Loading...

Site dependent response estimation by holistic record selection and bagging algorithm

Ghods, B ; Sharif University of Technology | 2023

0 Viewed
  1. Type of Document: Article
  2. DOI: 10.1016/j.jobe.2023.106163
  3. Publisher: Elsevier Ltd , 2023
  4. Abstract:
  5. An efficient two-step method is proposed for site-specific structural response estimation of buildings at multiple sites with different seismicity. In the first step, holistic record sets that cover a range of seismic characteristics are constructed for a few intensity measure levels. Next, using the holistic sets and their associated responses, any required response measure (e.g., structural collapse) is calculated using a bagging algorithm. The proposed method has two advantages to many similar suggested processes in this regard. Because the proposed process works directly with record selection methods and a computational method, it can consider the effect of different ground motion intensity measures and works for all engineering demand parameters. The performance of the proposed method is investigated using two 4 and 8-story steel moment-resisting frames in six cities with different seismicity. The collapse fragility curve, maximum inter-story drift ratio (MIDR), and peak floor acceleration (PFA) are considered for performance assessment. It was observed that when the number of selected records is about twice the common required number of records for response assessment, the proposed method, on average, leads to equal or better efficiency for collapse fragility estimation compared to conventional single site-building analysis. Also, the results indicate an efficiency similar to utilizing about 10–20 ground motion records at each IM level for MIDR and PFA response estimation. © 2023 Elsevier Ltd
  6. Keywords:
  7. Bootstrap aggregating (bagging) algorithm ; Conditional spectrum (CS) ; Fragility curve ; Performance-based earthquake engineering ; Regional seismic risk analysis ; Site-specific record selection
  8. Source: Journal of Building Engineering ; Volume 68 , 2023 ; 23527102 (ISSN)
  9. URL: https://www.sciencedirect.com/science/article/abs/pii/S235271022300342X