Level Ground Soil‐Liquefaction Analysis Using in Situ Properties: II

Abstract

Current methods for level ground soil‐liquefaction analysis involve the use of a stress ratio required to cause liquefaction obtained from laboratory testing of small element samples. Limited field evidence and considerable centrifuge model studies have shown that the soil liquefaction is a boundary value problem. There are numerous computer programs available that treat soil liquefaction as a boundary value problem. However, these analysis procedures do not take into consideration the properties of soil at low effective stresses, and most of the input parameters for these procedures have to be determined from test results of small “undisturbed” samples. These procedures have not been verified extensively. A method is presented for level ground soil‐liquefaction analysis, which is treated as a boundary value problem. This method uses a semiempirical, one‐dimensional, elastoplastic constitutive model and finite difference method to solve the governing differential equation for the prediction of pore pressure generation, dissipation, and settlement characteristics during and after dynamic excitation. A compressibility function incorporating effects at low effective stresses is included in this method, which is verified by centrifuge model tests. This procedure utilizes input properties representative of field conditions determined by a nondestructive electrical method. The proposed method for analyzing the generation and redistribution of pore pressure, therefore, provides a means for predicting liquefaction of level ground sites—a boundary value problem—based on the results of in situ testing.