After running the sonicFoam/fo
After running the sonicFoam/forwardStep case, I was interested to see how well the shock could be resolved by interpolating the old solution onto a finer mesh. Using icoFoam/cavityFine as an example, I increased cellDensity in both i and j directions, to 2i+1, 2j+1, then adjusted startTime to 10, then mapFields returned:
Source mesh size: 5250 Target mesh size: 21483 Consistently creating and mapping fields for time 10 interpolating p --> FOAM FATAL ERROR : FOAM exiting Any information concerning the cause of the problem and the means of resolving it would be much appreciated... the error message gives me no clues. Thanks, John Van Norman |
I get a similar error when run
I get a similar error when running mapFields. I have just like you, only refined my mesh and want to map a coarser solution to a finer mesh. I get the following with the flag "-consistent"
--> FOAM Warning : treeBoundBox::treeBoundBox(const pointField& points) : cannot find bounding box for zero sized pointFieldreturning zero Caught arithmetic exception and then the output from gdb is: #4 in Foam::meshToMesh::calcAddressing () from /users/tfd/f98faka/OpenFOAM/OpenFOAM-1.1/lib/linuxAMD64Opt/libsampling.so #5 in Foam::meshToMesh::meshToMesh () from /users/tfd/f98faka/OpenFOAM/OpenFOAM-1.1/lib/linuxAMD64Opt/libsampling.so #6 in mapConsistentMesh () However, if I run without the -consistent flag, it runs ok, but doesn't do anything, that is I get no new directories in the directory with the finer mesh. /Fabian |
Heya,
John: you have probab
Heya,
John: you have probably hit a boundary condition that the mapFields does not know about - could you have a look at the fields you're trying to map and tell me what boundary conditions you are using. Not a disaster, you'll just need to re-link the mapFields with some additional libraries (the one including the fvPatchField type you need). Could you try re-running the application under gdb and post the trace-back, e.g.: gdb mapFields r root1 case 1 root2 case2 where roots and cases are what you've used with the original mapFields When it falls over, you'll (probably) be able to type "where" which will produce the info Fabian: Hmm, is it possible that one of your meshes has no points in it? This is a failure to build the search trees, which means something is probably wrong with the mesh. Try running checkMesh on both meshes. Hrv |
Also it is helpful to run with
Also it is helpful to run with
setenv FOAM_ABORT 1 in your .cshrc (or equiv for .bashrc). This will cause OpenFOAM to abort (and produce a coredump) instead of exit. |
I ran the same case with FOAM_
I ran the same case with FOAM_ABORT 1 under gdb, with the same results. No stack or core dump, what am I doing wrong here? The fields are mapped (I think) from the same as in sonicFoam/forwardStep. Shall I forward the ~constant/* stuff to you? Perhaps my lack of understanding of OOP is the problem.
Thanks, JVN |
Hi John,
if you have the FO
Hi John,
if you have the FOAM_ABORT environment variable set (in OpenFOAM1.1) you should get the message FOAM aborting (FOAM_ABORT set) Abort (core dumped) Exit 134 and as long as you don't limit the coredumpsize you should get a core. (or you can run it inside gdb right away) |
Hi Mattijs,
I set FOAM_ABORT
Hi Mattijs,
I set FOAM_ABORT 1, then ran mapFields under gdb and got the following: (gdb) r $FOAM_RUN/tutorials/sonicFoam forwardStep $FOAM_RUN/tutorials/sonicFoam forwardStepFine -consistent Starting program: /home/john/OpenFOAM/OpenFOAM-1.1/applications/bin/linuxOpt/mapFields $FOAM_RUN/tutorials/sonicFoam forwardStep $FOAM_RUN/tutorials/sonicFoam forwardStepFine -consistent [Thread debugging using libthread_db enabled] [New Thread 4148004544 (LWP 4372)] /*---------------------------------------------------------------------------*\ | ========= | | | \ / F ield | OpenFOAM: The Open Source CFD Toolbox | | \ / O peration | Version: 1.1 | | \ / A nd | Web: http://www.openfoam.org | | \/ M anipulation | | \*---------------------------------------------------------------------------*/ Exec : /home/john/OpenFOAM/OpenFOAM-1.1/applications/bin/linuxOpt/mapFields /home/john/OpenFOAM/john-1.1/run/tutorials/sonicFoam forwardStep /home/john/OpenFOAM/john-1.1/run/tutorials/sonicFoam forwardStepFine -consistent Date : Jun 20 2005 Time : 15:25:23 Host : gibbon PID : 4372 Root : Case : Nprocs : 1 Source: "/home/john/OpenFOAM/john-1.1/run/tutorials/sonicFoam" "forwardStep" Target: "/home/john/OpenFOAM/john-1.1/run/tutorials/sonicFoam" "forwardStepFine" Create databases as time Source time: 10 Target time: 10 Create meshes Source mesh size: 5250 Target mesh size: 21000 Consistently creating and mapping fields for time 10 interpolating p --> FOAM FATAL ERROR : FOAM aborting (FOAM_ABORT set) Program received signal SIGABRT, Aborted. [Switching to Thread 4148004544 (LWP 4372)] 0xffffe405 in ?? () (gdb) where #0 0xffffe405 in ?? () #1 0xffff5c8c in ?? () #2 0x00aab955 in raise () from /lib/tls/libc.so.6 Previous frame inner to this frame (corrupt stack?) (gdb) q I'm not sure what this means, besides that libc.so.6 is problematic. I have the linux version of OpenFOAM, but my distro (FC3) is the AMDx86_64 and I did not install every package. Think this might be the cause? Thanks again, JVN |
Forgot you were running on the
Forgot you were running on the AMD64 where traceback does not function properly. It should have printed out all the functions called upto the error.
Does the tutorial showing mapping work? Check what your case is different and work your way back from that. |
can someone tell me step by step how i can use mapFields utility?
i use cavityFine tutorial but i cant run this utility and below error appeared: i make a folder in home by name cavityFine and copy all data of cavity folder (0 and system and constant) in cavityFine after change the mesh resolution go to cavityFine directory and type: mapFields cavity and the below error appeared: Please help me Quote:
|
hello,
from your log, you are not in cavityFine dir. Try "cd cavityFine" and after "mapFields cavity" if you don't want to be in cavityFine dir, use the -case option. regards, olivier |
Dear OliverG thanks for your answer
i did what u said but i cant mapFields cavity to cavityFine i thinks if i write here what i did step by step its better the question is I CANT MAPPFIELDS CAVTIY TO CAVITYFINE AND NEED HELP? i did: 1.copy cavity file to home/ashkan-2.1.0/run from openfoam tutorial 2.open terminal in Ubuntu and and make a directory by name cavityFine 3.copy cavity sub directory (0 and constant and systems) to cavityFine directory 4.change mesh resolution in cavityFine directory (constant/blockMeshDict) 5.go to cavityFine directory in Terminal 6.type "mapFileds OpenFOAM/ashkan-2.1.0/run/cavity" and the below error appeared Quote:
|
hello ashkan,
* after step 4, you should create the mesh again, so use blockMesh again. * step 6: try mapFields ../cavity (or mapFields -case /home/ashkan/OpenFOAM/ashkan-2.1.0/run/cavityFine /home/ashkan/OpenFOAM/ashkan-2.1.0/run/cavity if you use the full path). regards, olivier |
so thanks olivierG
i did what you said and below error appeared: Quote:
did i need make a sub directory by name mapFieldsDic? i so sorry for these questions and so thanks for your help |
yes, you must have mapFieldsDict in your cavityFine/system/ dir.
regards olivier |
dear oliver i do what u said again and again but i cant run mapFields.
how i can make mapFieldsDict? i really need it and i read OpenFoam tuturials but i cant run mapFields. :( |
can not run mapFields please help
Dear All,
I am trying to run mapFields for cavity tutorial . I am getting error as mentioned here. :confused: Can you help please. ------------------------------------------------------------------------- test@ubuntu:~/OpenFOAM/test-2.1.0/run/cavity$ cd cavityFine test@ubuntu:~/OpenFOAM/test-2.1.0/run/cavity/cavityFine$ mapFields -case /home/test/OpenFOAM/test-2.1.0/run/cavityFine /home/test/OpenFOAM/test-2.1.0/run/cavity -consistent /*---------------------------------------------------------------------------*\ | ========= | | | \\ / F ield | OpenFOAM: The Open Source CFD Toolbox | | \\ / O peration | Version: 2.1.0 | | \\ / A nd | Web: www.OpenFOAM.org | | \\/ M anipulation | | \*---------------------------------------------------------------------------*/ Build : 2.1.0-0bc225064152 Exec : mapFields -case /home/test/OpenFOAM/test-2.1.0/run/cavityFine /home/test/OpenFOAM/test-2.1.0/run/cavity -consistent Date : Apr 16 2012 Time : 22:25:19 Host : "ubuntu" PID : 2041 Case : /home/test/OpenFOAM/test-2.1.0/run/cavityFine nProcs : 1 sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE). fileModificationChecking : Monitoring run-time modified files using timeStampMaster allowSystemOperations : Disallowing user-supplied system call operations // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // Source: "/home/test/OpenFOAM/test-2.1.0/run" "cavity" Target: "/home/test/OpenFOAM/test-2.1.0/run" "cavityFine" Create databases as time --> FOAM FATAL IO ERROR: cannot find file file: /home/test/OpenFOAM/test-2.1.0/run/cavityFine/system/controlDict at line 0. From function regIOobject::readStream() in file db/regIOobject/regIOobjectRead.C at line 73. FOAM exiting test@ubuntu:~/OpenFOAM/test-2.1.0/run/cavity/cavityFine$ ------------------------------------------------------------------- Thanks in Advance, Indrajit |
Hi all,
I am running a 2D cavity simulation (3D region with spanwise cross section all the way across) I want to map these results into a 3D cavity flow (essentially a channel flow with a box cut into the middle of the bottom surface of the channel). When I try to run mapFields to map the 2D cavity results into the 3D cavity it does not create a 0/ directory (and of course, no files with it). However, there are no error messages and the output message reads as follows: /*---------------------------------------------------------------------------*\ | ========= | | | \\ / F ield | OpenFOAM: The Open Source CFD Toolbox | | \\ / O peration | Version: 2.1.0 | | \\ / A nd | Web: www.OpenFOAM.org | | \\/ M anipulation | | \*---------------------------------------------------------------------------*/ Build : 2.1.0-bd7367f93311 Exec : mapFields ../../cavity2D/cavityRes3 Date : Aug 08 2012 Time : 13:24:35 Host : "node6" PID : 21563 Case : /gpfs/home/eng/mauiie/OpenFOAM/mauiie-2.1.0/run/project/cavity3D/cavity3Da nProcs : 1 sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE). fileModificationChecking : Monitoring run-time modified files using timeStampMaster allowSystemOperations : Disallowing user-supplied system call operations // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // Source: "../../cavity2D" "cavityRes3" Target: "/gpfs/home/eng/mauiie/OpenFOAM/mauiie-2.1.0/run/project/cavity3D" "cavity3Da" Create databases as time --> FOAM Warning : From function Time::setControls() in file db/Time/Time.C at line 226 Reading "../../cavity2D/cavityRes3/0/uniform/time" from line 18 to line 24 Time read from time dictionary 50 differs from actual time 0. This may cause unexpected database behaviour. If you are not interested in preserving time state delete the time dictionary. Source time: 0 Target time: 0 Create meshes Source mesh size: 4864000 Target mesh size: 8128000 Mapping fields for time 0 End This looks quite promising and I am perplexed as to why it is not creating any files. |
Hello,
I am having a similar problem as you describe. I do a simulation in an empty box with blockMesh, then I want to map the fields to the same box with a finer mesh done with snappy. This time there is an obstacle in the box, so an additional patch. When I run the mapFields I do not get any output either. Source mesh size: 500000 Target mesh size: 919306 Mapping fields for time 0 End Did you find out what was the problem in your case? I tried indicating the patchMap ( ); and cuttingPatches ( ); in mapFieldsDict, but nothing seems to work, and no error messages of any sort. Thanks heaps in advance for any help on this. Carlos |
I think this can happen if you have already data in the target time directory. You can solve this option, to map it to a time-folder with some original files (like you have in 0/). If these files are empty, except for boundary conditions and uniform internal field, there is no mesh-information stored, and you should be able to use mapFields.
|
Thanks, Bernhard!
Yes, I did something similar to that, as it is suggested in this thread: http://www.cfd-online.com/Forums/ope...mapfields.html Best regards, Carlos |
All times are GMT -4. The time now is 06:58. |