I am thinking of simulating the flow of water on a tank through holes at the bottom. It can be hand calculated to estimate the time required to drain out the water but I want to do some CFD on it.
It won't have an inlet. I just want to visualize how flow exits and how long it takes to completely empty the tank. Is it even possible to do so in Fluent?
Hi Nikesh,
I think it is fairly simple to model it in FLUENT. Multiphase model - VOF approach is what you have to use. You can initialize the tank with fluid volume, in this way you will not require inlet for the tank.
Ahh that sounds interesting. How about the BCs for outlet? I used pressure-outlet (o gauze pressure) and I'm getting reversed flow there.
Oh! Most of the time pressure outlet BC works. Try using outflow BC. It would be useful, if you can attach some pics. So, that i have correct understanding of the problem.
Refer to the figure attached.
The box/tank is filled with water and has no inlet. Water exits from the pipes at the bottom. I can also use symmetry option to simulate only 1/4th of the tank.
I want to simulate a transient case to visualize overall draining of the water.
Pressure outlet boundary condition should work for your problem. You may get some reversed flows in the initial runs (which is OK), after which it would be normal to the flow direction. I would highly recommend you to use 1/2 symmetry. Since, gravity is important for your problem, 1/4 th symmetry may not work.
Help me out with the setups here, something is not working.
General: pressure-based, transient, Absolute velocity formulation, Gravity enabled 9.81 m/s2
Models: Multiphase VOF, 2 Eulerian phases, explicit scheme; Viscous model k-epsilon with Standard wall functions
Phases: Phase-1 Water, Phase-2 Air
Then I get a little messed up with the BCs. How do I allocate BC types to the zones and phases?
Gravity would be -9.81 m/s2. I would suggest you to define water in the domain as an initialization. This can be done as patch option in initialization tab. Define volume fraction of phase-1 as 1 or phase-2 as 0. I do not think you have make any changes with the other boundary conditions. Keep them as default values. Before running the simulation, just make sure you check whether you initialized correctly in the post processor.
Finally managed to work it out on 2D for now. Thanks a lot for your quick response.
I'm trying few cases and will then move to 3D flow case.
- Badges
- Science topic