This article discusses the Pros/Cons for different approaches that can be used to model erosion.
Erosion is a common phenomenon for many industrial applications. Repetitive impact of solid particles or liquid droplets can erode material resulting in reduced life time of components. There three are modeling approaches that can be utilized to model erosion, this includes Lagrangian Multiphase (LMP), Eulerian Multiphase (EMP) and Discrete Element Method (DEM). Each method has its benefits and limitations. Choosing the correct modelling approach can be vital in achieving fast and accurate results.
This article provides guidance on choosing the correct modeling approach by describing the pros and cons of each approach. See related articles section for best practices articles for EMP and LMP approaches.
The Eulerian multiphase models the particulate flow as dispersed phase. It captures particle-particle interaction and is not limited by particle loading. The mesh cell size is governed by the particle size. Since the particles are modeled as a dispersed phase, the particle size should be smaller than the cell size.
Pros |
Cons |
| Computationally inexpensive | Particles are not modeled explicitly but as dispersed phase |
| Includes the effect of particle-particle interaction. | Mesh cell size limited by particle size. Recommended Cell size/ Particle size = 5~8 |
| Not limited by particle loading | Can be challenging to converge for large particle sizes as it would require coarser mesh. |
| Provides quick results. Suitable for design exploration studies. |
The lagrangian multiphase can be used to explicitly model the particles. This approach doesn't captures particle-particle interaction and therefore is limited by particle loading.
Pros |
Cons |
| Explicitly models solid particles. | Can only predict impact wear. It will not capture scouring caused by abrasive wear. |
| Computationally inexpensive and can be modeled as steady state. | Does not captures particle-particle interaction |
| Can be modeled as one-way coupled for low particle loading and small particle sizes. | Requires two-way coupling for high particle loading and large particle size. |
| Not suitable for cases with high particle loading. Recommended to use where particle volume fraction < 10% |
The discrete element method is the most expensive and accurate method. It explicitly models the particles, it captures particle-particle interaction and is not limited by particle size or loading.
Pros |
Cons |
| Explicitly models solid particles. | Can only be modeled as unsteady. |
| Predicts both impact and abrasive wear | Computationally expensive. |
| Captures particle-particle interaction and is not limited by particle loading or particle size. | More expensive for cases with high particle loading. Small and stiff particles will requires a smaller DEM time-step |