Analysis of shear banding in true triaxial tests on sand

A series of true triaxial tests have been performed on rectangular prismatic specimens of Santa Monica Beach sand at three different relative densities to study the occurrence of failure, mechanisms that create failure, and soil behavior in the vicinity of failure. One mechanism is smooth peak failure, in which the soil continues to behave as a continuum with uniform strains, and smooth peak failure is followed by strain softening. Another mechanism is shear banding, whose occurrence in the plastic hardening regime limits the strength of the soil. Presented here are analyses based on theoretical conditions for localization and subsequent shear banding and on the results of the true triaxial tests. Thus, the strength increases as b [=(sigma (2) - sigma (3))/(sigma (1) - sigma (3))] increases from 0 to about 0.18, remains almost constant until b reaches 0.85, and then decreases slightly at b = 1.0. Shear banding initiates in the hardening regime for b-values of 0.18-0.85. Thus, peak failure is caused by shear banding in this middle range of b-values, and a smooth, continuous 3D failure surface is therefore not generally obtained for soils. The experimental results agree with the theoretical conditions for the occurrence of shear banding and its consequent effect on the 3D failure stress states for soil.