Science-based Optimization of Plastic Wood Composite Formulations for Extrusion

J. Xing, A. Kundu, A. Duhduh, J. Coulter
Lehigh University,
United States

Keywords: extrusion, PVC, wood, composite, optimization

Summary:

This study focuses on the development of science-based optimization strategies for PVC based plastic wood composite (PWC) formulations. The PVC composites are used for various applications such as door frames, outdoor decking and furniture etc. In the industry, the composite formulations are often developed empirically to achieve optimal price-performance balance. The composite formulations are fairly complex, may contain more than 10 different raw materials and the prices of each of the raw materials are diverse. An empirically derived composite formulation is highly unlikely to be the optimal solution. In this study, a scientific method is applied to understand the effect of the individual components in the starting mix on the processing and the properties of the final parts. This fundamental understanding is then utilized to determine the optimal formulations to meet specific price-performance criteria. In this research, PVC based plastic wood composite (PWC) formulations containing 10 distinct components were investigated. It was determined four specific additives, namely, impact modifier, processing aid, stabilizer, and foaming agent are critical components in these formulations. A set of distinct formulations were designed following a Taguchi optimization method and door frame profiles were extruded for each of the formulations. The impact resistance and the pull-out strength of the frames were measured and the correlation between the formulations and the properties were analyzed. These properties are the most relevant for this kind of applications. The correlations between the properties and the formulations were established from these experiments. With these results, it would be possible to predict the expected properties for a specific formulation and the associated cost to produce the parts. This scientific method could be integrated into a software platform which can be then utilized to determine optimal performance-formulation-processing-price combinations for specific applications.