高科技廠庫板牆系統耐震性能與破壞預測

Seismic Behavior and Damage Prediction of High-tech Fab Cleanroom Partition Walls

10.53106/101632122024060128007

姚昭智(George C. Yao)；葉晏熙(Yen-Hsi Yeh)；陳彥丰(Yen-Feng Chen)

庫板隔間牆系統 ； 高科技廠房 ； 有限元素法 ； 足尺振動台試驗 ； Cleanroom Partition Wall ； High-Tech Fab ； Finite Element Analysis ； Shake-Table Experiment

建築學報

128_S期：技術專刊（2024 / 06 / 30）

1 - 16

繁體中文;英文

台灣高科技廠房無塵室中，普遍以庫板牆系統安裝於懸吊式天花板下方作為空間分割。然而於2016年台南美濃地震後，發現庫板隔間牆系統地震時有板體往面外方向脫軌，使高科技廠房無塵室失效停擺，造成巨大的經濟損失的現象。本論文透過單向靜態試驗，了解不同設置下各位移階段中庫版牆的強度與破壞機制。根據靜態實驗結果，以ANSYS有限元素法模型建立數值模擬，模擬彈性階段強度與預測第一挫曲點發生值。動態分析部分，基於靜態分析所得有限元素法模型，並與2021於國震進行的足尺振動台試驗所得數據做為比對基礎，驗證動態模型的準確性。最後基於此動態模型，提供基於天花板振動頻率與加速度的破壞預測線。

Cleanroom partition walls play a fundamental role in Taiwan's advanced cleanrooms. However, during the Meinong earthquake that struck southern Taiwan on 6th February 2016, these partition walls sustained significant damage, leading to the shutdown of the cleanroom and production line, causing severe financial losses. This study focuses on investigating the out-of-plane performance of cleanroom partition walls. To assess the out-of-plane performance of different screw layout partition walls, we conducted a single-direction push-over experiment. Through this experiment, we collected valuable data on the partition walls' behavior in response to lateral forces. Subsequently, we developed a finite element analysis (FEA) model using ANSYS to simulate the buckling deformation shape and force of the partition walls. For dynamic analysis, we observed that the partition walls retained their elastic behavior after failure. Consequently, we employed the mode superposition (MSUP) analysis method for the transient model. To validate the accuracy of the transient model, we compared its output with data obtained from a full-scale partition wall shake table experiment conducted in 2021. Utilizing the validated model, we derived a ceiling acceleration and frequency failure prediction curve specific to cleanroom partition walls. This curve provides crucial insights into the performance limitations of cleanroom partition walls under seismic events, aiding in designing safer and more resilient cleanroom facilities. In conclusion, this study addresses the critical issue of cleanroom partition wall performance during seismic events. By combining experimental and numerical approaches, we contribute to enhancing the understanding of the partition walls' behavior, which is crucial for mitigating potential damage and financial losses in cleanroom facilities during earthquakes.