CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Transient two-phase flow in a pipe (

cpadam February 4, 2013 12:36

Transient two-phase flow in a pipe
I want to model the flow of a slug of water driven through a pipe by a high-pressure gas (hundreds of atmospheres). The gas and slug are initially at rest and discretely separated in a pipe with one end sealed and one end open to the atmosphere. The gas is at some initial pressure. At t=0 the gas is allowed to push (and mix with) the water slug. After traveling a short distance, both phases discharge into the atmosphere from the open end of the pipe.

The two things I want to get out of this simulation are the dispersion pattern of the liquid slug as it breaks up in and out of the pipe, and also the pressure drop in the gas at various locations.

How should I approach this problem in Fluent? That is, what multiphase model would be appropriate, and should I model the liquid with a dynamic mesh? I have only done single-phase transient models in Fluent so far.


msaeedsadeghi February 5, 2013 01:28

No need of dynamic mesh. I have already done liquid-gas breakup modeling before, take a look at my website, you may could find usefull documents.

cpadam February 5, 2013 16:45

Thanks Saeed. I will check it out.

cpadam February 15, 2013 12:37

I'm still having trouble with this problem. I need to model multiphase, which means I have to use the pressure-based solver. I cannot for the life of me get this solver to work, even with a minimal test case. My test case is a shock tube that discharges into the atmosphere (transient simulation). The tube is pressurized to 3 atmospheres initially. After that, I expect to see some flow out of the tube, but instead the pressure in the tube just drops to ambient instantaneously. I'm using a time step of 1e-6 seconds, and I have tried both the coupled and SIMPLE schemes. I have a suspicion that the x-velocity is not converging, due to the high residuals I'm getting.

Can anyone suggest what may be wrong with my model, or what else I can try? Thanks.

All times are GMT -4. The time now is 05:25.