Inlet, no Outlet
 

Hello All,

I am looking to see the time it takes, and the pressure created when a flow of oil is introduced to a previously air filled chamber. I am well versed in inlet/outlet flow and rotational closed domain flow, but this one stumps me.

Thanks

Can you describe it a bit more completely? Do you have a chamber filled with air at time zero, and then you add some oil to it (how? under pressure or a defined flow rate?) and the pressure in the air chamber increases? Does the chamber have any inlets or outlets? Is it steady state or transient?

Yes, a chamber is filled with air at t=0s. This chamber has no outlet, only an inlet. At t=0s oil is introduced through this inlet at P=2 atm (just random numbers), consistently applied. The air would be displaced, and some diffusion into the oil but i'm not worried about that so much. I suppose it could be modeled as an empty chamber at t=0s and then oil added and it would still be accurate.
While the boundary conditions are steady state, I would say the problem is a transient one. The goal of the analysis is to determine how long this is taking since the fluid is VERY viscous and the inlet is rather small.

This sounds like something which can be worked out with elementary thermodynamics as a pen and paper calculation, not needing CFD at all. Why are you thinking of doing it using CFD?

I would agree to be honest, pressure drop, mass flow rate, etc. all easy enough to look into. But I am often asked to do things which produce "pretty pictures". Honestly, I may do the hand calculations and then force a simulation to match. But ultimately, yes, knowing the entry conditions, fluid properties, etc. it is an extremely elementary problem.

The simulation you describe appears to be a straight forward transient compressible gas/oil free surface simulation. But to get a mesh fine enough to look "pretty" will require a fine mesh, and that means a long run time. So this model is going to take a while to develop and a while to run. If all you want is pretty pictures then I recommend photoshop :)

Hi

I would like to revive this topic by revisiting the no-outlet concept. I want to model a person releasing breath inside a room in steady state just to make him the source of a certain species. Even though the introduced mass-flow-rate through his mouth is small, I want to understand how this affects the continuity. Will it converge or do you think there is anything I should pay attention to. Energy equation and Species transport will be enabled in addition to flow.
I am asking this because there are many other features in this room which will cause a large mesh and I want to make sure of a few things before I start generating a mesh and running the simulation.

Thank you very much

If you are planning to do this as a transient simulation then this is straight forward. Generate an additional variable (or other approach) and watch it progress over time.

If you want a steady state simulation you will need to think about how to define this in steady state. Your person is breathing in and out. How are you going to model that? If you model it as (for instance) a time-averaged flow rate source, then where is that flow going to go? The air comes in through the person, it has to go out somewhere. You can't have the person's breath pressurise the whole room.

Thank you for your answer. Unfortunately, I don't have the access to resources for a transient run. I was hoping to have a steady case first, with an airflow field and a breath jet at the mouth and then enable the species transport model so that I have a species entering the domain at a certain rate. I did a preliminary run and indeed there was a pressure increase but also at the same time now I am thinking whether this is just affordable as the inlet BC is mass flow inlet.

If you want to model this as steady state then you need to define somewhere for the air to go. You don't pressurise a room when you breathe out - the room leaks air so the pressure does not change. So you need to put this air leak in somewhere.

Why are you doing this as species transport? Unless the change in species is having an effect on fluid properties (which sounds unlikely for air breathed out from a person) then the additional variable approach would be more suited and much cheaper.

I am calculating CO2 transport and indeed my assumption was the species transport did not affect the airflow (If I understand you correctly here).

I also went with species transport because it looked simpler than scalar transport.

And also, only now I realize I have been writing under the CFX Forum even though I have been working on Fluent! So sorry if this has been a waste of time for you!