Hi
In eulerian multiphase model, it is assumed that mass and momentum are transferred on a cell-basis.
So, if cell size is smaller than particle size, it means that the math is incorrect. But there have been studies in literature where authors have used smaller cell size than the particle size?
When I try to do that in Ansys Fluent, my simulations don't converge even with a smaller time-step size compared to the authors. e.g. the cell size is 2 mm, and the particle diameter is 5-7 mm.
Article Improved numerical approaches to predict hydrodynamics in a ...
Article Numerical study of hydrodynamics with surface heat transfer ...
The problem is there are two solid granular phases whose size difference is huge. so cannot make a coarse grid to solve for smaller particles.
Can anyone please tell, how can I solve this problem.
I have tried reducing relaxation factors, double-checked everything including boundary/initial conditions.
Thanks
I think there are some discrepancies about the grid cell size when Eulerian-Eulerian is used. However, I think you can use grid size smaller than your bubble size as selected control volume can be different from selected numerical cells.
But, for 2D simulation :
The cell size larger than a bubble diameter is required for 2D modelling and for 3D finer grid sizes can be used but more study is needed. (Homogeneous regime- FLUENT was used as the solver) (Robert Picardi-Thesis- Virginia Polytechnic Institute and State University).
Best,
Siamak
It is; the Euler-Euler model provides an ensemble-average of the phase holdup in each gridcell, without any reference to bubble size whatsoever. You can interpret it as "given the flow conditions, what is the average amount of gas I will find at this location, considering all possible realizations"
- Badges
- Science topic