BARKHAUSEN: A study on laser-processed grinding burn simulation and analysis based on Barkhausen noise measurement

Abstract

Barkhausen noise (BN) is widely used as a quality control method for grinding burn detection in ferromagnetic materials such as the case-hardened steels used in gears. The study of grinding burns with BN equipment requires, however, a calibration reference sample with a known BN response. The demand for better calibration samples is well recognised. In this study, artificial grinding burns with laser processing were produced in case-hardened steel. The laser was selected because it is a surface-sensitive heating method and it accurately simulates the heat generation during grinding between the grinding wheel and work-piece. A test series was made by changing the laser input power and traverse rate to process the surface of a ground gear wheel sample. The laser-processing parameters were optimised experimentally. The BN measurement data was used in the modelling of selected quality measures: hardness and residual stress (RS). The BN measurement device provided some features from the time-domain analysis of the signal and a new feature, BN entropy, was also used. The modelling was done with Matlab® using linear regression. The results showed that laser processing had a clear effect on the BN features and residual stresses.