Estimating birch veneer (Betula pendula Roth) moisture content using infrared technology

The potential of infrared (IR) technology in estimating the moisture content (MC) of birch (Betula pendula Roth) veneer was investigated. A total of 56 birch veneer sheets (42 x 130 cm(2)) were dried with a convective type laboratory-scale veneer dryer at drying times ranging from 1 to 6 min at 1 min intervals. After drying, IR data were recorded and the veneer sheets were sawn into 100 x 100 mm(2) pieces. IR data were averaged over the same area and then the true MC and density values were assessed gravimetrically. The dependency between temperature and MC was found to be non-linear and between temperature and density there was no dependency. The relationship between temperature and MC was modeled with a neural network and a Gaussian model. Both model types gave similar results. At a true MC below 10 %, the root mean square error of prediction (RMSE) was 1.5 % for the 100 x 100 mm(2) pieces and 1.2 %, for the larger veneer sheets, whereas at higher MC (above 10 %), the RMSE increased to 2.6 % for the 100 x 100 mm(2) pieces and 1.9 % for the sheets. It was observed that the IR measurement should be taken within 5 s, since after that the RMSE increased rapidly. Based on the results of this study, it was concluded that veneer MC can be estimated with reasonably good accuracy using IR technology.