Advancements in the slurry erosion predictions based on Computational Fluid Dynamics

Slurry erosion experienced in many engineering fields, including mining and petroleum, results in time wastage during plant construction and repair, and general high cost of maintenance. Such erosion occurs when surfaces are exposed to solid particles in slurries. Recently, researchers have made significant efforts towards the development of models that are capable of predicting slurry erosion. The step will ensure improvement in the design, maintenance, and management of the plants.

Various models for predicting slurry erosion have been developed. Models based on Computational Fluid Dynamics have been used to predict both simple and complex liquid-solid and gas-solid erosion processes. These models quantify the wear process by taking into consideration the abrasive particles and the target surfaces, and they rely on erosion correlations that express the amount of the eroded material removed as function of several fluid-dynamic and material-related parameters. However, the models developed so far are often unsuccessful in predicting accurately the slurry erosion for many reasons, including the difficulties in properly accounting for the self-induced changes in the geometry and the complex interactions between the phases. Furthermore, most of these models are based on Eulerian-Lagrangian approach, thus resulting in complex algorithms and computations. The Eulerian-Lagrangian approach employs an Eulerian description in solving the fluid flow and tracked trajectories of a given number of particles in simulating the solid phase.

Gianandrea Vittorio Messa and Stefano Malavasi from the FluidLab research group at DICA Politecnico di Milano in Italy developed an effective numerical methodology for more accurate prediction of the evolution of the slurry erosion process. The proposed strategy was applied to the benchmark case of abrasive jet impingement test and the results compared to those obtained by the traditional method, which assumes steady-state flow. Their research work is published in the research journal, Wear.

Briefly, the proposed technique involved the use of the Eulerian- Eulerian model for the steady-state slurry flow simulation and restricted the calculation of the trajectories of the individual particles, as obtained by solving the Lagrangian motion equations, only to the proximity of the solid walls. All the phase interactions occurring in the dense flow were well captured by the Eulerian-Eulerian model, hence eliminating the use of a time consuming Eulerian-Lagrangian model accounting for inter-particle collisions. An erosion correlation allowed displacement of the eroded surfaces before tracking new particle trajectories. Eventually, the authors made specific but reasonable assumptions concerning the slurry flow structure during the erosion process, which allowed eliminating the need for repeated and time-consuming solutions of the Eulerian-Eulerian equations.

The developed model improved the accuracy of the numerical simulation of several abrasive jet impingement experiments. Additionally, it was also efficient in capturing the effects associated with the complex interactions in the slurry flows, and both the used Eulerian-Eulerian model and particle equation of the motion contributed to this achievement.

From the several reproduced slurry abrasive jet experiments, the developed model resulted capable of correctly predicting the deceleration of the slurry erosion process as well as capturing the decrease in the erosion ratio as the concentration increases.

Generally, the method showed improvements in the slurry erosion prediction as compared to the other commonly used ones as it effectively describes the physical mechanisms involved in the slurry transport and erosion processes. Being also time-saving and less complex in term of computational effort compared to the standard methodology, its application to complex flows appears very promising. Therefore, it may be used in real life practice for various engineering fields.

Comment from the authors

This study is part of a broader research area in the FluidLab group at DICA Politecnico di Milano, whose long-term and ambitious goal is the development of methods for more accurate prediction of slurry erosion in hydraulic devices, such as valves and machinery, based on a synergy between laboratory experiments and numerical simulations. The authors are particularly proud of this work because, in addition to improving the reliability of the erosion estimates, the developed method allowed gaining more insight into the physical phenomena underlying slurry erosion and better interpreting the experimental results. Authors are very grateful to the Editors and the Reviewers of the Wear journal for the attention and for their useful comments, and to Advances in Engineering for highlighting this research.