Modeling Critical Ground-Motion Sequences for Inelastic Structures

The specification of earthquake loads as inputs to engineering structures is a crucial task in earthquake engineering. There are cases, however, where the site under consideration has limited or scarce seismic data, making this process a difficult task. Meanwhile, structural engineers are often concerned with the worst-case scenario that can happen to the structure during its service-life under possible future earthquakes. Repeated ground-motion sequences occurring after short intervals of time, resulting from mainshock-aftershock earthquakes, have been observed in many parts of the world. Such ground motion is capable of creating severe damage in the structure due to accumulation of inelastic deformations from multiple sequences before any structural repair is possible. This paper models ground motions of multiple sequences that produce the maximum damage in the structure. The ground acceleration is represented as Fourier series, with unknown amplitudes and phase angles, modulated by envelope functions. The unknown parameters are optimized to produce the maximum damage in the structure while the ground motion is constrained to the available seismic data at the site. The resulting inverse nonlinear dynamic problem is tackled using optimization techniques, nonlinear time-history analysis and damage indices. Numerical illustrations on modeling critical earthquake sequences for inelastic frame structures are provided. It is shown that critical repeated acceleration sequences produce larger structural damage compared to single critical earthquakes.