標題: 複合材料風力葉片之彈性常數識別
Elastic Constants Identification of Composite wind blade
作者: 鄭堪元
金大仁
Zheng.Kan-Yuan
Kam.Tai-Yan
機械工程系所
關鍵字: 彈性常數識別;最佳化方法;風力葉片;複合材料;Elastic constants identification;Optimization method;Wind blade;Composite material
公開日期: 2017
摘要: 複合材料結構在不同的製造過程及製造環境,皆會影響其材料的各種彈性常數。本文目的在針對複合材料風力葉片進行彈性常數識別。 本研究先經由靜態載重實驗、動態振動實驗,探討複合材料風力葉片在靜態負載下的應變、動態振動之頻率,並使用有限元素分析軟體ANSYS,對葉片變形狀況作分析,結果顯示分析貼近實際情形,以此驗證葉片模型之正確性。 本研究利用兩階段複合式最佳化方法針對2.5m複合材料風力葉片進行彈性常數識別。第一階段為隨機多起始點最佳化方法,其可有效於繁雜不規則的多局部極小值問題中獲得可靠度較高之總域極小值;第二階段為窮舉搜尋法,此方法可針對較小範圍進行全域搜尋進而獲得更為精準之識別結果。複合材料風力葉片彈性常數識別過程中,分別將量測的蒙皮表面應變與動態實驗之頻率作為目標函數之識別參數,透過兩階段識別方法進行識別後結果顯示利用蒙皮表面之應變作為識別參數之識別方法(誤差10%)優於利用頻率作為識別參數之識別方法(誤差41%),利用應變識別更能預測複合材料葉片之特性。
Composite materials structure fabricated using different manufacturing processes and under different manufacturing environments may have different the material's elastic constants. The purpose of this paper is to identify the elastic constants of composite wind blades. This study uses the static load test to measure the strain of composite wind blades and dynamic vibration test to investigate the frequency of composite wind blades. The finite element code ANSYS is used to analyze the deformation of the blade. The results have shown that the analysis can simulate the actual situation and the correctness of the blade model is verified. In this study, a two-stage compound optimization method is used to identify the elastic constants of 2.5 m composite wind blades. The first stage uses the random multi-start and –trajectory method to obtain multi-local minima. The second stage uses the exhaustive search method to find the global minimum. This stage allows the global search in a smaller region to find the global minimum. In the proposed elastic constants identification method, the measured skin surface strains or the natural frequencies of the blade are used to construct the objective function via the least square approach. The identification using the skin surface strains produces the elastic constants with 10% error. The identification using the natural frequencies produces the elastic constants with error 41%. Therefore, the identification method using the measured strains of skin surface can predict the material constants of composite blades more accurately.
URI: http://etd.lib.nctu.edu.tw/cdrfb3/record/nctu/#GT070451068
http://hdl.handle.net/11536/142909
Appears in Collections:Thesis