INTERFACE SHEAR CAPACITY OF CONCRETE SURfACES USED IN . - NIST

C HAP T E R2LITERATURE REVIEW2.1 Repair and Strengthening of Reinforced Concrete StructuresA number of reports have been presented by Sugano, \"lyllie,and others [14,16,19,20] discussing the repair and strengthening ofreinforced concrete structures for seismic resistance.Fieldobservations of buildings that had been repaired a,1d strengthened andthen subjected to later earthquakes have been reported by Wyllie andDean [20]. Experimental programs have also been developed to studythe effectiveness of various types of infilling and bracing techniquesand beam column connections [16,17]. The state of the art in seismicstrengthening of existing reinforced concrete buildings have beenreviewed by Sugano [14] and developed into a guideline for theretrofitting of existing structures for use in Japan [15].Different methods are currently being utilized for repairingand strengthening buildings for improved seismic resistance. Whilesome of the techniques used are similar, the objectives of the designengineers using these techniques in repairing and in strengthening astructure are different. By repairing a damaged structure an attemptis made to return the structure to no less than its original strengthand seismic resistance. When a building is strengthened, however, theobjective is to improve its seismic resistance by increasing strength,stiffness, ductility, or all three.There are three main reasons why the repair or strengtheningof a building would be undertaken.1.A building damaged in an earthquake may be repaired torestore its serviceability and possibly strengthened toimprove its performance in future earthquakes.2.An existing building may be strengthened to meet currentseismic provisions if its usage or occupancy changes.3.An owner's concern for the safety and protection of hisinvestment might entice him to voluntarily strengthen abuilding.It should be noted that while the strengthening of a buildingcan improve its performance it is not a guarantee of a damage freebuilding.causesWhen undertaking the repair of a damaged structure, theand extent of the damage must be thoroughly assessed.3

4Determination of the structure's performance and the type of failureobserved, whether it be shear, flexural, bar anchorage or any othertype of failure, is essential to the selection of an adequate repairor strengthening scheme.During the design of a strengthening scheme for a building,the engineer must consider both structural integrity and the user'sneeds. The foremost consideration by the design engineer would be thepublic's perception of and confidence in the strengthened building.The strengthening system's functional requirements must be met whilekeeping it aesthetically pleasing and economically feasible.The strengthening system selected will provide increasedstrength and may also be used to increase the stiffness to reducedamage to nonstructural elements of the building.The strengtheningtechnique must be examined to determine whether stiffness or strengthdiscontinuities have been produced which could cause a failure inanother element of the existing structure.Epoxy injection of exi sting cracks and partial or completereplacement of a damaged member are commonly used techniques for therepair of a structure. Some of the techniques used to strengthenbuildings include new cast-in-place or shotcreted infilled walls, theconversion of exi sting nonductile frames to an acceptable shear wallsystem, and the use of structural steel bracing. One such techniqueutil izing wing walls to strengthen the columns of a reinforcedconcrete frame is shown in Fig. 2.1.A successful repair and strengthening scheme, as reported byWY11 i e and Dean [2 0 ], w as use don 0 neb u i 1 din g who s e h 0 11 ow b 1 0 c kwalls and reinforced concrete colum;,s were heavily damaged in anearthquake. The repair scheme cost about one-third of the cost of theoriginal structure and included epoxy injection of cracks in damagedcolumns and construction of new reinforced concrete shear walls inelevator and stair wells. The repairs increased the stiffness andstrength of the building and prevented major damage, and subsequentrepair costs, from recurring when the building was subjected to asecond earthquake a few years later.Tests on di fferent strengthening schemes wree conducted bySugano and Fujimara [15,16] on one-story reinforced concrete frames.It was found that frames where the colu mns were strengthened by wingwalls, similar to the arrangement shown in Fig. 2.2, provided asignificant increase in lateral load strength.The effect of theinterface shear capacity at the wing wall connection, however, was notstudied.

5IIExisting BeamcDE::Ja&0u0".C. - Addi tion.Wi ng lola 11 sVIXu.JDIIFig. 2.1Frame with columns strengthenedfrom Sugano [15]by wing walls,LLDDowelExistingCol ullinAdditional Wing WallCast-in-Place or)( Precast Concrete'Anchor Reinforcement(Connected by Screws)Fig. 2.2Wing walls with dowelconnections, from Sugano[1

62.2Shear Transfer MechanismsFor successful strengthening of reinforced concretestructures an understanding of the shear transfer mechanism across theinterface between old and newly cast concrete is needed. A survey ofthe studies undertaken by many researchers to evaluate the effect ofaggregate interlock, friction, and dowel action on the shear transfermechanisms of a concrete interface is reviewed in the followingsection.2.2.1Previous Research.Research done by Hanson [6] in1960 was one of the first comprehensive studies done using push-offtype test specimens to evaluate shear stress-slip behavior of concreteinterfaces with different surface preparations. The effect of theinterface reinforcement and surface bonding was studied. These testsindicated maximum shear stresses would be increased when the interfacesurface preparation was varied from smooth and bonded to rough andbonded. Figure 2.3 shows the stress-slip curves reported in from thisinvestigation.The ACI Building Code requirements for reinforcement ofconcrete interfaces is based on a shear friction hypothesis presentedby Birkeland and Birkeland in 1966. A shear load when applied acrossan interface will produce both parallel and perpendiculardisplacements at the shear plane as shown in Fig. 2.4.Theperpendicular displacement produced when roughened surfaces slideacross one another will result in axial tensile stresses in thereinforcement crossing the interface. These stresses will inducevertical compressi ve stresses on the concrete interface which willprovide a frictional force that resists sliding.The ultimate shearcapacity will be developed when the yield strength of the interfacereinforcement is reached. The ACI Building Code (318-83) thereforegives the ultimate shear force across an interface as:whereVn nominal shear strength, lbsAvf area of shear friction reinforcement, in. 2fy specified yield strength of reinforcement, psit.L coefficient of friction along the interfaceThe following values of the coefficient of friction are gi ven in ACI318-83 for normal weight concrete:

00o100200Fig. 2.3 ! eno(l). :if)(l)o'-. 300if)(l)If)If)c 400a.500600 I-BOND ANDROUGHNESS.010Slip in Inches015020Shear-slip curves for various interface surface conditions, from Hanson [6].005JSMOOTHBONDEDROUGH UNBONDEDBOND AND ROUGHNESSWITH KEY/IStirrup Effect Subtracted For All CurvesFig. 2.4rCrack width1Sheardisplacementti. ,General orientation of crack Displacement along a cracked shearplane, from Park and Paulay [12]w4r-II 'Ik -

8monolithic concrete1.4intentionally roughened surfaces1.0untreated surfaces0.6Tests conducted by Mast [9] showed that this shear frictiontheory was based on static ulti mate loads after cracking and is onlyvalid when failure occurs by yielding of the reinforcement andtherefore full development lengths should be provided on both sides ofthe interface. It was also shown that tensile forces across theinterface affect the shear force that can be developed and that shearfriction is not applicable to connections subjected to fatigue orwhere slip is highly critical.Mattock et ale [10] have conducted several investigationsinto the shear strength of cracked and uncracked concrete interfaces.Some of the variables studied to establish their effect on theultimate shear strength of interfaces included: (1) the concretecompressive strength, (2) yield strength of the reinforcement, (3)different percentages and arrangements of interface reinforcement, (4)existence of addi tional stresses, such as moments, along and acrossthe interface, (5) construction joints, (6) aggregate type, and (7)the effect of cyclic loadin These tests demonstrated a di stinct difference in behaviorbetween ini ti ally cracked and uncracked speci mens.In the uncrackedspecimens, a concrete strut transferred stresses between the small,inclined cracks that developed near the shear plane at high shearstresses and relatively small displacements along the interface. Forinitially cracked specimens relatively large displacements occurredalong the interface at the maximum applied shear loads.It was foundthat specimens subjected to cyclic loading averaged 83% of the shearstrength measured for monotonically loaded specimens. Figure 2.5shows t.he effect of the a mount of reinforcement crossing the interface(Pfy) on the shear strength as establi shed by one of these studies.The shear transfer mechanisms acroSS a horizontalconstruction joint were studied by Paulay, Park and Phillips [13].Surface preparation and interface reiforcement percentage effects weretested by the application of monotonic and cyclic shear stresses alongthe construction joint. The mechanism of dowel action and the loadslip relationships for the dowel action of di fferent sized dowels isshown in Fi gs. 2.6 and 2.7, respecti vely. The load-s lip curves ofconcrete shear transfer for various surface preparations are shown inFi g. 2.8. These tests showed that the maxi mum shear stress increasedwith an increase in surface roughness and the interface reinforcementpercentag For low steel percentages, in the range of 0.31%, failure

9(100)14001200(80)Series 1Initially Uncrocked 1000(GO)0' -800"-Series 2Initially Crocked on Sheer PlaneVu ,f """ 4000 psi (280kgf/cmt )50 ksi (35 kgf/mmlf)fy -I ,o8001(GO)100011200(sKJ)psi (kgf lern')Fig. 2.5Variation of shear strength with reinforcementparameter, P fy, from Mattock et al. [10]

10IIM Vd -i-"!.f.---.(J Vd ITFLEXURESHEARVd :: .2!:!. ::4 d b Asfy.(3rt---The mechanism of dowel actionacross a shear interface, fromPark and Paulay [12].I. .LISQ10/00jQ0;. i V . r o I0.07I(1.50.02Fig. 2.7I.!JO.OJ1 .2Jfp. :0-04IJ.41500-II-1 Oar '4dl 0.05Slip0.06 2.4jQ.iOOO . .QIOarP" : 0.oo.J1 I I jQ.osa. 9I .,,/11 no oantlP.,.'";!2000Joi:nt I'r.QGrotlOnV'.4 .4-- I o.rTro., "'tI{y -.--o-41)01 !tII200-'" ;}O :: Asfy cos 0'.vT.:!'" JOO. . :: Asfy.(Fig. 2.6KINKINGI 1.2III0.07sen :i2.5 Imml2.00.04. pD.09Load-slip relationship for dowelaction, from Paulay et al. [13]D.a:l I in.J

11-REINFORCE"" Nr-"3/lQrsIf:: 4GOOpSlp. f : 0.006926lk9f/ mle.,700 ,* Joonf (If:: : J.Jt50 pSI)(If:: ,;36 kgf/cm')I-sol"k 500"'::::-- -':""' J-4(J"" Esao - i JS R0. ."". !: 400""" FII - ;50d.Sl(ln ltrHl9'It t.v.';00. - - -ir-----t-----if--- -- , -- 4. ---I;O JOG -- -----4 ----4 ---- ---- ---- ---- I---- ---- ---- {ls10051.00a0.01Fig. 2.8o.oJ1.52.0a.07a.a9Load-slip curves of concreteshear transfer for various surface preparations, from Paulayet al. [13]il.JO lin)

12consisted of the yieldings of the interface reinforcement. For highersteel percentages failure consisted of crushing of the concrete at theshear plane.The failure planes of specimens containing more than ACIBuilding Code (318-71) minimum reinforcement and a rough bondedsurface did not occur along the construction joint. Paulay et alesuggested that this indicates that the strength capacity would not begoverned by the surface condition along a horizontal constructionjoint.It was found, however, that ACI 318-71 conservativelypredicted the strength of these specimens.White and Gergeley [18] investigated dowel action andinterface shear transfer under cyclic loading. It was found that theload-slip behavior for dowel action alone is similar to that forinterface shear transfer except the residual slip after unloading isless for dowel action. Dowel action during the first cycle of shearloading differed sharply from that of subsequent cycles and resultedin crushing of the concrete around the bars, destroying the bond andthereby changing the restraint stiffness of the interface.Theapplication of axial tensile forces on the interface reinforcementalso resulted in large increases in slip at the interface for a givenapplied load.Liu and Holland [7] studied the influence of dowel spacing onthe dead load carrying capacity of repaired concrete. A series ofdowel pullout and shear transfer tests were conducted to discover anoptimum dowel spacing as a function of concrete thickness.Tests conducted by Luke, et ale [21] at The University ofTexas showed that dowel pullout strength per inch of embedmentincreased an average of 125% when embedment length of the dowels wasincreased from 3 to 6 in.When designing the interface connection of a strengtheningscheme for earthquake resistance, the effect of cyclic loading atlarge displacement levels and the residual sh. ear transfer capacityafter the initial peak strength should be known. To date, however, noinformation has been found on the post ultimate strengths and theresidual load-slip behavior of reinforced concrete interfaces.2.2.2 Summary. The research conducted on the shear transfermechanisms along a concrete interface as presented above have led tothe following conclusions:1.The principal mechanisms of shear transfer are: bond of theconcrete interfaces, dowel action of the reinforcement, andinterface shear friction along rough concrete surfaces.

132.Shear forces are initially transferred through ne uncrackedinterface by bond.Once a crack forms along or near theinterface the shear forces are transferred by the combinedaction of aggregate interlock, friction, and dowel action.3.The shear friction theory used in the ACI Building Code is alower bound to the experimental data available from sheartransfer tests.

C HAP T E R3EXPERIMENTAL PROGRAM3.1Test SpecimensThirty-three push-off type specimens were tested toinvestigate the interface shear transfer capacity between new concretecast against an existing reinforced member. Test specimen dimensionsare shown in Fig. 3.1. Figure 3.2 shows a test specimen prior totesting. The variables studied include:1.amount of interface reinforcement;2.embedment depth of interface reinforcement;3.compressive concrete strength of existing member and newmaterial;4.concrete interface surface preparation;5.reinforcement detailing in both existing and new elements;6.casting procedures;7.concrete interface area.andA detailed description of each specimen is given in Table 3.1 andillustrated in Fig. 3.3. Specimens 1A through 6A were identical. Inall of the other specimens some aspect of the specimen was varied.3.1.1

reinforcement, structural detailing of the concrete elements, and the compressive strength of both existing and-newly cast concrete elements on the shear capacity of reinforced concrete interfaces. Test results consisted of load-deformation relationships, maximum shear capacities, stress-slip envelopes, -and an evaluation of the