U.S. AIR FORCE PROJECT RAND RESEARCH MEMORANDUM ‘THE STRENGTH OF ANCHOR BOLTS ‘SET IN CONCRETE Charles A. Sandoval 2, November 1954 Assigned to This is © working paper. It may be expanded, modified, or with- drawn at any time. The views, conclusions, and recommendations expressed herein do not necessarily reflect the official views or icies of the United Stotes Air Force. 7 RAND coerce 1700 MAIN ST. + SANTA MONICA + CALIFORNIA Copyright, 1955 ‘The RAND CorporationRMLOL 11-24-54, ni ABSTRACT This menorandun reviews cur present knowledge of the strength of anchor bolts set in concrete. Experimental data on the differeat types of loading (Aves, tension, shear, tond) on anchor bolts is adequate, but very little date on their ultinate strength 1s avatlable. Many types of anchors have been designed by rule-of= tmuab methods, without knowledge of their true behavior, Since the design of structures on the basis of ultimate strength (or plastic theory) 4a becoming more comon, it hes become essential to cbtain accurate knowledge of the actual structural behavior of the anchorages. This knowledge can only be obtained ty a comprehensive series of teste.R101 11-24-54, INTRODUCTION Considering that the failure of anchor bolte is one of the most important factors in the response of structures to air blast loading, it is the object of this report to review some of the pertinent experimental data and methods of analysis used in determining the strength of these bolts. Very little data is available on the ultimate strength of anchor bolts. There has been some research on anchor bolts subjected to direct tensile Loads. The paucity of information is perhaps due to the fact that anchor bolts represent a very minor item in the cost of a structure and designers can easily afford to be generous so that failure under ordinary service conditions is unlikely. With the present interest in air blast Joadings, (which induces very high lateral loads) failure of the anchorages becomes cute importent and perhaps primary mode of failure. ‘Todsy, design criteria and knowledge of the ultimate strength of anchor belts ere based mainly on the theories and results of pull-cut bond tests on steel rods. This limits the applicability of these results to a few types of anchore.RM 11-24-54, eae SCOPE This report will cover only anchor bolts with heads and straight or hooked bars. Some reference will be made to moment resistant column bases (with emphasis on the anchor bolts). Expansion bolts, "shot" bolts and other patented connectors will be cmitted. The test results of anchor bolts set in holes packed with lead wool have only been included as being part of a series of testa, The tests reported and the methods cf analysis may be classified as follows: 1. Type of loading 2. Type of stress 3. Type of bolt All of the tests referred to in this report, except one, were perfermed under one type of loading - static. The one exception is series performed under dynamic leading conditions, There is no reference made to the effects of creep, sustained or fatigue loading. ‘The types of stress considered are: i. Tension in the 2. Shear ny Combined shear and tension 7 méterial and, 1. Bond >) in the 2. Shear foundation materiel 3. Bearing A great variety of types of anchor bolte are used in service, but only & few more common types will be considered here. These types are described in the next section.RML01 1-24-54, TYPES OF ANCHOR BOLTS (Refer to corresponding type numbers in Fig. 1.) 1, Straight bar, usually plein, threaded at projected end to receive nut. 2. Plein ber vith semi~cirouler or ninety degree hook for added strength. 3. Bolt or bar with bearing plete. Bearing plate may be cut from flst plate, angle or channel. 4. Machine bolt with standard head. More comprehensive descriptions and construction procedures may be found in references 1 and 2. TESTS AND RESEARCH ‘There have been numerous tests performed on the physical properties of a etecl used for anchor bolts. The most comonly specified steel for anchor bolts is designated as A.S.T.M,-A7. It has the following properties: Tensile strength, pest. €,000 to 72,000 Yield point, min. pest. 0.5 x tensile strength but in no case less than 33,000 Elongation in 8 in., min., per cent. 1,500,000 tensile strength Elongation in 2 in., min., per cent. 22 A typical stress-strain diagram for this material is shown in Fig. 2. The ultimate shearing strength of the bolt material is specified as three-quarters of the ultimate tensile strength. Anchor bolts are seldomRMALLOL 11-24-54 a a) Fig. ‘Types of Anchor BoltsStress-K.S.I. 60; 50, 40 30 20 10. RY-1401 11-24-54 5 hee —5 , ; = “a 3 Oh 05 gee ee Strain = inches per inch Fig. 2 Typical Strecs-Strain Curve for Anchor Bolt MaterialREUOL 11-24-54 ne subjected to pure shear or pure tension. The usual leading 1s combined sheer and tension. Higgins and Munse(?) nave reported the results of a series of tests on rivets subjected to this combined loading. Fig. 3 1s adapted from the interaction curves given in their report. The ultinate strength of steel anchors set in drilled holes in concrete (4) and Richera, (5) The results of their vests has been determined by Graham and the aodes of failure are shown in Tables I end I, Wnittenore, Nusbaum and Seaquist made extensive testa on the impact and static tensile properties of bolts in 1938,(6) ‘Tne discussion so far has been concerned with the properties of the belt aateriel, ‘The strength of concrete in bend, shear and bearing has been determined from testa primarily designed to fix allowable stress values for beans, cclums and highway slabs, Abrams, in 1913, made the first coaprehensive tests on bond.(7) In 1938, Gilkey, Chamberlin and Beal shoved, in a serics of tests, that bond stresses were not proporticnal to concrete stresses above certain limiting velues.(©) Fig. 4 shove the relation of bond strength to concrete strength as determined in their tests. Fig. 5 shows the relation of Length te dianeter ratio and bond strength, also determined in the same series of teste. Cierk, in 1947-1951, performed teste at the Bureau of Standerds on the bond strength of the new type of deformed ters. (2) (20) In 1928, Mylrea tested the carrying capacity of hocked bars. In 1947 he developed a method of computing the carrying capacity of both straight (22) (29) and hocked bers. Fishburn'*9) hes performed tests on this strength and slip of bent bar anchorages in haydite concrete.Ultimate Tension Strength-K.S.1.-7 FM-1401 11-24-54, =7 60) 50| & 5 8 x 10 | | 1 | a) TX 20 30 a 5 Uitimate Shear Strength-K.S.1.-X Pig. 3 Simplified Curve for Conbined Rivet Stress Acapted from "How Much Combined Stress Can a Rivet Take" iegine and tase, Engineering Yews-Record, Decotber 4, 1952. Copyright’ ) McGraw-Hill Publishing Company. (All ordinates appearing in the Engineering News-Record curves have been multiplied by a factor of 3.)*huedwog SurysTTaN TTH-Meapog *CNST auBTAkdon ‘PIOSO-TRON THSSUPRE ot wor poaurxdey RHL01 24-54, “8 “PaTF®S POT - Took pest sduoo UT seq *Joq “peTFes puoq - Toon poet sdmop uF su UTETE *emRTPEJ 4oxI Aq ponoTtos Fayesou0 Jo emTTeS TwHTUT - B4Bq IUTCOTS *paTTEs §nowd “puoq UT PTaY - szUq *zPzOH reat Fag Jo epoy - - grt esc] ot — ss ver] s [se om | z o _- — zet_utet| 6 |— — eo ver] 7 fos ese | it of ee ree | ot gm eutt{ e |so oe | 2 [eo rer] ¢ | — — 1 x ee eee | ——]| fer os — = _— ow ev] @ @ 9 car ay —t | Toon poet | fanoz TT “oom pay aor TT Baoudnog sors 9:1 anoay THT paxomdmop sdots ait gnozy THT oaeq. | ox szeg Jeon | saw frox sxmg [ow Szeq |roq | sououy saysoq fasoz} —auog [aseg] vans fase | gue fassz | -apaon [aesz | suoepecy | sung pousogag ‘oxwbs “uy TT T seq punoy urera “ay 2 | SONAMHAANOUTO HVE 40 HONT Wid SdIX NI SGVOT GH @ waNn suOTVA ONY LINN WVE CM a SATA NT AUNYASISHY TLONZG Y WEONN SHNTVA ‘WAVMCHLIM OL Suva 4 TONVISISTN SUAONO NI SHON CUTTING NI SHOHONY "TawZS a I TTGVLREUOL H=2i-54 — T do. 0 0.0 aT us er T —-uoug 9fog JO "Treg *su9y Ut 7s et i (@UTT pa930p) *aomsD —2/T LS 8 f Jo 9400 4no xe9y a 8 #3899 Jo *oN sant Fag JO apoy PROT “ATA (*UT)P soto ¢ aT spvoy 74 4 2/9 9 S4TOE € JO aOASHTD UO sAEay, re) t oo “ . Let or s T — *aoltos amos J sey.oH yno=TTY Ler 8 q cee ets . Let L P y wueys 39 santTeg *sudy, 1x a 83892 FO “ON sant Peg JO SPO peoy "ata = (*8T)P :e9T0g StHUTS wo 6ys9y gree = LGBT = $109 Aa FOR POT “VIA {TOMS FTO syraed o6z = OF, frou0g Tz = weg rpms g/t = S/A tanga SvLN SRIOH CHTITUC NI IAS SWIOd WOHONY JO HLONIMLS IT 9TavLRM-1401 7 noses i asaee ds leo Concrete Strength- fc! p. Pig. Plain Bars’ 4 Series VI Tests pg 105 "Boni Between Concrete and Steel" Gilkey et al., Iowa Engr. Exp. Sta. Bull. 147, Dec. 1940 Reference: 5 al al (00s TPrevd-sseng puog, oymmryTnregu, 11-24-54 -n- 30 20 Ultimate Bond, Percent of Mtinate Compressive Strength 500. | 0 0 Ultimate Unit Bond stress 10 35 50 T/D Ratio Pig. 5 Relationsht) Ratio a1 000 teal Casting Reference: Fy 7% "Bond Between Concrete and Stee!" Gilkey, et, al., Iova Engr. Exp. Sta. Bull. 147, Dec. 1940 Load m Planes of Failure — u Vn Pig. 6 Ding eon FailsRELL 11-26-54, re The use of steel dowels in transverse pavenent jointe has prompted Tesearch on their load carrying capacity. Westergeara, “!2) Friberg, (13) (4) and others have tested and discussed dowels. Kushing and Frenont! Friberg found that the stresses in the concrete around the devel at the face of che joint were extremely high. le reporte a bearing pressure of 14,050 pest. for a 3/4 in, diameter dowel subjected to a lateral lead of 5000 Ibs. Marcus discusses these results and develops formule for conputing the ultinate load carrying capacity for dowels."5) From the date of these tests it appears that the strength of dovele carrying lateral loads Le much greater than allowed tn the bullding codes, (1127118) Most of the above investigations were not directly concerned with anchor bolts but the results may be adapted to then, The Engineering Research Institute of the University of Michigan has outlined a research progran thet would adé to the knowledge of the strength of concrete anchors.(9) this program vould include: 1, Tests on the bond strength for various types of anchors. 2, Tests on the effects of compressive, flexural and shear loading. 3. Tests on various colum-base details, i.e., resistance to monent, frictional resistence, amount of rotation. METHODS OF DESIGN AND ANALYSIS There is little information on the design of anchor bolts. For example, (20) the AISC Mamuel'“"’ merely specifies that anchor bolts shell extend into the masonry not less than 2 ft.-6 in., and farther when necessary, The AREA manuel of rail~ ways) gaye that anciior bolts shell be embedded « distance sufficient toAM OL 11-24-54, 13 resiatl.2ttines the uplift forces. ven the butlding codes are I1ttle nore explicit. ” Several Southern California ccdes state only the above and give alloveble shear values in concrete and brick masonry, (1127138) In large, heavy butlding structures the design of anchor bolte is almost unecessary. The large dead loads and relatively small lateral loads reduce ‘the uplift, and there is sufficient frictional resistence in the bearing plates to carry the sheer loads. In the case of butldings with fixed column beses (crane columns, for example); machinery bases, chimneys, stacks, vertical tanks and blest furnace stories the anchorages must be designed to be moment resistant. Design nethods for mement-resistant column bases are presented in various text books (2252342452526)27) one mgineering Research Institute has ae steel desten. outlined « method to determine the ultinate moment capacity and the rotation of fixed bases. (28) Most of the resistance to overturning loads in stacks, tovers and tanks must be provided by the anchor bolts since the deed load resteting moment 1s usually fairly small, The methods of design are semi-empirical (in some cases ruleof-thumb).(2134) the dynamic nature of te lateral loads (wind or earthquake), the effects of repeated or reversed loads and sustained loading are taken inte account ty specifying low design stresses (usually 12,000 to 15,000 p.s.t.) without the usual one-third increase. This value ts used to proportion the anchor bolts. The length of enbednent in the concrete, according to most desten standards, 10 such that the allovable tensile value of the anchor belt may be developed An bond. If sufficient length to develop this value 18 not available, the bar may be hooked or plates or structural shapes may beRHLOL 1-24-54 Tae welded to the lower end of the bolt to develop tie required strength in besring instead cf bond. Fellure in ancherages may be due te: 1. Fatlure of anchor bolts 2. Failure of concrete ‘This may be subdivided further. Anchor belts may fail in tensicn, shear or combined tension and shear, The ultimate load in tension is given by the expression: wnerey A, = Wet area (at root of threads) of anchor bolt cross section, in’ Sup = Ultimate tensile strength of bolt materiel, p.sst. The strain at thie point may be taken fron a stress-strain dlagran of the material (Fig. 2).* The ultimate load in shear is given by the expression: where, Gross area of enchor belt cross section, in® Ultimate shear strength of bolt material, p.s.t- S, is usually taken as three-fourths of the ultimate value in tension. ‘The combined failure load may be computed from the interaction curve of * the effective length in tension must be used in determining strain. See reference 19.RELL 11-24-54 n15— Fig. 3. The simplified straight-line diegram may be used. The tensile load will govern when the shear stress is less than 18 K.S.I, and the shear may be neglected as far as the bolt material is concerned. Failure of the concrete may occur in several ways. The most probable being @ pull-out failure. The ultimate strength of a straight bar in bond is given by the expression: Uap ok where, Uyyy = Ultimate bond strength ef coneretey pesite Circumference of anchor bolt at gross cross secticn, in. L = Length of embedment cf bar, in. It has been si wm ty Gilkey ct al that the maxim effective length in bend (L) ts about 24 dtancters. (6) Yhe strength of @ hooked bar is due to a combination of enbedment and hooking. Myireat) has stated that a load which stresses a hooked bar to 20,000 p.s.i. represents the carrying capacity of a hook. Accordingly, the ultimate strength would be given by the expression: P, cL! + 20,000 ag n= Mune where, L' = the length of enbedment from the face of the concrete to the point of tangeney of the hook. Veines of U may be taken from Fig. 4 or the results of other tests. Mylrea has presented a method of computing the bond strength which takes into accountRMLOL U1-24-54, -16- the dimintshing of the bond strength as the length of embednent increases. (11) Fatiure of the concrete pier in diagonal tension (Fig. 6) 42 quite possible vere shear loads are high and the pier 4o narrow, A nethod of analyzing this case has been developed by the Engineering Research Institute. (28) Marcus, (25) as a result of his tests, has developed a formula for comput- Ang the crushing load of @ steel dowel subjected to a lateral load, as follows: psd Sted), corr where, 4 = Diemeter cf anchor bolt, in. h = Distance from center line of bar te edge of concrete, in. © = Distence between the point cf application of the load and the face of the joint, in, = Length of embedment of bar, in. o£, the tensile strength of the concrete, pest. he compressive strength of the concrete, p.s.t. This formula was obtained fer unreinforced concrete (no stirrups or ties). Where short anchor bolts with a great deal of holding power are used, fetlure may ccour by tearing cut a plece of concrete. This type of faiiure has been reported by Grahan and Richard (see Tables I end II). More experi~ mental dete cn this type of anchorage are required, Another possible mode of failure ie the rupture of an unreinforced concrete pier due to tension. Thie is = foundetion failure and is not considered here.RM UOL 1-24-54, The effect of dynamic loads on the failure ef anchor bolts hes net been considered, Tests and experiments might uncover @ "dynamic correction factor" ‘thet could easily be used in analysis. In the overturning analyeie of steel stacks subjected to air blast loads there are two conditions whieh determine the upper and lover boundaries of structural response. The upper boundary is determined under the condition ‘that the anchor bolts in tension vill stretch like "spaghetti" and offer con siderable resistance to overturning. The lover boundary 1s fixed ty the com dition that the anchor bolte break before the structure has moved apprestably and offer little resistance to overturning. Tt may be seen from the above that accurate knowledge of the actual structural behavior of anchorages 1s lacking. Since the design of structures on the basis of ultimate strength (or plastic theory) is becoming more common, ‘the design of anchor bolts on this basis is most desirable. A comprehensive series of tests for the purpose of obtaining this knowledge appears a necessity.RMLOL 1-24-54 =16- References 1, Dunham, C. W., "Foundations of Structures," McGraw-Hill Book Co. Ins., New York, 1950, pp. 203-13. 2. Structural Bureau, “Fastening and Bedding Column end Machine Bases," Portland Cement Association, Chicago, December 1949, ST 61. 3. Higgins, T. R., Munse, W. H., "How Much Combined Stre: Take?" ‘Engr. News Record 149(23):40-H2, 1952. can 8 Rivet 4, Graham, H. E., "Strength of Steel Anchors in Concrete," Engr. News Record, Vol. 130, 1943, pp. 560-1. 5. Richard, E, L., "Anchor Bolts Set in Drilled Holes," Jour. ACI, Vol. 19, No. 2, Sept. 1947, pp. 61-: 6. Wnittenore, H., Nusbaum, G. W., Seaquist, E. 0., "Impact and static ‘Tensile Properties of Bolts," National Bureau of Standards Journal of Research, Vol. 14, 1938. 7. Abrans, D. A., "Tests of Bond Between Concrete and Steel," Universizy of Illinois Engineering Experiment Station Bull. No. 71, 1913. 8. Gilkey, H. J., Chamberlin, 8. J., and Beal, R. W., "Bond Between Concrete and Steel," Iowa Engr. Exp. Sta., Bull. 147, Iowa State College, Ames, Towa, 1940. 9. Clark, A. P., "Bond of Concrete Reinforcing Bars,” Journal of Research of the Rational Bureau of Standards, Vol. 43, No. 6, Dec. 1949, pp.565-79 10. Mylrea, T. D., “The Carrying Capacity of Seni-Cireular Hooks,” ACI Proc. Vol. 2h, pp. 20-272, 1928. 11. Mylrea, T. D., "Bond and Anchorage," Jour. ACI, Vol. 19, No. 7, Mar. 1948. 12. Westergaard, H. M., "Spacing of Dowels," Proceedings, Highway Research Boaré, 1928. 13. Friberg, B. F., "Design of Dowels in Transverse Joints of Conerete Pavenents," Laclede Steel Co., St. Louis. 1A. Kushing, J. W. and Frenont, W. 0., "Design of Load Transfer Joints in Concrete Pavenents," Proceedings, Highvay Research Board, 1940. 15. Mareus, H., “Load Carrying Capacity of Dovels at Transverse Pavement Jour. ACI, Vol. 23, No. 2, Oct. 1951, pp. 169-84. 16. “Los Angeles County Building Lavs," Colling Publishing Co., 1949, Pp. 126, 193. 17. “Uniform Building Code,” Pacific Coast Building Officials Conference, 1949, pp. 92, 159.RA-LIOL 11-24-54, =. 1B. Loe Angeles City Building Code," Building Code Publishing Co., 1953, pp. 126. 19. Hu, L. 8., Byce, R. C., Johnston, B. G., “Steel Beams, Connections, Colums and Frames," Dept. of Civil Engr., Univ. of Mich., Mareh 1352. AISC, “Steel Construction,” Amer. Inst. of Steel Construction, New ‘York, 1950. 21. AREA, "Manual for Railway Engineering," Aneriean Railvay Engineering Association, Chicago, 1952, Vol. 2, pp. 15-22. Lothers, J. B., "Design in Structural Steel,” Prentice-Hall, Inc., Tew York, 1953. 23. Urquhart, Le C., “Civil Engineering Handbook," MeGrav-ill Book Co., Inc., Hew York, 1950, pp. 573-5. 24. Hook. G. A., Kinne, W. 8., "Structural Mexbers and Connections,” McGraw-Hill Book Co., Inc., New York, 1942, pp. 313-18. 25. Hool, G. A., Kinne, W. S., "Steel and Timber Structures," McGraw-Hill Book Co., Inc., New York, 1942, pp. 53-4, 416-17, 479-82. 26. Grinver, L. E., “Theory of Modern Steel Structures," The Macmillan Co., New York, 1949, Vol. 1., pp. 112-13. 2]. Shedd, T. C., "Structural Design in Steel," John Wiley and Sons, Inc., New York, 1950, pp. 407, kB. 28. Schenker, L., Salmon, C. G., Johnston, B. G., "Structural Steel Comections,” Dept. of Civil ingr., Univ. of Mich., June, 1954, 29. Fishburn, C. C., "Strength and Slip Under Load of Bent Bar Anchorages and Straight Enbedments in Haydite Concrete,” ACI Jout., Vol. 19, Wo. 4, pp. 289-305, December 19h7. Merriman, T., Wiggin, T, H., "American Civil Engineer: Wiley and Sons, Inc., New York, 1942, 5th Ba, pp. 1287~ Handbook," John 31. Fleming, R., "Anchor-Bolt Tension: Six Different Results from Six Books," Engineering News, April 1914, pp. 956-58, 1089.