https://linen.dev logo
Join Discord
Powered by
# les-ras
  • k
    Hey, I am having trouble with using the turbulentDFSEMInlet function for applying U, L and R boundary conditions. When i try to decompose (or run) my case, I get a few warnings regarding reynolds stresses saying it doesnt obey det(R)>=0. The case runs fine however i am concerned it might affect the results, any ideas what might cause it, and how to fix it? Attached is the decompose log and the R file where R is defined.
    logconversation
  • z
    The values you've provided for the stress tensor produce a determinant < 0 which isn't possible for a realistic flow - How have you come up with the numbers you've plugged in as the R field?
  • z
    I think the main problem would be issues taking the inverse of the matrix, but if it's not crashing then it might be fine? Definitely indicates your R field isn't appropriate though
  • z
    Negative determinant -> not positive semi-definite -> not realizable
  • z
  • k
    The R values are taken from a seperate model and extracted using a function object saving the field turbulenceProperties:R.
  • s
    Getting floating point error core dumped
  • s
    What things should I check to diagnose the problem?
  • s
    The checkMesh returns no problem
  • q
    We'll need to look at the context where it occurred, at what point of the simulation... The error is likely due to some divide by zero type logical error
  • s
  • s
    Here is the log
  • q
    Looks like there's some issue with initial/boundary conditions...
  • q
    epsilon shows clear indication of immediate blow up
  • s
    Thanks
  • u
    Could you share the fvSchemes file?
  • s
  • s
  • s
    pisoFoam gives an error stack. I have no idea how to diagnose the issue. It is flow through a device (the details are confidential , but it has an inlet and an outlet)
  • s
    Copy code
    GAMG:  Solving for p, Initial residual = 0.0503985, Final residual = 0.00323134, No Iterations 2
GAMG:  Solving for p, Initial residual = 0.0799968, Final residual = 0.00418961, No Iterations 2
GAMG:  Solving for p, Initial residual = 0.200912, Final residual = 9.82865e-07, No Iterations 50
time step continuity errors : sum local = 2.97211e-10, global = -1.05228e-13, cumulative = 9.51037e-09
#0  Foam::error::printStack(Foam::Ostream&) at ??:?
#1  Foam::sigFpe::sigHandler(int) at ??:?
#2  ? in "/usr/lib/libc.so.6"
#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 ??:?
#4  Foam::GaussSeidelSmoother::smooth(Foam::Field<double>&, Foam::Field<double> const&, unsigned char, int) const at ??:?
#5  Foam::smoothSolver::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const at ??:?
#6  Foam::fvMatrix<double>::solveSegregated(Foam::dictionary const&) at ??:?
#7  Foam::fvMatrix<double>::solve(Foam::dictionary const&) in "/opt/OpenFOAM/OpenFOAM-10/platforms/linux64GccDPInt32Opt/bin/pisoFoam"
#8  Foam::fvMatrix<double>::solve() in "/opt/OpenFOAM/OpenFOAM-10/platforms/linux64GccDPInt32Opt/bin/pisoFoam"
#9  Foam::SolverPerformance<double> Foam::solve<double>(Foam::tmp<Foam::fvMatrix<double> > const&) at ??:?
#10  Foam::RASModels::kEpsilon<Foam::incompressibleMomentumTransportModel>::correct() at ??:?
#11  ? in "/opt/OpenFOAM/OpenFOAM-10/platforms/linux64GccDPInt32Opt/bin/pisoFoam"
#12  ? in "/usr/lib/libc.so.6"
#13  __libc_start_main in "/usr/lib/libc.so.6"
#14  ? in "/opt/OpenFOAM/OpenFOAM-10/platforms/linux64GccDPInt32Opt/bin/pisoFoam"
Floating point exception (core dumped)
  • s
    Adding more information: I replaced the geometry with a simple cylinder (with the same inlet and outlet surface) and the simulation completes without any problem
  • q
    Well, in that case, do you think the error is likely due to some poor mesh elements? What's the mesh quality like?
  • q
    Also, for turbulence schemes, see if simple upwind instead of limitedLinear/minMod helps
  • k
    Also, check yPlus, kEpsilon will not like it much less than 30
  • s
    Copy code
    Checking geometry...
    Overall domain bounding box (-0.0561983 -0.00400735 -0.0625121) (0.0562626 0.00399265 0.0624879)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
    Boundary openness (4.01787e-17 -1.50628e-15 7.90755e-17) OK.
    Max cell openness = 3.85865e-15 OK.
    Max aspect ratio = 902.728 OK.
    Minimum face area = 8.71623e-10. Maximum face area = 3.30882e-05.  Face area magnitudes OK.
    Min volume = 1.56705e-14. Max volume = 5.056e-08.  Total volume = 6.71351e-05.  Cell volumes OK.
    Mesh non-orthogonality Max: 89.8525 average: 26.6205
   *Number of severely non-orthogonal (> 70 degrees) faces: 3393.
    Non-orthogonality check OK.
  <<Writing 3393 non-orthogonal faces to set nonOrthoFaces
    Face pyramids OK.
 ***Max skewness = 5.43393, 1 highly skew faces detected which may impair the quality of the results
  <<Writing 1 skew faces to set skewFaces
    Coupled point location match (average 0) OK.

Failed 1 mesh checks.
  • s
    I don't know where to find it. I did a search for the keyword
  • s
    This is pisoFoam RAS
  • s
    It errors out at the same spot
  • q
    There's a function object to generate it
  • q
    As you can see. The mesh is of poor quality. Maybe work on that and see how it goes.
1...424344...52Latest