Effect of earthquake duration on structural reliability

Modern design methodologies rely on accurate estimates of structural reliability indices for code calibration and design. In general, most methodologies do not explicitly account for earthquake duration when calculating reliability indices. This paper presents a summary of a study whose goal was to provide a basic method to better estimate the effect of earthquake duration on structural reliability using (1) an ultimate strength and a (2) low-cycle structural damage-based limit state function. The damage model developed by Park and Ang was coupled with the theory of order statistics all within a Monte Carlo simulation framework in order to estimate the probability of failure for hypothetical groups of structures. Structures were idealized using elasto-plastic oscillators in order to simplify the analysis and make it general enough to compare with future or existing studies. A response database was used to provide randomness beyond the number of time domain simulations initially performed. Existing suites of ground motion time histories were scaled using linear response spectrum scaling for soil conditions using a well-known attenuation relationship. This study is unique in that it allows variation in the peaks of the highly non-linear structural response without actually performing time domain analyses, which are commonly employed in earthquake engineering. A simple measure was introduced and termed the duration effect factor (DEFβ) and is defined as the slope of a best-fit line for multiple reliability indices plotted against duration. The DEFβ could help provide a means for comparison between earthquake duration studies, which has otherwise been difficult.