I highly appreciate Professor Mesri’s discussion regarding some of the salient details and limitations of the analytical approach presented in my paper (Sadrekarimi 2013). I have provided further clarification of these points in the following paragraphs. The objective of the paper is to illustrate the combined influence of state (cs) and compressibility (cs) parameters on the undrained critical shear strength, su(critical), and that cs on its own is insufficient to characterize and describe sand potential for liquefaction flow failure. As stated in the conclusions of the paper “The critical-state parameter method should be only used for identifying strain-softening sands (cs > 0) and the evaluation of su(critical) for a particular sand deposit rather than among several different sands. A state–compressibility ratio (cs/cs) is introduced in this study that is a more inclusive parameter for evaluating liquefaction susceptibility, postliquefaction strength, and the likelihood for the occurrence of a catastrophic liquefaction flow failure. In summary, although the outcome of this study does not provide any direct solution to any stability problem, it illustrates the significance of soil shearing–compressibility (cs) in the failure of cohesionless soils.” The intention of the paper is not to encourage the direct application of eq. (6) to field cases or design. I fully agree that the major limitation of using eq. (6) to estimate su(critical) is the difficulty of estimating cs and cs of in situ soils. Equation (6) is highly sensitivity to variations of cs and cs and small inaccuracies in assessing the field values of these parameters could result in significant errors in estimating su(critical)/nc from this equation. The discusser has suggested that eq. (6) be evaluated by comparison with su(critical)/vc mobilized in the field in terms of (N1)60 or qc1. This may not be possible at this time, as accurate information regarding the in situ cs and cs of soils involved in past liquefaction flow failures is very limited and possibly inaccurate. For example, compared to su(critical)/vc estimated from eq. (6), (N1)60based correlations (Mesri 2007) overestimate su(critical)/vc for the Suncor Tar Island Dyke (cs/cs = 1.014; (N1)60 = 4–15), and underestimate it for the North Dyke of Wachusett Dam (cs/cs = 0.700; (N1)60 = 4–10). The alternative forms of eq. (6), presented by eqs. (D1) and (D2) of the discussion, could be a promising future step towards the estimation of in situ su(critical)/vc . In the following plots (Figs. R1 and R2), I compare these equations with su(critical)/ vc –ucs/ vc pairs for Toyoura sand from direct simple shear (DSS), hollow cylindrical torsional shear (HCTS), triaxial compression shear (TxC), and triaxial extension shear (TxE) tests (Yoshimine
Read full abstract