| In order to obtain complete probabilistic properties of site-specific design parameters for future seismic events, a computational framework has been used for simulating seismic motion propagation and quantifying influences of associated uncertainties.;Regional ground motions are first simulated through the Stochastic Point Source Model using sets of geophysical and seismological parameters. With an analytical seismic wave transfer function process and the Domain Reduction Method, effective seismic forces for a project site are obtained at the interface of the weathered bedrock and an uncertain soil profile. Marginal cumulative distribution functions and statistical parameters of simulated seismic loads are then extracted using a nonstationary scalar-valued random process with the Polynomial Chaos Decomposition Method. After, synthetic seismic motions with probabilistic properties are propagated through an uncertain 1-D soil column using the Stochastic Elastic-Plastic Finite Element Method to obtain full spatial and temporal probability distribution functions and fragility curves of displacements for a specific interested local feature given a seismic event. In addition, by inducing different levels of uncertainty from elements of seismic motion propagation, illustrations are presented to quantify the relative influence of the associated uncertainties.;These efforts are intended to contribute to probabilistic seismic hazard analyses and provide a supporting tool to allow practicing engineers, regulators, insurers, policy makers and owners to develop risk-based performance analyses using rational mechanics and probability theory with higher accuracy and low Kolmogorov Uncertainty. |
