Learning Forum

[rbaheti]

Hi,

I'm running a simple fluid simulation in a porous domain where I have a box with 20 inlets extruding in from a face and 20 outlets on the opposite face (See image below). Assume left side is inlet face and right is outlet face.

[Amine Ben Hadj Ali]

What do you mean running again? Starting from same data? Same version on same number of core?

[rbaheti]

Running it with the same initial data, same version of ANSYS (workbench 2020R2) but not sure about the number of cores.

[Rob]

You initialised both and ran, or ran the second model from the result of the first? Pressure in & pressure out makes the mass flow part of the solution so it can be difficult to converge, given porous media will then add to the effect you may need a better starting point. Note, those resistance coefficients are very small, where did you get them from?

[rbaheti]

I ran 2 different modules on 2 different workbenches.
I agree with the Pressure in and out boundary conditions - I came to the conclusion that for a given pressure inlet and outlet boundary conditions may not converge to a mathematically sound solution. I'm thinking of implementing a mass-flow inlet and outflow outlet BC. I'm not a huge fan of this combination because: 1) Not sure if this combination will be a good fit. 2) I won't be able to get values at each outlet face during postprocessing (or so I think).
As for the resistance coefficients - My end goal is to perform this type of simulation in an organ where the perfusion coefficient is extremely low (some references say these coefficients are even lower than the ones I've used: in order of e-13).

[Rob]

Don't use outflow, it's for incompressible flows and I think I last used it in Fluent 4. For compressible flow mass in and pressure out are good, for incompressible flow velocity in and pressure out are favoured.