Machine tool condition monitoring system using tooth rotation energy estimation (TREE) technique

This paper introduces a tooth rotation energy estimation (TREE) technique and its implementation on a PIC Microcontroller based distributed machine tool condition monitoring system. The technique uses existing machine signals namely; spindle speed and spindle load for the purpose of data acquisition, analysis and decision making thus avoiding the use of any additional sensors. The paper discusses the evolution of this time domain technique, starting from signal acquisition, hardware filtering, the application of moving average for software filtering before going on to explore the signal´s variations for different tool conditions. The acquired data is analysed in terms of the energy per tooth. The strength of an energy index in the acquired signals under various cutting conditions can then be used for fault diagnosis and prognosis. The software and hardware system architectures and the test application on a Kondia B500 vertical axis milling machine are described