Improving defect detection of rolling element bearings in the presence of external vibrations using adaptive noise cancellation and multiscale morphology

Identifying the signals of defective rolling element bearings of industrial machines in the presence of external vibrations is a difficult task. Therefore, in this article, attempt has been done to improve the defect detection of rolling element bearings in the presence of external vibrations using adaptive noise cancellation (ANC), self-adaptive noise cancellation (SANC), and mathematical multiscale morphology (MMM). Circular defects (diameter 400 mu m), on either of the races of the test bearings, have been artificially created and external random vibrations have been imparted to the defective test bearings using an electromechanical shaker in the experimentation reported herein. The defective bearing signal-to-external noise (vibration) ratio has been significantly enhanced after the application of ANC, SANC, and MMM. This resulted in the clear identification of defect frequencies in the vibration spectrum. It is essential to mention here that in ANC technique the least mean square (LMS) algorithm with and without signum function has been used. However, in MMM technique, triangular structural elements are utilized during the closing operation and bottom-hat transform (BHat). In comparison to the LMS algorithm without signum function, the LMS algorithm with signum function has been proved more effective and efficient in minimizing the error. Authors noticed that the MMM filter is much effective for noise removal.