LAUSR

Abstract The results of a series of laboratory tests on unimproved and cement-improved specimens of two clays are presented, and the ability of a bounding surface elastoplastic constitutive model to predict the observed behavior is investigated. The results of the oedometer, triaxial compression, extension, and cyclic shear tests demonstrated that the unimproved soil behavior is similar to that of soft clays. Cement-improved specimens exhibited peak/residual behavior and dilation, as well as higher strength and stiffness over unimproved samples in triaxial compression. Two methods of accounting for the artificial overconsolidation effect created by cement improvement are detailed. The apparent preconsolidation pressure method is considerably easier to use, but the fitted OCR method gave better results over varied levels of confining stresses. While the bounding surface model predicted the monotonic behavior of unimproved soil very well, the predictions made for cyclic behavior and for improved soils were only of limited success.