THE SIGNIFICANCE OF THE STRAIN PATH-ANALYSIS IN THE INTERPRETATION OFPIEZOCONE DISSIPATION DATA

When a piezocone is driven into the ground, pore water pressures are s et up as a result of the stress changes. The dissipation of these pore pressures occurs in the subsequent con-solidation process in a manner dependent on the initial stress distribution as web as on the coeffic ient of consolidation. The process of analysis therefore has two disti nct components. The first is to identify the appropriate total stress distribution caused by driving the piezocone, which acts as the initia l condition for the consolidation. The second is to solve the consolid ation problem itself. This approach has been followed by Levadoux & Ba ligh (1986, J; Geotech. Engng, ASCE 112, No. 7, 707-726), Baligh & Lev adoux (1986, J. Geotech. Engng, ASCE 112, No. 7, 727-745) and by Houls by & Teh (1988, Penetration testing 1988 (ed. J. de Ruiter), Balkema, Amsterdam, Vol. 2, pp. 777-783). Each of the two sets of authors uses the same consolidation theory for the second stage of the analysis, an d a strain path method to identify the initial stress conditions. Howe ver, different assumptions about soil behaviour and the relevant soil parameters are made in carrying out the strain path analyses. A real s ituation mill differ from the theoretical one both by differences from the model assumed, and in the choice of numerical values of the relev ant parameters. For the two approaches to the piezocone analysis. Houl sby & Teh (1988) used an elastic-plastic model, in which the value of the rigidity index, G/s(u), may be specified in the analysis, while Le vadoux & Baligh's solution, with a similar theoretical basis, used par ameters specifically appropriate for Boston Blue Clay. This paper disc usses the question of an appropriate choice of numerical values of the soil parameters and reports a comparison between the two theoretical predictions and held measurements from two different piezocones deploy ed in Sarapui soft clay, Brazil. The values of the coefficient of cons olidation c(h) calculated from the field data by the two methods are t hen compared with high quality laboratory tests on the Sarapui clay. I t is shown that both theoretical predictions provide good agreement fo r the shape of the dissipation curve, but that the predicted C-h value s are different, and that thought should indeed be given to the use of an appropriate value of G/s(u).