基於地震境況之地表加速度歷時縮放方法研究

Abstract

美國ATC-58計畫於2012年提出房屋結構新一代耐震性能評估方法，將地震強度、材料性質、結構反應及損壞情況等參數，以機率式的概念引入評估方法，並以修繕金額、修繕時間及傷亡人數等業主較為關心瞭解之參數作為性能指標，以便工程師與業主能有效的溝通，協助業主做出適當決策。 上述新一代耐震性能評估方法共分基於地震強度、基於地震境況與基於地震危害度三種不同形式之評估，皆涉及縮放地表加速度歷時進行非線性動力分析，以求得結構反應分佈，進而決定房屋內各結構與非結構元件的損害情況以計算性能指標。其中B型基於地震境況之耐震性能評估，目的在推估特定地震境況(即某一地震規模與斷層距工址距離之組合)下房屋之性能指標，所使用之地表加速度歷時須能合理地反映此特定地震境況之加速度反應譜值之分佈(包括譜值大小、離散程度與適當譜形)，否則將獲得與實際情況差異甚大的評估結果，將導致業主及工程師決策錯誤。本文旨在建議一適用於此型之耐震性能評估的地震歷時縮放方法。 本研究各選取50筆地震紀錄模擬三種選定之地震境況，以一棟十五層樓與一棟五層樓特殊抗彎矩鋼構架為例，討論三種地表加速度歷時之縮放方法，包含Conditional Spectrum Method、Unconditional Spectrum Method、Distribution Scaling Method，透過非線性動力分析求得兩範例構架最大層間位移、最大加速度、平均樓層加速度反應譜等結構反應，推估其結構反應在真實值正(負)20%內之信心水準，以探討各縮放方法對結構反應分佈之影響，進而建議適用於B型耐震性能評估之縮放方法。

The ATC-58 project in the United States developed and published next-generation tools and procedures for seismic performance assessment of buildings in 2012. In the ATC-58 procedures, damage is measured in terms of direct economic loss, indirect economic loss and casualties rather than by building component deformations and accelerations. Uncertainty and randomness will be captured in every step of the performance assessment process, including characterization of earthquake hazard, simulation of building response, damage assessment, and loss computation. The ATC-58 procedures offer three different types of assessments, namely, intensity-, scenario- and time-based assessments. All three involve the use of response-history analysis. The objective of this study is to develop recommendations for the selection and scaling of ground motions for scenario-based assessment. The scenario-based assessment aims at predicting the loss for a specific scenario (representing by a combination of earthquake magnitude and distance). The ground motions scaled for the assessment should reflect the distribution of spectral acceleration for the specific scenario. In this study, a series of nonlinear response-history analyses were performed for a 15-story and a 5-story sample moment resisting frames subjected to ground motions scaled for three sample earthquake scenarios determined based on the aggregation data for the site of the sample buildings.Sets of ground motions were developed for each building and each earthquake scenario using three scaling methods, namely, Conditional Spectrum Method, Unconditional Spectrum Method and Distribution Scaling Method. The impact of 1) scaling procedures, 2) number of ground motions, and 3) ground-motion database on structure responses, including peak inter-story drift, peak floor acceleration, and floor spectral acceleration, was studied. Recommendations for ground-motion scaling procedures for ATC-58 scenario-based assessment were developed.Selections of 30 ground motions from scenario database for ground-motion scaling procedures(Unconditional Spectrum Method and Distribution Scaling Method) is recommended in this study.