CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   SST Model for Low Re ... some minor settings (http://www.cfd-online.com/Forums/fluent/111887-sst-model-low-re-some-minor-settings.html)

mrenergy January 17, 2013 06:55

SST Model for Low Re ... some minor settings
 
Hi everybody,
I am investigating the Low Re flow over a family of thin 2D Airfoils ...
Thickness from 0 to 5 %
Camber from 0 to 10 %
Re from 70 k to 200 k
AoA from 0 to 12 deg.

I got a mesh quality of about 0.8 in most cases.
I validate my computation work using some famous airfoils, such as SD-7003 and NACA 4412

I explored SST, k-w, S.A., and k-e models, and I found that they are as I organized them from the best to the worst.
I used the Turbulent Intensity as 0.1% as I found in the literature (Experimental)

But I still have an under estimated values for both Cl and Cd by about 15 % (for SST model)

Are there any other parameters to be used in validation?
How can I improve my results?
What about the Turbulent viscosity ratio?
Any suggestions or Advises?

thanks in advance

Best Regards for all

Mamdouh

cfd seeker January 17, 2013 13:29

Use turbulence intensity and turbulence length scale as the option for turbuelnce conditions. Estimate turbuelknce length scale as 0.4*$ where $ is B.L thickness. Use flat plate B.L theory to approximate the thickness of B.L and then reduce it by an order of magnitude.

mrenergy January 17, 2013 22:08

I appreciate you help drer CFD Seeker,

although ... I use a values 0.001 , 0.005 , and 0.01 for the turbulence length scale as a near guess ... it works good ... I reached about 95 % of the experimental values.

How can I use the order of magnitude?
Do you suggest any other parameters to be useful in validation rather than Cl and Cd?

thanks a lot

Best Regards

cfd seeker January 18, 2013 00:57

Quote:

How can I use the order of magnitude?
probably you didn't get my point. I said reduce it by an order of magnitude e.g if the B.L thickness using flat plat formula comes out to be 25mm and reduction in order of magnitude means 15/10 =1.5 mm. Use this value, 0.4*$= 0.4*1.5 =0.6 mm as the turbulence length scale value.

BTW 95% agreement with the experimental results is excellent.
Quote:

Do you suggest any other parameters to be useful in validation rather than Cl and Cd?
No. When I was doing this study some time back, I noticed the gigantic reduction in Cd value when I use the turbulence length scale value of 0.4*$.

BTW

mrenergy January 18, 2013 02:08

Done ...

thanks for your concern
thanks for your time
thanks for your help

Best Regards and Respect

Mamdouh

mrenergy January 18, 2013 02:28

multiple FLUENT runs ???
 
excuse me dear CFD Seeker

I have hundreds of runs with almost the same settings, are there any option in ANSYS FLUENT to perform successive runs for different cases?

I mean to repeat a certain loop starting from loading the mesh file, setting the BCs, choosing the model ... until saving the case ?

to saving time and effort, and to avoid human errors ...
I used ICEM reply control to do that in the meshing process.

I hope ...

Best Regards
Mamdouh

cfd seeker January 18, 2013 11:07

well set all the BC, solver settings, models etc and save the case file. For every run just change the Mach No and AOA for that case. This is the shortest possible one as per I know.

mrenergy January 18, 2013 21:13

thank you
very much


All times are GMT -4. The time now is 11:40.