[cristina]

Hi all,

I would like to run a simple problem of flow over a terrain using buoyantBoussinesqpisoFoam. I have problems with setting my boundary conditions.... my case colapses when I start runnin it. Is there any tutorial case I could consult?

Help please!!!!!!!

Thanks,

Cristina

[Thangam]

could you post the log file with the error please? since, there could be be many reasons that can cause your simulation to crash!

[cristina]

Hi Thangam,

please find attached the case. Note that it is the Hot Room case modified... with inlet and outlet patches.

If you look at the test.stdout.txt file you will see how the velocity increases after each time step until it crashes.

Thanks in advance! Hope you can help me!

Cristina

[Micha69822]

Hy Cristina,

which version from openFOAM you use?
In the new Version the piso-Solver for buoyant is changed in pimple.


Best regards
Michael

[cristina]

Hi,

I am using the 1.6-ext version...

¿did you get a case runnin with pimpleFoam?

Cristina

[Micha69822]

Hi,

no!
I tried to simulate free convection with "buoyantBoussinesqSimpleFoam" and "buoyantSimpleFoam" but i get no good flow pattern!

I think the p_rgh values are wrong. Or there is a bug for this cases - I don't know!


---
Michael

[Thangam]

Hi,
Just on a quick look at your case, there are a couple of things that needs major revisions. First, the mesh. The mesh is very very coarse. With 4000 cells, running a URANS is not a good idea. Second, the boundary conditions of U. The negative value of 800 in the ceiling. You are over constraining the boundaries by fixing these values at the ceiling and the floor. There are a number of boundary conditions that will suit your case and you need to do some searching. To me, the bc's look wrong. Next, the pressure bc's. same as U. you are fixing both the floor and the ceiling pressure values.Also,im not sure how this works without the p_rgh file. Similarly for other fields as well, you have got the bc's incorrect for the ceiling and the floor. My suggestion would be to revisit an simple steady state case with buoyantBoussinesqSimpleFoam and get running with the right bc's. In short, try not to constrain the boundaries by fixing values at the inlet and the outlet. Hope this helps.

[cristina]

Hi Tangham!

Thanks for your comments. I went a few steps back and simulated a really easy case (inlet/outlet/ground, sides and sky) with simpleFoam RANS and with simpleFoam LES. And it worked out! I think now I got the right BC. Now I am trying to run that same case with buoyantboussinesqsimplefoam...


Cristina

[cristina]

Hi again!

I am trying to run the simpleFoam and the buoyantBoussinesqSimpleFoam cases and I definitely do not find the right BC! . I really do not know how to go on....

I keep getting very high values for the velocity, and of course the simulation crashes...

¿can you please help me? I attach you the cases for the LES simplefoam and for the LES buoyantBoussinesqSimpleFoam.

maybe you can have a look and help me find where the problem is....


Cristina

[Thangam]

Hi,
I just went through your case. I think there are still plenty of things for you to work on. First, you need to figure out whether you need to run this as a steady state or unsteady state. The next is type of modeling, whether RANS, URANS or LES. Based on the above, you need to choose your solver. Whether simpleFoam or pisoFoam. There is fair bit of reading for you to do and understand the above terminologies. Otherwise CFD would give you nightmares and you would start hating it eventually! Having said that, I have still modified the files for you to have a look.
Hope this helps.

[cristina]

Hi Thangam,

Yes you are right!!! I must define correctly the cases.... I want to simulate a steady state RANS (k-epsilon) simulation. For this I am using the simpleFoam solver. After I would like to simulate the same case but including the temperature equation with the buoyantBoussinesqSimpleFoam.

I have set up the RANS case following your instructions but I still don´t get any good results! And really dont know why.....

Thanks!

Cristina