Theoretical modeling of bonding characteristics and performance of wood composites. Part II. Resin distribution

Further to the development of the inter-element contact model reported in Part I of this series (Dai et al 2006), this paper reports the development of a mathematical (analytical) model and a computer simulation (numerical) model of resin distribution. Based on theories of random coverage process and stochastic system, resin distribution is analytically defined by the average and the variance of resin coverage on constituent wood elements. To complement the analytical model. a numerical model using image digitization and Monte Carlo technique is developed on a computer to visualize and simulate the spatial variation of resin coverage. To validate the models, resin distributions on OSB strands are experimentally investigated using an image analysis technique. The analytical and numerical models are validated by close agreements with each other and with experimental results. It is proposed and modeled that resin coverage is classified into area coverage and mass coverage. The former follows an exponential relationship with resin content, while the latter has a linear relationship with resin content. Both area and mass coverage are strongly affected by element/strand thickness and wood density. The resin area coverage is further affected by resin spot thickness, density, and solids content. Resin spot size has no effect on the average but strong effect on the variance of resin coverage. Implications of the model predictions on improving uniformity and efficiency of resin application are also discussed.