On line monitoring system for rotation mechanical spindle fault based on virtual instrument

BING Zhigang1, LI Weilin2, CHEN Feng2, LI Shanghui2, Dong Chenchen2

(1. Zhejiang Test Academy of Quality and Technical Supervision, Hangzhou 310018, China;

2. Zhejiang Fangyuan test Group Co., Ltd., Hangzhou 310013, China)

Abstract: Aiming that spindle running state of machine tool during working has a great influence on product precision, cut tool lifetime, spindle condition monitoring technology during the process of machining was studied. A spindle condition monitoring and diagnosis system for rotator machine was established. The vibration, speed and temperature signal of rotator machine were selected as monitoring signal, the effective fault characteristic which was sensitive to spindle fault was extracted by signal analysis techniques, including time domain analysis, frequency domain analysis and wavelet packet analysis technology, a recognition model based on fuzzy Cmeans clustering algorithm was established, the condition of spindle can be detected by calculating the membership between the unknown samples and known state. Finally, a spindle condition monitoring and diagnosis software was designed by LabVIEW which is a virtual instrument programming tool, and the theoretical analysis and instance validation were carried on in the actual machine tool. The results indicate that the faults of the spindle included Impact, Friction, Loose and Imbalance, can be classified and identified by this system, and has high recognition success rate, up to 99%, and it has the features of fast and efficiently.

Key words: virtual instrument; fault diagnosis; wavelet packet analysis; fuzzy Cmeans clustering