[quantumwire]

Dear forum,

I'm by no means a knowledgeable openfoam user since I have no clue about CFD in general.

However,

my intention is to run a scalable openfoam benchmark for which I found this one:

https://github.com/do-jason/OpenFOAM...arks/motorbike

The benchmark looks good to me (I guess) since I could tune the size of the problem (MCells) and the number of cpus.

The problem with it is that the simulation explodes after the first few steps.

I've compiled two openfoam version (V2112 and V2106) and tested the 42 MCells setup using 16 cores and 1 TB of RAM.

With both openfoam versions I get this log in which the 'sum local' number in the 'time step continuity errors' line diverges extremely quickly leading to a simulation failure:

...
Time = 1

smoothSolver:  Solving for Ux, Initial residual = 0.287454, Final residual = 0.0159974, No Iterations 22
smoothSolver:  Solving for Uy, Initial residual = 0.670437, Final residual = 0.0578161, No Iterations 19
smoothSolver:  Solving for Uz, Initial residual = 0.713336, Final residual = 0.0384459, No Iterations 19
GAMG:  Solving for p, Initial residual = 1, Final residual = 0.131524, No Iterations 1000
time step continuity errors : sum local = 1.81912e+08, global = 861174, cumulative = 861174
smoothSolver:  Solving for omega, Initial residual = 0.113341, Final residual = 0.00964374, No Iterations 4
smoothSolver:  Solving for k, Initial residual = 1, Final residual = 0.0804192, No Iterations 4
ExecutionTime = 174.48 s  ClockTime = 174 s

forceCoeffs forceCoeffs1 execute:
    Coefficients
        Cd       : 2.2912e+14   (pressure: 2.2912e+14   viscous: 3.724e+07)
        Cs       : 1.5138e+13   (pressure: 1.51381e+13  viscous: -6.80952e+07)
        Cl       : 7.90224e+12  (pressure: 7.90232e+12  viscous: -8.50578e+07)
        CmRoll       : -3.08433e+13     (pressure: -3.08433e+13 viscous: 8.91239e+06)
        CmPitch       : 9.23788e+13     (pressure: 9.23788e+13  viscous: 2.55493e+06)
        CmYaw       : -1.7646e+13       (pressure: -1.7646e+13  viscous: -2.22275e+07)
        Cd(f)    : 8.37169e+13
        Cd(r)    : 1.45404e+14
        Cs(f)    : -1.0077e+13
        Cs(r)    : 2.52151e+13
        Cl(f)    : 9.63299e+13
        Cl(r)    : -8.84277e+13
Time = 2

smoothSolver:  Solving for Ux, Initial residual = 0.665592, Final residual = 0.0643762, No Iterations 47
smoothSolver:  Solving for Uy, Initial residual = 0.610635, Final residual = 0.0404073, No Iterations 58
smoothSolver:  Solving for Uz, Initial residual = 0.600655, Final residual = 0.0574642, No Iterations 45
GAMG:  Solving for p, Initial residual = 1, Final residual = 0.0199624, No Iterations 1000
time step continuity errors : sum local = 2.39263e+19, global = -7.23625e+16, cumulative = -7.23625e+16
smoothSolver:  Solving for omega, Initial residual = 0.107859, Final residual = 0.00710626, No Iterations 4
smoothSolver:  Solving for k, Initial residual = 0.381193, Final residual = 0.0353298, No Iterations 3
ExecutionTime = 341 s  ClockTime = 341 s

forceCoeffs forceCoeffs1 execute:
    Coefficients
        Cd       : 2.41032e+33  (pressure: 2.41032e+33  viscous: -7.07675e+19)
        Cs       : 5.40342e+33  (pressure: 5.40342e+33  viscous: 2.95738e+20)
        Cl       : 7.57467e+32  (pressure: 7.57467e+32  viscous: -3.32622e+19)
        CmRoll       : -2.66388e+33     (pressure: -2.66388e+33 viscous: -1.76725e+20)
        CmPitch       : 1.04103e+33     (pressure: 1.04103e+33  viscous: -3.59468e+19)
        CmYaw       : 8.57365e+31       (pressure: 8.57365e+31  viscous: 7.09699e+19)
        Cd(f)    : -1.45872e+33
        Cd(r)    : 3.86904e+33
        Cs(f)    : 2.78745e+33
        Cs(r)    : 2.61598e+33
        Cl(f)    : 1.41977e+33
        Cl(r)    : -6.62301e+32
Time = 3

[19] #0  Foam::error::printStack(Foam::Ostream&) at ??:?
[19] #1  Foam::sigFpe::sigHandler(int) at ??:?
[19] #2  ? in /lib/x86_64-linux-gnu/libc.so.6
[19] #3  Foam::GaussSeidelSmoother::smooth(Foam::word const&, Foam::Field<double>&, Foam::lduMatrix const&, Foam::Field<double> const&, Foam::FieldField<Foam::Field, double> const&, Foam::UPtrList<Foam::lduInterfaceField const> const&, unsigned char, int) at ??:?
...

Any advise about stabilizing this simulation is greatly appreciated.