Tool wear prediction in edge trimming of MD CFRP considering interlaminar effect

Abstract

The tool flank wear progresses rapidly in edge trimming of carbon fiber reinforced polymer (CFRP) components, affecting the machined surface quality and process efficiency. This paper presents a tool wear prediction method considering the unique interlaminar effect of the multi-directional (MD) CFRP in the edge trimming process. The experimental results show that different laminar configurations of MD CFRP change the tool wear progression of unidirectional (UD) plies at specific fiber orientations. The interlaminar effects from the adjacent layers with different fiber orientations are explained by their varying supporting strengths on the target layer due to the change of bouncing back height. Therefore, the interlaminar effect from two sides of a UD ply is not a simple summation of the contributions by the individual side. The machined surface roughness of unidirectional ply varies up to 30 % due to the interlaminar effect. A long short-term memory (LSTM) - back propagation (BP) network is developed to predict tool wear length, including the effect of different interlaminar configurations. With the proposed model, the effect of the interlaminar configurations on tool wear progression in edge trimming of MD CFRP is predicted quantitatively. The prediction of the wear length progression based on the developed machine learning model is validated by the experimental results.