相鄰建築地震碰撞機率分析

Abstract

本研究利用隨機振動理論及數值模擬方法，調查依建築規範設計之相鄰建築物發生地震碰撞之機率及避免地震碰撞所需分隔距離之機率分佈。基於頻譜分析法及數值模擬法，本研究分別提出建築物碰撞機率之線彈性理論解法及非線性數值解法。並提出數值例之解析方法及過程。 本研究假設結構為多自由度剪力梁型房屋結構，且結構依據1997年版美國UBC規範設計。非線性分析時，假設結構之力-位移行為模式為彈塑性遲滯行為模式。 本研究結果顯示，相鄰建築物之週期比為建築物碰撞風險之重要影響因子。然而，現行美國UBC規範有關建築間隔之條款並未將此因子納入考慮。另外， UBC規範用以計算建築間隔之方法並未適當考量相鄰建築物之振動相位差對建築物避免碰撞所需分隔距離之影響。因此，UBC規範採用的方法無法準確地估計相鄰建築物避免碰撞所需分隔距離。本研究結果亦顯示，相鄰建築物避免碰撞所需分隔距離之機率分佈為第一型極值機率分佈。從本研究數值例分析結果得知，不同的相鄰建築物具有不同的碰撞風險。換言之，依美國UBC規範設計之相鄰建築物之地震碰撞風險並不均勻。

The emphasis of this study is on the use of random vibration theory and numerical simulation approach to investigate the seismic pounding probability of adjacent buildings separated by a minimum code-specified separation distance and the probability distribution of required separation distance of adjacent buildings to avoid seismic pounding. Proposed herein is an analytical method based on a spectral approach for linear elastic analysis and a numerical simulation approach for nonlinear analysis. Detailed procedures of the analytical method and some numerical examples are presented. In this study, the structures are modeled as a multi-degree-of-freedom shear beam type model and designed according to the 1997 Uniform Building Code (UBC’97). For nonlinear analysis, it is assumed that the behavior of structures exhibits elasto-plastic behavior in the form of a hysteretic restoring force- displacement characteristic. It has been demonstrated that the period ratio of adjacent buildings is an important parameter that affects the pounding risk of adjacent buildings. However, there is no specific consideration for the period ratio in the related seismic pounding provisions of the UBC’97. It is noted that the method used in the UBC’97 provides poor estimates of the required building separation due to improper treatment of the vibration phase of adjacent buildings. Results also reveal that, for two adjacent buildings, the probability distribution of required distance to avoid seismic pounding fit very well with the type I extreme value distribution. From the numerical examples investigated in this study, it is found that significant differences of the pounding probability are observed for the adjacent buildings separated according to the minimum separation distance specified in the UBC’97. In other words, the pounding risks of buildings separated by minimum code-specified separation distance are not uniform.