CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   Very Low Re: SST vs. v2-f? (https://www.cfd-online.com/Forums/main/80428-very-low-re-sst-vs-v2-f.html)

CAVT September 25, 2010 04:34

Very Low Re: SST vs. v2-f?
 
Hello all.
I'm simulating the flow in an air-conditioned room having an extractor hood with reposition, in order to see the how much conditioned air is lost. So far I have made simple 2D transient simulations with laminar flow and then with SST turbulence model. I tried doing simulations with the v2-f model but it's harder to converge and there's no point in going so refined for a 2D simulation.
Now that I'm going for 3D, I would like to know if for these kind of flows an SST model would suffice, or if a v2-f model is more appropiate. According to the theoretical descriptions and some test results from other publications, the v2-f seems to be more appropiate, but I need a quite low y+, not so easy to achieve in my geometry if I want to keep simulation times down, and there the robustness of SST is tempting. I'm doing the runs in my PC, which is not exactly a number-cruncher, and I don't have access to a cluster.
The maximum flow velocity magnitudes are going to be less than 2 or 3m/s, and I also simulate convection with temperatures varying from 150║C (a heat source) to 20║C (intended air-conditioned temperature). I'm using Code_Saturne, and so far the 2D laminar and SST results are promising, even with coarse meshes and not differing much one from the other. My biggest problem is that I don't have experimental data to contrast with; I was only told "you should obtain no more than 10% or 15% of conditioned air lost", which is even an estimation, not sustained by measured quantities (despite that, I'm in that order of magnitude).
Such a low Re would indicate that a laminar attemp should suffice, but that's hard to afirm in reality given that in the ambient there will be people and stuff constantly in movement, dirt and who knows what else, that's why I suppose giving a try to turbulent models is justified, specially to see differences in convective flow. I don't need a great numerical accuracy, since I don't even have data to compare with. So... SST, v2-f (called phi-k in Code_Saturne) or some other?.
Thanks in advance.

CÚsar


All times are GMT -4. The time now is 07:12.