Yielding and failure of cement treated soil

Abstract

This study deals with the development of a constitutive framework for cement-treated Singapore marine clay. Element testing was conducted for a wide range of mix proportions and different curing conditions to shed light on the constitutive behavior of the cement-treated clay. Based on the experimental results, common trends were identified which forms the basis of a framework of behavior of this material. Finally, a constitutive framework was postulated for the primary yielding and early post-yield behavior of cement-treated marine clay. Coop and Atkinson¿s method was introduced to define the primary yield stress and the primary yield locus. The results showed that a reasonably consistent primary yield locus of cement-treated marine clay can be obtained. The results of isotropic compression tests showed that the post-yield compression index seems to be independent on cement content and curing period. The post-yield compression index is also independent on total water content and curing stress. However, the isotropic primary yielding stress increases with the increase in cement content, curing load and curing time as well as decrease in total water content. The experiment results showed that behaviors under triaxial compression are consistent with Chin¿s (2006) study. The results also showed that increasing the cement content, curing stress and curing period or decreasing the total water content all has a similar effect in increasing the peak strength of the specimens. The post-yield behavior of the cement-treated soil appears to be influenced by densification effects as well as breakage of inter-aggregate bonds. The yield locus of the cement-treated marine clay evolves into a shape which is well-fitted by an ellipse. Experiment results suggest that excess pore pressure and stress ratio may be better indicators for shear band initiation than the deviator stress itself. The short specimens with enlarged low-friction end caps showed significantly slower rate of strain softening and a more uniform post-peak behavior than that of conventional specimens showing a single shear band. The short specimens probably reached a critical state at a shear strain of about 20%, with a friction coefficient much higher than that of conventional long specimen. The experiment results showed that the isotropic compression curve of the remoulded treated marine clay depends mainly on the cement content and lies between the compression curves of the untreated marine clay and the corresponding intact treated marine clay. The undrained stress path and deviator stress-strain of remoulded cement-treated marine clay is similar to that of remoulded untreated marine clay. The excess pore pressure changes very small after peak point. Base on the experiment study, the remoulded cement-treated marine clay may be considered as a reference ¿remoulded¿ state towards which the microstructure of the cement-treated clay will evolve with continual shearing. The ultimate state of remoulded cement-treated specimen seems to be only dependent on the cement content, thereby making it a convenient reference state. Based on the observation that the cement-treated soils do have a structure surface (or primary yield surface) but that this yield surface can change shape and size during shearing, a new theoretical approach was attempted. In the theoretical framework, a new yield function was derived by introducing true cohesion parameter into a modified form of the modified Cam Clay energy equation. The proposed yield function can fit well the observed primary yield locus and evolution of yield locus for cement-treated marine clay specimens. The parameters of the proposed model can be obtained from conventional test.