Characterizing Mechanical Properties of Layered Engineered Wood Using Guided Waves and Genetic Algorithm

Sensors (Basel). 2023 Nov 14;23(22):9184. doi: 10.3390/s23229184.

Abstract

This study develops a framework for determining the material parameters of layered engineered wood in a nondestructive manner. The motivation lies in enhancing nondestructive evaluation (NDE) and quality assurance (QA) for engineered wood or mass timber, promising construction materials for sustainable and resilient civil structures. The study employs static compression tests, guided wave measurements, and a genetic algorithm (GA) to solve the inverse problem of determining the mechanical properties of a laminated veneer lumber (LVL) bar. Miniature LVL samples are subjected to compression tests to derive the elastic moduli and Poisson's ratios. Due to the intrinsic heterogeneity, the destructive compression tests yield large coefficients of variances ranging from 2.5 to 73.2%. Dispersion relations are obtained from spatial-temporal sampling of dynamic responses of the LVL bar. The GA pinpoints optimal mechanical properties by updating orthotropic elastic constants of the LVL material, and thereby dispersion curves, in a COMSOL simulation in accordance with experimental dispersion relations. The proposed framework can support estimation accuracy with errors less than 10% for most elastic constants. Focusing on vertical flexural modes, the estimated elastic constants generally resemble reference values from compression tests. This is the first study that evaluates the feasibility of using guided waves and multi-variable optimization to gauge the mechanical traits of LVL and establishes the foundation for further advances in the study of layered engineered wood structures.

Keywords: engineered wood; experimental mechanics; guided waves; mass timber; nondestructive evaluation; numerical simulation; optimization.