High precision suspension erosion modeling

To date, prediction of particulate erosion in suspension ows has been unreliable, with the state-of-the-art reporting errors upwards of 40% [Zhang et al., Wear 240 (2000) 40–51]. These errors stem chiefly from the underlying erosion models, significantly hindering understanding of erosion dynamics in suspension ows and design of plant equipment with improved erosion characteristics. Herein we present improved experimental methodologies and data analysis techniques based on computational fluid dynamics (CFD) to solve the associated inverse problem and generate high precision models of suspension erosion. These methods are applied to a test case of silica sand particles eroding an aluminium surface, and the resultant data is fitted to an erosion function. To test the predictive accuracy of the fitted model, CFD predictions are compared with an independent erosion experiment, and the peak erosion rate is found to agree to within 1%. These results suggest that the methodologies presented provide a sound basis for high precision suspension erosion modeling.