9. Conclusion
In order to study the seismic performance of structures more accurately, several ground motions of different intensity level should be considered. To this end, incremental dynamic analysis is an extremely useful method which can become a valuable additional tool for seismic engineering. However, it really takes a lot of time to be done in order to consider the record to record variability and massive study cases. One useful strategy to cope with this problem is earthquake records truncating. In this regard, this study has shown the effects of record truncation on the fragility curves and mean annual frequencies of different RC frames. The fragility curves and MAFs of IO and CP limit states are computed for two different damage indices of interstory drift and the modified Park & Ang. While maximum inter-story drift is based on deformation only, the modified Park & Ang is based on the combination of maximum deformation response and hysteretic energy dissipation. It has been concluded that truncating has little effect on outputs, in the case of normal and mid-rise frames, as long as deformation-based damage indices, like interstory drift, are chosen. On the other hand, investigating seismic performance of structures by energy-based or combination indices are more dependent on the duration of records and truncation may produce more errors. Additionally, high-rise structures with long periods are more sensitive to such changes. Records may be truncated after due consideration of these factors. Based on the computed errors in this study, the truncation effect for IO limit state is less than CP limit state regardless of damage index. Furthermore, this study has shown that more than half the execution time can be saved by truncating the records.
