You are on page 1of 10
Ss Seismic Sugar StreNGTH oF Reinrorced Concrete CoLumNs By M. J. Nigel Priestley,! Ravindra Verma,? and Yan Xiao" AAssrnAct: Aspects relating tothe shear strong of eitealar and rectangular ok tis under sets oaing ate presented. An ckamitaton of exting denn eau ‘tons evel ade diferenes i predited responses, Patclar emphasis laced e'modestapabieof representing the imcracion between flexral Juli aod Shea aengih, simple method te proposed whereby the strength enbanecment jronded by sual compreston seprate fom the eoerete component of shear [irengih and conidred fo real (rom arch action. The ssngih ofthe conse onmponcnt fs faured as sur dopaoment day meen, Peto of Shear strength rom the proposed aed atemativemcthus are compaved wih resus ito a wide range of tr of cols fang in shea. The proposed method 1 ‘own o prose ignfanty improved cretion with experimental rss. Com Eervalve modieatons ae made wo encble the method tobe used fr dese INTRODUCTION Satisfactory seismic response of reinforced concrete structures requires that brittle failure modes be inhibited, Since it is common practice to rely ‘on ductile inelastic flexural response of plastic hinges to reduce the strength requirements for structures responding to strong seismic attack, it is nec- essary {0 inhibit shear failure by ensuring that shear strength exceeds the shear corresponding to maximum feasible flexural strength (Ang etal, 1989). ‘This is the so-called capacity design philosophy (Park and Paulay 1975). "Although in reinforced concrete frame design, plastic hinges will normally be located at the ends of beams, column plastic hinges at the base of the structure are required to complete the plastic deformation mechanism. With bridges itis neither feasible nor desirable to locate the plastic hinges in the superstructure, and thus, the columns tend to be the prime source of energy Specal cate is needed when plastic hinges form in columns because the shen sttengthisa funtion ofthe flexural utility. As plastc-hnge rotations increase he widening ofexure shar racks rede the apt fo shear transfer by aggregate interlock, and the shear strength reduces. Examination Ufreapons of ge coum in eathguakes (riestly eta. 1992) enables 8 lea distinction fo be made between brit shear faiure ofthe column, Gccuring before the flexural strength ofthe column has been reached, and dhcile shear failure, where a degree of ductility develops hinges before Shear failure occurs. ‘This acknowledged inthe conceptual model for shear Strength proposed by the Applied Technology Counal ("Seismic Design 1991) and shown in Fig, 1, where shear siength assumed to decrease in "Prof. of Struct, Engrg., Univ. of California at San Diego, La Jolla, CA 92093-0085. "Sr Bridge Engr, CRSS Civ. Engr., Inc, 2500 Michelson Drive, Suite 100, Irvine, cA 92714 3Res, Sci., Univ. of California at San Diego, La Jolla, CA. Note, Discussion open until January 1, 1995. To extend the closing date one month, @ written request must be filed with the ASCE Manager of Journals. The ‘manuscript for this paper was submited for review and possible publication on July 15, 1993. Ths paper is part of the Journal of Struciural Engineering, Vol. 120, No. 48, August, 1994, ©ASCE, ISSN 0733-9445/94/0008-2310/82.00 + 8.25 per page Paper No. 6588. 2310 PS pecurat ve Gesidualy SHEAR FORCE, V 10 DISPLACEMENT DUCTILITY, » Fa. 1. ATC*s Model (Seem Gesgn 181 or meracton betwen Shee Sen and Duetitity. oa » ae a incr fshion asthe displacement duty increases, 1 he shear force tovresponding to Hexsral stent sles than the sess sear sta shetie tex respone is cnsued: if gence thaw uses Stenin'a bride Shear tae via: I shee ene inland esta hear ents ton shen fore sean ae see corresponding to the intersection of the strength and force-deformation charters, show i ig Aah it behnor asl wel acepe, hs ot ound is wah into conree desi codes, except n great pie fone: Aca 89 ("Building Code" 1989) contains a equireient that shear sucngth 1 attributed to concrete mechanisms should be ignored in ductile design when ihe anil oad rat is ese than P= GS/she where fe ie impression suength and ic the prow wes fie New Gonzi Code Code" 148) puts ee Ofer P< Uyearaaiey eat Plastic hgeretione: Neither cludes an expt eatoashp heen at tity an Searsvengt econ considerable experimental research particusly by Ang et (1988) nd Wong eta (983) on ea ols farbcon deed eae beter detnton ofthe shear stenghvdetiy team Anna sis fom tet by Petey eta 98th) on ci ad anaes Cols have soplemented hs database, Fares dts eom tne or Rae tek and Wang {a} lisa and Woodard (1980 and east Use) 1s well yan extnnive Japanese databae (Watanabe ad itis 4) Wi be refered fo int paper. Most of hese were nett Columns tat fed without Seveloing sigan dusty: oe cause ofa i, hear bonds of confinement) could nb ucts from he ava torsion, Guy those comms whee ther faites have boon included ths nvestigtion A Tal a ke shiaan i vaaiecoemhere Prey ot a 95) 2a ‘This paper examines a number of methods to predict shear strength of columns, and compares results with the existing database. A refinement ‘and simplification of the method developed by Ang et al. (1989) and Wor Ct al. (1993) is proposed, which results in a close agreement between pre dicted and measured shear strength over the full range of the experimental data base. PREDICTIVE SHEAR-STRENGTH EQUATIONS Codified shear strength methods of design cannot be considered as pre- dictive equations since they are intended to provide a conservative and safe lower bound to strength, and as such will not be considered in this paper. A detailed review and citique of the American Concrete Institute (ACI) design approach, which is the basis for most of reinforced concrete design in the United States, has been presented elsewhere (Priestley etal. 19940). ‘The ASCE-ACI Joint Task Committee 426 (1973) proposals for shear strength cannot be considered fully predictive equations in terms of the scope of this paper, sinee they do not consider the influence of ductility, and ate for- fhulated in-a conservative fashion appropriate for code. implementation Nevertheless they are based on a very extensive review of existing data and design equations, are widely respected for representing the influence of parameters not considered in the ACI approach, and have been adopted is biher codes [for example, in the New Zealand concrete design code ("Code {982)]. As a consequence, they will be reviewed here. ASCE-ACI Committee 426 Proposals "As with ACI design equations, the shear strength of reinforced concrete members is considered ta be the sum of the strength of the concrete shear- resisting mechanism, V., and the strength of the truss mechanism, V,, in- Yolving transverse reinforcement contributing transverse tie forces to a 45° truss, and diagonal concrete struts that carry the shear force. Thus Va = Vet Ve o “The truss-mechanism strength is given by AL , = Al ° where A, = the total transverse reinforcement area per layers = spacing longi axis = the steel yield strength; and d = the effective depth, ortally taken‘ 0 oF USD for restangular or circular columns, re- apeetvely hen the yeinforcement is dstributed around the ercumerence “Tuo aemate methods are given for evaluating Vin he “simple” method, se a(t ja) ot . where the base shea tess is given by: (0.067 + 109,)Vfi = 0.2VF, — (in megapascals) (4a) uy = (0.8 + 120p,)VFi s 2.4VFZ (in pounds per square inch) (4b) v. where py = the ratio of longitudinal tension reinforcement, taken as 0.5p, zai for clu the a hear and rectangular lamas, repcetvely tro on of elective depth (3) and (4) 40) ples a efecve shea oh A, = Osbh = 8A, tor estangul? colin and hee Bae 102A, for circular columns, where A, is gross column area, There is thus an apparent dierepaney between siclarand retagulr cole alhoreretned approach equates V, tothe leer he shea fouired omit Menara stacking Ys and web shear cracking Von Fea lunns subjected to actions induced only at the ends, V, can be" expressed cement ratio) as D and b for rtion with the det My Va = mya, + Me 6) where a = MIV = the ratio of moment to shear at the critical section ‘M, = the decompression moment, given by a Matt ) where J = second moment of area: and y, = /h/2 and D/2 for rectangular and circular columns, respectively. Substitutin and crcula colunes, respectively. Substituting for A, 1, and in(3) yells wt) + 0167 o vam wlan + 010528 ey ‘The web shear cracking force for a column is given by Ven = (0.29 ost (o29vn +034 bpd (in megapascals) (Ba) P Yue (250 +03) nud Gin pounds esque inch) (8) which, on substituting b4d = 0.84, oF 1.02, for rectangular an Columas, respectively, yields, for eéstangular éolummss et "4 SURE Vou = 0.29VFA, + 0.24P (MPa) = 3.8VF.A, + 0.24P (psi) (a) For circular columns: Vow = O.29VF-A, + 0.31P (MPa) = 3.5VfLA, + 0.31P (psi) (9b) Ikis apparent that (9a) and (9b) will never govern Both (7) (02)~(9) imply adel proportionality between shar stespte see aa Toad and (i) and (7) imply a near eltonsip betncen shear sent and ihe inverse ofthe shearspan to“ depth rae ah oa If we consider the upper limit of vy, = 0.20Vf, MPa (2.4VF. psi) in (3). 2313 it can also be manipulated into « form similar to (7a)~(7h) and (9a)—(9)) substituting for A, = bed, resulting in, for rectangular columns: 76P (ps) (1) For circular columns: 0.60? 7.20P, (MPa) 7 ps (mPa) Ar Psi) £20 MPa = f= 35 MPa only weakly dependent on Ve = tebe + (106) ‘Over the typical range of concrete streng (G,000 psi =f = 5,000 psi), (10a) and (106) Fe.and they may be rewritten as ve eas +0105) ea) = + 018 (9) 9 ay, with reasonable accuracy for both rectangular and circular columns. Com parison with (7a) and (76) indicates that the latter equation will, predict higher_values of V. only when p, is low (and thus vy < 0.2Vj- MPa [2.4VFi psi), andor when a/Hf or a/D is less than about 1.5. In light of the physical characteristics of the columns in the database considered in this paper, (11) {and thus in more complete form, (3)] governs for prediction of shear strength. ‘A comparison of experimental results and shear strength predicted by the ASCE/ACI 426 equation is provided in Fig. 2. Only data from columns failing in shear are included, though many of these cofumns achieved flexural strength before failing in shear at ductility levels noted in Fig. 2(a) for circular and rectangular columns, respectively. Experimental data in Fi 2, and in subsequent theory/expetiment comparison, included circular and rectangular columns with concrete compression strengths in the range 13.8 MPa = f. = 42 MPa (2.0 ksi = ft = 6.1 ksi), transverse steel volumetric ratio in the range 0 = p, = 0.024, longitudinal steel volumetric ratio in the range 0.0122 = p, = 0.05, yield strength of transverse reinforcement 300 MPa = f, = 455 MPa (43.5 ksi = f, = 66 ksi), and shear-span—to-depth ratio in the range 0.9 = MIVD = 2.7 In Fig. 2, data are organized by displacement ductility, axial-oad ratio, ‘and aspect ratio (shear-span~to-depth ratio), dividing the observed strength, Vas by the predicted strength. In all cases, predicted strengths are based on measured material properties rather than nominal values. All columns ‘were subjected to cyclic reversals of lateral displacements of gradually in- creasing magnitude. Displacement ductility is referred to the measured yield isplacement, extrapolating the line from the origin through the displace- ‘ment at first yield, to the theoretical flexural strength, based on measured material properties. The ASCE/ACI 426 equations are generally conser- vative, with the conservatism decreasing as the ductility increases, with several data points falling below the predicted strength at ductility levels freater than or equal to three. This behavior is expected based on predictions fof the model of Fig. 1, since the strengthiductility degradation isnot included in the ASCE/ACT 426 approach. ‘There is no obviously discernible trend of the measured/predicted strength io with increasing axial-load ratio, but the scatter is rather large, partic zane fo Nan eet + lstment Duc Dipleceient Sect 5 Yate 8 er ere &. geet ‘ape tate GH) a tern tate spect Reis (Ay) Coe FG. 2. ASCE/AC 426 Equation Compared wih Ex pared with Experimental Results Related to (a) Displacement Ductility; (b) Axial-Load Ration; and (c) Aspect Ratio Fetted lary for retangula columns. making the trends hard to identity. Pedic tans or rextanglr columns appear fo be moe agen ate th cular columns, with masimurn rts of 22 and 1-75: respectively, This arly due tothe nonconservative assumption of effciive sat agea of the Sitcularcolurns, noted earir, anda resuit ofthe greater numberof eneulor clu ing st moderate doy ie fn icteare i column aspect ratio (MIVD) from 1 to 3 typical causes gradual eduction inthe VeVi sat, ations these val abso affected by a simultaneous chage fy other parameters ‘Shear-Strength Equations of Ang et al an ea a ig et al. (1989) and Wong et al. Based on a large number of tests on circular cantilever columns un rial (Ang eta 198) and muliectona (Wong etal 1988) ey lateral displacements, a column shedr-srength model based on the ATC 2018, ding strength relationship (“Seismie Design” 1981) suggested in Fig {Ro bech developed. ‘The modes based onthe adv approach Of (1). wth the nel shear strength dened by vy = Abad: (12) and 3P ‘ PA, osta (1 + RA. (in megapaseals) (13a) v. 4.450 (1+ im (in pounds per square inch) (136) where « = 2\(MIVD) = 1.0; and D and D’ = the column diameter and Tore diameter measored tothe centerine ofthe transverse hoop Or spiral sich nana erosesecional area, Ay and yield strength fy, The effective Bitar aten Ac = OSA and other symbols have been previously defined. Eq, (12) is based on analysis of effective shear resistance provided by trans- lsd oops assuming a 45" ruse mechaniam (Ang eal. 1989). Eqs, (132) Sad (130) are of Siar form to (3) and (2) bat mp substantially higher SEneete contsbutton, particule for squat columns ‘The residual shear strength in Fig. Las based on a reduced concrete contabution, ut increased stength of the tse mechanism corresponding to nateeper inclination ofthe dagonal compressions struts, in accordance ‘wh te following equations V,, = 0.1858.VF:A, (MPa) = 2.23BV7A. (psi) ay Vy, = Vy cot 8 as) where the inclination 0 of the diagonal compression struts to the column taxis was based on plasticity theory as cot 0 = V(I ~ WM; and W = pula! (0.2f%) = the effective transverse mechanical reinforcement ratio. The ratio of transverse reinforcement is p, = 2A, /(D's) (.e., half the volumetric 4 wi SHEAR FORCE, V 1 ' ' 1 1 Ductity 1 ' 1 ‘Biaxial ' 1 1 1 1 ' ' a T z 7 Wr DISPLACEMENT DUCTILITY, FIG. 3. Ang/Wong's Model for Shear Strength versus Displacement Ductlity 2316 Diepactmest Decay o dapat Rao (HD) © FIG. 4. AngWong Equations Compared with Experimental Results Related to (a) Displacement Ductity;(b) Axlal-Load Ration; and (c) Aspect Ratio ratio) and B = 100p, = 1. A lower limit of @ = 25° or the corner-to-comer column inclination sas proposed. For columns subjected to uniaxial ductility, the shear strength inthe Ang! ‘Wong model decreased from the intial value ata displacement ductility of Value at the flexural ductility capacity (as shown ial ducility is expeeted, the ductility limts are each decreased by one unit, as is also shown in Fig. 3. Fig. 4 compares the experimental database for circular columns with predictions of the Ang/Wong model. Rectangular column data are not in- cluded since Ang et al. (1989) and Wong et al. (1993) did not produce predictive equations for rectangular columns, and in the absence of a defined Flexural ductility capacity, a value of uy = 4 was assumed. Fig. 4 indicates significantly improved prediction of shear strength compared to the ASCE! ‘ACI 426 model, particularly for columns subjected t0 uniaxial ductility zat ‘There isa tendency for the scatter to be greater for the Wong et al. (1993) columns tested under biaxial ductility demand, and for both seatter and Conservatism to inerease as displacement ductility and axial-load ratio in- Grease. ‘The Ang/Wong model appears to underestimate the influence of ‘axial load on shear strength, while the influence of aspect ratio seems to be adequately represented. ‘Shear-Strength Equations of Watanabe and Ichinose (1991) ‘Watanabe and Ichinose (1991) have proposed a design method for ree- tangular sections, based on superposition of arch action and truss action usinga lower bound plasticity approach, limiting diagor ssion stress resulting from the combined arch and truss act the permissible diagonal compressive stress is progressively reduced as th plastic rotation increases, thus indicating behavior similar (o that of Fig. 1 Mihis approach has been adopted in the recommendation of the Architectural Institute of Japan (Design 1988). Initial shear strength follows the recom- mendation of Nielsen et al, (1978), by limiting the diagonal compression stress 10 wolfe (16) were (inmenpacas) 8) a aerse re eee ce Fig. 5 shows assumptions relating to the truss and strut mechanisms. The shear force V, carried by the truss mechanism is given by = AAD con wy where jt = the distance between upper and lower chords of the analogous truss, and cots [eats = Aub 520 as) ‘The corresponding diagonal compression stress due to truss action alone ar “The shear force V. carried by arch action is based on an assumed compres- sion zone depth of half the member depth, as shown in Fig. 5(b), and is given by (1 + cot%4) (20) Va = (vee = fad tan 0 @ where 0 is defined in Fig, $(6). Eq. (21) conservatively ignores the different 2318 (6) FIG, 5. Analogous Models In Watanabefchinose Approach: a) Tues: and (b) inclinations of the concrete compression diagonals in the truss and arch ‘mechanisms, when determining the compression stress (nf? ~ fy) that ea be allocated to the arch mechanism. ‘The total initial shear strength is the given by Det a Eq. (22) is quadratic in cot . The implication of the critical diagonal Wf = (1 = 158)vef2 = 0.2500f2 3) ‘The tower limit of (23) is reached at a plastic rotation of § = 0.05 radians, ‘Also, the maximum allowable value of cot ¢ in the truss mechanism was reduced to allow for loss of aggregate interlock, in accordance with 1s cot <2 - 508 4) Watanabe and Ichinose (1991) also suggest that this approach could be used for circular columns using an equivalent square section of equal cross-sectional 2319 areato the cela eolran for preditng te diagonal compres arash rss mchanan tenth needy the for cov Te anson between peictons ofthe Watanabe and shinee ee ns an ea is bter than sith the ASCEIACI 426 approach, columns the Hecate Of Vag! Vay. = 1-85, though scatter is sil large FR eeu dgerible endfor inctening conservatism 8 the aa Tee nn Tren with vying duct evel o uec ae not arly discernple. Values predicted for circular colus a not clearly dere seater than with rectangular columns. PROPOSED PREDICTIVE EQUATIONS 1 design equations deserived here produce, with few exkltvout al iy consteauve predictions of shear stength, the seater "pageant Ft Tea ate rR Ba Gea a8 wD Sel rdain Rectang Colas » [Moun tne npnae a ‘ape Hate Q/) ‘ean Coa. o wns Compared with Experimental Results Re- FG, 6. Watanabechinose Equations Compared with Experimental Resi fated 10 (¢) Displacement Ductliy; (b) AxiaHLoad Ration 220 is rather large. and as a consequence the average ratio of measured to predicted strength is high. As noted before. the ASCE/ACI 426 approach (ASCE-ACI 1973) does not consider the influence of ductility. a result that is evident in the experimentipredictions ratios of Fig. 2. The Ane Wo: model works well for low ductiities. but scatter increases at moderate to high ductility levels. apparently as a consequence of the residual shear strength being assumed independent of the axial-load level and the aspect ratio. The Watanabe/Ichinose approach provides good predictions for re {angular columns at low axial-load levels, but the lack of specific consid: ration of axial load leads to increased conservatism as the axial-load level increases. For ductile shear strength, the strength of both arch and truss ‘mechanisms are reduced in the Watanabe/lehinose model. though it docs ‘not appear obvious why the truss-mechanism capacity should be reduced. Indeed, experimental data indicate a reduction in the inclination of diagonal cracking 10 the column axis as the ductility increases, implying an increase in the truss-mechanism capacity. AAs noted in the preliminary discussion of the ASCE/ACI 426 approach. existing design equations for concrete mechanism strength commonly used in the United States and elsewhere can be manipulated in the form Vo = Vig + AP 25) where V.. = the value applying for P = 0; and ky varies between about 0.1 and 03, depending on the base design equation chosen. Also, in ‘one equation {(7)], the factor , depends on the aspect ratio a/ff. ‘These observations ate incorporated into a simplification of the Ang/ Wong model in the following form. The shear strength of a column is considered 10 consist of three independent components: a concre ponent V_ whose magnitude depends on the level of ductility. an a component V,, whose magnitude depends on the column aspect a truss component V, whose jude depends on the transverse rein forcement content. Thus ea 26) Concrete Component V, ‘The conerete component for both circular and rectangular columns re- duces with increasing ductility in accordance with the form of Fig. 7. and Ve= kVA, en where & depends on the member displacement ductility level and the system OF unis chosen (megapascals or pounds per square inch) ay well se on whether thecolumn is expected tobe subjected io uniaxial or bina uti demand. In (27), the effective shear areais taken as Ay = 0.8A.,., for both Sirculrand rectangular columns: Fig. 7 i snplifiation of We Rng Won ductility’ model, and the upper limit to KVP of UOIVTE Moe DSU TE psi) is the same as implied by (9a) and (9b) for web sheat cracking in the ASCEIACL 426 approach. Eq, (4a) and (4 imply a dependency of Von longitudinal reinforcement ratio. The range of the database was insutficient to determine wheter sack 5ninluence was significant, and pending results fom current researc progress. the simplified form of (27) 1s segpested 2a21 Ductility (MPa units) 3004 Member Displacement Ductlity FIG. 7. Degradation of Concrete Shear Strength with Ductility ‘Axial Load ‘Axial Load * hy % h | * Ha L ip gli wy tie i (a) FIG. 8. Contribution of Axtal Force to Column Shear Strength: (a) Reversed Bend Ing; and (b) Single Bending ert . the axial-load level, whereas the Watanabe/Ichinose model is independent ot ihe na fad eve A ease to shear srengi isthe horizontal component ofthe pee diagonal compression strut, since this component directly resists the applied shear force. Thus Pune » @s) a where D = the overall section depth or diameter; c = the depth of the compression zone; and a = L for a cantilever column and L/2 for a column in reversed bending. There are similarities in form between (28) and (7a) (76). though (28) implies much greater dependency on the effective aspect ratio al(D-c) = MI[V(D-c)]. V, is not degraded with increasing ductility. Eq, (28) implies that as the aspect ratio of the column decreases. the load contribution to shear strength will increase, and that for very slender columns, the axial-load cont ion may be rather minimal, Since shear force is likely to be low in such cases, this may not be significant. Eq, (28) also implies that as the axial load increases, the effectiveness of the avial- load contribution to column shear strength will decrease sinee the depth of ‘the compression zone increases, ‘Truss-Mechanism Component V, ‘The contribution of transverse reinforcement to shear strength is based ‘on a truss mechanism using a 30° angle between the compression diagonals (i.e, erack pattern) and the column axis or the corner-to-corner inclination, whichever is larger. Thus both initial and residual strength of the truss ‘mechanism are given by, for circular columns v, = FAD con 30° (29a) 2S and for rectangular columns Ape v, = Abb corso eo $ where D' = the distance between centers of the peripheral hoop or spiral, 2 defined in Fig, 9. This is similar tothe. AnghWong, model for cesta columns, but implies a greater effective depth for rectangular columns than traditionally adopted. This is based on the supposition that the number of hoop layers crossed by the diagonal crack is more significant than the dis: tance between tension and compression resultants. In Fig. 10, predictions of this proposed model are compared with the experimental database for circular and rectangular columns, Comparing nh o ma FIG. 8. Definition of D" for Truss-Mechanism Strength 223 BL a foteton/tones ° ip ot oe Raia WD) ape Bate G1/¥D) Gree Come ecanatar Ceara FIG. 10. Proposed Model Compared with Experimental Results Related to (#) Blaplacement Ductiy;(b) Axlak-Load Ration; and (e) Aspect Rtio these results with Figs. 2, 4, and 6 indicates that a greatly improved pre~ cae a has resulted. For circular and rectangular columns, the maximum GAperimenvprediction ratios are 1.25 and 1.35, respectively. The strength care rtion factor of 0.85 effectively provides a lower bound to the experi- reoutalodata, ‘The influence of ductility level, axial load, and aspect ratio appear to be well represented by the proposed method. Te statstical comparison of the different methods discussed in this paper is provided in Table 1 in terms of mean (m) and standard deviation (9) for ine caperiment/predicted shear-strength ratios for both circular and rectan- thar golumns. (ne and m ~ 2a are the statistical lower bounds), The Sroposed method provides the closest agreement withthe data. with a mena Fregth ratio of 1.021 and a standard deviation of 0.124. This standard Seefttion is tess than 40% of that resulting from the ASCEJACI 426 and Watanabe/ichinose equations, and 61% from the Ang/Wong model {could be argued that the proposed method is a predictive equation, 228 ‘Terms of Experimental/Predicted Shear-Strength Ratio Statistical Comparison between Models and Data, TABLE 1. mee (3) | fo/22 28 *).]gg 88 8 e218 a z gree ‘pale é ealS1 #8 galBS a8 a e Arelug 33 3 i)-clgd 33 8 @ 7.209 1.060 1s 0.995 Model w ASCE-ACI AIS ‘Ang Wong ‘Watanabellh Proposed 5 whereas the alternative methods are design equations. As a consequence, inher average values of the strength ratio are not only to be expected but via pe lie as well. This is of course true, but the final determination of the Sppropriateness of the design approach is dependent on the lower limits to Re Ghhalprediction comparison. As is evident from Figs. 2, 4, 6, and 10, 2 Strength Feduction factor of , — 0.85 forms a reasonable lower bound to Silmethods, with a few isolated exceptions. Some data points are overpre~ dicted by the ASCE/ACI 426 equations, and four data points in the Ang! Wong model {all below g, = 0.85. For the Watanabellchinose model, all Gata points fall above %,, = 0.85. For the proposed method, three data points are marginally below ¢, = 0.85. "Also. in the comparisons provided in this paper, shear strength was in all cases predicted using measured concrete compression strength and trans, atige reinforcement yield strength. In the design situation, nominal material ‘Srength would be used, which in the vast majority of cases will result in stlaltonal conservatism in the predicted strength. It can therefore be argued that the proposed method is adequately conservative. However, decreasing the streagttrof each component of (26) by 15% would result in 9 design uation that sil hada lower mean strength ratio than the ASCE/ACI 420 Say Watanabe/lchinose equations (and thus would be more economical), wate providing substantially greater safety against failure of the extreme care ues of the data bases than provided by the other methods, since a Value of , = 1.0 would then provide a lower bound to the data. nis wold be effected by the following changes to the component strengths: 1. The upper and lower limits to k in (27) would become 0.25 and 0.085 using mepapascals (3.0 and 1.0 using pounds per square inch). Hee rial load contribution of (28) would become V, ~ 0.85P tan a 3. The truss mechanism would be based on an angle of @ = 35° rather than 30° CONCLUSIONS [A predictive equation for the shear strength of columns of rectangular ‘and deealar section subjected to cyclic lateral shear force was presented. ‘This equation separately considers the contributions of concrete mecha- ine: axial load mechanisms and truss mechanisms. Although separating teenetiaiload component from the concrete mechanisms initially appears Hovel it was shown that many existing design equations can be manipulated into a similar form, "The predictive equation was compared with an extensive database of column tests: It provided a very good simulation ofthe influence of flexural Suctlity, axial load, and aspect ratio, with a mean value of measured to predicted shear strength of 1.021 and a standard deviation of 0-124 ‘the proposed approach provided significantly improved prediction of sheat sength than alternative methods, particularly in the eritical aspect Siieduced scatter of the measured/predicted shear-strength ratio [A design approach adopting a slightly conservative modification to the proposed predictive equation would still result in a more efficient design rrepestisting methods (as measured by the mean predicted strength ratio) while providing better protection against shear failure for the low end of the experimental database. ‘Preliminary comparison of the proposed shear model with results from 2328 reinforced conerete structural walls indicates that the met pled. without moication to stuctral wally Ths j encourage see the boundary between walls and colu i rary. F fered 10 examine thsi et ets in progress on columns with low longitudinal reinforcem (Priestley and Benzoni 1993) should determine whether or not the strength mis ceaely arbitrary, Further work support of Caltrans under grant RTA $9626. the vey of Calon at San Diego» an the New Zealand Earhquske a Mir Banage Commision grateiyakvonedged. Te commons gstion ah oon mate nth paper owt seve ee APPENDIX I. REFERENCES ASCE-ACT Joi Task Commie 426, (1973) “Shea sgih of vise mentee Sac Engrg ASCE, 930) 1@Ie1I6) Tra tae BGM I od 8). ACI S159. ‘Am. Coner. Inst., Detroit, Mich. apie ‘Assodiion af New Zealand; Wengen, New eaands S277 Starts encert }). Structural Committee, Arch. Inst. of Japan, Tokyo, J i ra. and Windward KA (1980), Bena caaiaion of eet nonce yeti fC Engr ny oR an ee ee NO 802 ets subject to high axial compressive stress." ACI Siruct. J.. SL. 287-29. a vicar ARBs Gita co yn pense Bae ral Engineers, Zurich, ‘Switzerland, P-1S/78, 1-16. anes ‘ith kw inguinal eintorcengn eis’ Sng Sc Rear Por Report No. SSRP 93109, Univ. of California, San Diego, Calif. ae Priestley. M. J. N.,Seibe, F and Chai, Y_ H. (1998) "Des sesmen fi nd ea of bridge for oe pertormance Yi te Si ce ala aes a of California, San Diego, Calif. a retrofitting of reinforced conc brids ei hearse re I: a considerations and test design.” ACI. ‘smc 91, to ag one ley, N., Seible, F., Xiao, Y., and Verma, R. (19%c). eck rerting of reifreed concrete tg columns fr enaned shee seth _ ia Tetra empl ut eS sod Se mee os ee” aD TC soe Watanabe, F., and Ichinose, T. (1991), “Strength and ductility design of RC members aor guidelines for Srructral Ss3- of California, San Diego. subjected to combined bending and shear.” Proc., Workshop on Concrete Shear in Bowthquake, University of Houston, Houstin, Tex,, 429-438, Wong, YL, Baulay, Tay and Priestley, M. J. N. (1993). “Res Tatorced conetete columns to multidirectional seismic attack.” ACI Struct. J-. 9012), 180-191. Xone. omit: M., and Sakino, K. (1986) “Experimental study on design method Wo prevent shear failure of reinforced concrete short circular columns by confining fa ftet tubes.” Trans, Japan Concrete Institue, Tokyo, Japan, 8, 335-542. APPENDIX Il, NOTATION The following symbols are used in this paper: effective shear area; ‘gross-section area, frea of cross section of transverse reinforcement in form of hoop or spiral; total transverse reinforcement area per layers ‘MIV, ratio of moment to shear at critical section; web width; ‘compression-zone depth; circular column diameter; core diameter measured to centerline of transverse reinforcement; effective depth of member; diagonal compression stress; ‘compressive cylinder strength of concrete; Yield strength of transverse reinforcement; Tectangular column depth (also H); second moment of area; distance between upper and lower chords of analogous truss in ‘Architectural Institute of Japan (ALJ) model; factor for influence of flexural ductility on concrete shear capacity; axial load factor for V.3 column height; ‘decompression moment; axial load acting on member; spacing of transverse reinforcement along member axis; Shear force carried by arch mechanism in ALJ approach; Shear carried by concrete shear-resisting mechanisms; final concrete shear capacity at large ductility in Ang/Wong model; shear force required to induce flexure-shear cracking (ASCE-ACL 4426) initial concrete shear capacity in Ang/Wong model; value of V_ applying for P= 0; Shear force required to induce web shear cracking (ASCE-ACT 426); nominal shear strength of reinforced concrete member; Shear strength enhancement provided by axial compression; shear carried by transverse reinforcement; initial shear carried by transverse reinforcement in Ang/Wong model; final shear carried by transverse reinforcement in Ang/Wong model; shear force carried by truss mechanism in ALJ approach; ‘nominal concrete shear stress; Hz and DP for rectangular and circular columns, respectively; 2328 @ = 2(MIVD) = 1.0 = inclination of diagon: HARD) = 1.0 = inctinaton of eigonal compression struts (due 100p, = 1.0; blast rotaton in hinge; iclination of diagonal compression stress wi tntaton of diagonal compresion stress wth member axsin ass aieplacement duty factor: factor for limiting diagonal compression stress = 0.7 — Sas orgs da P tress = 0.7 — (1/200) in ratio of transverse reinforcement gross longitudinal reinforcement rat ratio of longitudinal tension reinfore effective mechanical reinforcer inclination of diagonal compressive struts to column axis. 2329

You might also like