A two-dimensional non-linear static and dynamic response analysis of soil structures

A method of analysis in two-dimensions to predict static and dynamic response of soil structures, including soil-structure interaction has been presented herein. The static and dynamic analyses can be performed in either effective or total stress mode or a combination of both modes. Non-linear stress-strain behaviour of soil has been modelled by using an incrementally elastic approach in which tangent shear modulus and tangent bulk modulus were taken as the two "elastic" parameters. The material response in shear was assumed to be hyperbolic coupled with Masing behaviour during unloading and reloading. Responses to changes in mean normal stress was assumed to be non-linear, elastic and stress dependent. Slip or contact elements have been incorporated in the analysis to represent the interface characteristics between soil and structural elements. The properties of the slip elements were assumed to be elastic, perfectly plastic, with failure at the Interface given by the Mohr-Coulomb failure criterion. In the static analysis proposed here, gravity may be switched on at once for the completed soil structure or the construction sequence can be modelled by layer analysis. The stress-strain conditions determined by the static analysis give the in-situ stress condition before the dynamic analysis. In the dynamic effective stress analysis, the residual porewater pressures are calculated using a modification of the model proposed by Martin, et al. (1975). The parameters, G[sub max] and[sub max] are modified for the effects of residual porewater pressure. The dynamic response study includes the prediction of post earthquake deformations. The predictive capability of the new method of analysis has been verified by comparing the recorded porewater pressure and accelerations of two centrifuged models subjected to simulated earthquakes, to those computed by the new method. This method has also been used to compute response of an offshore drilling island supporting a tanker mounted drilling rig. Results suggest that the common practice of neglecting soil-structure interaction may not be appropriate for islands which support heavy tanker type of structures. At present one-dimensional methods are used for computing the response of these islands. Comparative studies are also reported to asses the validity of this procedure.Applied Science, Faculty ofCivil Engineering, Department ofGraduat