[Hme]

Hello fellow foamers!

I am having trouble choosing the correct boundary-conditions for my gas dispersion case, where H2 gas and air is to flow freely out of the domain, into the atmosphere. Air should also be allowed to flow freely into the domain.

With gravity set to a non-zero value in "constant/g" I experience a wall-like effect as the gas cloud nears the boundary, as if the H2 does not escape the domain. The flow velocity is non-zero, albeit small in y-direction on the boundary, so some gas/air clearly escapes.
This problem is not present when gravity is set to "( 0 0 0 )", which I find odd...
The effect seems limited to the y-normal boundary (the top) which is also normal to the direction of "g".

I have tried different models for the 5 sides, including:

• totalPressure
• prghPressure
• prghTotalPressure
• fixedFluxPressure

About the case:

• The mesh is made using blockMesh + snappyHexMesh with jet + obstacle .obj
• Free atmosphere on all 5 sides, all except ground which has noSlip
• Only dispersion, hence "noCombustion" model and inert H2/Air components
• Case folder is included below

The pictures show a jet of pure H2, an obstacle, H2 threshold between (0.01, 1), ground as a solid colour and the outline of the domain.

"g" set to ( 0 -9.81 0 )



"g" set to ( 0 0 0 )



Vertical jet, to confirm gravity & buoyancy is working with g set to ( 0 9.81 0 )


Video, with better resolution and gravity enabled:
https://www.dropbox.com/s/ug3j1qi24zjbxhc/sim.ogv?dl=0


If anyone has some insight into what I should do differently, it would be greatly appreciated


Sincerely
~Hovrud

[slr]

Hello Henrik,


I am having the same issue when applying these boundary conditions in buoyantBoussinesqPimpleFoam and twoLiquidMixingFoam. I am guessing the reason may be the zeroGradient boundary condition (including other boundary conditions that switch from zeroGradient and fixedValue depending on the flow direction at the boundary, e.g. freestream) that I used for temperature (in buoyantBoussinesqPimpleFoam)/ alpha (the volume fraction in twoLiquidMixingFoam). What boundary conditions did you use for your H2 and air? Do you have any progress in finding the reason solving this problem?


Kind regards
SL

[kcjarvis56]

Try type symmetry for the BC of the sides you should get the results you are expecting. You will have to set the sides to symmetry in the blockMesh also.