Periodic Outlet Boundary Conditions ( nonequilibrium extrapolation method)

[Henderson_C]

Hello everyone,

I was wondering how OpenLB tackles the migration of certain object (droplet, bubbles, ….) through the domain outlet. I read some papers, and many has recommended the use of nonequilibrium extrapolation method. Is there an implementation of such method in OpenLB?

Regards,

[FBukreev]

Hello,

we have particle counter by the Lagrangian particles and the mass counter by Eulerian particles at the outlet. Which one do you want to use?

Greetings
Fedor

[Henderson_C]

Dear Fedor,

Thank you for replying. Can you elaborate on the difference between them?

Thanks

[FBukreev]

In first case you track the separate points moving with Newtonian equation system and in the Eulerian case you regard particles as a concentration cloud.

[Henderson_C]

Dear Fedor,

Can you show me both implementations as I am now indecisive of which method to use.

[FBukreev]

Look in the examples/particles/bifurcation/eulerLagrange and eulerEuler. There you can see it clear.

[Henderson_C]

Dear Fedor,

I want the particle to move to the flow inlet when it passes through the flow domain outlet. I have set the periodicity to true, but when the particle touches the outlet, it sticks to the outlet wall.

regards,

• This reply was modified 1 year ago by Henderson_C.

[FBukreev]

Which particles do you use? Lagrangian or Eulerian? If you have set the periodicity on the particle Lattice there must be no outlet and inlet at all.

[Henderson_C]

Dear Fedor,

I am not using the super particle system. I am simply using the single particle simulation. The particle is flowing through a channel with an inlet and outlet. To achieve fully hydrodynamic condition, I set the periodicity in the x-direction to be true (cuboidGeometry.setPeriodicity(false, true)). I thought once I enable this feature, any flow interruption due to the particle motion (Brownian motion) beside the outlet will be shown in the inlet of the channel. But that is not the case.

[FBukreev]

Hello,

dou you simulate the resolved moving particle?

[Henderson_C]

Dear Fedor,

Yes, similar to one used in dkt2d example.

regards,

• This reply was modified 1 year ago by Henderson_C.

[FBukreev]

You can add the periodicity vector (false,true) at the end of the ParticleManager, look also at the ParticleManger code in src.

[Henderson_C]

Dear Fedor,

By that you mean modify the source code (src/particles/particlemanager) with the periodicity vector?

regards,

[FBukreev]

I mean that you can look at the parameters of the particlemanager in the src and than put the periodicity vector as argument on the right place in the particlemanager declaration in your simulation setup.

[Henderson_C]

Dear Fedor,

I am not developer not I am expert in c++ but to understand correctly, I need to modify the ParticleManager constructor to accept the periodicity vector as an argument, and initialize the periodicity member variable with it. Is that correct? When you mention the parameters, what do you mean exactly? I am really confused on how to solve my issue.

Your continuous support is really appreciated.
Regards,

[Author]

Posts