| Popis: |
Soil-Structure Interaction (SSI) procedures for performance-based seismic design of building structures have been in existence in design guidelines and provisions for decades. However, several issues still remain regarding the application of these procedures to inelastic multi-storey buildings. Three main issues are identified and investigated in this research. Firstly, the gap between code-specified design response spectra and base shear demands of inelastic flexible-base multi-storey buildings is bridged by introducing a strength reduction factor RF and a Multi-Degree-Of-Freedom (MDOF) modification factor RM. The strength reduction factor RF, derived based on the combined (and similar) effects of SSI and structural yielding, allows base shear demands of a flexible-base yielding Single-Degree-Of-Freedom (SDOF) structure to be calculated directly from code design response spectra. The MDOF modification factor RM links base shear demand of a MDOF structure to that of its SDOF counterpart. Secondly, the effect of frequency content of ground motions on elastic and inelastic flexible-base buildings located on very soft soil profiles is examined. Results showed that normalising the equivalent period of a SSI system Tssi by the corresponding predominant periods resulted in more rational spectra for seismic design purposes. In the elastic response spectra, Tssi is normalised by the spectrum predominant period TP corresponding to the peak ordinate of a 5% damped elastic acceleration spectrum, while for nonlinear structures Tssi should be normalised by the predominant period of the ground motion, Tg, at which the relative velocity spectrum reaches its maximum value. It is shown that an actual SSI system can be replaced by an equivalent fixed-base SDOF (EFSDOF) oscillator having a natural period of Tssi, a viscous damping ratio xissi and a global ductility ratio of mussi. The EFSDOF oscillator performed well for linear systems while, in general, overestimated ductility reduction factor Rmu of SSI systems with high initial damping ratio, which consequently led to an underestimation of inelastic displacement ratio Cmu. The two issues stated above were addressed by results of a large number of response history analyses performed using a simplified SSI model where the foundation response was assumed to be linearly elastic and frequency-dependent. The soil-foundation model, developed on the basis of the cone theory, has been verified to be a reliable tool for simulating dynamic soil-foundation interaction. Finally, in order to take into account foundation nonlinearity in preliminary seismic design of building structures, a simplified nonlinear sway-rocking model was developed. The proposed model is intended to capture the nonlinear load-displacement response of shallow foundations during strong earthquake events where foundation bearing capacity is fully mobilised. Emphasis is given to heavily-loaded structures resting on a saturated clay half-space. The variation of soil stiffness and strength with depth, referred to as soil non-homogeneity, is considered in the model. Although independent springs are utilised for each of the swaying and rocking motions, coupling between these motions is taken into account by expressing the load-displacement relations as functions of the factor of safety against vertical bearing capacity failure (FSV) and the moment-to-shear ratio (M/H). The simplified model is calibrated and validated against results from a series of static push-over and dynamic analyses performed using a more rigorous finite-difference numerical model. Despite some limitations of the current implementation, the concept of this model gives engineers more degrees of freedom in defining their own model components, providing a good balance between simplicity, flexibility and accuracy. |
