37where:vxyV= (10.5)b ⋅ dvFrom Equations (10.2), (10.3) and (10.4) ε 1 can be expressed as:* 2ε = ε + ⎡⎣ε + ε ⎤1 x x x⎦⋅cotθ⎛ v⎞*xyε x = 0.002 ⋅ 1− 1− ⋅ ( tanθ + cotθ) ⋅ ( 0.8 + 170 ⋅ε1)⎜ ′⎟⎝fc⎠(10.6)Considering the MCFT as a truss model with a concrete contribution equal to the transversecomponent <strong>of</strong> the shear stresses v ci transferred across the crack (see Figure 70), the shear strengthmay be expressed as:Av⋅ fvyV = Vs + Vc = ⋅dv ⋅ cotθ+ vci ⋅b⋅dvsAv⋅ fvy= ⋅d ⋅ cotθ+ f ⋅cotθ⋅b⋅dsSubstituting Equation (9.19) into Equation (10.7):v c1v(10.7)whereAv⋅ fvyV = ⋅dv ⋅ cotθ+ β ⋅ fc′⋅b⋅ dv(10.8)s0.33⋅cotθ0.18β = ≤1+ 500 ⋅ε24 ⋅ w1 0.3 +a + 16(MPa,mm)(10.9)In order to use Equation (10.8) <strong>for</strong> the evaluation <strong>of</strong> the required amount <strong>of</strong> transverserein<strong>for</strong>cement, it is necessary to determine appropriate values <strong>of</strong> θ and β which must satisfyEquation (10.6). These values are generally given as function <strong>of</strong> the longitudinal strain ε x andshear stress level ν xy /f ’ c in the <strong>for</strong>m <strong>of</strong> tables or diagrams. The values given as example in Table23 ensure that the tensile strain in the stirrups is at least equal to 0.002 and that the compressivestress f c2 does not exceed the crushing strength f c2max <strong>of</strong> concrete. In determining these values itwas assumed that the amount and spacing <strong>of</strong> the stirrups would limit the crack spacing to about300 mm.13. CYCLIC LOAD MODELING THROUGH THE MCFTIt is worth discussing why some codes have maintained an approach based on the truss modeleven when more rational methods exist, such as the MCFT, which is now adopted, <strong>for</strong> example,in the Canadian, Norwegian and AASHTO LRFD codes. As observed in Duthinh and Carino[1996], the reason may be behind the fact that a whole generation <strong>of</strong> engineers has learned andused methods based on the truss approach. However, design engineers should take into account
38that there is a fundamental difference between the two approaches. According to ACI-ASCECommittee 326 [1962], “diagonal tension is a combined stress problem in which horizontalstresses due to bending as well as shear stresses must be considered”.Collins [1993] explains that this combination <strong>of</strong> flexure and shear is what has made the shearproblem so intractable. He contrasts the traditional type <strong>of</strong> shear tests (beam), which are simpleto per<strong>for</strong>m, but difficult to analyze, with the more recent tests (panel), which are more difficult toper<strong>for</strong>m but simpler to analyze. In the tests <strong>of</strong> beams simply supported with two concentratedloads, the behaviour <strong>of</strong> the member changes from section to section along the shear span andalso over the depth <strong>of</strong> the beam. In contrast, the state <strong>of</strong> stresses in a panel loaded in pure shear,or in a combination <strong>of</strong> shear and axial <strong>for</strong>ces, is uni<strong>for</strong>m.Because shear is studied independently <strong>of</strong> bending, the shear carried by the compression zone isnot taken into account by the MCFT. In addition, the shear carried by dowel action is neglected;only the shear carried by aggregate interlock is accounted <strong>for</strong>.The contribution to shear strength due to dowel action is rather small, while the contributionrelated to the shear carried by the compression zone can account <strong>for</strong> 25-30% <strong>of</strong> the totalconcrete contribution in beams with typical rein<strong>for</strong>cement ratios. Nevertheless, when the MCFTis applied to traditional beam tests, it per<strong>for</strong>ms rather well as a predictor <strong>of</strong> strength. The reason<strong>of</strong> this good agreement is related to the choice <strong>of</strong> the location (mid-span) <strong>of</strong> ε x , which reflects theredistribution <strong>of</strong> shear stresses transferred from the most highly strained portions <strong>of</strong> the crosssection to the less highly strained portions. However, it would be conservative to use the highestvalue <strong>of</strong> ε x , as an increase in ε x decreases the shear capacity. So, implicitly, the higher shearcapacity <strong>of</strong> the uncracked or least strained region is taken into consideration.The Seismic Shear Wall International Standard Problem documented by the Nuclear PowerEngineering Corp. <strong>of</strong> Japan [1996] has brought to the <strong>for</strong>e the inability <strong>of</strong> the proposed methods<strong>of</strong> analysis and modelling <strong>of</strong> concrete to give a reliable evaluation <strong>of</strong> the peak strength and theductility <strong>of</strong> structural walls subjected to reversed cyclic loading. Among these methods thesmeared crack approach 5 , which assumes fixed cracks, tends to be the most favoured. Okamuraand Maekawa [1991] and Sittipunt and Wood [1995], among others, have documented <strong>models</strong>assuming a fixed crack approach and have demonstrated a good agreement with experimentalresults. However, the fix cracked method requires separate <strong>for</strong>mulations to model the normalstress and the shear stress hysteretic behaviour. This is somewhat at odds both with the testobservations and with common elasticity approaches to constitutive modelling.An alternative approach based on the smeared rotating crack assumption was proposed byVecchio [1989, 1990]. The procedure was based on an iterative, secant stiffness <strong>for</strong>mulation,where the concrete is treated as an orthotropic material with its principal axes corresponding tothe directions <strong>of</strong> the principal average tensile strains and principal average compressive strains,modelled according to the constitutive relations <strong>of</strong> the MCFT. The secant stiffness <strong>for</strong>mulation ismarked by excellent convergence and numerical stability characteristics. Correlations toexperimental data <strong>for</strong> structures subjected to monotonic loading conditions are generally verygood. Vecchio [1999] has demonstrated that secant stiffness based <strong>procedures</strong> can be also usedto effectively model reversed cyclic load effects in rein<strong>for</strong>ced concrete structures. This result hasbeen obtained employing strain <strong>of</strong>fsets to model the plastic components <strong>of</strong> strain in the concreteand rein<strong>for</strong>cement, as hereafter illustrated in some details (Palermo and Vecchio [2003]).5 Smeared crack approach simulate cracking with a fictious constitutive model, thereby avoiding the need<strong>for</strong> changing geometry and mesh model.
- Page 1 and 2:
Simplified models/procedures forest
- Page 3 and 4:
Simplifiedmodels/proceduresfor esti
- Page 6 and 7: iiiTABLE OF CONTENTSLIST OF TABLES
- Page 8 and 9: vLIST OF TABLESTable 1 Mechanical p
- Page 10 and 11: viiLIST OF FIGURESFigure 1 (a) Rein
- Page 13: xFigure 68 Average stress-strain re
- Page 17 and 18: xivν xywxyyA gA hA vCC sD= Shear s
- Page 20 and 21: 11. INTRODUCTIONA number of importa
- Page 22 and 23: 3On the basis of these results, it
- Page 24 and 25: 53. NUMERICAL MODEL3.1 GENERALITIES
- Page 26 and 27: 7⎛ 8 ⎞ ⎛ s ⎞ ⎛ s ⎞α =
- Page 28 and 29: 9less clear if the numerical model
- Page 30 and 31: 11longitudinal reinforcement ratio
- Page 32 and 33: 13where A c is the concrete gross s
- Page 34 and 35: 15where B and H are the section wid
- Page 36: 17whereξW2πE=dd(5.10)χ yLEd = μ
- Page 39 and 40: 20Another series of tests carried o
- Page 41 and 42: 22Rectangular case, section 2-2:F =
- Page 43 and 44: 24concrete struts can form with the
- Page 45 and 46: 269.1 PRIESTLEY ET AL. [1994]Accord
- Page 47 and 48: 28is given by Equation (7.1), the p
- Page 49 and 50: 30edges remain straight and paralle
- Page 51 and 52: 32of principal stresses in the conc
- Page 53 and 54: 34vci max0.18 ⋅ fc′=0.3 + 24
- Page 55: 36F = F + C − C(9.40)k k−1 k1 k
- Page 59 and 60: 40where α i defines the reinforcem
- Page 61 and 62: 42where⎡N( Ec3 −Ec2)⋅Δε⎤f
- Page 63 and 64: 44where1βt= (13.37)1 + 1.15 ⋅rec
- Page 65 and 66: 46• Acknowledging that hollow sec
- Page 67 and 68: 48Menegotto, M., Pinto, P.E. [1973]
- Page 69 and 70: 50CSA Committee A23.3 [1994] Design
- Page 71 and 72: 52Wong, Y.L., Paulay, T., Priestley
- Page 73 and 74: 54Table 4 Values of the parameters
- Page 75 and 76: 56Table 13 Values of factor α s de
- Page 77 and 78: 58PART BTable 23 Values of θ and
- Page 79 and 80: 60100800Displacement [mm]500-504000
- Page 81 and 82: 6260004000Moment [kN*m]20000-2000-4
- Page 83 and 84: 64NumericalExperimentalSlice #3Figu
- Page 85 and 86: 66NumericalExperimentalSlice #4Figu
- Page 87 and 88: 680.04 (3)0.03 (3)(3)1.20.02 (3)0.4
- Page 89 and 90: 702010 -2181614121, 2, 31, 2, 32 3
- Page 91 and 92: 72μ2018161412108642χ y* H10 -28.0
- Page 93 and 94: 7420181614121 to 51 to 510 -28.0x10
- Page 95 and 96: 762010 -2μ18161412108642031212 3 1
- Page 97 and 98: 78μ2018161412108642χ y* H10 -28.0
- Page 99 and 100: 80μ201816141 to 53214513to52χ y*
- Page 101 and 102: 82μ201816141210864203213213213211
- Page 103 and 104: 840.400.35ρ = 0.0050.400.35ρ = 0.
- Page 105 and 106: 86Compression chordPART B2PTransver
- Page 107 and 108:
88Figure 64 Contribution of axial f
- Page 109 and 110:
90Figure 70 (a) Calculated average
- Page 111 and 112:
92Figure 76 Hysteresis model for co
- Page 113:
The mission of the JRC is to provid