[shiyun]

Dear guys:

Problem description:
I have done a 2D airfoil mesh (I followed the youtube video, "ANSYS 12.1 (Part 2 of 3) ICEM CFD Hexa 2D Airfoil meshing"), and I add one more step----convert the pre-mesh to unstructure mesh and extrude 1 layer in 0.1 chord length.
But when I output the mesh to "Output Solver-----ANSYS CFX, Common Structural Solver----ANSYS", there is a warning message occur :" Family FLUID has mixed of 2D and 3D elements. this is not allowed in CFX-5"
But this mesh still can import to ANSYS CFX 14.0

Question:
1.Is this warning important?
2.I also face the convergence problem, In order to increase the Cd accuracy I turn on the transition turbulence model within SST model (I run low Re 1e5 - 5e5), but I found in high AOA(>10 degree) it is very difficult to converge and I also try some different mesh, the convergence situations aren't so much different. The thing is if I only run SST, the convergence issue won't appear but Cd accuracy is low. I also change the mesh density and Y+ value, the latest mesh has displayed 2D quality>0.9 and 3D quality>0.35 in ICME quality check and the Y+ =1.6 in CFX post.

Anybody can please give me some direction? Thank you very much.

Anybody have idea about convergence?

[Far]

Family FLUID has mixed of 2D and 3D elements. this is not allowed in CFX-5

I have no idea about this error message. These threads may help you with this error message.
http://www.cfd-online.com/Forums/cfx...-elements.html
http://www.cfd-online.com/Forums/cfx...-icem-cfd.html

But I can tell you that " you can export the 2d mesh in Fluent format (.msh) and when you import into CFX-Pre, it will automatically extrude the mesh by one element. And to control the thickness of mesh in third dimension you can change the dafault settings in CFX-Pre"

[D.B]

Hi Shiyun,
According to me all your 2D elements should be under seperate names to provide the boundary condition, while the 3D elements only make up the volume. Since CFX is a purely 3D solver it might not allow to have 2D and 3D elements under the same name. if your mesh has 2D and 3D elements under the same name, how are you giving the B.C's ?

Let us know if you find out anything new about this issue.

[shiyun]

Hi Far

Thanks for your suggestions and those links.

[shiyun]

Originally Posted by D.B

Hi Shiyun,
According to me all your 2D elements should be under seperate names to provide the boundary condition, while the 3D elements only make up the volume. Since CFX is a purely 3D solver it might not allow to have 2D and 3D elements under the same name. if your mesh has 2D and 3D elements under the same name, how are you giving the B.C's ?

Let us know if you find out anything new about this issue.

-----------------------------------------------------------------------------

Hi D.B

Thanks for your reply.
I will upload the screenshot pictures and I am not sure how to check the 2D 3D elements name, do you mean part name?
I define the B.C. in CFX-Pre (follow the ANSYS official workshop "transonic flow over a NACA 0012 aifroil" ), On the other hand, I cannot define boundary conditions in ICEM, there is no selection let me select (for this, I also upload the pictures)
And do you think the convergence problem is caused by this warning? I mean if I solve this problem, do you think it improve the convergence situation?