Efficient parameter estimation for grinding force models using a Bayesian approach

To prevent mechanical damage caused by process forces when grinding turbine blades for aerospace applications, the prediction of process forces is of major interest. Macroscopic grinding simulations utilizing suitable force models are a common method to predict process forces. However, the prediction accuracy of the simulation highly depends on determined model parameter values. In order to enable an efficient and precise model parameter estimation, this study presents a Bayesian Markov Chain Monte Carlo (MCMC) method based on hybrid data sources. By only using few measurements combined with simulated data, force model parameters are conditioned leading to accurate simulation results.