本文使用最佳化方法來設計壓電複合結構之模態感測器。通常儀器所量測到的信號為所有模態的信號的加總,然而透過不斷的修正電極形狀來使所要感測的模態之信號遠大於其餘模態的信號。先利用有限元素法分析軟體ANSYS分析出全部鋪滿接地電位的電極之系統模態,再利用本文所提不影響系統模態的切割方法切割出所需的電極形狀,而所需電極形狀配合系統模態振形透過MATLAB進行形狀之最佳化設計,並利用信號後處理來感測特定模態,此時只針對某幾個特定模態來設計。反之,亦可以達到濾除某特定模態之信號。 為驗證本文所提方法的可行性,以鋁板黏貼壓電片為例,利用微影製程定義出最佳電極的形狀,使用衝擊槌來激振試片,將電流信號轉電壓信號並作運算,再將信號輸入頻譜分析儀,結果確能達到模態感測器與模態濾除器的效果。
Piezoelectric modal sensors can be created using the orthogonal property of eigenfunctions. Sensors have been designed for use with 1-D flexural structures by shaping electrode patterns proportional to modal strains. However, it is not easy to apply the same concept to general structures due to the difficulty in implementing the location weight needed for signals. Therefore, a nonlinear optimization scheme is proposed in this paper to design modal filters for general structures. For any given electrode pattern, the signal contributed from each mode is found by integrating the corresponding free surface charges on the sensing electrode. A modal filter is obtained by modifying the electrode pattern to achieve the desired relative signal strength for different modes. Sensors capable of capturing or filtering out the signal generated by a specific mode can be developed. A PZT coated aluminum plate is used as an example. Mode shapes are found with finite element analysis. Design of both a mode-pass and mode-reject filter is demonstrated. This method also works for three-dimensional structures, as only free surface charges are involved in the operation of modal filters.