'Test DMP Cyclone' simulation project by zakharov


#1

I created a new simulation project called 'Test DMP Cyclone':

Test DMP Cyclone


More of my public projects can be found here.


#2

@zakharov, glad to see you are using discrete phase model.

DPM simulations might be a little tricky to start with. Therefore, for anyone who is interested, I’ll add a few suggestions regarding similar DPM setups. These are general settings to ensure stability of the run to reach the interesting part, e.g. here 0.2 seconds when particle injection starts.

  1. It would help to use multigrid (GAMG) for pressure and particle cloud solvers. You could tighten up tolerances for better convergence, specifically for final solvers, e.g. pressure final solver.
  2. Similarly, use the smooth solver with a symmetric Gauss-Seidel as a smoother for velocity and turbulent quantities.
  3. Default numerical Schemes are mostly fine. Here, consider using linear and/or upwind divergence, and linear corrected laplacian schemes.
  4. Boundary conditions are correctly setup. However, if none of the above helped, you should consider using fixedFluxPressure for pressure wherever flux is set, e.g. here inlets and walls. To do that, you should use a custom BC though.

I hope this helps. Once the simulation reaches the point where particles are injected, we could see how it behaves. We will then focus on the actual DPM setup, if further tweaking was required. Let’s see.


How simulate DPM for Pressure-Swirl-Atomizers
#3

@zakharov, just to be sure, I tried your case with the improvements I mentioned above, and it successfully ran past the point it crashed before. Feel free to share a link to the project with your new runs to see what the problem is.


'sepor' simulation project by zakharov