Machinability Study of Fibre-Reinforced Polymer Matrix Composites

Abstract

The trend in the applications of advanced composite materials, namely the fibre-reinforced polymer (FRP) composites, ranges from high performance industrial products to the low end consumer goods. These composite products are commonly fabricated to near-net shapes with finishing steps that involve machining being the integral part of component manufacture. However, the composite machining becomes a challenge compared to that of the conventional metallic materials due to their inherent properties. The damage susceptibility of the FRP composites impedes the consistency of machining quality, whereas the abrasiveness of the workpiece material inflicts rapid wear on the cutting tools. As a result, extensive scientific research has been devoted to investigate the machinability of these materials in order to elucidate their fundamental machining characteristics. Although much attention on turning and drilling of FRP composites can be traced in the existing literature, only a handful of researchers have reported experimental results on limited aspects of FRP composites milling machinability indices. Hence, this thesis has embarked on a systematic machinability study of end milling glass fibre-reinforced polymer (GFRP) composites. A design of experiment methodology was initially employed to determine the effects of machining parameters on key machinability indices or outputs and the suitable operational or machining parameters (guided by the final applications). On the basis of this parametric study, experimental investigations under a wider range of machining parameters and material characteristics were conducted. From these experiments, the empirical relationships between tool performance (in terms of tool life) and the selected parameters were analysed using the traditional Taylor’s tool life equation. The useful life of the cutting tool was found to be well described by the Taylor’s equations. The cutting speed was identified as the key parameter in influencing the tool life followed by feed rate and fibre orientation. Surface finish, on the other hand, was found to marginally improve with a higher spindle or cutting speed, but rapidly deteriorated with the increase of feed rate. An acceptable machining quality could be achieved by machining along the fibre orientation despite a higher tool wear rate. It appears from the scanning electron microscopy that the machining induced damage comprises fibre fracture, pull-out or protrusions, delamination damage, and epoxy matrix brittle failure. All of these are attributed to the high machining force and reduction of tool sharpness. The constitutive relationships between the growth of tool wear and the measured machining forces were also studied as a pursuit to monitor the cutting tool condition during machining operation. Although adequate agreement between experimental data can be well achieved using multiple regression analysis, the application of fuzzy logic with neural network model demonstrated a significant improvement in the prediction accuracy. Notably, the accuracies of this model are pronounced as a result of nonlinear fuzzy membership function and its hybrid learning algorithms. This makes it attractive as an indirect tool condition monitoring during the machining operation. Machinability of GFRP composites has also been qualitatively evaluated in terms of chip forming mechanisms. This has been accomplished using a high-speed video camera and a quick-stop method. It is apparent that the heterogeneity and insufficient ductility of the composites have produced discontinuous and fracturing chips under the tested machining parameters. A layer of delaminated chip was formed (under the mild cutting speed) as the tool cutting edge fractured the workpiece material along the fibre orientation. However, the increased cutting speed and fibre orientation accelerate the fracture of chips into smaller segments, which make it difficult to denote any chip formation processes.