Data-driven identification and explanation of process changes due to batch fluctuations on the workpiece and tool side during surface grinding

• Applicant: Professor Dr.-Ing. Thomas Bergs
• Subject Area: Metal-Cutting and Abrasive Manufacturing Engineering
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 552756260

The process state and result variables of grinding processes are influenced by a multitude of process input variables. In order to explain the cause-effect relationships between the process input, process state and process result variables, mainly empirical and numerical models have been developed in previous research work in the field of grinding technology. A disadvantage of these models developed so far is that their validity is limited by the tool and workpiece used in each case, the test plan-specific process parameters, the cooling lubricant used and possibly also boundary conditions such as the machine tool, the ambient temperature or the machine operators. A change in the listed process input variables leads to a change in the process state and result variables and therefore requires a new or extended empirical or numerical process model. Since batch variations of the tools and workpieces used are not taken into account within the previous empirical or numerical models, the transfer of the models into practice is only possible to a limited extent. As a result, deviations between measured and modelled results often occur in grinding processes with workpieces or tools of the same specification. The systematic recording of workpiece- and tool-side batch fluctuations in the process input variables during surface grinding and their effects on the process result variables have not yet been researched. As a result, it is not known to what extent workpiece- and tool-side batch fluctuations cause a change in the thermo-mechanical load of the surface grinding process, can be identified by data-driven methods and resulting deviations in the process result variables can be explained and predicted. The aim of the present research project is therefore the development of a combined model for the identification of workpiece- and tool-side batch variations in grinding as well as for the explanation and prediction of the resulting process deviations. The research project is based on the quantitative characterization of tool- and workpiece-side batch variations by both material and topography analyses. Dressing and grinding investigations are used to describe and explain the changes in the process state and result variables as a function of the tool and workpiece batch. The process-parallel identification of the tool-side batch variations is made possible via the data-driven evaluation of machine control and sensor data during dressing. Based on this, the workpiece batches are identified in grinding investigations in a data-driven manner and the batch-dependent process result variables are predicted.

• DFG Programme: Research Grants