CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM (http://www.cfd-online.com/Forums/openfoam/)
-   -   How to chtMultiRegionFoam (http://www.cfd-online.com/Forums/openfoam/87466-how-chtmultiregionfoam.html)

Scot April 20, 2011 17:38

How to chtMultiRegionFoam
 
Anyone got a simple case or "how to" tip for getting started in chtMultiRegionFoam? My tutorials don't run. Pretty complicated. Thanks

val46 April 21, 2011 02:59

the tutorial case should run. Maybe your openfoam isn't properly installed?
What is the error message you get?

Tobi April 21, 2011 05:45

yes - the tutorials should run.!

i 'll looking for my old simple case - maybe i find it.!

Scot April 21, 2011 09:30

Thanks Val46 and Tobi. I'd like to start with something simple like a block of iron is heated at one end, how does the temperature at the other end rise with time. So I'd probably use chtMultiRegionFoam. But for any of the tutorials in chtMultiRegionFoam and chtMultiRegionSimpleFoam, I try running and get the error: Foam fatal error: Cannot find file "points" in directory "bottomAir/polyMesh" in times 0 down to constant". However, the tutorials don't look anything like the case of a block heated at one end. Anybody got a similar case? Many thanks, Danke Sehr. Scot

Tobi April 21, 2011 09:47

Hi Scot,

you started the tutorials with "./Allrun" ?

mirko April 21, 2011 09:51

Quote:

Originally Posted by Scot (Post 304576)
Thanks Val46 and Tobi. I'd like to start with something simple like a block of iron is heated at one end, how does the temperature at the other end rise with time. So I'd probably use chtMultiRegionFoam. But for any of the tutorials in chtMultiRegionFoam and chtMultiRegionSimpleFoam, I try running and get the error: Foam fatal error: Cannot find file "points" in directory "bottomAir/polyMesh" in times 0 down to constant". However, the tutorials don't look anything like the case of a block heated at one end. Anybody got a similar case? Many thanks, Danke Sehr. Scot

Use Allrun and check the log files. My bet is on log.splitMeshRegions that will point the way. All of my missing "points" file errors were traced there.

Mirko

Scot April 21, 2011 10:46

Thank you Tobi and Mirko, ./Allclean and ./Allrun made it run, but the results appear very bad. In directory 0, all fields look fine. But in directories 50, 100, 150, 200, I get subdirectories "bottomAir", "topAir", etc and these all have the usual files p, phi, T, etc but these files are filled with a lot of junk symbols instead of scalar numbers. Instead of recognizable numbers for Temperature, for example, I get one number like 1200 followed by many lines of symbols. The log files look fine.

mirko April 21, 2011 12:23

Quote:

Originally Posted by Scot (Post 304593)
Thank you Tobi and Mirko, ./Allclean and ./Allrun made it run, but the results appear very bad. In directory 0, all fields look fine. But in directories 50, 100, 150, 200, I get subdirectories "bottomAir", "topAir", etc and these all have the usual files p, phi, T, etc but these files are filled with a lot of junk symbols instead of scalar numbers. Instead of recognizable numbers for Temperature, for example, I get one number like 1200 followed by many lines of symbols. The log files look fine.

How do the plots look?

Scot April 21, 2011 13:10

Hi Mirko- No plots of any fields except for the initial state (directory 0), presumably because the field files in directories 50, 100 etc are filled with junk symbols. The only clue is the warning: 'Detected wildcard ".*" in changeDictionaryDict' in all 5 log.changeDictionary files. Curiously, in the field files in directory 0 such as T, the boundaryField is defined using this wildcard ".*", which appears to be used as a shorthand in place of the names of the various boundaries (frontAndBack, fixedWalls, etc). Are you using version 1.7 and have no problem viewing the Temperature, for example, of the chtMultiRegion tutorials? Thank you! Scot

mirko April 21, 2011 13:35

Quote:

Originally Posted by Scot (Post 304609)
Hi Mirko- No plots of any fields except for the initial state (directory 0), presumably because the field files in directories 50, 100 etc are filled with junk symbols. The only clue is the warning: 'Detected wildcard ".*" in changeDictionaryDict' in all 5 log.changeDictionary files. Curiously, in the field files in directory 0 such as T, the boundaryField is defined using this wildcard ".*", which appears to be used as a shorthand in place of the names of the various boundaries (frontAndBack, fixedWalls, etc). Are you using version 1.7 and have no problem viewing the Temperature, for example, of the chtMultiRegion tutorials? Thank you! Scot

On 1.7.1 I cannot run the chtMultiRegionFoam/multiRegionHeater tutorial. A keyword is undefined in one of fvSchemes files.

However I can run the chtMultiRegionSimpleFoam/multiRegionHeater

I also get the wildcard warning in the log files.

As for junk symbols, do you have compression set to yes in system/controlDict?

Scot April 21, 2011 15:53

Mirko- I changed system/controlDict from "uncompressed" to "compressed". The result was that the field files were gzipped. But upon gunzipping, I see the same junk in the files rather than scalars. Please look in chtMultiRegionSimpleFoam/multiRegionHeater/100/topAir/T. Do you not see a bunch of symbols, many of which are a black ovals with a white question mark in it? Thank you again, Scot

mirko April 21, 2011 15:59

Quote:

Originally Posted by Scot (Post 304632)
Mirko- I changed system/controlDict from "uncompressed" to "compressed". The result was that the field files were gzipped. But upon gunzipping, I see the same junk in the files rather than scalars. Please look in chtMultiRegionSimpleFoam/multiRegionHeater/100/topAir/T. Do you not see a bunch of symbols, many of which are a black ovals with a white question mark in it? Thank you again, Scot


Yes, I see junk too even though I am writing uncompressed.

But I can preview the results in Paraview.

Can you post (attach) a part of log.chtMultiRegionSimpleFoam for timesteps 0, 1, 2?

That might tell us how the numerics are going.

Mirko

Scot April 21, 2011 16:13

Mirko- you see junk in the field files as I do, but you are able to view the results using paraFoam? I can select fields only for time = 0. For later times, I only have the possibility of selecting "Solid Color", "vtkCompositeIndex" and "cellNormals". I can't get T or U or anything else to show up in ParaFoam's object inspector. Can you view temperature, for example, at the end of the simulation?

mirko April 21, 2011 16:18

Quote:

Originally Posted by Scot (Post 304635)
Mirko- you see junk in the field files as I do, but you are able to view the results using paraFoam? I can select fields only for time = 0. For later times, I only have the possibility of selecting "Solid Color", "vtkCompositeIndex" and "cellNormals". I can't get T or U or anything else to show up in ParaFoam's object inspector. Can you view temperature, for example, at the end of the simulation?

Yes, I can see the results in paraFoam.

One thought: You need a file whatever.foam in the run directory for paraFoam to read the files.

Scot April 21, 2011 16:27

In my "run" directory is only another directory "tutorials". Would you please say in more detail what your mean about needing a "whatever.foam in the run directory"? Thank you! Scot

mirko April 21, 2011 16:43

Quote:

Originally Posted by Scot (Post 304639)
In my "run" directory is only another directory "tutorials". Would you please say in more detail what your mean about needing a "whatever.foam in the run directory"? Thank you! Scot

No, somewhere you need to have a directory with 0, constant, system directories, and Allrun and Allclean. That is your run directory.

Scot April 21, 2011 16:59

My directory is /chtMultiRegionSimpleFoam/multiRegionHeater. What file is needed here for paraFoam to read the files? PS: I've had no trouble with other heat transfer solvers such as buoyantSimpleRadiationFoam. But in all those cases, the field files were filled with ascii numbers, never symbols. Thank you for your continued consideration! Scot

mirko April 21, 2011 17:05

Quote:

Originally Posted by Scot (Post 304641)
My directory is /chtMultiRegionSimpleFoam/multiRegionHeater. What file is needed here for paraFoam to read the files? PS: I've had no trouble with other heat transfer solvers such as buoyantSimpleRadiationFoam. But in all those cases, the field files were filled with ascii numbers, never symbols. Thank you for your continued consideration! Scot

Sorry, but I'm out of ideas :(

Scot April 21, 2011 19:48

Sometimes field files have format ascii, as seen at the the top of the file. The junky files I note have "format binary". So that's the answer to that, but I still can't view results in paraFoam. Will keep trying, Thank you for your help! Scot

Tobi April 22, 2011 05:29

1 Attachment(s)
hey - normally you can 'not open the files in 1 2 3 4 5 - like T or U - only in the "0" file you can open it correctly (it 's my experience).

Sometimes (if you using for example buoyantFoam) you got the files as a tar zip... so thats not a problem.

I am using OF-1.7.x
the tutorials are working fine...

i added the tutorial files...


Tobi


All times are GMT -4. The time now is 18:07.