CONSEQUENCE MODELS FOR SEISMIC RISK

PREDICTION

Tehran, Iran

This paper presents new probabilistic models intended for use with reliability methods for computing seismic risk on the component scale. In the present context, risk is defined as the probability of attaining or exceeding a specific level of monetary loss. The models are classified as explicit story-specific demand model and consequence models. The demand model predict maximum drift at the level of each story of low-rise to mid-rise steel moment resisting frames (SMRFs) and, consequence models estimate repair cost of some fragility groups in a building. A linear function of spectral acceleration at fundamental period in logarithmic space is considered for demand model. The demand model is coupled with a set of relations to explicitly estimate unknown statistical characteristics of the probabilistic demand model parameters. The consequence models are also formulated as a polynomial function respect to Engineering Demand Parameter in logarithmic space. Bayesian regression technique is employed to determine probability distribution of the consequence models parameters. Finally, as an application of the proposed models, seismic loss curve of an example building is developed for earthquake intensities.