|dc.description.abstract||Structure and soil act simultaneously under the application of loads, and their actions are interdependent and integral in nature. The true characteristics of these integral actions are quite difficult to model. The difficulties arise because of the nature and behavior of the soil itself and its integrated action with the structure. The very first complexity about soil is that its constitutive equations are nonlinear. Other than material nonlinearity, it is anisotropic, inelastic and nonhomogeneous. In the past, researchers had tried to model the integral action of the combined soil-structure system through the concept of modulus of subgrade reaction, such as, Winkler's model. Experience has shown that a correct numerical estimation of the subgrade reaction is a major problem. In recent years, many researchers have been using the finite element method to solve soil-structure interaction problems for more reliable and accurate solutions.
This study is aimed at developing a more sophisticated and versatile method which represents the true and real behavior of the soil-structure interaction phenomenon. For this reason, the nonlinear constitutive behavior has been considered for both soil and the structure. Often it is also necessary to consider many soil-structure problems using three-dimensional analyses, and in this research particular attention is given to the anal3'^sis of problems in three dimensions. A hybrid numerical technique has been developed by the combination of the two powerful methods, namely, the finite element method and the boundary element method. The most challenging part of solving an elastoplastic soil-structure interaction problem is the evaluation of its material parameters, such as the elaslic constants and the hardening parameter. A simple and practical technique has been suggested for the evaluation of the material parameters.
Two computer codes have been developed; one for finite element analysis alone and the other one for coupled finite element and boundary element analysis. Successful attempts have been made to solve a problem with real dimensions. The method has been tested on pier-soil interaction problem. The example tested is taken from the instrumented lateral load tests conducted by the Southern California Edison Company. The results have been compared with those of the finite element method alone and with those of the instrumented lateral load tests, and they are found to be quite satisfactory.||