Prediction of the Bending Strength of a Laminated Veneer Lumber (LVL) Using an Artificial Neural Network

The application of an artificial neural networks (ANN) to predicting the bending strength of a laminated veneer lumber (LVL) manufactured under different conditions is considered. First, experimental studies were conducted, and then an ANN model was developed based on the experimental data obtained. LVL specimens of walnut wood, glued with urea-formaldehyde resins containing a chemically modified starch and nanocellulose, were obtained by pressing for different times. Experimental results for them showed that the direct effect of the press time, the square effect of the modified starch, and the joint effect of them had the highest statistical significance to the bending strength of LVL. The ANN model developed gave good predictions for the bending strength, well agreeing with experimental data.