[skabilan]

Hi All,

I need some clarification with the Kepsilon turbulence model. I have computed a steady state solution for the following geometry (Figure A) using simpleFoam with Kepsilon model. The boundary conditions include a zero pressure at the inlet and a pressure of -10 pa at all the eight outlets. There was no problem with the convergence. On investigating the results, the pressure at the cross section AA` was -3.91 pa and the mass flux was 1.77e-04 m^3/sec.



I have another geometry (Figure B), which is a subset of the aforementioned model. When a pressure of -3.91 pa was applied at the two outlets with no change in any of the boundary or the solver settings, the expectation was that the massflux would be ~1.77e-04 m^3/sec (at the same location). Instead it was 1.94e-04 m^3/sec. So I fine-tuned the pressure at the outlets to match the massflux of 1.77e-04 m^3/sec. The final pressures at the two outlets which yielded a massflux of 1.77e-04 m^3/sec was – 3.47 pa.



My question is… why do I see a difference of 11.03% in pressure for the same massflux and the same geometry? Am I missing something? Any thoughts or suggestions?

Thanks
Senthil Kabilan

[olwi]

Hi Senthil,

One guess: If you look at the velocity distribution in the outlets of Geometry B, and compare with equivalent cross-section in Geometry A, the velocity distributions in the cross-section are probably not the same.

I am not chocked by the differences you see. The mathematical properties of a boundary condition does differ from that of an arbitrary cross-section in the middle of a developing flow. As a general rule, boundary conditions should be placed some distance away from the regions you're really interested in, and/or in a zone of well-developed flow.

/Ola

[skabilan]

Hi Ola,

Thanks for the feedback! So there is nothing wrong in what I see. It is just the way it is!

Hrv: Thanks for the email.

Regards,
Senthil