Masterand
CurrentTitle: Developing Adaptive CNC Grinding Strategies based on Machine Learning ModelsMotivation: Saving time and money by automation of manufacturing processesKey Achievement: Reduced rework of multiple turbine blades, up to 5 working daysIn-depth description:CNC grinding of turbine blades & vanes has multiple advantages over manual grinding, like shorter cycle times, better reproducibility, and lower heat impact on the material.As Nickel-based superalloys are commonly known as hard-to-machine materials, finding optimal machining parameters is error-prone and time-consuming, if possible.Moreover, describing the process physically is challenging on all length- and time scales.Hence, the development of an ML model that predicts the resulting cutting depth based on the geometry of the part is needed. It will be used backward, that is, it will suggest machining parameters for a given geometry and local cutting depth.Finally, with the complete 3D data of a part, it can be ground optimally to the customer's nominal geometry.