Multiscale simulation of volumetric wear of vitrified alumina grinding wheels

Abstract

Wear of vitrified alumina wheels is a major research issue. Although extensive experimental work has been presented in the scientific literature, wheel manufacturers and grinding users are still unable to predict the expected grinding ratio of their operations. This work presents a novel multiscale approach that integrates the mechanical behavior of vitrified bonds with the stochastic nature of grain location. The new multiscale model integrates a DEM microscale model (μSM) and a randomization of the μSM (RμSM) of the wheel. The μSM simulates the stress field in the region of the wheel in contact with the workpiece, while RμSM accounts for the actual and random location of the alumina grits. This approach drastically reduces computational time and effectively determines the actual number of grains lost under a set of given grinding conditions. The model successfully predicts the detachment of complete clusters of abrasive grains from the bonds, with a large degree of agreement with experiments considering the deviation introduced by the micro-cutting edges that modify the hypothetical spherical geometry of the Discrete Elements (DEs). This is the first model to predict the volumetric wear of grinding wheels. The results will be useful for informing the future analysis of grinding operations affected by volumetric wear.