Using Multivariate Quality Statistic for Maintenance Decision Support in a Bearing Ring Grinder

Grinding processes’ stochastic nature poses a challenge in predicting the quality of resulting surfaces. Post-production measurements for form, surface roughness, and circumferential waviness are commonly performed due to infeasibility measuring all parameters during grinding operation. Therefore, it is challenging diagnose root cause deviations real-time from variations machine’s operating condition. This paper introduces novel approach predict overall individual parts. The grinder equipped with sensors implement condition-based maintenance induced five frequently occurring failure conditions experimental test runs. crucial measured produced Fuzzy c-means (FCM) Hotelling’s T-squared (T2) have been evaluated generate labels multi-variate data. Benchmarked random forest regression models trained using fault diagnosis feature set labels. Quality T2 statistic preferred over FCM their repeatability. model, achieves more than 94% accuracy when compared ring disposition. predicted sensors’ against threshold reach trustworthy decision.