Poiseuille profil, Stokes , Uzawa

General Discussion
1

Hi everyone,

I am new to this amazing world of FreeFem users and I am having some issues with my work on a Poiseuille profil for Stokes equations with the Uzawa method.
My results are totally wrong but I don’t see what I have to change.
Could you give me some advice or recommandations to improve my work.

Thanks for your time.

1 Like
2

If you are looking for something much more efficient than Uzawa method, you can look at PETSc examples, e.g., stokes-fieldsplit-3d-PETSc.edp, which also solves Stokes equations with a Poiseuille inflow.

1 Like
3

Thanks for you answer, I have already solved the problem with another method, but i have to compare it with the Uzawa one (that’s the work I am ask to do). I think my problem is maybe link to how I apply boundaries conditions and the labels identification.

4

Have a look on my PhD dissertation, there is a complete Uzawa algorithm in the appendix

1 Like
6

Hello,
I’d be happy to help you troubleshoot your code. You can follow these suggestions:

  1. Check Boundary Conditions
  2. Mesh Quality
  3. Parameter Tuning
  4. Initial Guess
  5. Debugging Output
  6. Comparison with Analytical Solution
  7. Code Review
    You can also check this : FreeFem-sources/examples/hpddm/stokes-fieldsplit-3d-PETSc.edp at developgenerative-ai

I hope this will help you.

Thanks