Numerical Determination of Diffusion Coefficients in Wood Using Data From CT-Scanning

The radial moisture diffusion coefficient in Fick's law for a sample of Norway spruce (Picea abies) under isothermal drying conditions was determined in a parameterization of Arrhenius' equation type. Using X-ray CT-scanning, the wood density and moisture content distributions were obtained in the radial direction for the wood sample. An optimization scheme, based on finite element computation, was then applied to find the parameter values such that the difference between observed and computed moisture content was minimized. The combined numerical and experimental technique was developed to reduce known disadvantages of similar approaches, and a specific algorithm to determine diffusion coefficients was presented. A comparison of the calibrated diffusion coefficient with those given in the literature showed a good fit. The computed moisture content based on the obtained diffusion coefficient and the observed moisture content agreed well. Finally, the effect of measurement errors on the computed material parameter was found to be small