Dynamic soil-structure interaction using disturbed state concept and artificial neural networks for parameter evaluation
AdvisorDesai, Chandrakant S.
MetadataShow full item record
PublisherThe University of Arizona.
RightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
AbstractInteraction between the superstructure and foundation depends on the behavior of soil supporting the foundation. To study the behavior of interfaces, it is necessary to characterize the behavior at the interface, model constitutive relationships mathematically, and incorporate the model together with the governing equations of mechanics into numerical procedures such as the finite element method. Such an approach then can be used for solving complex problems that involve dynamic loading, nonlinear material behavior, and the presence of water, leading to saturated interfaces. In this dissertation, a general model, called the Disturbed State Concept constitutive model has been developed to model saturated Ottawa sand-Concrete interface and saturated Nevada sand. In the DSC, the material is assumed to transform continuously from the relative intact state to the fully adjusted state under loading. Hence the observed response of the material is expressed in terms of response of relatively intact and fully adjusted states. The DSC model is a unified approach and allows for elastic and plastic strains, damage, and softening and stiffening. The model parameters for saturated Ottawa sand-Concrete interface and saturated Nevada sand are evaluated using data from laboratory tests and are used for the verification of DSC model. The model predictions showed satisfactory correlation with the test results. In this dissertation, a new program based on concept of neural computing is developed to facilitate determination of interface parameters when no test data is available. The back propagation training algorithm with bias nodes is used to train the network. The program is developed in FORTRAN language using Microsoft Developer Studio. The reason for selecting FORTRAN as a programming language to develop Biased Artificial Neural Network (BANN) simulator is due to its proficiency in number crunching operations which is the core requirement of the ANN. A nonlinear dynamic finite element program (DSC-DYN2D) based on the DSC model is used to solve two problems, a centrifuge test and an axially loaded pile involving interface behavior. Overall, it can be stated that the DSC model allows realistic simulation of complex dynamic soil-structure interaction problems, and is capable of characterizing behavior of saturated interfaces involving liquefaction under dynamic and earthquake loading.
Degree ProgramGraduate College
Civil Engineering and Engineering Mechanics