Model based fault diagnosis in rotating machinery

Abstract

A continuing task in engineering is to increase the reliability, availability and safety of technical processes and to achieve these fault diagnosis becomes an advanced supervision tool in the present industries. Vibration in rotating machinery is mostly caused by unbalance, misalignment, shaft crack, mechanical looseness and other malfunctions. The objective of this paper is to propose a model based scheme for fault diagnosis of a rotor system. Presence of faults changes the dynamic behaviour of the system which is taken into account by equivalent loads acting on the healthy system model. In order to diagnose the faults in a rotor system the experimental time responses for healthy system as well as for faulty system were used. It was observed that the proposed scheme successfully detects and identifies the type, location and amount of fault in a rotor system for unbalance, misalignment and crack. This method has thus demonstrated the efficacy of the model based fault detection system for a simple rotor-bearing system.