[madad2005]

For some reason I am getting floating point exceptions when trying to convert foam data from the motorbike tutorial over to ensight data. It appears to crash out when trying to write the variables to file. Has anyone know how to get round this?

################################################## #



/*---------------------------------------------------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: 1.6 |
| \\ / A nd | Web: www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
Build : 1.6-f802ff2d6c5a
Exec : foamToEnsight -latestTime
Date : Sep 14 2009
Time : 17:33:57
Host : rg25
PID : 1013
Case : /spare/OpenFOAM/testing/motorBike
nProcs : 1
SigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time
Create mesh for time = 500

write case: motorBike.case
Translating time = 500
Converting field (binary) nut
#0 Foam::error:rintStack(Foam::Ostream&) in "/spare/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
#1 Foam::sigFpe::sigFpeHandler(int) in "/spare/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
#2 __restore_rt at sigaction.c:0
#3 writeEnsDataBinary(Foam::Field<double> const&, std::basic_ofstream<char, std::char_traits<char> >&) in "/spare/OpenFOAM/OpenFOAM-1.6/applications/bin/linux64GccDPOpt/foamToEnsight"
#4 void ensightFieldBinary<double>(Foam::IOobject const&, Foam::ensightMesh const&, Foam::fileName const&, Foam::word const&, int, Foam::Ostream&) in "/spare/OpenFOAM/OpenFOAM-1.6/applications/bin/linux64GccDPOpt/foamToEnsight"
#5 main in "/spare/OpenFOAM/OpenFOAM-1.6/applications/bin/linux64GccDPOpt/foamToEnsight"
#6 __libc_start_main in "/lib64/libc.so.6"
#7 _start at /usr/src/packages/BUILD/glibc-2.9/csu/../sysdeps/x86_64/elf/start.S:116
Floating point exception

[henry]

I tried foamToEnsight on the motorbike tutorial case in OpenFOAM-1.6.x and it worked correctly even with valgrind.

H

[madad2005]

Sorry, I should have updated this.

I forgot to specify the -ascii flag for the data, and I got it working in the end. The results I'm getting, though, are completely wrong, with almost every variable in every cell with a value in the region of 10^48 or more.

[henry]

It works correctly for me in OpenFOAM-1.6.x with or without the -ascii flag.

H

[carl]

Hi all,

being very happy seeing such a complex geometry in the testcase, i tried it out.

I'm having the problem of unstable computation in simpleFoam too. Without
having changed anything of the case, the pressure values are in the range of

+/- 1e+48 at step 300
+/- 1e+65 at step 400
+/- 1e+85 at step 500

paraview (parafoam) is reporting pressure as "+/- inf", what is no problem at
that point.

Maybe the problem does not exist in every OpenFOAM version. I'm using the
32bit version in an Ubuntu 8.04 system, and will try to check with the 64bit
version on another machine.

How do i find out, whether I'm using single or double precision? and where
do I set it? (might be another point to check before messing too much with
the setup.

The airfoil testcase seemed to work properly.

Carl

[madad2005]

Carl, the test case appears to work fine with OpenFOAM-1.6.x compiled on Ubuntu-Studio 9.04. I was using the downloadable binaries with the release version when I had issues.