Analysis of local wear variables for high-precision erosion modelling in complex geometries

Abstract

Particle-laden flows in complex geometries often require a more robust performance from erosion models in order to obtain accurate prediction of wear. Most erosion models developed over the last six decades have not been able to achieve the desired accuracy needed for these complex systems, primarily because they are based on single particle-wear material interactions, experimental correlations, and often do not capture the full physics of the erosion process. Recently, the capability of computational fluid dynamics (CFD) in resolving fluid-particle-wall interactions more realistically has been shown to have great potential in the development of high-performance erosion models, especially when combined with experimental data analysis. This paper adopts this combined CFD-experimental methodology to improve the prediction performance of some selected erosion models. Experimental data for the wear of an elbow in gas-solid flow from a previous study was combined with new CFD predictions of local wear variables such as particle impact angle, particle impact velocity, and particle mass rate. A geometric function is developed which can be combined with some common erosion models to significantly improve their wear prediction accuracy for gas-solid elbows. Also, the effects of target material surface roughness and particle rotation on local wear variables are investigated numerically. The simulation results reveal an increased dispersion of particles and a more even distribution of local wear variables at the elbow. In conclusion, this paper presents a method for improving performance of erosion models, and potentially a method to translate wear data between complex geometries.