[rjmcsherry]

Hi all,

I'm trying to set up a problem that will involve flow between two chambers, each of which will be partially filled with water. There will therefore be two free surfaces (water-air boundary), that will rise and fall as the respective chambers fill and empty.

Does anyone have any feeling as to whether or not this will be possible please, and what the likely pitfalls might be, particularly at the water-air interfaces? I have limited experience in both CFX and free-surface flows, and so would be grateful if anyone could share their experiences of similar problems.

Thanks in advance for any advice,

Richard

[ghorrocks]

You might be able to do this without a free surface and save all the hassle of multiphase modelling. One option is to put a pressure boundary on each side and control the pressure to match the static head above it. Another option is to mesh up to the free surface and use moving mesh to move the surface up and down. Both of these approaches simplify the simulation down to a single phase model.

But whether this is appropriate will depend on what you want to simulate, for instance does the free surface deform, or just move up and down. Also, is inertia of the water column important?

And you can also model this as a multiphase free surface model. Use the tips on the free surface flow over a bump tutorial with CFX to get started.

[rjmcsherry]

Hi Glenn,

Thanks very much for your detailed response, it is much appreciated. I hadn't thought about applying varying pressure boundary conditions to mimick the static head... This could be an ideal solution, as the deformation of the two free surfaces are not important at all (at least not at this stage of the work). However I suspect that the inertia of the water column may be important to the dynamic behaviour of the system (it's a wave energy device in which head difference between two chambers, separated by a turbine, drives the turbine).

I will have a think about the implications of applying the pressure boundary conditions as you've suggested.

Thanks again - this might prove to be very useful.

Richard

[AliTr]

in case of single phase modeling, you can also use a subdomain with :
Momentum Source Z component = - ro * g
to replicate hydrostatic pressure in a single phase model, this will drive the flow and keep the right balance between chambers, but still need to take care of mesh deformation at water surface.

[vishal.indomitable]

Dear,
There is a Research Paper on similar problem like yours- " Development of Free Surface Flow Between Cylinders with Vertical Axes" By T Watanabe, Y Toya and I Nakamura in IOP Science. You can use it for reference.