[jallison]

Hi,

I have a pressure based, steady, wall bounded flow, with pressure inlet and outlet, where flow is driven by radial blades fixed to rotate in a fluid volume in a rotating frame. Inlet and outlet are both at ambient pressure. The walls are smooth, no-slip, with no-shear walls at the inlet and exit to give a calming length. Flow enters axially then is driven radially outward through a high aspect ratio channel. The program uses the SA turbulence model. When the walls bounding the channel are narrowly spaced, the mass flow rate matches experimental results. However, when the walls are widely spaced, the Fluent mass flow rate is about 30% less than the experimental mass flow rate. I believe the error is in the Fluent model, but despite checking monitors and residuals and allowing the models thousands of iterations to flatline, the mass flow rate is not coming out as expected.

I have attempted a laminar model which has the residuals stuck in a limit cycle before converging and k-omega which also seemed to display some sort of limit cycle, but was too computationally expensive anyway.

I have also attempted refining the mesh, with ~20 grid points between the wall and the rotating blade. This gives a mesh of 3M cells, and I can't afford any more.

Does anyone have any thoughts as to what could be causing this discrepancy?

Thanks, Jon

[tanaji]

Hi Jallison,
I am also facing same problem of flow rate mismatch.

Let us know, if you have got the solution to fix this problem.

Thank you