IC11 - KI4SF

The plunge finishing process offers enormous advantages in the machining of complex geometries by grinding the entire component surface with loose abrasive granules. Complex machine kinematics, as would be necessary with conventional processes, are thus eliminated. Automated optical monitoring of the grinding machining of components is an important advance that can help improve the quality of machining and increase productivity.

Aim and approach

The key points for optical monitoring in stream finishing operations of components include selecting suitable measurement techniques, identifying component geometry, monitoring process parameters, integrating process data into a monitoring system, and implementing algorithms for early detection of anomalies in the grinding process.

• For this purpose, necessary measurement setups, instruments, and data infrastructure are employed according to the requirements, enabling automated identification of components and detection of geometric features through intelligent image processing.
• Simultaneously, the initial state of the surface structure and possible defects on the surface are identified. The machining condition is iteratively monitored using a camera and stored along with the process parameters, allowing for the analysis of temporal correlations in the machining process.
• Finally, a simulation of the process is performed, enabling the creation of a large database based on data augmentation without the need for extensive experimental investigations. By combining all data from experiments, simulations, and process models, an overall process model is created that holistically represents the machining process and enables predictive predictions for "first part right" production.

Benefit

Evaluating data from optical monitoring using cameras and artificial intelligence contributes to gaining a better understanding of the grinding process and optimizing it. The collected data can be used for process modeling to enhance production efficiency and improve the quality of machining. The results can also be transferred to related grinding processes widely established in the industry (e.g., drag grinding, centrifugal disk grinding) and serve as a solid foundation for knowledge transfer in research and industry.