Monitoring and Model Generation for Intelligent Optimization and Control of Surface Grinding Processes

Abstract

The grinding process involves more variables than most of the other machining processes. In the past, grinding process has been viewed as an art more than an exact science. This paper presents a monitoring and model generation strategy developed to allow science-based optimization and control of the grinding process. The monitoring solution involves simultaneous acquisition of power, forces, acoustic emission and vibration data generated during surface grinding. A custom build data acquisition program helps capture the information and visualize the process condition. Dressing consistency and spindle condition are monitored through the same system. Part of the data is processed off-line to determine coefficient values for generalized equations that model main monitored parameters. An optimization relative to cycle time or cost can be conducted based on the results gathered for each combination of grinding wheel, workpiece material and metalworking fluid. The procedure requires a minimal number of experimental runs to determine the model coefficients. The solution opens the path towards the development of a model-based condition monitoring system with adaptive control.