Ferrite.jl icon indicating copy to clipboard operation
Ferrite.jl copied to clipboard

Improve pressure handling in incompressible Navier Stokes example

Open termi-official opened this issue 2 years ago • 5 comments

Pressure handling does not work very consistently for now, because the current choice for the solver does not resolve the constant consistently.

termi-official avatar Oct 05 '22 19:10 termi-official

We will work on this once https://github.com/SciML/OrdinaryDiffEq.jl/issues/1570 is resolved

termi-official avatar Nov 09 '23 09:11 termi-official

I thought #917 solved this, but seeing the following failure here? Is that related?

KnutAM avatar Aug 04 '24 13:08 KnutAM

I thought #917 solved this, but seeing the following failure here? Is that related?

Technically yes, but the solution is not very clean yet.

Also, I cannot reproduce the error and master CI is also green. Will try to take a look next week, but I am really tight on time, so do not expect too much from my side on this.

termi-official avatar Aug 04 '24 14:08 termi-official

Should we just disable this test then, and use this issue to track its solution? Possibly also add a warning to the example?

It's a bit annoying with sporadic unrelated CI failures...

KnutAM avatar Aug 04 '24 14:08 KnutAM

Should we just disable this test then, and use this issue to track its solution? Possibly also add a warning to the example?

It's a bit annoying with sporadic unrelated CI failures...

If we change something which triggers failure in the CI we should first investigate why the CI fails.

While I agree that sporadic CI failures are annoying, this issue is actually related to the linked example drove gmsh into a faulty state. See https://github.com/Ferrite-FEM/Ferrite.jl/pull/799#discussion_r1703618608 .

termi-official avatar Aug 05 '24 06:08 termi-official