Learning Forum

[manelaero2020]

Hi friends ;

I am supposed to use FENSAP-ICE. The aerodynamical computation must be essentially performed by FENSAP, Fluent can do the job but I prefer FENSAP.

I have made my geometry as following: solid frozen body enclosed in a sphere ( so two frozen bodies) then I have applied boolean function, finally one body (fluid).

Meshing is performed by Ansys Meshing.

The solution on Fluent has converged, but in FENSAP: an error message is displayed: execution error

then the software stopped and is closed automatically.

Have you an idea please why FENSAP refuse to run ? May be the error is the way that I have created the geometry ? ( I don't have an idea but I think that in the case of FENSAP , i must leave the two bodies frozen (separated ) without applying bolean function (please look at the following video (2.32)

note : the file that FENSAP read is the same as FLUENT (.msh)

https://www.youtube.com/watch?v=mSiUp8MR3lU&t=175s

[Shoaib Shah]

Hi
Is the video you shared your geometry? I don't see any sphere. Or is it something you followed to create your own geometry setup.
If its a cube (geometry) in a sphere (fluid), than that should not be a problem for FENSAP to run. What's the execution error you are getting ? It could be anything. Open out file (logs). And see what error is printed.
Also for the sphere how do you setup the boundary condition: As 1 Farfield or into 2 BCS: Inlet/Outlet ?
Thanks. Let me know. Cheers

[manelaero2020]

Hi Mr. Shah ;
Thank you for the reply.
This video is just an example from Youtube, in the order to give an idea of how I have set my fluid domain and boundaries conditions.
My geometry is a plane into a sphere domain. In the order to save time I have performed an axisymmetric geometry. Computation diverged on FENSAP but easily converged on Fluent.
I am not sure but I think that the mesh is not good ( because of the presence of some small surfaces on my geometry); the question is so: why computation has converged on Fluent?
The error message is: Zero iterations in the iterative solver ( Fensap)
Have you an idea about the reason and solution please (especially for FENSAP)?
Thank you a lot
Sincerely regards

[Jian Chen]

hi,
Can you do the test as following:
after your fluent run successfully (let's say 1 or 2 iterations), save your fluent run as *.cas (or *.cas.h5), then import your fluent cas file into Fensap-Ice and try to run it again. See if this way works , and if this way gives any other error issue. (Please copy the error message to us if it appears.)
Thanks

[manelaero2020]

Hi Jian Chen,

it seems a good idea, I will try it and return to you with the response

Thanks

[Shoaib Shah]

Hi again
Yes as you mentioned that mesh has some issues. Every solver is different. Fluent has some extra poor mesh numerics to help run with poor meshes, but it does not necessarily mean your final solution is good or the right one, especially in areas where you have poor mesh points. You should verify your solution near those poor mesh cells, especially if it is in area of interest. Also, it depends on how you have setup Fluent, i am not aware of your solver settings and choices and/or convergence.
In general, you should try to fix the mesh first, that way you ensure you have a good final solution and a good starting point.
As far as FENSAP is concerned, there could be many reasons why your case is not working. It could be because of poor FENSAP setup? Poor mesh? Did you setup the inlet boundary conditions correct? What CFL are you using? How bad is the mesh near the surface, and in particular the prisms near the area of bad mesh points? Also how have you set up your inlet as a Farfield BC or an inlet/outlet BC? Also what mesh type are you using ? - Tet only? Hybrid tet/prism ?
Please also share some images of the mesh, where the surface mesh is poor. But also if possible share some pictures of the mesh, in general e.g. near the leading edges, trailing edge, of some cutting planes showing prisms and general mesh transition from boundaries to volume.
A problem like this, it would be best to share the setup and mesh to help troubleshoot better.
Let us know. Cheers.