. The maximum bond resistance came at a slip of about 0.1 in. and averaged 745 lb. per sq. in. The tests indicated that failure was preduced by piecemeal shear- ing of the concrete surrounding the threads. The pull-out tests on threaded bars made with the 1912 beams (see Table 34) gave somewhat lower values than those above, but the 1912 specimens were not reinforced against bursting. 41. Twisted Square Bars.—Pull-out tests were made on 1% and l1-in. cold twisted square bars. The characteristics of the bars may be found in Table 13. The load-slip curves for these tests have been included with the deformed bars in Fig. 23 and 24. Load-slip curves for the individual tests and for the plain rounds made from the same concrete are shown in Fig. 26. A mean curve has been drawn for each size of twisted square bars. The tests exhibit some peculiar phenomena. The load-slip curves are not unlike those for plain bars up to an end slip of about 0.01 in., except that they are lower on the scale. After a slip of about 0.01 in. has been reached there is a decided drop in the curves. This drop is found in all the curves in varying degree. It is more pronounced in the 14-in. than in the 1-in. bars. The mean curve for the 1-in. bars shows only a slight depression after a slip of 0.01 in., but the slip is increased to about 0.05 in. before there is any further increase in the bond resistance. From this point the curve continues to rise at an increasing rate. In the 14-in. bars the mean curve drops from about 373 lb. per sq. in. at a slip of 0.01 in. to 330 lb. per sq. in. at a slip of 0.04 in. and the end slip reaches about 0.10 in. before the bond resistance exceeds that corresponding to a slip of 0.01 in. After this point the course of the curve is similar to that of the 1-in. bars. The load required to pull out these bars would probably depend on the amount of restraint against the bursting of the concrete block. The maximum loads considered in the table were (as in the case of the deformed bars) those corresponding to an end slip of 0.1 in., although the load-slip curves show that a maximum point was reached at an end slip of about 0.01 in. Some of these bars were pulled out as much ag 74 ILLINOIS ENGINEERING EXPERIMENT STATION 2 or 3 in. before reaching their highest resistance. The apparent bond stress at these large amounts of slip was very high, but of course such stresses and slips can not be developed in a structure, and are entirely meaningless under a rational interpretation of the tests. However, values obtained in this way have been commonly reported as the bond resistance of twisted square bars, and such tests have been generally used as the basis for the bond resistance of such bars. The load-slip curves for twisted square bars are quite similar to those for polished bars with wedging taper given in Fig. 18. A little consideration will show the reason for this similarity and will also reveal important differences between the two cases. A longitudinal sec- tion of a twisted square bar presents an outline which consists of a series of equal arcs as shown in Fig. 26. The length of the radius and of the arcs will depend upon the amount of twist. The twisted bar is essentially a combination of a wedging taper and a non-wedging taper. As the bar begins to slip through the concrete the wedging tapers are drawn more firmly against the concrete, while at the same time the non-wedging tapers are separated from the concrete with which they were originally in contact. This separation of about one-half of the surface of the bar from its original contact is one important difference between the action of the twisted bar and that of the bar with wedging taper. Of course the separation of the surface of the twisted bars does not occur at once, on account of the curvature of the surfaces, and it may not be expected to occur as suddenly as with the polished bars on account of the difference in the texture of the surfaces. The load-slip curves for the twisted bars seem to indicate that this separation becomes pronounced at an end slip of about 0.01 in., which is about the slip at which plain bars have been found to reach their maximum bond resist- ance. The gradual drop in the curves suggests that the loss in bond resistance resulting from the separation of the non-wedging tapers and the continued sliding of the flatter portions of the longitudinal section have more influence in reducing the average bond resistance than the increased bearing of the wedging tapers against the concrete has in increasing the bond resistance. Continued slip of the bar finally allows the bond resistance due to the wedging action to rise and all further resistance must be due to this cause. It is evident that this action must produce a large twisting moment in the bar as well as a high splitting stress in the block. In the tests in which the bars were withdrawn, say ¥4 in. or more, this tendency to untwist was quite pronounced. This Bond Srress- Lb per 5q./n, 700 500 8 8 8 500 100 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 1-2-4 Concrere, Limbedmert 6 in, Age 76 days Llochks relatorced with £-(01, mire Spiral, Axial Section of a twisted Square Bar Qo Oo O be 04 06 a Jo dé 4 46 AO Sip of Bar -lnches Fic. 26. Loap-siir Curves FoR TWISTED SQUARE Bars. £0 79 76 ILLINOIS ENGINEERING EXPERIMENT STATION was shown in two ways: first, the cavity left in the top of the concrete block by the withdrawal of the bar presented the same spiral surfaces which were formed by the bar in its original position, the bar having to untwist somewhat during the movement; second, upon releasing the load at the completion of the test the blocks gave a sudden lurch in the direction of the twist in the bar. It should be borne in mind that this tendency to untwist was not pronounced until the bar had been with- drawn 1% in. or more, and that it had no effect at the loads given in the table, and may not be expected to have any influence at working stresses. On the basis used in comparing deformed bars (the bond resistance at a slip of 0.001 in., as compared with the resistance of plain round bars) the rating of the 14-in. twisted square bars is 111% and the 1-in. bars 88%. Ata slip of 0.01 in., which represents the first maximum in the load-slip curves for the twisted square bars, the bond stress for the 1-in. bars is 343 lb. per sq. in., or 80% of the value for plain rounds at the same slip. The bond stress on these bars at a slip of 0.1 in. is 93% of the maximum bond resistance for plain rounds. It has been frequently stated that cold twisting is effective in rais- ing the yield point of the bar by over-stressing a portion of the metal, and at the same time it furnishes a very severe test on the quality of the steel itself. However, it has been shown by tests that the elastic limit has been raised on only a portion of the section (the outside) and that for stresses above the original yield point the modulus of elasticity of the whole section is considerably smaller than the normal walue for steel within the elastic limit. In other words, for stresses above the original yield point the metal in the interior of the section will be stressed beyond its elastic limit, and the rate of change in tensile deformations in the bar as a whole will be larger than at the lower stresses. The tests here recorded show conclusively that the bond resistance of twisted square bars is inferior in characteristics to that of plain round bars of similar surface, and that these bars have little or no advantage in bond resistance within limits of slip which would be useful in struc- tures. It seems strange that the twisted bar has gained such a wide popularity as a reinforcing material. 42. Hlements of the Bond Resistance of Deformed Bars.—In Att. 3% and 39 the general phenomena of the tests of deformed bars were described. It seems desirable now to consider the secondary stresses which are developed in the concrete by deformed bars of the usual kind ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 77 and to learn the causes of the variations in the bond resistance of bars of different types. Such a study will indicate certain limitations of the deformed bar and furnish suggestions for the rational design of a bar which may be expected to give the best results. We have seen that the projections do not become effective in taking bond stress until the bar has slipped an appreciable amount, and that during the later stages of the tests bond resistance is due principally to the bearing stress developed between the projections and the concrete in front of them. The pull-out tests showed that high bursting stresses were developed in the concrete blocks by the components of this bearing stress normal to the axis of the bar. This bursting action would be very destructive in a smaller mass of concrete which is under no restraint. This makes it apparent that if the bearing area of the projections is to be most effective in taking bond stress and least destructive in causing bursting, the planes of the bearing faces of the projections should be as nearly as possible at 90° to the longitudinal axis of the bar. The low bond resistance of the twisted square bar and certain deformed bars is due to the small angle which the so-called bearing faces make with the axis of the bar. In many of these tests the entire excess of bond resistance over that of plain bars was due to the restraint offered by the concrete block. Having determined that the bearing faces should be as nearly per- pendicular to thg axis of the bar as practicable, the projections should be of such height and number that the proper relation between bearing stress on the concrete and shearing stress over the enveloping surface of concrete will be preserved. The ultimate bearing strength of con- crete has not been studied, but these deformed-bar tests and the tests on anchored bars described in Art. 51 indicate that small areas which are restrained by a larger mass of concrete may be expected to carry a very high stress in bearing. These tests show that the concrete momentarily supported a bearing stress of from 8000 to 14 000 Ib. per sq. in. Under these conditions the concrete was probably reduced to a powder as it yielded slowly under the stress. The shearing value of concrete under these conditions has not been determined, but these deformed-bar tests show that it is higher than has commonly been assumed, and certain tests reported in Illinois Engineering Experiment Station Bulletin No. 8 indicate that it is nearly as great as the compressive strength. It may be noted that the values for the highest bond resistance given in Tables 14 and 15 were not the & 78 ILLINOIS ENGINEERING EXPERIMENT STATION ultimate values of the bond resistance and that the shearing stresses finally developed over the surrounding concrete were larger than the values for bond stress which are given in the tables. However, the tests with threaded bars were the only ones in which failure occurred by shear in the concrete. In the threaded bars the bearing area was 80 large that the bearing stress was not great enough to produce an appreciable deformation in the concrete ahead of the threads and the elongation of the bar caused a piecemeal shearing of the concrete at an apparent stress much lower than was developed in the other tests. These considerations indicate that while high values of bearing and shearing stresses may be developed in such tests, the shearing resistance of the concrete will probably prove to be the limiting factor. The action of the threaded bars suggests that it is not desirable to keep the bearing stress too low in comparison with the shearing stress, since a slight yielding of the concrete due to a higher bearing stress will give a more uniform distribution of shearing stress and thus reduce the tendency to fail by shear. Practically all experiments in reinforced concrete in which the ultimate shearing stress was developed have emphasized the desirability of avoiding failures of this kind. In order to study the bearing stresses developed in some of the deformed-bar tests, the values have been worked out for certain bars as shown in Table 15. The corrugated bars have been used since with the different types of bar they are instructive in making comparisons of bearing pressures, sliding resistance, and shear. For obvious reasons the lug bars and some other types could not be used in such a com- parison. It was seen above that, in general, the projections have little effect previous to an end slip of 0.01 in., and at a slip of 0.1 in. they may be considered to take all the bond stress except that taken by sliding resistance. If we consider the sliding resistance of a deformed bar to be the same as for a plain bar at the same slip, we may calculate what portion of the total bond stress is being carried by the projections at this stage of the test. It is appreciated that this method may be subject to error, since the secondary stresses in the concrete may be expected to cause an increased sliding resistance in the deformed bar, but on the other hand, this increase is probably largely counteracted by the bursting stresses developed in the concrete block. The sliding resist- ance of the plain bars in this series at an end slip of 0.1 in. was 250 Ib. per sq. in. If we use 90% of this value, on account of the area lost ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 79 by slip of bar, we may say that about 225 lb. per sq. in. of the highest bond stress of deformed bars was due to sliding resistance. The column in Table 15 headed “Bond Stress Carried by Projections at the Maxi- mum Load Considered” shows the values obtained by subtracting 225 lb. per sq. in. from the bond stresses at an end slip of 0.1 in. for the TABLE 15. BEARING SrrREsses DEVELOPED BY DEFORMED Bars. Stresses are given in pounds per square inch. cI a a As BSh Gl Sa Gilets toad 3 2 2843 Bond Stress at Pac a 38 # z gua 3 3 3 B23 End Slip of = [O® gy S| Fy S [ys eg e a & & BBs aan =| BS S\suge a 3 Ss 8. $24 2 a a las s. 2, | Pee Bae 2 ag a 3" er BE] sd] 28 ge | S82 | oorin | o1i EaeE| wees [Sse8 8 a6 ac ad 3 3 o 01 in. .1 in. 3 pag 4 aor aeens a & a 3 a= ie Bones a= #3 eee Sem Corrugated Square (Type A). 3% .37 .080 .700 10.7 469 861 636 5850 940 1 :70 105 875 11.4 588 1025 800 7000 880 1% 1.07 1125 1.185 10.04 || capesertses foes etecanny eteadleeteie ce ie Ulead cataneek 1000 Corrugated Square (Type B). 4 -56 .078 1.10 6.0 462 710 485 8100 1660 1 1.00 | 112 1.50 6.5 497 797 572 8800 1540 1% 1.56 114 1.75 BUD: Voazice enance [edies seca) Geiser wcieccd| ne oe Sanus 1820 Corrugated Round (Type C). 4 44 082 1.30 6:5. \eadeway vesilinanas estew liens va gare las es nea 1820 1 .78 :050 1.50 DID |encesescte saznl tact we veal see 4g wna ae er aeoestes 4550 1% 99 .087 1.60 4.2 334 824 600 14000 2400 Round Bar with Standard V-shaped Threads. 1 55 082 «125 70.0 679 745 520 740 145 bars shown. The bearing area of the projections per unit of area of the entire surface of the bar is shown in the fifth column of the table. The areas given are the areas of the normal projections of the bearing faces, without any reference to the angle at which they are placed. It will be seen that there is a wide variation in the ratio of the bearing area to the area of the surface of the bar. This value is about 11% in 80 ILLINOIS ENGINEERING EXPERIMENT STATION the type A bars, 6% in the type B bars and 2 to 5% in the type C bars; for the threaded bars the ratio is 70%. Considering that these bearing areas take all the stress in excess of that carried by adhesive resistance, the bearing stresses given in the table were computed. The bearing stresses for the corrugated bars, computed on this basis, vary from 5800 to 14.000 lb. per sq. in. The bearing stresses are seen to be inversely proportional to the bond stresses developed by these bars at an end slip of 0.01 in., when the projections were just beginning to take effect. These considerations show that the ratio of the normal projections of the bearing areas to the area of the surface of the bar is the proper criterion for judging the bond resistance of a deformed bar. The bear- ing stresses developed at this stage of the tests also show the absurdity of seriously considering the values that are usually reported as the bond resistance of many such bars. The values in Table 15 also give some indication as to the proper ratio of the bearing area to the superficial area of the bar in order that the best results may be obtained. In order to obtain a notion of the bearing stress developed at ordinary bond stresses the values given in the last column of the table have been computed on the basis of a bond stress of 100 lb. per sq. in., all of which was considered to be taken in bearing by the projections. We may readily conclude that 11%, as in the type A bars, and the percentages found for the other corrugated bars are all too small, since the bearing stress is entirely too high; 70%, as in the threaded bars, is too high since the bearing stress is absurdly low and failure came by piecemeal shearing. The proper value lies between 11% and 70%; it seems probable that values of, say, 20% to 25% would give satisfactory results. Of course, a satisfactory value could readily be determined experimentally. With a ratio of 20% the bearing stress would be about four times the shearing stress over the enveloping surface if we disregard the sliding resistance; considering sliding resistance, the bearing stress would be about three times the shearing stress, which appear to be suitable values. A bar with projec- tions extending around the circumference, practically perpendicular to the axis and having a height of, say, 1/10 of the diameter of the bar and spaced, say, ¥% the diameter apart, would answer the requirements mentioned above. The spacing now ordinarily used would require so high a projection in order to secure the desired bearing area that the height would interfere with the practical requirements of manufacture and it would not give a good distribution of stress along the bar. A ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 81 comparatively close spacing would give the necessary bearing area with the use of a minimum amount of metal. Advocates of deformed bars would do well to realize that a certain amount of metal must be sacri- ficed to the projections in a bar of proper design and to recognize the fact that if high bond resistance is desired, it must be paid for in much the same way as is done in dealing with tensile stress. e. Effect of Age and Miz. 43. Preluminary.—To determine the effect of age and mix, several series of pull-out tests were made, using 34-in. plain round bars and 34-in. corrugated square bars (type B), at ages varying from 2 days to 15 months and 314 years, with the following mixes: 1-4-8, 1-3-6, 1-2-4, 1-114-3 and 1-1-2. These tests are summarized in Table 16. All the bars were embedded 8 in. in 8-in. cylinders of concrete; they were all stored under the same conditions and tested in the same manner. All specimens tested after 4 days were stored in damp sand upon the removal of the forms; but the storage sand was allowed to dry out after about 2 years. The blocks with corrugated bars were reinforced against burst- ing by means of a spiral consisting of 6 or 7 turns of 14-in. round wire. Generally the specimens were tested in sets of 5 at each age. The specimens of each mix were made from two batches of concrete, dis- tributed as follows: two specimens with plain rounds and three of corru- gated squares for each age were made from the first batch and the remainder from the second batch. In the case of the 1-114-3 group a part of the tests with plain round bars was duplicated. The numbers of the batches used are given in the table; for further information re- garding the concrete reference may be made to Table 4. The “Highest Bond Stress Considered” for the corrugated bars is based on the stress at an end slip of 0.1 in., if a maximum had not been reached at a smaller slip. In only a few cases was the maximum load reached before a slip of 0.1 in. had occurred. The values given in Table 16 are the averages for the number of tests noted. For convenience of reference, the values from the com- pression tests of 6-in. cubes for each mix have been included. As a ready method of comparing the effect of age on each mix, the percentage of the bond stress for the several ages at an end slip of 0.01 in. to the bond .. stress for the same slip at age of about 60 days is given in the table. It will be seen that it would not make any material difference in the percentages given if the bond stresses developed at another slip or at the ILLINOIS ENGINEERING EXPERIMENT STATION 82 “gUlod plaré puoAaq 10 07 posses BIB » “XTUI OUTES OY} TIT 67694 ABP-09 243 Jo 48q} GIA posedurod se “Ur 10'0 Jo drys pua ue 4B soUBYsiser puog aq} JO 4099 Jed BY, 4 “dis Jo yUNOUTE Ja]|SUIs & 48 PUNO SBA PRO] UUNxwUT oN} Esa/ON ‘JEq 9q4 JO pus doz] aq} 48 “UT [0 JO die e 04 Surpuodses0| , avatalae | satecacari atte ll selease Gel eids ]! aneie.| Il aeater | ete aceee!| | oath set |4858 1+ | gee | gpg Z “uke cer ogeP 9 681 | 46b9T | “"" | ESET} O8IT| O8OT| FFB | 406 | O&Z & ae es arent st roe ol eS, rr “1k 4G out eer i Oeste Pee sen) ies Meee eee eee el eee: [ae ees o ost \atse | oe2! 62g! ste} oo! coz | ¥ | omar SOT OLE g raw QO@FT | “"" | °°" | °°" | OLOT] 996 | O28 | SbL ¢ SOI | 982 | FEL | LIL | £99 | II9 | 8T¢ g ‘oul g sol £686 & O@L 6LOT | SZOT} 128 | 292 | 202 | 829 | 929 | Leo g Fel 809 | 969 | ZOS | IIS | 694 | 6IF ¢ OzT oot 926 9 oot G8 | 028 | 289 | L429 | GSS | GLE | PEF | O6E g oot Goh | OSF | Gem | POE | E9E | ZEEE g 09 22 ge9T 9 SL TH9 =| 209 | OLS | LF | HIF | LE | 908 | F1z g 06 FOF | 268 | 09E | STE | 88Z | 99% g 82 9¢ Shel 9 gg LLY | OFF | TOF | ShE | 06% | OFZ | OIZ | 002 g 99 OOE | 682 | 842 | 192 | LEZ | 127 g *1 oF 896 9 L¥ 6Sh | Fah | 8h | G62 | 6FZ | 602 | LET | 99T g og 9% | 91% | 10% | #8T | SST | IFT g Z ce 9ZL 9 Ié » SOE ILG | 622 | 96 | OLT | SI | 62T | 80T ¢ ¥E sot GST | OFT | T€l | OIL | 26 g ¥ 1% o8y 9 <4 612 1O@ | SLT | SST | S2t | SIT | 26 LL g 13 Sel Sel | PIL | 46 | 68 GL g & wie sis si Giese. | teeter: || seane Pf aareses eRe cake Saeed dance Ih ration ozt | zze | eze| zze| $92 | cee | o8z e ou oy 691 0922 € J} TIT 2424 | 889 | 889 | 00S | OFF | BLE | 2zE | 692 & pepe oe] Gene (cecnalih teal tee lire ee 7 “oul g €21 OFoT $ Sit 9FL £69 | 989 | 809 | 98h | E98 | 9zE | 992 ¢ elT 98S | Sto | 86h | IhF | 868 | 2gE g cet oot Osst 9 00T 829 | 269 | 209 | OS | 86E | ZTE | 082 | The g oor TIS | 808 | £82 | 8hZ | L422 | 802 g 09 98 SPIT 9 69 cop | 98h | EOE | SIE | 69% | ITS | BLT | SST g LL T¥S =| O83 | 902 | GST | O&T | BIT g 8% o¢ LPL 9 OF 98% | 690 | TIS | SLT | SFI | IZE | FOL | 16 g €¢ got 8ST | SFT | 12 | Stl | FOL g L 8h 19 9 0€ 6&@ | €1Z | SOT | FEL | TIT | 88 | 2 69 g SZ LL 69 | 29 | a ey ) Le g y 98 GLY 9 81 ZSt | G&l | OTT | €8 | 89 | Gc | FF | 98 g LI e¢ lb) 1) | be | te | OF g 3 Cie ie = £03 8@Z | 689 | €8¢ | 00S | 9FF | BE | Zz | 69e & 96T SLE | $98 | EE | GZ | ESz | es £ “owl GT 701 060T 9 aa OLF | SP | SFE | 942 | 28z | GST | BST | TFT g TIT | OL | 280 | GAT | OLZT | ZT | OOT & Ser 00T 6F0T 9 oot T6E | $98 | ZOE | 9FZ | 0% | SST | BET | 6OT q OOT | O61 €8t | TLT | OST | SSI | Far ¥ $9 £9 199 9 99 S12 | OFZ | SBT | 29T | Sx | TIT | 46 | OL g 6L 6FL | SFI | GET | 82r | OZT | 90 g 8% 98 SLE 9 82 eet SIL} $8 | OL | 99 | Ib | Ge | 08 g 63 ¥¢ es | Lb | 88 | v& L& g L 92 ILG 9 rea Ort 48 | 99 | & | 6& | 6% | e% | LT ¥ 9% 6F Sh | oh | 98 | 2 | 8 g ¥ 8T 161 9 Il #9 Ig | 248 | 8% | 12 | Zt | et | 2 ¥ ¥1 Le S@ | 8 | 6t | ZT | OF g z «Peso waQUMN | WIFTe7yg} By697, 44aq “PIs | GO" | ZO" | 10° | 00°} Z00"| 100° )$000°0| 8389.1, 44aq_ | 889199] 10° | $00°} 200°] 100°)$000°0| #389, [suo |Zurqsni9| jo -umy | -w09 jo -umy | puog jo " lodoig | a#eiaay | Jequmyy || peony | 882.99 73q jeuor | uma eq sep -1od | puog (seqout) -umy sod |-rxe yy (seyout) -umN 482], “01g a Jo dyg pug 48 sseyg puog Old jo dyg pug 48 ssang puog ay. ~qatH BAND “UI-9 Jo #169, gieg elenby payesdnu0g “ul-% sreg punoy WIejy “UI-% ZI pus y soyneg Fel 6I paws oye, oT 9z UE TZ sane srl eazeg ‘OT HTAV.L “your aienbs Jad spunod a uaara are sasser3g “((Q) 1 ‘3tq) [ertds 8 Jo uns0; 943 UI am “UI-F4 Jo suN}y J 10 g Jo sueeM Aq Zury8mq qeuyede padsoyuted ala BIBq pezBsNLI0d qa suouoeds ayy, ‘pues durep ut poloys o194 sXep % 104Je pajse} suaUTIOadg a4aI0T00 PaxTWI-puLy ‘yusUIe. [BSIBATUE) “JapUT]A “UI g UY TI peppequie aIEg “XI. GNV GOyY fo Load 83 \ ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL *XTUI OUTS OY} ILA 8780} ABP-99 947 Jo 48Y} GIA poredmi0d se ‘at [Q'O JO dijs pus uv yw aoUB,sIeA7 PUOg oq} Jo 4Ua0 Jad ayy, | “qutod ple aq} puoseq Jo 07 passers aIVg 4 “dys JO JUNOUTE Ja][VUIE B 48 PUNO; SEM PLO] UMUTTxeUT OG} ssaTUN ‘JBq OY} JO PUD deIJ 94} 4B “UT [°C JO dijs B 07 SuIpuodeauosy , a fuss ¢ “6 Slee. see Hicoe tl ies [ages th eS uo 342 See t || ea z || 2et | 622 | 602} sez | $89 | 090 | #9% | g | ‘omer ae ee : e || ozt | 999 | o#0| eto | 199) ze | cer | z 121 ee pas oz 00T ecg a1 g oot | oss | 929 | gem | O€F | eee | FSe | ¢ 19 saqnyeg 6o GZIz 9 ¢ || 2 | eee | gze| soe | sze| ooe | azz | 9 8% a Ze 6SIT 9 ¢ Ly | 96% | SF | OFZ | 122 | coz | Zar | F L ert itd LL 9 $ ze | 2er | OBt | OBt | 121 | ost | eet | g ¥ &I 06 9 9 || 22 | tet | Tet | set | stl | 20r | 66 g z it pos . ch, || | ee eel) teach asepeel sora fall bch “1k 34e : ce 2 : zor | ee | 408 | $92 | 809 | 9 "14g cor ene : 1 a i 0 kg ee a : zor 4968 | 9z8| 168] 268 | oz8} SOL | ‘our FT ee pu et ‘z LEI 8268 g z || oct | 209 | 09 | sc9 | seo | seg | ay | 4 Boyde oot 182 9 g || oor | ¥sg | ¥s9| eco | Bos | zeR | Tce | OF 20 : 0g 8822 9 g 66 | OFS | Sho | 4z¢ | 28h | ZoF | sze | og ze SALT 9¢ FF01 9 9 || ss | ooe | 962 | sez | zoz | osz | Tz | 8 Z 9% sll 9 9 a | Tez | $22 | gee | O1% | 961 | Ozt | 2 y 9 Os 9 $ 6% | 6st | Sot | ost | Fe | eet | att | 8 z d20q PereE| oy. | a9: | 10: so0°| z00°| 100" |s000"o| #9824, at ssaxyg| 10° | $00'| Z00'| 100°|$000"0| #1891, aan qy3ueng "L-com saan jo. || peur | puog jo serigey area aqamany perror9 Teor ; det |) toe. | ania w0q ee “tod 489 (saqout) -umN || -01g |-xeyy (saqout) “any 9821, azas000d, “Hd | asc Jo dig pug 98 svanjg puog jo dug pur 48 seang puog 48 t : esy BOND “UI-9 Jo 848], reg erenbg payednu0y “w-% Beg punoy UIE[g “-% ‘XI. GNV GOV 40 Logaay “dH ONILNOD—9OTL ATAVL 84 ILLINOIS ENGINEERING EXPERIMENT STATION maximum load were used as the basis of these values. Two groups of tests were made on mixes of the forms 1-2-m and 1-n-2n, where m varied from 0 to 8 and n from 1 to 5. The tests were made with 34-in. plain rounds embedded 8 in. in 8-in. cylinders. 800 70 2-in. Plain Rounds Embedmert 8in. § 8 8 Bond Stress - Lb per 5q./n g 8 100 —----Gond Stress atEnd Slip of ‘Maximum Bond Resistance o 20 40 60 G0 120 140 160 180 00 100 Age - Days Tic. 27. Errect or AGE AND Mix on Bonp RESISTANCE OF PLAIN Rounp Bars. 44, Effect of Age on Bond Resistance——Fig. 27 shows the bond stresses developed in the pull-out tests with 34-in. plain round bars with the different mixes of concrete at ages up to 6 months, at an end slip of 0.0005 in. and at the maximum. Load-slip curves for the 1-2-4 concrete at the various ages using plain bars are given in Fig. 28 and for the corrugated bars in Fig. 30. The load-slip curves for other mixes are omitted, but the values plotted for the 1-2-4 mix may be con- sidered representative of the effect of age in the entire series. The tests at early ages gave surprisingly high values. In the 1-2-4 concrete tests with plain bars at 2 days, slip did not become appre- ciable until a bond stress of 75 lb. per sq. in. was developed. The maxi- mum bond resistance at this age was 123 lb. per sq. in—27 % of the ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 85 maximum at 60 days. At 4 days slip became appreciable at 97 lb. per sq. in.; maximum bond resistance 153 lb. per sq. in—34% of the maximum at 60 days. The rate of growth of bond resistance was great- est from 2 to 7 days of age; for the older tests the rate of growth grad- ually became less. From 4 days to 28 days the growth in bond resist- 900 Concrete 600 Plain Rounds 700 § 8 8 Bond Siress -lh per SGM g 0 07 08 09 JO Qo Ol 02 03 04 OF O06 Slip of Bar- Inches Fic. 28. Loap-stip Curves FoR PLain Rounp Bars EMBEDDED IN 1-2-4 CoNCRETE. ance for the various amounts of slip is nearly proportional to the age, as shown by the straight lines in Fig. 27 and 29; for 28 days to 6 months the growth in resistance is again proportional to the age. The rate of growth of bond resistance at early ages will probably depend largely on the quality of the cement used. At 60 days slipping began in the 1-2-4 concrete at a bond stress of 332 lb. per sq. in.; the maximum bond stress at this age was 452 |b. per sq. in. These values correspond closely to those found for 1-2-4 concrete in the tests reported in University of Illinois Engineering Ex- periment Station Bulletin No. 8. The bond stresses developed are about 86 ILLINOIS ENGINEERING EXPERIMENT STATION the same as found in the 1912 beam and pull-out tests for 1-2-4 con- crete with 1-in. plain round bars. The bond stresses at a slip of 0.001 in. for ages 120 days, 6 mo. and 17 mo. are 129%, 168% and 220%, respectively, as compared to that at age of 60 days. It seems probable that the bond resistance for small slips, such as 0.001 in., for concrete of the kind used in these tests may be expected ultimately to reach a value as much as twice that developed at 60 days. For mixes leaner and richer than 1-2-4 the relation between the bond resistance developed and slip of bar for the plain round bars is much the same as that found with the 1-2-4 concrete. The maximum bond resistance for the plain bars was found at a slip of about 0.01 in. and the bond stress at larger slips drops off in a manner similar to that shown in Fig. 28. In all cases the bond resistance for a given end slip increased with age, except for certain irregularities at the older ages. The values in Table 16 will enable the reader to make a further study of these tests. The tests on 34-in. corrugated squares formed a group parallel to those with 34-in. plain rounds discussed above. The load-slip curves for the 1-2-4 concrete are given in Fig. 29. The load-age curves for all the mixes used are given in Fig. 30. During the interval between 2 days and 7 days the bond resistance of the 1-2-4 concrete increased about 100%. From 7 days to 6 mo. there is a nearly uniform increase in bond resistance with age. During this interval the rate of growth in bond resistance is about 110 lb. per sq. in. per month, for the beginning of slip, and 175 lb. per sq. in. per month at the maximum. The values for the 1-2-4 concrete at 180 days seem abnormally high as compared with those at earlier and later ages and as compared with the other mixes. For the smaller amounts of slip at ages under 6 days the bond resistance developed by the corrugated bars is not materially different from that found in the tests of plain round bars. For an end slip of 0.001 in. in the tests at 2 to 28 days, inclusive, the corrugated bars aver- age 8% higher than the plain rounds; for ages of 60 days and over the average is about 25% higher. As the specimens were made from the same concrete atthe same time, the results give comparative values for the two forms of bar. Absolute values will vary, of course, with the concrete used. At an end slip of 0.01 in., corresponding to the maximum bond resistance of the plain round bars, the corrugated bars developed from 10% to 25% more bond resistance than the plain rounds. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 87 /400 4300 1200 H00 1000 1-2-4 Concrete 900 ae Corrugated Square Vocks reinforced &§ s 800 & 3 wn gH & 700 g 600 500 400 300 200 /00 o 0 0 40 60 60 100 20 140 160 480 200 Age - Days Fic. 29. Errect of AcE anp Mix on Bonp RESISTANCE oF CoRRUGATED Bars. 88 ILLINOIS ENGINEERING EXPERIMENT STATION 1300 1200 00 1/000 900 gs 8 8 Bond Siress -Lb per 59.1. a 8 400 300 100 oe 3”. Corrugated Squares oO o Ol 02 03 04 05 06 07 08 09 a Sip of Bar-inches Fic. 30. Loap-s_tip Curves ror CorruUGATED Bars EMBEDDED IN 1-2-4 CONCRETE. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 89 With 1-4-8 concrete, for the smaller amounts of slip, the corrugated squares give lower values than the plain rounds for all ages under 15 months. This is true for slips up to and including 0.001 in., and in case of the 2 and 28 day tests the values for corrugated bars were lower until a slip of nearly 0.01 in. was produced, a slip corresponding to the maximum bond resistance of the plain rounds. The bond resistance for the 15-month tests is about twice that of the 60-day tests. TABLE 17. CoMPARATIVE VALUES ror Bonp on Piain Rounp Bars. Compiled from the pull-out tests with 34-in. plain rounds given in Table 16. The values in each of the two principal divisions of the table are given as percentages of the corresponding value for the 60-day tests with 1-2-4 concrete. At End Slip of 0.001 in. At Maximum Bond Resistance Age at Test 1-4-8 | 1-3-6 | 1-2-4 |1-114-8] 1-1-2 || 1-4-8 | 1-3-6 | 1-2-4 |1-114-3] 1-1-2 DGB: e u 8 8 /00 50 o O 002 004 006 G00 0 00? O04 006 008 O10 Ol? Olt WE OK 220 Slip of Bar - Inches Fic. 42. Loap-stip Curves FOR SPECIMENS OF THE Form SHOWN. 59. Autogenous Healing in Concrete-——The term “autogenous healing” may be applied to an interesting group of phenomena which were brought out in the tests on pull-out specimens loaded one or more times after having been previously tested to their maximum resistance, and in the compression tests on concrete cylinders which had been loaded to failure at an earlier age. These tests were suggested by certain ob- servations made by the writer during the summer of 1911, while con- ~ 108 ILLINOIS ENGINEERING EXPERIMENT STATION ducting a load test on a 40-ft. through girder reinforced concrete bridge for the Illinois Highway Commission.* This test bridge was built by convict labor in December, 1907, inside the prison yard at the Southern Illinois Penitentiary near Chester, under the direction of Mr. A. N. Johnson, State Highway Engineer. In May, 1908, a test load amounting to 106 tons was placed on the bridge floor and removed. This load was sufficient to cause numerous diagonal tension cracks in the web near the ends of the girders. These cracks were very minute and were not observed until subsequent weathering brought them out. Before begin- ning the 1911 test all such cracks were carefully mapped, and the girders whitewashed. At this time crushed stone and pig iron were piled on the bridge floor and on the girders sufficient to give a live load of 309 tons on one girder. During the progress of the loading, some of the original cracks re-opened at loads much higher than that applied in the first test. Many of the original cracks, however, did not 1e-open under the final load, although new cracks sometimes formed in the immediate vicinity. The bridge was about four months old when first loaded, but it is probable that the concrete strength was not greater than would have been found at, say, two months, under more favorable tempera- ture conditions. The behavior of the girders could be partially accounted for in the increased strength and stiffness of the concrete acquired during a period of more than three years since the first test; but this would scarcely account for the concrete of the girders being able to develop, without re-opening some of the cracks, a diagonal stress which was equiv- alent to about 6 times the stress which originally caused them. The only reasonable explanation was that the fractured surfaces of, the concrete at the cracks had “knit” together and entirely healed during the interval between the tests, giving a joint which was in many in- stances stronger than the unbroken concrete. The tests described in the following articles on the effect of loads reapplied after failure in bond and compression were planned to throw further light on this sub- ject. The phenomena discussed in Art. 60 and 63 are only other mani- festations of the effect of retarded or interrupted hydraulicity, which have frequently been observed in tests on concrete setting under low tem- peratures, retempered concrete, cement reground after setting, etc. 60. Effect of Loads Reapplied after Failure of Bond.—Tests were made on 63 pull-out specimens to determine the effect of reapplying *“Test of a 40-ft. Reinforced Concrete Highway Bridge,” by D. A, Abrams. Proceedings of the American Society for Testing Materials, 1913, s ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 109 loads to specimens which had been tested to failure some time before the retest was made. Both plain and cold-rolled round bars were used. The specimens were stored under various conditions. See Table 21. Some of these specimens have been loaded as many as five times, and the tests will be continued. The load-slip curves for two of the groups of tésts in Table 21 are given in Fig. 43 and 44. The numbers in parentheses adjacent to the curves correspond to the number given to the tests in the table and indi- cate the order of loading. All specimens in a group were made from the same batch of concrete. Group (a) was a preliminary series. At the age of 5 days part of the bars were pulled out 0.001 in. at the free end and the remainder to the maximum load, which came at an end slip of about 0.01 in. Some of the specimens were then stored in water and some in air. In the sub- sequent tests on these specimens all bars were pulled out 0.1 in. It makes little difference in the subsequent tests whether the bars are pulled out 0.001 in. or 0.01 in. at the first test; in fact those pulled to 0.01 in. gave somewhat higher values in the second tests. With one exception (No. 3) the specimens stored in water during the interval between the first and second tests show considerably higher values than those stored in air. For all the specimens in this group the maximum bond resistances for the third test average 24% greater than those for the second tests made 4 or 5 months earlier when the concrete was 1 or 2 months old. The fourth tests, after another interval of 4 months, when the concrete was 10 months old and the bars had previously been pulled out over 0.2 in. gave maximum bond resistances which average the same as the second test at 1 or 2 months—about 290 lb. per sq. in. It is interesting to note that the period of 1 or 2 months in air follow- ing the first tests had the effect of temporarily retarding the increase of bond resistance which was largely overcome by the subsequent period of water storage. Group (b) consisted of two parallel series of tests on 1-in. plain round bars stored in air and in damp sand. In all these tests the bars were pulled to the maximum load, corresponding to an end slip of about 0.01 in. Two specimens for each condition of storage were tested at 5 days, 30 days, 3 months, and 1 year. At each test period, all speci- mens which had been tested previously were loaded to the maximum 110 Pull-out tests; embedment 8 in. in an 8-in. cylinder of concrete. ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 21. Errect or Loaps REAPPLIED AFTER FAILURE oF Bonn. given in Fig. 43 and 44. Stresses are given in pounds per square inch. Load-slip curves are ip of (inch Max. Ref. No. Age as Bond Stress at End Slip of (inches) Bond No. | Tests] Test 0005} .001 |.002 |.005 | 01 | .02 | 06 | io: Pease (a) 1-in. Plain Rounds; 1-2-4 Hand-mixed Concrete (1912). Bar pulled out 0.1 in. in each test unless otherwise noted. 1 1 | 5 dayal(1) Pulled to 001 in.............. T48 | TA TH eectstine| oases feacne [atstenll desta staan deus aes Pa Eee 6 mo. {0-ame, |(4) joored 1a saber Shier cod 233 | 254 | 286 | 298 | 300 | 278 | 247 | 217 | 305 21-2mo,.| 5): 'fy ere ee ee ee 223 | 239 | 243 | 260 | 265 | 250 | 229 | 203 265 2 2 | 5 daya|(1) Pulled to 001 in.............. DGD.) ASD: | aciiccastsa| eetactvl seca ece' te ateatsel] aiesites | nora ae |cieleenteos . ea ee 111 | 198 | 233 | 287 | 202 | 206 | 182 | 160 | 257 5 mo. ‘ 7 10 mo. | fee in water after oe 161 | 183 | 214 | 297 | 241 | 290 | 205 | 171 | 241 21 mags. 15). J Neen eines ee aoe tessa eases 185 | 209 | 224 | 246 | 251 | 241 | 216 | 191 253 3 3 | 5 days|(1) Pulled to 001 in.............. 4D: | MED | cusctes, 5 |evenceay | etdeeaas [Geese cos Sncicae ls araettar Pegteaeads 72 days|(2) ioe | abe sly 211 | 236 | 251 | 265 | 240 | 230 | 225 | 182 278 5 mo. |(3) Stored in water.. 260 | 290 | 326 | 342 | 343 | 295 | 256 | 215 343 10 mo. |(4) Stored in water.. 182 | 209 | 237 | 266 | 276 | 261 | 230 | 204 276 21 mo. |(5) Stored in water... 187 | 197 | 214 | 231 | 236 | 230 | 214 | 191 238 4 1 | 5 days|(1) Pulled to maximum........... 154 | 161 40 days|(2) Stored in air .| 139 | 199 +6 ae ‘a iain in water after noma i ed AAPM: CS) ec ee hs EE 159 | 175 5 2 | 5 days|(1) Pulled to maximum. 202 40 days|(2) Stored in water........ 282 6 mo. |(3) Stored in water........ 328 10 mo. |(4) Stored in water........ 256 21 mo. |(5) Stored in water............... 234 6 3 | 5 days|(1) Pulled to maximum 189 72 days|(2) Stored in water.......... 295 5 mo. |(3) Stored in water.......... 275 10 mo. |(4) Stored in water.......... 229 21 mo. }(5) Stored in water............... 212 (b) 1-in. Plain Rounds; 1-2-4 Machine-mi 113 2 | 5 days’ ie ( | 113 | 124 1 mo. Stored in air 192 | 216 3 mo. Gy Pulled to maximum 177 | 220 1 yr. | (4) 39 | 55 17 2 | 1 mo, | (1) 162 | 173 3 mo. | (2) do. 165 | 208 lyr. | (8) 41 | 69 121 2 | 3 mo. | (1) 175 | 194 ie lien do: | 38 | 67 125 2 |1yr. | (i) do. 81} 119 115 2 | 5 days} (1) 134 | 154 1 mo. | (2) pasa in damp sand 265 | 301 3 mo. | (3) Pulled to maximum 4272 | 310 1 yr. | (4) (| {407 | 425 119 2 |30 days} (1) 230 | 238 3 mo. | (2) do. 307 | 348 1 yr. | 3) 390 | 430 123 2 | 3 mo. | (1) 354 | 378 | 395 | 397 |.....J.000,...0.f.00. 400 1 yr. | (2) } do. { 407 | 453 | 478 | 481 |...0.)0000f0..0J.00. 485 127 2 {1 yr. | (1) do. 424 | 443 |} 446 | 450 J.....4..00.[.0004 0. 451 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL TABLE 21—ContiInvED. 111 il: No, Age 2 & Bond Stress at End Slip of (inches) Max, 0 al emar! ‘ i No. | Tests] Test 0008} .001 02 005 | .01 | .02 | .05 | .10 |Resist- (c) 1-in. Polished Rounds; 1-2-4 Machine-mixed Concrete.” (1912). 101 | 2 | 5 daye|(1) ; 119 ch | Sealer fe eeaz lebeae elles ane 119 1 mo. |(2) Stored in air 108.) WUT Vice sex Alisenttcoea lena nef artese:cll enna wea 2 111 3 mo. |(3) Pulled to maximum 80 95 lyr. |(4) 31 32 105 2 1 mo. i | 192 3 mo. |(2) do. 120 lyr. |.3) ( 23 109 | 2 | 3 mo. |(1) } do. { 152 1 yr. |(2) 47 103 2 | 5 daya|(1) 122 1 mo. |(2) Stored in damp sand | 170 3 mo. {8 Pulled to maximum 164 1 yr. |(4) { 210 107 2 | 1 mo. |(1) ) 157 3 mo. |(2) | do. 147 lyr. |(3) } 178 1 2 | 38 mo. |(1) } do. { 224 1 yr. |(2) 133 (d) 1-in Plain Rounds from 1912 Beam Series; 1-2-4 Hand-mixed Concrete.° Bar pulled out 0.1 in each test. 1054.1 | 3 | 2 mo. |(1) Stored in air................. 329 | 373 | 398 | 433 | 450 | 447 | 400 | 348 | 459 7 mo. |(2 259 | 298 | 339 | 373 | 389 | 370 | 343 | 394 389 11 mo. |(3): }Stored in water after first test;/ 215 | 240 | 264 | 296 | 318 | 312 | 295 | 265 319 21 mo. |(4) 251 | 271 | 288 | 307 | 323 | 338 | 314 | 292 338 1051.1. | 3 | 2 mo. (1) Stored im air................. 300 | 331 | 361 | 375 | 382 | 368 | 334 | 284 386 7 mo. 220 | 252 | 284 | 316 | 337 | 309 | 275 | 243 337 11 mo. |(3) }Stored in water after first test;| 196 | 220 | 245 | 275 | 287 | 287 | 267 | 231 291 21 mo. |(4) F 203 | 212 | 220 | 247 | 262 | 266 | 246 | 227 266 1056.1 | 3 | 2 mo. |(1) Stored in air................. 332 | 405 | 427 | 447 | 454 | 417 | 364 | 304 454 7 mo. |(2 ({} 198 | 233 | 265 | 292 | 306 | 296 | 274 | 249 306 11 mo. |(3) }Stored in water after first saa 171 | 184 | 206 | 234 | 252 | 257 | 240 | 221 257 * 21 mo. |(4) 191 | 199 | 208 | 229 | 241 | 248 | 234 | 219 248 1055.1; 3 | 2 mo. |(1) Stored in air................. 344 | 397 | 417 | 441 | 417 | 379 | 323 | 262 443 7 mo. |(2) Stored in water after first test..| 168 | 194 | 220 | 254 | 265 | 264 | 236 | 209 267 11 mo. |(3) caer J} 188 | 218 | 233 | 247 | 251 | 243 | 229 | 205 248 BL mor {cay ;Stored in air after second test,| 755 | “77 | “3s | 110 | 132 | 140 | 144 | 136 | 147 1059.1] 3 | 2 mo. |(1) Stored in air................. 421 | 525 | 564 | 586 | 584 | 547 | 473 | 409 591 7 mo. |(2) Stored in water after first test...) 205 | 252 | 278 | 315 | 334 | 339 | 321 | 296 | 339 11 mo. |(3) | oe 200 | 236 | 255 | 273 | 282 | 288 | 277 | 259 290 21 imo; |(G) }Stored in air after second test{} "G9 | 753 “96 | 117 | 134 | 150 | 183 | 186 | 187 1046.1+| 3 | 2 mo. |(1) Stored im air................. 233 | 258 | 275 | 306 | 326 | 330 | 316 | 289 | 336+ 7 mo. |(2) 185 | 220 | 266 | 301 | 316 | 295 | 270 | 238 316* 11 mo. |(3) }Stored in water after firat test} 166 | 183 | 202 | 233 | 245 | 250 | 234 | 213 252+ 21 mo. |(4) 179 | 190 | 198 | 211 | 225 | 282 | 215 | 202 232+ * Groups (b) and (c) were made from the same batch of concrete. Compressive strength of 6 6-in. cubes, tested at 3 mo., 2720 lb. per sq. in. ° Compressive strength of 18 6-in. cubes, tested at 63 days, 2360 lb. per sq. in. + 1-in. plain square bars. 112 ILLINOIS ENGINEERING EXPERIMENT STATION again. Thus, as indicated in the table, some of the specimens have been loaded to their maximum as many as four times. The load-slip curves are given in Fig. 43. The notable feature of these tests is that the water-stored specimens which had been previously loaded to a maximum once, twice, or three times, gave values greater than or equal to those found in the specimens, which at the same age, were loaded for the first time. The same statement is true of the air-stored specimen, except that all the 1-year tests, both on specimens which had been previously loaded and those loaded for the first time, gave lower values. The two pairs of specimens tested at 5 days gave nearly identical results, as may have s 8 Bond Srress-Lo. per Sqn. 8 100 oO Slip of Bar- Inches Fic. 43. Loap-stip Curves For Loaps REAPPLIED AFTER FAILURE OF Bonn. been expected, since the difference in storage conditions would not affect the relative strengths at that age. During the interval between the first and second tests of these specimens (5th to 30th day) the relative in- crease in maximum bond resistance is about the same for air storage and damp sand storage—37% and 30%; during the interval between the first and second tests of specimens No. 117 and 119 (1st to 8rd month) the increase was 9% and 31%, respectively. Group (c) was similar in every way to group (b) except that pol- ished bars were used. It was thought that by comparing the behavior of plain and polished bars in these tests, additional information as to the components of bond resistance would be obtained. Generally, only one observation on slip of the end of the polished bar within the maxi- mum load could be obtained. The maximum bond resistance came in \ ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 113 400 300 1051.1 |-In, Plain 100 Bond Stress -Lb per 5g./n. 100 All specimens were stored in water affer fest Ol 06 Slip of Bar- Inches Tic. 44. Loap-strp Curves ror Loaps REAPPLIED AFTER FaiLureE oF Bonn. 114 ILLINOIS ENGINEERING EXPERIMENT STATION all the tests at a slip of about 0.001 in. The results of these tests are not as consistent as those on plain bars, but some interesting facts are brought out. A study of the tests in groups (b) and (c) indicates that the bond resistance which is developed before the free end of the bar of the s.ze used in these tests has slipped, say, 0.001 in. is due to adhesion. The value of this adhesion is dependent on the amount of moisture present and the age of the concrete. In the case of the polishd bars the original bond resistance was generally regained in the specimens stored in damp sand, but the increase with age was small. The air-stored specimens show about the same bond resistance as the sand-stored specimens in the first tests, but upon reapplication of load there is a material falling-off in strength. With the round bars of ordinary surface the adhesive re sistance was entirely restored, except in the 1-year tests on air-stored specimens. The amount of increase in bond resistance apparently de- pends upon the presence of the water necessary for the continuation of the hydraulic action of the cement. The drying-out of a specimen after a certain period interferes with the phenomenon here discussed. Group (d) included 6 sets of pull-out specimens from the 1912 beam series. All of these specimens were stored in air up to the time of the first test at the age of 2 months. In each of the tests the bar was pulled to an end slip of 0.1 in. The load-slip curves are given in Fig. 44. These tests are of interest in showing what may be expected when the specimens have attained a considerable age before the first failure of bond. In no instance were the values from the second to the fourth tests, as high as found in the first tests. The average values for all the specimens with plain round bars (disregarding minor varia- tions in storage conditions) are 406, 246, 220 and 167 Ib. per sq. in. for the first, second, third and fourth tests, respectively, at a slip of 0.001 in.; and 467, 327, 281 and 237 lb. per sq. in. for maximum bond resist- ance. It will be recognized that the conditions in this group were the most adverse of any of the groups in which the load was reapplied after failure. The bond resistance did not subsequently reach the amount found in the first tests, but it is noteworthy that even at the fourth loading comparatively high values of bond resistance were found. The foregoing tests are of value in showing that a considerable displacement of the bar, and a frequent disturbance of the bond be- tween the concrete and steel, even several days or weeks after placing the concrete, does not necessarily produce permanent weakness. If ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 115 disturbance occurs after a longer period the final effect will probably depend upon whether sufficient moisture is present in the concrete. These results give added confidence to the permanency of bond and indicate that plain bars may properly be used in certain classes of work in which they have sometimes been considered unsuitable. It should be noted that the specimens were unstressed in the interval between loadings. 61. Bond Resistance of Concrete which Set under Pressure. As a means of gaining further knowledge as to the nature of bond resistance, a few tests were made on specimens in which the concrete was caused to set under pressure. 1-in. plain rounds, 1-in. cold-rolled rounds and 11%-in. corrugated rounds were used. The pressures used were 0, 6, and 100 Ib. per sq. in. The specimens were molded in the usual way, except that in the specimens which set under a pressure of 100 Ib. per sq. in. the long end of the bar was upward. The pressure was applied immediately after placing the concrete in the form. A circular plate with a central hole which allowed the rod to pass through was placed inside the form on the fresh concrete. The 6-lb. per sq. in. pressure was obtained, by piling weights on the cover plate; the 100-lb. per sq. in. pressure was obtained by setting the form in a pan on the bed of a testing machine and running the head down onto a nest of springs, which transmitted the load to the cover plate. These loads remained on the specimens for 5 days. After removal from the forms, the specimens were stored in damp sand. The tests are summarized in Table 22. The load-slip curves are given in Fig. 45. The increase of maximum bond resistance and other properties with the pressure under which the concrete set are shown in Fig. 47. The maximum bond resist- ance for plain bars in concrete which set under pressures of 6 and 100 lb. per sq. in. are 9% and 91% higher, respectively, than the correspond- ing values for concrete setting under no pressure. In the case of the cold-rolled rounds, the increase due to the pressure was slight. The value for 6 lb. per sq. in. pressure seems abnormally high. The corrugated bars show a large increase in bond resistance due to the pressure. These values cannot be compared directly, since some of the specimens were tested without removing the steel-pipe forms in which they were made. However, the corrugated bars show an increase in bond resistance of about 100% as compared with specimens tested under similar conditions which had set without pressure. The specimens which set under a pressure of 6 lb. per sq. in. (No. 91) show an in- 116 ILLINOIS ENGINEERING EXPERIMENT STATION crease of 66% to 90% as compared with the values for the same amount of slip in the specimens setting without pressure. The concrete blocks split in these two sets of tests at the same slip—0.02 in. Specimens No. . No. 89 shows the effect of leaving the steel form in place during the test of a specimen which set without pressure; No. 95 shows the effect of casting the specimen with the long end of the bar upward. For a discussion of other tests with corrugated bars showing the effect of reinforcing the block against bursting, see Art. 64. 93 made and tested in a steel form gave very high values. TABLE 22. Bonp REsIsTANCE OF CONCRETE WHICH SET UNDER PRESSURE. 1-2-4 concrete, machine-mixed; embedment 8 in. Stresses are given in pounds per square inch. Age at test 80 days. Bond Stress at End Slip of—inches Maxi- mum Ref. | Num- Qa Bond No. | berof Characteristics Re- Tests .0005 } .001 | .002 | .005 | .01 02 -05 10 sist- ance ¢ (a) 1-in. Plain Rounds. 71 2 Without pressure............. 273 302 | 328] 345) 351} 350] 320] 290 353 73 6 Concrete set 5 days under pres- sure of 6 lb. per sq. in....... 221 256 | 288) 382] 363 | 378| 351] 297 378 77* 2 Conerete set 5 days under pres- sure of 100 Ib. per aq. in..... 387 457 | 546] 634] 671 | 673 | 570] 477 674 (b) 1-in. Cold-Rolled Rounds. 79 2 Without pressure............. LEE Nexis sis leoveaaallaann aa reneuale arewaless tolled s 141 81 1 [Concrete set 5 days under pres-| sure of 6 Ib. per sq. in....... DOD isanee set eet satus lhe yer see caeneccnl| oad desl lesava ug allecinaseiy 222 83 2 Concrete set 5 days under pres- > sure of 100 Ib. per sq. in..... Io ester Nisesedasnai owrase asl cease] da“ udes 94 81 163 (c) 11%-in. Corrugated Rounds. 85 2 Without pressure. Nospiral..| 234 248 | 258; 278] 312] 341 ]......]...... 375 87 2 Without pressure. Spiral— See Fig. 1 (b)...... 244 279 | 314] 359 | 412] 517] 809} 1115 | 1115 89 2 Without pressure, WED 6 save salle ta 28 4aren a6 347 | 404) 420] 438] 516] 605] 855 | 1070 | 1070 91° 2 {Concrete set 5 days under pres- sure of 6 Ib. per sq. in....... 380 413 | 468) 503] 617] 672] .....|...... 720 93° 2 |Conerete set 5 days under pres- sure of 100 lb. per 89. in..... 682 937 | 1180 | 1417 | 1600 | 1783] .....)...... 1960 95 4 Cast with long end o} “bar up- ward as in Fig. 1 (¢); spiral..} 408 508 | 580) 687] 793 | 881 | 1140 | 1250 | 1250 * Specimens molded in forms made of 8-in. sections of 8-in. steel pipe. The forms were cut off at age of 5 days. ° Tested with steel pipe form in place. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 117 62. Compression Tests of Concrete which Set under Pressure.— In order to make a further study of the effect of allowing concrete to set under pressure, a series of compression tests was made on 8 by 16-in. cylinders. It is believed that these tests are of sufficient interest to warrant their inclusion in this report. Seventeen cylinders were 48300 Na71, I-in, Plain Round ; without pressure, Na7RZ #4 i ser under 6 1h par sgn. 1600 No77. is joo” NGS. Bhankend 3 i 70 ) pital, MAST 148 8 ph ral, M69 * *" a : ithout, rp NMaglo* * 7 iver or 6 Spin 1400 Nave" # # Oe tip eel ne No9S " " * oe Gof ber ipmares £ = 1200 & 8 % 3 1000 & 5 8 /-2-4 Concrere, § 600 Lrabedmery 6 tr, Age G0 days. Al specimens Fron Same COPCrere, Qo Q a C2 OF o4 OF 06 O7 1008 og SO JSlia of Bar-lnches Fic. 45. Loap-s~tip Curves ror MIScELLANEoUS PULL-oUuT SPECIMENS. made from a single batch of concrete. Two of these were allowed to set under normal conditions; three set under a pressure of 6 lb. per’ sq. in.; five set under 20 lb. per sq. in. and five under 100 lb. per sq. in. Some of the specimens remained under pressure for 1 day, some for ? days and some during the entire period of storage. Two specimens set for seven days under 3 lb. per sq. in., after which they were placed in a testing machine and loaded at the rate of 100 lb. per sq. in. per - 118 ILLINOIS ENGINEERING EXPERIMENT STATION week until a total load of 500 lb. per sq. in. had been applied; this load then remained on the cylinders until they were tested at the age of 80 days. The method of applying the pressure during the setting and hardening of the concrete was similar to that described in Art. 61 for the pull-out specimens. Details of these tests are given in Table 23. Five of the cylinders were tested at 7 days; and the remainder at age of 80 days. The retest of the five cylinders which were originally tested to failure at 7% days will be discussed in Art. 63. Deformations were measured over a 10-in. gage length by means of a wire-wound instrument which was a modified form of the Johnson extensometer. The weight of the concrete, the compressive strength, and the initial modulus of elasticity of the cylinders are given in the table. Typical stress-deformation curves are given in Fig. 46. The variation in compressive strength, initial modulus of elasticity, and density with the change in pressure are shown in Fig. 47. The first notable feature of these tests is that it makes little or no difference in the strength and properties of the cylinders, whether the concrete remained under pressure for 1, 7 or 77 days. In other words, if the concrete takes its final set and hardening begins under pressure there is nothing to be gained by continuing the pressure for a longer period. For this reason the values for all the cylinders for each pressure which were tested for the first time at an age of 80 days were averaged in computing the percentages in. Fig. 47. The consist- ency of the values indicates that the pressure probably had the effect of producing an unusually homogeneous concrete and justifies us in placing confidence in the results of the tests. The stress-deformation curves in the lower right section of Fig. 46 show the effect of different pres- sures on the compressive strength and the modulus of elasticity of these cylinders. The compressive strength was increased from 1840 lb. per sq. in. for no pressure to 3140 Ib. per sq. in. for a pressure of 100 Ib. per sq. in.—an increase of 73%. The compressive strength of the 8-in. cylinders setting without pressure is 91% of that for 6-in. cubes tested at the same age. The curve in Fig. 47 indicates that over one-half the increase occurred at pressures below 20 lb. per sq. in. It seems prob- able that the strength of the concrete which was subjected to a pressure of 100 lb. per sq. in. was considerably reduced by the loss of the water which was forced out when the pressure was first applied, owing to open- ings around the cover plate. On account of the danger of getting the ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL TABLE 23. 119 Compression TESTS oF CONCRETE WHICH SET UNDER PRESSURE. The test pieces were cylinders 8 in. in diameter, 16 in. long. Concrete 1-2-4, machine-mixed. All specimens were made from the same batch. The cylinders were stored in damp sand after seven days except those setting under pressure for a longer period and those tested originally at age of seven days. The cylinders which set under pressure longer than seven days remained in the forms until they were prepared for testing; those tested at seven days were placed in water until the time of the second test. The compressive strength of 6 6-in. cubes tested at 77 days was 2010 lb. per sq. in. Weight Comrress- Tnitial Pressure Duration of ive Modulus Ref. | Age at During of Concrete Strength of No. Test Setting Pressure Ib. per Ib. per Elasticity days Ib. per sq. in. days cu. ft. aq. in. Ib. per sq. in. 6501 7 0 0 142.5 725 1 500 000 501 80 0 Oo cig 1850* 2 370 000* 502 80 0 Oe seared 1820 3 100 000 503 80 6 Uo OD) lo plepate 2220 3 480 000 504 80 6 ann 2170 3 600 000 505 80 6 17 142.5 2220 3 820 000 506 7 20 1 148.5 1240 2 000 000 } 506 80 20 BO eee 2450* 3 800 000* t 507 80 20 Lt Ot. wethese 2350 3 880 000 508 7 20 7 148.0 1087 2 250 000 508 80 20 rr a (i9see 3 240 00%} 509 80 20 tI) state 2670 3 800 000 510 80 20 7 148.0 2850 3 520 000 511 7 100 1 148.2 1570 2 500 000 511 80 100 Te tein 3060* 4 400 000* 512 80 ed 1 147.0 ae 4 400 000 ele) @ | | Be | ae) aie 514 80 100 7 149.0 3140 4 000 000 515 80 100 77 148.0 3160 4 400 000 16° 80 143.0 1865 2 640 000 517° 80 145.0 2000 2 720 000 * Second test on the same specimen. ° Set under pressure of 3 Ib. per sq. in. for seven days; forms removed and cylinders placed in testing machine, under pressure as follows: 100 Ib. per sq. in. 7th to 14th day; 200 Ib. per sq. in. 14th to 21st day; 300 Ib. per aq. in. 21st to 28th day; 400 Ib. per sq. in. 28th to 35th day; 500 Ib. per sq. in. 25th to 80th day. The storage of these cylinders differed from the others in that they were exposed to the air after age of seven days. 120 ILLINOIS ENGINEERING EXPERIMENT STATION £800 2400 £200 2000 1600 1400 1200 1/000 600 600 COMPRESSIVE STRESS ~LB. PEP Sa./N nN 8 O Ll? L0Ob LOE LOE 0 OME 20D) LO0h L008 Lil0 Lethe L066 L016 00U8 2020 Da Unit DEFORMATION Fic. 46. STRESS-DEFORMATION CURVES FOR 8 BY 16-IN. CONCRETE CYLINDERS. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 12] cover plate wedged in the form in applying the pressure to the fresh concrete, a close fit was impracticable. Under a pressure of 100 lb. per sq. in. the length of the 16-in. cylinder was shortened nearly 1 in. and as much as a quart of water was forced out. The initial modulus of elasticity increased from 3100000 lb. per sq. in. for the cylinders without pressure to 4 300 000 lb. per sq. in. for those setting under 100 lb. per sq. in.; an increase of about 37%. As in the case of the compressive strength, over one-half of this increase occurred below a pressure of 20 lb. per sq. in. The cylinders which set for 7 days under a pressure of 3 lb. per sq. in. and were gradually loaded to 500 lb. per sq. in. (No. 516 and 517) gave about the same values as those setting without pressure. Apparently the difference in storage conditions nearly counteracted the effect due to pressure. The increase in compressive strength and modulus of elasticity with the pres- sure under which the concrete sets is more pronounced in the 7-day than in the 80-day tests. The density of the concrete as determined by the weight of the cylinders was increased about 4% by the pressure of 100 lb. per sq. in. 63. Hffect of Loads Reapplied after Fatlure in Compression.— An interesting instance of “autogenous healing” in concrete was found in the retest of five cylinders at the age of 80 days which had been loaded to their ultimate strength in compression at 7 days. One of the cylinders had set under no pressure; two set under 20 lb. per sq. in. and two under 100 lb. per sq. in. During the period from the 7th to the 80th day the cylinders were stored in water. The results of the first and second tests have been bracketed together in Table 23. Load- deformation curves for three of the cylinders are given in Fig. 46. In all the 7-day tests loading was continued until there was a distinct drop in the beam of the testing machine and the extensometer showed a rapid increase in deformation. The fact that the load-deformation curves are nearly horizontal and that a unit-deformation greater than 0.0012 in. was produced in all these cylinders show that they received their ultimate loads in the 7-day tests. Numerous vertical cracks and surface flaking could be seen on most of the cylinders. The compressive strength and modulus of elasticity for the 80-day tests averaged 93% and 64%, respectively, greater than for the 7-day tests of the same cylinders. The average compressive strength and modulus of elasticity for the retested cylinders are 91% and 92%, respectively, of the values for similar cylinders which were tested for the first time at 80 days. It is note- 122 ILLINOIS ENGINEERING EXPERIMENT STATION worthy that the cylinders which set under pressure behaved in much the same way as those setting under normal conditions. Apparently a compressive stress equal to the ultimate resistance of concrete applied as much as 7 days after mixing (as long as the speci- men is not entirely shattered) does not necessarily permanently destroy its usefulness, if this failure is followed by a period of rest in the pres- ence of sufficient water to permit hydraulicity to continue. Pressure sg 8 8 g nN 3° at Concrete Setting without 8 8 f jnitial modulus al bale 8 8 Percent Increase over Values for 0 O 10 20 30 40 50 60 70 G0 90 100 Pressure under which Concrete Set - Lb. per Sq In. Fic, 47.. Errect oF Pressure DurING SETTING ON THE PROPERTIES OF CONCRETE. 64. LHffect of Reinforcing the Pull-out Specimen against Bursting. —In many of the pull-out tests with deformed bars, the concrete blocks were reinforced against bursting by means of 6 or 7 turns of a 4-in. wire spiral placed inside the forms. It was desired to determine what effect this restraint had on the load-slip relation and on the maximum bond stresses developed. Several specimens made with corrugated square and corrugated round bars were reinforced as indicated above and others from the same batch were without reinforcement. The results of these tests are given in Table 24. In nearly every case the specimens rein- forced with the wire spiral gave higher bond stresses at all stages of the test than those without reinforcement. The difference is not very great for ages of 2 to 7 days. For the tests made at 28 days or over the bond resistance at a slip of 0.001 in. for the specimens with the spiral aver- aged about 15% higher than those not reinforced; at the maximum load the values are about 50% higher. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 123 TABLE 24. Errect or REINFORCING THE CONCRETE BLOCK AGAINST SPLITTING. Pull-out tests, embedment 8 in. All specimens were stored in damp sand. Stresses are given in pounds per square inch. Age | Num. Bond Stress at End Slip of— e ip of—Inches wee my ber of Remarks daya | Tests .0005{ .001 | .002 | .005 | .o1 | .02 | .05 | .o1 (a) 34-in. = Squares, 1-1-2 Hand-mixed Concrete (Batch 20). 1 2 2 86 96 102 121 136 159 | 207] 250 2 2 2 97 111 117 131 154 174 | 220] 236 3 4 2 145 157 174 | 208 | 234] 268] 315] 351 4 4 2 108 129 148 177 196 | 230 | 258 | 259 5 7 2 167 177 198 228 273 317 | 396 | 447 6 7 2 154 166 174 | 218 253 298 | 322} 322 7 28 2 309 329 343 377 | 422 507 683 827 8 28 2 258 | 286 | 316 393 | 480 | 527 ]...... 527 9 60 5 445 498 | 555 655 793 965 | 1101 | 1145 10 60 4 338 | 399 | 419 | 499 ]......]............. 771 11 [15 mo. 2 709 892 | 1015 | 1165 | 1340 | 1480 ]...... 1525 12 |15 mo. 2 645 778 888 layne vel ces Heese lis vas 1025 (b) 34-in. Corrugated Squares, 1-1-2 Hand-mixed Concrete (Batch 23). 13 2 1 Spiral ofarbefoustspesbrase- cunt dhacshatanietage, Ge 86 92 102 138 166 200 | 270] 312 14 2 1 In 8-in. steel pipe fcrm §. ‘ 91 114 119 140 166 195 271 326 15 2 2 Without spiral................ 93 105 112 129 142 180 | 211 214 17 4 1 Spirals acciiasicicw cateteinie uso 146 158 182 233 279 | 325 408 | 444 18 4 1 In steel pipe §... ~f 131 145 169 206 | 249] 300) 405 | 463 19 4 2 Without spiral 101 118 123 151 178 | 213 258 264 20 7 1 192 | 216] 246] 310] 375 450 | 533 555 21 7 1 148 171 190 | 243 | 300] 376 | 495] 562 22 7 2 150 159 177 194 210 | 214 | 242 | 274 23 39 1 Spittal sec sactacamasancaeeaaaes 390 | 4383 | 458 562 657 796 | 894 906 24 39 1 In steel pipe §.. 351 378 | 434 554 695 | 861 } 1075 } 1152 25 39 2 Without spiral 386 | 413 | 453 | 527] 579]......}...... 700 26 61 2 Spirals ps ace 3c cactssers oe jasarciaann 577 | 660 | 749 | 887 | 1019 | 1104 | 1134 | 1152 27 61 2 Without DITA sssccsiyai3, Sis duals 378 | 439} 489 607 OEE | sstiaciie 3a sphess 819 28 |5 mo. 2 Spiral ss ccssica vias ee ee seen 516 | 586 657 | 829 | 1023 | 1195 |...... 1329 29 |5mo. 2 Without Alta ices parses se gaceiee 474) 564 | 639] 770] 894] 1024 ]...... 1070 (c) 11%-in. Corrugated Round, 1-2-4 Hand-mixed Concrete. 30 76 5 Spiral sc v3 seven ei mates: 281 | 306] 328 | 374] 438] 535] 677] 846 31 76 1 Without spiral...............5 235 260-| 270 286 307 328 | 392 | 457 (d) 114-in. Corrugated Round, 1-2-4 Machine-mixed Concrete (1912). 244 | 279 | 314) 359] 412] 517{ 809 | 1115 347 | 404} 420] 438 | 516] 605] 855 | 1070 234 | 248) 258] 278} 312] 341 ]...... 375 32 80 2 Bpiral,..wnec es yee 33 80 2 ra 8-in. steel pipe Ee 34 80 2 Without spiral... § Blocks were molded in an 8-in. length of 8-in. steel water pipe which remained in place during the test. 124 ILLINOIS ENGINEERING EXPERIMENT STATION In order to study the effect of a greater amount of restraint against bursting, part of the specimens were molded and tested in forms consist- ing of 8-in. lengths of 8-in. steel water pipe. The specimens with solid pipe forms gave values about the same as those with the spiral, during the earlier stages of the tests, but as may have been expected, they gave much higher values of maximum bond resistance. These tests indicate that the restraint of the spiral reinforcement is effective in raising the bond resistance in pull-out tests with deformed bars. Fig. 45 gives load-slip curves for a few tests which show the effect of reinforcing the concrete block against bursting, the effect of causing the concrete to set under pressure and the effect on the corrugated bar tests of molding the specimen with the long end of the bar upward. B. REINFORCED CONCRETE BEAM TESTS. 65. Preliminary.—lt is in the design of reinforced concrete beams that a correct knowledge of the bond resistance between the concrete and the steel is most important. The study of reinforced concrete beams with special reference to bond stresses was begun at the University of Illinois in 1909; additional series of beam tests designed for the study of bond stresses were made in 1911 and 1912. The results of 110 beam tests are given in this bulletin. The dimensions of the beams were: width 8 in., depth 12 in. (10 inches to the center of the steel) and span length 5 to 10 ft. Typical forms of reinforced concrete beams are shown in Fig. 2. The longitudinal reinforcement usually consisted of a single straight bar of large diameter placed in the middle of the width of the beam. These bars extended to the ends of the beams. Vertical stirrups were frequently used as web reinforcement. With a few excep- tions which are noted in the tables, the beams were loaded at the one- third points of the span. The usual span length was 6 in. less than the total length of the beam; in some of the 1911 tests beams 714 and 814 ft. long were loaded on a span of 6 ft. with the ends overhanging the supports 9 in. or 15 in., instead of 3 in. as in the other beam tests. Bars of large diameter and beams of comparatively short span lengths were used in order to develop high bond stresses in comparison with the other stresses in the beam, and thus to produce bond failures. _ 66. Measurements of Slip of Bar—In the beam tests the slip of the ends of the reinforcing bar was measured by means of Ames gages as in the pull-out tests. The instrument was carried by a wooden or . ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 125 metal yoke which was attached to the ends of the beam in such a position that the movable head-piece of the instrument had a direct bearing against the end of the bar. As the test progressed, the amount of slip at each end of the bar was noted. In many of the beams of the series of 1911 and 1912 the slip of bar was measured at numerous points along the length. Openings about 1 in. in diameter were cut or formed in the concrete under the reinforcing bar at points where measurements of slip were desired. Plugs of 3¢-in. square steel were screwed firmly into threaded holes in the bar. The gage was carried by a metal bracket attached to a steel plate which was fastened to the concrete by plaster of paris in two places on either side of the plug. The plunger of the gage rested against the steel plug so that a movement of the reinforcing bar in either direction with respect to the concrete would be indicated by the pointer. As many as 13 such instruments were used in some of the tests. The positions of these instruments are shown for typical tests in Fig. 56. The amount of slip which has been termed “first slip of bar” in the tables corresponds to a movement of about 0.0002 in. We are justified in using a smaller quantity as a measure of first end slip in beams than was used in the pull-out tests, for the following reasons: (1) Measurements from the yoke to points on the concrete of the end of the beam near the bar do not show appreciable deformation in the concrete at any stage of the test. This is probably due to the unstressed condition of the concrete at the end of a beam. (2) The load-slip curves for the beam tests show that as soon as the slip at the end of the bar reaches about 0.0002 in., it rapidly increases with a further applica- tion of load. (3) Tests made on beams by allowing the load which pro- duced a very small amount of slip at the end of the bar to remain con- stant for several hours or days indicate that the ultimate bond resistance probably would be developed under the indefinitely continued application of loads which produce at first an end slip as small at 0.0002 in. The tests in which slip of bar was measured at intermediate points are dis- cussed in detail in Art. 94. 67.. Computation of Stresses in a Reinforced Concrete Beam.— An analysis of the stresses in a reinforced concrete beam was given in Bulletin No. 4 of the University of Illinois Engineering Experiment Station, “Tests of Reinforced Concrete Beams,” by A. N. Talbot, and reference may be made to that bulletin for methods of computing the stresses given in the following tables. In that analysis it was shown 126 ILLINOIS ENGINEERING EXPERIMENT STATION that for the assumptions of beam theory the bond between concrete and steel in a reinforced concrete beam is a function of the vertical shear. The bond unit-stress was expressed by ”v mod Where v = the total vertical shear, m = the number of bars, o = the perimeter of one bar, d’ = the effective depth of the beam, or the distance between the center of the steel and the centroid of the com- pressive stresses. The formula assumes that the reinforcing bar ends at the point of support. As in the test beams the bar extended beyond the point of support 3 in. or more, the amount of available bond surface will be greater than that assumed in the analysis, but generally in the calcula- tions this additional bond surface will be disregarded and the formula used as given above without modification. It is recognized that this analysis does not consider all the phenom- ena of bond action. It will be seen below that after slip of bar becomes appreciable, the bond stress in a beam is not distributed as indicated by the formula. However, instead of attempting to take into account the effect of slip of bar-and the cracking of the concrete, the values com- puted by the above formula will be used. These nominal values of bond resistance form a useful basis of comparison in a series of tests in which the dimensions and make-up of the beams are similar. 68. Phenomena of Beam Tests—If a bar which is embedded in a prism of concrete is subjected to a pull at its ends a part of the tensile force is transmitted to the concrete and a bond stress is developed be- tween the steel and the concrete which may be said to be due to the stiffness of the concrete or its resistance to stretch. When the concrete has been distended to its limit of stretch or to its limit of tensile strength, a break occurs and the concrete adjoining the break springs back toward the unbroken concrete, there is a slipping along the bar, and a minute crack forms at the break. The size of the crack and the amount of the slip will depend upon the relative dimensions of concrete and steel. It is evident that some tensile stress will remain in the unbroken concrete and some bond stress between concrete and steel even after slip has occurred, and that with increased tension in the bar there will be further ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 127 stretch in the concrete or further slip of the bar or more cracks in the concrete. For want of a better one, the term “anti-stretch slip’ may be used for this slip of bar which is due to the stiffness or springiness of the concrete after cracking. At what strain this anti-stretch slip occurs and how far apart the cracks will be may be expected to be a function of the size and periphery of bar and the condition of its sur- face and the quality of the concrete and the section of concrete which is tributary to the reinforcing bar. It is well here to call attention to the fact that a tension break in the concrete may occur without its presence being visible to the eye. The stretched concrete to the side of the break will act as a spring and tend to pull the concrete back from the break. If this elastic force is greater than the bond résistance slip will occur and the crack will widen and become perceptible. Such cracks are very fine. The white- washing of the surface of the beams materially assisted in detecting these cracks at an early stage of their development, and doubtless many of them would not have been seen on the natural surface of the concrete. It may be added that the term “anti-stretch slip” has been used some- times in the bulletin to cover the elastic force which goes with the slip. In the case of a beam which is loaded symmetrically at the one- third points, the region between the two loads carries no vertical shear and there is no bond stress due to normal beam action as it is usually analyzed. However, this is the region of greatest longitudinal stress and hence the region in which a phenomenon similar to that just de- scribed will exist to a considerable degree. At the beginning of load- ing, the concrete itself will carry the greater part of the tensile stress and a part will be transmitted to the reinforcing bar. In the later stages, after cracks have formed, it may be simpler to consider the ten- sion to be carried primarily by the bar and the concrete to form the web and the enveloping medium. At the formation of tension cracks, the phenomenon of anti-stretch slip described in the preceding paragraph will exist. The reinforced concrete beam tests to be described indicate that a slip of bar took place in this region at loads below those at which minute cracks on the whitewashed face of the beam appeared, and con- siderably below the loads at which they have been noted in beams having only the usual concrete surface for inspection. In general, this early interior slip of bar was noted at a tensile stress in the bar of about 6000 lb. per sq. in. The distance apart of these cracks and their rela- tion to the form and size of bar will be discussed with the details of the tests. 128 ILLINOIS ENGINEERING EXPERIMENT STATION For the region outside the load points, that is, in the outer thirds of the span length, for the method of loading generally used, shearing and other web stresses exist, and due to the beam action bond stresses between the concrete and steel are developed which may be termed beam bond stresses. By the usual analysis, for the loading used, these bond stresses are nominally uniform from load point to support. The anti- stretch slip may be expected to exist here also, especially in the part near the load points, and the existence of this slip and bond stress concurrently with the beam bond stress makes a considerable complica- tion and may be expected greatly to modify the distribution of the bond stresses over the length of the bar, and to affect resistance to beam bond stresses. In the tests of beams loaded at the one-third points, the beam exhibits considerable stiffness up to the load which causes the first ten- sion cracks in the concrete. After this point, the beam deflection increases more rapidly and continues at a nearly constant rate until failure is imminent. A marked change in the rate of deflection follows the diminution in effective tensile resistance of the concrete and is coin- cident with the stage when anti-stretch slip first develops prominently in the middle third of the beam. Due to the inequalities in the tensile resistance of the concrete and to the development of anti-stretch slip the tension cracks form at certain points instead of being closely spaced, and a large part of the increase in length of the lower fibers is concen- trated at these cracks. An examination of the photographs and sketches of the beams after failure will show that tension cracks form through the region of the middle third and generally a short distance outside with usually other tension cracks farther out. For the stage of the test at which cracks first appeared at the load points or a short distance outside, the bar showed a measurable slip just beyond the crack (nearer the end). It was necessary to increase the load 50% to 300% before slip was produced at the end of the bar; the amount of increase varied principally with the distance between the load point and the support. We may expect then that a bond stress nearly as great as the ultimate bond resistance was being developed for a short distance beyond the crack, much of which was due to the condition which produces anti-stretch slip of bar. At this stage of the test the bond stress at the supports was only a small part of the maxi- mum bond resistance. The tests indicate that when the maximum bond stress was first developed outside the load points the bond stress at ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 129 the support was not more than, say, 15% to 40% of the maximum bond resistance. As the bar was further stressed by an additional load applied to the beam, the bond stress near the crack decreased on account of excessive slip of bar and the region of full bond resistance was thrown more and more toward the support. Later in the test, due to the com- bination of beam bond stress and anti-stretch slip, another crack was formed a few inches nearer the support. The opening of a second crack had the effect of reducing the bond stress between the cracks and the tensile stress in the bar was increased toward the support. As the load- ing progressed this process continued with the piecemeal development of the maximum bond resistance and the subsequent reduction of bond due to excessive slip over the portion of the bar affected, until the ef- fective embedded length of the bar was no longer able to withstand the bond stresses developed and failure from slip of bar soon followed. It is clear that the unbroken embedded length of bar at the ends of the beam which takes the principal portion of the total bond stress during the last stages of the test, will depend upon the relative dimensions of the bars and the beam and upon the bond, tensile and shearing resist- ance of the concrete. The beam tests discussed below show how this distance varied for a variety of conditions. Fig. 57 to 63 show the appearance of representative beams after testing. The surfaces of the beams had been whitewashed before the tests in order to facilitate the tracing of cracks. The positions of the load-points and supports are shown by vertical arrows. The surface cracks, which were generally very fine, were traced with black paint on the surface of the beams in order that the location of the cracks may be shown on the photographs. The numbers near the cracks indicate the extension of the cracks as the test progressed expressed in thousands of pounds load. ‘In the beams in which slip of bar was measured at intermediate points, the positions of the instruments are indicated by the numbers inside circles; the points at which tensile deformations in the longitudinal bar were measured in the tests of certain beams are indicated in the same way. The load-slip curves for representative beams in each group are given in Fig. 69 to 76, inclusive; load-deflection and end-slip-of-bar curves are given in Fig. 77 to 86, inclusive. These diagrams give impor- tant indications as to the effect of the different variables in the make-up or loading of the beams. 130 ILLINOIS ENGINEERING EXPERIMENT STATION a. 1909 Beam Tests. 69. Outline of Series—In 1909 eleven reinforced concrete beams were tested with special reference to a study of bond stresses. Data of the beams and tests are given in Tables 25 and 26. These beams were considered as a preliminary series; no companion pull-out specimens were made. TABLE 25. Data oF REINFORCED CONCRETE BEAMS—1909 SERIEs. 1-2-4 hand-mixed concrete; Chicago AA portland cement. Each beam was reinforced with a single longitudinal bar. All beams were 8in. wide; total depth 12in.; depth to center of steel 10 in.; length 614ft. Longitudinal Reinforcement Compression Testa 1 eB himipd Mixture Beata a Beam (Round bars y rom ge ‘verage Per . ‘i Net Kind of Bar oon | SEA ME | CONE | eee | tee | he days sq. in. 120 1-in. plain round... .-| 0.98 ¥ in. 1-2.35-4.15 121 82 1420 121 1-in. plain round. . .-| 0.98 5 in. 1-2.35-4.15 120 82 1420 201 1-in. plain round.... 0.98 % in. 1-2.53-4.40 202 62 1757 84 | 134-in. plain round........... 1.53 5% in. 1-2.34-4.27 85 65 1722 208 1}4-in. plain round.... .-| 1.58 ¥ in. 1-2.52-4.47 204 63 1573 204 14-in. plain round........... 1.53 ¥ in. 1-2.52-4.47 203 63 1573 M7 Lin. CUPS cccenccnies ee se sees 1.25 5% in. 1-2.23-3.93 118 87 1960 62 114-in. corrugated round...... 1.25 5 in. 1-2,44-4.23 te 65 1762 202 1}4-in, corrugated round......| 1.25 %% in. 1-2.53-4.40 201 62 1757 85 34-in. corrugated square...... 0.70 &% in. 1-2.34-4.27 84 65 1722 118 1-in. twisted square.......... 1.25 5 in. 1-3:23-3.93 117 87 1960 The average compressive strength of 6 sets of 6-in. cubes waa 1700 Ib. per sq. in. All the beams were identical as to materials and external dimen- sions. 1-2-4 hand-mixed concrete made with Chicago AA cement was used. Each beam was reinforced with a single bar which extended flush with the ends of the beam. The size and spacing of vertical stirrups is given in Table 25. The age at test averaged about 100 days. All beams were loaded at the one-third points of a 6-ft. span. In all but one test (Beam No. 84) the load was applied progressively to failure. In the test of Beam No. 84, load was applied until one end of the bar showed an appreciable slip; the load was then released and reapplied. This test is discussed in detail in Art. 90. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 131 70. Bond with Plain Round Bars.—End slip began in the beams reinforced with a 1-in. round bar at an average bond stress of 192 lb. per sq. in.; and for the 114-in. round at 211 Ib. per sq. in. The average maximum bond resistance was 279 lb. per sq. in. for the 1-in. bars and 303 lb. per sq. in. for the 144-in. bars. The load-deflection and load- slip curves for Beams 121 and 204 are given in Fig. 77. TABLE 26. Tests or REINFORCED CoNCRETE Brams—1909 SzERizs. All beams were loaded at the one-third points of a 6-ft. span; overhang 3 in. at each end. Loads are given in pounds; stresses in pounds per square inch. In computing unit stresses, the weight of the beam was considered. At First Slip s ice Load of End of Bar® At Maximum Load Beam| ,at First Com- Tensile |Vertical] Com- Failure of Beam No. | Test | Outer | Applied | puted Applied | Stress | Shear- | puted days Crack Load Bond ad in ing | Bond Stress Steel Stress | Stress 120 112 11 000 10 000 198 11 700 22 100 89 230 | Slow bond failure at S. end 121 112 10 000 10 000 198 14 500 27 100 111 281 Bond at S. end. 201 88 10 000 10 000 180 16 900 31 400 128 327 | Bond at S. end Av. 104 10 300 10 000 192 14 370 26 870 109 279 84" | 104 17 000 15 000 243 20 500 24 900 158 327 | Bond at S. end. 203 91 13 000 11 200 185 19 300 23 500 149 309 | Bond at 8. end. 204 90 13 000 12 500 204 17 000 20 800 132 273 | Bond at 8. end. Av. 95 14 300 12 900 ait 18 900 23 070 146 303 117 118 19 000 19 000 292 33 000 48 100 250 500 | Bond at 8. end. 62 56 12000 | .. ... xe 22 500 33 200 172 390 Hisgena’ tention and bond 202 88 14 000 12 000 213 19 000 28 100 146 332 | Bond at §. end. Av. | 72 | 13000 |.......f..-| 20750 | 30650 | 159 | 361 85 105 | ..... 6 000 128 28 700 72 500 209 568 | Tension in steel. 118 118 20 000 15 000 233 32 100 46 700 241 488 J Bond at S. end. ° Corresponding to a slip of 0.0002 in. * The load was released and reapplied. See Art. 90 and Fig. 51. 1. Bond with Deformed Bars.—One beam in the 1909 series was reinforced with a 1-in. cup bar, one with a 34-in. corrugated square bar (type B) and two with 114-in. corrugated rounds. Beam No. 118 reinforced with a 1-in. twisted square bar will be referred to in the discussion of other beams of the same kind in the 1912 series (see Art. 84). 132 ILLINOIS ENGINEERING EXPERIMENT STATION Beam No. 117, reinforced with one 1-in. cup bar, carried the highest load in this series—33 000 lb. End slip began at an applied load of 19 000 lb. and a computed bond stress of 292 Ib. per sq. in. This is about the same load that caused the first diagonal crack. The maximum bond resistance was 500 lb. per sq. in., corresponding to a steel stress of 48100 lb. per sq. in. The load-end-slip and load-deflection curves for this test are given in Fig. 77. The slip measured at the end where failure occurred was 0.006 in. at the maximum load; the slip at the opposite end was 0.001 in. Beam No. 85, reinforced with one 34-in. corrugated square bar (type B) showed first end slip at a computed bond stress of 128 lb. per sq. in. The load-slip curve for this beam (Fig. 77) shows a very rapid slip at one end following an applied load of about 7000 lb. The end slip had reached 0.001 in. at a load of 8000 Ib.; at the maximum load the slip was about 0.02 in. at one end and 0.004 in. at the other. While this beam failed finally by tension in the steel, it is plain that a bond resistance much higher than the value given—568 lb. per sq. in.—could not have been developed. ; Beams No. 62 and 202 were each reinforced with one 114-in. cor- rugated round bar. Slip of bar was not measured in Beam No. 62. In Beam No. 202 end slip began at a load of 12 000 lb. and bond stress of 213 lb. per sq. in. After a load of 12 000 Ib., slipping at both ends was very pronounced. At failure, at a load of 19 000 lb., bond stress of 332 lb. per sq. in., one end of the bar had slipped 0.014 in.; the other 0.007 in. The bond stress at the beginning of end slip of Beam No. 202 was about 5% higher than the average of the six beams in this series which were reinforced with plain rounds. The average bond stress at the maximum load for the two beams reinforced with 11¢-in. corrugated rounds (361 lb. per sq. in.) was 24% higher than the average of the six beams with plain rounds. However, the beams with corrugated bars were tested at a somewhat earlier age; on the other hand the cube tests show them to be made of concrete of somewhat higher compressive strength than those with plain bars. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 133 b. 1911 Beam Tests, 7. Outline of Series—Thirty-six reinforced concrete beams were included in the 1911 series. Plain round and corrugated bars were used for longitudinal reinforcement. All the beams, except two, were tested with a 6-ft. span. Beam No. 1045.2 and 1045.3 were tested with an 8-ft. span. Several of the beams in this series were tested with ends over- hanging the supports 9 in. or 15 in. In one group, the depth of con- crete below the steel was varied. Fifteen of the beams had no web rein- forcement; the remainder were provided with V-shaped stirrups of ¥Y4-in. or ¥-in. plain rounds. In the 1911 beams the concrete was hand-mixed; the age at test averaged about 8 months. Pull-out specimens and 6-in. cubes were made from the same materials as were used in many of the beams. The pull-out specimens were stored in the open air with the beams, until tested; the cubes were stored in damp sand. Details of the make-up of the beams, their dimensions, the mate- rials used and the strength of the concrete will be found in Table 27. Table 28 gives the data of the tests and some of the calculated stresses in the beams as well as notes on failure. The computed bond ‘stresses developed in the beams and in the pull-out tests in this series at various amounts of end slip will be found in Table 29. A summary of the bond stresses in the beam and pull-out tests is given in Table 30. In several of the beams in this series measurements were made on slip of bar at points other than the ends. Discussion of the slip found at internal points will be given with the discussion of similar tests in the 1912 series (Art. 94). 73. Basis of Comparison of Bond Resistance in Beam and Pull- out Tests.—In the 1911 and 1912 beam series, the corresponding beams and pull-out specimens were made from the same batch and stored under the same conditions. In comparing the bond resistance in beam and pull-out tests, it is evident that the bond stresses corresponding to definite amounts of slip in each case should be considered. The amount of slip in both kinds of specimens was measured at the free end of the bar, and at other points in some of the beam tests. Differences in bond resistance developed in the two forms of test specimens may be expected to be due principally to differences in the secondary stresses in the ILLINOIS ENGINEERING EXPERIMENT STATION 134 ut ‘be sad ‘qi 1662 ‘yaeares qaqa] wo 8408 g ‘ut ‘bs Jad “q] OOTE ‘JaauIe0 [esIOATU() Wor 8499 [] (9&1 abpd uo panuyuos Lg 91901) “UI-OT [238 Jo 209009 04 Gydep ‘ar FT GIdep {0,4 "I OT 19048 jo 29}009 OF UAdap tur TT Aap [eIOL 4 SBMOT[O} SB BEA HYIWOU g yNOGE 48 po}8e} GAqNd ‘UI-g Jo 4}FUAIIS aBvIOAT ayy, , Tossed uI09, eis i 2LL° 8-80 8-1 yamet of au0N 86'0 ured “at-T on ol FPOL 119% 8 19° €-80' 2-1 qatyeT £°ShOT aon 86°0 meld “m-T 09 FOL 9 L998 L OF £-G0° 2-1 [esz3Aray) TFF01 auoN 86°0 wield “m-] 0 oS F501 4 8 99° 8-L0°S-T qatqeT 8° 2F01 au0N 86'0 ureyd “ar-] uO +8 FFOT 0262 KE 89° £-80°2-1 qarqaT 2° SFOI auoN 86°0 eid “wi-T 49 42° FP0I LSTE L OF E80 2-1 yeexeara) 3° 501 au0N 86°0 ured “a-] 9 ol FFOT $ eesz “8 BL EST ST yaiqe'T £°OF0T m1, 860 med “aI-] 9 ©2801 L¥98 L 99° E21 2-1 [ssuaaray) € L801 mH 86°0 wed “Ul-T 9 £°Te0r + ozze SAL 19° S811 yeeroatay) T'SF0I “wh 860 ureyd “ar-T 49 2 °2e01 £822 6 29° €-01'S-1 [eexoaray) T2801 ur 86°0 meld “ary 9 1'Ze0r 208% 8 BL 8-08 1 [esrastay) &°S€0r “W345 86'0 ured “aL-] 9 @'10t g SLE 6 LOETL' U1 jeszeatay) T'¥Ee01 my 86'0 uwrepd “at-] Ho T TOL LI L 29°€-90' 2-1 qarqeT £" S801 “Wr 86°0 areyd “mn-T 9 £°6€01 L168 K8 9F'£-80' 2-1 qeexoaTu) 2 Le0l “WK 86'0 ured “aI-] 49 2° 6801 t 019% 6 £0° 8-20 U-1 yeexearay) T8401 “wr 86°0 ued “ut-T 0) T6€0T ££9% ws ZL ESOT qaqa £° 2801 au0N 88°0 ureyd -ut-] uo £° OFT L¥9E L 99° E21 2-1 jesvaaray) 3° eh01 au0N 86°0 wreyd “-T %9 £° L801 LL6E Ks OF £-80' 3-1 Jwexaatay) 36801 au0N 86'0 areyd “ur-] a) 2° LE01 I SEL% 6 19° 8-01 2-1 Teeuaatay) T°Ze0L anon 86'0 med “uy 549 T2801 ‘at ‘bs sed “qi our'ysaJ, (qzede ‘ul g quae near nanOseie meal, ¢ jo adeleay | ye o3y 3431044 Aq quauie, qoeg oures og a panow) reg ‘qf panoy jo pary 4225 ‘on wsag lanorp oS3QNY “UI-g Jo 83897, oIngyy wou WeIg pamesiomey pempanzooy q33ue'] “ULQT ‘12938 Jo Jayu39 07 YIdep ‘pazou astw9q}0 ssarun “ur ZT YIdep [8104 ‘epLA “UI g alam suIvaq [TV “SHINES [TI6I—SNVEG WLAYONO’) AHOUOANIAY AO VLVCG ‘22 ATAVL *]20}8 UOqI¥O-YSIY JO 1eq o]FUls B JO poysisuod MBA TOUS Ul JUsUTGOIJOJUIEI [BUIPNyTUO] ey, *9}910U00 paxIul-pusy F-Z-T 135 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL ‘O]Pplur Jo apis Gove ‘ur g syurcd ye papeoy » (LET ebod uo ponurjU0s 9g 2190,],) "8G “BLT pUe OG “HV 296 ‘yno par[nd eq oq} [HUN “q] 000 9Z Ie PoULEIMIEUT SEA E'GEOT “ON Weag UO peo] aU, y ggg 61Z oores | ooeez | eee 0022 | oooot | 8 “pas "N 48 puog £8 08% 009 | oo20e | £0F ooore f wT" fob “pus “N 38 puog #98 eG 00g #8 | 00262 | 85% o00%2 | oooor | 2 ‘uo N18 puog | T2e aes 90% ooros | OorZ2 | 6&F oooez | "| OF 1g8 THI oozte | osset | #22 ooe#T | oonoT | 2 “pus *g 48 puog #98 98l oosze | osgzt | #82 ooozt | ooog | 2 ‘pus N48 puog |" ore S81 oosze | O0Zt | 222 000%! | ooogt | 2 ‘pua "N98 puog | Ie 888 asl oos9e | 0000 | Lze 000zT | oooet | Z BLE 61 oorse | oozer | 26r oooot | cose {sz “pus "S98 puog | Eze *868 Sst oot ze | oosoz | 26r oooot | ooog | 2 “pue “g 98 puog Aiquqorg |" 2898 eH oose® |, 0068T | 26r oooot | ooort | 8 829 Tee opr oorze | oozzt | Tze oozit | 0068 | 8 ‘pus "g 98 puog | 249 IF 201 gosee | oostz | 62 ooost | onset | £ ‘pug 98 puog |” 08% SIT 00822 | ooost | Z6r oo00T | 0009 | 8 ‘Berp Aq pomopjoy puog | 999 218 Sat OOF 0 | OOF OT 621 000 6 000 4 8 "PH “N48 Yo pond seq ‘puog | ALF 208 el oos6z | oosst | Iz oooTt | 0008 | 8 1s 8S 6Ll ooser | oozes | 262 ooest | ocoot | 8 pus ‘g 38 puog | 189 +965 961 009 2% | 40009% |. 88 0002 | covet |z ‘pus ‘G48 puog | 899 628 60% oogos | 00822 | 222 000+t | ooog | 8 ‘pus g 38 puog | JSF wee 2el oooze | O0eZT | sez o00zt | oovoT | 8 OLF ace OFT ooose | osrst | 6ez oozzr | cosor | 8 ‘pue "N48 “ou9} “Bip pus puog | gue 298 a o0gse | 06st | 98% ooozt | oooor | 2 eee ae «PUR 'g 48 OmTITYY pug | **" ory zor ogee | ooriz | T62 00st | oooor [3 ‘paa “g ‘puog Aq emjrey wappng | ggg gge ig ooze | O0¢8T | 6c 0008 | oooor | 8 Puoq 4q pamopoy worsey “seq | = *** 162 aI 00822 | ooost | zzz 00st | 00g | 8 say, | mong | sang | jeag ssanig rat akteig | jena | gmat’ | PMY |pofnidog| PHY | EmB | oe wivag jo ompreg ae Paynduroy | peory18 A sua I, Poy 0 Jane pr puog qeg jo pw 48 peoy | a3 aMatyy peor UNUTZEYy FY jodiesenny : Oe + meDerce 0D 6D 69 OD 1 pug qoea 48 sueq -10AQ uedg 98a], - -aBezaay 9 L°¥¥01 549 9° FF0I 9 io o°FFOI cree) -98er2ay 59 £ FFOT 9 2 FPOI g 9 T¥F01 sone --eBeieay 9 of 2801 ¥ So | 8° Te0r cores) -@8gleAy at) 8° 801 549 T°2e0T KO 'Te01 g 549 T1800 oo) @eraAy S49 | +8°6£0I KO "6801 t 9 T6€01 vests) -98eIaAy 9 © OFO 49 €°LE01 K9 &" LEOT 1 m9 T° L801 4993 “ON qysueT weg dnoiy “ParepIsuco sea Urea 9} JO F4SI0M ay} Sassorys y1UN SuyNduIod uy ‘your eenbs aed spunod ut wears are sosserys ‘spunod ut uaald oie speoT “pajou estmseyjo ssaqun ‘ueds ay} Jo sjurod piryy-auo 34} 3e perdde o1am speoy oy, ‘SHINES [IGI—SNVEG DLAYONOD AGOUMOANIDY 4O SLsay, ‘86 WTAVL ILLINOIS ENGINEERING EXPERIMENT STATION 136 sar ‘bs ad ‘q] 266g “yaeUIEd qArYe] WOT] 948 9 fur ‘bs sad “q] OOTE ‘3U9uIE0 [BBIBATUL) WOIy 8498 [] :SMO[[O} 8B SVM STRUOUT g YNOge 4e payse} saqnd “uI-9 JO 49308148 aBe10Ae OUT, , ey 0082 8 99° €-L0°S-T qsreT &°¥POT “ur-34 ¥o'1 paqedns0o “WI-$4 7 HL £°Gr01 LO8E 8 2G SEL ST yesvearay) £ "F801 auON boil payedn109 “Ul-HT AL 3° OF0T el 029% 8 GSB" S-8L ST Teszearuy) @ ¥EOT auoN ¥0'T payesn409 “UI-$4T HAL TOF01 LOBE 8 BG SET T yeasearay), 2° OF0T “1-34 ¥o'1 payesns109 “U1-$4T a9 "e0r £02 8 GB SEL ST yeesearay) 1/0601 hie! 521 pagednsi09 “ul-$4T 9 2" FE0T at elie 6 19° STL UT [sazeaTaQ), 1 T80t “WH ¥o'T payesnisod “Ut-$4 1 Ho T'¥e0r L1G 8 1S" £-80° 2-1 qareT 9° FF0T “Or eo'T ured “ut-%4T 48 e°SFOL 026 8 89° £-80' ST argo 3 ¥F01 wr est uyefd “UI-AT “8 3° SPOT 11 022 HL TG" SET O-T [sar9aTaQ) 2°2e0l “wh est med “WI-F4T 8 T'S¥OT LELZ 48 89° SE ST qatqeT & @F01 amon es'T aed “Ul-F4T w8 £°SF01 or Ly L 29° 8-90°O-T aaa £° 6801 isa: es'T ured “WI-¥4T v9 €"se0r 2082 8 ZL €-00' OT yesxeatay) 2 1801 “urd, eo'T ured “Ut-F4T “49 "sor 6 £182 6 £9° SUT ST yeereatay) T-2P0r “a4 eo ayer “UAT “9 stor L¥OE L 99° E-21' ST [estat] 8° 180l auON 86°0 wed ‘UI-T “8 a" SPor 019% 6 £9° 820° CT leszaatay) T6801 amon 86°0 wield “wl-T HB T'501 8 LEZ "8 Bo SET ST qaryeT £01 a, 860 med “ar-T vb G'avor S182 6 £9 E-Z1' ST yesroaray) 1 Se01 “m-% 86°0 areyd “Ut-T AL T@F0r L “ut “bs zed * *OUL ‘489, : queo 87682], € JO ieaay 8 aay quya1aq £q qoveg aures (qaede ‘ur 9 19q seg punoy Jo Pury 993 | -on meog (a : aang xt queue) Woy wag ae punoy) W308T 5 eqD oe a : nays quawmadioyuley [eUrpnyzu0'T “UI-QT ‘12098 Jo 4ay090 0} Y3dep ‘pojou estmsaqyO ssafUN “Ul ZT yydop 1840} ‘epIM ‘Ul g o19M suTeeq [TV “9938 UOQIBd-YsIY Jo req a[SuUIs ¥ Jo poysisMoo UrBAq YOeS ul yuauladr10sUter [SUIPNZUo] ey], “9301900 paxtu-pusy F-Z-T ‘SHIUGY TI6I—SWVEG BLAYONOD CHOUOANIGY 40 VLVG ‘pl ADVd WOUT AHANILNOD—1Z ATAVL 137 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL J &6F LIZ 002 IF 009 8% L8¢ O0F 2% OOTIT (42 (Po uoreee sp eneraay “12098 OT TorsuaL, ok 6s¢ 9F% O0E LF 00S cE £8F 000 8 000 OT L 6 9 KL £° eror “uorsua} [euoser IL? 8lh O12 00S OF OOL 23 POE 000 12 000 OT 8 6 9 KL &'OF01 sl “puog Aq paaorjoy “sue “ser bPPl ThF ¥61 008 4é 00S $2 PIE 000 81 00F &1 Z 6 9 AL TOF0T re 867 612 O0€ 2% 006 82 92% 008 21 008 IT ET (abiiantadh jiaseda) ine ) eaereay “8pud 470q 4B “su9} “Zep pue puog TLP LLg TSS 000 6% 009 £8 621 000 01 000 OT 6 g 9 “9 £° yeor “pue “N 48 puog PPPL ceP O61 009 98 000 $@ rd 00S I 000 OT ZL s 9 49 & Peor at “pue "N 48 puog PLEL 98h FIZ 002 1% 00 82 SFE 000 #1 000 #1 8 g 9 9 T¥80r coe LLE ¥8I 00% 8 00L &% LZ 000 9T 0008 §=fHOT | Pp op etary “pue "N 38 puog im PLE egt 002 88 009 2% LOZ 000 FT 000 8 or g 8 at) £° SFO “pua "N 98 puog Th 08e 98T 009 8¢ 006 &@ 882 000 8T 000 8 iW g 8 8 2 S#OT Il it SOF 926 00F SE OSF 62 16 000 $2 000 21 Ge fee Pe coe) -@8eaay “puog £q pamor[oy “893 “Berd oFS ol 123 009 &F O0F 9€ 444 000 8% 000 21 L ST 9 Ks T'St0T “molema} jeuoReIq “i. 8ge PLT 002 12 00S 2% TS€ 000 2 0002I = |%01 ST 9 “Es &° Sh01 or Ole SIP 102 00F TE OOT 92 902 002 31 OOLOT 32 Jo pp cr] eseray “puog Aq paaorfoy “sua4 “Berg ge¢ 86¢ 16 00S SF 000 8 922 000 #1 000 21 L € 9 9 £° 9801 “pa "g 38 puog O8F 9ee 891 00S 9 OOT 12 sot 000 OT 000 8 8 g 9 9 3° S80T 6 “pua "N 38 puog S68 20 6FT 008 & 002 61 92% 000 FT 000 21 8 g 9 a9 T'Se0t os SLY 281 009 SP 006 42 OFF 000 008 @1 Bop ppc eseray “pus “g 48 “8u9} “ZeIq ae 864 161 000 8% 002 92 967 000 92 00001 |%8 ST 9 Ks & SPOT “puog Aq pasorjoy “8u93 “Beg Loh 19% BLT 002 & 009 & ¥8E 000 02 009 $T 8 oT 9 “8 T Shor 8 ae 00s 161 000 8% | 09292 618 oszer | oose [36 ne feteice voos-|--ogmoay “pua “g 38 puog ve TIS 206 002 6% 006 9% Soe 000 61 000 Tt IL 6 9 AL f° P01 “pua “g 48 puog SLE 887 861 008 9F 009 S¢ F6E 00S 02 000 8 8 6 9 HL 4 70) | L 8389], S891} 889195 124g psoT 880145 peo “at qnoqng | puog | Smwayg | ur esang | poyddy | paog | payddy | yowg | com | pad | gaa meeg Jo mpeg w payndurog} jsonz,A | 9ytsuay, payndurog, yng | 489, | yous |e 3095 “ON aoa reg Jo pug sd ee ee | | ee | aay puog peoy umanxeyy 4V yo dys 98114 IV 48 peoT | ady | dueq iL wnUONxe yy : : “1940, *Ppasapisuoo SEM UIvEq ay} JO FY SIeM 94} Sosserys 41UN SuIyndw0s uy = ‘yout eivnbs sed spunod ut uaats ov sassolys ‘spunod ut WeAIS 918 SpBo'T ‘ueds 94} Jo squiod pirqj-au0 04} 48% paydds o10M sp¥oy egy, ‘SHIUDG [I6[—SWVEG BLAYONOD GHOMOANIGY dO SLSAT, "GET Hvd WOU CHANILNOD—8Z WIAVL ILLINOIS ENGINEERING EXPERIMENT STATION 138 ey need meee mop " uso) use { eo [ 809 { 18h [ eee | 8 ff ee voor) amesaay Pe Bee ee sini! eso | ose | Teo | oss | szb | sor | 2 ued ‘uy | 2 bPOT oe ee as gos | 19g | zg | seo | ozo | gob | 2 ured “uit | 9° FFOT 9 ope £08 | Fz tet | 9 zs | 21g | 98h | O9F | SH | 6eF | OF ured -u-y | ¢°FF0T yee i eak eneclll see mei Wics feos |(sees Cowen pees | ome | ae foe vil adossayy Se oe se me s FPS PPE OFE ogee STs 88% L vo mre Oly £°PP01 cecal tas ae a ies ge | “° | °° | ope | ze | we | 2 cuted ult | 2° FPOT ¢ 18 Ors PLS T6T 9 888 zt, e88 98 L¥E 00€ L ured “a-y TV Pr0r Bae est Bie IN ae Pave dog aie nol gse ll gee) anges | der eae. || Shs seinen oats pean eze | goe | ege | see | sez | 92 | ceo | FOr | 2 eee | gee | we | ase | ez | zor | 2 coreeesompd a-T | §°Z80T ; ele | soe | e9e | 86 | 182 | FIG | fee | Yor eee Ree | eee aoe eae haere resumen apy | BELGE sz¢ | cre | eg | mtg | she | Tee | gee | 092 |%2 tee | «*° | e62 | goa | 96¢ | ate | g iP | eBereay zs | org | ong | 629 | 58% | 98h | ze | 90E | 6 iy | zor | ese | see | gze | res | 2 ‘med “wi-T | 2° Ze0T mee. mame ae Oa Nes eosin, Yee eet 062 | ose | ecz | Tez | ov | zor | 8 ‘ured “ur-T | T°Ze0T 909 | qo¢ | g9¢ | geo | som | gee | re | OI | 8 ue | °° | tz | gs2 | sez | ost | 8 ‘ured “UT |Z" TS0T & uy | cop | ozy | 19m | ee | GE | 682 | SEZ | g zoe | ose | scz | sez | Oo | ost | 8 ‘ured “ut-T | [TOT 19s | gece | ogg | 629 | Per | OF | Sze | 982 | 8 ry | oo | co | ge | ee | ore | 8 pereesesesesss9°° 1 QBROAY 189 | p99 | goo | 2e9 | goo | aso | ter | ose | 8 gop | cc | ct | gee | gue | tee ieae ilo raed wrt | ¢°6e0r esa | geo | geo | StS | OFF | Be | SPE | 082 | 6 te¢ | zo | org | oze | zoe | gez | g |fci1iiiciiiareid art | 2 6eor z sy | pep | ogp | Lee | we | vee | Tez | Fes | 9 yee | 082 | ere | ore | zzz | see | g i| Medan | 1 6e0r ous | zov | 6sh | gee | zoe | 9ze | oes | zez | 8 og yo fs | oes ore: | eee dpe |S eee cores s | oBgraay eze | goe | g9e | gee | see | esc | ees | ber | 2 e908 so awe Pees gpg: | ggg. | “med w-T | 8°OFOT aie Vl dere fl cinesenn Leas ah fete Lal eae hes | Ge As SP OF “+ | $68 | oe | Tez [548 uiejd “ut-T | ¢°L£0T : gos | ocg | gcc | gig | ope | Boe | Bre | O82 | 6 gge | “° | ‘7 | ope | Tez | 6st | 8 weld wit | Z°280T Bog | ggg | Seo | BTS | OFF | B6E | BFE | OBC | Be we | an bf OL Oe ee |e =gte doer oe peoyT aoue | ozo’ | oto’ | soo’ | 200° } 100° | 000° | 2000" umut | oo" | zoo" | Too" | $000" | 2000" 4818 cou || -Ixey ‘our 3u 4891 eon 489, puog qe S qe (seqout) (seqou) : seg punoy jo : awh jo dig pug 48 sse1g puog ot pene jo dyg pug ye stag puog sey pury pue azig ON meeg = [dnoiy -m100) SLSAL LAO-TINd SLSAL WVdd ‘qout aienbs 1od spunod ur ears are sassalyg “Ul g poppaqure siBq TO £4894 20I]} Jo oBeIIAG aq} ary suaUITOads yno-[INd eq} JO} WeATS sassaz}s Puog OU, ‘djs 1994801 ‘66 ATAVL 3 aq} SULMOYS pus of} 09 Jojar 6380} UIA oq} UI eq Jo dijs Jo sjueuleInsvou 94 L, ‘SHINDS TT6I—SLsa]L LN0-TINg INV SAVadgd NI BONVLSISaY GNOG AO NOSIYVdWON 139 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL ‘Jends & jo VAIO} 94} Ul aM "M-F7 Jo SUIN} g GWA paosOjMiel SHOOTY 4 406 Tog Ish oly LL& Iv 286 lle 4 tor 2, elp 88 “ge jad || Fae “" | exeieay oe Bike sae oa a as ee fan c 6c¢ ane ees | ozs esr | 2 || ---payesnasoo “0-341 © GFOL Th SIP CIP 688 TSE 0¢E Sle 108 ZL SLP GLE OsP oeP $9E 8 “+ payesnisoo “ul-8¢ 7 2 OF0T gl “PPFT | S8S 16h Ost $0F £98 682 B6e L ThP eeP Oly 098 Fre Z “**pazesnss09 “uI-84] T0701 960T | 26S 16h oeF ace 008 obs WSl [iL 967 ers LT Tle 18 8 Poe Dh Poder oO oe = ases0Ay TL sir olP 688 Ig¢ 068 SLE 103 L 91g Osh L9E 0g $96 621 6 “* payesnis0d “ur-84 7 £° F801 aPPPT | 989 16h Ose FOr £98 682 66S Z 087 £98 eI 966 09% OFG Z “* paqednssod “Ul-84 7 & Peo él #PL81 | PLL TL¢ och 008 91z col Ter 6 e8P olF ose Oe 068 Sh 8 “*"pazesnssoo “o1-84 1 Teor obs oss Org 66 ¥8P oF 618 908 6 91s ad pi ese 0ZE 19B [ROT JP | aseraay SPS ogg OFS 69 PSP ger 6LE 908 6 PLE SOE sce Ore 682 LES Or “1 taped “m-4T & SHOT ll Chg oes Org 66S ¥8h 9eF 6LE 90€ 6 BLE yo ar 998 198 882 iW ured “Or T &° SPOT ae sa i iis a8 3a y's ges i coF 4 wat ee oF wee | 6 |[cccctcceee | oBezeay obs oss OFS 60S $8P 98h 618 90€ ZL GLE pes PLP FHP L Oe oes as ured “ut-Y41 T'SF01 01 PRP SUSE A ACE TT ees yal Lanter calleremeneid Meee Ui cae | Io gce ai gee | ose [or |) ured “UAT 8° €F0T S6F Sop kes 98h +0F GFE 162 91% [KL ely $8 ea itd 90% [tL vereeeeesesesccceess 1 QgQIOAW geg | STS | 92G | 20S | 29m | 6Ib | 89g | F82 [342 geo | *** | 669 | OOF | 6I8 | 232 “med “W-F4T £°S80T 08F 5 ae 8Lb 6FF 868 9¢8 0G L oes 61s Slo 896 LES SOT “ureyd “UI-F%T & S80 6 slg | 48 | #98 | 82 | 8h2 | 802 | O8T 09st 8 Log | LOE | 0% — 08% 9% “med “OKT T S801 th ea sb aa se ie 5 nee : #le | OFF a $97 fete OFF vitteeesetees | g@ezaay ree Lee bts ae a ee me oe ‘ 86h 867 LLP 86h = 96% 28 Carey “ur-y & Shor 8 LS Leh Ost LE? PLE Fee 183 F0e 9 Ts £0F 0tF OIF Ter P8E 8 cos ured “aT 1 &Por ee Si aa ibe Say a ages F og eee 297 ct | IF ose lee ||: vettetercseee | ggeraay 2 ae ee es iid ae : 11g 98% vr Og? cOP o9e TE Jo ureyd “ar-T & SPOT L o1e P08 68 803 Ost oot 8 887 Bee 08h 6h SOP F6E 8 co uperd “urT 1Gr0r peoT eOn8 | 029° oto" $00" 00° 100" | $000° | 2000" umur | 00° 00° 100° | $000° | 2000° 4818 ‘om || -xey ‘ou pase 383.L oe 489, 7 48 78 umnur (saqout) (seqour) reg punoy jo ; “Tey Jo dig pug 38 svayg puog ay, Fad jo dyg pug ye ssa1yg puog aey, Pury pus azig ON WBeg | non “W100, SLSHL LNO-TTNd SLSdL WVGd ‘Saragg [I6I—SLsSa], LNO-1I1INg ANV SWVag NI GONVISISAY ANOg do NOSIUVdWO)) ‘dH ONILNOO—6Z ATAVL 140 ILLINOIS ENGINEERING EXPERIMENT STATION specimen and in a much less degree to the form of the specimen. The anti-stretch slip mentioned in a preceding paragraph is a manifestation of one of the secondary stresses. Since the beams were loaded sym- metrically and failure was practically always due to bond or some cause involving bond, the bar pulled out at the end where the bond resistance was the weaker. In other words the beam tests give in each case the smaller of two possible values of bond resistance resulting from the same combination of materials. In a few tests one end of the bar showed a greater slip during the earlier stages of loading, only to be overtaken by the other end as the loading progressed. In two or three tests both ends of the beam behaved almost identically in this respect. These observations apply to beams reinforced both with plain and de- formed bars. The pull-out tests for comparison with the reinforced con- crete beams were made in sets of three. Since we have automatically rejected the higher of the two bond resistances in the beams, it will make the comparison more nearly cor- rect to reject the highest one of the pull-out specimens. The average values for the lowest two from each set of pull-out specimens are given in Tables 29 and 30. These are the values which have been used for com- parison in the discussion of beam and pull-out tests. It will be seen that this method gives values for the pull-out tests which are a little too high; but the error due to this cause is not of enough consequence to justify further refinement in the computations. 74. Bond with Plain Round Bars.—In the 1911 series 23 beams were reinforced with 1-in. plain rounds and 7 beams with 11/-in. plain rounds. The effect of size of bar, vertical stirrups, depth of concrete below the steel, overhang of ends, etc., in beams reinforced with plain rounds will be discussed in detail in the following articles. It will be seen later that the presence of vertical stirrups has no considerable effect on the bond resistance of beams of this kind, hence for purposes of the present discussion all beams in the 1911 series reinforced with 1-in. plain round bars, except those having a total depth of 14 in. and those with unusual overhang of ends, will be grouped together. The groups with 1-in. plain rounds include 16 tests as shown in groups 1 to 5, Table 30. End slip became appreciable at a bond stress of 247 |b. per sq. in., 67% of the maximum bond resistance. The largest amount of end slip which was observed in every test in these groups was 0.001 in., corresponding to a bond stress of 344 lb. per sq. in., 98% of ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 141 TABLE 30. SumMaRY oF Bonp StTrREssES IN BEAM AND PULL-ovuT Trests— 1911 SpRizs. Stresses are given in pounds per square inch. Bond Stress at End Slip of Computed No. | Age (inches) Bond Group Characteristics of Test Stress at Tests! ro. |.0002|.0005] .001 |.002 |.005 | .01 ee Beam Tests; 1-in. Plain Rounds. 1-5 | All beams except as noted*................ 16 | 76] 247 | 310 | 344 ].....}..... 371 7 Beams with 9-in. overhang............... 2 | 914] 380 | 414 | 445 | 462 ]..... 500 8 Beams with 16-in. overhang............... 2 8 | 440 }..... 454 ]..... 440 474 6 Beams 14 in. total depth.................. 3] 8 | 433 | 481 | 508 | 523 ]..... 557 \ Pull-Out Tests; 1-in. Plain Rounds. 1-8 | Lowest two from each set of three tests®..| 14 |..... 225 | 298 | 353 | 399 | 457 | 481 493 All pull-out tests....... 0.0.0.0... ccs eee 21) 7 | 242 | 312 | 365 | 410 | 469 | 491 500 Highest from each set. . © \lrsweapsss 275 | 340 | 390 | 433 | 491 | 509 511 Lowest from each set................0.00. Cl eseson 201 | 284 | 338 | 385 | 430 | 475 491 Beam Tests; 114-in. Plain Rounds. 9 6-ft. beams with 3-in. overhang............ 3 744) 206 | 259 |..... B54 levees: le sarc 413 10 8-ft. beams with 3-in. overhang... . 2] 84] 397 | 416 |.....}.....J..... Jo... 465 11 6-ft. beams with 15-in. overhang........... 2 | 10%] 267 | 320 | 353 |.....)..... seed 376 Pull-out Tests; 114-in. Plain Rounds. 9-11 | Lowest two from each set of 3 tests®....... 8 | on bes 231 | 301 | 342 | 389 | 446 |.... 474 All pull-out tests................0.. 0c eee 12 | 8 | 239 | 313 | 365 | 412 | 459 | 483 483 Highest from each set................0-- Bel ices. 3 253 | 338 | 413 | 459 | 487 |..... 502 Lowest from each set................605. 4 hep eas 226 | 266 | 313 | 360 | 410 |..... 441 Beam Tests; 114-in. Corrugated Rounds. 12 Beams with 3-in. overhang................ 3 8 221 | 271 | 317 | 348 | 402 |.... 496 13 Beams with 9-in. overhang................ 3 | 714} 387 | 438 | 473 |.....]..... sect 493 Pull-out Tests; 114-in. Corrugated Rounds. 12,13 | Lowest two from each set of 3 tests®..... 6 ...| 184 | 283 | 270 | 313 | 400 | 440 + All pull-out tests...............-0e eee ee 9 | 74} 181 | 242 | 300 | 354 | 432 | 491 * Highest from each set. . 8 a8 say 141 | 262 | 360 | 464 | 526 | 598 + Lowest from each set...........-.... 0000 So lov eee 190 | 242 | 280 | 312 | 367 | 409 + * Includes all beams reinforced with 1-in. plain rounds, except those in groups 6, 7 and 8. +The maximum bond resistance for 5 beams made from Lehigh cement averaged 390 Ib. per sq. in. See zube tests, Table 27. © Use these values for comparison with bond stresses in beam tests. See Art. 73. + Omitted on account of part of the specimens being reinforced against bursting and part unreinforced. See Table 29 for details of these tests. 142 ILLINOIS ENGINEERING EXPERIMENT STATION the maximum bond resistance. It is evident that when the end slip of the bar has reached 0.001 in. the stress distribution is such that the beam generally has passed its usefulness for carrying load. In the case of five beams reinforced with 114-in. plain rounds (Table 30) end slip began at a computed bond stress of 282 lb. per sq. in., 65% of the maximum resistance. In this group an end slip of 0.0005 in. was measured in all the tests, corresponding to a bond stress of 86% of the maximum bond resistance. A comparison of the bond stresses developed in the beams rein- forced with 1-in. and 114-in. plain rounds in Table 30 and Fig. 49 with the pull-out tests on bars of the same sizes (the average of the lowest two tests from each set) shows values for the two kinds of tests which are similar at each amount of slip given in the table up to an end slip of about 0.001 in. The maximum bond resistance for the beams in these two groups is about the same as the bond stress developed in the corresponding pull-out specimens at a slip of 0.002 in. Fig. 78 shows the load-deflection and load-slip curves for typical beams reinforced with plain round bars. The tests do not show any material difference in bond unit resistance due to size of bar. A summary of the bond stresses developed in all the 6-ft. beams reinforced with plain round bars and loaded at one-third points is given in Table 35. 75. Effect of Vertical Stirrups on Bond Resistance——Beam groups 1, 2 and 3, in Table 28, may be used to see whether the presence of vertical stirrups influences bond resistance in beams of this kind. The beams in group I had no stirrups; those in groups 2 and 3 were rein- forced throughout the outer thirds with V-shaped stirrups of 14-in. and ¥,-in. rounds, respectively, spaced 6 in. apart as shown in Fig. 2 (b). Each of these beams was reinforced with one 1-in. plain round, and was tested on a 6-ft. span. The relation of bond resistance due to the presence of web reinforcement is not well defined. The beams with Y-in. stirrups gave higher values for bond resistance at all stages of the tests than either of the other groups. In all these tests end slip of bar became appreciable at about 65% (range 64% to 67%) of the ultimate load. At an end slip of 0.0005 in. the average bond resistance of the beams is 82% of the ultimate. It will be seen in Tables 27 and 28 that both the cube tests and pull-out tests for the beams in group 2 indicate a better quality of concrete than that in the beams of groups ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 143 1 and 3. The results indicate that bond stress is the primary cause of failure in all these tests. Although the diagonal tensile stresses were fairly high, it would appear that if the reinforcement had not slipped, failure by diagonal tension would not have occurred until a higher load had been applied. Of course stirrups of this kind may not be expected to be effective when the bond stresses are high. Attention is called to the discussion in Art. 78. Load deflection and load-slip curves for beams with and without stirrups are given in Fig. 78 and 79. The appearance of some of the beams after failure is shown in Fig. 57. 76. Effect of Depth of Concrete below the Reinforcement.—In general, the depth of concrete below the center of the longitudinal rein- forcement was 2 in. In two groups of tests this depth was made 1 in. and 4 in., respectively; that is, the total depth of the beams was 11 and 14 in., with the depth from top of beam to center of bar 10 in., as usual, These beams were not provided with vertical stirrups; they were each reinforced with one l-in. plain round bar and were tested on a 6-ft. span. See groups 5 (1-in. depth), 1 (2-in. depth) and 6 (4-in. depth), Table 28. It is seen that there is little difference in the values for the beams with depths of 1 and 2 in., but the beams with 4 in. of concrete below the steel show a very great increase in strength. The maximum bond resistance for the 1-in. and 2-in. depth averaged 356 lb. per sq. in., while for the 4-in. depth the maximum was 557 |b. per sq. in. The load-deflection and load-slip curves for these beams are given in Fig. 78. The deflection of beam is greatest for the beams with 1-in. thickness and least for those with 4-in. thickness. Beams in group 6 show great stiffness up to a load of about 6000 lb. The photo- graph, Fig. 58a, shows the appearance of these beams after test. The numbers opposite the cracks indicate the growth of the cracks at loads expressed in thousands of pounds. An interesting feature of the tests in group 6 is the absence of cracks in the outer thirds of the beam length. It seems evident that the anti-stretch slip has occurred from the cracks near the load points and that slip has progressed outward from these cracks toward the ends of the beam during the remainder of the test instead of from the several cracks which usually appear in the outer thirds subsequently in the tests. The first end slip came at a materially higher load in this group. It seems probable that the distribution of 144 ILLINOIS ENGINEERING EXPERIMENT STATION bond stress along the bar beyond the outer cracks in this case is not materially different from that found in a pull-out test. The full bond resistance for the embedded length of bar beyond the crack is developed, or at least that for a small amount of slip, instead of having part-of the bond resistance consumed by stresses in opposite directions due to anti- stretch slip. The presence of the large mass of concrete below the bar acts to maintain the integrity of the concrete; its total tensile strength is greater than the available residual resistance to slip along the bar, and hence the concrete will slip along the bar, probably still maintain- ing some tensile stress. In the case of a small depth of concrete the total tensile strength of the concrete is insufficient to overcome the bond resistance, the concrete breaks in tension at another point and springs back both ways from the point of rupture, forming a crack. With further addition of load further slip will occur at the crack, and with still more load other cracks may appear. In this case a part of the bond resistance is consumed by stresses in opposite directions due to anti-stretch slip and due to the slip the bond resistance utilized is less. For the beams of ordinary depth in groups 1 to 5 in Table 28 the outer cracks at the ends where failure occurred cross the plane of the steel at distances ranging from about 12 to 22 in. (average 17 in.) from the end of the beam. For the beams in group 6 with 4 in. of concrete below the center of the reinforcement, the principal failure cracks came at 21 to 30 in. (average 26 in.) from the end of the beam. It would appear that after the cracks have lengthened, the main effective bond resistance lies on the part of the bar between the outermost crack and the end of the beam. 7%. Effect of Span Length—The beams in group 11, Table 28, were loaded with an 8-ft. span. The beams were reinforced with 114-in. plain round bars and ¥-in. vertical stirrups. These tests may be com- pared with the beams in group 9, loaded on a 6-ft. span. In the 6-ft. beams the first outer cracks appeared at an average load of 10 700 lb., and the first end slip of bar at 12 700 lb.; the corresponding loads for the 8-ft. beams were 8000 Ib. and 16 000 lb., respectively. These loads are significant. The first outer crack appeared in both groups of beams at loads corresponding to a stress of about 13 500 Ib. per sq. in. in the longitudinal steel at mid-span. The steel stresses at mid-span at be- ginning of slip at the end of the bar are about 15 600 and 26 000 Jb. per sq. in. for the 6-ft. and 8-ft. beams, respectively, corresponding to ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 145 a computed bond stress of 206 and 257 lb. per sq. in. The average distances of the cracks from the ends of the beams were about 17 in. and 25 in., respectively. This suggests that when end slip of bar began the main effective bond resistance for beams of both lengths lies princi- pally in the portion of the bar having an ubroken embedment. The fact that the maximum bond resistance for the 8-ft. beams is lower than for the 6-ft. beams is probably due to the opening of other outer cracks than the first one, nearer the supports, causing a further concentration of the bond resistance toward the ends of the bar. Tests of other beams in which the span length varied are discussed in Art. 86. %8. Effect of Length of Overhang of Ends of Beam.—It has been the practice at the University of Illinois in designing test beams of the kind described herein to make the beam 6 in. longer than the test span. In making the usual calculations of bond stress no allowance has been made for this additional 3 in. of embedment of the reinforcing bar. In order to study the effect on bond resistance, some of the beams in the 1911 series were tested with the ends overhanging the supports for 9 in. or 15 in., as shown in Fig. 2 (c) and with the bars extending to the ends of the beams. 1-in. and 114-in. plain rounds and 11,-in. corrugated rounds were used in these tests; see groups 7, 8, 10 and 13 in Table 28. Figs. 58 and 59 show representative beams after failure. A summary of the tests is given in Table 30. These beams differed somewhat in web reinforcement, even in the same groups, and these differences must be kept in mind in making comparisons. The effect of varying the length of overhang may be seen by com- paring the values for the different functions given in Table 28 and also the bond stresses corresponding to different amounts of end slip in the beam and pull-out tests in Table 30. The applied loads at first outer crack show whether the appearance of anti-stretch slip was affected by overhang. The applied loads and the bond stresses developed at first end slip and at the maximum load indicate the influence of the over- hang on the bond resistance of the beams. One-inch plain rounds were used in beams with the ends overhang- ing the supports 3, 9 and 15 in. The tests with 3-in. overhang include 16 beams; the others two or three each. 114-in. plain rounds were used in beams with 3 in. and 15 in. overhang, and 114-in. corru- 146 Applied Load -La Computed Bond stress - Lb. per 5g.In. ILLINOIS ENGINEERING EXPERIMENT STATION 40000 r 275 ; 7 35000 ae - 250 oi as : o —Laggin 30000 | @5%° aa QV s a? 2007 ea 8 25000 & ea L175 5 \p , wand i, on ae 1 -Maximum Load. & pie 2- First Slip of End of Bar. a Ss -First Outer Crack. 20000 F 199 & a oak wD i 8 - 125 . 15000 -— + 100 § = é an, Plain Round _ -4 g athe’ 10000 % &3 Jé-in. Cor. Round & Fin. Plain round + 50 5000 - 25 oo 600 d - 500 A o 400 300 gun 4 /- Maximum Load. e- First Slip of End of Bar. 200 3- First Hip of Bar corrected for Overhang. /-2-4 Concrete Age about & mo. 100 al 6-f¢ Span. All Beams reinforced with 2-in Stirrups 6 in.apart o 3 9g 1S Overhang of Ends of Beams -inches Fic. 48. Errecrt or Lenctu oF OverHanc or Enns or BEAMS. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 147 gated rounds in beams with 3 and 9 in. overhang. The values for bond stress given for these beams in the tables have all been computed in the same way without reference to the length of overhang of the ends. A study of the data shows that the amount of the overhang of the ends had practically no influence on the load causing the opening of the first outer cracks and that the load at which first outer cracks appeared was independent of the kind and size of bar. The load at which first end slip of bar was found was greatly influenced by the length of overhang. In beams without stirrups the additional overhang seems to have little, if any, effect upon resistance to diagonal tension. For the beams recorded as failing in diagonal tension, the vertical shear- ing stress averaged 178 lb. per sq. in. for beams without stirrups and 288 lb. per sq. in. for beams with stirrups. The relation between the bond stresses is somewhat different from that of the shearing stresses on account of the different diameters of the bars used. In Fig. 48 the values of the computed bond stresses have been plotted for the beams which were reinforced with vertical stirrups. It will be seen that the beams with 9-in. and 15-in. overhang developed a considerably higher computed bond stress than the beams with 3-in. overhang. If we reduce the computed bond stresses in these beams by considering that the bond stress at any load is uniformly distributed over the overhanging portion of the bar in the same way as is assumed for the outer thirds of the span (equivalent to multiplying the computed bond stress in beams with 3-in. overhang by 24/27, that in beams with 9-in. overhang by 24/33, and that in beams with 15-in. overhang by 24/39), it will be found that such corrected bond stresses will generally be greater for the longer overhangs. It should be noted that this method of correcting the bond stresses is subject to error in one respect, since slip was measured at the ends of the bars in all cases. The beginning of end slip with 15-in. overhang may be expected to represent quite a different state of stress in the beam proper than exists in a beam with 3-in. overhang when end slip begins. 79. Bond on Corrugated Bars.——114-in. corrugated round bars were used for longitudinal reinforcement in 6 beams of the 1911 series— groups 12 and 13. These tests were referred to in the preceding articles ; they will be included also in the discussion of similar tests from the 1912 beam series, Art. 85. 148 ILLINOIS ENGINEERING EXPERIMENT STATION 500 475 L hee Lae 450 Za at IIS g — /-in, Plain Pounds S W-- (fin * ” ‘ J50 —ieinCor + 3 Q 52s g s § wO0 27s ~ 250 ees 1-2-4 Concrere, Age about 8 months. All beams were rested with eras overhanging (he SUQpATS FIN, one The curves for pulf-our tests are the averages Of I/7e values for (he lowest two fromm each WS Jer of ree specimens, ——— 150 Oo OW O00? 003 :O08 OOS .006 :007 1008 209 O10 Sp of nal of Gar-lrches. Fic. 49. Comparison oF Bonp Stresses In BEAMS AND PuLL-out SPECIMENS— 1911 SERIEs, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 149 80. Bond Resistance in 1911 Beam and Pull-out Tests—Table 30 gives a summary of the bond stresses in the 1911 beam and pull-out tests. The load-slip curves for many of these groups have been plotted in Fig. 49. As noted in Art. 73 the average stresses from the lowest two from each set of three pull-out specimens is used in all comparisons between the beam and pull-out tests. For the 1-in. plain rounds in beams with 3 in. overhang the bond stress up to an end slip of 0.001 in. is about the same as for the pull-out specimens for the same end slip. The bond stress at an end slip of 0.001 in. is 93% of the maxi- mum bond resistance for the beam tests and 72% for the pull-out tests. The 1 and 114-in. plain rounds give nearly identical values in the pull- out tests. It should be borne in mind that with the uneven distribu- tion of bond stress developed in beam tests the bond stress developed with a given end slip is dependent upon the relative dimensions of bar and beam and hence that the relation between the values of bond in beam tests and in pull-out tests here given may be to a certain extent accidental. In the tests with 114-in. plain rounds the 6-ft. beams with 3-in. overhang gave values which are somewhat lower than the values from the pull-out specimens at the earlier stages of the tests; at the maximum load the average computed beam bond stress is 87% of the maximum bond resistance found in the pull-out tests. For the 114-in. corrugated round bars the bond stresses in the beam tests at end slips up to 0.005 in. are higher than in the pull-out tests for the same end slip. The maximum bond stresses in the two forms of specimen can not be compared directly, since part of the pull-out speci- mens were reinforced against bursting; in the one set not so reinforced the maximum bond resistance was 471 lb. per sq. in. as compared with 496 lb. per sq. in. in the beam tests. The corrugated bars in the beam tests give stresses which are about intermediate between those for the 1-in. and the 114-in. plain rounds for end slip less than about 0.002 in. The maximum bond resistance for the corrugated bars is about 25% higher than for the plain rounds. Considering the pull-out tests only, the plain rounds give higher bond resistances than the corrugated rounds for all stages of the test up to the end slip which corresponds to the maximum bond resistance in the plain bars. A discussion of the relation of bond resistance in beams and pull-out specimens from another series of tests is given in Art. 92. 150 ILLINOIS ENGINEERING EXPERIMENT STATION c. 1912 Beam Tests. 81. Outline of Series—Sixty-three reinforced concrete beams were included in the 1912 series. Generally the span length was 6 ft.; one group of beams was tested on each of the following spans: 5, 7 and 8 ft.; four groups were tested on a 10-ft. span. The loads were generally applied at the one-third points; groups of 6-ft. beams were tested with two symmetrical loads spaced the following distances apart: 2, 21%, 3, 814% and 4 ft. Plain round, plain square, twisted square, and corru- gated round bars were used for longitudinal reinforcement. In all but six of the beams the longitudinal reinforcement consisted of a single bar. Three beams were reinforced with four 5¢-in. and three with three 34-in. plain rounds. The beams in this series were made in groups of three; the first beam in each group was made of hand-mixed concrete, the remainder of machine-mixed concrete. All the beams were provided with vertical stir- rups of 14-in. rounds. In the beams reinforced with a single longi- tudinal bar, the stirrups were V-shaped; in the other beams U-shaped stirrups were used. See Fig. 2. The average age at test was 63 days. Pull-out specimens and 6-in. cubes were made with nearly all of the beams. The pull-out specimens were stored in the open air with the beams; the cubes were stored in damp sand. Details of the make-up of the beams, their dimensions, the materials used and data on the strength of the concrete will be found in Table 31. Table 32 gives data of the tests and some of the calculated stresses in the beams, as well as notes on the failures. The bond stresses corre- sponding to given amounts of end slip for both beam and pull-out tests will be found in Table 33. A summary of the bond stresses developed in this series of tests is given in Table 34. In comparing the results of pull-out tests for the square, twisted and the deformed bars in this series with those made earlier it should be borne in mind that the concrete cylinders in this series were not reinforced against bursting. 82. Bond Resistance with Plain Round Bars.—Thirty-seven beams in the 1912 series were reinforced with 1-in. plain round bars. The tests to determine the effect of span length, effect of variation in the position of the loads, and on the sizes of bar are discussed elsewhere. Table 35 gives a summary of the results of tests of 6-ft. beams reinforced with 1 and 114-in. plain round bars from the series of 1909, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 151 1911 and 1912. The 1909 tests gave lower values than the other series, but the concrete was of lower strength, as may be seen from the cube tests. All but one of the beams included in this table failed in bond. Disregarding the differences in the age of the beams, the mean values for bond resistances are found as given in the table. First slip of bar was measured at the end of the beam at a mean bond stress of 236 lb. per sq. in—66% of the maximum bond resistance. The mean bond stress at an end slip of 0.001 in. for the beams in these three series was 324 lb. per sq. in—91% of the maximum bond resistance. The mean maximum bond resistance for the 34 beams considered was 356 lb. per sq.in. These values seem to be representative for the conditions present in the tests. However, it should be noted that the bond resistance developed in beams, as given by the ordinary methods of computation, appears to be dependent upon the relation of size of bar to the dimen- sions of the beam and to the span length and that the computed values found in these tests may not apply to bars of smaller diameters or to beams of other size and span. 83. Bond Resistance with Plain Square Bars.—Six beams were reinforced with 1-in. plain square bars. In group 29 the bars were placed with their sides horizontal; in group 30 the bars were placed with a diagonal line horizontal. Comparing the bond stresses in the beam and pull-out tests for groups 29 and 30 in Table 32 and in the sum- mary, Table 34, it will be seen that while the bars placed with a diagonal horizontal give a higher bond resistance, the pull-out and cube tests show that on the average the concrete in the second group of tests was inferior in strength to that in the first group. Beam No. 1046.5 gave abnormally low bond resistance. It is probable that no difference in bond resistance can be ascribed to the difference in the position of the bars. The bond stresses at various end slips in the beams reinforced with 1-in. square bars are about 75% of those for similar beams rein- forced with 1-in. rounds. See groups 14-15 and 29 and 30, Tables 32 and 34. A comparison of the pull-out tests gives about the same relation. End slip of bar begins in the pull-out tests at a lower stress than in the beam tests; at a slip of 0.001 in. the stresses are about the same. At the maximum loads the calculated bond stresses in the beams average 86% of those in the pull-out tests. This is about the same ratio as for the tests with 1-in. round bars in this series. 152 1-2-4 concrete; Universal portland cement, sand, and crushed limestone. ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 31. Data oF REINFORCED CONCRETE BEaMs—1912 SERIEs. The concrete for the first beam of each group was mixed by hand; all others were machine- mixed. All beams were 8 in. wide; total depth 12 in.; depth to center of steel, 10 in. Each beam, except those in Groups 27 and 28, was reinforced with a single longitudi- nal bar. See Fig. 2 (f). All beams were provided with vertical stirrups of 14-in. plain rounds, spaced 6-in. apart, outside the load points. Longitudinal Reinforcement B OnE ee of eam : Beam | Length from Mixture Group] “No. feet Per ae. by Weight Age | Average of 3 Size and Kind of Bar cent | PA da Teste Y8 llb. per aq. in. 1052.1] 634 | 1-in. plain round.......... 0.98 | 1051.1} 1-1.95-3.34 | 70 2120 14 | 1052.2] 644 | 1-in. plain round.. 0.98 | ...... 1-1.95-3:34 | 61 2730 1052.3 | 634 | 1-in. plainround...........| 0.98 | 1057.3 | 1-1:93-3123 | oi 15 | 106.1] 634 | Lin plain round...........] 0.98 |{1087-1}) 11.0¢8.20 | 62 2140 1056.2 6% | 1-in. plain round.......... 0.98 | ...... 1-2.00-3.20 68 3500 1056.3 614 | 1-in. plainround...........]| 0.98 | 1047.2 | 1-2.00-3.36 es svekeng 1052.4 | 634 |)1-in. plain round 0.98 | 1055.1] 1-2.07-3.48 | 61 2065 16 | 10525] 634 |} 4° 84in, auxiliary barsi| 0.98 | ...... 1-1.92-3.20 | 66 2640 1052.6 | 634 |] asteachend of beam.... || 0.98 | 1058.3] 1-1.93-3.32 | 60 2960 1057.1 | 634 | 1-im. plain round...........] 0.98 |{t923-1}| 1-1.94-3.29 | 62 2140 17. | 1057.2 6% | 1-in. plainround...........] 0.98 | ...... 1-1.98-3 34 1057.3 | 634 | Lin. plainround...........| 0.98 | 1052.3 | 1-1,93-3.23 1058.1] 6) | 1-in. plain round...........) 0.98 |{1088-t)) 11,043.29 | 62 2140 18 | 1058.2 6% | 1-in. plain round........... 0.98 | ...... 1-1.96-3.27 66 2730 1058.3 | 634 | 1-in.plainround...........| 0.98 | i052.6| 1-1.93-3.32 | 60 2960 1059.1 | 634 | 1-in, plain round........... 0.08 |{iOes'1}| 1-2.07-8.46 | 60 2680 19 | 1059.2} 634 | 1-in. plain round O08" || secre 1-1.99-3.32 | 66 2600 1059.3 | 634 | 1-in. plain round.. 0.98 | 1062.3] 1-1.923/31 | 67 2600 1060.1] 644 | 1-in. plain round........... 0.08 |{1058 1] 1-2.07-3.46 | 60 2680 20 | 1060.2 | 614 | 1-in. plain round... 1055.5'| 1-2.01-3.23 | 64 3140 1060.3 | 624 | 1-in. plain round lobe} 1-1.98-3.30 | 63 3250 1051.1 | 534 | 1-in. plain round 1052.1 | 1-1.95-3.34 | 70 120 21 =| 1051.2 | 544 | 1-in. plainround...........] 0.98 | ...... 1-1. 94-3 26 66 2800 1051.3 | 534 | 1-in. plain round.. 1053.3 | 1-1.983.28 | 68 2760 1053.1 | 734 | 1-in. plain round 1050.4 | 1-1.95-3.09 | 60 22 | 1053-2] 74 | Lin. plain round...........] 0.98 |... 1-1.98-3:15 | 62 2200 1053.3 74 | 1-in. plain round 1051.3 | 1-1.98-3.29 Bite 1054.1 | 844 | 1-in. plain round . 1046.1 | 1-1.93-3.26 62 23 | 1054.2 814 | 1-in. plain round. . BS | oacctme 1-1. 96-3 28 65 2800 1054.3 814 | 1-in. plain round.......... 0.98 | 1061.6 | 1-1 98-3.30 63 2870 * The average compressive strength of 36 6-in. cubes of hand-mixed con oubes of machine-mixed concrete, 2800 lb. per aq. in. crete was 2200 Ib. per sq. in.; of 93 (Table 31 continued on page 154) ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 153 TABLE 32. Tests or Reinrorcep ConcreETE BEaMs—1912 SERIzEs. All beams had 3 in. overhang at each end. In computing unit stresses the weight of the beam was considered. Loads are given in pounds; stresses are given in pounds per square inch. At First Slip 4 Maxi- Dis- lof Endof Bar| At Maximum Load mum Test | nce] Age Load Bond Group] Beam | Span | yea | atte | whe Verti Rent railure of B roup ween | Test | Firs -1,| Verti- ance | Failure of Beam No. | feet | Loads| j.. | Outer Ap- Come Ap- | Zensilel cal | Com: | Yin days i puted| Stress puted feet Crack] plied Bond lied in | Shear- Bond | Pull- oad Stress oad Steel Strens Stress jute 1052.1 6 2 67 | 8 000) 8 000) 154 /19 940/36 800] 149 | 383 | 356 |Bond at S. end 14 1052.2 6 2 59 11 000/11 000} 211 {18 000/33 200} 136} 347] 405 |Bond at N. end 1052.3 6 2 64 | 6 000/16 000} 307 |20 000/36 800) 150] 385 ]...... Tension in steel Avs lusnewsl tase 3 63 | 8 300/11 700) 224 |19 300/35 600] 145 | 372) 395 1056.1 6 2 65 |14 000/13 000) 250 {15 750/29 200} 119] 305 | 454 |Bond at N. end 15 1056.2 6 2 65 | 6 000/12 500} 240 |19 700/36 300} 149] 379 392 |Bond Cay tension in steel 1056.3 6 2 65 {10 000/13 500) 260 |21 000/38 600] 158 | 403 | 443 |Bond at N. end Ave (isseccsleur cas 65 |10 000}/13 000) 250 |18 800/34 700) 142] 362] 430 1052.4 6 2 61 |14 000/18 000) 347 |20 600/37 900} 155 | 395 | 443 |Bond at N. end 16 1052.5 6 2 63 |10 000/11 500) 220 |18 500/34 100} 139 | 356] 415 |Bond at N. end 1052.6 6 2 65 {10 000/14 000) 269 |18 800/34 700] 142] 383 | 381 [Bond at 8. end Ay: lesshle ei aeaax 63 {11 300)14 500) 279 |19 300/35 600] 145 | 372 | 413 1057.1 6 2144 | 68 | 8 000/12 000] 231 |19 500/31 400] 147 | 375 | 323 |Bond at N. end 17 1057.2 6 244 | 64 | 8 000/13 000) 250 |19 100/30 800) 144] 368 ]...... Bond at §. end 1057.3 6 2144 | 63 | 8 000/13 000] 250 |25 000/40 000] 187 478] 322 |Tension in steel AVG ©: |speecms aipaaaantees 65 | 8 000/12 700) 244 |21 200/34 100] 166 | 407] 322 1058.1 6 3 68 8 000/14 000} 272 18 900)26 100] 142 364 454 |Bond at N. end 18 1058.2 6 3 63 6 000/10 000} 198 |14 000/19 500] 106 | 272 | 252 |Bond at both ends 1058.3 6 3 GR IL gikdiees 13 000) 254 |18 600/25 700] 140] 359 | 381 |Bond at N. end AYE 9 || secae Sloanawi 66 7 000/12 300) 241 |17 200/23 800} 129 | 332] 363 1059.1 6 3% 64 |26 000/12 000} 285 |33 750/38 400} 251 642 | 591 |Bond at N. end 19 1059.2 6 344] 59 |...... 10 000] 198 17 700/20 400} 134] 341]...... Bond at S. end 1059.3 6 34 62 |14 000}12 000) 285 |22 000/25 200] 165] 421 416 |Bond at N. end Avs |vaasexlaccess 62 |20 OO0}11 300) 223 |24 400/28 000] 183] 468) 504 1060.1 6 4 70 {21 000}16 000} 309 |32 000/29 100} 238 | 608 | 591 |Bond at S. end 20 1060.2 6 4 62 {21 000)18 000) 347 |24 700/22 600) 185 | 472 | 599 |Bond at S. end 1060.3 6 4 Bho | senzen 17 000} 329 |30 000|27 400] 224] 573 | 579 |Bond at N. end Avy |Pteansheega es 64 {21 000/17 000} 328 28 900/26 400) 215 551 590 1051.1 5 1% 68 | 8 000/16 000) 290 |16 800]25 800} 128 | 324] 386 |Bond at S. end 21 1051.2 5 134 60 | 6 000/13 000) 252 |20 000}30 600) 151 383 | 342 |Bond at S. end 1051.3 5 1% 63 | 8 000/15 000) 290 |20 100}30 800) 152) 385} 480 |Bond at N. end Ave flows san fies pene 64 | 7 300/13 700} 277 |19 000/29 100) 144 | 364] 403 1053.1 7 2% | 62 | 8 O00\(No rejadings)/17 100/37 000) 130 | 331 | 366 |Tension in steel 22 1053.2 7 2 63 8 000/12 000) 236 |16 600}36 000) 126 | 322] 400 /Bond ane tension in stee 1053.3 7 244 | 63 | 8 000/16 000) 310 /18 400/39 800} 139 | 356]...... Tension in steel Avi. liaise forse ey 63 8 000)......]...... 17 400)37 600} 132 | 336] 383 1054.1 8 224 | 60 | 6 000/10 000) 210 {16 000/40 000) 121) 312] 459 {Tension in steel 23 1054.2 8 2% 63 | 5 000/12 000) 227 /15 100/37 800) 114 | 296} 525 |Tension in steel 1054.3 8 224 | 60 | 8 000/13 000} 256 |15 400/38 500) 117 | 301 | 596 |Tension in steel Ave. NecissareMes sista 61 '6 300/12 000! 231 '15 500'38 600! 117 | 303) 527 (Table 32 continued on page 166.) 154 ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 31—ContTINnUED FROM PAGE 152. Data or RetnrorceD Concrete Brams—1912 Smrizs. , C ion T f Longitudinal Reinforcement Beam bear es Cul a - Group Beam Length Se from Mixture No. feet ; on Banh by Weight Age | Average of 3 Size and Kind of Bar cent days |p, ae ey. 1055.1 1044 | 1-in. plainround........... 0.98 | 1052.4 | 1-2.07-3.48 61 2065 24 | 1055.2] 10% | 1-in. plainround........... 0.98 | oo... 1-1.99-3 28 — 1055.3 | 104 | 1-in. plain round........... 0.98 | oo... 1-1.98-3.37 1055.4 | 1044 | 1-in. plain round 0.98 | 1048.4 | 1-2.03-3.28 61 1940 25 1055.5 | 1044 | 1-in, plain round.. 0.98 | 1060.2 | 1-2.05-3.13 stats 1055.6 | 1044 | 1-in. plain round 0.98 | wi... 1-1.97-3.27 1055.7 | 10)4 | 1-in. plainround........... 0.98 | 1050.1 | 1-2.01-3.32 60 2220 26 | 1055.8 | 1044 | 1-in. plain round........... 0.98 | 1.1... 1-2.05-3.44 65 2980 1055.9 | 1044 | 1-in. plainround........... 0.98 | ww... 1-1.99-3.31 67 2870 1050.4 6% | 4 5-in, plainrounds...... 1.53 | 1058.1 1-1.95-3.09 60 2170 27 ~+| 1050.5 64 | 4 %%-in. plain rounds...... 1,53) sacas.s 1-2.02-3.14 62 2700 1050.6 642 | 4 5%-in. plain rounds.,..... 1.53 | 1061.3 1-1.95-3.32 60 2390 1050.1 6% | 3 %-in. plain rounds...... 1.66 | 1055.7 | 1-2.01-3.32 60 2220 28 | 1050.2 6% | 3 3-in. plain rounds...... 1.66 | ...... 1-2.02-3 .29 66 3290 1050.3 64% | 3 34-in. plain rounds...... 1.66 | 1047.3 | 1-1.95-3.36 me as 1046.1 61% | 1-in. square bar............ 1.25 | 1054.1 | 1-1.93-3.26 62 2580 (Side horizontal) 29 | 1046.2 6% | 1-in. square bar.. eggs! eg 2b aniatongsa 1-1.91-3.20 68 3790 (Side horizontal) 1046.3 644 | 1-in. square bar............ 1.25 1048.3 1-1, 93-3 25 59 2800 (Side horizontal) 1046.4 614 | 1-in. square bar............ 1.25 | 1047.1 1-1.96-3 .29 60 2190 (Placed on edge) 30 | 1046.5 6% | 1-in. square bar............ D258 IS nsecscvess 1-1.96-3.35 68 1890 _ (Placed on edge) 1046.6 614 | 1-in. square bar............ 1.25 | 1048.6] 1-1.98-3.39 64 2080 (Placed on edge) 1047.1 64 | 1-in. twisted square bar..... 1.25 | 1046.4 | 1-1.96-3.29 60 2190 (1 twist per lineal ft.) 31 1047.2 644 | 1-in. twisted square bar.....| 1.25 | 1056.3 1-1. 98-3 .34 71 2980 (1 twist per lineal ft.) 1047.3 | 6} | 1-in. twisted square bar.....| 1.25 | 1050.3 | 1-1.95-3.36 59 2670 (1 twist per lineal ft.) s 1059.1 1048.1 6% | 1}4-in. corrugated round...| 1.24 1060 1}| 1-2-07-3.46 60 2680 32 | 1048.2 614 | 1}¢-in. corrugated round...) 1.24 | ...... 1-1.96-3 24 61 3150 1048.3 64 | 1)4-in. corrugated round...| 1.24 | 1046.3] 1-1.91-3.30 an cates 1048.4 614 | 1}4-in. corrugated round...| 1.24 | 1055.4 | 1-2.03-3.28 61 1940 33 | 1048.5 644 | 1)-in. corrugated round...| 1.24 | ..... 1-2.05-3.19 65 3260 1048.6 64 | 1}4-in. corrugated round ...| 1.24 | 1046.6 | 1-1.98-3.34 ead 1049.1 | 1014 | 1}4-in. corrugated round...| 1.24 | ...... 1-1.97-3.28 63 2420 34 | 1049.2 | 1044 | 1)4-in. corrugated round.. | 124 | ...... 1-2,01-3.19 62 2440 1049.3 | 10% | 1}4-in. corrugated round...| 1.24 | ...... 1-1.95-3 32 ee gam * The average compressive strength of 36 6-in. cubes of hand-mixed concrete was 2200 Ib. per sq. iv.; of 93 cubes of machine-mixed concrete, 2800 lb. per sq. in. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 155 TABLE 32--ContTINUED FROM PAGE 153. Trsts or REINFORCED CONCRETE BEamMs—1912 SERIEs. At First Slip 2 : Maxi- Dis- of End of Bar At Maximum Load mum Test | tance | Age | Load Bond Beam | Span}, be | at at Resist- Group tween | Test | Firat Verti- No. | feet | Toads Our | Ap. | Com} ap. [Tensile Com- p- p- cal in feet | 2298 | Crack| plied | BU | plied | St7e55| shear-| BUted | py. on! in ° Bond Load | Stregs | L994 | Steel Stes Streas ou 1055.1] 10 | 314 | 64 | 4 500|(Noen/d slip) 12 300/39 300] 94 | 242 | 443 |Tension in steel ance | Failure of Beam 24 | 1055.2] 10 3% | 60 | 4 000/12 000 12 300/39 300} 94; 246] ... |Tension in steel 1055.3 | 10 344 | 2yr. | 4 000/10 400) 205 11 100/34 300} 86 | 220] 560t/Tension in steel Ave kiasssliarasahseanee 4 000/11 200} 222 /11 900/37 600) 91) 236] 501 1055.4 | 10 334 | 64 | 4 000/10 800} 218 |10 800/34 800) 83 | 218] 459 |Tension in steel 25 | 1055.5) 10 314 | 65 | 4 000] 9 500) 190 | 9 900|32 200] 76} 201 | 599 |Tension in steel 1055.6f) 10 3% | 2yr. | 4 000/10 400] 215 |12 100/38 700) 92) 236 | 497{|Tension in steel Ady heweealraenes ea nene 4 000/10 800) 208 |10 900/35 200] 84] 218] 518 1055.7 | 10 314 | 62 | 4 000/12 100} 242 |12 100/38 700} 92 242] 562 |Tension in steel 26 | 1055.8") 10 93 | 4 000/10 000) 203 |12 800*/24 300) 97 | 255 | 415 |Bond at S. end 85 | 4 000|12 000) 240 |16000°/30 200) 121 | 315] 349 |Bond at S. end 6 6 1050.4 6 2 60 |10 000]24 000) 195 |33 700/40 300} 259 | 269 | 405 |Tension in steel 2 2 27 (| 1050.5 6 62 {12 000/19 000} 156 |34 000/40 700} 261 | 271 | 591 |Tension and bond 60 | 8 000/20 000] 164 |27 900/33 500} 215 | 225 | 503 |Bond at N. end Sta het 61 |10 000/21 000) 172 |31 900/38 200} 245 | 255 | 500 6 2 61 |12 000]24 000) 217 |34 500)38 600) 265 | 308] 524 |Bond at S. end 28 | 1050.2 6 2 62 [12 000/20 000) 183 |26 000/29 200) 201 | 235 519 |Bond at N.end 6 2 57 |10 000|19 000] 174 |31 200/34 900) 241 | 278 |) 642 |Bond at S. end si gtbintacs 60 {11 300/21 000} 191 |30 600/34 200) 235 | 274 562 | 6 2 62 | 8 000|11 000) 174 |19 000/27 800} 146 293] 336 |Bond at S. end 29 | 1046.2 6 2 62 |10 000|11 000) 174 |19 300/28 200] 148 | 298] 287 |Bond at S. end 1046.3 6 2 57 |10 000|14 000) 219 |20 500/30 000} 157] 316] 465 |Bond at S. end ae ee eee 60 | 9 300/12 000) 189 /19 600]28 700) 150 | 302| 363 1046.4 6 2 61 | 6 000/12 000} 189 |20 000/29 200) 154 | 308] 340 jBond at S. end 30 | 1046.5 6 2 62 | 8 000/11 000} 177 |12 000/17 800} 94] 188 | 275 |Bond at N.end 1046.6 6 2 64 {10 000/13 000) 204 |16 700/24 500] 129 | 263 | 325 |Bond at N.end AY eevee Peetess 62 | 8 000/12 000) 190 |16 200/23 800) 129) 253] 313 1047.1 6 2 62 [11 000/15 000) 237 /21 800/31 800) 167 | 335 | 447 |Bond at N. end 31 | 1047.2 6 2 65 | 6 000/15 000] 233 |23 400/34 100) 179 | 359 510 |Bond at N.end 1047.3 6 2 60 | 6 000/12 500] 196 |20 600/30 100} 158 | 317 | 443 |Bond at N.end ANG. Mhutiaee Ua seat 62 | 7 700/14 200) 222 |21 900/32 000} 168 | 337 | 467 1048.1 6 2 70 13 000/15 000) 258 |31 550/48 000) 239 | 541 | 781 |Bond at N. end 32 | 1048.2 6 2 59 | 8 000/15 000} 258 |27 600/40 300} 209 | 475 | 652 |Bond at N.end 1048.3 6 2 60 | 8 000/18 000] 310 |28 300|41 300) 215 | 486 | 696 |Bond at S. end AVS, Wsget ss eae 63 | 9 700}16 000) 275 |29 100/41 900) 231 | 501 | 710 1048.4} 6 | 2 62 | 8 000/12 000] 206 |28 000/40 800] 210 | 482 | 721 |Bond at S. end 33 | 1048.5] 6 | 2 63 | 8 Q00|14 000] 240 }27 700/40 500} 209 | 477 | 706 |Bond at N.end 1048.6) 6 | 2 64 | 8 o00|17 000] 292 |25 000/36 600/ 191} 431 | 614 |Bond at N.end AX: fvaa cobonea ss 63 | 8 000]14 300) 246 |26 900/39 300) 203 | 463 | 680 1049.1] 10 344 | 61 | 6 000/18 000] 310 |21 000/52 500| 160 | 369 | 587 |Tension in steel 34 | 1049.2; 10 34% | 63 | 8 O0O{L5 000] 258 [21 100/52 700) 161 | 371) 521 |Tension in steel 1049.3f) 10 3% | 400 16 000| 275 |23 200]58 300) 175 | 400 | 619 |Tension in steel * A load of 10000 lb. (which caused a slip of 0.0002 in. at S. end), was maintained for 32 days; the load was then gradually increased until the bar pulled out on the 60th day at 12800 Ib. See Art 90. : 5 ee load of 16000 Ib. (which caused a slip of 0.0002 in. at S. end), was maintained till the bar pulled out. ee Art 90. + Tested at about 89 days. , / + Each load was released before a higber load was applied. See Fig. 64 and 67. ILLINOIS ENGINEERING EXPERIMENT STATION 156 yg 86P 08+ vr $98 T9T 9 89F ang T0r LRE GLO £86 oo vereeecesecccees ss s( aiglas oly OIF 968 soe 90€ IST 9 Top oan ler $88 608 $e eg |j°° °° panoz areyd “at-[ €°6S0T ae as aie ae ae a ed Te ae $08 Lo SEs 861 6¢ ovo" pumod urefd “UI-t 2° 6S01 61 16g 98¢ 9g Soo ler OLT $9 oH9 L29 LLt $88 18¢ SEs 9 “-o*pumnor arefd “Ut-T T° 6S0T £98 8se che T&é 882 OFT 89 eee ea L108 883 1s IFS 99 DEES EATS RE EERE: ee I8é Sle 098 She 80€ Tt TL 6ce 6F8 Tee org 163 $9 L9 “punos weld “ul-T €°8S0T La14 8S 62 Tho SG 611 89 &Lz 193 hS 826 802 861 89 punol uefd “aI-T & 8SOt 8T PSP L¥P Loy SOP See 6S1 99 598 -< LPS Tee BIE GLE 89 “punod ured ‘ui-T T8901 Sok L0E 186 $13 6&3 66 #9 LOF Ai as 868 yee HFS gg fcr) ameaay aoe 80€ 883 GLE OFS 60T #9 8LF as ad 8LF 9LE 0S¢ £9 “punod ated “ui-T €° LOT ae Be a cat nas A y 89E ae coe LE L1€ |. 083 #9 “punol ured “ul-T 3° L901 LI &ZE 208 182 TlZ 6&2 88 9 GLE CLE L498 ese 608 18% 89 “pumod orefd “Ur-T T2901 ely 90% $8E e9€ Ose L9T 69 Tae Pie oe 098 EE 626 €9 verreeeseeececesss ss) OBI9AY 18 Le a9€ Ste 808 Tel IL e9e 698 6S OE 6¢8 692 g9 “77 *pumor ured “uI-T 9°2¢Or STF SOF OLE LhE LOE 6FI TL gce Pon noe 9g 8cE 02% eg |j°7*punos urejd -ar-T $°3S00 or StF Th LIP L6E PE 1a $9 S6E Te coo ¥8e Le LPS {9 or pumor ured “at-f ¥ SSO OF 6lF 968 GLé €08 5 § 99 o98 SS ne 688 608 ase $9 il" po aoe adeiaay chr oer 888 L498 99% 82I 59 £0F he os £0% 8b 09% go “""**punos ured “aI-] €°9S0T 268 88E 698 HSE ag ost 69 618 L9E che FIE 862 OFS cg ||-°**"punoz arefd “ai-T @'9SOT SI oP LbP Lob SOF See 6ST 99 SOE ane rier 008 162 Oss G9 so *pumog wreyd “Ut-T T° 9S0T S6e L8& TLE 6FE 808 98T 99 GLE ee oe Gee S82 FOS €9 PEAS EVENS RES ERE! SARITA, re seer oe noe Byes ion wih se ggg ere ee ase 8EE LOE $9 “'"punog ured *ut-7 €° 3ST cOF 00% rach $98 91g I9T 99 LPE Cer ee 828 00g IIé 6S || °°° panos ured -ar-T @ SOT IL 98g ole 198 Tee 00g 11d 19 £88 — 6FE 162 91% FCT 19 “panos Urey “Ut-T T° 3sor : : : : = lars SO ; : : ; lane aouEzeLEay ¢00 00 100 ¢000 2000 i 'P cauncen i ¢00 200 100 S000 2000 a P inne te seanig puog eh weg 30 ‘on wag |dnog; PEN (soyout) 30 dyg pug 3 sseng puog | ogy || pomnduoy (soyomr) jo dug pug ye sang puog | ogy || Pury pus ezig 5 SLSHL LAO-TTNd SLSGL WVdd “yout arenbs Jad spunod ur wests are sesselg “Il g peppaqure seq WO 8389} say} Jo advIOAe 04} OTe SuAUTTOGdS 4NO-[[Nd oY} JO} WATS Sassel4s PU oy.) “dys 1038012 aq} SULAOYS pus 94} 0} Jajer 8480} TIeEq O44 UT Jeq Jo dys jo syueweMEEeM ayy, “GE PUB TE 89[qGE, 998 ‘8}89} TABAq Bq} UI Speo, oy} Jo TOMIsod oy} puB Meds oy} ZurMIeON0D ByEp 10,7 ‘SHIUTG ZIG6(—SLSH], LAO-TINg ANV SWVag NI GONVISISaY GNOg 40 NOSIUVaWOD ‘88 WIGVL 157 ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL (89I 2bpd uo panuruod $¢ 3190) *1@ Anosx) ul sUIAq aq} JO YOCS Ul pasn aie sIeq MoT y “06 “IV 32g “BAB [VIAAAS JO} PONUTZUOS PEOT » 00s 88h | O9F ( sep | gue | Ize | $9 ggz as ce a go | Zt | to f[ ct aderay £09 cop | 14h | pep | gze | Ost 19 92z eg | S61 | eet | zat | FOr | 09 |] ‘*'sPunor urepd “ui-$g 9° 0SOT 16¢ gig | #9 | 629 | FF | COE | £9 IZ “" 1 gog | gee | €02 | 9ST | 29 sunos ured “ui-84 $"0S0T 13 SOF e6e | 898 | oe | 908 | FIZ | OL 692 Poe A ou Tee | S6r | 09 +punor ated 14 ¥ OSOT ate ose | or | toe (pee. | toe..| #9 oe es 2d Bat is ah oa. || Seti eee oBe22Ay 6re gee | 60g | cez | 293 | 2er | 39 cig ae ‘ Pes ste | oF | $8 ‘punos ured ‘ur-T | 46" SS0T STF ree | IZ | spe | 26% | Zzt | 49 992 e2z | £02 | £02 | 02 | 803 | £6 ‘punos ued “ur-T | 48° SSOT 9% z9g T9¢ | 229 | ze | Gee | FBT | 29 B¥e oy “| ape | 29 “punod ureyd “Ut-{ L°SS0T sig vos | ob | Sep | sce | gat | 69 81z at an nae a cretesesssessess) o8gIgAy 16% csp | 2h | gop | sie | Fr | I8 98% ae “| goz | 808 | 4 @ || 077s punos urerd “ar-y 9°SSOT 66¢ ¥6S | g9¢ | Zoo | SOF | 98 | #9 102 56 oe a ns oer | $9 “punog ureyd “u-T g°Ssor 4 6S sey | s6e | eze | 628 | FEI | £9 SIZ oo ms mae wee siz | 9 “panos ureyd “Ur-] % SSOL Tos eer | 19h | pop | oe | 808 | 2 98% ei ‘ = aa a vreeeseeeceesscess se" Q8RIAV ogg ors | soo | ccm | Foe | 902 | 28 08% et ne soz | “4 2 |} 777: “punox ured “ary €°SS0L m see Bee ee a oe 4 942 i abe : cee 0% | 09 “punod ureyd “UIT 3° SGOT ¥% StF le | lt | 268 | FE | TIS | #9 are ; = em ae ing +9 “punos ureyd “ur-T T'Ssor Lag vos | oLb | 62 | o9e | Sot | £9 £08 os cee Ti = Teg | 19 ff ccc) eBereay 96g oes | 199 | Gos | sib | Hor | #9 T08 : ae eee ae 99% | og || “°°: ‘panos aed -mt-7 & FSOI zg ose | os | 90h | see | GrI | 99 962 ; ae “* 1 gez | 2@% | @9 || 77°) ‘panos meld -at-T 2° PSO1 €&@ 6SF esp | g6e | e246 | 628 | I9T | 09 aIe “| gre | 26% | gz | O1@ | 09 |] ‘°° punos ured “u-T T'¥S0T £8 g9¢ | zee | 616 | 282 | 99f | 29 6ge ee at “ ) oe | e2e | eg ff] eBereay a nee mas ane ae = as ose wea ii | gre | ore | eg. || cco spunos ured “a7 e°Ec0T 00% ese | e9e | ese | eee | 20s | &9 ze ote | oog | 26% |. eze | 982 | 9 |].77***punoa ured “uty Z€s0L a 998 we | soe | 642 | Os% | Tet | 69 188 synema| nega [ON gg |i °°" punog areyd “ui-T TeS0T 807 sse | goe | gee | s6z | oZT | 49 7908 a8 “" ) pee | gre | 223 | FO fp cc) asemay 08% vor | cep | Gor | ove | Fer | 99 ogg see | oze | 968 | aze | O62 | €9 || ©7°*‘puMoz meld ‘U-T 8 1S0T oe gre | ¥6z | Sez | 992 | 92 | 09 £88 bes eae 9 | goe | 262 | 09 || °°**‘punos meld “O-y 2 1S0T 1% 98 Sle 198 Tee | 008 11z | 29 Pee oze | Ite | 608 | oog | 062 89 |] “*** "panos arejd -m-T TIS0T ose ee | zee | aoe | cor | pat | #9 199 ee Bi a isp | ze | #9 [ccccccccccee | eBezay 629 gc¢ | 11g | 29m | see | Sor | 9 Lg “7 | 089 | 88% | erp | 628 | 19 “punoz med “OT-] £0901 669 reg | 99g | 20g | 80P | 98I | #9 LP 09 | Ih | STb | gee | Zhe | 29 “punos med “U-T 30901 0% 16s 98s | Fog | sz Tap | O2T 9 809 en ees id oep | 608 | 02 “punoz ured “at-T T0901 eae S103 soo’ | zoo’ | Too" | s000" | 2000" | *4ep peoyT s00° | zoo" | 100° | goog" | 2000" | esp puog 489], pean, 4a, eae 48 eB - ee (seqour) jo dyg pug ye sang puog | ay || paynduog (eayout) Jo dug pug ye ssong puog | ony puny pus orig ON meg |dnorp SLSHL LAO~-T1Nd SLISHL WVAd ‘dU ONIINO}—Ee ATAVL NGINEERING EXPERIMENT STATION fe ILLINOIS 158 *gz dnosx) Ul suresq 94} JO Yowa UI posn alam seq VIG], 4 18 90s | o9F ize | ave | cer aL ie arate ies sete anh ies by vevtereeeeesersereee] aBgzaay 619 29¢ sig osP goe ToT 6L 00% ve “1 OLE 828 GLB 00F “punod “109 “UI-$4 £°6F0T Teg Fly 068 998 908 OFT 69 we £98 ese 62E 208 893 £9 punod “109 “aI-$4 @' 6701 ve 06g Leo 064 SP gse OFT 39 698 oF oa 698 POE SI 19 “***punod “109 “UI-34T T 6F0T 089 28 Tes osy 68e Sst +9 80F 82 868 198 12 682 89 scewetinarssiielprinstr | (@SBIDAY, ya Lye | 168 | Z98 | 682 | 6 $9 18h lb | Ol | 268 | FE | 162 | 79 ‘punod "109 “UI-$4T 9°8F0I 90L 619 ggg 164 ose 691 $9 LLY Ish SOF ole 128 44 £9 punod “109 “I-84 S° 80 ee 12k ¥69 | 209 | PIS | L468 | 00% | 29 8h gee | we | Bee | 88% | 12 | ZO |} © “*punor -z00 “UI-$4T ¥ 8F01 OTL 829 e9g SlF 608 Sor sg Tos elF Leh 168 L¥E G82 £9 sresesssccsecscsess9) 93eI2AY 969 eL9 409 Ts 08s OFT 99 98F 087 8h 90% FOE sie 09 ****pUumod “109 “UT-$4T &°8P01 ego POP 68h 168 92€ Sst 99 GLP StF SCP 188 0ge £92 6s "+" pumod “109 “UI-84T o°8h01 ee 182 AIL 979 T8¢ 00F OLT #9 Ts 16? 6eF SOF L¥E Ole OL “*¢*pumod “09 “UI$4T T'8P0T 195 SOF (or3 oes 082 6FT 99 188 oes Org 682 #SS 22S oo preeeess se seseee se) Q8BIQAV chr 6LP otF Oly See cg 6g L18 Ore FOE 282 64S 961 09 ‘rarenbs poze} “Ul-] €° LPO ery 988 gee 608 992 TOL TL 6g€ SPE Ste £63 £96 S&S 99 “vazenbe poze} “Ul-T &° LOT 18 Lv 6re | S0f | C42 | 8h | SOT | 69 see oge | 00g | 862 | 6h | 28% | ZO || ‘“exeNbe pozsung “UIT 1240 or 822 | 62 | 823 | FOG | LIT 39 ese ice oo sez | 612 | O6F | ao ff ttt) eBeray pee ¥08 GLE gse S2e SI $9 £96 T¥Z £3 618 $0Z #9 “+ -arenbs ured “Ul-] 9 9F0T $23 9st 623 ZIZ ¥61 1&1 69 881 Ae PLT PLT LL 29 “7 -exenbs ured “Ur-T S°9F0T 0€ FE 062 obs LZ c6r 901 19 808 a 863 £9 681 19 “*-azenbe ure[d “Ol-T F901 e908 Ove sig $82 883 IST 89 08 ¥62 213 &9% 626 681 09 cepeeessesescecsess9) QBBIQAy SOF Ls | eh wg | 622 | SLT | 69 os ove | 862 | $82 | 9% | 612 | 2g |] °** ‘exenbe ureid “ur-T €°9F01 183 133 £9 HOG 808 a4 | 69 863 £63 082 £92 4 FL eo “+ sagenbe ured “aI-T o 9401 62 9e8 908 GLE 896 £82 LsT 09 £62 612 593 8&2 | 02 7A 2 “+ *-arenbs ured “aly T9F0T 29g obs Ig bLP gor 02% £9 Fle ae ee ove as T6T 09 vrteeeecssceecececes) aBBIOAy a9 ee9 | 009 | 9FS | Gem | BZ | 6S 8LZ 692 | 28% | 622 | 00 | PLE | 29 || “4panor ured -ar-%% £OSOE 61g 664 GLY SP £0F ad 19 98% of’: ak L03 00% §8T 2 “**4ptmog areyd “aI-% @'0S0T 82 Fee zos 199 ost LLE 161 2 808 tee 183 693 Seo LT 19 “**4punod areld -a1-% T'OSOT aouezsIsey $00" 200° | 100° | $000° | 2000" | sxep a Lana ae $00" | 200° 100° | 000° | 2000" | sep crate tee ssayg puog vee seg yo : (saqouy) Jo dyg pug 4e seang prog | aay payndmop (seqout) Jo dug pug ye sseNg Puog | ogy Pury pus ezig ‘ON Wweg jdnosp SISAL LNO-TINd SLSGL Wvdd *dHONILNOO—EE ATAVL ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 159 84. Bond Resistance with Twisted Square Bars.—The beams in group 31 were reinforced with one 1-in. twisted square bar. These bars were twisted cold (one twist per lineal foot) from the same stock as was used in the beams reinforced with plain square bars. Three pull- out specimens were made for each of the beams; the blocks were not reinforced against bursting. At an end slip of 0.0005 in. the bond stresses were 254 and 266 lb. per sq. in. for the beam tests and pull-out tests, respectively. Table 36 gives a summary of the bond stresses developed in the tests with twisted square, plain square, and plain round bars in the 1912 series. For the two stages of the tests included in the table the twisted bars give values about 20% higher than the plain squares in the beam tests and about 34% higher in the pull-out tests. The values for the twisted bars are about 12% lower than the plain rounds in the beam tests; in the pull-out tests the twisted bars are 8% lower than the plain rounds at a slip of 0.001 in. and show about the same difference at the maximum. The maximum load for the pull-out tests with the square twisted bars came at a slip of about 0.1 in. and was accompanied by the bursting of the concrete blocks while the maximum load for the plain rounds came at a slip of about 0.01 in. ‘The bond stresses for the square twisted and the plain round bars at a slip of 0.01 in. were 407 and 408 lb. per. sq. in., respectively. These values are about the same as the maximum bond stresses for the plain rounds in the 1909 pull-out tests (see Table 14). Load-deflection and load-slip curves for these beams are given in Fig. 85. It is seen that after end slip of about 0.001 in. is developed the action of the beams with twisted bars is quite similar to that in beams with plain square bars. Beam No. 118, series of 1909 (see Fig. 77), reinforced with one 1-in. twisted square bar acted in much the same way as the beams just discussed, except that after an end slip of 0.004 in. the bond re- sistance began to increase and a second maximum load about 3000 lb. higher than the first was carried. This phenomenon is similar to that observed in pull-out tests of twisted square bars as described in Art. 41. 85. Bond Resistance with Corrugated Bars.—Six 6-ft. beams and three 10-ft. beams were reinforced with 11,-in. corrugated rounds (groups 32, 33 and 34). All the 6-ft. beams gave bond failures; the 10-ft. beams failed in tension. 160 ILLINOIS ENGINEERING EXPERIMENT STATION TABLE 34. Summary or Bonp Stresses IN Bream anpD Putt-ouT TEsTs— 1912 Surrus. Stresses are given in pounds per square inch. A Bond Stress at End_Slip of Computed No. | “8° (inches Bond Group Characteristics of | Test ‘Stress at Tests 4 1 398) ooo 01s 00 002 | .005 | or | Loa Beam Tests; 1-in. Plain Rounds. 14-16] 6-ft.span; Loads 2 ft. apart......... 9 64 369 17 6-ft. span; Loads 2% ft. apart 3 65 407 18 6-ft.span; Loads 3 ft. apart.. 3 66 332 19 6-ft. span; Loads 34 i apart. . 3 62 468 20 6-ft. span; Loads 4 ft. apart 3 64 551 21 5-ft. span; Loads 13¢ ft, apart......... 3 64 364 22 7-ft. span; Loads 2} ft. apart 3 63 336 23 8-ft. span; Loads 22¢ ft. apart. . 3 62 308 24-26 | 10-ft. span; Loads 33¢ ft. apart T levees 230 26 | 10-ft.span; Loads 6 ft. apart......... 2 89 285 Pull-out Tests; 1-in. Plain Rounds. 14-26 | Lowest two from each set of three tests.°....] 54 |..... 144 | 306 | 362°] 389 | 410 } 417 426° All pull-out testa................ 408 | 434 | 447 448 Highest from each set aed 450 | 480 | 491 494 Lowest from each set..............-2.-0005 368 | 384 | 392 399 27 «(| All beam tests.............. 0.00000 e eee 3 61 | 172 | 205 | 285 |.....).....]..... 255 Pull-out Tests; 54-in. Plain Rounds. 27 | Lowest two ian each set of three tests”. 6 337 | 394°) 424 | 455 | 461 467° All pull-out tes - 9 375 | 435 | 460 | 490 | 490 500 Highest from oak set. 3 452 | 505 | 531 | 560 | 561 565 Loweat from each set.. 3 293 | 345 | 363 | 400 | 402 415 Beam Tests; 34-in. Plain Rounds. 28 | All beam tests.................0..0. eee ee 3 60 | 186 | 212 | 242 |.....].....]..... 274 Pull-out Tests; 84-in. Plain Rounds. 28 | Lowest two from each set of three testa’.....| 6 |..... 179 | 390 | 455°) 490 | 519 | 536 537° All pow: PSt8e. ic gesees or se Raeeae: 9 63 | 220 | 403 | 474 | 518 | 545 | 558 562 Highest from each set...............0.0005 Bi Hecieurs 301 | 430 | 510 | 559 | 595 | 604 611 Lowest from each set............0-...00 00s Bs lessees 125 | 365 | 441 | 476 | 510 | 524 524 ° Used for comparison with bond streases in reinforced concrete beams. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL TABLE 34—ConrTINvuED. 161 ‘Age Bond Stress at End Slip of Computed oz, No. | “at (inches) Bond Group Characteristics i Test Sizes at SEE tava Maximum 0002] .0005].001 |.002 |.005 | .01 Load Beam Tests; l-in. Plain Squares. 29-30 | All beam tests...........000...00002 2c 6 61 | 190 | 224 | 248 |..... lena | 278 Pull-out Tests; 1-in. Plain Squares. Lowest two from each set of three tests®.....| 12 |..... 115 | 209 | 241°] 266 | 294 | 309 324° 29-80 All pull-out tests...........0.....00. 18 64 | 134 | 221 | 256 | 281 | 312 | 328 340 Highest from each set Go sao 171 | 244 | 285 | 312 | 367 | 366 372 Lowest from each set...................-.. Ge ees. 104 | 194 | 220 | 241 | 266 | 283 298 Beam Tests; 1-in. Twisted Squares. 31 | All beam tests......... 0... ce cece eee eee 3 61 | 222 | 254 | 289 | 310 | 330 ]..... 337 Pull-out Tests; 1-in. Twisted Squares. Lowest two from each set of three tests®.....] 8 |..... 131 | 266 | 325°! 351 | 386 | 407 428° 31 All pull-out tests.........0....-... 0020s eee 12 65 | 145 | 278 | 330 | 362 | 403 | 433 467 Highest from each set.............-...05. 4 ....| 167 | 311 | 348 | 407 | 459 | 493 543 Lowest from each set...............-.200- 4 113 | 252 | 303 | 325 | 363 | 382 405 Beam Tests; 114-in. Corrugated Rounds. 32-33 | 6-ft. span; loads 2 ft. apart. . 6 63 | 261 | 334 | 382 | 418 | 458 | 465 482 34 | 10-ft. span; loads 34 ft. apa: 2 62 | 293 | 385 | 849 ].....].....h..... 370 Pull-out Tests; 114-in. Corrugated Rounds. Lowest two from each set of three tests®.....| 18 |..... 142 | 332 | 429°] 490 | 548 | 571 630° 32-34 All pull-out tests.............-....5. ..| 27 67 | 154 | 353 | 452 | 516 | 574 | 595 654 Highest from each set. ace OP Necks tee 177 | 394 | 498 | 570 | 625 | 639 698 Lowest from each set..............2.000005 9. fe sexs 132 | 304 | 407 | 470 | 522 | 546 617 Pull-out Tests; 1-in. Rounds with Standard V-shaped Threads. Lowest two from each set of three tests......| 16 |..... 181 | 380 | 470 | 535 | 592 | 613 877 All pull-out tests.............--..0006- ..| 24 66 | 206 | 410 | 505 | 580 | 642 | 685 725 Highest from each set... 8 Jews 254 | 473 | 576 | 670 | 738 | 757 813 Lowest from each set..........-..+--.0-05- Br fecacag 177 | 351 | 431 | 488 | 530 | 551 614 ° Used for comparison with bond stresses in reinforced concrete beams. 162 ILLINOIS ENGINEERING EXPERIMENT STATION Load-deflection and load-slip curves for the 6-ft. beams are given in Fig. 86. The ends of the bar in the 6-ft. beams give a slip of 0.001 in. at a computed bond stress of 382 lb. per sq. in. Beams with plain round bars give a bond stress of 345 lb. per sq. in. at the same amount of slip. The maximum bond resistance of the 6-ft. beams was 482 lb. per sq. in. as compared with 369 Ib. per sq. in. for similar beams with l-in. plain rounds. The load at an end slip of 0.001 in. was 79% of the maximum for the beams with corrugated bars and 94% for the beams with plain round bars. The loads at first outer cracks are about TABLE 35. Bonp ResiIsTaNcE IN 6-FT. BeaMs REINFORCED witH Puain Rounp Bars. Each beam was reinforced with a single round bar 1 in. or 114 in. in diameter. All beams were loaded at the one-third points of the span. Stresses are given in pounds per square inch. i Com- Number| Age Bond Stress at Slip Macinum ranaive: Series | Groups| _ of at of End of Bar of Bond Strength Tests Test Resistance of 6-in. 0002 in. -0005 in. .001 in. Cubes 1909 j........ 6 100 days 202 244 265 291 1575 1911 | 1-5,9 19 8 mo. 240 306 335 377 3030 1912 | 14-16 9 63 days 251 312 345 369 2630 Mean hess gdys [eeean' ss leateewrneey 236 295 324 356 pie the same for the 6-ft. beams reinforced with corrugated bars as for other beams tested on the same span, whether plain round, plain square, or corrugated round—about 9000 lb. The load at first end slip of bar averaged 15 200 lb. for the beams with corrugated bars and 13100 lb. for the beams with plain rounds. The average bond stress at first slip of end of bar was 261 lb. per sq. in. for the corrugated bars and 251 lb. per sq. in. for the plain rounds. It ig seen that the vertical shear- ing stresses developed at first outer crack noted are about the same for the two forms of bar, regardless of the difference in area of bars. At the maximum load the bond resistance of the beams reinforced with corrugated bars was about 30% higher than those with plain rounds. The beams tested on a 10-ft. span showed end slip to begin at about the same bond stress as in the 6-ft. beams. The 10-ft. beams reinforced with corrugated bars showed cracks much nearer the supports than were found in the beams with plain bars; see Figs. 59¢ and 63d. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 163 86. Effect of Span Length.—Besides the beams tested on a 6-ft. span, a group was tested on each of the following spans—d, 7, 8, and 10 ft. All the beams were reinforced with 1-in. plain rounds and were loaded at the one-third points of the span. Load-slip curves are given in Fig. 71 and 72. See the photographs in Fig. 63 for the appearance of these beams after failure. Nearly all the beams with spans of 7, 8 and 10 ft. failed in tension and hence did not develop their maximum resistance to bond stresses. Notwithstanding this, the longer beams give data on the slip of bar at ends and at intermediate points and on the deflections and general behavior of such beams. TABLE 36. Bonp RESISTANCE WITH TWISTED SQUARE, SQUARE, AND Rounp Bars 6-ft. beam tests from series of 1912; see Tables 31, 32, 33 and 34. Stresses are given in pounds per square inch. Reinforced ee Beam Pullout Testa* Size and Kind : of: Bar Number At End At Number At End At of Slip of Maximum of Slip of Maximum Tests 0.001 in, Load Tests 0.001 in. Load 1 in. twisted square 3 289 337 8 325 428 1 in. plain square 6 248* 278" | 12 241* 323° 1 in, plain round 9 345 369 54 362 426 * As stated in Art. 73, the bond stresses considered in comparing the pull-out tests with the beam teste are the averages obtained after rejecting the strongest specimen from each set of three tests. See Table 34. * Beam No. 1046.5 gave abnormally low results. Omitting this test, the values for the beam tests with plain square oars are 283 and 295 Ib. per sq. in., and for the pull-out tests, 265 and 355 Ib. per aq. in., respectively. The computed bond stresses at first slip at the end of the bar were as follows:—5-ft. beams, 277; 6-ft., 257; 7-ft., 273; 8-ft., 231; 10-ft., 222 lb. per sq. in. If correction is made for the added embedded length due to the 3 in. overhang of the ends of the beam the following values of bond stress at first end slip are obtained :—5-ft., 235; 6-ft., 222; 7-ft., 244; 8-ft., 209; 10-ft., 206 Ib. per sq. in. The stresses developed at the maximum loads were:—5-ft. beams, 364; 6-ft., 369; 7-ft., 336; 8-ft., 303; 10-ft., 230 lb. per sq. in. The last three values, however, do not measure the maximum bond resistance, since the beams failed by tension in the longitudinal steel. It is seen that there is not much difference in the computed bond stresses for beams of the span lengths tested, during the early stages 164 ILLINOIS ENGINEERING EXPERIMENT STATION of the tests. This indicates that the distribution of bond stress is similar in beams of these lengths during the early stages of the test. However, it should be borne in mind that these computed values for bond represent only the average bond stress and that they give no in- formation as to the distribution of the bond stress or the actual bond stresses developed at various points. The discussion in Art. 68 and 95 indicates the distribution of bond stress in beams of this kind. 8%. Effect of Position of Loads on Beam.—The beams in groups 17, 18, 19 and 20 (span 6 ft.) were tested in order to study the effect of the position of the loads with respect to the beam supports. The beams in groups 21, 22 and 23, which were tested by loads applied at the one-third points of spans varying from 5 ft. to 8 ft., will be in- cluded in this discussion. All these beams had the same cross-section— 8 in. wide and 10 in. effective depth—and were each reinforced with a single 1-in. plain round bar. The position of the loads with respect to the supports may be expressed in terms of the ratio of the effective depth of the beam, a, to the horizontal: distance from the support to the load point, b: This ratio for the beams included in the discussion varied from 1.2 to 3.2. To make the comparison more nearly accurate, it will be desirable in the calculations to take into account the length of embedment of the bar beyond the point of support, which in all the beams was 3 in. To do this it is assumed that the full calculated tensile stress of the bar at the load point is taken off by a bond stress which is uniform in intensity from the load point to the end of the bar. The values so calculated will be less than the computed bond stresses given in Table 32. Fig. 50 has been plotted with the average computed bond stress for each group of beams as ordinates and with the ratio b/a as abscissas. The closed circles represent the bond stresses developed at first end slip of bar; and the open circles represent the maximum bond stresses. It will be seen that the computed bond stress at beginning of end slip is nearly the same throughout the range of the tests. This confirms the conclusion stated in the preceding paragraph in the discussion of another series of tests, that during the early stages of the test the bond stress was probably distributed in a similar manner in these beams in which the loads were applied at different points on the span. The computed bond resistances at the maximum load are greater for the smaller values of the ratio b/a; except for group 19 and 20 this differ- ence is not large. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 165 A study of the position of the cracks in these beams shown in Fig. 60 and 61 gives no indication that the location of the cracks is pro- ductive of these differences in the maximum bond resistance developed by the loads in different positions. Added friction due to the increased bearing pressures obtained with the beams having smaller values of b/a may increase the bond resistance somewhat, but this would be expected to affect the static friction element of bond at the beginning of slip and there is no evidence of such effect. It would seem, however, that the variation must be due to the presence of other than normal beam action. a we taeto" T een _b=l2"10 32" | 4 | Computed Bond Stress-Lb. per Sg. in. sat © Bond Stress af First End Slip ° Stress at Maximum Load / 2 b 3 4 Ratlo 2 Fic. 50. Errrct oF Position oF Loaps oN REINFORCED CoNCRETE BEAMS. 88. LHffect of Auailiary Reinforcement in Ends of Beams—In addition to the main longitudinal reinforcement, consisting of 1-in. plain round bars, the beams in Group 16 were provided with two loops of 3%-in. round bars at each end extending from a point 3 in. inside the load points to the ends of the beam. The arrangement is shown in Fig. 2 (g). This increased the tension reinforcement outside the load points by 56%. The purpose of these tests was to discover the effect produced on the bond resistance as a result of retarding the opening of the usual outer cracks. The load-slip curves for these beams are shown 166 ILLINOIS ENGINEERING EXPERIMENT STATION in Fig. 69, and the load-deflection curves in Fig. 81. Group 16 may be compared with Groups 14 and 15, which were the same except they did not have the auxiliary bars. The load at first crack outside the load points in Group 16 averaged 11°300 lb. as compared with 9100 lb. for Groups 14 and 15; an increase of 24%. ‘The bond stress at first end slip of bar, computed on the area of the main bar only, is raised from 240 to 282 Ib. per sq. in., an increase of 17%. The maximum bond resistance, based on the main bar only, is practically the same for each of these groups of beams. The photograph of these beams after failure (Fig. 60b) shows a different distribution of the cracks in the beams having auxiliary bars. Generally speaking, the main cracks came inside the load points near the end of the auxiliary bars. The characteristic sloping cracks of such beams are generally absent, although it seems probable that interior cracks which did not appear on the surface may have opened to an extent that permitted the same distribution of bond stress as was found in the other beams. The most striking effect arising from use of the auxiliary bars is found in the uniformity and regularity of the slip of bar at intermediate points of the bar as shown in Fig. 69 and discussed in Art. 94. 89. Effect of Variation in Number and Size of Reinforcing Bars. —Four %%-in. plain rounds were used for longitudinal reinforcement in Group 27 and three 34-in. rounds in Group 28. This corresponds to 1.53% and 1.66% of longitudinal steel, respectively, as compared with about 1% for the 1-in. plain rounds used in most of the beams. These beams were provided with U-shaped stirrups as shown in Fig. 2(h). The measurements of end slip were made on the two outside bars. There was a marked uniformity in the behavior of the two out- side bars in nearly all the tests; generally the two instruments at the same end of the beams gave exactly the same reading; in other cases the larger value was used in plotting the load-slip curves in Fig. 84. The relation of the bond stresses developed in the beam and pull-out tests may be seen in Table 34. In the beam tests the 5¢ and 34-in. bars show end slip to begin at a much lower computed bond stress than for the beams with 1-in. bars, but it is worth noting that this first end: slip came at loads which gave the same tensile stress in the steel at mid- span and hence presumably about the same amount of stretch of bar between load point and support. For an end slip of 0.001 in. (using the values in Groups 14 to 16 for the 1-in. bars) the average computed ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 167 bond stresses are: 5g-in. bars, 235 |b. per sq. in.; 34-in. bars, 242; 1-in, bars, 345 Ib. per sq. in. At the maximum loads the values are: 255, 274 and 369 Ib. per sq. in., respectively. From this it would seem that slip is a function of amount of stretch of bar as well as of bond surface. The loads at first outer crack are about the same for each of the groups of beams. The concrete may be expected to fail in tension at about the same load, and it seems that the anti-stretch cracks form at once. The appearance of one group of beams after test is shown in Fig. 62a. By comparing the photographs of the beams in Group 27 with Group 14, it will be seen that the progress of these outer cracks may assist in explaining the difference in the bond stresses developed. For the beams reinforced with 1-in. plain rounds (Groups 14 and 15), the average distance of the outermost cracks which had appeared pre- vious to the first slip of bar at the end where bond failure occurred, was 18 in. from the end of the beams; in Group 27, reinforced with four 5-in. rounds, the corresponding distance is 11 in.; in group 28, reinforced with three 34-in. rounds, 13 in. A consideration of the position of these outer cracks indicates that after such cracks are formed in beams of this length a large proportion of the bond stress is carried by the embedded length lying between the crack and the end of the beam, while a smaller proportion is carried by the part between the crack and the load point. This assumption makes the bond stresses of the outer embedded length more nearly the same for the three groups. It seems probable that the actual bond resistance developed in the outer embedded lengths approaches the values developed in the pull-out tests. After cracks have formed in the concrete near the load points, it may be expected that an increasae of load will produce other cracks still nearer the support due partly to anti-stretch slip. This cracking progresses step by step toward the end of the beam until the remaining unbroken embedment of the bar is no longer able to furnish necessary resistance, and the beam finally fails by the bars pulling out. ‘The progress of the formation of cracks is very well illustrated in th beams of Groups 27 and 28. In Beam No. 1050.2 a crack opened about 3 in. outside the N. load at 8000 lb.; another appeared 5 in. farther out at 12 000 lb.; and a third at 16.000 lb. In Beam No. 1050.3 cracks appeared at the 8. end at 14000, 18 000, 29000 and 31000 lb. These tests indicate that the distribution of bond stresses in beams is influenced by the relation of the dimensions of the cross-section of the beam to the diameters of the reinforcing bars. 168 ILLINOIS ENGINEERING EXPERIMENT STATION 90. Effect of Repeated and Continued Loads on Beams.—In test- ing Beam No. 84 of the 1909 series, reinforced with one 114-in. plain round, the load was applied progressively as in other tests, until one end of the bar had slipped about 0.0002 in. under an applied load of 15 000 lb. The load was then released to 500 lb. The residual center deflection was 0.015 in. (see Fig. 51). At first loading, 6400 Ib. load was required to produce this deflection. The load was then increased to 17000 lb., causing a slip at the north end of the bar of 0.0006 in. Upon releasing the load to 900 Ib. the residual deflection was over 0.02 in. There was no recovery in the amount of the end slip of the bar upon release of the loads, although the bar had slipped an appreciable amount. The load was then reapplied up to a maximum of 20 500 |b., when the bar pulled out at the end which at first had shown the smaller slip. This phenomenon was observed in a few tests in the later series, showing that the bond resistance at the two ends was so nearly equal that failure was likely to come at either end, or at both ends at the same time. In most of the beam tests slipping was much more pronounced at one end than the other. The lack of recovery in the slip at the end of the bar in this beam test and in the repeated load tests on pull-out specimens is significant, and seems to be characteristic of plain bars. It will be seen in the test of Beam No. 1049.3 that corrugated bars do show some recovery of slip upon release of load. Beam No. 1055.9, reinforced with one 1-in. plain round bar, was loaded on a 10-ft. span at two points 6 ft. apart. The test progressed as usual until a load of 16 000 Ib. (bond stress, 315 lb. per sq. in.) had been applied, which caused a slip of 0.0004 in. at the south end of the bar. This load was maintained until failure occurred after about 34 hours. The changes in slip of bar with the time are plotted in Fig. 52. As was seen in the pull-out tests, after slipping becomes general, there is nearly a constant difference between the total slip at the end of the bar and at point (3). The slip at (1) was probably affected by the large diagonal crack which opened at that point. Beam No. 1039.3, reinforced with one 1-in. plain round bar, was loaded at the one-third points of a 6-ft. span. This beam was 714 months old at the time of test. A load of 26000 lb. caused a slip of 0.001 in. at the north end. This load was maintained constant until failure occurred after 66 hours. The end slip during the period the beam was under load is shown in Fig. 52. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 169 Beam No. 1055.8, 10-ft. span, reinforced with one 1-in. plain round bar, was loaded at two points 6 ft. apart. The. load was increased as usual until the north end of the bar showed a slip cf 0.0007 in. under a load of 10000 lb. The load was maintained at 10 000 lb. for 32 days, perSq.in. Beam No. 84 1-2-4 Concrefe Age 33 morths One lg -in, Plain Round Applied Load -Pounds i 5 2 3 ° O Ol 002 003 004 005 006 007 .008 009 O10 Ol Ol2 O13 04 O15 016 Slip of End of Bar-Inches HW 33 14 IS 16 Ol e@ 03 04 O05 06 07 oF JO dl eae oe Center Deflection -Inches Fic. 51, Loap-stip anp LoaD-DEFLECTION CURVES FOR A BEAM UNDER REPEATED Loaps. during which frequent observations of slip of bar and deflection were taken. After 32 days the load was increased 100 lb. per day until failure occurred on the 60th day of loading under a load of 12 800 lb. The slip of bar and the center deflections are plotted in Fig. 53. Slip- 0 Slip of Bar-inches ILLINOIS ENGINEERING EXPERIMENT STATION ole O10 Beam No.1055.9 008 /-2-4 Concrere, A HY S. Ge ‘in, Pain Round. ThIs bearn wes loaded upto 006 16000 1b, [band stress 315 1h per sg.) which caused a slip of 00004: M2. ot the Send. This load was main- Jained until the bar pulled our 004 oO 0140 0120 Beam No. 10393 1-2-4 Concrete. Age 73 months. 2100 One /-in, Plain Round. This bear was loaded up to 26000 la bone 317655 406 [ape per 23) which caused a slp of 0.001 10, ai Nend. This load wes maintained until 0080 the bar pulled our oO 6 2 1a 24 30 36 ae 48 sa €o Duration of Load -Hours Fic. 52. Stip of Bar in BEAMS UNDER Constant Loaps. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 171 0260 0240 0200 0180 0160 25 30 » Duration of Load uv £ & = 0140 5 center 5 2 © Q ol20 <= — Vv yn £ 0100 _ Center Deflection — Beam No /055.4 1-2-4 Concrete) 4 Age 93 days ak begiprir Of lest. oh ¢ One /-in, Plain Round, Span loft; Loads 6/7. aqparr 0020 | 7000 Increased 100 la per aay. a ie Moartinutn Load (2800/b. 10 20 30 40 50 60 Duration of Load -Days Tic. 53. Strep or Bar AND CENTER DEFLECTION OF BEAM No. 1055.8. 172 ILLINOIS ENGINEERING EXPERIMENT STATION ping continued at a nearly constant rate while the load was maintained at 10000 lb. There was no marked change in the rate of slip until the load had been increased by about 1000 lb. Fig. 54 shows the load-slip curves for Beam No. 1055.8. The curves have been plotted from the same origin in terms of the slip at the points where measurements were taken and the computed bond stress during the later stages of the test. The small diagram in Fig. 53 indicates the ratio of the end slip to the slip at points 6 in. and 12 in. from the end. When the load was first applied the slip of bar at the end was 70% and 40% of the slip at points 6 and 12 in., respectively, from the end; while near failure the slip at inner points was nearly the same as at the end. The slip- of-bar curves indicate that, after slip became general, there was a nearly constant difference between the end slip and the slip at inner points 8 a 6 8 Beam No. 10556 Bond Stress -Lb perSq.in, a 8 ° 005 oo oS 220 025 030 Slip of Bar -Inches Fic. 54. Loap-stip Curves ror Beam No. 1055.8. where the concrete was unbroken. It seems probable from the way slip was progressing in this test that failure would have been produced under a load of 10000 lb., if this load had been maintained for two or three months. The center deflection of the beam increased about 0.11 in. during the period of about 40 days in which the load was practically constant. This increase in deflection was probably due principally to the slip of the bar at the end of the beam where bond failure occurred. This slip amounted to about 0.007 in. at the south end after the beam had been under load 40 days. The tests described above in which the load causing a very small amount of slip at the end of the bar was maintained constant indicate that with beams reinforced with plain bars the load which produces a ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 173 small end slip will, if acting indefinitely, endanger the permanency of the beam. 91. Pull-out Tests with 1912 Beam Series——A summary of the results of the tests on the pull-out specimens made with the 1912 beams is given in Table 34. The load-slip curves have been plotted in Fig. 55. The tests were made at the age of about 65 days. The number of specimens of each kind is indicated by figures in parentheses. In these curves the average values from all the tests have been used; while for comparison with the bond stresses developed in the correspond- ing beams the average of the values from the lowest two of each set of three pull-out tests should be used. The 1-in. plain rounds averaged 448 lb. per sq. in. at the maximum ; the 5g-in. rounds about 11% higher, and the 34-in. rounds about 25% higher. The 1-in. plain square bars gave values which are about 75% (range 69% to 78%) of those for 1-in. rounds. In the tests with 1-in. twisted square bars end slip began at a bond stress of 278 lb. per sq. in—about a mean between the values for the 1-in. plain rounds and the 1-in. plain squares. At an end slip of 0.02 in. the twisted bars gave a bond resistance of 448 lb. per sq. in., which is about the same as given by the plain rounds at their maximum at an end slip of 0.01 in. This is a somewhat better showing than was found for the twisted square bars in the 1909 series of pull-out tests (see Art. 41). After a slip of 0.02 in., there was very little increase in the bond resistance of the twisted square bars, which reached 467 lb. per sq. in. at an end slip of 0.1 in.; this has been considered the maximum bond resistance. The concrete blocks split at a slip of about 0.1; they were not reinforced against bursting. The 114-in. corrugated rounds gave an average bond resistance of 452 lb. per sq. in. at an end slip of 0.001 in.—about 20% higher than the 1-in. plain rounds at the same slip. At an end slip of 0.01 in, corresponding to the maximum bond resistance of the plain rounds, the corrugated bars gave 595 lb. per sq. in——33% higher than the 1-in. plain rounds. The blocks split in all the corrugated-bar tests; splitting seemed to have begun soon after a slip of 0.01 in. was developed. 1-in. round bars, with standard threads (8 per inch), showed the highest resistance at beginning of slip; at a slip of 0.001 in. the bond resistance was 505 lb. per sq. in., as compared with 377 for the 1-in. rounds and 452 lb. per sq. in. for the 114-in. corrugated rounds. At an end slip of 0.01 in. the threaded bars gave a bond resistance of 685° 174 ILLINOIS ENGINEERING EXPERIMENT STATION lb. per sq. in. These bars split the blocks soon after developing an end slip of 0.01 in. 92. Bond Resistance in 1912 Beam Tests and Pull-out Tests.— The summary in Table 34 shows the relation of the bond stresses for the beam and pull-out tests for the 1912 series. The dimensions of the specimen and the embedded length may be expected to affect the bond resistance corresponding to different end slips, hence it will be of interest to compare the results of the beam and pull-out tests at the same amounts of end slip. In this comparison the average values from the lowest two specimens from each set of pull-out tests was used. 600 8 68 UE Bond Stress -Lb per Sq.|n. 8 3 200 1-2-4 Corcrere. Lmbedmert Bir, Age 65 Aay5s. 100 o Q al 02 .O3 05 .06 o7 08 o9 40 04 Slip of Bar-Inches Fic. 55. Loap-stip Curves ror PuLt-ourt Tests rrom 1912 Beam SERIES. The 6-ft. beams reinforced with 1-in. plain rounds, and loaded iv the usual way (Groups 14-16), showed about the same resistance to slip during the early stages of the tests as was found in the pull-out tests. At an end slip of 0.001 in. the values were:—9 beam tests, 345 lb. per sq. in.; 54 pull-out tests, 362 Ib. per sq. in. Ata slip of 0.002 in. the pull-out tests gave a bond resistance of 389 lb. per sq. in.; the maximum bond resistance in the beams, which came at an end slip of about 0.002 in., was 369 lb. per sq. in. For the tests with 1-in. plain rounds the bond ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 175 stress at an end slip of 0.001 in. was 85% of the maximum in the pull- out tests and 94% of the maximum in the beam tests. For the 5-in. plain rounds at an end slip of 0.001 in. the beam tests gave a computed bond resistance of 235 lb. per sq. in.; the pull-out tests» 394 lb. per sq. in. For the 34-in. bars the corresponding values were: beam tests, 242 lb. per sq. in.; pull-out tests, 455 lb. per sq. in. For the 5g-in. bars the maximum computed bond stress in the beam tests was 55% of the maximum bond resistance in the pull-out tests; for the 34-in. bars, the corresponding value was 51%. The wide differences in the computed bond stresses in the beams reinforced with 5g and 34-in. bars and the corresponding pull-out tests are due to the changes in the dis- tribution of the bond stresses in the beams as the test progresses, as was pointed out in Art. 68, and to the differences in the relative dimen- sions of the specimens. The 1-in. plain square bars gave about the same relative values for the beam and pull-out tests as was found for the 1-in. plain rounds; at an end slip of 0.001 in. the bond stresses were: 6 heam tests, 248 Ib. per sq in.; 12 pull-out tests, 241 lb. per sq. in. The bond stress at an end slip of 0.001 in. was 75% of the maximum in the pull-out tests and 89% of the maximum in the beam tests. 93. Variation in the Results of Cube and Pull-out Tests——The cube and pull-out tests made of 1-2-4 concrete in the 1912 beam series offer an opportunity to study the variation which may be expected in a large number of tests made under conditions which are fairly uniform. These -specimens were made from 44 batches of concrete and covered a period of about 3 months. This discussion includes 129 cube tests and 57 pull-out tests. Table 37 shows the average strength and the variation in strength for the individual cube tests and for the averages of the groups of three tests for hand-mixed and machine-mixed concrete. It will be noted that for the compression tests the mean variations of the individual tests and of the group averages are about the same for both methods of mixing. The higher strength of the machine-mixed con- crete is in accordance with the usual experience, but the favorable show- ing as to the variation of the hand-mixed concrete is contrary to the relation usually assumed. Table 38 shows values for the pull-out tests of 1-in. plain rounds embedded 8 in. at an end slip of 0.001 in. and for the maximum load. In this table only the machine-mixed specimens have been considered. The values for a slip of 0.001 in. and for maximum load show about the 176 ILLINOIS ENGINEERING EXPERIMENT STATION same variation. The pull-out tests show a greater mean variation from the average than the cube tests (16% and 10%, respectively) ; this justi- fies the practice of making five pull-out specimens in a set. However, it is instructive to note that the mean variations of the individual tests and of the groups for both cube tests and pull-out tests do not differ materially. TABLE 37. VARIATION IN THE COMPRESSIVE STRENGTH OF 6-IN. CONCRETE CUBES. 1-2-4 concrete from the 1912 beam series. Age at test. about 65 days. Hand-mixed Concrete Machine-mixed Concrete Average Mean Average Mean Crushing Variation Crushing Variation Ttem Moabes Strength, Honbee Strength, Testa Ib. per Ib. per Per Tests Ib. per Ib. per Per sq. in. sq. in. cent sq. in. sq. in. cent Individual tests 36 2200 204* 9.2 93 2800 294+ 10.5 Average of seta of three tests 36 2200 194* 8.8 93 2800 281* 10.0 + Plus or minus, TABLE 38. VARIATION OF THE VALUES OF Bonp RESISTANCE IN PULL-OUT TESTS. Tncludes 57 tests with 1 in. plain round bars from the 1912 beam series. 1-2-4 machine-mixed concrete. Embedment 8 in. Age at test, about 65 days. Stresses are given in pounds per square inch. . At End Slip of 0.001 in. At Maximum Bond Resistance Mean Mean Item Highest| Lowest|Average Naveen Highest} Lo west|/Average Variation Ib, per Per Ib. per Per Bq. in. cent sq. in. cent Individual tests 548 213 371 55* 14.8 690 225 440 73* 16.6 Average of sets of three tests 509 241 371 49* 13.2 599 254 440 66* 15.0 + Plus or minus. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 177 d. General Discussion of Reinforced Concrete Beam Tests. 94. Slip of Bar at Internal Points in Reinforced Concrete Beams. —In many of the beams in the 1911 and 1912 series, measurements were made to determine the amount of slip in the reinforcing bar at points other than the ends. It has been found that these load-slip rela- tions, in view of the information derived from the pull-out tests, give valuable indications as to the bond stresses developed at different points along the length of the bar. Load-slip curves for a few representative tests are given in Fig. 69 to 76. The curves in the figures are plotted in such a way as to indi- cate the amount and direction of movement of bar with respect to the adjoining concrete; they are plotted to the right or left with respect to the vertical line on the diagram which lies upon or nearest to the given point. Each horizontal division in the figures represents a slip of 0.001 in. The position of the loads, reinforcement, cracks, and other general features of the beams are indicated in the figures. The numbers within circles in the diagrams and on the photographs of the beams after failure are opposite the positions of the instruments during the test. A survey of the load-slip curves will confirm the following observa- tions: Slipping of the bar through the concrete was a phenomenon of all the beam tests in which observations were made. The slip was not much influenced by the form of the reinforcing bars or by variations in the points of application of the load. Slip-of-bar was quite pronounced over the middle region of the beams at loads well below those causing the appearance of the first visible cracks in the concrete; this probably indicates that tension cracks were present in the concrete some time before they became visible on the whitewashed surfaces of the beam. Since there is no bond stress due to beam action in this region of the beam, it is apparent that this is what has been termed anti-stretch slip in Art, 68. The load-slip curves for points on the beams outside the region affected by cracks are regular and show that slip increases continuously under an increasing load, or with the lapse of time under load; the curves for points within the region affected by cracks are in- fluenced by the proximity of cracks in the concrete, and show frequent irregularities and reversals in direction. Slip-of-bar is greatest in the immediate vicinity of a crack. The indicated direction of slip in the vicinity of a crack depends upon which side of a crack the instrument was 178 ILLINOIS ENGINEERING EXPERIMENT STATION carried. The instruments on the near side of cracks somewhat removed from the middle of the beam and those approximately midway between the cracks near the middle of the beam generally show little or no slip of the bar. In general slip-of-bar is greatest on the far side of cracks near the load points. Slip-of-bar progresses from the load points or from the outermost cracks to the end of the beam at a rate nearly propor- tional to the increase of load. The rate of movement of the bar depends on various conditions, but it is approximately constant throughout this length after the end of the bar has slipped about 0.001 in. It was found in these tests that the small openings in the concrete which were made in the bottom of the beams to admit the plug against which the extensometer worked, invited tension cracks, except near the ends of the beams. These cracks frequently formed directly under the plate carrying the extensometer and no doubt affected the character of some of the curves. In some cases it was impossible to tell (except as may be inferred from the form of the resulting curve) which side of the crack was carrying the instrument. It is clear that cracks cannot form in a reinforced concrete beam without more or less slip of bar, and that the instruments near cracks will indicate the greatest amount of slip at any point in their vicinity rather than the average amount over some length of bar. The curves for points near the ends, where the concrete is unbroken, are’ generally regular, and may be taken to represent the true load-slip relation at these points. Generally slip-of-bar was more pronounced at one end than at the other, although in some of the tests the two ends behaved nearly alike. In many of the tests it could be predicted from the observed differences in the slip measurements, some time before the completion of the test, at which end of the beam bond failure would occur. Table 39 gives the loads corresponding to beginning of slip at the points where measurement was taken in the tests of 6-ft. beams in the 1912 series. For one-third point loading, slip of bar at the middle be- came appreciable at about the same load for all the beams, regardless of the kind of bar used; this load averaged about 3700 lb., corresponding to a tensile stress in the concrete of 200 lb. per sq. in., if the concrete be considered as taking the entire stress. or the same beams slip of bar began under the loads at about 4400 lb: Tension cracks in the concrete were not generally visible until the load was more than double this amount. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 179 The load-slip curves for Beam No. 1056.2 (Fig. 69) may be con- sidered typical of those obtained in the tests of 6-ft. beams reinforced with 1-in. plain rounds. In this test, slip was first observed in the middle portion of the beam at a load of about 4000 lb. At the middle, instrument (1), the bar is represented as moving toward the right, but the direction of movement would have been reversed had it happened that the instrument was carried on the other side of the crack. The slip at points south of the middle is greater in every instance than at the corresponding point at the north end of the beam. A general slipping of the bar occurred at the S. end at a load of about 14 000 lb.; this slip varied from about 0.0003 in. at the end to 0.005 in. at the south load point. The computed bond stress at this load was about 270 lb. per sq. in. Slipping at the north end became general at a load of about 16 000 lb. The beam failed by bond at the south end, as shown by the flattening of the curves at (3), (5), (7), and 8. In Beam No. 1052.6 (Fig. 60b and 69), in which auxiliary bars were used outside the load points, slip-of-bar in the outer thirds began at a higher load than in other similar beams without these bars. This was probably due to the auxiliary bars taking a part of the tensile stress dur- ing the early stages of the test and thus relieving the tensile stress in the main longitudinal reinforcing bar at the points where slip-of-bar was measured. These auxiliary bars had the effect of reducing the number of large cracks outside the load points (or preventing them entirely) as shown in the photograph in Fig. 60b. The. prevention of large cracks seems to have had the effect of causing a more uniform distribution of the bond stress along the bar during the early stages of slip. Slip be- came general at both ends at a load of about 16 000 lb.; this corresponds to an average bond stress of 308 lb. per sq. in. The slip at this load varied from 0.0005 in. at the north end to 0.0012 in. 1 ft. from the supports. The pull-out tests (average of lowest two from each set) gave a bond resistance of 306 Ib. per sq. in. at an end slip of 0.0005 in. The presence of the auxiliary bars had no influence on the maximum bond resistance. The load-slip curves for the 1-in. square bars given in Fig. 74 show about the same general characteristics as those for beams with round bars. In testing Beam No. 1046.1 measurements of slip of bar were made at eleven points. The low loads at which slipping began at points (4), (6) and (8) is noteworthy; the reversal of direction in the curves for points (4) and (6) was due to the opening of cracks near these points ILLINOIS ENGINEERING EXPERIMENT STATION 180 ‘a[Pprur oY} UlOIy “Ul OF GIOM S}USUINIYSU] , ‘ureaq 04} JO a[pprur oq} WOIJ “Ul gy atom SUBUINIYSUY 4 or ; z 3 : ' ct eL 008 go |irttteetteetterseeeteeeee ees panos “109 “ut 8% F'erol St 8 ‘ : ¢ . fe oe ooe'ez gq |e BAe punoz ‘109 “Ut 841 e'srot aI 61 9 5 5 5 or al 009 0% oA Wee as axes pazetany “UT T £2501 g1 gL 9 9 ¥ 9 9 ¥I SI 008 12 GS iis caamexey wreak ea azenbe paystag “Ur J TLy01 st 8 9 ¥ 5 8 8 a 000 02 GROEN Pe untae were eee Se orenbs weld “Oy T ¥ OFT ol or 9 5 ¥ or ar 008 6T Be PRR Ec eee Ce orenbs mred “Uy T & 9501 WW xl z g g 5 ¥ 51 000 61 eo err arenbs uyed “Uy f T9F0I a 41 ot or i OL or 1% 6t 000 0& Yio 3) ete? Siena ee punoz ured “ur 7 £0901 or 8 5 5 z or g ol or 004 21 ef ee punog arefd “ur T &'6S01 Or | 48 9 # a # 9 +01 gl 009 81 Bee ae Seat es punoz uyeyd “ur 28901 I ul 5 % 5 é 8 ol ot 000 #1 Be ee cree emer punoz ureyd “ary & 8S0r 81 ar ¥ 8 9 é 8 al eI OOT 61 BAG, Pr TSE aire tavernas Pe punog ured “ut | 2° 2S01 #1 aL #1 al at #1 008 SI Ce ec eee punos uteyd “ur T 9° 2501 ar SI 8 9 9 5 8 41 9 002 61 Ge WR ee ey se Punos ureyd “ar y & 9901 9[ppt ce ee eS 2S a1PPIW, PN ee Dx ON a 499} : : : : : : edie ase juauadiojuraY [SUIpNysa0T aoe (40008 Jo sa, UT) NS | eee ‘eyUIog WarepIq ye syusuINysUy Aq payeorpuy dug yer 98 peoT parjddy ‘paysorput syutod 943 48 ‘UI ZQN0'O noge Jo dis @ pasneo a[qe} 94} UI USAIS speOoT oy, ‘ueds “3j-9 8 UO paqseq a1aM sUIvaq [[V ‘SNVEQ GQLAYONO/) GHOUOMNIDY NI SLNIOG WLVIGHWUALNT LY UVg dO dIIG LSUly LY AvOT dailddy ‘6 CHIAVL ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 181 at a load of about 7000 Ib. In spite of the large amount of slip at point (8), failure occurred in the beam at the opposite end. The curves for corresponding points at the opposite ends of Beam No. 1046.2 show an unusual uniformity. It will be seen by reference to Fig. 62b that cracks had formed near each of the instruments (2), (3), (4), and (5); the symmetry of the curves is due to the fact that the instruments were in each case carried by the concrete outside the crack. The curves for Beam No. 1046.4 show the bar to have been slipping in one direction throughout nearly two-thirds of its length; however, the presence of a crack at the middle makes the indicated direction of movement of the bar at this point depend upon which side of the crack the instrument happened to be carried. Slip-of-bar measurements at intermediate points were made on two beams reinforced with 1-in. twisted square bars (see Fig. 75). The bond stresses developed were lower than those found with beams rein- forced with 1-in. plain rounds. The curves for these tests exhibit about the same characteristics as in the beams with plain bars. Vertical and diagonal cracks formed during the progress of the tests with about the same frequency as in the other beams. The nearly vertical direction of the curve for instrument (2) in Beam No. 1047.1 is due to the pres- ence of a crack at this point. The load-slip curves for the north ends of Beams No. 1047.1 and 1047.3 and for points 6 in. from the north end showed a large increase of slip at an applied load of about 15,000 1b., immediately after slip at these points became appreciable. Measurements of slip of bar at intermediate points were taken on four beams in which 114-in. corrugated rounds were used for longi- tudinal reinforcement. Two beams (1048.3 and 1048.4) were tested on a 6-ft. span and two (1049.2 and 1049.3) were tested on a 10-ft. span. Load-slip curves are given in Fig. 75 and 76. Slip began at points near the middle of the beam at approximately the same loads as in the beams reinforced with plain bars. In the outer thirds, after slip began, the corrugated bar moved at a slower rate under a somewhat higher bond stress than the plain bars. Cracks in the beams reinforced with corrugated bars occurred with about the same frequency as in the beams with plain bars. The curves for points (2), (3), (4), and (5) on Beam No. 1048.3 (see Fig. 76) illustrate the effect of cracks on the amount of movement of the bar at the various points with respect to the adjacent concrete. Instrument (2) was carried by the concrete outside the crack, hence registered a very large movement; instrument (3) was carried by the concrete inside the crack, hence showed prac- 182 ILLINOIS ENGINEERING EXPERIMENT STATION tically no slip. The crack near instrument (4) crossed the bottom of the beam in such a way that the instrument was carried by the concrete on both sides of the crack. Instrument (5) was about 3 in. outside a crack until near the maximum load, hence indicated about the normal amount of slip. The instruments at the ends and at points (6) and (7) were undisturbed by cracks and hence more nearly indicate the relation between load and the amount of slip, which may be expected under these conditions. The influence of span length on the slip of the reinforcing bars at internal points is shown for typical tests in Fig. 71, 72 and 73. The span varied from 5 to 10 ft.; the points of application of the load are shown in the figures. For convenience of reference the curves for Beam No. 1056.2 (6-ft. span) are repeated here to the same scale as used in the beams of other spans. It should be borne in mind that in general the 8 and 10-ft. beams failed in tension. The curves for Beam No. 1051.2 (5-ft. span) are about the same as found for the 6-ft. beams; slip became appreciable near the middle at a load of about 4000 Ib. and gradually extended toward each end. The bond stresses developed in the 5-ft. beams at first end slip and at the maximum loads are about the same as for the similar beams of 6-ft. span. Slip began near the middle of the 7-ft. beams at a somewhat lower load than in the 6-ft. beams; slip near the ends of the bars began at a bond stress a little higher than in the 6-ft. beams. In the 8-ft. beams slip began near the middle at about the same load as in the ?-ft. beams. The slip at the ends of the bars did not amount to more than 0.001 in. in any case. In the 10-ft. beams loaded at the one-third points, slip of the bar became apppreciable near the middle at applied loads below 2000 lb. Slip began at points 3 in. inside the supports at about the same load as in the 6-ft. beams. At the ends of the bars slip was only beginning at the maximum load. The 10-ft. beams loaded at two points 6 ft. apart behaved in much the same way as the 6-ft. beam of the same kind loaded at the one-third points. This is as was expected, since the relation of the dimensions of the beam (as far as bond and web stresses are concerned) were the same in both cases. These tests show that local slip of bar near the middle of the beam is found at loads which are lower than those at which ten- sion cracks appear in the concrete; this probably indicates that cracks were present some time before they became visible, and that beginning of slip was probably coincident with the opening of the most minute tension cracks. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STE (a) Showing Instruments for Measuring Center Deflection and Slip of Bar at Intermediate Points. (b) Showing Instruments for Measuring Slip of Bar at End and at Inter- mediate Points. Fic. 56. ReINFoRcED ConcreTE Beams iN MacuHINe Reavy For Tesvt. 184 ILLINOIS ENGINEERING EXPERIMENT STATION (a) Group 1. One 1-in. Plain Round Bar; No Stirrups. (b) Group 2. One 1-in. Plain Round Bar; %-in. Stirrups. (c) Group 3. One 1-in. Plain Round Bar; 14-in. Stirrups. Fic. 57, Retnrorcep ConcRETE BEAMS AFTER TEST—1911 SERIES, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 185 ed rn dM a! fey (a) Group 6. One 1-in. Plain Round Bar; 4 in. of Concrete Below Bar; No Stirrups. 1042.1 | (c) Groups 8 and 10. Ends of Beams Overhanging Supports 15 in. i Fic. 58. Rertnrorcep Concrete BEAMS AFTER TEST—1911 SERIES. 186 ILLINOIS ENGINEERING EXPERIMENT STATION (a) Groups 1 and 13. Beams No. 1040.1 and 1040.2 were Each Reinforced with One 1%-in. Corrugated Round Bar. (b) Groups 10 and 11. One 1%-in. Plain Round Bar. (c) Group 34. One 1%-in. Corrugated Round Bar; Span 10 ft. Fic. 59. REInFrorcep Concrete BEAMS arrer TEST. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 187 (a) Group 14. One 1-in. Plain Round Bar. (b) Group 16. One 1-in. Plain Round Bar; Auxiliary Bars at Each End of Beam. 2 { 12 2 16 | us J20 ig ps i 3 2\0 6 3 6 (c) Group 17. One 1-in. Plain Round Bar, Span 6 ft.; Loads 214 ft. Apart. Fic. 60. Remnrorcep ConcreETE BEAMS AFTER TEST—1912 SERIES. 188 ILLINOIS ENGINEERING EXPERIMENT STATION (c) Group 20. One 1-in. Plain Round Bar, Span 6 ft.; Loads 4 ft. Apart. Fic. 61. REmnrorcep Concrete BEAMS AFTER TEST—1912 Srrigs, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 189 (b) Group 29. One 1-in. Plain Square Bar. (c) Group 32. One 1%-in. Corrugated Round Bar. Fic. 62. REINFORCED CONCRETE BEAMS AFTER TEST—1912 SERIES. 190 ILLINOIS ENGINEERING EXPERIMENT STATION (a) Group 21. One 1-in. Plain Round Bar; Span 5 ft. t (b) Group 22. One 1-in. a Round Bar: Span 7 ft. (a) Caines 24 and 25. One 1-in. Plain Round Bar; Span 10 ft. Fic. 63. Retnrorcep Concrete BEAMS AFTER TEST—1912 SERIES, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 191 The curves for the groups of 6-ft. beams in which the loads were applied at varying distances apart exhibit some distinctive character- istics (see Fig. 69 and 70). These tests show with unusual distinctness the progress of slip of the bar from the middle toward the ends of the beams. There are many illustrations in this group as to the effect of cracks on the slip of the bar. The curves for point (1) in Beams No. 1056.2 and 1057.2 show the amount of slip which may be expected very close to a crack at the middle of a beam; the curves for points (1) and (2) in Beam No. 1060.3 show that little or no slip occurred at points midway between two cracks in the portion of the beam where vertical shear was absent; the curves for points (2) and (4) in Beam No. 1057.2 and for points (2) and (3) in Beam No. 1056.2 show that slip of bar may be erratic at points on the near side of a crack. Slip at points outside the loads was abnormally large on the far side of the cracks. 95. Distribution of Tensile Stress along Reinforcing Bar.—The usual analysis of the stresses in a reinforced concrete beam indicates that in a beam loaded as in these tests (disregarding the weight of the beam) the tensile stress in the reinforcement is uniform between the loads and decreases at a uniform rate from the load points towards the ends, becoming zero at the supports. The slip of bar measurements which were made in many of the reinforced concrete beam tests gave some indication of the bond stresses being developed as long as the amount of slip was small and showed the general variations in bond stresses, but it will be recognized that after the slip of bar has reached an amount approximating that corresponding to the maximum bond re- sistance, such measurements are of no further value in indicating the amount of bond stress being developed. There are indications that the amount of slip corresponding to the maximum bond resistance of plain bars was not nearly so well defined in the beams as it was in the pull-out tests. In order to learn the amount of bond stress being de- veloped over any given length of har in a reinforced concrete member, we must first determine the exact stress developed in the reinforcing bar at each point over this length for a given load. If we can accu- rately determine the variations in the steel stress the bond stress will also be known, since the bond stress developed in any length of bar rep- resents the change in tensile stress over that length. In the tests of three beams in the 1912 series careful measurements were made to determine the stress in the longitudinal reinforcing bar at frequent intervals throughout the span length. These tests will be 192 ILLINOIS ENGINEERING EXPERIMENT STATION discussed in detail in the following paragraphs. All of these beams were 8 by 12 in. in section, 10 in. deep to the center of the reinforcing bar; they were tested by loads applied at the one-third points of a 10-ft. span. Beams No. 1055.3 and 1055.6 were each reinforced with one l-in. plain round bar; Beam No. 1049.3 was reinforced with one 114-in. corrugated round bar. In Beam No. 1049.3 the tensile stresses were measured at 10 6-in. gauge lengths over only one-half the span length; in the other two tests the stresses were measured at 19 gauge lengths over the entire span. In all the tests the gauge lines formed a con- tinuous series. A non-fixed strain gauge was used in determining the change in stress.* Load was applied in increments of about 2000 lb. In the test of Beam No. 1055.3 the deformations in the reinforcing bar were measured at each increment of load; in the other two tests each load was released and measurements were taken to determine the residual stresses, before applying a higher load. In Fig. 64 to 67 the distribution of bond and steel stresses are shown for different loads on the beams. In these figures the ordinates or abscissas to the curves show the amount of the stress in the reinforcing bar at any point for the loads shown, while the slope of these curves is proportional to the bond stress being developed. In Fig. 64 and 65 the curves for Beams No. 1055.3 and 1055.6 have been plotted by averaging the stresses at the corresponding gauge lines at the two ends of the beam and plotting the resulting values as the stresses for one-half the span length. This probably accounts in some measure for the greater regularity in the curves for beams reinforced with plain bars. It should be pointed out here that the values for steel stresses were determined from measure- ments over a 6-in. gauge length, and it is evident that the stress may vary considerably over this length of bar, especially at points where the bar has not slipped an amount corresponding to the maximum bond resistance. The stresses given on the diagrams are the averages over the entire gauge length. The plotted points indicate the location of the middles of the gauge lengths. It seems probable that a 6-in. gauge length was too long to accurately locate the points of maximum bond stress in the region affected by cracks, but it was not practicable to use a shorter gauge length. Of course, the usual errors of observation may be expected to affect measurements of this kind. The presence of tensile cracks in the concrete was another disturbing factor. *For a detailed discussion of this strain gauge see “Tests of Reinforced Concrete Buildi Under Load,” by Arthur N. Talbot and Willis A. Slater, University of Illinois Magee Experiment Station Bulletin No. 64, 1918. Also, “Use of the Strain Gage in Testing Ma- terials,” by Willis A. Slater and Herbert F. Moore; “Proceedings of the American Society for Testing Materials,” 1918. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 193 Table 40 gives some of the significant bond stresses for these beams as computed by the usual method and as determined from the steel stress measurements. In this table and in the figures only the stresses due to the applied loads have been considered. The first column of ob- served stresses are generally the average stresses over a length of about 12 in. in the portion of the beam about 4 to 16 in. outside the load points. The other column shows the average observed stresses over a length of 9 to 15 in. at the ends of the beam. The photographs in Fig. 59 and 63 show the location of cracks in thesa beams and their growth with increase of load. The numbers inside the circles at the level of the reinforcing bars are opposite the points at which measure- ments of steel stress were made. TABLE 40. DisTRIBUTION OF BOND STRESS IN REINFORCED CONCRETE BEAMS. Beams 8 by 12 in. in section, 10 in. deep to center of reinforcing bar. Loaded at the one-third points of a 10-ft. span. All beams failed by excessive tensile stress in the reinforcing bars. Observed Bond Stress , Average oer Beam Size and Kind of Bar = Toad on Rad | Over Region Near No. es Test eam Siress | just Qutside | Ends of pounds jp, per sq. in.| Pointg* th ee : Ib. per sq. in.) * per sg; in: 2 000 38 100 16 4 000 76 125 34 1055.6 | One 1-in. Plain Round...............- 2 yr. 6 000 114 191 36 8 000 152 226 64 10 000 190 201 117 11 700 222 165 238 2 000 38 48 15 4 000 76 75 54 1055.3 | One 1-in. Plain Round................ Qyr. 6 000 114 155 95 8 000 152 141 100 10 000 190 200 130 10 700 203 140 156 2 000 34 80 20 4 000 68 137 45 6 000 102 226 95 8 000 135 285 135 10 000 170 250 150 1049.3 | One 1}4-in. Corrugated Round........ 13 mo. 12 000 204 315 150 ' 14 000 236 350 225 16 000 270 385 260 18 000 306 400 290 20 000 338 450 315 21 000 355 200 360 21 900 370 Baie 390 * These stresses are, in general, the average bond stresses developed over a length of about 12 in. in the portion of the beam about 4 to 16 in. outside the load points + The average observed stress over a length of 9 to 15 in. at the ends of the beam. 194 ILLINOIS ENGINEERING EXPERIMENT STATION . Stresses after Stresses under Load Release of the Load Applled Load - Pounds Beam 1055.6 Tensile. Stress in Longitudinal Bar- Lb. per Sa.In. (a) Load-Stress Curves for Longitudinal Bar. 60000 Beam 1055.6 20000) 1-2-4 Concrete Age 2 Years One 1-in. Plain Round apo00 Loaded at 4% Points of 10-ft Spon. 30000 20000 10000 Stresses in Steet-Lb. per Sq.in. Stirrups 48 60 Distance from Center of Beam -Inches (b) Distribution of Tensile Stress in Longitudinal Bar. Fic. 64. Dracrams From Test oF Beam No. 1055.6. One 1-1n. PLAIN Rounp Bar; Span 10 Ft, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 195 Beam No. 1055.6, reinforced with one 1-in. plain round bar, failed by tension in the steel, although slip at each end of the bar at the maximum load amounted to about 0.002 in. From measurements of steel stress in this test it is seen that there is a wide variation in the bond stresses developed at a given load along the bar through the region where beam bond stress is present. At a load of 2000 lb. the steel stress through the middle third of the beam was nearly constant, with evidence of the highest stress about 6 in. outside the load point. It is significant that the photograph of this beam, Fig. 63d, shows a crack at this point on both ends of the beam. This crack was not observed antil a load of 4000 lb. had been applied, but it is evident that the measurements at 2000 Ib. load indicate that stresses higher than usual were being developed at these points. At a load of 2000 lb. the average bond stress developed over the 6-in. length between the gauge lines (6) and (5) was about 100 Ib. per sq. in.; a stress of about the same value was being developed between gauge lines (3) and (2), and it is signifi- cant that cracks opened at these points at low loads. It seems probable that the highest bond stress developed at a load of 2000 lb. was con- siderably higher than 100 lb. per sq. in., since the measurements over 6-in. gauge lengths probably did not show the most rapid changes in steel stress. Over the 15-in. length from (2) to the end of the beam the bond stress at this load was about 16 lb. per sq. in. At this load the computed bond stress was 38 lb. per sq. in. With a load of 4000 lb. the average bond stress developed over the 12-in. length from (6) to (4) was 125 lb. per sq. in.; from (3) to (2) about 200 lb. per sq. in.; from (2) to the end, 34 lb. per sq. in. The computed bond stress due to an applied load of 4000 lb. was 76 Ib. per sq. in. In other words, an applied load of 4000 Ib., which developed a steel stress in the middle third of about 13000 lb. per sq. in., developed a bond stress over cer- tain regions of the span length equivalent to 90% of the computed bond stress at the maximum load carried by the beam. A load of 6000 lb., which produced a tensile stress in the bar at the middle third of the span of about 20000 Ib. per sq. in., developed a bond stress between (6) and (5) of 191 lb. per sq. in.; at the end of the beam the stress at this load was 36 lb. per sq. in. A load of 8000 lb. pro- duced a bond stress from (6) to (4) of 226 lb. per sq. in., and 64 lb. per sq. in. at the end; the computed bond stress at this load was 152 lb. per sq. in. With a load of 10000 lb. on the beam the average bond stress over the length of 12 in. from (6) to (4) was 201 Ib. per sq. in., or lower than was found in this region at a load of 8000 lb.; from (3) 196 “ILLINOIS ENGINEERING EXPERIMENT STATION to (2) the indicated load stress was 390 lb. per sq. in., but it is pos- sible that the points plotted for steel stress are somewhat erratic. There was a marked shifting of the stresses following a load of 10000 Ib. At the maximum applied load of 11700 lb. the bond stress at points just outside the load had fallen still lower, while the stress near the ends of the beam was more than double what it was under a load of 10.000 lb. At the maximum load the highest bond stress was devel- oped between points (4) and (2) and amounted to about 294 lb. per sq. in. It will be seen from Fig. 64h that the region of maximum bond stress was thrown from near the load point toward the end of the beam as the loading continued. It should be borne in mind that this beam was 2 years old at the time of test and did not fai] by bond. The values in Table 40 show some of the changes which the bond stresses under- went as the load was increased. The bond stress near the load point did not show any marked increase after a load of 6000 lb., while at the end there was a continuous increase up to the maximum applied load. Still higher bond stresses could have been developed near the ends of the beam had the bar been of steel with a higher yield point. Fig. 64a shows that the residual stress in the reinforcing bar upon release of load was about proportional to the stress under load for the lower loads. After a load of 6000 lb. there is no further increase in the residual stresses. At loads near the maximum the residual stresses were nearly equal at all points except over the gauge lines nearest the ends. Beam No. 1055.3, reinforced with one 1-in. plain round, was simi- lar to No. 1055.6. In this test the loads were not released. Fig. 65 shows the distribution of the tensile stress in the reinforcing bar for different loads; the plotted points are the averages of the measurements at corresponding points in the two ends of the beam. This beam did not show as wide variations of bond stress as were found in the test of Beam No. 1055.6 discussed above, but many features of the tests are similar. It is noteworthy that the bond stresses developed near the ends of the beam were never more than about 75% of the com- puted bond stress. The bond stresses near the end are shown to be low by the fact that at the maximum load the end slip of the bar was only 0.0003 in. ‘This beam failed by tension in the reinforcing bar. A further examination of the results of these tests on beams of similar make-up, as shown in the tables and diagrams, will indicate how much variation may be expected in the action of beams which are of the same dimensions and made of the same materials, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 197 Beam No. 1049.3 was reinforced with one 114-in. corrugated round bar of high-carbon steel. This beam was tested at the age of 13 mo. on a 10-ft. span. Measurements were made to determine the variation in stress in the reinforcing bar at 10 points along one-half the span length and to determine the amount of slip of bar at five points of the other half of the span. A complete set of observations were taken after the application of each load and another upon releasing each load. The observed values of steel stress and slip of bar are plotted in Fig. 66 and 67. It will be noted that the load-stress curves from this test show much greater irregularities than were found in the other tests. This is probably largely due to the fact that in this test the $0000 Beam 1055.3 as 1-2-4 Concrete Age 2 Years One I-in. Plain Round y i i“ 30000 Loaded at % Points of to-ft. Span 20000 Stresses in Steel - Lb. per Sq.in. 40000 Stirrups: - oO 12 24 36 48 60 Distance from Center of Beam-In. Fic. 65. DistriBuTion oF TENSILE Stress IN Bar 1n Beam No. 1055.3. One 1-1n. PLatin Rounp Bar; SPAN 10 FT. points are the result of a single measurement, whereas in the tests described above the points represent the average of the values from both ends of the beam. At a load of 4000 lb., about one-fifth the maximum applied load, a bond stress of 137 lb. per sq. in. was being developed over the region (6) to (4) and an average of about 45 lb. per sq. in. over the 21-in. length at the end of the beam. The varia- tion of bond stress throughout the test can’ be studied from the dia- grams in Fig. 67 and from the values in Table 40. The observed bond stress over the region just outside the load point continued to rise until it reached 450 lb. per sq. in. at a load of 20000 Ib. Near the end of the beam the bond stress continued to rise until it reached 390 lb. per sq. in. at the maximum load of 21900 lb. The highest bond stress observed over a considerable length of the bar in this test occurred he- 198 ILLINOIS ENGINEERING EXPERIMENT STATION 24000 20000 Age 19 Months: in.Corrugaled Pees fort Sen, 16000 vn uu 5 & 12000 U g a vu © 6000 & tz 4000 O .002 004 006.008 010 02 0/4 O16 .O18 .020.022. +028 030 032 034.095 Slip of Longitudinal Bar -Inches (a) Load-slip Curves for Longitudinal Bar. 4000 20000 U. 12000 Applied Load-Po Beam 10493 0000 o 0000 oO 20000 o 20000 Oo ©0000 oO 20000 O 20000 oO €0000 Oo 20000 oO €0000 Tensile Stress in Longitudinal Bar - Lb. per Sq. In (b) Load-stress Curves for Longitudinal Bar. Fic. 66. Dracrams From Test or Beam No. 1049.3. OnE 1%-1n. CorrucaTep Rounp Bar; Span 10 Fr, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 199 24000 Applied Load - Pounds i 3 g§ 8 é Stresses under Lood Fesldval Stresses afler Felease of he Load 4000 Beam 10493 oO £0000 40000 60000 60000 oO 20000 40000 Tensile Stress in Lonoitudinol Bar - Lb. per 5q.In. (a) Load-stress Curves for Longitudinal Bar. Beam 10493 60000 1-2-4 Concrete. Age 13, fas one /g-i7. laadat i tais cd. t $ POINTS OF) Ob ae Span : Sess Loads Shown é : Stirrups Load Load Stresses in Steel — Lb. per Sq.in. (7) 2 a4 36 48 60 Distance from Center of Beam— Inches (b) Distribution of Tensile Stress in Longitudinal Bar. Fic. 67. DiacramMs From Test oF Beam No. 1049.3. ONE 114-1N. CoRRUGATED Rounp Bar; Span 10 FT. 200 ILLINOIS ENGINEERING EXPERIMENT STATION tween (3) and (1) at the maximum load, and amounted to 580 lb. per sq. in. It is evident that stresses as high as this must have occurred at all points along the bar from the load point to (3) at some stage of the test, but the measurements were not sufficiently refined or the observations taken at sufficiently frequent intervals to detect such changes. This makes it apparent that there must be very rapid changes in the distribution of bond stress along the bar in a test of this kind. Fig. 66a shows the slip of bar at five points in Beam No. 1049.3. The amount of slip measured at any point depends to a large extent upon the proximity of cracks in the concrete and the load-slip curves may be expected to show considerable variation due to this cause. The residual slip after release of load is equal to about one-half the total slip up to a lcad of about 16000 lb. The residual tensile stresses in the bar exhibit about the same characteristics as the residual slip-of- bar measurements. It is noteworthy that the residual tensile stresses over the larger portion of the length under observation did not increase much with increase of load after a load equal to about one-half the maximum had been applied. It is probable that a period of rest fol- lowing the release of a load would have shown a material reduction in the residual stresses measured. These tests show that the actual bond stresses developed in beams of this kind vary widely at all stages of the tests and that the bond stress calculated in the usual way represents the average stress, but does not indicate the extremes of stress in different portions of the span where beam bond stresses are present. The actual bond stresses developed varied from less than one-half to more than twice the cal- culated bond stress as determined in the usual manner. As was pointed out in Art. 68, “Phenomena of Beam Tests,” these tests indicate that at the early loads which develop the maximum beam bond resistance over a short length of bar outside the load points in beams in which the distance from the load to the support is as much as four times the effective depth, the bond stress developed near the ends of the bar may not be more than, say, 15% of the maximum bond re- sistance. The ratio of these stresses was not definitely determined, but the measurements of the tensile stress in the steel and the slip of bar indicate that the ratio given is approximately correct. It is probable that for longer beams the value would be found to be lower; in shorter beams the bond stress at the end of the bar when the maximum bond resistance was first developed outside the load points was probably as high as 40% of the maximum bond resistance. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 201 In Fig. 68 is shown the distribution of bond and tensile stresses which may be expected in a reinforced concrete beam which fails by bond under two symmetrical loads. The curves indicate the changes which the stresses at different points undergo as the load is increased. In the region between the loads the tensile stress in the bar would be constant (disregarding the weight of the beam and the effect of anti- stretch slip) as indicated by the horizontal lines in the left portion of the figure. Between the load point and the support, the tensile stress at any point would be represented by the ordinates to the curves and the oblique lines, and the bond stress by the slope of these lines. The dia- gram indicates that the maximum beam bond resistance is first developed a short distance outside the load point at a load of 40% of the maximum load. The region over which the maximum bond resistance is being developed by a given load is indicated by the heavy solid lines. Since a bond stress much higher than the average is developed over a portion of the span, it is evident that at other points the bond stress must be less than the average. For a load 40% of the maximum the computed tensile stress outside the load point would be indicated by the ordinates to the dotted line KS, and the bond stress by the slope of this line, while the actual distribution of the tensile and bond stresses would be represented somewhat as shown by the curved solid line KS. This curve indicates that for a short distance outside the load point and near the support the bond stress is small, and that over a short length the maximum bond resistance is being developed. As the load is increased the region just outside the load point over which the bond stress is small is gradually extended and at the same time the portion of bar over which the maxi- mum bond resistance is being developed is lengthened and pushed nearer the support. At the load causing failure by bond the actual stress in the bar is somewhat as shown by the line TLOS. In other words, the maximum bond resistance developed is represented to scale by MR in- stead of by LR as computed; and at failure this maximum stress was being developed over the length of bar PS, instead of over the length RS, as assumed by the usual methods of calculation. The parabolic line TS represents the theoretical stresses for a uniformly distributed load which produces the same maximum tensile stress at the mid-span as that produced by the concentrated loads shown. This curve suggests that a uniform load gives a more favorable condition for bond than concen- trated loads. This is also evident from other considerations. 202 ILLINOIS ENGINEERING EXPERIMENT STATION It is not intended that Fig. 68 should give a quantative indication of the distribution of stresses in any particular reinforced concrete beam. The presence of anti-stretch slip and the bond stresses due to this cause will greatly modify this distribution of stresses over certain regions of the span. However, it is clear that this somewhat idealized sketch does suggest the explanation of the discrepancy between the com- puted bond stresses in beam tests and the bond resistance found in other ways. The more accurate determination of the actual distribution of these stresses for beams of different make-up is a proper subject for further experimental study. 96. Relation of Slip of Bar to Diagonal Tension Cracks.—About 80 per cent of the beams reported in this bulletin failed in bond or an obvious combination of bond and diagonal tension. In many tests it was difficult to definitely assign the primary cause of failure, since the bar had shown considerable slip and at the same time there were evi-: dences of failure from excessive diagonal tensile stresses in combination with the large slip. In assigning the manner of failure as shown in the tables, all the evidence of the test was considered—the slip of bar at the ends and at intermediate points, the size and position of cracks in the beam and the calculated stresses in the steel and concrete. In some of the tests the measurements gave indications of bond failure, but the calculated tensile stress in the bar and the cracks in the beam showed that the longitudinal steel was over-stressed at the middle of the span. At failure nearly all of the beams except those of the longer spans showed one or two prominent diagonal cracks at one or both ends about midway between the load and the support. In Tables 28 and 34 the applied loads on the beam at the time of the observation of the first outer crack and at the first slip of the ends of the bar have been given in parallel columns. Reference to Fig. 69 to 76 and to Table 39 will show that slip of bar becomes appreciable at points about midway between the loads and the supports at loads about one-third the maximum for the beams reinforced with plain bars and loaded at the one-third points. For 25 tests on such beams the first visible crack outside the load points appeared at loads averaging 57% of the maximum and slip at the end began at 70% of the maximum load. These tests show that there is a considerable slip at the point where the diagonal cracks appear before these cracks became visible and indicated that the opening of these outer cracks was probably. due primarily to slip of bar. The proportion of this slip due to beam ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 203 bond stress and to anti-stretch slip will depend upon the dimensions and reinforcement of the beam. The bars used in these test beams were relatively large. The differences between the conditions in the middle third of the beam length and in the outer thirds should be kept in mind. In the middle region beam bond stress is not brought into action and the slip is wholly of the nature of what has been termed anti-stretch slip. In the outer thirds there is a combination of beam bond stress and of bond stress due to anti-stretch slip. The first outer crack generally became visible at a slip of bar of 0.002 to 0.005 in. It is evident that diagonal cracks may open at a very small slip of bar and that they open at loads which give very small end slip and even before end slip is noted. In the beams of Group 6 in which 4 in. of concrete was placed below the center of the bar, the cracks were restricted to the region within the load points or to a short distance outside (see Fig. 58a). The vertical shearing stresses developed were considerably higher than in other beams which were not reinforced with vertical stirrups. 3 r 100% L . . 90% ~~ S ge: = & Ss, x. Qe AN s8 | 80% SSS Se SS S§ ; 70% S > on Sg Se 0 Bs 60% Nae § g ok \Y ge 50% oN gs 10% K wey SS ba NS RS 30% SSS ACS 20% awe ~ ~ | | | Sy eS oS, 10% < 8 oes ‘R. _ a : SS SS SAS lin 2 € Fic. 68. IpEALIzED D1AGRAM SHOWING THE DISTRIBUTION OF TENSILE AND Bonp STRESSES IN A SIMPLE REINFORCED CONCRETE BEAM. 204 ILLINOIS ENGINEERING EXPERIMENT STATION 97. Influence of Slip of Bar upon Beam Deflection.—Several treatises on reinforced concrete develop formulas which are intended to give expressions for the center deflection of a reinforced concrete beam, when the dimensions of the beam, the condition of the ends and the loading are known. These formulas will not be considered here, except to say that they are based on the usual deflection formula for homogeneous beams but modified to involve the assumptions that are commonly made in the analysis of reinforced concrete beams. These formulas show that beam deflection is a function of the loading, con- dition of the ends of the beam, span length, the elastic properties of the material and the dimensions of the section. With a combination of two materials having such different properties as concrete and steel and owing to the variation in the modulus of elasticity of concrete as the compressive stress increases, the deflection of a reinforced con- crete beam is an extremely complex function. However, formulas have been devised which give fairly accurately the value of the deflec- tion up to working loads if the proper value is assigned to the modulus of elasticity of the concrete. One of the assumptions referred to above is that a plane section before flexure is a plane section after flexure. We have seen that during the greater part of the test of a reinforced concrete beam, slip of bar is an important phenomenon, and we may expect slip of bar to exert a marked influence on the deflection. Turneaure and Maurer,* in an investigation of deflection of reinforced concrete beams found that their formula gave values for deflection which were close to the true ones up to about one quarter of the maximum load carried by the beam. This is about the load at which slip of bar became pro- nounced through the middle portion of the span and we may expect that from this point the deflection is considerably influenced by slip of bar. As long as slip of bar is not general, there is no distinct change in the course of the load-deflection curve due to this cause, although its effect must be present from the time slipping begins. In homo- geneous beams the additional deflection due to shear is neglected. In reinforced concrete beams the deflection due to slip of bar is an analogous quantity which has an important influence on the total deflection of the beam. As soon as slip of bar becomes general, as indicated by slip at the ends, we may expect the deflection to respond more rapidly to this cause. Many of the load-deflection curves in Fig. 7% to 86 show a decided in- “Principles of Reinforced Concrete Construction, 1909, ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 205 crease in deflection immediately following the beginning of slip at the ends of the bar. Noteworthy examples of this action are found in Beam No. 1052.6, Fig. 81; Beam No. 1050.1, Fig. 84; Beam No. 1048.4, Fig. 86. 98. Critical Bond Stress in Reinforced Concrete Beams.—The fore- going discussions of beam tests in which slip of bar was measured indi- cate that slip of bar is found at relatively low loads. It was seen that slip of bar occurred first in the middle portion of the span at early loads. With increase of load slip of bar progresses through the outer thirds toward the ends. With the appearance of slip of bar at the ends of the beam, slip had become general through the region of beam bond stress and the beams were not able to permanently withstand a load apprecia- bly greater than that causing first end slip. The tests indicate that a very small end slip may correspond to a critical bond stress. The average bond stresses computed on the basis of the vertical shear are not the actual stresses being developed. As has already been pointed out, the tests show that maximum bond resistance is developed at certain points of the bar at loads much below that causing bond failure in the beam. Certain observations indicate that the bond stress first reaches the value of the maximum bond resistance in the portion of the span in which beam bond stresses are developed, at points a short distance outside of the load points. Anti-stretch slip may act to increase the bond stress developed over certain regions of the span. While slip of bar gives an indication of the bond stress being developed at any point as long as the amount of slip is small, it should be noticed that after the slip of bar has exceeded an amount corresponding to the maximum bond resistance further slip gives no indication of the amount of bond stress being developed. 99. Working Stresses for Bond.—The Committee on Concrete and Reinforced Concrete,* appointed by four of the leading American engin- eering societies (commonly known as the “Joint Committee”) recom- mended the use of a bond stress equivalent to 4% of the compressive strength of the same concrete at 28 days of age for plain bars of ordinary mill surface and values one-half as great for drawn wire. The com- pressive strength of the concrete is to be determined from tests on 8 by 16-in. cylinders. For concrete giving a compressive strength of 2000 Ib. per sq. in., at 28 days, the allowable bond stress is 80 Ib. per sq. in. for *See report in Proceedings of the American Society of Civil Engineers, Feb., 1913. Also Proceedings of the American Society for Testing Materials, 1913, p. 228. 206 ILLINOIS ENGINEERING EXPERIMENT STATION plain bars. These values have been generally accepted by American designers. The British Joint Committee on Reinforced Concretet recommended the use of a bond stress of 100 lb. per sq. in. in concrete having a compressive strength of 1800 lb. per sq. in. when tested in the form of cubes not less than 4 in. on a side or cylinders not less than 6 in. in diameter and of a length not less than the diameter. Tests re- cently made at the University of Illinois on compression specimens of different forms show that the compressive strength of 6-in. concrete cubes is about 27% higher than that of 8 by 16-in. cylinders. On this basis the British Committee’s concrete would have a compressive strength about 70% of the American Committee’s concrete and the allowable working bond stress would be about 80% higher than the corresponding values of the American Committee, considering the quality of the con- crete. It was seen in Art. 48 that the average bond stress in the pull-out tests when end slip began was about 17% of the compressive strength of 6-in. cubes made from the same concrete. Reducing this to the basis used above, we may say that slip began in the pull-out tests at a bond stress equal to about 13% of the compressive strength of 8 by 16-in. cylinders. In the same way the ultimate bond resistance in the pull-out tests discussed in Art. 48 was about 19% of the compressive strength of 8 by 16-in. cylinders. Certain of the beam tests indicate that the actual bond resistance of a bar embedded in a reinforced concrete beam was not materially different from that of pull-out specimens made in the same position. The working stress recommended by the American Joint Committee is equal to about 14 the bond stress corresponding to first slip of bar and to about 1/5 the ultimate pull-out resistance of plain bars. Apparently this indicates a factor of safety of about 3 against slip of bar, if we consider first slip of bar in the region where beam bond stresses are developed as the critical stress. However, we have seen that on account of the unequal distribution of bond stress in beams loaded at the one-third points, the bond stresses actually developed over a short length outside the loads is probably near the ultimate bond resistance even for working loads at the age of the beams tested. This in reality may not be a serious consideration since there should be a sufficient reserve of bond resistance in the embedded portion of the bar where the tSee Second Report Joint Committee on Reinforced Concrete, published by the Royal Institute of British Architects, 1911. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 207 bond stress is low, and an additional reserve in the increasing bond re- sistance with the increase in the age of the concrete. From the way in which the steel stress is distributed in Fig. 64 to 67, it would seem that the stress conditions of a beam under a uniformly distributed load are more favorable to bond resistance than the system of concentrated loads used in the tests. It would appear that the values of working stresses for bond recommended by the British Committees are entirely too high con- sidering the grade of concrete. The values of the American Committee are as high as should be used under average conditions of workmanship. The data for beams reinforced with deformed bars are not as com- plete as for plain bars, but on account of the undesirable secondary stresses which are introduced in the concrete as a result of high bond stresses, it seems the part of wisdom to use the same values for deformed as for plain bars. However, it should be recognized that properly de- signed deformed bars may be expected to give a greater uniformity of steel stress than plain bars, and they may be counted upon to guard against local deficiencies in bond resistance due to lack of care in plac- ing the concrete around the bars or to other defects due to poor work- manship. IV. Summary. 100. Summary.—tThe tests covered a wide range of conditions and the results have a significant bearing on the nature of bond resistance, the action of bars of different forms under bond stress, and the behavior of beams subjected to high bond stresses. The load-slip determinations have given definite information on the nature and distribution of bond resistance. The following is a resumé of the principal observations and conclusions which have been stated and discussed in the text. Para- graphs 2 to 34, inclusive, refer primarily to the results of the pull-out tests. (1) Bond between concrete and steel may be divided into two prin- cipal elements, adhesive resistance and sliding resistance. The source of adhesive resistance is not known, but its presence is a matter of uni- versal experience with materials of the nature of mortar and concrete. Sliding resistance arises from inequalities of the surface of the bar and irregularities of its section and alignment together with the correspond- 208 ILLINOIS ENGINEERING EXPERIMENT STATION ing conformations in the concrete. The adhesive resistance must be overcome before sliding resistance comes into action. In other words, the two elements of bond resistance are not effective at the same time at a given point. Many evidences of the tests indicate that adhesive re- sistance is much the more important element of bond resistance. (2) Pull-out tests with plain bars show that a considerable bond stress is developed before a measurable slip is produced. Slip of bar begins as soon as the adhesive resistance is overcome. After the adhesive resistance is overcome, a further slip without an opportunity of rest is accompanied by a rapidly increasing bond stress until a maximum bond resistance is reached at a definite amount of slip. (3) The true relation of slip of bar to bond stress can best be studied by considering the action of a bar over a very short section of the embedded length. The difficulties arising from secondary stresses made it impracticable to conduct tests on bars embedded very short lengths. The desired results were obtained by varying the forms of the specimens in such a way that the effect of different combinations of dimensions could be studied. (4) Pull-out tests with plain bars of the same size embedded dif- ferent lengths furnish data which suggest the values of bond resistance over a very short length of embedment, or indicate values of bond resist- ance which are independent of the length of embedment. ‘Tests with bars of different size which were embedded a distance proportional to their diameters give the true relation when the effect of size of bar is elim- inated. Two series of tests of this kind on plain round bars of ordinary mill surface gave almost indentical values for bond resistance after eliminating the effect of length of embedment and size of bar, and we may consider that these values represent the stresses which were de- veloped in turn over each unit of area of the embedded bar as it was withdrawn by a load applied by the method used in these tests. These tests showed that for concrete of the kind used (a 1-2-4 mix, stored in damp sand and tested at the age of about 60 days) the first measurable slip of bar came at a bond stress of about 260 lb. per sq. in., and that the maximum bond resistance reached an average value of 440 lb. per sq. in. Tf we conclude that adhesive resistance was overcome at the first measur- able slip, it will be seen that the adhesive resistance was about 60% of the maximum bond resistance. This ratio did not vary much for a wide range of mixes, ages, size of bar, condition of storage, etc. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 209 (5) Sliding resistance reached its maximum value for plain bars of ordinary mill surface at a slip of about 0.01 in. The constancy in the amount of slip corresponding to the maximum bond resistance for a wide range of mixes, ages, sizes of bar, conditions of storages, etc., is a noteworthy feature of the tests. With further slip the sliding resist- ance decreased slowly at first, then more rapidly, until with a slip of 0.1 in. the bond resistance was about one-half its maximum value. (6) Pull-out tests with plain round bars show end slip to begin at an average bond stress equal to about one-sixth the compressive strength of 6-in. cubes from the same concrete; the maximum bond resistance is equal to about one-fourth the compressive strength of 6-in. cubes. These values were about the same for a wide range of mixes, ages and conditions of storage. In terms of the compressive strength of 8 by 16-in. concrete cylinders these values would be about 13% for first end slip and 19% for the maximum bond resistance. (7) The tests indicate that bond stress is not uniformly distributed along a bar embedded any considerable length and having the load applied at one end. Slip of bar begins first at the point where the bar enters the concrete, and the bond stress must be greater here than else- where until a sufficient slip has occurred to develop the maximum bond resistance at this point. Slip of bar begins last at the free end of the bar. After slip becomes general, there is an approximate equality of bond stress throughout the embedded length. (8) Small bars gave a bond resistance somewhat higher than the large bars during the early stages of the test. This was probably on account of greater irregularity of section and alignment of the smaller bars. The maximum bond resistance was not materially different for bars of different diameters. (9) Computations based on the elastic properties of the materials indicate that in the pull-out tests the tensile deformation in the bar had a much greater effect on the amount of bond stress which permitted a given slip of bar than had the compressive deformation in the concrete block in which the bar was embedded. (10) Rusted bars gave bond resistances about 15% higher than similar bars with ordinary mill surface. (11) The tests with flat bars showed wide variations of bond resistance and were not conclusive. Square bars gave values of unit- stress about 75% of those obtained with plain round bars. 210 ILLINOIS ENGINEERING EXPERIMENT STATION (12) T-bars gave lower unit bond resistance than plain round bars, but gave about double the bond resistance per unit of length that was found for the plain round bars of the same sectional area. (13) With polished bars the bond resistance is due almost entirely to adhesion between the concrete and steel. Numerous tests with pol- ished bars embedded in 1-2-4 concrete and tested at 60 days indicated a maximum bond resistance of about 160 Ib. per sq. in., or about 60% of the bond resistance of bars of ordinary surface at small amounts of slip. This value agrees closely with tests reported elsewhere, and appar- ently represents the value of the tangential adhesion between any clean steel and concrete of this quality. The sliding resistance of polished bars was very low. (14) Tests with polished bars with wedging and non-wedging tapers showed that adhesion was broken for both types of bar at about the same bond stress as in the polished bars of uniform section. (15) The tests with polished bars with wedging taper showed that after the adhesion was broken a considerable movement of the bar (as much as 14 in. with the smallest tapers) was required before the bond resistance again reached the amount which was at first carried by the adhesive resistance. The amount of movement necessary to restore the bond stress to the value of the original adhesive resistance was inversely proportional to the amount of taper. This indicates that a definite normal compression must be developed in the surrounding concrete before a longitudinal component equivalent to the original tangential adhesion is produced. (16) It was noted in the tests with plain bars that sliding resist- ance was due to inequalities of the surface of the bar and to irregularities of its section and alignment. The projections on a deformed bar give an exaggerated condition of inequality of surface or irregularity of sec- tion. Adhesive resistance must be destroyed and the usual sliding resistance largely overcome and the concrete ahead of the projections must undergo an appreciable compressive deformation before the projec- tions on a deformed bar become effective in taking bond stress. The tests indicate that the projections do not materially assist in resisting a force tending to withdraw the bar until a slip has occurred approxi- mating that corresponding to the maximum sliding resistance of plain bars. As slip continues a larger and larger portion of the bond stress is taken by direct bearing of the projections on the concrete ahead. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 211 (17) In determining the comparative merits of deformed bars, the bar which longest resists beginning of slip should be rated highest, other considerations being equal. The bond stresses developed at an end slip of 0.001 inch furnished the principal basis of comparison for the different types of deformed bars. At an end slip of 0.001 in. 12 sets of deformed bars of %4-in. and larger sizes embedded 8 in. in 1-2-4 concrete, tested at about 2 months, developed an average bond resistance of 318 lb. per sq. in., 4% higher than the corresponding value for plain bars. At this stage of the test, two sets of deformed bars gave practically the same bond resistance, five sets gave lower values, and five sets higher values than the plain rounds. At an end slip of 0.01 in., corresponding to the maximum bond resistance of plain bars, the average bond resistance of the 12 sets of deformed bars was 445 lb. per sq. in., 10% higher than plain rounds. At this stage of the test two sets gave about the same values, two sets gave lower values, and eight sets gave higher values than the plain bars. The hooping used in these specimens had a marked effect in increasing the bond resistance even at small amounts of slip. (18) The concrete cylinders of the pull-out specimens with de- formed bars were reinforced against bursting or splitting, because it was desired to study the load-slip relation through a wide range of values. The bond stresses corresponding to an end slip of 0.1 in. are the highest stresses reported for the deformed bars. In only a few tests was the maximum bond resistance reached at an end slip less than 0.1 in. It should be recognized that, in general, the bond stresses reported for de- formed bars at end slip of 0.05 and 0.1 in., could not have been developed with bars embedded in unreinforced blocks. These high values of bond resistance must not be considered as available under the usual conditions of bond action in reinforced concrete members. In the tests in which the blocks were not reinforced, evidence of splitting of the blocks was found at end slips of 0.02 to 0.05 in. (19) The normal components of the bearing stresses developed by the projections on a deformed bar may produce very destructive bursting stresses in the surrounding concrete. The bearing stress between the projections and the concrete in the tests with certain types of commercial deformed bars was computed to be from 5800 to 14000 lb. per sq. in. at the highest bond stresses considered in these tests. For bars having projections of different heights and spacing, the bearing stresses on the projections at the highest bond stresses considered were inversely propor- tional to the bond stress which had been developed by the har at an end 212 ILLINOIS ENGINEERING EXPERIMENT STATION slip of 0.01 in., the slip at which the projections were beginning to be effective. These considerations show that the ratio of the area of the projections measured at right angles to the bar to the superficial area of the bar in the same length is the proper criterion for judging of the effective bond resistance of a deformed bar. In some forms of bar the bearing stresses must have been much higher than the values given above. The large slip and the high bearing stresses developed in the later stages of the tests show the absurdity of seriously considering the extremely high values that are usually reported to be the true bond resistance of many types of deformed bars. (20) Round bars with standard V-shaped threads gave much higher bond resistance at low slips than the commercial deformed bars. The average bond resistance at an end slip of 0.001 in. was 612 lb. per sq. in. The maximum bond resistance was 745 lb. per sq. in. These were the only deformed bar tests in which failure came by shearing the surrounding concrete. : (21) In a deformed bar of good design the projections should present bearing faces as nearly as possible at right angles to the axis of the bar. The areas of the projections should be such as to preserve the proper ratio between the bearing stress against the concrete ahead of the projections and the shearing stress over the surrounding envelope of concrete. Failure by shearing of the concrete should be avoided. The tests indicate that the areas of the projections measured at right angles to the axis of the bar should not be less than, say, 20% of the super- ficial area of the bar. A closer spacing of the projections than is used in commercial deformed bars would be of advantage. Advocates of the deformed bar would do well to recognize the fact that in a deformed bar which may be expected to develop a high bond resistance, a certain amount of metal must be used in the projections which probably will not be available for taking tensile stress. (22) The 1-in. twisted square bars gave a bond resistance per unit of surface at an end slip of 0.001 in., only 88% of that for the plain rounds. Following an end slip of about 0.01 in., these bars showed a decided decrease in bond resistance, and a slip of 5 to 10 times this amount was required to cause the bond resistance to regain its first maximum value. After this, the bond resistance gradually rose as the bar was withdrawn. Some of the bars were withdrawn 2 or 3 in. before the highest resistance was reached. The apparent bond stresses at these slips were very high ; but, of course, such stresses and slips could ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 213 not be developed in a structure and could not have been developed in the tests had the blocks not been reinforced against bursting. Such values are entirely meaningless under any rational interpretation of the tests. (23) The load-slip curves for twisted square bars are similar to those for polished bars with wedging taper. The twisted bar is essentially a combination of the wedging and non-wedging taper. As the bar is drawn through the concrete the wedging tapers are drawn more firmly against the concrete ahead, while at the same time the non-wedging tapers are separated from the concrete with which they were originally in contact. The drop in the load-slip curves after an end slip of about 0.01 in. shows that the separation of about one-half of the surface of the bar from its original contact and the continued sliding of the flatter portions of the bar, until a large slip has occurred, have a greater in- fluence in reducing the average bond resistance than the increased bear- ing of the wedging tapers has in raising the bond resistance. The results found with the twisted square bar do not justify its present wide- spread popularity as a reinforcing material. (24) The tests with plain round bars anchored by means of nuts and washers and with washers only showed that the entire bar must slip an appreciable amount before these forms of anchorage come into action. Anchorages of the dimensions used in these tests did not become effective until the bar had slipped an amount corresponding to the maximum bond resistance of plain bars. With further movement the apparent bond resistance was high, but was accompanied by excessive bearing stresses on the concrete. (25) The load-slip relation for bars anchored by means of hooks and bends was not determined. The high resistance given in these tests was probably a result of the bearing stresses developed in the concrete ahead of the bends. (26) Tests on specimens stored under different conditions indicate that concrete stored in damp sand may be expected to give about the same bond resistance and compressive resistance as that stored in water. Water-stored specimens gave values of maximum bond resistance higher in each instance than the air-stored specimens; the increase for water storage ranged from 10 to 45%. The difference seemed to increase with age. The presence of water not only did not injure the bond for ages up to three years, but it was an important factor in producing conditions which resulted in high bond resistances. However, it was found that 214 ILLINOIS ENGINEERING EXPERIMENT STATION specimens tested with the concrete in a saturated condition gave lower values for bond than those which had been allowed to dry out before testing. The bars in specimens whith had been immersed in water as long as three and one-half years showed no signs of rust or other deterioration. (27) Specimens made out-doors in freezing weather, where they probably froze and thawed several times during the period of setting and hardening, were almost devoid of bond strength. (28) Pull-out tests made at early ages gave surprisingly high values of bond resistance. Plain bars embedded in 1-2-4 concrete and tested at 2 days did not show end slip of bar until a bond stress of 75 lb. per sq. in. was developed. Bond resistance increases most rapidly with age during the first month. The richer mixes show a more rapid increase than the leaner ones. The tests on concrete at ages of over one year showed that the bond resistance of specimens stored in a damp place may be expected ultimately to reach a value as much as twice that developed at 60 days. (29) The load-slip relation of leaner and richer mixes was similar ‘to that for 1-2-4 concrete. For a wide range of mixes the bond resist- ance was nearly proportional to the amount of cement used. This rela- tion did not obtain in a mix from which the coarse aggregate had been omitted. (30) When the application of load was continued over a consider- able period of time or when the load was released and reapplied, the usual relation of slip of bar to bond resistance was considerably modified. The few tests which were made indicate that’the bond stress correspond- ing to beginning of slip is the highest stress which can be maintained permanently or be reapplied indefinitely without failure of bond. The effect of continued and repeated load, impact, etc., may well be the sub- ject of further experimental study. (31) Little difference was found in the pull-out tests whether the load was distributed over the entire face of the block or over a narrow ring at the center of the block or around the edge of the face of the block. (32) Specimens molded in a horizontal position gave lower bond resistance than those molded in a vertical position; when settlement of the bar with the settlement of the concrete was entirely prevented, the bond resistance was reduced to about 60% of that found for similar specimens which were molded with the bars in a vertical position. Plain ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 215 bars tested by being pulled in the same or the opposite direction from the settlement of the concrete during setting gave about the same bond resistance, but in the tests of certain deformed bars this was not true. (33) The term “autogenous healing” is used to designate phe- nomena observed in pull-out tests and in compression tests of concrete cylinders in which the hardening of the concrete was interrupted by loading the specimen at early ages to its ultimate resistance. Up to an age of one year the bond resistance of specimens stored in damp sand was not affected by as many as four loadings at intervals during the period of storage up to the ultimate resistance. For specimens stored in air and tested in the same way, the bond resistance was less than for damp-sand storage, but the tests showed a steady increase in bond resistance with each loading up to three months. Specimens which had been stored in air for two months before the first test and in water thereafter showed a.decrease in bond with each subsequent loading, although the bond resistance in the last test was fairly high. The pres- ence of water apparently permits the continuation of the hydraulic action of the cement for several months after the mixing of the concrete. (34) Bond resistance of plain bars is greatly increased if the con- crete is caused to set under pressure. With a pressure of 100 lb. per sq. in. on the fresh concrete for five days after molding, the maximum bond resistance was increased 92% over that of similar bars in concrete which had set without pressure. The greater density of the concrete and its more intimate contact with the bar seems to be responsible for the increased bond resistance. Light pressures gaves an appreciable increase in bond resistance. With polished bars the effect of pressure was slight. (385) As might have been expected, the compressive resistance of concrete setting under pressure was increased in much the same ratio as the bond resistance. At the age of 80 days the initial modulus of elasticity in compression for concrete which set under a pressure of 100 Ib. per sq. in. was about 37% higher and the compressive strength was increased by about 73% over that of concrete which had set without pressure. The density of the concrete, as determined by the unit weights, was increased about 4% by a pressure of 100 lb. per sq. in. on the fresh concrete. The increase in strength and density was relatively greater for the low than for the high pressures. A pressure continued for one day, or until the concrete had taken its final set and hardening had begun, seems to have produced the same effect in increasing the 216 ILLINOIS ENGINEERING EXPERIMENT STATION strength and elastic properties of the concrete as when the pressure was continued for a much longer period. (36) Concrete cylinders tested in compression at age of 80 days after having been loaded to failure at 7 days gave compressive strengths nearly as high as those tested for the first time at the same age. Retests of cylinders which had set under pressure gave similar results. (37) Beams of comparatively short span reinforced with bars of large size were used in order to develop high bond stresses and give bond failures. Most of the beams failed in bond; a few failed by a combina- tion of bond and diagonal tension or by tension in the steel. (88) The usual method of computing the bond stress in a rein- forced concrete beam does not take account of all the phenomena of bond action. Slip of bar due to beam bond action and the presence of anti-stretch slip may be expected to greatly modify the distribution of bond stress over the length of the bar, and otherwise to affect resistance to beam bond stresses. However, the nominal values for bond resistance, computed by the usual formula, form a useful basis for comparison in beams in which the dimensions and general make-up are similar. (39) Slip of bar was a phenomenon in all beam tests in which careful slip observations were made. These load-slip relations give important indications as to the bond stress developed at points along the length of the beam. (40) Slip was first observed in the middle region of the span at. loads producing a tensile stress in the steel of about 6000 lb. per sq. in. In this region the shear is zero and hence beam bond action, as usually understood, is absent. As the load was increased, slip of bar progressed through the outer thirds toward the ends of the beam at a rate nearly proportional to the increase of load. After slip occurred at the ends, the outer thirds of the length of the bar moved toward the middle of the span relative to the adjacent concrete. Slip of bar was probably partly responsible for the opening of outer cracks, since slipping was observed in the outer thirds of the beams before the cracks became visible. (41) The mean computed values for bond stresses in the 6-ft. beams in the series of 1911 and 1912 were as given below. All beams were of 1-2-4 concrete, tested at 2 to 8 months by loads applied at the one-third points of the span. Stresses are given in pounds per square inch. ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 217 Number First End End Slip Maximum of Tests Slip of Bar of 0.001 in. Bond Stress 1 and 1}4-in. plain round.............. 28 245 340 375 M-in. plain round.................... 3 186 242 274 Se-in plain round.................0.05 3 172 235 255 1-in, plain square...........0..... 000s 6 190 248 278 1-in. twisted square...............0005 3 222 289 337 1¥%-in. corrugated round.............. 9 251 360 488 (42) In the beams reinforced with plain bars end slip begins at 67% of the maximum bond resistance; for the corrugated rounds this ratio is 51%, and for the twisted squares, 66%. (43) The bond unit resistance in beams reinforced with plain square bars, computed on the superficial area of the bar, was about 75% of that for similar beams reinforced with plain round bars of similar size. (44) Beams reinforced with twisted square bars gave values at small slips about 85% of those found for plain rounds. At the maxi- mum load, the bond-unit stress with the twisted bars was 90% of that with plain round bars of similar size. (45) In the beams reinforced with 114-in. corrugated rounds, slip of the end of the bar was observed at about the same bond stress as in the plain bars of comparable size. At an end slip of 0.001 in., the corru- gated bars gave a bond resistance about 6% higher and at the maximum load, about 80% higher than the plain rounds. (46) The beams in which the longitudinal reinforcement con- sisted of three or four bars smaller than those used in most of the tests gave bond stresses which, according to the usual method of computation, were about 70% of the stresses obtained in the beams reinforced with a single bar of large size. The progressive opening of cracks with increase in load was well shown in these tests. These beams showed cracks nearer the ends than usual. The distances of the outermost cracks from the ends of the beams suggest that the unbroken length of embedment has an important bearing on the maximum loads which the beams may be expected to carry before failing by bond. It seems probable that the lower computed bond stresses in these tests are due to errors in the assumptions made as to the distribution of bond stress and not to actual differences of bond resistance in the bars of different size. (47) The tests on beams with the loads placed in different posi- tions with respect to the span gave little variation in bond resistance during the early stages of the tests. The maximum bond resistances increased rapidly as the load approached the supports. These tests indicate that the variation in the maximum bond stresses must be due to the presence of other than normal beam action. 218 ILLINOIS ENGINEERING EXPERIMENT STATION (48) Nearly all the beams tested on span lengths of 7 to 10 ft. failed by tension in the steel and did not develop the maximum bond resistance, although high bond stresses were obtained. The bond stress corresponding to first end slip of bar did not vary much with the span length. (49) The bond stresses developed in the beam tests indicate that with beams of the same cross-section the bond stresses are distributed in the same way during the early stages of the test in beams varying widely in span length and loading. During the later stages of the test, the distribution of bond stress seems to depend largely upon the conditions of stress in the concrete through the region of the span where beam bond stresses are high. The distribution of bond stresses in beams of different cross-section apparently varies with the relative dimensions of the beam and the reinforcing bars. (50) ‘The use of auxiliary tensile reinforcement in the outer thirds of the beam served to modify the distribution of bond stress during the early stages of the test, but did not have any influence on the maximum bond resistance. While the auxiliary bars seemed to prevent the opening of outer cracks, the tests indicate that interior cracks which did not appear on the surface of the beam may have opened to an extent that permitted the same distribution of bond stress as was found in other tests. (51) Increasing the thickness of the concrete below the reinforcing bars beyond the depth usually employed caused a very large increase in the resistance to bond and web stresses. The added stiffness of the beam and the increased flexural strength through the outer thirds of the span, prevented the formation of cracks in these regions. In the other beam tests such cracks were found to interrupt the continuity of bond action and to be an important factor in producing lower average bond resistances. (52) Increasing the length of overhang of the ends of the beam beyond the support did not increase the resistance to web stresses as indicated by the opening of outer cracks, but it had an influence on the bond resistance. The bond resistance at first end slip was greater in the beams with the longer overhang. The maximum bond resistance was materially increased by the additional overhang. (53) In the reinforced concrete beams it was found that very small amounts of slip at the ends of the bar represented critical condi- tions of bond stress. For beams failing in bond the load at an end slip of 0.001 in. was 89% to 94% of the maximum load found in beams ABRAMS—TESTS OF BOND BETWEEN CONCRETE AND STEEL 219 reinforced with plain bars, and 79% of the maximum load for similar beams reinforced with corrugated bars. As soon as slip of bar became general, other conditions were introduced which soon caused the failure of the beam. (54) The bond stresses developed in a reinforced concrete beam by a load applied as in these tests varies widely over the region in which beam bond stresses are present. High bond stresses are developed just outside the load, points at comparatively low loads. The load which first developed a bond stress nearly equal to the maximum bond re- sistance in the region of beam bond stresses produced a stress near the support which was not more than about 15 to 40% of the maximum bond resistance. As the load is increased, the region of high bond stress is thrown nearer and nearer the support, and at the same time the bond stress over the region just outside the load point becomes steadily smaller. This indicates a piecemeal development of the maximum bond stress as the load is increased. The actual bond stresses in certain tests varied from less than one-half to more than twice the average bond resistance computed in the usual manner. (55) Slip of bar in a reinforced concrete beam has a marked in- fluence in increasing the center deflection during the later stages of loading. (56) The comparison of the bond stresses developed in beams and in pull-out specimens from the same materials is of interest. Such a comparison should be made for similar amounts of slip. In the pull-out tests the maximum bond resistance came at a slip of about 0.01 in. for plain bars. The mean bond resistance for the deformed bars tested was not materially different from that of the plain bars until a slip of about 0.01 in. was developed; with a continuation of slip the projections came into action and with much larger slip high bond stresses were developed. The beam tests showed that about 79 to 94% of the maximum bond resistance was being developed when the bar had slipped 0.001 in. at the free end; hence the bond stress developed at an end slip of 0.001 in. was used as a basis of the principal comparisons in the pull-out tests. However, it is recognized that, under certain conditions, the stresses developed at larger amounts of slip may have an important bearing on the effective bond resistance of the bar. (57) The pull-out tests and beam tests gave nearly identical bond stresses for similar amounts of slip in many groups of tests, but it seems that this was the result of a certain accidental combination of dimen- sions in the two forms of specimens and did not indicate that the com- 220 ILLINOIS ENGINEERING EXPERIMENT STATION puted stresses in the beams were the correct stresses. However, it is believed that a properly designed pull-out test does give the correct value of bond resistance, and gives values which probably closely repre- sent the bond stresses which actually exist in a beam or other member as slipping is produced from point to point along the bar. The relative position of the bar during molding may be expected to influence the values of bond resistance found in the tests. (58)