EVALUATION OF A REGRESSION PREDICTION MODEL FOR SURFACE ROUGHNESS OF WOOD–POLYETHYLENE COMPOSITE (WPC)

Abstract

Milling processing is an important way to obtain wood–polyethylene composite (WPC) end products. In order to improve the processing efficiency and surface quality of WPC and meet the practical application requirements, this paper focussed on morphology and roughness of the WPC-milled surface and studied surface quality changes under different cutting parameters and milling methods through multi-parameters milling experiments. The milling surface morphology and roughness of WPC were analyzed and measured during cut-in, cutting and cut-out sections. It also revealed the affect rule of different cutting parameters and milling methods on milled surface morphology and roughness. The results show that the milling surface roughness of WPC products with wood powder content of 70% is significantly larger than the one whose wood powder content is 60%, and defects such as holes are also relatively more. Finally, a surface roughness prediction model was established based on the mathematical regression method and its multi-factor simulation was carried out. A comparative analysis of predictive and experimental values was performed to verify the reliability of the model. It could also provide theoretical guidance and technical guarantee for high processing quality of WPC milling and cutting.