Title: A hybrid fault diagnosis and monitoring system for precision rotor systems

Authors: Shih-Ming Wang, Ming-Je Luo, Da-Fan Chen, Kai-En Chang

Addresses: Chung Yuan Christian University, Chung-Li, 32023 Taiwan. ' Chung Yuan Christian University, Chung-Li, 32023 Taiwan. ' Chung Yuan Christian University, Chung-Li, 32023 Taiwan. ' Precision Machinery Research & Development Center, Taichung, 407 Taiwan

Abstract: Rotor and its corresponding system play an important role in a precision machine. However, accuracy deterioration, machine damage, or malfunction could also happen to the machine due to various abnormalities of the rotor system. Thus, a monitoring system that can early detect and identify abnormalities so that proper actions can be taken to ensure the machine works properly is necessary. In this study, fault diagnosis/monitoring method that can online monitor the three major abnormalities (impact by unusual force, abnormal friction and unbalanced rotation) of a rotor system was developed. Both of wavelet transform and fast Fourier transform were employed in the method to analyse the permanent change and transient change of system vibration signals for abnormality diagnosis. Characteristic values of converted vibration signals of a rotor system were calculated and used to form a diagnosis map. Subsequently, occurrence possibility of an abnormality can be predicted with the use of the degree of approach model. To verify the proposed monitoring method and system, both simulation and experiments were conducted. The results have shown high feasibility and reliability of the detection system.

Keywords: failure prediction; fault diagnosis; fault monitoring; rotor systems; wavelet transform; Fourier transform; precision machinery; force impact; abnormal friction; unbalanced rotation; vibration signals; simulation.

DOI: 10.1504/IJCAET.2011.042357

International Journal of Computer Aided Engineering and Technology, 2011 Vol.3 No.5/6, pp.480 - 491

Published online: 06 Sep 2011 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article