Distinct element method analysis of true triaxial test with membrane boundary and its application to failure criteria of soil

Several types of true triaxial test apparatus have been used for real materials with a membrane boundary (flexible boundary). However, the friction at the end face of the specimen have influence on the true triaxial stress state. In this study, a true triaxial test with a membrane boundary was reproduced without end-face friction using the distinct element method, and the true triaxial tests with constant Lode angle stress paths were simulated. The obtained stress-strain relationship of specimens is similar to that obtained under the wall-element boundary (rigid boundary) condition. This indicates that the difference in boundary conditions does not significantly affect the macroscopic behavior. On the other hand, the micromechanical properties for the inside of the specimen indicate that the membrane boundary contributes to the uniformity of the boundary force distribution, the void ratio distribution near the boundary, and the uniformity of the void ratio inside the specimen. These results suggest that the present analytical conditions reproduce a more ideal true triaxial test. Furthermore, from the examination of three different failure criteria, it was found that the Matsuoka-Nakai’s criterion could be the most suitable for the critical state under an intermediate magnitude of confining pressure used in this study.