具週期性電極之壓電纖維複材的材料性質研究

Abstract

以上下對稱之指差電極驅動的壓電纖維複材是一種新興的智慧型材料，具有高變形與制動能力，壓電纖維複材的可撓性高，適宜貼合於具有曲率的結構表面。可作為功能性材料，應用於結構健康監測、航空結構等。本研究依據混合律，將壓電纖維複材模擬成一種等效均質材料。部份材料常數 s33、s13 與 d33、d31 在單軸拉伸試驗及高電壓負載試驗中，可以直接藉應變規量測獲得，其餘的材料常數則以準牛頓法為基礎之反算程序決定。本研究採用有限元素套裝軟體ANSYS，計算壓電纖維複材軸向共振頻率，反算的二次目標函數是懸空壓電纖維複材軸向振動最低共振頻率之量測值與理論值的差值平方和。 根據偏光顯微鏡的觀察，壓電纖維複材的纖維與指差電極排列並非均勻，會導致材料常數量測的偏差。壓電常數d15對於軸向振動的共振頻率影響較小，並在參數的敏感度研究中呈現發散特性，不適合以本研究的反算方法求得。反算所得的材料常數大多與混合律估算的等效常數吻合。

Piezoelectric fiber composite (PFC) actuated by symmetrically aligned interdigital electrodes on its opposite surfaces is an emerging smart material developed in recent years. The PFC is capable of providing large deformation and stiff exciting force. The flexibility of PFC makes it possible to conform to curved surfaces of the host structures. These properties make PFC possess great potential to become a functional material in applications of structural health monitoring, aerospace structures, etc. The effective material constants of PFC are estimated by rule of mixtures. Some material constants, s33, s13 and d33 , d31 , are directly measured by stain gauges through uniaxial tensile test and applying high voltage to its opposite electrodes. The remaining constants are determined by an inverse scheme based on quasi-Newton method. The quadratic objective function is defined by the sum of squares of deviations between the measured resonant frequencies of the lowest extensional vibration modes for the finite-length overhung PFC and those calculated by a commercial finite element code ANSYS. It is observed by a polar light microscope that the piezoelectric fibers and interdigital electrodes are not uniformly aligned in some specimens. The deviation found in measured material constants is caused by the inhomogeneous material properties. The piezoelectric constant d15 is not suitable to be determined by the present method since it has weak influence on resonant frequency of extensional mode and divergent properties in seneitivity study of parameters. Most material constants are in agreement with those obtained by effective model estimated by mixture rule.