Development of a Simple Method for Record Selection to Assess the Seismic Risk Index

From the various intensity measures that may be applied to evaluation of the seismic risk of structures, the acceleration response spectrum, Sa(T), is the most famous. As a key assumption in usual risk assessment procedures, like as PEER methodology, the structural response depends only upon the applied intensity measures, and not on any other properties of the ground motion. This required condition has termed “sufficiency” of used intensity measure. The limited “sufficiency” of Sa(T) has been emphasized in the recent researches and as a result, different methods have been proposed to modify the structural response analysis. In this paper, the problem has been re-defined and then the recent studies have been surveyed. This paper is mainly focused on the spectral shape concern. It has been discussed how the spectral shape of a ground motion affects the structural nonlinear response. Epsilon, as a well known seismological parameter is introduced as a convenient indicator of spectral shape. Epsilon as an indicator on the spectral shape has significant influence in the structural collapse risk assessment. The epsilon is defined as a measure of the difference between the spectral acceleration of a record and the mean value obtained from a ground motion attenuation model for a given period. As a direct approach for the consideration of the spectral shape in the record selection, a target ε value, associated with a selected hazard level, is first obtained from the hazard disaggregation procedure, and then records with a closer epsilon value to the target value can be chosen. The major challenge in considering the spectral shape for the selection of records lies in the finding of different sets of ground motion records for each level of hazard for calculation of the MAF of a limit-state for a given structure. Due to the dependence of epsilon on period, it may not be practical to select different specific ground motion sets for any specified period (T1) corresponding to a given site with a particular hazard level. A simple alternative has been proposed in the ATC63 project which could be used instead of the direct selection approach. In this approach, a general set of ground motion records could be used for assessment of the collapse fragility of any structure, without considering the spectral shape of the records. Theresulting mean collapse capacity can then be adjusted to meet the hazard-related target epsilon value. The major challenge of this method is that the objective structure shall be necessarily analyzed via a huge number of ground motions which is unfavorably time consuming task. Another more straight-forward approach is defined in this paper for the mentioned issue. In this method, an efficient range of epsilons is proposed to select a limited number of ground motions for collapse risk assessment of the objective structure. The proposed efficient range depends on the structural major parameters; period and ductility as well as on the seismicity level of the site. The initial results confirm the validity of this simple approach.