Learning Forum

[user deleted]

Subiter 18 : residual 4.09558e+00

Warning: AMG preconditioner correction is of poor quality.

Subiter 19 : residual 4.09553e+00

Warning: AMG preconditioner correction is of poor quality.

Subiter 20 : residual 4.09552e+00

(Ref: Subiter 0 : residual 4.09578e+00)

59 7.2506e-02 1.8208e-01 1.4449e-01 2.2747e-01 7.8608e-03 3.3508e-03 04:23:02 442

Subiter 0 : residual 4.09552e+00

Warning: the residual minimization scheme stalls, therefore the scheme is initialized and the adjoint advancement is terminated.

** Suggestion: increase the number of modes in the residual minimization scheme if there is enough memory.

Good afternoon, I'm solving a topological optimization problem. When solving, a warning is issued and further solution does not go. What does this mean? I have been working in Fluent Fluent for 1 year and there has never been such a problem. Mesh triangular

after that the solution stops

[Rob]

It's the multigrid and may mean your optimiser routine/approach is struggling. Can you check you're not about to lose a surface or fold something so the topology is changing?

[user deleted]

Thanks Rob, I have converted the grid into hexes. The error is gone. The stream spins and turns 90 degrees. I redid the grid into hexes and everything was fine. The quality message appears but does not give an error. Did I want to ask you a fluent that solves the optimization problem with all types of meshes ?. Sorry for my English"

[Rob]

The optimisers should work well with tets, and also hex depending on the remeshing. The only problem I can see with the above is if the restriction towards the outlet distorts too much at a corner for the remeshing tools.

[user deleted]

How can this influence be reduced? Decrease mesh size? or make a thickening?

[Rob]

Have a look at the mesh to see what's happened to the model. The flow/shape may not be much use for engineering data but it'll give some hints into what is going on.

[user deleted]

Thank you so much !!!!

[phuocdinh]

Hello sir!
I am having the same problem as you said. It distorts too much at some angle for the rework tools.
How should i solve this problem?

[Rob]

That depends on what's happening. You may need to manually remesh and then continue the run or add limits to the model deformation. Some pictures would help.

[phuocdinh]

Thank you so much !!!!

[phuocdinh]

Hello sir!
I am optimizing the model with the observed pressure drop. But the distortion at the corners is too great, I've re-made the mesh many times but the sample still doesn't look like the picture below. How should I solve this problem?

[Rob]

That may be how you've set the constraints and observables. What are you getting?

[phuocdinh]

i didn't' set constraints. my observable is drop pressure. i am using direct Interpolation morphing method . smoothness equal to 2.

[Rob]

What's to stop the block ballooning out or collapsing?