Mass not being Conserved in Porous Solids

[MDMeredith]

Is there a response to my earlier thread 'Mass Not Being Conserved in Porous Solid"? 

[srhusain]

Hi:

Conservation of mass may not be exact but vanishingly close as your solution proceeds to a fully converged solution. You first objective is to set up the model as a well-posed problem and take the calculation process to tight convergence.

If your analysis fails to converge, mass balance locally or even globally will not be satisfied. For your problem setup to be numerically well-posed you need to observe the following:

1) The boundary conditions have to be compatible. For example, you cannot specify pressure and velocity at the same boundary for incompressible flow.

2) For a contiguous set of parts belonging to the same fluid, if you have flow specified one one boundary, at least one other boundary must have pressure specified.

3) Avoid placing distributed resistances at or near the boundary of the flow domain.

4) Mesh- ensure that the mesh size does not transition from one element to the next. For example, do not allow a size transition with a factor greater than 5, especially in the pricipal direction of the expected flow. Also, ensure that the element aspect ratio is less than 10 away from the boundary layer.

Hope this helps.

[MDMeredith]

I've had time to return to this problem again.  I have simplified the geometry to "catch the problem in the act."  Air flow downward at a uniform bounday condition of 0.1 ft/sec across a gap of about 1 foot then encounters a bed of resistance.  Resistance is permeqability type with a coefficient of .0001 sq mm inall directions.  Direction of flow is Global Z.  After 6 feet of bed it again goes through a gap of a foot to discharge at a constant total pressure of 0 psi gauge.

As you can see from the attached screen shots the velocity increases near the center.  The overall mass increases from 3018 lb/hr to 3587 lb/hr.  The system is well converged after 5000 iterations

[srhusain]

Can you post the support file from the successful run?

Thanks

Syed

[MDMeredith]

It has changed from the one I sent pictures of in that I switched to compressible which reduced by did not eliminate the creation of matter.  I also filled in the central pipe with solid.  Previously I had put reistance material in the central pipe.

The purpose of the exercise to to distribute gas evenly to a permeable bed.  The example given here has no reason not to be perfectly uniform.

Thanks for your help

[MDMeredith]

Tell me how to post the file.  I get an error when I try to attach the cfz share file.

[srhusain]

rename the "cfz" extension to "zip" and retry posting