Abstract

Reliability analysis assessment of seismic performance for reinforced concrete buildings was investigated in this work. This was performed through the response surface methodology in order to derive explicit expression of the failure function. Two limit states defined in terms of the total building roof displacement and the maximum inter-story drift were considered. The seismic behaviour of the building was examined by using conventional pushover analysis through finite element computations conducted by means of ZeusNL software package. Three random variables characterizing material resistance variations of concrete and reinforcement steel as well as member’s sections dimensions were introduced. A complete factorial design of experiment table having three levels was used to define a finite set of data points where the failure function was evaluated, before using these results to perform identification of the building response surface model via polynomial regression. An application of this procedure was illustrated on a five story building and analysis of reliability in terms of the actual ductility coefficient was performed. Discussion was carried out about the effect on reliability resulting from the distributions of probability modelling the uncertainties affecting parameters and from using the approximate methods: Monte Carlo based sampling analysis and FORM.

Key words: Earthquake, seism, reinforced concrete buildings, pushover, FEM, reliability, FORM, Monte Carlo, response surface.