this fluent problem...when i am calculating solidification in shell-tube storage system which is alien in Y-direction, i am facing this problem.
What do you mean, alien in Y-Direction ? There are many reasons that AMG solver error can occur. Can you tell me what are the boundary conditions you gave at the inlets and outlets ? And have you created a good mesh ? Please elaborate on your model.
could be due to incorrect meshing or incorrect boundary conditions.. u need to give more details
Hey !
Try changing the courant no. and check the quality of the mesh. If its poor, try using the mesh repair command and see if it works!
If you give more information about the case details that will be more helpful to see why is it diverging!
good luck !
Thanks all of you for trying to help me, problem is when i am taking the tube vertical(y direction),still it is happening, it is 2d problem, during initialization such error is showing
Error: cx-set-real-entry: wta[2](float)
Error Object: -1.#ind
Nitesh attach a picture of your mesh. And explain how you have employed your boundary conditions. Have you used any UDFs? If you have, you need to look at it , even though it does compile properly ,inconsistent definition may lead to divergence in your model.
it could be due to (1) incorrect boundary conditions (2) need to look at the scalar values as well as the vector directions in geometry (3) if sing any model, material or custom function ... look at the values specified there .. r they all reasonably real ?
Solidification problems do tend to diverge. There are many reasons to it. I will list them one by one with some remedial actions.
1. Assuming that you have set in the correct boundary conditions and material property and you have a good mesh. Try transient solution with a very small time step. See whether you are getting divergence still. If yes we may need to check the quality of mesh.
2. For solidification problems, mesh quality should be very good. Avoid mesh if the skewness ratio is above 0.75. In case you have 2d geometry as you have suggested, avoid using elongated rectangles as mesh. Make square shaped meshes and as small as your machine permits.
Solidification is a problem where large amount of latent heat is released. it is this release of latent heat which may cause a sudden jump in errors whenever a cell solidifies. To check that, invariably solidification problems are solved with very small step sizes over a good quality mesh domain.
If the above two steps doesn't help, then we may need to look into the physics of the problem and check the correctness of boundary conditions and material properties.
From the error messages which you have posted
Error: cx-set-real-entry: wta[2](float)
Error Object: -1.#ind
It appears that you are using a UDF and somewhere you are having a 0/0 calculation. Second line says (indefinite). Please look into it.
If you are using axi-symmetric, you have to make X-Axis as yours axis of symmetry.
Errors like the one you are getting can arise due to meshing issues. After importing your mesh in FLUENT do a mesh check and see if you get any errors. If there are errors you can rectify them by doing a mesh repair from the console.
I meet such problem in my 3D simulation on bluff body cantilever,
Quality mesh and BC are OK so far, initially it is running well but after 3373 iteration "reversed flow" warning appear and then "
Divergence detected in AMG solver: x-momentum -> Increasing relaxation sweeps!"
I use UDF for this case.
Please, your suggestion to solve those problems.
thanks a lot in advance
- Badges
- Science topic