Machinability of polypropylene biocomposites reinforced with natural fibers and biocarbon particles

Abstract

Biocomposites are increasingly being developed and applied in many engineering fields. Parts made of biocomposites are mostly produced close to their final shape. However, the secondary machining operation is required to achieve dimensional accuracy and surface finish. The machinability of biocomposite is significantly affected by the processing method, the reinforcement, and the matrix used in the biocomposite. Therefore, it is necessary to understand the machining behavior of a particular biocomposite. It could thereby help to choose suitable cutting parameters to obtain the desired surface quality during machining. This study investigates the key factors that affect the machinability indicators during the dry drilling of new biocomposites. These biocomposites are made of the same matrix (PP/POE/MAPP) reinforced with different weight ratios of chopped miscanthus fibers and biocarbon particles: biocomposite M1 (30wt% biocarbon), biocomposite M2 (30wt % miscanthus), and hybrid biocomposite M3 (15wt% miscanthus, 15wt% biocarbon). A full factorial design was used for the experimental design to study machining behavior during the drilling of biocomposites. The effects of the drilling parameters on the machinability indicators of biocomposites are measured and analyzed. The results from this research work demonstrated that the cutting parameters and the tool diameters significantly affect the machining process performance indicators (thrust force, specific cutting energy for thrust force, surface roughness, and fine particle emission) for all biocomposites. The type (nature and shape) and the weight ratio of reinforcement used in biocomposite have a significant effect on the machinability indicators. The drilling parameters are not statistically significant for ultrafine particle emission. Surface roughness (Ra) and thrust force in the drilling of biocomposites were predicted using regression analysis and ANFIS-based models. The results showed that predictive models have high predicted accuracy.

Date	16 Dec 2020
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Victor Songmene (Supervisor) & Anh Dung Ngô (Co-supervisor)