An inverse moisture diffusion algorithm for the determination of diffusion coefficient

The finite difference approximation is applied to estimate the moisture-dependent diffusion coefficient by utilizing test data of isothermal moisture desorption in northern red oak (Quercus rubra). The test data contain moisture distributions at discrete locations across the thickness of specimens, which coincides with the radial direction of northern red oak, and at specified times. Also, the rate of moisture variation at each specified time and location must be known or correctly estimated. The functional form of the diffusion coefficient as well as the boundary conditions at the surfaces are not known a priori. The resulting system of finite difference equations defines an inverse problem, whose solution may be sensitive to small changes of input data. Results indicate that the diffusion coefficient increases with increasing moisture content below the fiber saturation point, which defines the upper limit applied by the diffusion theory.