A Note on Optimization-Based Strategies to Identification of Material Parameters in Thermo-Hygro-Mechanical Problems

The interdependence between temperature, moisture content and dimensional changes is generally associated with thermo-hygro-mechanical problems. The importance of such problems lies on the fact that non-homogeneous temperature and moisture content lead to localised material swelling/shrinkage and growth of hygroscopic stresses with possible material failure. The present work addresses application of inverse problem techniques to identification of wood parameters based upon thermo-hygro-mechanical modelling of drying processes. The direct problem constitutes a transient, fully coupled solution of the energy (temperature), mass (moisture content) and momentum (displacements) conservation equations. A constitutive model including hygrostresses is also presented. The inverse problem is based on unconstrained optimization and adopts a global–local hybrid strategy. Particle swarm optimization is adopted as the global stage, whereas the Nelder-Mead scheme is used for the local stage. The strategy was successfully applied for the proposed problem, in which the total sample mass and geometrical distortion were used to determine moisture diffusive parameters and swelling/shrinkage coefficients.