P-Alert系統於結構物健康監測之評估與應用

Abstract

臺灣位於歐亞板塊及菲律賓海板塊交界，地震頻繁，若強震事件發生時可能導致房屋倒塌、結構物受損等影響，因此結構物健康監測逐漸受到重視。然而，傳統結構物健康監測因為硬體設施和實驗成本高昂，不易全面推廣至民間。為了解決此問題，本研究採用國立臺灣大學P-Alert地震預警系統，透過P-Alert儀器成本較低的優勢，評估其地震儀在結構物健康監測的可行性。過去研究中，透過振動台實驗已證實P-Alert可利用易損性曲線分析結構物損傷程度，並在分析層間位移時需使用互相關分析進行時間校正。本研究進一步測試，P-Alert與高精度加速度儀Setra在層間位移與模態振形(Mode shape)上的表現，結果顯示P-Alert識別出來的結果與Setra相符。 此外，本研究亦分析不同類型建物，包括P-Alert單站型結構物、P-Alert多站型結構物以及中央氣象署結構物陣列。並透過不同樓層對地面測站反卷積，求得結構物基本模態。在預估層間位移方面，使用雙重積分法將加速度資料轉換成位移資料，計算出各樓層的層間位移角。在強度測量(Intensity Measure)的參數分析中，則採用最大地表加速度、加速度反應譜進行分析。 研究結果顯示，P-Alert單站型結構物中，PGA ≥ 80 gal才能穩定識別結構物基本模態；在P-Alert多站型結構物與中央氣象署結構物陣列分析中則發現：（1）利用結構物基礎模態的變化，可以快速識別損傷程度。（2）在強度參數分析中，P-Alert與中央氣象署結構物陣列所測得最大地表加速度、加速度反應譜皆呈現線性關係。（3）在強度測量參數相同下，結構物振動週期越長，層間位移越大。(4)層間位移角的變化可以快速識別結構物損傷狀態。綜合以上分析結果，P-Alert 不僅在振動台測試結果與高精度加速度儀Setra相符外，透過中央氣象署結構物陣列的比較，顯示其在強度測量上的可靠性。另一方面，基本模態與層間位移角的變化，P-Alert 能夠快速識別結構物損傷程度，展現出其系統在結構物健康監測中的應用潛力。

Taiwan is located at the junction of the Eurasian Plate and the Philippine Sea Plate, where earthquakes are frequent, and strong earthquakes may lead to house collapses, structural damage, and other impacts, so structural health monitoring is gradually being emphasized. However, due to the high cost of hardware facilities and experiments, it is not easy to fully promote traditional structural health monitoring in the community. In order to solve this problem, this study adopted the P-Alert system developed by the National Taiwan University, and evaluated the feasibility of the P-Alert system for structural health monitoring (SHM) by taking advantage of its lower cost. A previous study demonstrated that P-Alert can analyze structural damage by using damage curves, but time correction using cross-correlation is required to analyze inter-story drift. In this study, the performance of P-Alert and the high-precision accelerometer Setra on inter-story drift and mode shape was further tested, and the results showed that P-Alert recognized the same results as Setra. In addition, different types of buildings are analyzed, including P-Alert single station structures, P-Alert multiple station structures, and the Central Weather Administration (CWA) building array. The fundamental mode of the structure is calculated by deconvolution, and the inter-story drift and inter-story drift ratio angle are predicted by double integration method. In the parameter analysis of Intensity measure (IM), the peak ground acceleration (PGA) and response spectral acceleration (Sa) are used. The results showed that in P-Alert single-station structures, PGA needs to be higher than 50 gal to recognize the fundamental mode of the structure stably; in the analysis of P-Alert multiple-station structures and building arrays, the findings include: 1. The frequency of the fundamental mode decreases when the structure is damaged. 2. P-Alert showed a linear relationship with both PGA and Sa (T1) as measured by the building array. 3. With the same IM parameters, the longer the vibration period of the structure, the larger the inter-story drift. 4. The Variety of inter-story drift ratios can immediately recognize the status of structural damage. In summary, P-Alert not only matches the results of the vibration test with the high-precision accelerometer Setra, but also demonstrates the reliability of the strength measurements by comparing the building array of the Central Weather Administration. On the other hand, with the change of basic mode and inter-story drift ratio angle, P-Alert can expeditiously recognize the damage level of the structure, which demonstrates the potential application of its system in structural health monitoring.