Experimental analysis of change detection algorithms for multitooth machine tool fault detection

This paper describes an industrial application of fault diagnosis method for a multitooth machine tool. Different statistical approaches have been used to detect and diagnose insert breakage in multitooth tools based on the analysis of electrical power consumption of the tool drives. Great effort has been made to obtain a robust method, able to avoid any needed re-calibration process, after, for example, a maintenance operation. From the point of view of maintenance costs, these multitooth tools are the most critical part of the machine tools used for mass production in the car industry. These tools integrate different kinds of machining operations and cutting conditions.