Abstract:
The research findings made in recent years now mean that the prospect of accounting for soil-foundation-structure interaction within seismic design is becoming a viable reality. By examining the cyclic response of a parameterized set of shallow foundations, simulated using a recently developed macro-element model that accounts for rotational-vertical-horizontal motion interaction and which considers coherently possible uplift behaviour, new degradation curves for the stiffness and damping of shallow foundations are developed. The improvements included in these curves with respect to previous proposals are: i) the uplift mechanism, a non-dissipative nonlinear mechanism, is taken into account and ii) the overturning moment and the corresponding simultaneous horizontal load are applied on the footing so that the effect of shear force on the overall response is investigated.
It is found that rotational stiffness degradation is more severe when shear demands are relatively large compared to flexural demands. Moreover, the stiffness degradation becomes more intense as the static factor of safety for centred vertical loads on the foundation reduces, since the response tends to be dominated by hysteretic behaviour in contrast to an increasingly rigid-body rocking response for larger factors of safety. Hysteretic energy dissipation evolution is represented via equivalent viscous damping curves, obtained from quasi-static cyclic analyses.
Finally, the new set of stiffness and damping curves are included for use within the direct displacement-based design framework. By using the improved curves, the bearing capacity of the foundation will be automatically respected since each point of the developed curves will correspond to a solution lying inside or on the ultimate load surface of the foundation system. The benefit of this approach is illustrated through the design of 6-, 8- and 12-storey buildings with and without taking into consideration soil-foundation-structure interaction. Nonlinear dynamic analyses are used to gauge the performance of the design solutions, and it is found that, even though the prediction of foundation rotation demands can be further improved, the direct displacement-based design method provides good control of storey drifts and displacements, suggesting that it could be a valuable procedure for performance-based earthquake engineering in the future.