Determination of CMC Grinding Processes with Reduced Workpiece Damage through Data-Driven Process Analysis

 

Ceramic Matrix Composites (CMCs) are innovative materials with high growth potential in the aerospace industry. Due to manufacturing-related geometrical inaccuracies, grinding of the workpieces is required in many cases. The influence of the process input variables during grinding (grinding process parameters, grinding tool) and the fiber orientation of the workpieces on the machining behavior of these composites and the resulting surface qualities of the workpieces have been insufficiently investigated. Due to the lack of knowledge of the material-specific machining behavior, the reproducible grinding of damage-free fiber-reinforced ceramics is only possible to a limited extent. This results in high costs in final workpiece testing and grinding-related scrap.

The acquisition and evaluation of grinding process data from the machine drives and external sensors offer the potential to identify damage during grinding. In this context, machine learning models enable both the classification of components into quality classes and the evaluation based on concrete damage parameters.

The aim of the project is to develop a method for data-driven damage identification in the grinding of CMCs, thus enabling process design to be optimized and scrap to be reduced.