CFD Online Discussion Forums - Floating point exception error on dambreak?

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)

- - Floating point exception error on dambreak? (https://www.cfd-online.com/Forums/openfoam-solving/212767-floating-point-exception-error-dambreak.html)

Floating point exception error on dambreak?

Hello all!

I'm getting a floating point exception error on the dambreak using 4 cores on the same machine. At first I didn't have a swap enabled and the machine would freeze up then I added one and the sim ran up to this part

Code:

PIMPLE: iteration 1

smoothSolver:  Solving for alpha.water, Initial residual = 0.000862611, Final residual = 1.98603e-09, No Iterations 3

Phase-1 volume fraction = 0.0224762  Min(alpha.water) = -4.83592e-08  Max(alpha.water) = 1

MULES: Correcting alpha.water

MULES: Correcting alpha.water

Phase-1 volume fraction = 0.0224762  Min(alpha.water) = -8.61764e-05  Max(alpha.water) = 1.00006

DICPCG:  Solving for p_rgh, Initial residual = 0.0821221, Final residual = 0.0038929, No Iterations 1

time step continuity errors : sum local = 0.000242214, global = -7.39702e-10, cumulative = 2.2925e-08

DICPCG:  Solving for p_rgh, Initial residual = 0.0046728, Final residual = 0.000226397, No Iterations 6

time step continuity errors : sum local = 1.32212e-05, global = -7.6524e-10, cumulative = 2.21598e-08

--------------------------------------------------------------------------

mpirun noticed that process rank 3 with PID 3622 on node turbulence exited on signal 8 (Floating point exception).

I ran it using
blockMesh
setFields
decomposePar
mpirun -np 4 interFoam -parallel > log &

Below is the project itself.
Attachment 67187

More details, apparently this was sent to the console in bash

Code:

#0  Foam::error::printStack(Foam::Ostream&) at ??:?

[3] #1  Foam::sigFpe::sigHandler(int) at ??:?

[3] #2  ? in /lib/x86_64-linux-gnu/libc.so.6

[3] #3  Foam::PCG::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const at ??:?

[3] #4  Foam::fvMatrix<double>::solveSegregated(Foam::dictionary const&) at ??:?

[3] #5  Foam::fvMatrix<double>::solveSegregatedOrCoupled(Foam::dictionary const&) at ??:?

[3] #6  Foam::fvMesh::solve(Foam::fvMatrix<double>&, Foam::dictionary const&) const at ??:?

[3] #7  ? at ??:?

[3] #8  __libc_start_main in /lib/x86_64-linux-gnu/libc.so.6

[3] #9  ? at ??:?

[turbulence:03622] *** Process received signal ***

[turbulence:03622] Signal: Floating point exception (8)

[turbulence:03622] Signal code:  (-6)

[turbulence:03622] Failing at address: 0x3e800000e26

[turbulence:03622] [ 0] /lib/x86_64-linux-gnu/libc.so.6(+0x354b0)[0x7f21297c54b0]

[turbulence:03622] [ 1] /lib/x86_64-linux-gnu/libc.so.6(gsignal+0x38)[0x7f21297c5428]

[turbulence:03622] [ 2] /lib/x86_64-linux-gnu/libc.so.6(+0x354b0)[0x7f21297c54b0]

[turbulence:03622] [ 3] ~/OpenFOAM/OpenFOAM-v1806/platforms/linux64GccDPInt32Opt/lib/libOpenFOAM.so(_ZNK4Foam3PCG5solveERNS_5FieldIdEERKS2_h+0xd04)[0x7f212aaff6a4]

[turbulence:03622] [ 4] ~/OpenFOAM/OpenFOAM-v1806/platforms/linux64GccDPInt32Opt/lib/libfiniteVolume.so(_ZN4Foam8fvMatrixIdE15solveSegregatedERKNS_10dictionaryE+0x15b)[0x7f212d5eafeb]

[turbulence:03622] [ 5] ~/OpenFOAM/OpenFOAM-v1806/platforms/linux64GccDPInt32Opt/lib/libfiniteVolume.so(_ZN4Foam8fvMatrixIdE24solveSegregatedOrCoupledERKNS_10dictionaryE+0x3d3)[0x7f212d256e53]

[turbulence:03622] [ 6] ~/OpenFOAM/OpenFOAM-v1806/platforms/linux64GccDPInt32Opt/lib/libfiniteVolume.so(_ZNK4Foam6fvMesh5solveERNS_8fvMatrixIdEERKNS_10dictionaryE+0x23)[0x7f212d2014c3]

[turbulence:03622] [ 7] interFoam[0x446872]

[turbulence:03622] [ 8] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7f21297b0830]

[turbulence:03622] [ 9] interFoam[0x44ed79]

[turbulence:03622] *** End of error message ***

Quote:

Originally Posted by massive_turbulence (Post 718435)

Code:

PIMPLE: iteration 1

smoothSolver:  Solving for alpha.water, Initial residual = 0.000862611, Final residual = 1.98603e-09, No Iterations 3

Phase-1 volume fraction = 0.0224762  Min(alpha.water) = -4.83592e-08  Max(alpha.water) = 1

MULES: Correcting alpha.water

MULES: Correcting alpha.water

Phase-1 volume fraction = 0.0224762  Min(alpha.water) = -8.61764e-05  Max(alpha.water) = 1.00006

DICPCG:  Solving for p_rgh, Initial residual = 0.0821221, Final residual = 0.0038929, No Iterations 1

time step continuity errors : sum local = 0.000242214, global = -7.39702e-10, cumulative = 2.2925e-08

DICPCG:  Solving for p_rgh, Initial residual = 0.0046728, Final residual = 0.000226397, No Iterations 6

time step continuity errors : sum local = 1.32212e-05, global = -7.6524e-10, cumulative = 2.21598e-08

--------------------------------------------------------------------------

mpirun noticed that process rank 3 with PID 3622 on node turbulence exited on signal 8 (Floating point exception).

I ran it using
blockMesh
setFields
decomposePar
mpirun -np 4 interFoam -parallel > log &

Below is the project itself.
Attachment 67187

hi, pls have a look at the Min alpha value which actually negative, it is not physically acceptable. try to change the time step ( small time step) and see if it helps

Quote:

Originally Posted by xneop (Post 718540)

hi, pls have a look at the Min alpha value which actually negative, it is not physically acceptable. try to change the time step ( small time step) and see if it helps

Thanks for replying. So far what I've noticed is that scotch and metis are both really slow compared to simple decomposition. Could you explain why that might be. I managed to get a good run with simple decomposition so that's why I'm asking if this particular case is not right for that (scotch or metis) decomposition. When is it a good idea to use either scotch or metis btw?

I'll trying running with smaller time steps using simple as the two other ones I mentioned are slow just as they are. Am I missing something if one decomposition is faster than another using the same timestep size?