GREMAN UMR 7347, Université de Tours, INSA Centre Val de Loire, CNRS, 3 Rue de la Chocolaterie, Blois, France.
Ultrasonics. 2022 Feb;119:106619. doi: 10.1016/j.ultras.2021.106619. Epub 2021 Oct 18.
This paper describes a non-contact method to characterize isotropic and anisotropic planar multilayer structures using a genetic algorithm. The method is based on the determination of critical angles, where the maxima of the modulus of transmission coefficient of the structure appear, and which correspond to the generation of guided waves. The optimization process minimizes the error between the reference critical angles and associated amplitudes of the transmission coefficient, with the corresponding estimated ones. The estimation of elastic parameters is demonstrated for acrylic and oak plates as well as for a bi-layered structure composed of oak and a thin layer of gesso. It is shown that to obtain satisfactory optimization results, it is necessary for guided modes of higher order than the zero ones to be taken into account. Results also show that some elastic constants such as C and C retrieved from the transmission coefficient are very sensitive to the optimization.
本文提出了一种使用遗传算法对各向同性和各向异性平面多层结构进行特征描述的非接触方法。该方法基于临界角的确定，当结构的传输系数模的最大值出现时，就会产生导波。优化过程最小化了参考临界角与传输系数的关联幅度之间的误差，以及对应的估计值。弹性参数的估计是针对亚克力和橡木板以及由橡木和薄薄的石膏层组成的双层结构进行的。结果表明，为了获得令人满意的优化结果，有必要考虑高于零阶的导波模式。结果还表明，从传输系数中获取的某些弹性常数，如 C 和 C ，对优化非常敏感。
