Stochastic numerical model for conventional kiln drying of timbers

Abstract

A numerical model that predicts the stochastic dispersion associated with industrial kiln drying of timber was adapted to conventional drying and evaluated with experimental data. The theoretical aspects of the model are briefly explained, a selection of the calibration parameters was carried out, and a new empirical dispersion factor is proposed to account for all unknown sources of random behavior. The model was calibrated with six experimental runs of western hemlock and amabilis fir (116 mm² timbers) to an average moisture content (target) of 14%–20%. It was found that with implementation of the dispersion factor, the number of required simulations is considerably reduced, the calibration results are consistent for all the experimental runs, and the target moisture content along with its standard deviation can be well reproduced using the all-run average parameters.