The potential of colour measurement for strength prediction of thermally treated wood

This paper investigates the effects of thermal treatment of birch with respect to colour and strength. Birch wood was treated at 175 degrees C and 200 degrees C for 0 h, 1 h, 3 h and 10 h. In bending-strength experiments, treatment was also performed at 185 degrees C for 2 h. Both static bending strength and impact bending strength were investigated using multivariate statistics (PLS) for correlation to process parameters, density, EMC, position in board, modulus of elasticity (only in static bending), colour and dimensions of samples. In static bending, two PLS models were designed, one based on process parameters and the other based on colour and EMC. From these models it was concluded that colour is not a useful parameter for prediction of strength. In impacted bending, the correlation was too small to give useful results. One test of static bending strength with matched samples was performed, and it showed a strength reduction of 43% when treatment was conducted at 200 degrees C for 3 h. Measurement of colour homogeneity of the treated boards showed that the colour is not homogeneous.