Boundary condition problem

Submitted by shahab.zaman on Wed, 08/22/2018 - 16:45

Dear Users,
I am using the periodicChannel tutorial to simulate my case. Therefore I need to change the velocity boundary condition in U file from cyclic to nonuniform List for the inlet and define the outlet as zeroGradient. I also defined for the pressure in outlet fixedValue, uniform zero and for the pressure inlet zeroGradient. But at the end the velocty profile do not appear smoothly and there is pressure drop in inlet. I have done the same boundary for the simpleFoam solve in openFoam and the result sounds acceptable. I also set only 1 particle in the whole geometry to avoid their effect on the flow field.
I would appreciate any help

Best regards
Shahab

neniu's picture

neniu | Thu, 08/23/2018 - 16:55

Hi Shahab,

you are starting from a periodic channel, i.e. an infinitely long domain which is modeled by a piece of a certain extent with -- as the name states -- periodic boundary conditions. The flow in the domain is driven by a volume force term specified in system/fvOptions. This force term is adapted in every time step to drive a certain mass flow through the domain.
One should note here, that the flow is NOT driven by the boundary conditions which also means that you do not need an inlet condition with an inlet velocity or inlet mass flow. One benefit is that you do not need to take care to set a specific velocity profile at the inlet as as the velocity profile (in the entire domain and thus also at the inlet patch) develops over time due to the friction at the walls.

From here I see two options to fix your case:
1. Restore the periodic boundary conditions. With this setup you should not get discontinuities in streamwise direction.
2. If your desired setup should not use periodic condition you also need to deactivate the volume force term and should then be able to run the simulation with inlet and outlet conditions. In a second step you also need to adapt the input script for the DEM to continuously insert particles with an insert/stream (in case you want more particles).

Best regards
Martin

Martin Niemann, DCS Computing

shahab.zaman | Tue, 08/28/2018 - 15:27

Hi Martin,
Thanks a lot for your complete and nice information. Actually my goal is to minimize my geometry to be able simulate a case with micro particles. But I want to have complete profile velocity within the geometry. For I am using symmetry boundary in 2 dimension direct (y & z) and decreasing the length.
Do you have also any infos about effect of Ksl, Us and void fraction boundary condition, on simulation!?

Best regards
Shahab