Strategies for Identification of Elastic Constants in Highly Anisotropic Materials Using Lamb Waves

Abstract

Information about exact material properties may be of great importance in many areas where CAD/CAE software is used. It is also a key component of properly operating model-based SHM systems. Unfortunately, composite laminates producers are not providing sufficient and/or precise enough materials data sheets to meet such requirements. This is the reason why material properties identification techniques are attracting considerable interest.This paper presents a new, non-destructive elastic constants identification technique based on Lamb wave phenomenon. Experimental dispersion curves are obtained by 3D Fourier transform of full wavefield time responses registered in a tested sample by scanning laser Doppler vibrometer. Numerical dispersion curves, generated by a semi-analytical element model, are optimized to match experimental dispersion curves. By minimizing the discrepancies between two sets of data, the elastic constants are identified.Two approaches are tested, where the Genetic Algorithm is used to fit dispersion curves in the wavenumber-frequency domain for chosen propagation angles or angular profiles in the wavenumber-angle domain for chosen frequencies. The direct approach was used in which C-tensor components where optimized.