Application of the discrete wavelet transform to the detection of catastrophic tool breakage during drilling of aeronautical structural components

Abstract

This work presents an approach to the detection of catastrophic tool breakage in automatic drilling processes that take place during the assembly phase of aeronautical structural components. During the drilling phase, fuselage components are arranged in hybrid stacks, including carbon fiber composite and titanium alloy. Machining conditions vary along each drilling operation depending on the material and number of components, their thicknesses, cutting parameters, the cooling and/or lubrication system, and workpiece clamping. In the machining process of these hybrid stacks, the level of complexity can lead to critical conditions that result in catastrophic tool failure. This scenario is particularly problematic due to the cost and time involved, especially in automatic processes. To detect the moment when these events occur in real-time, a detection method based on the time-frequency domain analysis of the main spindle power consumption signals is proposed. Specifically, the feasibility of using the Haar, Daubechies, Symlet and Coiflet discrete wavelet transforms with different decomposition levels is studied to detect breakages considering signals corresponding to a 4.6 mm diameter tool in the production system.