應用低價位的微機電系統地震儀在建築物健康監測的實驗研究

Abstract

國立臺灣大學的地震預警團隊和一家科技公司合作開發出名為Palert的微機電系統加速度計作為地震預警用途，其最大的優點為價格相對低廉。另一方面，當地震來發生時，結構物的健康監測也十分重要，但由於硬體設備的價錢昂貴，使得結構物健康監測的發展受到了限制。因此，若將Palert應用於結構物健康監測，或許可以使結構物健康監測普及化。此研究作為以此目的的前期研究，以驗證Palert用來做結構物健康監測的可行性。 為了要診斷結構物的狀況，我們應用兩種方法。第一種方法是根據Nakata et al.的方法去評估國家地震工程研究中心裡振動台實驗室的鋼架構的基頻。結果顯示用Palert所評估的基頻和國家地震工程研究中心的高性能加速計和資料擷取系統所評估出的基頻相似，並且其頻率符合結構物的受損狀態。第二個方法是應用易損性曲線在相同的目標結構物上。將每一層樓在實驗中測到的最大樓層位移比做為易損性曲線的輸入參數，我們可以得到各個樓層發生不同等級損傷的機率。結果顯示Palert偵測損傷位置的能力較差，但是作為整體結構物的損傷存在與否是可接受的。 此研究證實了Palert用來做結構物健康監測的可能性，然而最需要改進的是Palert記錄的時間不太準確的問題，並且未來仍需更多的研究。

The Earthquake Early Warning (EEW) research group at National Taiwan University (NTU) and a technology company have been developing a Micro Electro Mechanical Systems (MEMS) type of accelerometer named Palert designed for EEW purpose. The main advantage of Palert is that it is a relatively low-cost seismometer. On the other hand, due to the high price of commercial hardware of Structural Health Monitoring (SHM) systems, the application of SHM to buildings is limited. This study serves as a pre-study of Palert for SHM application. In order to monitor the health of the building, we apply two different methods. The first method proposed by Nakata et al. is used to estimate fundamental frequency of a steel building in the laboratory of the National Center for Research on Earthquake Engineering (NCREE). The results show that the fundamental frequencies estimated using the Palert are quite comparable to the ones estimated using the high-performance accelerometers and data acquisition system. The second method is to apply fragility curve to the same buildings in NCREE. With the peak inter-story drift ratios of each floor during the experiments as the inputs of fragility curve, we can get the probabilities of different damage levels of each floor. The results of the Palert show poor indication of damage locations but acceptable detection of damage existence. The Palert illustrates the possibility to be used to monitor the health of a building. If we want to do further SHM studies by Palert, then the time-shift problem may be the most important thing to correct.