Validating IM-based methods for probabilistic seismic performance assessment with higher-level non-conditional simulation

Probabilistic seismic performance assessment of structures can be carried out with different methods. One possible classification of the latter is into two main classes: IM-based methods (with IM being the well-known acronym for Intensity Measure, a scalar or vector quantity expressing the local intensity of ground motion at a site) and full-blown simulation methods. Two different probabilistic representations of the ground motion are employed in the two classes. In the former case, real recorded motions are used to determined the fragility curve, or more generally the structural response given the IM. Then an hazard curve for the chosen IM is used with the fragility curve to determine the mean annual frequency (MAF) of exceedance of any performance level of interest. In the latter case, a synthetic ground motion model is employed to produce time-series starting from basic (random) macro-seismic parameters that are part of the simulation, such as event magnitude and location. Whether the two classes of approaches lead to the same results is a question first asked in Jalayer and Beck (2008), where the Atkinson and Silva synthetic ground motion model and the subset simulation methods where employed. This paper presents a comparison along the same lines but using the importance sampling with K-means clustering method (Jayaram and Baker 2010) for the analysis and, most importantly, the parametrized random process model by Rezaeian and Der Kiureghian (2010) to generate ground motion time series. This latter model claims to reproduce not only the mean of real ground motions but also their natural variability. This is instrumental in evaluating the correct MAF of structural response, which depends on the total variability (including that of ground motion). Indeed, the lower than natural variability exhibited by the Atkinson and Silva model was to reason for Jalayer and Beck (2008) to introduce a "correction term" that inflated ground motion variability. F. Jalayer and J. L. Beck ‘Effects of two alternative representations of ground-motion uncertainty on probabilistic seismic demand assessment of structures’, Earthquake Engng Struct. Dyn. 2008; 37:61–79 Jayaram N and Baker J W (2010), ‘Efficient sampling and data reduction techniques for probabilistic seismic lifelines assessment’, Earthquake Engineering and Structural Dynamics, 39, 1109-1131. S. Rezaeian and A. Der Kiureghian ‘Simulation of synthetic ground motions for specified earthquake and site characteristics’ Earthquake Engng Struct. Dyn. 2010; 39:1155–1180