Unstable behaviour of model Jamuna micaceous sand

V. N. GEORGIANNOU (2008).Ge´otechnique 58, No. 10, 825–829D. Wanatowski, University of Nottingham andJ. Chu, Nanyang Technological UniversityThe discussers would like to point out that the statement‘The instability observed under fully drained conditions doesnot involve collapse or loss of strength’ made by the authorin the conclusion (c) may not be generally true for tworeasons.(a) A specimen can collapse under a drained condition ifthe test is conducted under a load-controlled loadingmode as shown and discussed in detail by Chu L Lade, 1992, 1993; Sasitharan etal., 1993; Leong et al., 2000; Chu & Leong, 2003a,2003b; Chu et al., 1993, 2003; Chu & Wanatowski,2009). As the author did not specify otherwise, thediscussers assume that the term ‘drained’ is used torefer to a condition where water is allowed to flow inor out of the specimen under a constant back pressurecondition. Under such drained condition, the effectiveconfining stress or the effective stress path can becontrolled by varying the cell pressure. With today’slaboratory testing techniques, it is possible to carry outa special ‘drained’ triaxial test to trace the curvedeffective stress path followed in an undrained test inwhich the specimen has collapsed. Assume this specialtest is conducted under a load-controlled loading mode,would the specimen in this test collapse or not? If itwould, then this collapse is under a drained condition.If it would not, then what are the other factors thatcontrol the collapse in addition to the effective stresses?In the discussed technical note, the author states correctlythat ‘unstable behaviour is associated with the sudden in-crease in the accumulation of plastic strains at an accelerat-ing strain rate’. However, no graph indicating theacceleration in strain rate has been shown in the discussednote. The figures shown in the discussed note indicateyielding behaviour only. It should be pointed out thatalthough yielding is a necessary condition for the instabilityto occur, it is not sufficient, as shown by Chu et al. (2003).This is because yielding means the development of a largestrain for a small change in stress, which does not implythat the specimen will become unstable. For example, thesame yielding points can be obtained for two ‘identical’specimens tested under strain-controlled and stress-controlledmodes. However, only the latter may become unstable, asstudied in detail by Chu & Leong (2001) and Chu &Wanatowski (2009). For the same reason, a comparison of adeformation-controlled undrained compression test (WT1)with a load-controlled drained probing test (WT5–WT7) inFig. 4(b) of the discussed note is not meaningful.REFERENCES