[Doug68]

Hi All,

It's be 3+ years since I last posted here, a long break from looking at OF and now I'm working mostly on trying to remember how I used to do things, but I'm not sure this is one of those issues.

Referring to the attached picture...

This is from a simple wind tunnel model,
You can see the forces in the X resolve to a basically steady condition but Y and most importantly Z keeps moving around. I don't believe the model can be that unstable.

This is from the last time stamp in the SimpleFoamParallelLog file:

smoothSolver: Solving for Ux, Initial residual = 0.000495791, Final residual = 2.66623e-005, No Iterations 3
smoothSolver: Solving for Uy, Initial residual = 0.00722292, Final residual = 0.000460959, No Iterations 3
smoothSolver: Solving for Uz, Initial residual = 0.00239546, Final residual = 0.000142308, No Iterations 3
GAMG: Solving for p, Initial residual = 0.0561813, Final residual = 0.00251192, No Iterations 2
time step continuity errors : sum local = 4.60976e-006, global = -6.60764e-009, cumulative = -1.52998e-005
smoothSolver: Solving for omega, Initial residual = 8.3893e-005, Final residual = 6.61511e-006, No Iterations 3
smoothSolver: Solving for k, Initial residual = 0.00104573, Final residual = 5.31296e-005, No Iterations 3
ExecutionTime = 45348.4 s ClockTime = 45348 s

Which to me looks like the forces have resolved ok, the check mesh passes ok also, so I'm not sure where to start to get a grip on this. Any ideas? I'm wondering if the settings on the solver for p might be where I need to research? I'm currently using the setting below a-la the motorbike tutorial:

p
{
solver GAMG;
tolerance 1e-7;
relTol 0.1;
smoother GaussSeidel;
nPreSweeps 0;
nPostSweeps 2;
cacheAgglomeration on;
agglomerator faceAreaPair;
nCellsInCoarsestLevel 10;
mergeLevels 1;

[Doug68]

Since the last post i've found the effect of relaxationfactors in fvsolution, running with:

relaxationFactors
{
fields
{
p 0.2; //0.3;
}
equations
{
U 0.3; //0.7;
k 0.3; //0.7;
omega 0.3; //0.7;
}
}

As opposed to the commented values I'm getting the results below, which are a lot more stable, I suspect I've over done it though.
I think probably need to run the thing again with twice the iterations to prove a smaller number of iterations needed to get a decent steady-ish value of Z.