MODELLING OF THE IMPACT WEAR PRODUCED BY DENSE LIQUID-SOLID SLURRIES

Many processes in the oil&gas industry involve massive quantities of solids being transported with the flow. When the plant components are hit by the particles they can experience loss of material, and this phenomenon, known as impact erosion, is often cause of considerable economic damages. At present, the erosion prediction follows a three-stages procedure, namely fluid flow computation, particle tracking, and application of an erosion correlation for estimating the material removal caused by each particle-wall impingement. However, the high computational burden of this methodology limits its application to dilute mixtures. On the other side, two-fluid models have been developed for handling dense fluid-solid flows, but they inherent structure preclude their coupling with an erosion correlation. For some years, our research has been aimed at developing innovative post-processing techniques for employing two-fluid models for wear estimation, with the final goal of allowing prediction of the erosion produced by dense mixtures. The present work fits into this research field. By reproducing numerically the impingement of an abrasive sand-water jet agasist a surface, a benchmark case for which experimental data available in the literature can be employed for validation purposes at least for low particle loading, we investigate the way in which the many submodels of a twofluid model affect the quality of the numerical results. This allows assessing the parameters which mainly influence the erosion prediction, in terms of both the identification of the area subjected to wear and the quantification of the loss of material.