Machinability Measurements in Milling and Recurrence Analysis of Thin-Walled Elements Made of Polymer Composites

Abstract

The milling of polymer composites is a process that ensures dimensional and shape accuracy and appropriate surface quality. The shaping of thin-walled elements is a challenge owing to their deformation. This article presents the results of milling polymer composites made of glass and carbon fibers saturated with epoxy resin. The milling of each material was conducted using different tools (tools with polycrystalline diamond inserts, physically coated carbide inserts with titanium nitride and uncoated carbide inserts) to show differences in feed force and deformation after the machining of individual thin-walled samples. In addition, the study used recurrence analysis to determine the most appropriate quantifications sensitive to changes occurring in milling different materials with the use of different tools. The study showed that the highest forces occurred in milling thin-walled carbon-fiber-reinforced plastics using uncoated tools and the highest feeds per revolution and cutting speeds. The use of a high feed per revolution (0.8 mm/rev) in carbon-fiber-reinforced plastics machining by uncoated tools resulted in a maximum feed force of 1185 N. A cutting speed of 400 m/min resulted in a force of 754 N. The largest permanent deformation occurred in the milling of glass-fiber-reinforced composite samples with uncoated tools. The permanent deformation value of this material was 0.88 mm. Low feed per revolution (0.1 mm/rev) resulted in permanent deformations of less than 0.30 mm for both types of materials. A change in feed per revolution had the most significant effect on the deformations of thin-walled polymer composites. The analysis of forces and deformation made it possible to conclude that high feed per revolution were not recommended in composite milling. In addition to the analysis of machining thin-walled composites, the novelty of this study was also the use of recurrence methods. Recurrence methods were used to determine the most appropriate quantifications. Determinism, averaged diagonal length and entropy have been shown to be suitable quantifications for determining the type of machined material and the tools used.