It is common practice nowadays with critical machinery to measure, either on-line or off-line, the vibration response at a number of suitable locations and to assess the state of health of the equipment based on observed changes in one or more of the response parameters. Early diagnosis of potentially damaging faults or defects can lead to substantial cost savings. Frequently, in practice, interpretation of the vibration data is performed more on a qualitative as opposed to quantitative basis which may be of limited benefit. The work presented here shows that by utilising a simple approach Involving the (SVD) Singular Value Decomposition method, changes in measured vibration data can be incorporated into an appropriate rotordynamic model and identification of the cause of the modified vibration response achieved, The technique is intended to be used in conjunction with available information related to the machine's construction and/or operational history. The proposed approach consists of two separate components: the first used to identify bearing-related problems and the second employed in the diagnosis of rotor balance problems. The effectiveness of the proposed procedure is assessed by recourse to analysis of a real representative rotor system.
展开▼