Learning Forum

[Átyla Matheus]

I have a question about my results. I am simulating a fluid injection into a porous medium with another resident fluid. I used ANSYS CFX and these were the results. I don't understand why these preferential paths are being formed, because the porosity is the same throughout the radius. I made several refinements to the mesh, and the greater the number of elements, the greater the number of these preferred paths. Since the porosity is the same throughout the geometry, should these preferred paths form?

3.5millions mesh 4 millions mesh4.7 millions mesh 5.2 millions mesh

[NickFL]

Can you plot the lines for the cells on these? Are you also using higher-order numerics for ALL of the equations? And I would assume this is laminar flow conditions, correct?

[Átyla Matheus]

Yes. Higher order numbers for all equations and laminar flow. I don't understand these preferred paths because the porosity is the same throughout the domain (isotropic)

[NickFL]

My first point was about your mesh. Based upon the images, my expectations are that it is a tet mesh. I will ask again, can you plot the contours with the mesh superimposed?

It is clear what your question is, but we cannot simply answer it without more information.

[Rob]

To add to Nick's question. Is that a multispecies or multiphase (VOF) model? 

[Átyla Matheus]

I don't know how to plot these contours with the mesh superimposed. Could you guide me on this please?

I used Model Multiphase homogeneous 

[Rob]

I think it's an option on the surface/plane you put the contour on. I don't use CFD Post (or CFX). 

With the multiphase model, did you set any surface tension values? 

[Átyla Matheus]

Yes, I set the value to 0.072 N/m2

[Átyla Matheus]

Would it be like this? I zoomed in because the elements are too small and it wouldn't be possible to see them if the geometry was complete

 

[NickFL]

 

Very interesting that it is Hex elements, as I visualized it with tets seeing these “fingers”. What do pressure and velocity contours look like over the same area? I am a bit suprised that it takes 8 cells to go from 1 to 0. A sharp interface is expected, correct? What timestep are you using and after how many timestep is this? A part of me wants to say this could be errors building up from solution that does not converge well each timestep. What does the convergence look like?

One last thing I would try is decreasing the porous media resistance to see if they still develop in the same way.

The cheat would be to look at a quasi-2D model. Then you guaranteed not to get these ;)

EDIT: And how are you specifying the PM resistance?

 

[NickFL]

Congratulations, your problem kept my mind going all night. Here are some more thoughts:

• In regards to the sharp interface, what are you using the for the Interface Compression Level? The value of 2 should probably be used here to keep the sharp gradient.
• One possible cause of the fingers is walls placed at the outlet. Are you seeing any messages about this? This would be consistent with increased fingers with the increased element count.
• What are you using as an initial condition? The best solution would be to have a converged single phase steady state solution for Fluid 1 before starting the transient multiphase run. Any uncertainty in the initial condition could manifest in these fingers. You would specify the initial condition in the solver step.
• Perhaps the first thing to do is look at two different timesteps. Do the fingers grow in size, or once they are there they convect as a stable front?  This knowledge would help us identify other potential sources.

[Átyla Matheus]

I need some time to check your questions. I confess that I don't know how to answer most of them and I will look for ways to analyze everything you proposed, I'm still starting in the area and perfecting myself. Thank you very much for all your notes, they will certainly be of great value. I hope to come back with better answers and information.