[VentinS]

Dear Foamers,

I am trying to set up a transient case. The velocity is slow: Re =17
The mesh is a 3D Mesh, basically a U-shaped channel. I never tried to solve a case in openFoam with transient conditions.
My setup is:
U:
Inlet : (0 -0.0076 0), everything else 0
p:
Outlet: fixedValue 0;
In the controldict:
adjustTimeStep is activated
max Co 1,
Solver is the pimpleFoam in the Piso-mode

What I am trying to achieve is something like that:
https://www.youtube.com/watch?v=jx3DQn33MHE

But the time steps are getting smaller and smaller my deltaT is about 1e-10.
And when I deactivate the maxCo and the timesteps, the courant number rises.
And the velocity profile doesn’t look like the video at all.
Shall I wait or is everything wrong?
Kind regards

[simrego]

Hi!


Your outlet is a zeroGradient for U and the inlet is zeroGradient for p?
Try to use a bit smaller Co (maybe 0.8-0.9) to be sure you wont exceed 1.
What about your mesh? Run a checkMesh so if you have a really poor mesh it'll complaining!

[VentinS]

Hey,

Originally Posted by simrego

Your outlet is a zeroGradient for U and the inlet is zeroGradient for p?

Yeah, exactly, I've checked

Originally Posted by simrego

Try to use a bit smaller Co (maybe 0.8-0.9) to be sure you wont exceed 1.

Ok, I set the maxCo =0.8 but the timesteps are still getting smaller...

Originally Posted by simrego

What about your mesh? Run a checkMesh so if you have a really poor mesh it'll complaining!

The Mesh is OK and the steady state worked fine...

[simrego]

Can you share your case?
I'm out of guesses, and if no one will give you the right hint, I'll check it. It doesn't seems to be a hell complicated case.

[VentinS]

Yeah sure,


It would be wonderful, because I am pretty stuck....


https://www.dropbox.com/s/m9kr4vzsm9...Daily.zip?dl=0


Thanks!

[RobertHB]

Try inletOutlet for your velocity outlet condition.

[VentinS]

Hey,
I've tried it but the deltaT ist still around 2e-7.


Maybe I should start with a 2D case first....

[arjun]

Originally Posted by VentinS

Yeah sure,


It would be wonderful, because I am pretty stuck....


https://www.dropbox.com/s/m9kr4vzsm9...Daily.zip?dl=0


Thanks!

i opened the mesh and it shows very very small volume of 1E-18 m3 types. Also does it really have those crumbles on the surface as this image shows.

[VentinS]

hmm.. I guess this small volumes are the result from trying to flatten the mesh in icem and also this crumbles. I guess I was lucky, that steady case worked...


But even when I chop off the unstructured part, the result remains the same.


Here is the "short" version, if it helps:


https://www.dropbox.com/s/3b47zh3xs7...short.zip?dl=0

[simrego]

Hi!


Sorry for the late answer.
So I checked your case. inletOutlet can be a bit dangerous. You can have really unphysical results.
Your 1st case seems to be really weird. I mean the mesh. Also the max nonorthogonality is ~83 if i remember well, which is pretty high.
But I think the mesh is also problematic in the second case you provided. It seems to be nice, checkmesh -> no error, but my first guess was that the mesh is bad. So i made a really fast rubbish mesh with snappyHexMesh, and the case ran perfectly as it should.
I suggest you to recreate the mesh. Try to create a simpler mesh, or just a coarser one, and if it works refine it.
I also made some modifications in the files. You can find your modified case in the following link.

https://drive.google.com/file/d/1cV5...ew?usp=sharing


Again!!! This mesh is really ugly, and you should doublecheck everything in the provided case, maybe i did some silly thing, but it is working without any problems for me.
You should tighten the tolerances, etc. Also there is a comment in the fvSolution file, I haven't used the foundation OF for a long time and i had no more time for that so sorry for the crap case.

[VentinS]

Thank you soooooo much!



I guess I was lucky that the steady case worked.

I'll try and report you back


And again sorry for your trouble!!!!!!!!!!!