PROCEEDINGS PAPER Connections in Precast Concrete Structures—Strength of Corbels by L. B. Kriz and C. H. Raths* SYNOPSIS This paper describes a project directed toward development of design criteria for reinforced concrete corbels. Part 1 contains these design criteria, together with design aids and design examples. Part 2 describes the tests on which the proposed criteria are based, involving 124 corbels subjected to vertical loads only and 71 corbels subjected to combined vertical and horizontal loads. Part 3 contains the discussion and analysis of the experi- mental data and the derivation of the design equations. Detailed test data are given in an appendix. INTRODUCTION A series of investigations of con- nections in precast concrete struc- tures is in progress at the Research and Development Laboratories of the Portland Cement Association. The three previous papers in this series, collectively entitled "Connec- tions in Precast Concrete Structures," have been concerned with the strength and behavior of continuity connections in double-tee floor con- struction', with the bearing strength of column heads supporting precast beams 2 , and with the strength and behavior of scarf joints in beams and * Formerly, Development Engineer and Associate Development Engineer, respec- tively, Structural Development Section, Portland Cement Association Research and Development Division, Skokie, Illi- nois. columns3. This paper deals with the development of design criteria for the strength of corbels which pro- trude from the face of a column. PART 1—DESIGN OF CORBELS Background Corbels projecting from the faces of columns are used extensively in precast concrete construction to sup- port primary beams and girders. Typical applications of corbels may be found in the Prestressed Con- crete Institute manual of connection details4. Until recent years little research had been available on the strength of corbels. In the United States it has been customary to design them as short cantilevers, using the flexural and shear design equations derived for beams of more normal propor- 16 PCI Journal
46
Embed
Connections in Precast Concrete Structures—Strength of Corbels … · 2018-11-01 · PROCEEDINGS PAPER Connections in Precast Concrete Structures—Strength of Corbels by L. B.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
PROCEEDINGS PAPER
Connections in Precast ConcreteStructures—Strength of Corbels
by L. B. Kriz and C. H. Raths*
SYNOPSISThis paper describes a project directed toward development of design
criteria for reinforced concrete corbels. Part 1 contains these design criteria,together with design aids and design examples. Part 2 describes the testson which the proposed criteria are based, involving 124 corbels subjectedto vertical loads only and 71 corbels subjected to combined vertical andhorizontal loads. Part 3 contains the discussion and analysis of the experi-mental data and the derivation of the design equations. Detailed test dataare given in an appendix.
INTRODUCTION
A series of investigations of con-nections in precast concrete struc-tures is in progress at the Researchand Development Laboratories ofthe Portland Cement Association.The three previous papers in thisseries, collectively entitled "Connec-tions in Precast Concrete Structures,"have been concerned with thestrength and behavior of continuityconnections in double-tee floor con-struction', with the bearing strengthof column heads supporting precastbeams2, and with the strength andbehavior of scarf joints in beams and
* Formerly, Development Engineer andAssociate Development Engineer, respec-tively, Structural Development Section,Portland Cement Association Researchand Development Division, Skokie, Illi-nois.
columns3. This paper deals with thedevelopment of design criteria forthe strength of corbels which pro-trude from the face of a column.
PART 1—DESIGN OF CORBELS
BackgroundCorbels projecting from the faces
of columns are used extensively inprecast concrete construction to sup-port primary beams and girders.Typical applications of corbels maybe found in the Prestressed Con-crete Institute manual of connectiondetails4.
Until recent years little researchhad been available on the strengthof corbels. In the United States ithas been customary to design themas short cantilevers, using the flexuraland shear design equations derivedfor beams of more normal propor-
16 PCI Journal
lions. Since the assumptions made inderiving these equations are notvalid for deep beams, it is not sur-prising that corbel brackets designedby these equations can have varyingsafety factors. The tests described inPart 2 of this paper show that designon this basis will lead to question-ably safe designs when the amountof tension reinforcement exceedsabout one percent, and also if shearreinforcement is necessary and isprovided in the form of vertical stir-rups. In addition, corbels have ingeneral been designed for verticalloads only, although horizontalforces caused by restrained creep,shrinkage, and temperature deforma-tions of the beams supported by thecorbels are often important indeed.Tests described in Part 2 of thispaper have shown that such hori-zontal forces can substantially re-duce the vertical load-carrying ca-pacity of corbels. This effect has alsobeen evidenced in the field wheresome corbels carrying light verticalloads were damaged by horizontalrestraint forces.
In Europe the design of corbelshas been based mainly on the inves-tigations of Rausch 5 ' 6. These designprocedures involve the "straight-line"method of design for flexure, and theprovision for bent bars to resist allshear forces.
In 1961, Niedenhoff7 suggestedthat a corbel acts essentially as asimple truss composed of two mem-bers: a horizontal tension member,i.e. the tension reinforcement, andan inclined concrete compressionstrut. On the basis of an experi-mental investigation, Niedenhoff pro-posed that the depth of the equiva-lent truss be taken as 0.8 times thetotal depth of the corbel. These as-sumptions form the basis of Nieden-hoff's working load design proce-
dure.A series of tests conducted at the
University of Illinois 8 ' 9 ' 10.11 involvedthe strength of deep beams. A deepbeam, loaded by a concentrated loadat midspan and supported by con-centrated reactions at the ends, actsessentially as a double corbel pro-truding from opposite faces of acolumn. However, the number ofspecimens tested under concentratedloads was not sufficient to lead todesign procedures for corbels. Thesetests, together with recent tests ofshort cantilevers made at the Uni-versity of Texas 12, will be referredto later.
The tests recently carried out inthe PCA Structural Laboratory, andreported in this paper, have beenspecifically concerned with corbelsin which the ratio of the shear spanto the effective depth of the bracketat the column face was less thanunity. One hundred ninety-five cor-bels were tested, of which 124 weresubject to vertical load only and 71to combined vertical and horizontalloads. The variables included in thetests were: size and shape of corbel,amount of main tension reinforce-ment and its detailing, concretestrength, amount of stirrups, ratioof shear span to effective depth, andthe ratio of the horizontal force to.the vertical force.
The design criteria set out beloware based on a study of the results ofthese tests; they have also beenchecked against the results obtainedfrom the tests at the Universities ofIllinois 8 ' 9,10.11 and Texas 12 . In the de-velopment of such design criteria,numerous plots and numerical com-putations were made to compareobserved performance with variousempirical expressions. Considerableuse was made of electronic compu-tation to arrive at suitable ultimate
February 1965 17
strength design equations
Proposed Criteria for theDesign of Corbels
1. Notation
A3 = area of tension reinforce-ment, in.2
A, = total area of horizontalclosed stirrups, in.2
a = shear span, i.e. distancefrom column face to re-sultant of vertical load, in.
b = width of corbel, in.d = effective depth of corbel
measured at column face,in.
= concrete cylinder strength,psi
Vf' = relationship expressed inpsi, so that \/f = 60 psifor f = 3600 psi
(a) These provisions apply to cor-bel brackets having a shear span todepth ratio, a/d, of less than unity.
(b) Provisions of the ACI BuildingCode (ACI 318-63) not in conflict
with the provisions of these proposedcriteria should be considered applic-able to the design of corbels.
3. Safety Provisions andDesign Loads
(a) Strength should be computedin accordance with the provisionsof section 4.
(b) The coefficient 0 should be0.85.
(c) The strength capacities of cor-bels so computed should be at leastequal to the total effects of the de-sign loads required by Section 3(d).
(d) The design loads to be usedin the design of corbels should equalthe design loads specified in Section1506 of the ACI Building Code (ACI318-63), multiplied by 4/3.
4. Strength Computations
(a) When special provisions aremade so that a corbel is subject tovertical loads only, the ultimate de-sign load capacity may be calcu-lated by:
Vu = ca [6.5bd Vf (1— 0.5a/a)
where1000 1/3 1p p(A3 + A,,)/bd does not ex-
ceed 0.02, and A„ does not exceed A8.(b) In all other cases the ultimate
design load capacity may be calcu-lated by:
V. = 016.5bd A f- (1 - 0.5d/a)
(1000p)(1/3 + 0.4H/V) 1
"`^l'100.8 H/V (2)
where p = A8/bd does not exceed0.013.
5. Minimum reinforcement
(a) The amount of tension rein-forcement A 8 should be not less than0004bd.
(b) Closed horizontal stirrupsshould be provided having a total
18 PCI Journal
Note:Distance "x' should be great
•weldenough to prevent contact Same
Bar between outer corbel edge Sizeand beam due to possible rotations Detail A
Unrestrained RestrainedPrecast
Beam Precast
Beam BeamReinforcementElastomeric Pad Steel R Not Shown
2"min.- 5"min. 2min., }I I e 1
FA.m see A 8 min. ' see A s
min. h s rh/2 min. s h y j— h/2 min. s
vy
f d
(A) Corbels Subject To (8) Corbels Subject To
Vertical Load Only Vertical Load And RestrainedCreep And Shrinkage Force
Steel lk s Welded Or Not Welded
Fig. 1—Recommended Corbel Details
cross section A not less than 0.5A8
6. Detailing of Corbels°`
(a) The tension reinforcementshould be anchored as close to theouter face of the corbel as cover re-quirements permit, by welding across-bar to the ends of the tensionreinforcing bars. The size of thecross bar should be at least equalto the maximum size for bar usedas tension reinforcement.
(b) The closed horizontal stirrupsshould be distributed over the up-
* The requirements of Section 6 are illus-trated,in Fig. 1.
February 1965
per two thirds of the effective depthat the column face.
(c) The total depth of a corbelunder the outer edge of a bearingplate resting on the corbel shouldbe not less than half the total depthof the corbel at the face of the col-umn.
(d) The outer edge of a bearingplate resting on a corbel should beplaced not closer than 2 inches tothe outer edge of the corbel.
(e) When corbels are designed toresist horizontal forces, steel bearingplates welded to the tension rein-forcement should be used to transferthe horizontal forces directly to thetension reinforcement.
19
7. Bearing Stresses
(a) The bearing stresses at ulti-mate strength beneath a bearingplate resting on a corbel should benot more than O.5 f. '..
Discussion of ProposedDesign Criteria
Safety Provisions and Design Loads
The proposed safety provision anddesign loads are in agreement withthe philosophy concerning safetyprovisions and design loads of PartIV-B, Ultimate Strength Design, ofthe ACI Building Code (ACI 318-63).Since a corbel is primarily a sheartransfer device, and since its ulti-mate strength is governed by shearstrength, it is considered appropriateto use the value (A = 0.85 specifiedin ACI 318-63 for ultimate strengthgoverned by shear and diagonal ten-sion.
The design loads specified forcorbels are made one third greaterthan those specified for the designof members in ACI 318-63 for tworeasons. First, in corbels having lessthan about one percent of tensionreinforcement, yield of the reinforce-ment occurs before the ultimatestrength of the corbel is developed.The ratio of the load at which yieldoccurs to the ultimate load can vary
between % and 1. The load factorsproposed will provide an adequatefactor of safety against yield of thereinforcement, thus insuring service-ability of the corbels under moderateoverloads. Second, it is consideredgood practice that the strength ofa precast concrete structure shouldbe governed by the strength of themembers and not by the strengthof the connections between mem-bers. Since a corbel forms part ofthe connection between a beam anda column it should be made strongerthan either the beam or the column.Use of the proposed design loadswill assure this.
Strength Computations
The equations for ultimate strengthpresented in Section 4 and Part 3 arebased on a study of the results oftests of 195 corbels carried out at thePCA Structural Laboratory. Eq (2)reduces to Eq. (1) when H/V is zero.However, the different definitions ofreinforcement ratio p in Eqs. (1) and(2) should be noted. Whereas stirrupsmake a considerable and consistentcontribution to the strength of acorbel subject to vertical load only,their contribution to the strength ofa corbel subject to combined verticaland horizontal loads is smaller andmore variable. It is therefore con-sidered sounder for the present not
Table 1—Comparison of Test and Calculated Strengths
SourceI
Type of SpecimenI Number of
Specimens
Average H/V I V test
I V„ talc StandardI Deviation
PCA Corbels without stirrups 78 0 1.02 0.119PCA Corbels with stirrups 10 0 1.11 0.084PCA Corbels without stirrups 25 1/z 1.05 0.132PCA Corbels without stirrups 21 1 1.21 0.216PCA Corbels with stirrups 4 1 1.42
U of P'10 ' 1' Deep beams 23 0 1.01 0.134U of Pa Beams with aid = 1.33 14 0 1.14 0.168U of T' Short cantilevers aid < 1.10 6 0 1.03 0.066
20 PCI Journal
to rely on their contribution whendesigning a corbel subject to com-bined loading.
Eqs. (1) and (2) have been usedto calculate the strengths of 181members tested at PCA, the Uni-versity of Illinois, and the Universityof Texas. Tests involving local fail-ures resulting from inadequate re-inforcing details were excluded. Asummary of the results of this appli-cation of the proposed equations isset out in Table 1. In these calcu-lations, 0 was taken equal to 1.0,since accurate values of materialproperties and of dimensions wereknown.
The application of these equationsis simplified considerably by the useof design aids which are presentedfollowing this discussion.
Minimum ReinforcementThe minimum amount of tension
reinforcement is specified to insureagainst too rapid opening of cracksafter first cracking. The lower theamount of tension reinforcement, thelower is the ratio of load at yieldof tension reinforcement to ultimateload.
Closed horizontal stirrups are re-quired in all corbels to eliminatethe possibility of a sudden explosive-type failure of the corbel, which canoccur in a corbel without stirrups.
Detailing of CorbelsThe correct detailing of corbels is
fully as important as the over-alldesign of the reinforcement. Almostinvariably, distress of corbels in thefield can be traced to poor detailing.If the tension reinforcement is noteffectively anchored close to theouter face of the corbel, the fullstrength potential of the reinforce-
ment cannot be developed and fail-ure will occur at a lower load thanindicated by Eqs. (1) and (2). Therecommended form of anchorageusing a bar welded across the endsof the tension reinforcement isshown in Fig. 1. A frequently useddetail for the main tension reinforce-ment is shown in Fig. 2(a). However,in order to conform to Section 801of the ACI Building Code whichspecifies minimum bend radii forreinforcing bars, the bars are actu-ally bent as shown in Fig. 2(b). Fail-ure has then been observed, both inthe field and the laboratory, to occuron the surface indicated in Fig. 2(b),the tension reinforcement being by-passed completely. Welding of thebearing plate to the main reinforce-ment when horizontal forces act isspecified to eliminate the possibilityof a local failure of the concrete be-tween the bearing plate and the re-inforcement.
The horizontal stirrups are locatedso that they will be as effective aspossible, both from consideration ofultimate strength and for control ofdiagonal cracks. A suitable spacingof stirrups, s, is given by
3(n a1
where n is the number of stirrupsused. The stirrups should be placedin the corbel beginning at a distances from the tension reinforcement.Horizontal stirrups are used ratherthan vertical stirrups because of thesteep inclination of the diagonalcracks. These cracks can in somecases be almost vertical.
The limiting proportions of a cor-bel, and the limiting location ofthe bearing plate, are both recom-mended to insure against local fail-ures of the concrete before the p0-
February 1965 21
V V
H — H ^—
PotenSiafFailureSurface
(a) As Drawn (b) As Bent
V V
-H^— HE--
(c) Cracking In Corbel (d) Cracking In CorbelWith Too Shallow With Outer FaceAn Outer Face Of Sufficient Depth
Fig. 2—Corbel Details
tential strength of the corbel hasbeen developed. If the outer face ofthe corbel is made too shallow, theprincipal diagonal crack will take acourse as shown in Fig. 2(c), andwill intercept the sloping face ofthe corbel, resulting in instantaneousfailure. If the outer face is suffi-ciently deep, however, the principaldiagonal crack will take a course asshown in Fig. 2(d). In this case adiagonal concrete compression strutis formed as indicated, and a furtherincrease in load may be possibleafter formation of the crack. Loca-tion of the bearing plate too close
to the outer face of the corbel canresult in a bearing failure beneaththe plate at relatively low intensitiesof stress. This is particularly the caseif the load on the bearing plate be-comes eccentric. It is essential to in-sure that rotation of the end of abeam due to deflection under loadshall not result in the beam bearingon the outer edge of the corbel.
Bearing Stresses
Use of the maximum bearing stressof 0.5 f ' is contingent upon compli-ance with the requirements of Sec-tion 6(d). Bearing failures were ex-
22 PCI Journal
perienced at stresses lower than 0.5fin corbels loaded through bearingplates located closer to the outerface than two inches.
Design Aids and Design Examples
Design Aids
Design aids have been preparedto facilitate the use of Eqs. (1) and(2).
Eq. (1) may be written:
Vu = cpbd Vf^ Fl F2 (1a)
where Fl = 6.5 (1 — 0.5d!'), and
F2, = (1000p)113
Values of Fl and F2 are listed inTables 2 and 3.
Similarly, Eq. (2) may be written:
Vu = cbbd '/f ,' Fl F3 (2a)
where F3 = (1000p s
aiv H/V )
Values of F3 are listed in Table 4.Using Eqs. (1a) or (2a), and Tables2, 3, and 4, Vu may be readily eval-uated for given values of b, d, f fp, a/d, and H/V. The use of thetables is illustrated in the followingexamples.
Since both p and a/d can bevaried independently, design of acorbel must be by successive trials.This process is simplified by use ofthe design chart given in Fig. 3. Itis proposed that corbels be designedby successive trials using the designchart, and that the strength of thefinal design be checked using eitherEq. (1a) or (2a), whichever is ap-propriate. Use of the chart andequations in this manner is illus-trated in the examples.
Example 1
A typical interior corbel shown inFig. 4(a) projects from a 14 x 14-in.
February 1965
square tied column. It supports a50-ft span prestressed girder carry-ing a live load of 1500 lb/ft and adead load of 960 lb/ft. Design thecorbel for the vertical reaction fromthe girder, assuming that suitablebearings are provided to eliminatehorizontal restraint forces, and thatthe corbel does not have to resistwind or earthquake forces. Inter-mediate grade reinforcement is usedand f = 5000 psi. Tolerance gap be-tween beam end and column face isone inch.• Design Loads.
• The details of this corbel areshown in Fig. 4(a).
Example 2
Redesign the corbel of Example1 assuming that a bearing shoe inthe prestressed girder is welded tothe corbel, and because of this, ahorizontal force of 45 kips will occurdue to restraint of creep and shrink-age deformation of the girder. Thisexample is illustrated by Fig. 4(b).• From Example 1, V,, = 138 kips,and a = 4 in.• Section 1506(a)5, of the ACIBuilding Code (ACI 318-63), re-quires that the effects of creep andshrinkage be considered on the samebasis as the effects of dead load,when calculating the design ultimateloads. Hence the load factor for thehorizontal restraint force will be:
4 (1.5) = 2.0
H. = 2.0(45) = 90 kips
therefore
H/V 90/138=0.65
• From the design chart, the valueof v,, corresponding to the maximumallowable p (= 1.3%), H/V of 0.65,an assumed a/d of 0.3, and f ' of5000 psi, is about 460 psi.
Therefore
d = Vu - 138,000 _ 21.4 in.v,,b 460 x 14
now
a/d = 4/21.4 = 0.1926 PCI Journal
Z^ x 4' x 14". I °
Elostomeric Lubricatedx 14°
PodSteel
Weldedd To2 I" Pod 2" *8 Bars
gt^
j4574_
3" Welded #5
48
3" Welded #82"15°
44 TiesAt 3" c to c 23"
^. 7 #4 Ties
/At3"ctoc
#q
#4
14x14 in. 14x14 in.Col. i Col.
Example # 1 Detail Example *2 Detail
(a) (b)
Fig. 4—Corbel Details for Example Problems
• Determine p from the design chartfor vu = 460 psi, f = 5000 psi, a/d =0.19 and H/V = 0.65.
2.88 in. Use/3-in.. cen ters.Assuming a 1-in, thick bearingplate welded to the main tensionreinforcement, over-all depth ofcorbel:h 1+0.5+21.4== 22.9 in.Use 23 in,
• The details of this corbel areshown in Fig. 4(b).
Special Note
It should be noted that the addi-tion of the horizontal restraint forcehas necessitated an increase in depthof the corbel of 53 percent and anincrease in main tension reinforce-ment of 162 percent. It is clear,therefore, that for safety, a realisticestimate must be made of any hori-zontal forces that may act on acorbel. If special provision is not
February 1965 27
made to eliminate the horizontal re-straint forces by using lubricatedsandwich pads at one end of eachgirder, it is proposed that H/Vshould be assumed in design to beat least 0.5, unless the horizontalforce is calculated.
PART 2—TESTS OF CORBELS
Scope
Three series of tests were made:(a) exploratory tests, (b) tests of cor-bels subjected to vertical loads only,and (c) tests of corbels subjected tocombined vertical and horizontalloads. The exploratory tests involvedtesting procedures and reinforcing
detailing. The other two series in-volved a systematic investigation ofthe effect of different variables onthe strength and behavior of corbels.
The variables considered in thetests were: reinforcement ratio, con-crete strength, ratio of shear span toeffective depth, amount and distri-bution of stirrup reinforcement, sizeand shape of corbel, and the ratioof the horizontal applied load to thevertical applied load. The range ofthe variables is indicated on Fig. 5.
Test Specimens
All specimens consisted of a lengthof 8 x 12-in, column with two corbels
12
^ a a
Jh
Range of Variablesa - 2.75 to 12.5 in.
54° b -Sin.
72„ h - 18 to 45 in.h' - 6 to 26 in.ti - 6 to 24 in.f^- 2110 to 6680 psi
p-0.21 to 1.86%
fy- 39.9 to 95.8 ksi
Fig. 5—Corbel Test Specimen
28 PCI Journal
arranged symmetrically, as shownin Fig. 5. With the exception of cer-tain specimens in series (a) the maintension reinforcement consisted ofstraight deformed bars anchored bybars of equal diameter welded acrosstheir ends, as shown in Fig. 6. Cor-bels with horizontal stirrups weredetailed as shown in Fig, 6(b). Cor-bels to be subjected to combinedvertical and horizontal loading wereprovided with grooved bearingplates welded to the tension rein-forcement as shown in Fig. 6(c). Thedetailing of the reinforcement of thecorbels in the exploratory series (a)was as indicated in Fig. 7.
The dimensions of the individualspecimens and the material proper-ties are set out in Tables Al throughA4 appended to this paper.
Materials and FabricationAll concrete was made with Type
I portland cement. The coarse aggre-gate was a gravel of 3/4-in, maximumsize, and the fine aggregate wasElgin sand. The concrete slumpsvaried from 1% to 3 in. An air-en-training agent was added to produce4 to 6 percent air. One batch ofconcrete was used for each speci-men, with the exception of twolarge specimens, which required twobatches each. Three 6 x 12-in, cylin-ders were taken from each batch fordetermination of concrete strength.The specimens and test cylinderswere moist cured for three days un-der a plastic cover, and then storedat 70'F and 50 percent relative hu-midity, and were tested at six days.The concrete cylinder strengths
Fig. 6-Reinforcement Details of Test Corbels
February 1965 29
a.nus+ea.^ ^r.avffi.eoso^-
Type BI, Gent Reinf.
(a)
Type WC, Welded Cross Bar Reinf
(b)
Type WI, Welded Cross BarAnd Welded
Inclined Bar Reinf.(c)
Fig. 7—Detailing of Corbel Reinforcement in Auxiliary Test Series
varied from 2110 psi to 6680 psi, asgiven in Tables Al through A4.
The reinforcing steel conformed toASTM Designation A305 for defor-mations. The steel yield strengthswere determined from tension testsof 30-in, coupons taken from eachreinforcing bar used; the yieldstrength varied from 39,900 psi to95,800 psi, and are given in TablesAl through A4.
Instrumentation
The corbels were instrumentedwith SR4-A-12 strain gages mountedon the reinforcement and with SR4-A-9-4 strain gages mounted onthe concrete. This instrumentationvaried according to the purpose ofindividual tests.
Test Procedures
For convenience all corbels weretested in an upside-down position........A heavily-reinforced U-frame cen- -
tered under the loading platen ofa million-pound testing machine wasused to support the corbels. To as-sure adequate bearing capacity ofthe legs of the U-frame, the top ofthe legs was armored by steelplates. These plates were carefullyaligned in the forms of the U-framebefore placing the concrete to pro-vide parallel bearing surfaces.
The corbels were subjected to var-ious combinations of vertical andhorizontal loads. The loads were in-creased in increments until failure.After each load increment the de-velopment of cracks was observedand marked on the specimens. Allstrain measurements were recordedcontinuously by strip-chart strain re-corders
Vertical Loading OnlyThe corbels were loaded through
steel. bearing 'plates placed symmet-
30 : - PCI Journal
MachineLoad
8 Plywood
Load ISteel l^-'sMeasuring Half Round
Cell On OppositeSide
2 Plywood
Fig. 8—Vertical Loading of Corbelsrically on the top of the corbels asshown in Fig. 8. The length of thebearing plates was equal to thewidth of the corbels. The width wwas either 3 or 5 in. and the thick-ness of the plates was 1 or 1% in.To eliminate restraint of deforma-tions, a half-round and a round barwere placed between the bearingplates and another set of steel plateswhich rested on the supportingU-frame. The load was applied tothe bottom of the column stub bythe testing machine platen.
To assure uniform load distribu-
Top refers to the position in a structureand not to the position of the. specimenin the testing machine. This conventionis used throughout this paper.
tion on all bearing areas, new ply-wood inserts were used in each test.A 3/a-in, plywood sheet was placedbetween the column bottom and thetesting machine platen, '/s-in, ply-wood sheets between the corbels andthe bearing plates, and ½-in. ply-wood sheets between the U-frameand the second set of steel plates.After the application of the first10,000 lbs, the machine platen wasblocked to prevent its rotation.
In the first five tests the load ap-plied to the corbels was checked byload measuring cells to establishthat the load was distributed equallyto the two corbels. Since the twoloads did not differ by more thantwo percent, the use of these load
February 1965 31
measuring cells was discontinued infurther tests. The two test setupsare shown schematically in Fig. 8.Fig. 9 shows the test setup used forthe tests involving vertical load only.
Fig. 9—Test Setup for Vertical Load Only,H/V = 0
Three tests were made to deter-mine whether a column load carriedfrom upper floors influenced thestrength of the corbels. In thesetests a load was applied to the topof the column stub by a 100-tonhydraulic ram. A constant ratio ofthe machine load to the ram loadwas maintained throughout each ofthese tests. The loading of the ramwas controlled by the oil pressureindicator but the load was also con-tinuously monitored by a load-meas-uring cell placed between the ramand the column top. This test setupwas similar to that shown in Fig. 9,except for the 100-ton hydraulic ramwhich was within the U-portion ofthe test frame.
Combined Vertical andHorizontal Loading
The horizontal forces which de-velop in precast beams as a resultof restrained volume changes weresimulated by horizontal forces ap-plied at the level of the top of thecorbels. To permit a direct transferof the horizontal forces to the ten-sion reinforcement, the 3-in, bearingplates were welded to the reinforc-ing bars. The horizontal forces wereapplied by four or six hydraulic ramsto a set of loading plates, and trans-ferred to the bearing plates throughmilled shear keys. The hydraulicrams were positioned on each sideof the corbels in such a manner thatthe resultant of the ram loads wasat the level of the top of the corbel.The frictional restraint to lateral de-formations was eliminated by plac-ing 2-in, diameter round bars be-tween the loading plates and thesteel plates on the supportingU-frame.
The rams used for applying thehorizontal forces were calibrated sothat the loads could be correlatedwith the oil pressure. The operationof the rams during testing waschecked by load measuring cellswhich indicated that the errors inthe load as determined from the oilpressure were less than one percent.Therefore, the use of the load-meas-uring cells was discontinued.
The vertical load was applied inthe same manner as in the tests ofcorbels subjected to vertical loadsonly. A constant ratio between thevertical and the horizontal loads wasmaintained throughout each test.
The loading system for combinedhorizontal and vertical loading isshown in Fig. 10.
Test ResultsThe principal data obtained in
32 PCI Journal
Fig. 10—Test Setup for Combined Horizontaland Vertical Loading, H/V does not equal zero
these tests have been listed in TablesAl through A4 appended to thispaper. Other data are reproducedwhere appropriate in the discussionof the behavior of the corbels setout in Part 3.
PART 3—BEHAVIOR OF CORBELS
Series (a)—Exploratory Tests
Effect of Additional Column Loads
Three tests were made on pairs ofidentical specimens. One of each ofthe companion specimens was sub-jected to vertical loads applied tothe corbels only, while the otherspecimen was subjected also to anadditional load applied at the topof the column stub. The pertinentdata are given in Table 5. Thesetests show that the strength of thecorbels is not significantly influencedby the additional load carried by the
column. Therefore, subsequent testswere performed with loads appliedto the corbels only.
Detailing the Corbel Reinforcement
Test of corbels reinforced conven-tionally according to Fig. 7(a) haveshown the weakness of such detail-ing when loads were applied closeto the outer edges of the corbels.These corbels failed along a surfacefollowing the bends of the reinforce-ment, Fig. 11, indicating that the
Fig. 11—Conventionally Reinforced Corbel, TypeBI, Loaded Near Outer Edge. Failure PlaneFollows Bend Radius
Table 5—Effects of Additional Column Load
Concrete Load per AdditionalSpecimen Strength, f' Effective Depth Corbel at Column
reinforcement was not fully effectiveand that it even created a possiblesource of weakness. Measurementsof the strains in the reinforcementalong the compression side indicatedonly small compressive stressesthroughout the length of the rein-forcement.
Previous tests of corbels and ofdeep beams 8 ' 9.11 and the tests re-ported herein, show that the stressesin the tension reinforcement of acorbel do not vary significantly alongits length between the face of thecolumn and the point of load appli-cation. Consequently, high bondstresses exist in the outer parts ofthe tension reinforcement and maylead to bond failures. Such bondfailures were observed in tests ofdeep beams$. The anchorage of thebars can be assured by cross-barswelded to the ends of the tensionreinforcement as shown in Fig. 7(b).This method proved satisfactory andsubsequent tests were made on speci-mens reinforced with straight ten-sion bars anchored by the weldedcross-bars.
Tests of corbels with inclined com-pression reinforcement welded tothe ends of the tension reinforce-ment, Fig. 7(c), show the compres-sion reinforcement contributes littleto the strength of the corbels. There-fore, compression reinforcement wasnot used in further tests.
The strength of corbels with thethree types of reinforcement is com-pared in Table Al. The specimensdesignated by letters WC had ten-sion reinforcement with welded crossbars,. Fig. 7(b), specimens BI hadbent reinforcement, Fig. 7(a), andspecimens WI had compression re-inforcement and cross-bars weldedto the ends of the tension reinforce-ment, Fig. 7(c).
The- arrangement and amount of
reinforcement in the column haslittle influence on the strength of thecorbels projecting from the column,as may be seen in Table Al. Thus,the amount of column reinforcementused in subsequent tests was thatwhich would prevent failure of thecolumn portion of the test speci-mens.
Series (b)—Corbels Subject toVertical Loads Only
Behavior Under Load
Initially the corbels behaved elas-tically, and the stress in the maintension reinforcement was propor-tional to load. In all the tests, thefirst cracks to appear were flexuralcracks starting at the junction ofthe horizontal face of the corbel andthe face of the column. After for-mation of these cracks the tensionreinforcement stress increased muchmore rapidly. Typical relationshipsbetween applied load and force inthe tension reinforcement are shownin Fig. 12. Subsequent developmentof the cracks depended primarily onthe reinforcement ratio and the ratioof the shear span to the effectivedepth, and was also closely relatedto the mode of failure.
`Four principal types of failurewere observed, as described below.
• Flexural Tension—A flexuraltension failure occurs by crushingof the concrete at the bottom of thesloping face of the corbel after ex-tensive yielding of the tension rein-forcement. Such a failure is illus-trated in Fig. 13(a). The appearanceof a corbel after a flexural tensionfailure is characterized by very wideflexural cracks.
• Flexural Compression—A flex-ural compression failure occurs whencrushing of the' concrete takes placeat the bottom of the corbel before
34 PCI Journal
CCDC-
0
200
180
v va °
0
160hI d
o•
o.o /
1400
0 0
• o /
0 • • e
120o 0 0 •
0 0 ° • •
/
e
• o 0
1NY 100
P i e° ° • .• /e o
^
/• • m//\\/mo • • • e e Eq.(4)
00 • • • e.^ a/d %/ a/d! ///0•••/ a e. e 0.14
0
/e40w e n^ h = 18". 0.30--
0 ///
// i/o 0.20
h = IS !• 0.30--
20f'4000 psi ec^ 0.53--- t^= 4000 psi a .0.37---NO STIRRUPS
extensive yielding of the reinforce-nnent has occurred. The tension re-inforcement stress at failure is eitherbelow or just at the yield point andthe flexural cracks, while well de-veloped, have not opened excessive-ly. Such a failure is illustrated inFig. 13(b).
• Diagonal Splitting—The diag-onal splitting mode of failure isshown in Fig. 14(a) and 14(b). Theflexural crack pattern was well de-veloped before the diagonal splittingof the concrete, which occurredalong a line extending from the bear-ing plate toward the junction of thesloping face of the corbel and theface of the column. A corbel withsuch a crack usually fails by shear-compression of the concrete com-pression zone, as in the corbel shownin Fig. 14(b).
• Shear Failure—Shear failureswere characterized by the develop-
ment of a series of short inclinedcracks along the plane of the inter-face between the column and thecorbel, as may be seen in Figs. 15(a)and (b). The final failure was byshearing along this weakened plane,and the appearance after failure canbe seen in Fig. 15(b).
• Secondary Modes of Failure—Failures which did not involve thedeepest section of the corbel at thecolumn face were considered . see-ordary modes of failure. These wereof two types: (a) the splitting awayof a portion of the concrete due toa major crack intersecting the slop-ing face of the corbel, as seen in Fig.16(a), and (b) bearing failures ofthe concrete beneath the bearingplate, as seen in Fig. 16(b). Bothtypes of secondary modes of failureoccurred at loads lower than thoseat which failure would have oc-curred by one of the principal modes
36 PCI Journal
(a) (b)
Fig. 14—Diagonal Splitting Failures (DS), H/V = 0
(b)
Fig. 15—Shear Failures (S), H/V = 0
February 1965 37
(a) Corbel End Failure (CE) (b) Bearing Failure (B)
Fig. 16—Secondary Modes of Failure, H/V = 0
of failure had the secondary fail-ures been prevented.
Discussion of Behavior
To understand the behavior ofcorbels and to arrive at design equa-tions, extensive plotting of test datawas made. During such studies,further tests were conducted to coveradequate ranges of the significantvariables. Empirical design equa-tions were gradually arrived at bynumerous comparisons of observedproperties to those computed byvarious expressions. An LGP-30 elec-tronic computer was used in thesestudies.
The relationships between tensionreinforcement force and applied loadshown in Fig. 12 are for corbelsmade from concrete with a strengthof about 4000 psi. Similar relation-ships were found to hold for corbelswithout stirrups made from 2000 and6000-psi concrete. It was not con-
38
sidered necessary to test corbels withstirrups made from concretes havingstrengths other than 4000 psi. It wasfound that the tension reinforcementforce, A S f s , is a function of the ap-plied load, V, of the ratio of shearspan to effective depth, a/d, and ofthe concrete strength f r . The rela-tionship between load V and tensionforce A s f s can be idealized as shownin Fig. 17. The linear part of the
UltimateVa
InitialYield
VY. V=V,+mAsfs
v, .InitialCracking
Asfs — FORCE IN TENSION REINFORCEMENT
Fig. 17—Idealized Relationship Between AppliedLoad and Force in Tension Reinforcement
PCI Journal
relationship between first crackingand yield of the reinforcement canbe represented by the equation:
V=V„+mA 3f8 (3)where V = applied load
Vo = nominal cracking loadm = slope
The nominal cracking load, V0, andthe slope, m, are both functions of fand a/d. These functions can beexpressed as:
44 f 1/3V0 — bd— ± — (a/d)v (a%d) Ci
and m— 2 f^ 13 1000 C2
Substituting for Vo and m in Eq. (3)above yields:
V 4.4 1/3v= — - f' C^
bd (a/d)^ a/d
2 f^ pf (4)+ 31000 C^
where C, = 1 for vertical loads onlyC2 = 0.8(10)¢13d when there
are no stirrups= 0.25(10)'" when there
are stirrupsEq. (4) may be used to calculatethe nominal shear stress, v, at work-ing load by substituting the allow-able steel stress for f8j and can beused to calculate the nominal shearstress at yield of the tension rein-forcement, v,, by substituting fy forf8.
Eq. (4) has been used to calculatethe nominal shear stress, v, at yieldof tension reinforcement in thosecorbels tested in which yielding oc-curred. The average value of (v0test/va calc) given in Table A2 is1.06 and the standard deviation is0.135. When the computed steelstress, fsu, given by Eq. (5) belowwas less than the yield point of thesteel used, no value for vy calc is
given in Table A2.Eq. (4) can also be used to de-
fine whether or not the tension re-inforcement will yield prior to thecorbel developing its ultimatestrength. If the nominal shear stressat ultimate strength is v,,, (= V,,/bd),then transposing Eq. (4) and sub-stituting v for v yields:
4.4 f' 1/3f8u = vu (a/d),, ( a d ) Cl ]
1.5C^
X p'/f'/1000 (5)
in which the stress, vu may be calcu-lated from Eq. (7).
The tension reinforcement willyield if f3 .0 calculated using Eq. (5)is equal to or greater than the yieldpoint stress f,.
To facilitate the use of Eqs. (4)and (5), values of C l and C2 havebeen listed in Tables A7 and A8appended to this report.
For purposes of practical design,it should usually not be necessary tocheck the stress in the tension rein-forcement. As indicated in the dis-cussion of design criteria in Part 1,yield of the tension reinforcementwill usually take place at % to 1times the ultimate load. The pro-posed ultimate strength procedureaccounts for this by specifying loadfactors 1/ greater than those usedfor the individual precast members.
Ultimate Strength
The ultimate strength equationmust of necessity be empirical be-cause of the complexity of the stateof stress in the corbel. Several con-clusions concerning the effect of in-dividual variables on the strength ofcorbels can be drawn on the basisof the experimental data presentedherein. These conclusions, togetherwith the requirements of the laws of
February 1965 39
similitude, lead to a suitable formfor the ultimate strength equation.
The ultimate strength of a corbel,Vu, is a function of its width b andeffective depth d, of the reinforce-ment ratio, p (= A$/bd), of the con-crete strength f ' and of the ratioof the shear span to the effectivedepth, a/d. From the laws of simili-tude it is concluded that the ultimatestrength, V,4, must be directly pro-portional to the width b and to theeffective depth d. The tests haveshown that the strength is also pro-portional to. Accordingly, thestrength may be expressed interms of the non-dimensional ratioVu/bdV/f^. This ratio must be a func-tion of the remaining two variables,a/d and p.
The tests show that increasing thea/d ratio lowers the corbel strength,V,,. The maximum strength is ob-tained for a = 0, while a = QO repre-sents the condition of pure bending..Hence, V,, = 0 when a = -. Thevariation of the strength with a/dcan be represented by a term of theform Kl (1 — K2dia) where K2 is lessthan unity.
These tests also show that thestrength increases when the rein-forcement ratio increases. The effectof the reinforcement ratio can beexpressed by the term K3pK4. Theforegoing analysis leads to the ex-pression:
Vu
bd \'T= , = Kl (1 — Kea/a) K3px4 (6)
The constants Kl and K3 need notbe known separately and may becombined into a single coefficient.Statistical analysis of the test dataresulted in the following equation:
v______ V.
'j- 'lf'= 6.5 (1 - 0.5d/a) (I000p) 1 "3 (7)
Multiplying both sides of Eq. (7)by bdVV,' and introducing thestrength reduction factor di yieldsEq. (1) of the proposed criteria fordesign of corbels.
Eq. (7) was used to calculate thenominal shear stress at ultimatestrength, vu, for all corbels subjectedto vertical loads only, and the re-sults of these calculations are listedin Table A2. Excluding those speci-mens which experienced secondaryfailures by bearing or splitting offof the corbel end, the average valueof (v test/vs calc) was found to be1.02, and the standard deviation0.119.
Analysis of data from tests of cor-bels with horizontal stirrups showsthat the stirrups are as effective inresisting vertical loads as is the maintension reinforcement. Accordingly,the strength of a corbel with hori-zontal stirrups and subject to verticalloads only can be calculated usingEq. (7) but calculating p on thebasis of the total cross section oftension and stirrup reinforcement,i.e. p = (A, + A)/bd. The calculatedultimate strengths of corbels withstirrups and subject to vertical loadslisted in Table A4 were determinedin this manner. The average value of(vu test/vu calc) was 1.11 and thestandard deviation 0.084.
Fig. 18 shows a graphical repre-sentation of Eq. (7), together withthe corresponding test values. Thetest results from corbels which ex-perienced secondary failures are notincluded in this figure.
In Table A5 comparisons havebeen made between data obtainedby other investigators at the Uni-versities of Illinois and Texas, andthe ultimate strengths calculatedusing Eq. (7). A satisfactory agree-ment is found.
40 PCI Journal
3.4
3.2
3.0
2.8
2.6
24
2.2
2.0
•l.8LU0
1.6tr 1.4
1.2
1.0
0.8
0.6
04
0.2
q
V u ft (1/3+0.4 H/Vi
L
bd f-6.5(I- 0.5d/a)
=(10)o.8H/V
o
o 00
0
0
000
0 0 0
0 0 .0
0 00
00
0000 0
8 0•
80g
0
^00
•
.
8o 0 0
0•°
0• ZS
0 •
• •
•
4.\1 ^
s/°• •
I • • H/V
0t 0 0 0
I/2
f/s I/I
2 4 6 8 10 12 14 16 18 20I1000 p
Fig. 18—Ultimate Strength of Corbels
February 1965 41
Series (c)—Corbels Subject toCombined Vertical andHorizontal Loads
Discussion of Behavior
The addition of outward hori-zontal forces to the vertical loadsdoes not change the essential char-acteristics of behavior, which canstill be represented by the idealizeddiagram of Fig. 17. However, thefunctions for V0 and m must be mod-ified to account for the lower valuesof the nominal cracking load V0 andof the slope m observed in data fromtests of corbels subject to combinedloading. Typical relationships be-
tween applied load and tension re-inforcement force for corbels sub-jected to combined loading areshown in Fig. 19.
The function for v derived fromthe data shown in Fig. 19, and fromother similar data not presentedhere, takes the form:
VV =
bd _—
4.4 f • 2 f' pfs'2 a%d)1/ Cl + 3 1000 C2
1 + 2H7 (8)3 V 1^ 1000120
V V
loo H --►H °
h d °
80E4• (8)
°
V 260 h_18„ o . p
'° ° /°fJ4000 psi °
/ '40
N
No Stirrups o ^/
o • ^G A^^'o o ^c
20—Ql
0.37 — A 0.62 ---
0
H - I Eq. (8)60 V
h=18° /•0 0
f^= 4000 psiao o o _ ^^^
No Stirrups o s/ o—
° ° ^ a—20 °°
a.v11
0 10 20 30 40 50 60 70 80 90 100A sfs , KIPS
Fig. 19—Relationship Between Applied Load and Tension Steel Force,Combined Vertical and Horizontal Loading
42 PCI Journal
where C1 = 1.5 (a/d)2 / 3, and C2 =0.7 (10) 012d, whether stirrups arepresent or not. Eq. (8) reduces toEq. (4) when H/V = 0, i.e., for verti-cal load only. However, it should benoted that coefficients C l and C2must then be as defined earlier forEq. (4).
Eq. (8) has been used to calculatethe nominal shear stress„ v, at yieldof the tension reinforcement in thosecorbels tested in which yield of thetension reinforcement occurred. Theresults are given in Table A3. Theaverage value of (vy test/vu calc)was 1.04 for H/V = 1/z, and 0.92 forH/V = 1h, the standard deviationsbeing 0.088 and 0.084 respectively.
As before, by equating Eq. (8)to the nominal shear stress at ulti-mate strength, vu, and transposing,the reinforcement stress at ultimatestrength, f can be determined.
f8 =[
vuf1 +2H f 4.4l 3 V1000} (a/d)'
1/3 1.5502(fl Cl (9)
a/d p Vf'/100u
where Cl and C2 are as defined forEq. (8) above, and v,, is obtainedfrom Eq. (10) below. Values of C1and C2 are also listed in Tables A9and A10 appended to this report.
For purposes of practical design,yield of the tension reinforcementmay again be accounted for by theuse of load factors 1/3 greater thanthose specified for individual mem-bers.
Ultimate Strength
The principles used in the deriva-tion of the ultimate strength equa-tion for corbels subjected to verticalloads only apply also to the deri-vation of an ultimate strength equa-tion for corbels subject to combined
horizontal and vertical loads. Theultimate strength Vu must againbe proportional to b and d, andit may be assumed that it is alsoproportional to Vf'. The ratioV„/bd \I f,; is then a function ofa/d, p and H/V, which should re-duce to Eq. (7) when H/V = 0, i.e.for vertical loads only. The follow-ing equation ! was established afterstudy of the test data, having inmind the above requirements.
vu = bd = 6.5 elf, (I — 0.5d /¢)
/1000,,.,)(1/3 + 0.4H1V)l ( !11
0)0.sH /v (10)
Eq. (10) was used to calculate thenominal shear stress at ultimatestrength for all corbels subjected tocombined vertical and horizontalloads, and the ultimate shear stressesso calculated are set out in TableA3. Eq. (2) of the proposed designcriteria is based on Eq. (10). Exclud-ing those specimens which experi-enced secondary failures (i.e., bybearing or by splitting off of thecorbel ends), the average value of(vu test/vu talc) was 1.05 for H/V =%, and 1.21 for H/V ='h, the stand-ard deviation being 0.132 and 0.216,respectively.
The appearance of typical corbelsafter failure under combined loadingis shown in Figs. 20 and 21.
A limited number of corbels withstirrups were tested under combinedloading, and the results are given inTable A4. It was found that thestirrups did not increase the resist-ance of a corbel to combined loadingby as large a proportion as was thecase with a corbel subject to verticalload only. Also, the contribution ofthe stirrups was more erratic, viz.corbels 13S and 14S with 0.34% and
February 1965 43
Fig. 20—Flexural Yielding Failure Followed byCrushing of the Concrete (FT), H/V = 1
Fig. 21—Shear Failure (S), H/V = 1/2
0.93% of stirrup steel, respectively,and all else the same, gave ultimateshear stresses of 260 and 273 psi.The effectiveness of the stirrups isalso apparently a function of thea/d ratio and of the H/V ratio. Aconsiderable program of tests wouldbe necessary to assess the influenceof the various factors which appar-ently influence the effectiveness ofstirrups in a corbel subject to com-bined loading. For the present itwas decided that any contributionfrom the stirrups should be regardedas reserve strength, and should notbe taken into account in design.Stirrups do lead to a more ductileform of failure, and hence it wasconcluded that a minimum amountof stirrups should always be pro-vided.
Secondary Failures
The following comments apply toboth vertical load only and to com-bined vertical and horizontal load-ing.
Corbel End Failure—In certain ofthe tests the depth of the outerface of the corbel was deliberatelyvaried in order to determine theminimum depth necessary to preventthe occurrence of a secondary failureby splitting away of a portion of theconcrete at the tip of the corbel. Itwas found that this type of failure,as shown by Fig. 22, did not occurin those corbels having a depth be-low the outer edge of the bearingplate greater than about 0.5 thedepth of the corbel at the face ofthe column.
Bearing Failure--Crushing of theconcrete below the bearing plate oc-curred in some of the tests. Thebearing stress, fbu, a t ultimatestrength of the corbels is listed inTable A6. Bearing failures occurred
44 PCI Journal
Fig. 22—Corbel End Failure (CE), H/V = 7
at stresses as low as 0.34f', when theload was applied near the outer edgeof the corbel in a combined loadingtest. However, if the outer edge ofthe bearing plate was at least 2 in.from the outer face of the corbel,then bearing failures did not occur atbearing stresses less than 0.5 f' N de-tailed study of bearing stresses wasnot made. It is believed that 0.5 f '.
is a suitably conservative value.
CONCLUDING REMARKS
The experimental evidence pre-sented in this paper indicates thatthe nominal ultimate shear stress, vu,in corbels with a shear span to effec-tive depth ratio less than one mayexceed the maximum shear stressallowed by Chapter 17 of the ACICode (ACI 318-63) for beams witha/d ratio greater than one.
The nominal ultimate shear stressin a corbel is a function of the ratioof the shear span to the effective
depth, of the reinforcement ratio, ofthe concrete strength, and of theratio of the horizontal and verticalcomponents of the applied loads.
Horizontal forces acting outwardfrom the column significantly reducecorbel strength, and must be con-sidered in the design of a corbelunless special provisions are madefor free movements of the supportedbeams.
Tension reinforcement and hori-zontal stirrups are equally effectivein increasing the strength of a corbelsubject to vertical loads only. How-ever, the effective amount of rein-forcement is limited.
Loads carried by a column do notaffect the corbel strength, nor doesthe amount or arrangement of col-umn reinforcement.
The results of this investigationhave been used as a basis for theformulation of "Proposed Criteria forthe Design of Corbels" which is pre-sented in Part 1 of this paper.
ACKNOWLEDGMENTS
The work described herein wascarried out in the Structural Labora-tory of the Portland Cement Asso-ciation under the direction of EivindHognestad and Alan H. Mattock.Contributions were made by severalmembers of the laboratory staff. Par-ticular credit is due Bernard J.Doepp, William Hummerich, Jr.,David C. Yates and Kenneth Hirtefor the laboratory work involved.
NOTATION
The notation of the ACI BuildingCode (ACI 318-63) is used whereverapplicable. The letter symbols usedin this paper are defined below:
A8= area of tension reinforce-ment, in.2
A„ = area of horizontal stir-rups, in.2
February 1965 45
a
bd
= shear span measured fromthe face of the column tothe resultant of appliedload, in.
= width of corbel, in.= effective depth of the cen-
troid of tension reinforce-ment at the column face,in.
fbe = bearing stress at ultimatestrength, psi
f$ = stress in tension reinforce-ment, psi
fs^ = stress in tension reinforce-ment at ultimate strength,psi
Iv = stress in horizontal stir-rups, psi
fl, = yield stress of reinforce-ment, psi
fC = concrete cylinder strength,
psi
Vile = relationship expressed inpsi, so that /f " = 60 psifor f' = 3600 psi
H/V = ratio of horizontal andvertical components ofapplied loads
h = over-all depth of corbel atcolumn face, in.
h' = depth of corbel outerface, in.
n = number of horizontalclosed stirrups
p = reinforcement ratio = (A8+ A)/bd when H/V =0= A,/bd when H/V doesnot equal zero.
V5 = vertical load at initialyielding of tension rein-forcement, lb
va = nominal shearing stress atinitial yielding of tensionreinforcement = V5/bd,psi
w = width of bearing plates,in.
cA = capacity reduction factor
REFERENCES
1. Rostasy, F., "Connections in PrecastConcrete Structures—Continuity inDouble-T Floor Construction," Journalof the Prestressed Concrete Institute,Vol. 7, No. 4, August 1962, pp. 18-48;PCA Development Department Bulle-tin D55.
2. Kriz, L. B., and Raths, C. H., "Con-nections in Precast Concrete Structures—Bearing Strength of Column Heads,"Journal of the Prestressed Concrete In-stitute, Vol. 8, No. 6, December 1963,pp. 45-75; PCA Development Depart-ment Bulletin D73.
3. Gaston, J. R., and Kriz, L. B., "Con-nections in Precast Concrete Structures—Scarf Joints," Journal of the Pre-stressed Concrete Institute, Vol. 9, No.3, June 1964, pp. 37-59; PCA Devel-opment Department Bulletin D79.
4. "Connection Details for Precast-Pre-stressed Concrete Buildings," PCICommittee on Connection Details, Pre-stressed Concrete Institute, October1963.
5. Rausch, E., "Berechnung der Abbie-gungen gegen Abscheren" (Design ofBent Bars for Shear), Der Bauingeni-eur, Vol. 3, No. 7, Berlin, April 1922,pp. 211-212.
6. Rausch, E., "Beanspruchung auf Ab-scheren im Eisenbetonbau," (Shear inReinforced Concrete Structures), DerBauingenieur, Vol. 12, No. 32/33, Ber-lin, August 1931, pp. 578-581.
7. Niedenhoff, H., "Untersuchungen UberDas Tragverhalten von Konsolen andKurzen Kragarmen," (Investigations ofBehavior and Strength of Corbels andShort Cantilevers), Desertation, Tech-nische Hochschule Karlsruhe, 1961,115 pp.
8. Austin, W. J., Egger, W., Untrauer,
V0Vuvu
46 PCI Journal
R. E., and Winemiller, J. R., "An In-vestigation of the Behavior of DeepMembers of Reinforced Concrete andSteel," Air Force Special Weapons Cen-ter Report AFSWC-TR-59-18 (1960).(Also Civil Engineering Studies, Struc-tural Research Series No. 187, Depart-ment of Civil Engineering, Universityof Illinois, Urbana, Illinois, January1960, 103 pp).
9. Untrauer, R. E., and Siess, C. P.,"Strength and Behavior in Flexure ofDeep Reinforced Concrete Beams Un-der Static and Dynamic Loading," AirForce Special Weapons Center ReportAFSWC-TR-61-47, Vol. 1 (1961). (AlsoCivil Engineering Studies, StructuralResearch Series No. 230, Departmentof Civil Engineering, University ofIllinois, Urbana, Illinois, October 1961,167 pp.; Untrauer, R. E., Universityof Illinois PhD Thesis, 1961).
10. dePaiva, H. A. R., and Siess, C. P.,"Strength and Behavior in Shear ofDeep Reinforced Concrete Beams Un-der Static and Dynamic Loading," AirForce Special Weapons Center Report
AFSWC-TR-61-47, Vol. 2 (1961). (AlsoCivil Engineering Studies, StructuralResearch Series No. 231, Departmentof Civil Engineering, University ofIllinois, Urbana, Illinois, October 1961,252 pp.; dePaiva, H. A. R., Universityof Illinois PhD Thesis, 1961).
11. Dill, A. F., and Siess, C. P., "Behaviorof Simple and Restrained Deep Rein-forced Concrete Beams Under StaticLoading," Research and TechnologyDivision, Air Force Weapons Labora-tory, RTD TDR-63-3092, March 1964,250 pp. (Also Dill, A. F., Universityof Illinois PhD Thesis, 1963).
12. Unpublished data on tests of shortcantilever beams supplied by the Uni-versity of Texas.
13. Moody, K. G., Viest, I. M., Elstner,R. C., and Hognestad, E., "ShearStrength of Reinforced ConcreteBeams, Part 1—Tests of SimpleBeams," Journal of the American Con-crete Institute, Proceedings, Vol. 51,December 1954, pp. 317-332; Rein-forced Concrete Research Council Bul-letin No. 6.
Presented at the Tenth Annual Convention of the Pre-stressed Concrete Institute, Washington, D.C., September 1964
February 1965 47
cc Table Al-Exploratory Test Results
No. Type h, h', a, d, p, a/d fc, fy, fsutest, vytest, vycalc, vntest, vucalc, vytest utest v utest cf'in. in. in. in. % psi ksi ksi psi psi psi psi vycalc vncalc (vltest)WC/
WC = welded cross-bar tension reinforcementBI = inclined bar formed by bending tension reinforcementWI = inclined bar welded to WC tension reinforcement
ro CL-% additional column load, % indicates ratio of column load to corbel loadCR = No. following CR indicates reinforcing bars in 8xl2" column
1 w = 5 in. (in all other cases w = 3 in.)z ? column failed
= 12 in. and b = 8 in. for all specimens unless otherwise noted* not measured or inconclusive test data
=6in.? fsu calculated smaller than fy
ut
5,w
01CA
Table A4-Test Results for Corbels with Stirrups
h, h', a, d, StPr^r^ Stirrups a fc' fy^vytest utest
No. and ins. d Spacing Stirrup Tension f test, v test,y , v calc, v testu , v caic, v calcups Reinf. c/c ksi Reinf. su y u vycalc uin. in. in. in. ps i in. psi psi psi psi psi psi