[fedez91]

Hi everyone,

I am implementing a solver where as a part of it I introduce a advection diffusion equation.

fvScalarMatrix alphaAirEqn
(
    fvm::ddt(alphaAir) 
  + fvm::div(phiAir,alphaAir)
  - fvm::laplacian(kappaEff, alphaAir)
);

Info << "Before Solve" << endl;
alphaAirEqn.solve();
Info << "After Solve" << endl;

The simulation seems to run fine until suddenly it blows up with no type of warning of indication of instability. Here I add the output for the last two time steps.

Courant Number mean: 0.114512 max: 0.610134
Interface Courant Number mean: 0.00675904 max: 0.498688
deltaT = 0.00714286
Time = 0.0571429

PIMPLE: iteration 1
smoothSolver:  Solving for alpha.water, Initial residual = 0.00521154, Final residual = 4.30428e-09, No Iterations 3
Phase-1 volume fraction = 0.130194  Min(alpha.water) = -9.8495e-15  Max(alpha.water) = 1
MULES: Correcting alpha.water
MULES: Correcting alpha.water
Phase-1 volume fraction = 0.130194  Min(alpha.water) = -4.79255e-06  Max(alpha.water) = 1
DICPCG:  Solving for p_rgh, Initial residual = 0.0190907, Final residual = 0.00031443, No Iterations 2
time step continuity errors : sum local = 0.000880848, global = -8.58739e-06, cumulative = 1.10099e-05
DICPCG:  Solving for p_rgh, Initial residual = 0.000426879, Final residual = 1.66463e-05, No Iterations 15
time step continuity errors : sum local = 4.57798e-05, global = 4.31204e-06, cumulative = 1.53219e-05
DICPCG:  Solving for p_rgh, Initial residual = 0.000124589, Final residual = 8.68277e-08, No Iterations 53
time step continuity errors : sum local = 2.48986e-07, global = -2.59522e-08, cumulative = 1.5296e-05
smoothSolver:  Solving for epsilon, Initial residual = 0.0125835, Final residual = 7.23496e-07, No Iterations 6
smoothSolver:  Solving for k, Initial residual = 0.0853305, Final residual = 5.3777e-07, No Iterations 9
Before Sa
smoothSolver:  Solving for alphaAir, Initial residual = 0.135861, Final residual = 3.69369e-09, No Iterations 12
After Sa
ExecutionTime = 0.31 s  ClockTime = 0 s

Courant Number mean: 0.144905 max: 0.789675
Interface Courant Number mean: 0.00910855 max: 0.600491
deltaT = 0.00857143
Time = 0.0657143

PIMPLE: iteration 1
smoothSolver:  Solving for alpha.water, Initial residual = 0.00697895, Final residual = 3.22405e-11, No Iterations 4
Phase-1 volume fraction = 0.130194  Min(alpha.water) = -4.24805e-12  Max(alpha.water) = 1
MULES: Correcting alpha.water
MULES: Correcting alpha.water
Phase-1 volume fraction = 0.130194  Min(alpha.water) = -1.17147e-05  Max(alpha.water) = 1
DICPCG:  Solving for p_rgh, Initial residual = 0.0287047, Final residual = 0.000485994, No Iterations 2
time step continuity errors : sum local = 0.00177274, global = -2.53438e-06, cumulative = 1.27616e-05
DICPCG:  Solving for p_rgh, Initial residual = 0.000654371, Final residual = 2.85206e-05, No Iterations 15
time step continuity errors : sum local = 0.00010096, global = 1.20234e-05, cumulative = 2.4785e-05
DICPCG:  Solving for p_rgh, Initial residual = 0.000197295, Final residual = 9.52018e-08, No Iterations 54
time step continuity errors : sum local = 3.55484e-07, global = -3.56348e-08, cumulative = 2.47494e-05
smoothSolver:  Solving for epsilon, Initial residual = 0.0170451, Final residual = 8.69067e-07, No Iterations 7
smoothSolver:  Solving for k, Initial residual = 0.0966996, Final residual = 9.68412e-07, No Iterations 10
Before Sa

--> This is where it just stop running

There clearly a problem with the line alphaAirEqn.solve();, but I have no idea what type of error I may be facing. Maybe the linear system of equations has no solution?

Additional Info:

The ADE should have a source term (S), but in order to have more control of how alphaAir is created/destroyed, I decided to calculate the creation/destruction of alphaAir before the equations (S*DeltaT), add it to the field, and then just solve the ADE without the source term. I don't know if this may be an issue when trying to solve the linear system.

[HPE]

Hi,

- Does it crash without any error message?
- Or is it frozen? May be some infinite loop?
- What settings are you using for the linear solver? Have you tried any other linear solver for that equation?
- Can you constrain the number of iterations for the linear solver with `maxIter` entry (or maxIters?)?
- Can you compile the solver with the debug flag and rerun the case?

[fedez91]

Hi HPE,

- Does it crash without any error message?
- Or is it frozen? May be some infinite loop?

It crashes immediately. No message.

- What settings are you using for the linear solver? Have you tried any other linear solver for that equation?

I tried it now and with GAMG it runs ok.. I don't know if I want to dig deeper to find what was wrong with the smoothSolver. Thanks!