Uncertainty analysis using fuzzy randomness method towards development of fragility curves formoment-resisting steel structures

Seismic fragility analysis is one of the main steps of consequence basedearthquake engineering process. accurate uncertainties modeling involved in this methodology,aects the nal results of seismic fragility analysis and hence assessment of decisionvariables which are the nal products of performance-based seismic analysis. one aspectof such eorts is to incorporate the sources of uncertainties associated with various factorscontrolling seismic loads on the buildings as well as structural responses to such excitations.probabilistic approach is usually used to model quantitative sources of such uncertainties,however, there are other factors with descriptive nature which probabilistic approach maynot well incorporate them. in this paper a fuzzy randomness approach is used to modelepistemic uncertainties as an alternative to the conventional probabilistic method. theapproach is used to model those uncertainties which have not been addressed by theothers, in particular the denition of the collapse limit state. to illustrate the eciencyof the proposed approach, fragility curves for a sample moment-resisting steel frame aredeveloped. the results demonstrate the superiority of fuzzy solution in comparison withexcising probabilistic methods to incorporate epistemic uncertainty in view of much lesscomputational eort.