Convergence Problem
 

Hi all,

i am currently trying to simulate a roof in OF 2.3. The Problem is as mentioned in the topic my variables does not converge. i changed the boundary condition and tried a lot but nothing helped...my results are non physical and not so good as i wish to be...i do not know how to handle it. I hope someone can help me cause i have no ideas anymore...

thanks in advance

Hi,

1. What solver do you use? (from fvSchemes and fvSolution I guessed simpleFoam, is it right?)
2. Are you sure your mesh is OK? (provide checkMesh output)
3. Did you try upwind for all div(phi, *) lines?

does it include for all fluids fvSolutions??

Hi,

thank you for your response! I am quite new in OpenFoam. I am trying to simulate a roof for that i used the motorbike case. My target is to get pressure on surface of the roof.
1. Yes it is a simpleFoam solver
2. checkMesh is ok i put it in attachment
3. velocity is still increasing, even when its better with upwind now.

Are the boundary condition i choosed ok or did i something wrong??

Hi,

unfortunately my telepathic abilities are still quite weak, I didn't get "I am trying to simulate a roof for that i used the motorbike case". Please describe the problem you are solving in more details: geometry, flow conditions. Without this information, one can just try guessing if there is a problem with IC/BC or not.

Hey Alexey,

you have to work on your telepathic skills :) , it would help me much better, cause my english is really bad. To my project:
I used a flat roof in wind tunnel to get aerodynamic coefficients. Now i want to solve it numerical, to compare the results. For that i took the input files of the motorbike case of openfoam tutorial and changed it, as you can see in the attachment above (u,p,k,etc.). Simplefoam is a steady-state solver. My simulation has to converge after a while but it doesnt. This is a 2D Simulation, i want to simulate it later as 3D, with much more cells. So it must converge less than 10000 Iteration. Otherwise it needs too long time for computation.
I put some picture in attachment. For the velocity result in the picture i needed 50000 Iterations. This is too much. It would be better if i could achieve it earlier. As i said before, i am really new in openfoam and not so familiar with it. I hope you understand my problem now?

Hi,

well I'd suggest

1. Change BC for U top patch -> slip (for vector it's slightly different than zero gradient).
2. Increase nCellsInCoarsestLevel in p solver up to 255.

everything else seem to be OK (or I'm missing something from the description of the case).

Hi,

thanks a lot for your help. I will try it now!
1. Do you know how to plot my pressure over time? I tried to do it with "PlotSelectionOverTime" (see Picture) but it did not work.
2. I want openfoam to output the forceCoeffs. It says all of the coeffs (lift/drag) are zero. There is no error output? I changed forceCoeffs file as in other threads of CFD recommended.

Do you want to take a look over my case? I can send it to you...if you has time for it?

Hi,

1. You've selected a point but in paraview tree on screenshot you've got only cell arrays. So there's nothing to plot. Usually for plotting over time I use probes and gnuplot.

2. I don't know you geometry, I've got nothing to say about Cd and Cl calculation.

Also are you sure there's a steady state solution with your geometry and Reynolds number?

what do you mean when you say "are you sure there's a steady state solution with your geometry and Reynolds number?" Normally there must be a steady state solution...

with slip bc the solution is getting worser :(

Not quite sure about "must" ;) What about, for example, vortex street?

yeah you are right.