Flow through porous medium

Submitted by Boyang on Wed, 05/30/2018 - 08:27

Hi CFDEM users,

I just started to use cfdem-coupling and try to simulate a pressure driven fluid flow through randomly packed spheres with certain distribution. The fluid is flowing from top to bottom with particle settling along with the fluid direction. The flow rate profile I've got seems a bit counterintuitive to me. I've attached the screen shot to this post. The flow rate appears to increase when it reaches the particle domain where I would expect volumetric flow rate would decrease when the fluid flow through the porous medium. I'm not sure where went wrong or if I interpret the result incorrectly.

The known variables would be inlet pressure and outlet pressure.

I modified the tutorial in tutorial/cfdemSolverPiso/voidfractionTest with the following boundary condition
0/p:
inlet {type fixedValue; value uniform 200;}
outlet {type total pressure; p0 uniform 0;U U;phi phi;rho rho;psi none; gamma 1; value uniform 0;}
0/U:
inlet {type pressureInletOutletVelocity; value uniform (0 0 0);}
outlet {type zeroGradient;value uniform(0 0 0);}

U is in the unit of m/s, I couldn't figure out how to output a reasonable volumetric flow rate which I expect would decrease as fluid encounter with more particles. I tried to specify inlet type as flowRateInletVelocity but the code won't even run, I also tried to add 0/phi with similar setup like 0/U but when I plot the results in paraView, it didn't have correct format.

I'm looking forward to your comments and suggestions,

Thanks and regards,

AttachmentSize
Image icon flow rate profile107.18 KB
alice's picture

alice | Tue, 06/05/2018 - 08:58

Hello Boyang,

by prescribing a pressure drop you prescribe a flow rate between inlet and outlet at the same time since the fluid is is incompressible. Due to the presence of the particles the volume available to the fluid is decreased, which means that the flow velocity will have to increase in this region.

Concerning the flowRateInletVelocity-BC: since you have prescribed a pressure at the inlet and the outlet, your system is allready fully determined. Adding another BC will lead to over-determination and a crash of the simulation.

Best,

Alice

Boyang | Fri, 06/08/2018 - 19:10

Hi Alice,

Thanks very much for the reply. Now I've learned the BC I tried to use would over specified the system and cause crashes.

A follow up question would be how could I output a volumetric/mass flow rate at the outlet patch. Is there some built-in function that does the work? I thought 0/phi would contain that information. Since I didn't find a exact example for this file, I created similar to the 0/U setup and add phi in fvschemes as well. I did get results but I can't seem to visualize it in paraview like all the others. I'll keep working on it.
I wrote a post-processing script averaging the velocity U across certain surfaces above the particle region and times it with the surface area A to treat it as volumetric flow rate. Is this a reasonable approach?

Thanks in advance!

Best regards,