Reliability-based evaluation of load factors using unzipping and Bayesian methods

In this study, various sources of uncertainty in demand and capacity, and theirdirect and indirect dependencies according to their signicant function in performancebased earthquake engineering (pbee), were incorporated proposing a practically newsimple procedure. this procedure distinguishes and evaluates the collapse probability ofelementary mechanisms, as well as computes their dependencies, utilizing commonly useddeterministic computer software through a full probabilistic methodology using -unzippingand bayesian methods. this procedure has been benchmarked for a typical 2d reinforcedconcrete moment frame, and the reliability index of the structure for the most probablecollapse mechanism was obtained within a reliability-based framework. moreover, thecoecients of dead, live and earthquake loads for a aci load combination were reevaluatedutilizing -unzipping and bayesian methods through the proposed procedure, and werecompared with the code's coecients. good compatibility between obtained coecientsand those of the code ones was perceived, and the probability of collapse, which wasimplicitly considered in the code, was revealed. it has been concluded that the coecientsextracted from the bayesian method are closer to those of the code compared to thosefrom the -unzipping method. however, as it requires less primary computational eortand due to slight dierences between the results, -unzipping is usually more preferred byevaluators.