Seiner Hanus, Landa Michal
Department of Materials, CTU Faculty of Nuclear Sciences and Physical Engineering, Trojanova 13, 120 00 Prague 2, Czech Republic.
Ultrasonics. 2005 Feb;43(4):253-63. doi: 10.1016/j.ultras.2004.07.004.
The elastic coefficients of anisotropic solids are often evaluated from measurements of phase or group velocities of ultrasonic bulk waves by the usage of inverse optimizing procedures. This paper discusses the effects of various factors on such procedures results for transversely isotropic solids with considerably strong anisotropy. First, the inverse determination of all elastic coefficients of unidirectional CFRP composite is briefly outlined. Then the results of the optimization are treated as exact values and the sensitivity of the optimizing process versus main considered sources of inaccuracies is analyzed. Results of extensive simulations are presented to illustrate the effect of input data distortion, input data incompleteness, and geometrical conversion from experimentally obtained group velocities into corresponding phase velocities used as input data for the optimizing procedure. The paper takes note of how information about the elastic coefficients can be extracted from the different segments of the phase velocity surface. The stability versus input data distortion for inversion from group velocities and phase velocities is compared and the importance of reliable geometrical converting from group into phase velocities is illustrated. An novel method for geometrical conversion of distorted group velocity data into corresponding phase velocities based on affine combinations of low-order polynomials is presented and compared with piecewise or high-order polynomial fitting.
各向异性固体的弹性系数通常通过使用反向优化程序，根据超声体波的相速度或群速度测量值来评估。本文讨论了各种因素对具有相当强各向异性的横向各向同性固体此类程序结果的影响。首先，简要概述了单向碳纤维增强塑料（CFRP）复合材料所有弹性系数的反向确定。然后将优化结果视为精确值，并分析优化过程对主要考虑的不精确来源的敏感性。给出了大量模拟结果，以说明输入数据失真、输入数据不完整以及从实验获得的群速度到用作优化程序输入数据的相应相速度的几何转换的影响。本文注意到如何从相速度表面的不同部分提取关于弹性系数的信息。比较了从群速度和相速度进行反演时相对于输入数据失真的稳定性，并说明了从群速度可靠地几何转换为相速度的重要性。提出了一种基于低阶多项式仿射组合将失真的群速度数据几何转换为相应相速度的新方法，并与分段或高阶多项式拟合进行了比较。
