Reliability Analysis of Flexural Steel Frames by Using the Weighted Simulation Method and Radial Basis Function Interpolation

Flexural steel frames are one of the most commonly used building’s structural systems. The safety of this structural system could be affected by the uncertainties such as imperfectness in construction (especially for imperfect rigid connection), material properties and effect of unexpected loads. By considering these uncertainties and combination them with the probability theories and structural engineering, structural reliability investigates the safety level of structure. In this paper by employing the weighted simulation method, reliability and reliability-based design optimization of the flexural steel frames is investigated. For this purpose, by simultaneously using the firefly algorithm and the Radial Basis Function Meta-Model, a hybrid algorithm to approximate the failure probability and design point is proposed. The efficiency of proposed framework firstly evaluated by solving several numerical examples and then the safety and parameters sensitivity of a three-story steel frame under an artificial acceleration is investigated. The results of the safety evaluation presented good agreement with the accurate results obtained by the Monte Carlo simulation. Results of the safety evaluation and sensitivity analysis show that the imperfectness in connections rigidity highly affects the failure probability of the frame.