This paper describes a vibration based diagnostic method for estimating the surface roughness on the race in ball bearings. The surface roughness has been measured by a stylus which directly traverses the surface of the ring obtained by taking apart the ball bearing. We developed a new method to accurately estimate the surface roughness by analyzing the short-length vibration signal which is excited when balls contact with the surface of a defective rotating ring in a ball bearing. Our experimental results confirm that the roughness estimated by the proposed method agrees with that measured directly by using a stylus even in the case of crack μm wide. We applied this new method to the diagnosis of surface roughness in small-sized ball bearings and inferior samples were detected with a 94.7 % accuracy rate.
ASJC Scopus subject areas
- Acoustics and Ultrasonics