Channel3d

[Bran]

Dear developers,
I can’t match the results you guys published on Large-eddy simulation coupled with wall models for turbulent channel flows at high Reynolds numbers with a lattice Boltzmann method — Application to Coriolis mass flowmeter. Especially when I use the Musker profile, the u_tau I get is always lower than the expected value(0.05). My u_tau is around 0.045. When I increase the resolution, it gets worse. So what’s the trick there, how can I get it right? And why did you guys implement a bodyForce after computing u_bc? I tried without it, almost no difference. Also, in the paper I mentioned above, you guys said that you used Musker profile to approximate the AVERAGE wall shear stress, why is it AVERAGE? In my point of view, I think it should be INSTANT. As for the power-law function, I can’t see how you guys derived formula(22) from formula(21). The formula you guys used in the code is not the same one you guys mentioned in the paper. In the end, it will be appreicated very much if you guys can answer my silly questions. That will be really helpful.
Live long and prosper.

// save tau_w for next step
cell.template setField<descriptors::TAU_W>(&(tau_w[0]));
// STEP 6 : compute velocity vector at the boundary
u[0] = e_x_loc[0] * u1[0];// u_bc
u[1] = e_x_loc[1] * u1[0];
u[2] = e_x_loc[2] * u1[0];

if (_wallFunctionParam.bodyForce) {
auto force = cell.template getFieldPointer<descriptors::FORCE>();
for (int iDim=0; iDim<DESCRIPTOR::d; ++iDim) {
u[iDim] -= force[iDim] / 2.;
}

[FBukreev]

Hello Bran,

in the release version of the wall model it is not possible to choose the height of the velocit ymapping point. Close to the wall u_tau is always calculated lower then DNS value.
Body force is firstly substracted and then added during the simulation step, so at the end the velocity is not influenced by the force.
If you need more help please write us at info@openlb.net

[Bran]

Danke!
I still have a question. In the paper–Large-eddy simulation coupled with wall models for turbulent channel flows at high Reynolds numbers with a lattice Boltzmann method — Application to Coriolis mass flowmeter, the u_tau you guys used to plot, is it the DNS value(0.05), or the value you calculated? If I use the DNS values, my results look fine. But if I used the u_tau I calculated, my results are not good.

[FBukreev]

Yes, the results are evaluated with DNS shear velocity. The shear velocity that is calculated locally will variate depending on the wall distance. You can also compare the not normalized velocit yvalus with DNS to understand how far are the curves from each other.

[Author]

Posts