CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   About icoFoam case (http://www.cfd-online.com/Forums/openfoam-solving/60377-about-icofoam-case.html)

liugx212 December 14, 2005 12:09

Dear all, I am trying to ru
 
Dear all,

I am trying to run the icoFoam cavity case with the block size of 92*92*32 and defult sets. But I was told the error of time as follows. Could you please help me see and give advice?

Thanks a lot.

Errors:

Time = 0.02

Mean and max Courant Numbers = 0.109749 132.97
BICCG: Solving for Ux, Initial residual = 0.769511, Final residual = 9.41873e-06, No Iterations 38
BICCG: Solving for Uy, Initial residual = 0.725626, Final residual = 3.13877e-06, No Iterations 36
BICCG: Solving for Uz, Initial residual = 0.8086, Final residual = 9.13794e-06, No Iterations 35ICCG: Solving for p, Initial residual = 0.953251, Final residual = 9.12515e-07, No Iterations 265
time step continuity errors : sum local = 3.04285e-07, global = 1.32364e-18, cumulative = 1.65714e-18
ICCG: Solving for p, Initial residual = 0.841697, Final residual = 9.02924e-07, No Iterations 242
time step continuity errors : sum local = 1.65841e-06, global = 5.28474e-18, cumulative = 6.94188e-18
ExecutionTime = 192.11 s


Time = 0.025

Mean and max Courant Numbers = 1.37811 15395.5
BICCG: Solving for Ux, Initial residual = 0.997305, Final residual = 0.0301471, No Iterations 1001
BICCG: Solving for Uy, Initial residual = 0.997675, Final residual = 15.8377, No Iterations 1001BICCG: Solving for Uz, Initial residual = 0.999334, Final residual = 0.00554499, No Iterations 1001
ICCG: Solving for p, Initial residual = 0.999822, Final residual = 9.36367e-07, No Iterations 124
time step continuity errors : sum local = 0.00903109, global = -3.98103e-15, cumulative = -3.97409e-15
ICCG: Solving for p, Initial residual = 0.998574, Final residual = 9.58777e-07, No Iterations 132
time step continuity errors : sum local = 5.25563, global = 4.08839e-11, cumulative = 4.08799e-11ExecutionTime = 748.68 s


Time = 0.03

Mean and max Courant Numbers = 1.61369e+06 5.99557e+10
BICCG: Solving for Ux, Initial residual = 1, Final residual = 1, No Iterations 1001
BICCG: Solving for Uy, Initial residual = 1, Final residual = 1, No Iterations 1001
BICCG: Solving for Uz, Initial residual = 1, Final residual = 1, No Iterations 1001
ICCG: Solving for p, Initial residual = 1, Final residual = 51.6963, No Iterations 5001
time step continuity errors : sum local = 2.80935e+17, global = 2.2091, cumulative = 2.2091
ICCG: Solving for p, Initial residual = 1, Final residual = 56.148, No Iterations 5001
time step continuity errors : sum local = 7.65932e+26, global = -1.49149e+09, cumulative = -1.49149e+09
ExecutionTime = 1827.63 s


End

ploceus December 14, 2005 12:52

I may be wrong, but looks like
 
I may be wrong, but looks like your mesh is too fine. Maybe you can try to reduce time step.
Regards.

liugx212 December 14, 2005 13:01

Dear Charles Ju: Thanks for
 
Dear Charles Ju:

Thanks for your advice. my using time step is 0.005 and stop on 0.03. writeInterval is 1. Please give more details.

Thanks deeply.
Guoxiang

ploceus December 14, 2005 13:34

Hi. Maybe time step = 0.001 o
 
Hi.
Maybe time step = 0.001 or 0.0001?
Or you could increase cell sizes.
Also wrong boundary conditions will cause diverging solutions. Check them too.
Regards.

liugx212 December 14, 2005 13:47

Dear Charles Ju: Thanks ag
 
Dear Charles Ju:

Thanks again. I will check as you said. By the way, the boundary conditions were checked which was ok by run checkMesh. I am not sure whether it is ok or not.

I will tell you after test.


Thanks deeply.
Guoxiang

liu December 14, 2005 14:14

Boundary condition are checked
 
Boundary condition are checked by yourself, not checkMesh.

Do some coarse mesh calculation to see what will happen.

liugx212 December 14, 2005 14:15

Dear Charles Ju: I have ch
 
Dear Charles Ju:

I have change as you said, but the problem is same. Do you have another advice?

Thanks deeply.
Guoxiang

liugx212 December 14, 2005 16:40

Dear Xiaofeng: Thanks a lot
 
Dear Xiaofeng:

Thanks a lot. Just now, I checked the boundary file. My nFaces is: 28704, but the startFace is: 798192 So I think something is wrong here since startFace is further more the nFace. Could you please give some advice?

Thanks deeply,
Guoxiang

liugx212 December 15, 2005 16:59

Dear All, I still do not so
 
Dear All,

I still do not solve this questions yet. I have changed to small time step and checked the boundary. It likes ok. But the "time step continuity errors :" problem is still here. Could you please help me?

unoder December 17, 2005 14:44

Obviously something is wrong w
 
Obviously something is wrong with one of your files, so why don't you just copy a whole new/fresh icoFoam cavity case from inside /tutorials directory and then run blockMesh and icoFoam.

AFAIR there's also an .allrun file which works straight "out-of-the-box".

liugx212 December 17, 2005 14:49

Dear Martin Jørgensen: Than
 
Dear Martin Jørgensen:

Thanks for your help.
You mean I should use Allrun to instead of the blockMesh and icoFoam. Right?

Thanks again.
Guoxiang

unoder December 17, 2005 15:29

Hi Guoxiang, I mean: (under
 
Hi Guoxiang,

I mean: (under linux), standing inside your working directory - that would probably be something like (can't exactly remember if this path is correct, however here we go):

cd /home/guoxiang/OpenFOAM/OpenFOAM-1.2/guoxiang-1.2/run/tutorials

I would then do a recursive remove on this directory that doesn't work "rm -rf icoFoam/cavity" - notice that you should make sure you have a backup somewhere of the original directory which in my place would be in something like /OpenFOAM/OpenFOAM-1.2/run/tutorials. If you did like in the readme file that ships with openfoam, you would also have a backup.

So, standing in your working directory:

"cp -r /OpenFOAM/OpenFOAM-1.2/run/tutorials/icoFoam/cavity (place you want your working copy like for instance .)"

Then ./allrun worked for me (blockMesh and icoFoam should also work) as far as I remember and once you made sure everything is okay, you can experiment by changing mesh grading etc (in blockMeshDict-file).

I assume you once already have solved this case, right?

liugx212 December 17, 2005 18:43

Dear Martin Jørgensen: Tha
 
Dear Martin Jørgensen:

Thanks a lot.
Just now, I have tries as you said, But the problem is still same. Could you have another advice?

Thank you very much.
xiang

unoder December 18, 2005 10:42

Yep, attach your case here in
 
Yep, attach your case here in a compressed file so others can download it and try for themselves.

liugx212 December 18, 2005 14:28

Dear Martin Jørgensen: I ha
 
Dear Martin Jørgensen:

I have a test when I reinstalled. But the question is still here. It is very strange. I use the cavity in icoFoam case. I have really no idea. I do not how to do?

Thanks a lot.
xiang


All times are GMT -4. The time now is 02:22.