Sunteți pe pagina 1din 96
MASTER OF ENGINEERING REPORT ERC UNIVERSITY ‘ORNIA art es Center HORIZONTAL CONSTRUCTION JOINTS “YIN CAST-IN-SITU CONCRETE, by M.HLPHILLIPS FEBRUARY 1972 | 4, DEPARTMENT OF CIVIL ENGINEERING UNIVERSITY OF CANTERBURY. CHRISTCHURCH, NEW ZEALAND HORLZOWTAL CONSTAUCTIOW JOINTS | Ih CASTHI ITU CONCRETE. A report submitted in partial fulfilment of the requirements fer the degree of faster of Engineering at the University of Canterbury, Christchurch, New Zealand. by . He PHILLIPS FEBRUARY 1972. ABSTRACT In this project, the strength and behaviour of horizontal concrete construction Joints in pure shear were studied. Aspects uhich influence the strength of the joint were investigated by testing thirty "push off" type specimens. The main variables considered were: Surface preparation of the joint 2. Dowel action of the reinforcing 5. Reinforcing content 4. Cyclic loading The tests indicate that construction joints subjected to shear need not be the governing factor in determining the size of the members in which they occur. However they do require special design consideration. Tt was found that the shear friction clauses in the ACI 316-71 code provide a satisfactory method for the design of construction Joints, However in shear wall structures the code does not explicitly Tequire these relevant clauses to be applied. ACKNOULEDGEMENTS Tam indabted to my supervisors, Professor R. Park and Or. T. Paulay. My special thanks go to Or. T. Paulay for his valuable guidence, optimism and encouragement throughout this project. 1 am grateful for the facilities made available by Professor H.3. Hopkins as head of the department. My thanks are extended to: ir. K,L. Marrion, Senior Technician, for constructing the test rig and concrete mould, and fir. $.R. Robinson, Technician, for making the test specimens. The financial assistance given by Morrison, Cooper and Partners is very much appreciated. To my wife, Bov.,for typing this script and for her love and encouragement given throughout. CHAPTER 1. INTRODUCTION AND SCOPE OF PROBLEM 1,1 Introduction 1.2. Scope of this Project CHAPTER 2 REVIEW OF LITERATURE 2e1 General 2.2 Bond 2.3 Dowel. Action 2s4 Surface Roughness 245. Concrete Compression Strength 246 Agoregate Size and shape 247 Aggregate Interlock of Shear Friction 2.8 Reinforcing Contant 2.8 The Complete Joint 2.10 Load Conditions 2e11 Length of Joint CHAPTER 3 TESTING AND INSTRUMENTATION 3.1 Introduction 3.2 Specimen Details 3.3. Surface Preparations 3.4 Instrumentation 3.5. Test Procedure 10 12 16 16 as 2a 23 23 25 28 28 32 35 38 CHAPTER 4 TEST RESULTS 4.2 Casting 4.2. Dowel Action 4.3. Bond 4.4 Joint Surtace Preparation 4.5 Concrate strength 4.6 Reinforcing Percentage 4.7. Slip 4.8 Cyclic Loading Tests CHAPTER 5 AN EXAMINATION OF THE DESIGN OF CONSTRUCTION JOINTS 5.1 Limiting Shear Stress Values 8.2 Reinforcing Content CHAPTER 6 CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH 6.1 Conclusions 6.2 Suggested Future Research APPENDIX A MISCELLANEOUS GRAPHS BIBLIOGRAPHY 39 40 46 50 54 55 56 59 74 74 9 82 83 Be fareermnicn As cross-sectional area of reinforcement across a joint, in square inches 9r0ss concrete cross-section araa af a joint, in square inches natt exial load per unit area, in pounds per square inch compression cylinder strength of concrete, in pounds per square inch yield strength of reinforcing steel, in pounds per square inch moment on joint, in pound inches reinforcing ateel ratio (or content) horizontal reinforcing steel ratio in walls vertical reinforcing steel ratio in valle applied shear on joint, in pounda ultimate applied shear on Joint, in pounds shear stress on joint, in pounds per square inch ultimate shear stress on joint, in pounds per square inch reduction factor friction coefficient for a surface CHAPTER ONE INTRODUCTION AND SCOPE oF PROJECT 2.1 INTRODUCTION Although there is a lerge anount of Literature on concrete construction Joints, much of it is based on opinion rather than on experimental evidence. Tho object of this project was to review the Present state of knowledge and to continue the investigation by carrying out tests on various joints. The examples of the joints shoun in Fig. 1.2.1 after they have been subjected to an earthquake, denonstrate the need for sone research into the problem of construction joints. 142 SCOPE OF THIS PROJECT This project was orientated to investigate horizontal conatruction Joints subjected ta both monotonic and cyclic shearing forces. Horizontal Joints were considered as these are generally wosker than vertical Joints. They aro important in earthquake zones, particularly where they occur in shear wall structures, Verticel Joints occurring in beans are usually not as critical as horizontal ones in walls. They can normally be located at points auay from areas where high shear stresses exist. The usual object in the design of @ construction Joint is to ensure that the surrounding concrete acts as a monolithic part of the structure. The aim of this report vas to see if joints could be designed end formed Of sufficient strength to withstand the loads the codes allow in the design of concrete structures, To commence this investigation, thirty test specimens were cast and tested to study some of the main variables which nay affect the strength of a construction joint. The variables locked at vere: Surface preparation of the joint. Reinforcing percentage passing through the Joint. Dowel action of this reinforcing. Bond. Cyclic loading. The loading on all the specimens was a shearing force across the Joint with no applied manent or axial load, As the tine for this project was limited, an exhaustive search of the effects of all the variables was not possible, The tests carried out for this report end the section on suggested future work should aid future Fesearch to be channelled into other important aspects of construction Joints. CHAPTER Tuo REVIEW OF LITERATURE Much of the present knowledge on construction Joints has been Gained by different researchers looking at specific problems of construction Joints. This review will be based on these rather than on a chronological review of all known work undertaken. 2.1 GENERAL Grook” has revieued the literature on construction Joints and suggests verious itens uhich future researchers could look at. In the past very little work has been carried out directly on the problem Of construction joints in cast-inesitu concrete, fost of the studies were concerned with precast to casteinasitu joints. This work! together with studies on shear transfer by aggregate interlock”*3#2112 gake up most of the literature available on construction joints. Anderson”, Hanson’®, and tat” investigated construction joints Using pushoff tests. A sketch of their different test specimens is shoun in Fig. 2.2.1. The results from these tests are referred to throughou this chapter. 2.2 pono The earliest knoun research on construction joints was that by Davey® in 1930 who looked at the aspect of bond between tuo layers of concrete As these results were for hand tampering methods they will not be produced Cast~in-place Precast ANDERSON Cast~in-place Precast mast "push OFF" Test Specimens 2 ancora heres More recent researchers examined tho espect of bond between tue leyers of concrete under various load conditions. The results of Bate’, + Watere®, and the United states Engineer Wateruays Experiment Station’, and the 1oading conditions used are present in tables 2.2.1, 242.2, and 242.3 respectively, The ganeral agreement for bond is that Setting the face oF the joint results in lower bond strength. Applying mortar or cement grout does not affect the strength of the Joint greatly. Ss Roughaning the surface gives the greatest bond. EH. Waters? Proposed two hypotheses to explain why the bond strength of Joints was the highest for a dry joint surface. These were: (4) The absorption of water by the old dry concrete fron the fresh concrete decreases the uater/coment ratio of the latter and consequently increases its strength, (44) The Finer particles of cenent are draun into the interstices of the eld concrete as it absorbs water from the fresh concrete. eters carried out sone tests on tiles” which showed that the second hypothesis is unlikely, Ho sign of tha coment peste entering into the pores of the tiles could be shoun in the tests. The first hypothesis aithough not proved appesre to be the most likely cause for the greater strength, There are different opinions on the treatment of the concrete to obtain a rough surface, Waters points out that in his experinents of seabbled” joints, many of the Faflures did not occur along the joint but in the first section cast close to the joint, It was thought that this * Scabble - "to rough dress stone" ~ oxford Dictionary TABLE 2.262 Effect of various treatments upon the vond strength af construction Joints, by gate’, Loading Condition: pure ttoment on the Joint. PREPARATION OF SURFACE OF |—CONDITICW OF SURFACE] AATIO-OF JOINT. JomTeD To un-so1uTED BEANS. none ho treatment 0.60 fone surface dry O51 none surface ust a2 exposed aggregate fo treatment 0.93 vertical | exposed aggregate surface ary 0.98 eae exposed aggregate surface wet 0.89 exposed aggregate Bin. layer of 0.76 | cement mortar applied imreciately prior to new conerete. scabbled no treatment 0.72 scabbled surface dry 0.80 scabbled surface wet 0.56 horizontal none no trostmant 0.79 Joints none” surface dry o.74 none” surface wet 0.53 Top surface, including laitence layer, left as cast. No Treatment: The neu concrete was c 24 hours after the first half. Surface dry: The first half of each bean uas removed fron the moule 24 pours, cured in water for 7 days and then dried et room femperature for 7 days. The eecond half vas then cast against the dry surface, Surfaco wot: 24 hours after casting, the exposed surface was wetted with & brush, and the new concrete was then placed. Sxpesed aggrecate: 3 hours after casting Lightly brushed uitn vater to expose the coarse aggregate, Tensile Bond Strength of Horizont Load Condition: Tension on the Joints instruction Joints, T. Jaters®, PREPARATION OF SURFACE OF JOINT. | CONDITION GF RFALE. RATIO OF STRENGTH OF JoINTED To ungoTRTEO SPECTII one _ Ce scabbled treatment scabbled mortar applied | a,55 wire brushed wet sand blasted to remove 0.53 laitance wet sand blasted to expose 3" of augregate 0.76 leitance removed by water jetting 0,54 surface brushed 3 hours after ij placing 0.49 no Limestone chippings inserted by treatmant hand into the fresh concrete 0.56 dry sand throun on the fresh concrete, washad off before casting éecond stage 0.42 Fetarder applied after 1 day, wire brushed after 7 days 0.42 | retarder applied to fresh concrete} wire brushed after 1 day. 0.46 none dried out 0.78 scabbled Oried out 0.72 wire brushed | dried out 0.76 dried out in sicator TABLE 24263 Sond Strenath of Joints Subjected to Different Trestnents, U.S. Army Engineer Waterways Experiment Station’, Lover Ralative strongth of 3oint ea percentage Surface | Loading of Strength of Concrete Condit ons No Wortar | Thiek wortar | Tnin master Breanod | Flowed | Brooned | Fiscad fewre | 74 7 ary | hoa a fon | ola average | 72.8 | 708 flexure 2 | 0 | 0 | 06 ry 8 | sheer | 9 | 0 | on young overage ry 735 | 90 | ous fiewral | 63 % wat | shear er ss oad average | 78 76.8 Alexiral x | oe | as | on wot | aneer | 66 | 6s ae young average el 64 82 xi) Not an airewater jet, All the joints were left rough and treated with was due to damage to the concrete caused by the scabbling operation, even though more care was taken than in normal practice, However testa by Hanson? indicate that the shear carrying capacity of joint io not sensitive to the method of preparing the surface roughness. Waters® has found that the age of the old concrete at the tine of Placing the second layer does not affect the bond strength across the Joint. 243° DOWEL ACTION Only tuo known researchers have looked at the douel acticn of the Teinforcing crossing the joint. These are Mattock? and Henson!®, Dowel actien is the resistance supplied by the bending and direct bearing of the reinforcement which runs through the joints. ffattock eliminated this effect by wrapping the reinforcing at the joint in rubber sleeves. See Fig. 2.3.1, This method of eliminating devel action is not the best as it reduces the atiffness of the reinforcement by providing jeinforcing Gar Reinforcing Gar Rubber No Bond over, Joint ‘Smooth Joint this Length No Bond matTock HANSON Figs 2.3.2 Methods for Measuring Dowel Action @ length of reinforcenent adjacent to the Joint surface uhich is not Bonded te the concrate. Mattock found that dowel action wae insignificant in initially uncracked specimens but significant in initially cracked specimens, Hanson”? Locked at dowel action by testing specimens vith Joints which were smooth and unbonded. See Fig. 2.3.1. With this method, dovel action can be measured for any rainforcenent pattern. Hanson's results are produced in Fig. 2.3.2. Also in this figure are the rasults obtained from a series of tests produced in this report. 1/2" Dia, Bar 5 fonce 4 - PER cat (kIPS) 3 1/2" Dia. Ser Phillips 3/8" Dia. 1/4" dia. 0 OL +02 +03 SLIP (INCHES) Figs 26562. Load Slip Curves for Dowel Action Part of the reason for the discrepency in the curves is that in peeevie, ceuber Sher joint woe (mot oul jected calrasl sheer tet a of fe sae @monent acting on the joint dus to the eccentricity of the applied doad. See Fig. 2.3.3. Another factor which could account for some of the discrepency 4s the conerete strength, Hanson used a concrete strength of approxinatel, 5000 psi for the lover section end 3500 psi for the top section. The specimens tested for this report would have @ much lover local concrete Strength. The greatest height of the concrete when placed affects adversely the strength of the topnost layer. This will be discussed further in thi report. 204 SURFACE ROUGHNESS Hanson? carried out a series of pusheoff shear tests for which he compared different surface preperations. He concluded that for the transfer of shear the preparation of the joint was not of vital importance. He Givides surface preparations into three categories (4) smooth 1.0, trovelled (44) Rough (413) Keyed See Fig. 2.4.1, The Tough surfaces used in these tests varied from 1/16 inch to 3/8 inch above and below the average level. No difference in the shear carrying capacities of thess surfaces uas found. The difference betusen She smooth and the rough surfaces is measured by the friction coefficient For the surfaces. This will be discussed later under the heading Aggragate Interlock. For the majority of the 8 ' pe, teats bw az, + - Other values of L ei were 6" and 24", Note: Use L = 12" Test Specimen Hanson’ Pe fp Forces Acting on Joint Applied load p Moment on Joint = Px 2 Tensile force in stirrup = T Compression force in concrete = ¢ For equilibrium T = ¢ Friction coefficient = 41 Shear transferred across Joint, P = Dowel Action + Ae the compressive force is relatively small on the concrete, sssume that it has a Sinch moment arm, therefore equating moments ce Ps 1f f+ 49 assumed to be 0.7 for @ smooth concrete aurtace, Shear transferred across joint, P = Dowel Action + 0,14 p Therefore Dowel Action is only 86% of the applied shear. Fide 23.3 Effect of moment on Hanson's Dowel Action Tests Shear Stress (pst) 600 } 500 L. 4o0 f 300 200 100 / Oowel Action subtracted For All Curves ' 1 1 ond & Roughness Sond & Roughness with Key Unbonded with Key Rough Unbonded Smooth Bonded 0 a +005 +020 +015 +020 Slip (Inches) Figs 2.4.1 Load-Slip Relationship for Various Surfaces by Hanson? Manson faund that shear keys are no more effective than = reugh surface with ond. Initially the monelithic concrete of the keys resists all the load. As the area of the keys is only a portion of tne total area of the Joint, the stress across the keys 1s large. A Grack eventually forme through the keys. hen this occurs the keys eg es ee dl the ettat ion tte etter) ec leet cf ctiei eer @ Fough surface, as shoun in Fig. 244.2, crews vf sheer Keys Failure Plane Figs 26462 Failure of @ keyed Joint 2.5 CONCRETE COMPRESSION STRENGTH CONCRETE COMPRESSION STRENGTH Mattock” considered the effect of the concrete compression strength on shear transfer by aggregate interlock. He postulated a relationship for the ultimate shear strength transferred across a cracked joint as shown in F49.2.5.1. The concrete strength only controle when the Feinforcing content is highs The ACI 318-71 coda linits the ultinate shear stress carried by 9 wall section to 10/T%E of which 2/FTE is sssimed £0 bo carried by the concrete. Fron the equations in the code, the following ratio can be formulated for the required steel retic for shear. vy = 207% + pty £10 Fre pfy <8 / Fie whence 400 < pry =566 poi where 2500 = fc =so0opsi From Wattock's proposed curve St can be seen thet the concrete strength sili not normally control the strength of the joint if the code limite for the maximum shear stress on the section are observed, Hanson?? also examined the effact of the concrete compressive strength on construction joints, His tests show that the concrete strength has @ definite affact on the initial peak shearing stress for @ Joint. It would appear that the weaker of the tuo concrete strengths #8 the controlling Factor. In a nconel castein-oitu joint in a vali, ‘hie would be in the lower section poured due to 1aitance and other effects which will be discussed later, 266 AGGREGATE SIZE AND SHAPE AGGREGATE SIZE AND sHapE The effect of aggregate size and shape has been investigated 2500 | Shear Strength Not Governed by Concrte Strength vLTiMaTe 1000 SHEAR sTHESs (PSI) fre =2500 Shear Strength Governed by Concrete Strength 500 pfy (Pst) Figs 2.541 Possible Influence of Concrete Strength on the Ultimate Shear Strength of a Crack Joint By tattock® for the transfer of shear across joints in concrete pavements by tiovlen!*, Colley"? and Humphrey. These teots were for a Joint that hed been Formed by cracking a monolithic concrete slab, They found in thie situation that the Joint efficiency was higher as a, the aggregate size increased b, the particle angularity increased The results ere shoun in Fig. 2.641 Losber?? in his teste on aggregata interlock found.that, there was Little difference betwean the load displacement responses for specinen: with 3/8 inch round, 3/6 inch round, or 3/4 inch crushed maxinun eize aggregate . The maximum crack width used in Losber's tests was +02 inches. 200 +035" Joint Opening a0 }— 60 a EFFECTIVENESS PERCENT = ag L (One miion Load Cycles) ot alam Opening obi, 0 1 2 3 MAXIMUM AGGREGATE SIZE (INCHES) Fee 26.1 Influence of Aggregate size on Joint Erficioncy It would appear from these tests, only concrete with a maximum size aggregate of over 14 inches ‘ould affect the strength of a joint, This 1s probably due to the nature of the failure plane being able to find @ path around the smaller aggregates, but for the larger ones the Plane is forced to pass through the aggregate increasing the strength of the joint. 2.7 agcRt ATE INTERLOCK OR SHEAR FRICTION Aggregate interlock and shear friction are tue different concepts luhich are used to explain how sheer is transferred across a crack in concrete. The aggregate interlock mechanism relies on the condition that when a crack is formed in concrete the surface is rough and irregular. The majority of the larger particles remain enbedded but protrude across the crack, These particles are then considered as dowels across the crack. A shear applied across the crack will cause a movement slong the crack bringing these particles to bear on the opposite surface. Provided gome rastraint is available to prevent large increases in crack width, the shear can be transferred across the crack by aggregate interlock. A substantial shear displacemant is a prerequisite. The shear friction concept assumes that shear is transferred across @ crack or joint by the friction betueen the tuo surfaces. Consider tuo rough surfaces in contact with each other as is shoun in Fig. 2.7.1 luhen @ shear force and a consequent displacement is applied across the interface the two surfaces tend to separate. By applying 2 clanping force across the interface the shear can be transferred across the crack Sy friction, In concrete the clamping force is provided by the SHEAR FORCE, Vv v NORMAL po T v — —— —_> Fig. 2.7.1 Shear Friction Concept reinforcement across the interface, Therefore the shear force V, is associated with a normal force N, which ie + Ve yn and ultimately Vy = prhoty or y, 1 VSR = pty where V, = ultimate shear force As = total area of reinforcement across interface fy = yield strength of reinforcing Ag = gross area of interface V, = ultimate shear stress on interface P = steel ratio, As/Ag y= feiction coefficient for the surface For different surfaces the value of 4s changes. Various researchers 2,3,12,16 have measured the friction value for different surfaces. Values given by Birkeland’ and the ACI 318-71 code are: act BIRKELAND monolithic concrete aa 7 artificially roughened joints 10 1d for ordinary construction joints O68 = 160 conerete to steel interfaces 0.7 O-8 ~ 100 For convenience in this report the concept of shear friction will be used rather than that of aggregate interlock. It fs applicable to smooth as well as rough joints, the difference being in the value of the friction coafficient, 268 REINFORCING CONTENT Hanson!® found that for each par cent of reinforcement the Joint was capable of carrying 175 psi shear stress. This was additional to the allowance for bond and joint roughness. Birkeland’ has collated the results of other researchers”?910, Fon the ultimate shear strass on joints for various reinforcenent ratios. These have been reproduced in Fig. 2.81 together with the test roculte obtained for this report. The ultimate shear stress design limit in this Glagram has been imposed by the shear capacity of the concrete. This has beon discussed earlier and is shown diagramatically in Fig. 2.5.1. The prediction and design lines on the graph are based on the shea: friction theory using a friction coefficient of 164. For the tests “without bond the ultimate shear values are very close to the prediction Values. For the cases uith bond, the results are nore scattered especially for the lower staal ratios, This is due to the variation in the bond strength. 400 Anderson, cleat hast x Phi eo . 200 a COGS) 010005 eee) STEEL RATIO, peAs/Ag Fine 2.8.1 Test results of Ultimate Shear Stress Reinforcement Contents DESIGN LIRIT 1 x 7 scan : Fon Tt os coh” x" PH vuswure sess: wilenson, Rough Sonded Q Rough Untondted r Rough Beaded Rough Untonded Kaar for Various | | 249 THE COMPLETE 301NT Hanson’® and Anderson? have considered the complete construction Joint. Anderaon only gives the ultimate shear stress of the joint for various reinforcing ratios in his paper. These are also shoun in Fig. 2.8.1, Hanson's tests were orientated to composite structures such a. Prestressed girders with a cast-inesity slab. This type of Joint is subjected to different load conditions to those Joints in sheer wall structures. The shear to be carried by these joints ds usually of smaller magnitude than that occurring in certain shear walls. If the Construction Joint is weaker in @ shear wall structure than that of the monolithic concrete surrounding it, then it might limit the uhole design of the structure, Hanson gives the following data for the ultimate design of Jointe: 1. Rough bonded surface 500 psi shear stress 2. Snooth bondaurtaco 300 pei shear stress 3. For avery per cont of stirrups an additional shear stress on the interface of 175 psi. Comparing these design valuss with the test results in Fig. 2.841 Shows that Hanson's design figures are unsafe for low steel ratios. 2.10 LOAD CONDITIONS The earliest research on bond strength was to test the tensile strength of bond. Later researchers uho have investigated construction Joints have only tested then in shear. In the case of Hanson's tests this is obviously tha critical loading case as they were designed specifically to investigate composite girder action. The leading chosen for the tests in this report was direct shear across the joint, This is the most eritical loading case except for shear plus axial tension, Consider the shear friction concept for the jointe with various loading cases. For the case with moment only the joint does not affect the shear strength, as all the forces are normal to the joint's surface. ‘The case for moment and shear acting on the joint is usually no more critical than the case for shear only. For the equilibrium of forces, the tensile force in the reinforeing bers is balanced by an equal Conpressive force on the concrete. Thus the sane totel internal compressive restraining force normal to the joint is still available for the transfer of the sheer across the Joint. However the loading condition may be critical due to a possible failure by diagonal tension. If the moment causes a crack to open, all the shear must be carried by the remaining uncracked portion of the joint, and failure occurs by Giagonal tension as shoun in Fig. 2.10.1. Uncracked Portion Failure Fig. 2.10.2 Failure by Diagonal Tension A Joint subjected to shear and moment may be stiffer as the moment Clamps the surfaces together at ene end. when the shear 1s applied. only @ rotation due to the moment is required to increse the restraining force on the joint. In the case for shear only, movenent along a roug! Joint produces a component normal to the Joint uhich tenses the reinforcing thus a shear displacenent is required before the reinforcing will restrain the surfaces, This is not the case for shear and monent: If axial compression acts on the joint, this helps in the transfer Of shoar across the Joint as it increases the restraining force nornel to the Joint, The opposite effact is the case for axiel tension Sufficient reinforcement to carry all the tensile force, plus that necessary to transfar the shear across tha Joint is required. ven with this amount of reinforcoment the joint will be more flexible than if subjected to shear only. If the tensile force is larger it will cause the concrete to crack at the joint. For sheer to be transferred across this crack, movement slong the crack is necessary Assuming that the law of super-position holds, the reinforcing Tequirements for any load condition can be calculated if the requirenants for shear alona are known, It must be stressed here that this is an assumption, as no known research has been done on ather loading conditions. 2011 LENGTH OF JOINT Hanson’ has noted that the shortor the length that the shear acts on, the higher the initial peak stresses are for the surfaces, especially for the ones with bond, This effect is shoun in Fig. 2.1141. Hanson Concludes that a high bond stress can only exist over a relatively short Initial Peak Average Shearing Stress (PST) 800 F 700 F 600 + 500 400 300 200 100 Fig. 2.12.2 Oowel Action subtracted where Present © Specimen with stirrup © Specimen without stirrup 6 12 1s 24 Shear Length in Inches Effect of Shear Length Be pees Tere eonlication| of the lesa a ren ee tt Progressive along the joint tovards the free end, This phenomenon has also been observed! in pull-out tests on smooth reinforcing bara The Joints iqvestigates by Hanson had a moment ecting.on thea = 890 Figs 2.3.34 A moment acting on a joint increases the ctiffness of the Joint in shear, This was discussed in section 2.10 of thie Peport. The flexural resiste Of @ monber varies as the depth squared, shite the shear capacity varies linearly with the depth, Thus the constant eccentricity of the applied load in the case of Henson's test. “ould have a greater effect on the stiffness of the shorter joints the aes ory Pe the recace ortine niger ne obtained on the shorter joints. The distritution of the reinforcing could affect the shear strength of a Joint. To obtain a uniform shear stress on the surtace a uniform Gistribution of the reinforcing would be necessary. In Hanson's test. ee ee cece) on) ON8| controlar, the) joint ucctuses vevat ta Jonger Joints the non-uniform shear stress may have effected the results. CHAPTER THREE ING_ AND. STRUMENTATI 3.1 INTRODUCTION Thirty push-off specinens were tested in pure shear uith load, slip and stesl strains being measured. The size of the specimens was kept constant while the reinforcing percentage and surface preparation vere varied. The properties of all the specimens are shown in table 3.1.1 The test rig and the specimen held’ in it are sketched in Figs Sell ‘The loading ran could ba used to apply load to the specimen from either ond of the rig. The specimen was lowered into the rig and a screw jack sat to obtain the correct level of the joint inch above the top of the studs. The moveable stub was then pushed up firmly against the specimen end steel wedges clamping the stub against the specimen uere inserted and tightened by driving them in with a hammer. The stub was finally clanpad to the test rig to stop it being lifted out of the test frames 3.2 SPECIMEN DETAILS The size end reinforcing detaile of the specimens ara shoun in Fig. 5.2.1. The specimens were prepared in batches of six with the concrete being supplied commercially, The casting sequence vas such that Bach batch of concrete completed six bottom and aix top sections. ith this sequence the top saction was cast nine to thirteen days after the bottom section, except for the last betch for which tuenty-five days elapsed, The concreta supplied had a certified strength of 2500 psi at 4 Hydraulic Ram, Fig. Sele1 Sketch of Test Rig 1/a" aan Spiral / Nore: The Specimens wera Inverted for Testing Type & Type € Fig. 5.2.1 Detail of Test Specimen paringn Typy & 1/4" 29 x (Continued overieat) Specimen Properties TABLE 3.141 oe £zto* v6 z03u 69 ug0uviay 6900° S65 zoey £9 te00* orz zoun ag O3A34 Ope Tos so oa 1sayvas 092 Toss: icf a3Hs¥m Oze Tosm 99 XXX = TaMOuL soo TOBxL on Tamu Lee TOBL st €2to* Spo XTOON a u30uv Lay 6900° GLb xTOSY 99 TaMmowL 6900° “urn xTOSL 9 Teo" xTOWL (sAva) (Sava) dOL |wo1so8 ous | 1sva A (sa) win | “Ravn | State | oe | mown | ym | oxy | See | 2 | ' } | | 32 twenty-eight days. A 4 inch slump was specified in an attempt to produce similar conditions in the specimen which would occur at tha top of a concrete pour 6 to 10 feet high. The aggregate for the concrete was naturally occurring river worn greywacke from the Christchurch area, with a maximum size of } inch. ‘The mix proportions of the concrete by weight ware: cement : aggregate : sand = 10: 4+52 : 3-09 water : cement = 1-0 : 1+60 The specimens were cast in a plywood linad timber mould. After casting, the specimens were cured for saven days under damp hassian, then stored in the laboratory with no further curing, For each batch of concrete six No. 12 X 6 inch test cylinders were cast. These uere tested at the same time as the specimens, The age of the concrete at esting is shown for each specimen in table 3.1.1. 3,3 SURFACE PREPARATIONS The Following are the various preparations used on the joint surfaces, with the identification code, in brackets, used to designate the specimens, Izouel (1) The surface was screeded and left for spproximately four hours after uhich it was floated with a steel trovel to form a ‘smooth surface, This is defined as a smooth surface preparation in this report, see Fig, 3.3.2. Retarder (R) The surface was screeded and after one hour a retarder (Fabol) was sprayed on the surface, The next day the surface was washed and scrubbed with a firm brush, see Fig. 3.3.3. Trowellad and scrapsd (T*) The surface was screeded then scraped | | | i Figs 3.5.1 Figs 3.3.2 Figs 3.3.3 Trowelled Surface Waxed Trowelled Surface Surface Treated with Retarder Figs 3.364 Fig. 3.5.6 Washed Surface Scabbled Surface Retarder Surface Treated with Varnish 35 with @ pointed trowel in a criss-cross fashion (XXXX). The indentations were about } inch deep. Washed (W) Approximately four hours after screeding the surface was washed and if required brushed with a firm brush to expose the coarse aggregate. Fig, 3.3.4. Scabbled () The surface was screaded and left. after four days the surface vas chipped using @ hammer and chisel. Fig. 3.3.5. Keyed (K) Tuo timber blocks 4 inches wide and 1}-inches deep were set in the surface for the full width of the joint, The remainder of the surface was only screeded. when removed from the Formuork the blocks were removed to leave grooves which wore filled when casting the top section. No Bond (x) For the trouslled surface with no bond, candle wax vas melted and painted on with a paint brush. (Fig. 3.3.2). For the Tough surface with no bond the retarder type surface was used, the bond being destroyed in this case by three coats of varnish which was sprayed onto the surface. (Fig. 3.3.6). Bend For joints with bond, the surface was suept with a soft brush before casting the top section, In all cases the surface was left dry and no mortar or grout used. 3.4 INSTRUMENTATION ‘Tha load was measured by a 60 ton Philips bridge type load cell connected to a "Budd" amplified bridge. The steel strains of the six bars extending through the joint were measured using 4 inch Demac gauges. Stool studs were welded onto 4 the reinforcing bars and covered with plastic tubing. Before testing the plastic tubing was renoved and pro=drilled locating discs were attached to the studs with sealing wax. These may be seen in Fig. 4.1.2. The slip along the Joint was messured in thrae ways initially, but this was found to be unnecessary and for the cyclic loading tests the slip was only measured uith the dial gouges. The three methods of measuring the slip are shoun in Fig. 3.4.1. Reasonable agrecnent wa obtained fron all three methods The first method used 8 inch Denac gauges to measure the displacement of the specinan above and below the joint relative to the testing frana. This method was ineffective uhen the specimen did net crack along the Joint and also for slips over a tenth of an inch, The second method used 4 inch Oanac gauges in the form of a triengle fas shoun in Fig, 3.4.1, By assuming that one side of the triangle ronained parallel to the joint, the slip along and the apening of the Joint wae calculated. The average value of the slip was taken of the four triangles, tuo in the form of a ster each side of the specinen, A useful check using thie mathod was to compare the computed vertical ‘opening of the joint using the triangles to that measured directly from tha apex of the tuo triangles. ‘The third method for measuring the slip was the most satisfactory as this could be used to measure large slips. Also the gauges usually remained effective uhen the conerete cover over the reinforcing at the Joint started to break off, The dial gauges used had a 2 inch trave! with 001 inch divisions, 8y using tuo gauges each side of the Joint, any Totation about the joint could be compensated for and still offer no visual 37 ‘Demac Point Fixed to Test Frame Slip Measurement Using 8" Demac Gauges NOTE: d = Extension over gauge length Assume this line remains parallel to joint > As a check: (9, + dy ~ 4) / VT = (dg 44, = bd, )//E Slip Measurenent Using 4" denac Gauges Sy Ota gauges tines 0 test specimen S, to test op m@2| £>>tever cistacee Joint a Fixing of orn to test specimen rement Using Dial Gauges obstruction of the Joint. It was considered that using dial gauges was the best mathod as this was virtually unaffected by local straine and cracks in the concrete, 3.5 Test proceoure The specinen was mounted in the test rig and zero readings taken. The load wes then applied in increnents. For each increment the procedure sas to apply the load, then allow one minute to elapse before the readings were taken for that increment, The load was measured aftor the elapse of a minute and also at the end of the readings. The load considered for a particular increment was obtained from the average of the tuo Feadings. The ultimate load of the specimen was teken as the maximum load reached without allowing for any time Lapse. The test was stopped after @ slip of 0+5 inches was recorded or when the slip gauge mounting was dislodged. | | | | } CHAPTER FOUR TEST RESULTS 4.1 casting In most of the lower sections of the specimens cast, surface settlenent cracks formed. Also in a few cases fine cracks were visible on the side of the specinens, just below the top horizontal reinforcing bar. See Fig 4.1.1. No cracks ware visible in specimens uhich had a smooth trowel finish, as the aurfaces of these were trovelled efter settlement nad occurred. After the surface had been prepared by scabbling, Fetarder, washing etc., the cracks were not visible. ‘These settiensnt cracks are typical in wall sections. For walls & to 10 feet high they could be larger than those which occurred in these teste, In casting the top sections any grout leakage showed up as a porus Patch on the concrete surfaces, See Fig. 4.1.2. Teainforcing sare Elevation of Specimen Fine Cracks Plan on Joint Vertical Bars Fide 4el.1 Settlement Cracks in Specimens Figs ded Fig. 4.1.2 Effect of Grout Leakage In practice grout leakage can be @ problen, especially when an unobtrusive joint is required, It can be prevented by careful preparation Of the formwork, A greater fault thet sometimes occurs, though not in these tests, is honeycombad concrete due to poor compaction and segregation of the concrete layer just above the joint. The effect of this can be seen in Fig. 1.141, This can be avoided by the use of suitable compacting devices used by experienced operators, 462 DOWEL ACTION The dowel action for three different reinforcing groups was measured by testing joints with smooth unbonded surfaces. The results are shown in Fig, 442.1 and 4.2.2, Two tests were completed for each group of bars, 6 ~ 1/4" diameter, 6 = 3/8" diameter, and 6 = 1/2" diameter, with reinforcing ratios of p = +0031, +0069 and -0123, Taspectively. The results of both tasts for each group of bars have close correlation even though the concrete strengths differ by 400 psi, uor90y Temog soy senang drs - peoy Zee OTT (s3HoNI) OTIS 2 ost sz* oz" st ort so* 0 0 , 7 T T T T t 0 pug ON UaTR paTTemorL a eoxoy dnaayag --— TOTUENGEE ae 7 Norivavased INToC vif ening atts ~ peoy — on 2 dFTS = peor Wes (sar) 0% OT saa0g angaris (isd) Wior oc lost SSauIs Wins ov, os 09, This would indicate that for dowel action the steel properties are more important than the concrete properties, Three possible mechanisms which could be effective in transferring the shear across the joint are: (4) Bending of the bars (4i) Shear resistance of the bars (444) Kinking of the bare. These are shown in Fig. 4.2.4, The total shear on @ joint, using the same steel ratio but different bar sizes, is shoun in Fig. 4,2,3, These curves, based on the results shoun in Fig. 442.1, were drawn assuming that the shear resisted by each bar is proportional to the area of the bar, It illustrates that the wel action is proportional to the area and not to the elastic modulus of the bars, Force in —> v Bar = fA ee ye v A shear, v = 2 Shear, V = Aw, Shear, V= fA. Cos A : C o v8 ’ 3's (4) Bending of the (44) Shear Resistance of (iii) kinking of the Sars the Bare Bers Fig. 422.4 Possible tiechanisns of Oowel Action 45 which would have been the case if the bending of the reinforcing was the major mechanism. Considering the ultimate shear strength of the bars, and assuming a uniformly distributed shear stress of fy//3 for the steel, the following values are obtained: 1/4" diameter 1.42 kipe per bar 3/8" diameter 2-68 kips per bar 1/2" diameter 525 kips per bar The shear values measured in the tests are close to these values at a slip of O2 inches. However for tha smaller bars, 1/4 and 3/8 inch diameter, the load continues to increase at lerger slips. This would indicate a different mechanism, that of kinking. ith a kinking mechanism jocel effects can occur in the reinforcing adjacent to the joint. The effacts ere; a change in slope of the inclined portion of the bar, and local strain hardening in the bar. goth these effects can cause an increase in the load carried. The most likely mechanism for dowel action therefore is direct shear Tesistance of the bers at low slips and kinking of the bars at higher slips. The total stirrup force-slip curves illustrated in Fig. 4.2.2 appear to lag behind the load-slip curves, although they are similar in shape. ‘The 1ag is greater as the bar size decreases. This could be caused by the more efficient anchoring of the smaller bars. It is possible that local yielding in the bars occurred, which would not be detected with the technique used for measuring the steel stress. The steel stress was calculated by measuring the extension over a 4" gauge length and by assuming constant strain over this length, The local effect would thus be averaged over the 4" gauge length, resulting in a lower force being calculated than is actually present, At larger slips the effect of kinking i in the bars would not be detected as it is also a local effect, i | No tests were investigated where the only machanism of transferring the shear across the joint would have bean the bond between the concrete, Tests to measure the bond strength by this method could be useful to compare with the results obtained here. In these tests, the mathod used to measure the effect of bond on the strength of the joint was to consider the tuo series of tests with trowelled joints, one with bond (Fig. 4.3.1), anc the other without bond (Fig. 4.2.1). The effect of bond is expressed by the difference between the load-slip curves of the tuo series of tests. These sre shoun for the three different steel contents in Fig. 4.3.2. for the rough and bonded Joints, the failure plane was not along the joint. See Fig. 4.3.3. Hence the effect of bond could not be measured by comparing the tests on Joints with rough surfaces with and without bond. In these tests the ee eer ee | ar | Figs 4.3.3 Failure Plane for a Rough Sonded Joint Pept 8qUTOC pepuog Y300us soy senang dTTS - peOT (s3H: Too" = 4 este “P70 w¥/t\, i [Sea oor ening dtTs = peol (1sd) e020) dnaztqg-——— : ae ssaais ieee z031 © 7 awe ‘Worivavases INTOT zo. + ; oor zou + i TaN TSS i geto. +d saeg “era wz/t / id 09 Sa bond strength was greater than the strength of the surrounding concrete. The lead-slip curves showing the role of bond would indicate that the bond is more effective with higher reinforcement ratios. Hanson? tested Joints with bond but no reinforcement across the Joint. The peak shear stress for these varied betueen 90 and 150 psi, except for one test uhich nad a peak stress value of 230 pei. The specimens with trouelled joints and bond used for the cyclic loading tests had different ultimate shear stresses to those used for the monotonic case, This is shoun below. | ~S SS ‘SPECIMEN STEEL RATIO ULTIMATE SHEAR STRESS (PSI)} TAOZ +0032 210 | | TAML +0031 320 | TBOL "0069 347 802 +0069 403 | | Tem. *0069 500 | a | re02 sons as LL rem, 0123 395 Fron this comparison it can be seen that there are large differences in the ultimate strengths of the specinens with the sane steel content. This variance in the results does not seen unusual uhen compared with the tests on bond reported in tables 2.241, 2.242 and 2.2.3. In these cases the bond varied considerably for only enall changes in the surface preparation of tha joints. Hanson!” produces a load-slip relationship for a smooth bonded surface, and this corresponds closely to the curve in Fig. 4.3.2 for the steel ratio of p = +0031 (Hanson's curve is shaun in Fig. 2s4s] of this sport), However Hanson does not discuss or show whether this is an erage or a lower bound curves From Fig. 4,3.2 it would appear that the bond strength may depend on the reinforcing percentage. This is unlikely as the bond strength is due to the adhesin of the coment paste to the surface of the lower ection of the concrete, As the failure of the adhesian is very brittle, any movement slong the joint would destroy it. The magnitude of the slips recorded in tests with bonded joints imply that this adherent bond was broken, A much more Likely effect is that a fine crack is Formed at low slips, The surface of this crack would be rough due to sone of e smaller particles adhering to or breaking from the trouelled urface. The joint then behaves similarly to a Joint with a rough surface. However the failure plane is along the trovelled surface which is a straight line. 7 is causes a lerge drop off in load, The drop off ni load is Larger for a smooth bonded Joint than a rough bonded one. Avtar the drop off in load, the load fs still greater for the bonded, somelled Joint than the unbonded trowelled one. This is due to the srger roughness of the trowelled bonded joint. 4.4 JOINT SURFA PREPARATION In all cases a rough surface preparation increased the effectiveness cf the joint. Also for a rough surface the joint effectiveness is not Cependent on bond. Fig. 4.441 shous the effect of different surface sions on the load deflection curves for a constant reinforcing Although each preparation is shown seperately here, the curves could be divinad into four classes: (4) Smooth with no bond (44) Snooth with bend (441) Rough with no bond (iv) Rough with bond Hanson? aiso came to this conclusion and has pracuced load-slip curvas for the various types of joints. The stirrup effect (in this report referred to ae dowel action) was measured by testing @ anooth join without bond, This effect was subtracted from all the other load-slip curves. The lead~siip relationships which Hanson obtained are shoun in Fig. 24.1. Hanson concludes that the load-slip relationships are independent of the reinforcing percantage, The test results completed for this report do not confirm this, They indicate that the load-slip relationships are dependent on both the surface preparetion and the Peinforcing content. For the joints with a rough surface the failure plane was generally not along the joint but about one inch below. Even for the cese with shear keys the failure plane was not through, but along the bottom of the keys. See Fig. 4.4.2 Tha tuo surface preparations used uhich could have affected the position of the failure plane were scabbling and washing. doth methods produced very sound joints. With surface chipping, cracks below the surface may have been produced but none were visible in the tests performed here. For a washed surface, the strength of the mortar matrix around the surface pebbles could have been affected by the jetting action. This may loosen the pebbles or weaken the bond holding them, thus affecting the failure plane of the joint. Hanson's?” test results and those obtained for this report indicate that these two aurface preparations Smooth Joint \ fo ars cri Rough Joint Eig. 4.4.2 Failure Planes For Various Joints 54 do not result in different strength or behaviour. If the joint has @ roughened surface, the failure plane is not along but below the joint, See Fig. 4.5.3, Tha method usad to roughen the surface does not appear to be important. The reasan for the failure plane being below the joint will be discussed in the next section. 4,5 CONCRETE STRENGTH A systematic 4 vestigation into the effect of the strength of concrete was not studied in this report. The concrete strength which is important is that adjacent ta the Joint, This is likely to be different m that obtained from test cylinders, This is due to the compaction of the concrete causing water gain at the surface, Also the larger aggregate particles tend to sink and the smaller ones rise towards the surface. This produces week concrete near the lower surface of the joint. The failure plane of all the joints with rough surface preparations formed in this weak portion of concrete, See Fig. 4.4.2. The layer of concrete cast against the joint is also important. This layer often contains honeycombed concrete dus to poor compaction, segregation and grout leakage, The results of this are shoun in Fig, 1.1.1. This greatly reduces the strength of the joint, The larger aggregate particles from the honeycombed concrete may act like ball bearings, lubricating the movement along the joint. To measure the strength of the concrete uhich affacts the strength of the joint would require sections to be cut out of the joint, The concrete strangth is more important with the higher steel ratios, At lower reinforcing contents the controlling factor is the yielding of the bars, but for the higher steel ratios the concrete strength controls the strength of the joint. This is shoun in Fig. 2.5.1 by curves proposed by Mattock for the limiting values of concrete strengths, The ultimate shear strength allowed by the ACI 318-71 code limits the shear strength of any section to /¥Te uithout shear reinforcement and to 10 /f¥e with shear reinforcement, The concrate used had a certified strength of 2500 pai but this was exceeded in all cases, the average being 3750 psi. If this figure is used, then the ultimate allouable shear stress is 612 psi, All the Joints with a rough surface and a reinforcin ratio of +0123 exceeded this value as did most of the Joints with a stee: content of -0069. The Limiting value of 10 /F¥G, which the ACI 31871 code imposes on the ultimate shear stress for a section, gives a close approximation to the limiting ultimate values for the Joints investigated in this report The limit of O+2 ftc which the shear friction clauses of the ACI 318-7: code imposes on the ultimate shear stress on a joint was also exceeded for all the tests with a steel content of +0123. However the stiffness degradation of the joint at this load intenaity was greet. Seo Figs. 4.8.5.and 448.6, (This is discussed in section 4.8 of this report.) 4.6 REINFORCING PERCENTAGE The type of failure of the joint depended on the reinforcing content. For lightly reinforced joints the failure was cenfined to the reinforcing which yielded while the concrete around the joint renained in a sound condition, for large reinforcing percentages the failure wes due to the concrete breaking up which affected a large area around the joint. 56 In ali cases, except in the specimens designed to test dowel action, the reinforcement across the joint yielded at the peak shear stress. Considering the shear friction hypothesis es discussed in section 2.7 this is as expected as the reinforcing provides the restraining force between the two surfaces. Once the steel yields, cracks open causing failure of the aggregate interlock, Thus the reinforcing content for a particular type of surface preparation determines the ultimate strength of the joint, provided the concrete strength is adequate For the lower steel ratios, bond has e lerger effect on the Joint. and for this reason the strength of the joint is often much higher than expected. for the unbonded cases the results are nore consistant as shoun in the dowel action tests end also the rough surface teste uithou bond. The rough surface tests, plotted in Fig. 2.8,1,confirm that cases without bond closely approximate the shear friction theory, if 2 value of 14 is used for the coefficient of friction, The test uhich fell below the expected value had a steel ratio of +0123 and the ultinate shear for this case was Limited by the conerete strength, The Limit of F¥ would indicate a maximum shear stress in the order of 620 psi which is below the value obtained from this test. Figs. 4.6.1 & 4.5.2 show the load~slip curves obtained for various steel contents 407° SLIP For ell cases the joints in this project were significantly nore flexible than those investigated by Hanson?© (see Fig. 2.4.1). The typical slips at mexinun load recorded here were +02" uhile those of Hanson were +0012", There are tuo possible effects which may cause this In Hanson's tasts there was a moment acting on the joint, and the concrete 58 Load - Slip Curves for Rough gonded Joints 22 Flo. 406 op——r mo ; : r Toe eo (s3noni) 37 i 4! ok LF | fh eo teo0+ = 4 sxep vera w4/t sos soa0y masts ye | b- -—- — —- > ee — / | es BUTWIEZP BIEM sanind asey TY wry e4TNSex 4803 eu TION. 7, HW poupuseyep pa ese Yorye wosy eatnses 9803 ou FHI Googe Sy? ono+ = d saeg *eyp ya/e toy sosoy dnaayyg—__ OY Ady 6900 “ETP w8/e 04 sox05 dnayas— aden ek -— — — —-— —— — — — -- c (sen) Te00- = d s4eg “erg wy/T 304 enang Arts ~ peOr / 33905 — anys Sis s J 4 oo» Soo Paes ks 7 usd) 6900 3 Asdyee: 4 seau1s Ws Copan f Joos os saeg "870 u8/¢ Soy anang drts = peor | TR “Store o Siew Tere AE/e a3 500) OS eee feo" = 09 F _ 3 J 009 afar # ie 4 ove ertar = 4 e209 tesa u2/t 20s enang dFt5 - pe0T~ pepusg pus uonoy ETT | -—__. 4 ova — — } — 008 strengths used were higher. The effect of the moment on the joint increases the strength of the Joint as is shoun in Fig. 2.3.3. The joint also becones more rigid as discussed in section 2410 of this report, The concrete strengths used in the lover sections of Henson's tests were between 4000 to 6000 pei, These are higher than those used in the teats reported here. Also Henson's test specimens uere shallower, being only 12" deap. This could preduce stronger concrete at the joint, because less concrate of inferior quality accumulates below the joint. 468 cyel LOADING TESTS Each specimen was subjected to @ number of loading cycles from both directions. The loading for the cyclic tests with a rough surface preparation can be obtained from Fig. 4.8.1. This loading sequence wa: followed except for test ACI which reached its maximum load capacity before all the sub-ultimate load cyles were conplated. See Fig. 4.8.5. The load chosen for the first few cycles vas 75% of the peak load based on the tests completed for the identical monotonic loading specimens These are shown in Table 4.8.1. Also shoun in this table are the ultimate loads predicted using the shear friction hypothesis assuming a friction Value of 1¢4 for the rough surface, Comparing these loads it can be seen that the loads used for the initial cycles of the test far exceeded ‘the design working load for the joints which would be approximately 75% of the specified ultimate load for the joint. 58 TABLE 4,8,2 LOADS FOR CYCLIC LOADING ON ROUGH JOINTS 15% Peak Shear Stress | Theoretical Ultimate|75% of Theoretical from Monotonic Loading] Shear Stress Ultimate Shear Steel Ratio Tests Predicted by —_|strese Predicted P (pai) Shear Friction |by Shear Friction (psi) (psi) +003, 320 220 165 | +0069 425 416 310 * a +0123 695 798 598 “this is Linited by O42 Pte = 696 pos this ie Limited to 75% of 0+2 fte = 522 psi The load=slip curves for the cyclic loading on the joints with rough surfaces are shoun in Figs 448.2 to 4.8.6, Load intensities during the cycles in the pra-ultimate range were above the normal working load intensities expected on a structure, For this loading intensity some degradation of the stiffness was recorded in all cases. The degradation for the specimens ACM and ACH2 was very large, but the intensity of the loading was greater than’ that normally allowed y on a section in shear, i.e, greater than 10/f'c. If the loading had been only up to the normally expected working load, the degradation would have been much less in all cases. From the load-slip curves it can be seen that the ultimate loads predicted by the shear friction method are sustained well into the inelastic range for the specimens with steel ratios of +0031 and -0069. For the specimens with a steel ratio of -0123 the drop off in the load THEORETICAL WORKING LOAD ve for Cyclic Test RCM2 Fige 460.6 Load = Slip cur as much quicker, This is dua to the deterioration of the concrete e gives a close approximation ta tha maximum ultimate The limit of 10y/ load developed in these tests. Fig. 448.10 shows the crack pattern for a typical cyclic loading test. In the first cycle soma very small shear cracks form. A cycle in the opposite direction caus further shear cracks running in the opposite direction, with further cycles the cracks mest and form a crack which runs continuously across the joint. In cycle 11 the load vas increased until the peak load was reached, Increment 69 in Fig. 4.8.10 shous the state of the Joint after the peak load has beon passed. The cracking is more intense and the crack widths are much wider. with further cycles the cover concrete deteriorates and the failure plane becomes much more noticeable (cycle 14). Cycle 16 shous the joint a the ond of the test with and without the cover concrete, The displacement in cyele 16 is half an inch elong the Jeint. If this joint is conpare: to those in Fig, 1.1.1 it can be seen that the types of failures are different, The joint here has failed as 9 whole, affecting the area surrounding the joint, The failure of the Joints in Fig. 1.2.1 have a failure plane along the joint and not in the surrounding concrete. The load-siip curves for the cyclic leading tests on joints wit trovelied surfaces aro shown in Fige. 4.8.7 to 4.8.9, These curves show that after the peak load hae been reached, the Joints becone very flexible, The drop off in load is also very larga, much greater than for the co: ssponding rough Jointed specimens. The failure plane is along the smooth surface of these trowelled joints. See Fig. 4.8.11. This is similar to the failure plane of the Joints SLI (INCHES) SUL crit Fin. 4.8.10 Typical Propagation for Cyclic Loading Tests (Continued Overiear) Ti ci ar Crh Crane Coe JOINT ed aa Fig. 4.8.11 Failure Plane for a Smooth Joint shoun in Fig. 1.1.1, This type of Joint is obviously not satisfactory for use where large cyclic loading is likely. CHAPTER FIVE AN EXAMINATION OF THE DESIGN OF CONSTRUCTION JOINTS The ACI 318~71 code contains a section on design using the shear Action concept. The conditions in this section of the code will bo compared with the other sections on shear design. Sel LIMITING SHEAR STRESS VALUES The ultinete shear stress, v,, allowed in manters designed by the shear friction method is: Vy 1062 Pre <= 800 psi end the corresponding value, related primarily to diagonal tension in the general design section, vy = 20VFe Values for these tuo restraints are compared in Table S.2e1. From this comparison it can be seen that for the normal range of concrete strengths used the second restraint of 10/Fre 4s more restrictive. Hence if a construction joint is designed by the shear friction method it will not control the design of the member. 5_2 REINFORCING CONTENT The shear frict’ fon method has no maximum or minimum restrictions on the limit of the reinforcing ratio, The steel ratio required is given by: pty = ie CONCRETE STRENGTH 02 Fle w/t (Pst) (Pst) (pst) 2500 500 500 3000 600 548 4000 800 632 5000 00 706 6000 00 775 6400 800 800 7800 800 866 whore sis the friction coefficient for the surface = 14 monolithic concrete = 1-0 concrete against hardened concrete 0-7 against rolled structurel steel For the sske of a comparison of steel requirements the value of the capacity reduction Factor of f may be taken as unity. ‘The corresponding relationships are shoun in Fig. Ss2ele For a section of @ Flexural menber subjected to shear the code normally requires pry su, = 2/ Fe and pfy= 50 if v= Vio The 1971 code contains a separate section on the design of walls subjected to shear in hich the ultimate sheer stress allowed is also 10 VFte. The horizontal shear reinforcement for walls is in general the sane as stirrups in normal slender flexural members except that: a. the mininum steel ratio 1s p = +0025 bs the shear stress carried by the concrete can be larger if the wall is in compression. ‘The minimum verticel reinforcement for a wall is given by: Py * 10025 + O¥5 (265 ~ H/0) (p, ~ +0025) my, p= +0025 uhere p, = vertical wall reinforcing Pp = horizontal wall reinforcing 4 = overall height of wal D = horizontal length of wall Thus for walls with 2 greater neight to depth ratio than 2+5 the minimun vertical reinforcement ratio required is p = +0025 The various steel ratio requironents for a conerote strength of Fie = 4000 pei, and a steel yield stress of, ty = 40,000 are shoun in Fig. 5.2.1. The chart illustrates thet for a monolithic structure the steel ratio depends on the dosign mathad used. This ie not 96 inconoiatan as it firet appears. For general cross sections diagonal tonsion cause: by the shear normally controle the design. For cases where design by shear friction is applicable, the failure is not 11ksly to be caused by diagonal tension. The shear friction section of the ACI code provides @ sound basis For the design of horizontal construction joints in wells. Kovever Joints may be over designed if large compressive forces are acting on the Joint. 6 ALLOWABLE 900 00 700 600 00 400 300 200 190 Concrete Strencth f'c = 4000 pai Steel Yield Stress fy = 40,000 pet Limit for Shear Friction Oesign \ mt 7 Shear Friction = 1.4 0 / Monolithic Concrete / (Requiring vertical Ny / steel) a Se Limit 10 Fre / — LL Normal Shear Design (Requiring Vertical Steel) Ze i“ fe—_ shear Friction 4 = 1.0 Construction Joints (Requiring Vertical Steel) Vertical Reinforcing Required for walls /, (Note: Not Specifically for Sheer) 200 400 600 00 2000 PSI :1 RCT 318-71 Code Reinforcing Requirements for Shear Perhaps 2 better equation for design would be: ory = -F where F is the nett axial load acting per unit area across the assumed crack, positive for compression, negative for tension, Oesigners understanding the concept behind the shear friction method would probably make this adjustment automatically, The value of the friction coefficient of 1+0 may appear to be conservative if compared with the test reaulte contained in this report. However it should be noted that these teste were performed under laboratory conditions. 79 CHAPTER SIX CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH 6.1 CONCLUSIONS To study the important components of construction joints, four type: of joints were used. These were: 1, Snooth surface preparation without bond 2. Smooth surface preperation with bond 3. Rough surface preparation without bond 4, Rough surface proparation with bond Each type of joint had a characteristic behaviour that was dependent on the bond, surface preparation, concrete strength, and reinforcin: content which were the main variables. These affect the joint in the following manner: a. Surface Preparation: As expected a rough surface preparation was superior to a smooth one, The loads sustained by rough surfaces vere larger and the drop off in load, after the peak had been attained, was smaller than that for smooth surface preparation. The strength of the Joint vas not sensitive to the method used to obtain the rough surface. ALL the preparations used} acabbling, washing, retarder end scraping, in these tests gave similar results. be Gond: Joints with bond were more rigid and had a higher strength then those without bond. The magnitude of the bond strength varied greatly for no apparent reason. an ®. Concrete Strength: The concrete strength within the narrow Tange used in this project appeared to limit the ultimate shear strength Of the Joint. This limit was found te be in the region of 10//F¥e. Reinforcing Content: For shear stresses less than 10/#TS the steel content controlled the strength and type of failure of the join for a specific surface preparation, For low steel contents a crack formed along the failure plane and with the addition of further load the steel yielded. The damage to the concrete was localized to the failure ack, Far high steel contents a large area of concrete surrounding the joint was crushed. when the cisplecenent along the Joint was increased the area of concrete affected by crushing was also increased. In this case the failure was due to the breaking up of the concrete, which was not limited to a definite failure plane DESIGN OF Jo1NTS A recommended method for the design of joints 4s the shear friction concept. This de conservative if @ suitable low friction coefficient for the surface is chosen, for the tests completed and reviewed in this Teport a satisfactory friction coefficient appears to 1.4 for a roug bonded surface. However the ACT 318-71 cade clauses reconnend that the friction coefficient for construction joints uith rough surfaces should be 1.0. For design this is considered to be more suitable as all the known test results are for those completed in a laboratory and not on an actual construction site, The ACI 318-71 code does not explicitly require these relevant cleusos to be applied. If only the minimum vertical reinforcement is used in 9 shear wall structure, @ premature failure may occur along a construction joint. eyELIC Loaotna The cyclic loading tests showed that, the mode of failure for joints with rough surface preparations was ductile, and the failure for those with smooth surfaces wes brittle. & ductile failure is a very important property in earthquake zones. for this resson alone, all important construction joints should have a rough surface preparation. For joints, the stiffness degradation is large after the peak loa: has been attained, so this should not be approached, No reliance should be placed on the ductility of a construction joint, to absorb energy from a structure during an earthquake, The cyelic loading tests also showed that the joints with rough surface preparations sustained the ultinate loads predicted by the shaar friction theory well into the post-elestic range. In some cases the peak lead attained was much higher than the predicted ultinate load. However this occurred only in the first cycle. The peak load value would be unsafe to use for design purposes as it is only reached once, after which the strength value drops close to that predicted by the shea: friction method. summary The type of joints used in practice usually have either trouelled exposed aggregate or scabbled surfaces. Generally all joints are bonded. although this is sometimes affected by poor cleaning of the surface of the Joint. The other major factors in practice are that of poor compaction and segregation. However with good site supervision, joints of similar standard to those that performed well in these tests can be obtained, Such Joints in 2 well designed structure should not be the cause of any premature failure. However joints of much lesser quality do occur aa can be saen by the examples in Fig. 1.lsl. 6.2 SUGGESTED FUTURE RESEAR’ The Following items have been revesled during this project uhich could be investigated further a, The problem of fatigue due to many reversals of the applied load. Load reversal in tell shear wall structures is very commen due to wind loading, In thess cases it 4s important to know if fatigue at the construction Joints is 1ikely to affect the structure, and if eo at what stress intensities. by The effectiveness of the Joint may be improved with different reinforcing arrangements. ith the configuration used for this report the cover concrete cracked and eventually spalled off, This effectivel reduced the area of the joint by 30 to 50 percent. It may be better t have a single row of bars in the centre of the Joint. Alternatively, confining stirrups eround the bars may increase the strength and ductility of the joint. c. It is considered that for future test programs, specimens cast as one unit snould be included, A diract comparison will then be able to be made between the performance of monolithic and Jointed concrete. Other types of loading could be considered. The case of axial tension and shear on a joint mey be more critical than expected, due to the tendency for the axial tension to open up any cracks formed. 83 APPENDIX 8 Asl NISCELLANEDUS GRAPHS The graphs included in this aection are: (a) Load - Strain curves for the reinforcing bars (b) Plot of the experimental date, if not ahoun in the text, used to obtain the mean curves. SS3HI8 was 86 zoe ¥ TovY 89801 0g senang dyTs ~ peor EV OFT (s3Honr) dis or sot nat a wor got zor tot a ® t t T T T r i — t / ot ening 80x04 dnasr: 7 fi 19 soos dnazr9s = tn (18a) 02 zow ° Teoo. = d , aa tow + fang teja a/t Sur2so.UT0 | aeaTTs THT i WOT g¢ pepuog pue yénoy woriwavases TNTOC or fenang dtts ~ peoy oor, oy 00s 09 Load - Slip Curve for Tast a4 r 1 1 7 77 0 60" 80° Lo* 90° so* vo" cor zor To° f (Swany ars ! t ' oot 1 # — + aos BHODK, Kee. t anc Concrete Association Technical Report TAA 414, May 1969, maST, RAF, wuxiliary Rei ‘orcemant in Concrete Connections" Proceedings ASCE, Vol. 94, STS, June 195i, pp. 196: MATTOCK, AL. CK, 2.F.y and IBRAKIN, I.0.° "Shear Transfer in Reinforced Concrete", ACI Journal Proceedings, Vol. 66, No 2, Feb, 1969, pp 119 - 128, DAVEY, Ne “Construction Joints in Concrete, Bonding ew Concrete to Old", Ospartinent of Scientific and Industrial Research, Building Research Special Report No 16. BATE, Es. “Some Experiments With Concrete". Reinforced Concrete Review, Vol. 4, No. 7, Sept. 1957, pp 421 = 447, "A Study of the Tensile Strength of Concrete Across Construction Joint: flagazine of Concrete Research Vol. 6, No 18, Dec. 1954, pp 151 - 153. UNITED STATES ARRY ENGINEEA AT CORP'. OF ENGINEERS. RWRYS EXPERT NT STATION, “Investigation of Methods of Preparing Horizontal Construction Joints in Concrete. Tests of Joints in Large Blocks", Report 2. Vicksburg, July 1963. pp 20. Technical Report No 6 - 518. lo. le 12. 1a, UATERS, E. HANSON, te "A Note on the Tensile Strength of Concrete Across co Tuction Joints", lagazine of Concrete Research Vol. 11, No. + Nov. 1959 pp 163 ~ 164, "Composite Designs in Precast and Cast~in-Place Concrete". Progressive Architecture Vol. 41, + 9, Sept. 1960 pp 172 = 179. "precast - Prestressed Concrete Bridges. 2, Horizontal Shear Connect jon: Journal of the PCA Research anc: Development Laboratories Vol. 2, No. 2, pp 38 - 58 1961, of Bulletin 035, BIRKELAND, FU, and BIRKELAND, H.w. LOESER, P.3. “Connection in Precast Concrete Construction". A.C.T. Journal Proceedings vol. 63, No. 3, March 1966, pp345'- 363, "Shear Transfer by Agyregate Interlock". Unpublished i.€. Thesis at University of Canterbury, N.Z, (1970) COLLEY, B.E. and HUNIPHREY, HAA. NOULEN, "Aggregate Interlock at Joints in Concrete Pavements", PCA Research Uevelopment Laboratoria: Bulletin 0124, “Influence of Ayuregate Properties o Effectiveness of Interlock Joints in Concrete Pavenents". PCA Research and Oevelopnent Laboratories Vo. 10 No. 2, pp 2= 8 lay 1968 or Bulletin 0 139,

S-ar putea să vă placă și