High Re Flows Calculation Fault

[steed188]

Hello
I’m new in LBM and trying to use SmargorinskyBGKDynamics to simulation a High Reynoalds flow but the result is always diverging.
I consider the material as air crossing a 1m*1m*5m wind tunnel , parameters like belows
CharL=1
CharU=5
CharNu=1e-5
latticeL=0.02
latticeU=0.1

I doubt there is any fault in the parameters. If I change the CharNu higher such as 1e-3, the result seemed good.

By the way, I imitate examples outflow as below:

sLattice.defineDynamics(superGeometry, 4, &bulkDynamics);
bc.addPressureBoundary(superGeometry, 4, omega);
AnalyticalConst3D<T,T> rhoF(1.);
lattice.defineRho(superGeometry, 4, rhoF);

I’m not quite understand it ,does it mean that the outflow is a constant pressure boundary?

In my past CFD simulation, I usually define the outflow a free pressure boundary that the pressure gradient is zero. How can I realize it?

your best
steed188

[mathias]

Periodic BC might be more stable. We also have free outflow conditions. Please have a look in the dogygen documentation. It is called “convection” BC and simular to the one paper of Junk form Konstanz.

Best
Mathias

[Author]

Posts