本研究乃針對軌道列車之車輪踏面扁平問題進行診斷評估，以建立一套基於行車安全且不中斷營運之長期監控系統。當列車通過監測區段時，利用安裝於鋼軌底板上之陣列式光纖應變計(Array-type FBG)，擷取鋼軌應變歷時資料，並透過現地監控電腦快速進行數據分析。而本研究在數據分析上採用兩種分析方法—特徵線法(Feature Line Method)及衝擊比值指標(Impact Ratio Index)。其概念在於帶有車輪扁平之擦傷車輪(平輪)與軌道持續滾動接觸時，將於擦傷部位產生車輪滯空之現象，使得應變量下降，又因滯空後之撞擊而造成應變量上升，利用此一應變轉折之特徵圖形作為判別機制，形成損傷特徵線之可視化圖形作為圖像判別；並透過剪力法計算因平輪所造成之衝擊力以及健康車輪下之平均輪重，並將該衝擊力與平均輪重之比值作為平輪之損傷指標(IR index)。 平輪損傷警戒層級之界定辦法，在本研究則介紹一種快速計算平輪撞擊之數值模擬方法進行制定，參考Mazilu[15]利用格林函數(Green’s Function)之數學概念計算平輪撞擊之力學行為，該方法簡單且快速，若能得到現地軌道結構之實驗參數，則可利用該方法建立出IR index之警戒層級趨勢，且搭配現地監控設備建置長期歷史資料庫並進行每日車輛平輪損傷追蹤，進而達到即時列車災害防護之最大功效。本論文詳細說明了實驗設備之安裝配置，應變訊號處理，分析方法、分析成果及平輪撞擊數值模擬結果。 ;Flat Spot is a defect of running surface on railway wheel , giving rise to noise and safety problems. This type of damage occurs when the wheel locks and slides along the rail, leading to the wheel surface presenting a lack of material. The manual visual inspection in depot is the commonly adopted detection to identify wheel defect ,but consumes a lot of time and manpower. Therefore, an automatic, reliable and accurate monitoring system for wheel tread is must desirable. The research describes a real-time wheel flat monitoring system based on the Fiber Bragg Grating (FBG) sensors of array-type, installed near the root of the rail for measuring the strain caused by each wheel load. Feature Line Method (FLM) and Impact Ratio index (IR index) has been developed to express the wheel condition. The FLM can generate the graphic visualization about non-contact period of flat spot, but cannot quantify a damage level of wheel flat. The IR give a ratio about peak load and average load to be an index that essentially indicated the condition of wheel defect, but the measured range reduced to 60% of wheel circumference. The alarm level of the IR index caused by wheel flats requires the application of numerical simulation - Green’s function and the historical database in the field experiments to formulate.
