An implementation framework for condition-based maintenance in a bearing ring grinder

Abstract

The demand on a bearing ring grinder, as any other machine tool, is to produce parts as per the specification and desired quality. A failure to achieve the quality can be due to functional issues or mechanical failure modes. Ultimately this results in lower productivity and higher production costs. Despite the increased emphasis on practicing condition-based machine maintenance (CBM) in manufacturing applications, it is still considered a challenge to fully deploy CBM in production machines due to diversity in equipment and variety in machine configurations as well as complex characteristics of failure modes. Although there exists extensive literature on CBM for machine tools and subsystems, the issue remains with realization of a technically capable and cost effective CBM system, specifically for a bearing ring grinder. Therefore, sensor(s) selection, data acquisition setup, data processing and analysis are the essential factors considered in the proposed framework to ensure a systematic and organized CBM implementation. The CBM setup is evaluated against production of bearing rings under different process and failure conditions. A machine type independent data acquisition system is designed to capture both machine and process dynamics. The data gathered from sensors at strategic locations exhibits its effectiveness in capturing the process and condition variations in relation to time and operating modes. The presented results of data analysis support the capability and effectiveness of the proposed framework. The utility of this framework can be extended for any number of scenarios including predictive maintenance or adaptive process optimization where solutions using machine learning and artificial intelligence tools can benefit from high dimensional structured dataset. The proposed framework provides a strong foundation to fast track the adaption of CBM in other production machines having similar subsystems.