Multiscale prediction of viscoelastic properties of softwood under constant climatic conditions

Abstract

This paper covers the development and validation of a multiscale homogenization model for linear viscoelastic properties of wood. Starting point is the intrinsic structural hierarchy of wood, which is accounted for by several homogenization steps. Using the correspondence principle, an existing homogenization model for the prediction of elastic properties of wood is adapted herein for upscaling of viscoelastic characteristics. Accordingly, self-consistent, Mori–Tanaka, and unit-cell-based techniques are employed, leading to pointwise defined tensorial creep and relaxation functions in the Laplace-Carson domain. Subsequently, these functions are back-transformed into the time domain by means of the Gaver-Stehfest algorithm. With this procedure the orthotropic macroscopic creep behavior of wood can be derived from the isotropic shear behavior of the lignin-hemicellulose phase. A comparison of model predictions for viscoelastic properties of softwood with corresponding experimentally derived values yields very promising results and confirms the suitability of the model.