A PORE PRESSURE MODEL FOR ELASTIC-PLASTIC FINITE-ELEMENT ANALYSIS
Abstract
Existing pore pressure models are limited either to the cases where the pore pressure component produced by the shearing stress is neglected (usually restricted to static analysis) or where the pore pressure component produced by the mean confining stress is disregarded (usually restricted to cyclic loading, as in earthquake and offshore engineering). A pore pressure model including the simultaneous effects of the shear and mean confining stresses is presented in this dissertation. The model is simple and applicable to static as well as cyclic loading. The conceptual relationship between strains and pore pressure is recognized in soil mechanics. However, it is difficult to determine the strains by FEM with the necessary accuracy to ensure satisfactory pore pressure predictions. The proposed model assumes that the influence of the strains can be satisfactorily represented by path dependent calibrating functions to be coupled to an elastic-plastic procedure, allowing the pore pressure increment to be calculated before the iterative elastic-plastic routine is activated. Besides the increased accuracy in the pore pressure predictions, the proposed procedure allows an overall improvement in the modeled effective stress path during undrained analysis, with a corresponding improvement in the strain predictions. The pore pressure model was incorporated into the Purdue version of the cap model. In order to improve some of the limitations of this plasticity model, the concepts of "sub-yielding hardening" and "virtual pore pressure" are introduced. These concepts allow to extend the model applicability to any direction of stress path in 2-D space (axisymmetric and plane strain formulations), for the situations where the vertical stress is larger than the horizontal stress. Also included in the model are some peculiar loading paths that can generate negative pore pressure as in extension loading. The conceptual effect of the pore pressure component produced by the shearing stresses is investigated in this work. Comparisons of the proposed pore pressure model to the common practice where the pore pressure component due to shear is neglected are presented. It is shown that small errors in the pore pressure predictions may cause a pronounced deviation in the modeled effective stress path. This effect is usually not evident when comparing only the modeled stress-strain curves to experimental results.
Degree
Ph.D.
Subject Area
Civil engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.