Nonlinear system identification on shallow foundation using Extended Kalman Filter

Abstract

This study employs system identification using the Extended Kalman Filter to investigate variations in the stiffness and damping of shallow foundations during earthquakes. System identification results showed that the elastic stiffness of different foundations was significantly smaller than specifications proposed by FEMA 356 for the S-E site class. As the earthquake load increased, a partial uplift of the foundation occurred. Following this uplift, the time domain inelastic stiffness decreased due to variations in contact area between the foundation and sub-soil. The inelastic stiffness at the maximum response was less than the elastic stiffness, according to the effective peak ground acceleration (EPGA) and the contact area ratio. After uplift in the foundation, the EPGA increased, the contact area ratio decreased, and the damping ratio increased by up to 20%. On the basis of these system identification results, we determined relationships between elastic stiffness and the ratio of bearing stress demand to the soil-foundation system capacity.