Predictive grinding process optimisation and monitoring

Grinding is one of the oldest and most important metal removal processes, and is capable of high dimensional and surface finish tolerances. It is a complex and expensive process; industry has much to gain by increasing production rates to reduce cost. The major limitation to higher production rates is the risk of thermal damage of the workpiece. This is now being challenged by developments in “High Efficiency Deep Grinding” which has been proven to produce low grinding temperatures at extremely high material removal rates. In order to take advantage of these developments, whilst maintaining the integrity of the workpiece, it is necessary for production engineers to have tools available to them that allow the selection of optimal process parameters and monitor grinding conditions to sustain this optimum. A review of current research efforts in predictive and reactionary methods of optimising grinding process highlight a number of failings. This study leads to the development of a new system that employs analytical and empirically derived indicators of thermal damage to enable an operator to select optimal but safe grinding conditions. The system also provides a monitoring function that can warn of the onset of thermal damage and make recommendations to the machine operator. A demonstration of the systems possible benefits in an industrial context is presented. Validation via simulation is also performed. Predicted finished workpiece temperatures are compared against measurements taken using embedded thermocouple and the PVD coating melt depth method. The ability of the system to predict bum is also tested across a range of grinding conditions. The possibility of using the system as part of an adaptive controller is also reviewed and directions for further work are identified.