[appa]

I am trying to simulate a turbulent flow over a naca profile. I am working with both simpleFoam and pisoFoam solvers, but I didn't get any good results until now. Can someone help me please and tell me what should I do ???
http://www.4shared.com/file/Bdx2tB1G/blockMeshDict.html

[mihaipruna]

maybe you can attach the text file instead of (I assume involuntarily) sending us to a spammy link?

[appa]

Originally Posted by mihaipruna

maybe you can attach the text file instead of (I assume involuntarily) sending us to a spammy link?

it is not a spammy link
here is the blockMeshDict file, I put it in the .c format so as to be able to attach it here.

[mihaipruna]

I think it would be easier to create an STL of the airfoil with a bit of thickness (s stubby wing) ,but obviously more than the width of the domain, and then use Snappy Hex Mesh to automatically mesh the domain around the wing/airfoil and to refine the mesh.
You can create the STL with a freeware program such as FreeCAD.
I used BlockMesh with a width of 0.1 and 1 cell width, empty patches on the side, like you have, and was able to get OKish results for Cl.

In your blockmesh dictionary, I do believe you are using parentheses instead of brackets in several areas, and I don't think you can have spaces in the patch name (empty frontandback there I think is a patch name, not a patch name and BC)

Refer to the airfoil tutorial for simpleFoam.

[kid]

Hi,
I did a NACA0012 turbulence incompressible model using simpleFoam (K-epsilon). It did run beautifully although, for zero AOA simpleFoam runs but when i change the AOA to 5 degree, it crashes after few iterations.
@appa
Regarding your result, i have few questions like can you share some plot with standard
resluts, as you have got some result but you are worried about accuracy?
Also , can you suggest me how to over come my problem. See, at inlet Vx=43.8.(approx) and Vy= 3.88.( approx) for AOA 5degree , but solver crashes after few iteration.
Though for zero AOA it solves.
regards
CFDkid

[mihaipruna]

maybe running out of memory?
what does it say when it crashes?

[appa]

Originally Posted by mihaipruna

I think it would be easier to create an STL of the airfoil with a bit of thickness (s stubby wing) ,but obviously more than the width of the domain, and then use Snappy Hex Mesh to automatically mesh the domain around the wing/airfoil and to refine the mesh.
You can create the STL with a freeware program such as FreeCAD.
I used BlockMesh with a width of 0.1 and 1 cell width, empty patches on the side, like you have, and was able to get OKish results for Cl.

In your blockmesh dictionary, I do believe you are using parentheses instead of brackets in several areas, and I don't think you can have spaces in the patch name (empty frontandback there I think is a patch name, not a patch name and BC)

Refer to the airfoil tutorial for simpleFoam.

I don't know how to create an STL, I found this blockMeshDict on the internet and I tried to add some changes to make it work but when I run chechMesh there is always a failure. I am preparing my own model with gmsh right now. How to use the Snappy Hex Mesh ??? is it a console command ?

Originally Posted by kid

Hi,
I did a NACA0012 turbulence incompressible model using simpleFoam (K-epsilon). It did run beautifully although, for zero AOA simpleFoam runs but when i change the AOA to 5 degree, it crashes after few iterations.
@appa
Regarding your result, i have few questions like can you share some plot with standard
resluts, as you have got some result but you are worried about accuracy?
Also , can you suggest me how to over come my problem. See, at inlet Vx=43.8.(approx) and Vy= 3.88.( approx) for AOA 5degree , but solver crashes after few iteration.
Though for zero AOA it solves.
regards
CFDkid

yes sure, here is what I get using pisoFoam (K-epsilon model also) with the mesh of the airfoil2D file .
For the change of the AOA, I don't know really. I'll have to try it first and I'll let you know.

Originally Posted by mihaipruna

maybe running out of memory?
what does it say when it crashes?

no it saids "exception en point flottant"
I guess there are some nodes faces that he didn't know where to belong, when meshing under OF they were added to a new boundary where the solver crashes.
"

[mihaipruna]

to use SnappyHexMesh, you create your STL file, the dictionary file in the System directory, and run it with the command snappyhexmesh.
Take a look at the motorcycle tutorial, the STL version.

[appa]

@mihaipruna: thanks , but I still don't know how to create an STL file
@CFDkid: I tried to change the Vx and Vy values in the u file, Vx*cos(5) , Vy*sin(y) , I put the values and it worked perfectly.
It didn't work when I used a geometry with an angle of attack different from 0 and I don't know why :/ (I tried to run this example on Elmer and it worked !)

[mihaipruna]

Originally Posted by appa

@mihaipruna: thanks , but I still don't know how to create an STL file

that is the easiest part. You can draw your airfoil in a free cad program such as this one:
http://sourceforge.net/apps/mediawik...itle=Main_Page
extrude it a bit, making sure it will extend beyond the blockmesh domain spanwise,save it as STL and then feed it to SHM.

[wiedangel]

Hi Appa, maybe you should use a bigger mesh. It looks to me that you are still in the situation where your simulation is influenced by the boundary conditions.
Hope it works.

[emarchandise]

Originally Posted by appa

@mihaipruna: thanks , but I still don't know how to create an STL file
@CFDkid: I tried to change the Vx and Vy values in the u file, Vx*cos(5) , Vy*sin(y) , I put the values and it worked perfectly.
It didn't work when I used a geometry with an angle of attack different from 0 and I don't know why :/ (I tried to run this example on Elmer and it worked !)

Hello Appa, I am just trying to solve flow around naca airfoil and have some troubles with my elmer sif file (no convergence, problme with boundary condition, ...). Could you help me by posting your Elmer example ? (emilie_marchandise@skynet.be)
This could be very nice ...
Emilie