Learning Forum

[christian.katsavos]

Hello everyone,

I have a simple question to ask, which seems to be one of the hardest to answer. ?

What is the most correct way to model a pipe dispersing a fluid into the atmosphere, in CFX? Just imagine a cylindrical pipe which releases a subsonic given flow rate of fluid into an open environment (the problem is monophasic, so the fluid is air dispersed into the earth atmosphere without any wind , or water dispersed into a sea without any considerable current). I'm interested in studying how the flow paths behave in the free atmosphere, very close to the pipe exit line (and not far away from it, which must be too difficult or even impossible!) .

Maybe the right approach to follow is to model the atmosphere as a big volume limited by a surface where an "open" boundary condition is used. But what else? How big must this volume be? What about it shape? And, most of all: what about the "open" boundary condition settings?

Thanks so much for your help, I'm sure this can be extremely useful to a lot of users.

Christian

[UniversityofQLD]

Your model design and boundary conditions really depend on what exactly you want to find out. Have a look at relevant papers which might be modelling something similiar to what you would like to achieve. Check what equationsand setups they are using and if you can apply them to your problem. Some suggestions:
Erinc Dobrucali, Selma Ergin: An investigation of exhaust smoke dispersion for a generic frigate by numerical analysis and experiment
Oliver Mărunţălu, Gheorghe Lăzăroiu, Elena Elisabeta Manea, Dana Andreya Bondrea, Lăcrămioara Diana Robescu: Numerical Simulation of the Air Pollutants Dispersion Emitted by CHP Using ANSYS CFX
Teixeira, Jose C. F.;Teixeira, Senhorinha F. C. F.;Cunha, Pedro;Silva, Angela M.: THERMAL DRIVEN DISPERSION OF SMOKE IN A PARKING SPACE

[christian.katsavos]

Thanks so so much for your answer.
The first paper is unfortunately written in turkish;
In the second paper, the atmosphere is described as modeled with an "Open" boundary condition, which confirms what I thought, but no other detail is provided;
The third paper (at least that's what it looks to me) is not distribuited for free and I can't buy it, unfortunately.
I'm going to try and search deeper. In the meantime, I wish someone who has already encountered this kind of problem will help. Any kind of advice is more than appreciated.
Many thanks again

[Rob]

As says, you need to give more detail. Is this a continuous slow release, a high speed (near supersonic) high pressure pin hole, full bore breach etc Simulation is applied to the (sometimes very specific) problem so there may be several possible solutions based on what you need to know.

[christian.katsavos]

Rob, and thanks for your help.
The application I'm studying foresees a not continuos but periodic release (sinusoidal to zero in the time domain). Top speeds of the fluid inside the pipe are around 1m/s, so definitely it's not a supersonic flow . The pipe diameter is around 50 -75 mm, it will be a full bore breach. The pressure peak inside the pipe is around 1.5 barg.
Really thanks for your help!

[Rob]

Are you emptying the pipe or is pressure pulsed?

[christian.katsavos]

it's pressure pulsed

[christian.katsavos]

It's Pressure pulsed

[Rob]

Long weekend in Europe :)
Assuming CFX behaves the same as Fluent (can correct me later) I'd run a starting case at minimum, mean and maximum pressure in steady state. Pressure inlet venting into an open area. Use these to fine tune settings and the mesh. Flow will be compressible, so ideal gas density, and check the tutorials for settings, the NACA airfoils will be a good source of information.