CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [GAMBIT] High aspect ratio (http://www.cfd-online.com/Forums/ansys-meshing/80075-high-aspect-ratio.html)

RGRUIZ September 14, 2010 10:53

High aspect ratio
 
Hi.

I am working determining transition point over an airfoil using SST Transition model.

I have been doing the mesh with gambit (2D airfoil) but when I export the mesh to fluent it says: HIGH ASPECT RATIO (over 615 times). I have been reading some papers and stuff and everyones agree on the fact that the aspect ratio should be between 1 and 10, is that correct??.

I tried everything but i think i have some limitation on dimensions on C-grid.

If anyone could help i would be great, i can send you pictures of it.

-mAx- September 15, 2010 01:24

Once you loaded your mesh into fluent, do a check mesh, and see if you also get any warning regarding aspect ratio
Else in Gambit go to Examine Mesh.
Select Range as Display Type, then Aspect Ratio in Quality Type
Enable Show Worst Element option.

RGRUIZ September 15, 2010 08:10

Quote:

Originally Posted by -mAx- (Post 275249)
Once you loaded your mesh into fluent, do a check mesh, and see if you also get any warning regarding aspect ratio
Else in Gambit go to Examine Mesh.
Select Range as Display Type, then Aspect Ratio in Quality Type
Enable Show Worst Element option.

Hi MAx, thanks for taking your time to answer me

I've done that and i have seen the worst element, and i have modified the edge mesh parameter so i can have a good aspect ratio but still have some problem. I need to have a y+ near 4-5 because i have to calculate the sub-laminar viscous layer also. I will send you pictures of my mesh so you can see more details of what I am trying to tell you

best regards

-mAx- September 15, 2010 09:11

did you use Boundary Layer?

RGRUIZ September 15, 2010 09:40

No max i did not use that because i do not know how to do it..how do I do to make a boundary layer that follows the airfoil profile???

-mAx- September 15, 2010 09:45

in Gambit there is a Boundary Layer Tool.
Just select the edges from your airfoil for attaching the BL.
Check Gambit Online help, it is described how to use it (see also tutorials)

RGRUIZ September 15, 2010 20:13

Hi max, I have been trying to do what you said but to be honest, i have no idea how to do a boundary layer using gambit but I cannot attached some pictures about what i have done. do you have any mail so i can send you?? I am not familiar with this type of image attachment

Hope you can help with any advice so i can try to do it

jlichtwa September 16, 2010 00:20

Ruben,

You may want to first calculate the first layer viscous grid spacing. Using your flow freestream flow properties and mean aerodynamic chord, you can calculate the wall spacing for the first layer, for any y+ you desire.

Use a site such as

http://geolab.larc.nasa.gov/APPS/YPlus/

Once you have your initial spacing, and a growth ratio and number of layers (you want ~10 cells within the boundary layer), you can then use the boundary layer tool in GAMBIT to create a boundary layer.

If you are using a fully structured C-grid, you can get around using the boundary layer tool by supplying the initial height by selecting a mesh law, on a specified edge, with initial and/or final spacings option.

Hope this helps.

-mAx- September 16, 2010 02:57

Regarding attachment you have to "Quote" and not "quick reply". Then click on the paper-clip icon.
If you are interested I have a fluent tutorial about BL (pdf file)

sidspark12 September 16, 2010 05:08

hello max...

can u please send me that BL tutorial at siddharthjain.2007@rediffmail.com

RGRUIZ September 16, 2010 06:22

4 Attachment(s)
Quote:

Originally Posted by -mAx- (Post 275390)
Regarding attachment you have to "Quote" and not "quick reply". Then click on the paper-clip icon.
If you are interested I have a fluent tutorial about BL (pdf file)

thank you max...of course i am interest on it..my email is rgruiz@gmail.com...hope it can help with my 2d airfoil..thank you i attached images so you can see what i have donde...any suggestion you have about what i am doing is welcome..i am trying to improve everything i do...

best regards

RGRUIZ September 16, 2010 06:28

hi sidspark12

i attached some pictures about what i have done up to now...how you can help on HOW to do a boundary layer that can be attached to the rest of the geometry..

best regards

RGRUIZ September 16, 2010 06:29

hi max...thanks for the advice..i attached some pictures about it..hope you can explain me how to attached a boundary layer to may existing mesh..

best regards

-mAx- September 16, 2010 06:40

I send you the pdf file on your email. It should be sufficient for understanding how it works (very simple)

RGRUIZ September 16, 2010 06:57

Hi max what do you think about the pictures i published..

-mAx- September 16, 2010 06:59

I would refine the mesh at the trailing edge

RGRUIZ September 20, 2010 10:33

Hi max


Thank you so much for the tutorial...I have done the boundary layer tutorial, also I have problems with the convergence of results on fluent but that's another thing to do.

I would like to asked you about why most of the people split the edges of the airfoil on x=0.3c..is this applied to any airfoil?? is that a rule??? Any recommendation????

Another question, a good aspect ratio should be review only on the airfoil?? I mean if I have an aspect ratio of 10 on the airfoil, should I not care if I have an aspect ratio of 800 on the farfield??

Hope you can help me

Best regards

-mAx- September 21, 2010 02:24

You can try runnng your stuff with aspect ratio over 800.
Do a check Mesh in Fluent, if it is ok, then compute.
It is possible you get convergence issur.
But I am not experienced in airfoil.
Splitting edges at 0.3c may give your the possibility to plot chart on specific edges (ending at 0.3c for instance)
But maybe there is another reason, for instance on the meshing side.
But once again, I have no experience


All times are GMT -4. The time now is 06:36.