Statistical Modeling and Response Surface Optimization of Extruded HDPE/Wood-Flour Composite Foams

In this paper, a regression model was constructed to optimize the relationships between the void fraction of foamed HDPE/wood-flour composites, the processing conditions (extruder\u27s die temperature and screw speed rate), and the formulation compositions [chemical foaming agent (CFA) concentration and the moisture content of wood flour] by applying a four-factor central composite design (CCD) statistical approach. Design Expert® software was employed to carry out the experimental design, statistical analysis, and numerical optimization. The analysis of variance (ANOVA) of the model showed that the void fraction of HDPE/wood-flour composites was a strong function of the extruder\u27s die temperature, the screw speed rate, and the moisture content of wood flour, while independent of the CFA content within the studied range (0 ∼ 1%). In addition, the moisture content of wood flour and the extruder\u27s screw speed rate exhibited significant interaction effect on the void fraction of the foamed composites. The response surface optimization generated using the regression model suggests that HDPE/wood-flour composite foams with the highest void fraction can be achieved at the highest levels of extruder\u27s screw speed rate (120 rpm) and moisture content of wood flour (12%), together with the lowest levels of extruder\u27s die temperature (170°C) and no chemical foaming agent