This research will investigate developing a machine diagnostics system based on sensors that are typically found in a machine tool. This approach will be applied to machine tool drive systems. The signals that are readily available are motor voltage, current, velocity, and position; these signals will be employed in this technique. The research is concerned with detecting either a change in the load imposed on the drive system (e.g. load caused by a failed bearing, gear, or belt or a slide/weighs fault) or a change in the dynamic behavior of the drive system (e.g. caused by motor characteristics or slide/weighs changes). The research will concentrate on mechanical component faults. The major tasks of the research include development of a dynamic mathematical model of the system to quantify the relation between the load and the remotely measured load signal, identification of the number of sensors required to detect machine faults, and off-line algorithms to detect and identify faults.
