Integration of process monitoring and machine condition diagnostics to improve quality prediction in grinding

Abstract

Bearing ring grinding incorporates sensors to control the grinding cycle in real time. Prediction of output quality is difficult due to the complex combination of process settings and machine characteristics. Causal relationship of machine performance with varying operating conditions was studied with reference to the produced quality by adding condition monitoring setup to the machine. Data driven diagnostics of machine condition through integration of condition and process monitoring sensor data at the completion of the grinding cycle improves quality cognisance. This can be used to tune control parameters to achieve more predictable quality in successive cycles.