Tool Wear Analysis in the Composites Machining through Hierarchical and Non-Hierarchical Clustering Approaches

The knowledge on the machining of ductile metals is not suitable for fibre reinforced plastic matrix composites where chip is powder-like or discontinuous and the work material structure is non-homogeneous and anisotropic. When optimisation problems are concerned, a deep knowledge of the mechanisms responsible for material removal and tool wear development is required. In this paper, sensor monitoring of tool conditions was carried out through detection and analysis of cutting force signals during orthogonal cutting of different types of composite materials. Decision making on tool wear state was performed using different methods: hierarchical and non-hierarchical clustering algorithms. These different paradigms are evaluated in terms of cluster identification performance