MODELING THE WATER ABSORPTION RATE OF WOOD IMPREGNATED WITH SILICONE-BASED CHEMICALS USING AN ARTIFICIAL NEURAL NETWORK


GÜRGEN A. , Kilic C., YILDIZ Ü. C. , YILDIZ S.

DREWNO, cilt.62, ss.55-69, 2019 (SCI İndekslerine Giren Dergi) identifier identifier

Özet

In this study, the water absorption rate of wood impregnated with silicone-based chemicals was predicted by an artificial neural network (ANN). For this purpose, spruce (Picea orientalis L.) and beech (Fagus orientalis L.) wood samples impregnated with five commercial silicone-based chemicals were tested. Wood specimens were impregnated with these chemicals in concentrations of 10% and 50%, and the water absorption rates of samples at different times (2, 4, 8, 24, 48, 72, 168 and 336 hours) were calculated. These results were then modeled by an ANN. Wood species, silicone-based chemical, concentration and time in water were used as the input variables, and water absorption rate as the output variable. The results show that an ANN can be used successfully for predicting the water absorption rate of wood impregnated with silicone-based chemicals.