Hello Friends:

I am currently involved in the modeling of two-phase flow and heat transfer in a structure of small round tubes whose inside diameters range between 2-10 mm. My work involves modeling a two-phase flow [of a known inlet flow regime] that first passes through a small diameter vertical adiabatic tube, then bends by 90 degrees via a connecting elbow and enters a larger diameter horizontal tube. Then, the flow undergoes condensation while passing through an inclined row of horizontal tube passes arranged in a serpentine fashion using U-bends.

I would greatly appreciate any information on the two-phase flow capabilities of the commercially available CDF tools/codes that could help me model the steady-state pressure drop and heat transfer characteristics of this particular flow geometry in various flow sections. I will perform flow visualization studies, and will experimentally verify the computational results, later. I would also like to know if the CFD tools have capabilities for simulating the transient response of such a system for a time-dependent inlet void fraction.

If my desired CFD capabilities for modeling steady-state two-phase flow and heat transfer are not commercially available, what would be an appropriate rationally based modeling approach ? Would a correlation/flow-map based semi-analytical approach be the only possible recourse ? I would appreciate any guidance on tools, specific literature references, open-source algorithms, etc. I have conducted a preliminary literature survey, and have not found any CFD type two-phase flow modeling of the geometry and flow scales of my interest.

Thanks a million in advance for your help.

Dr. Eric Poindexter

Eric:

The only commercial code that I have any extensive experience with is PHOENICS. I am pretty sure that most of the available FVM codes will do this. But have a look at:

http://www.simuserve.com/cfd-shop/vol12-4.htm#Al

where an example like yours is considered. There is also a PHOENICS users group page where more specific postings should be directed.

Hello,

I have used Fluent in the past looking at the filling characteristics of thin section castings and StarCD for particle transfer through valves and also the cleanability of turbomachinery eg looking to see if one particular fluid (say milk) can be forced out using another fluid (say steam). It depends on how you two phase flow is configured. Are you trying to track particles through your system or does your two phase flow have a clearly identifiable interface.

Most commercial CFD codes can now do this using a Lagrangian method for particles or methods such as MAC or VOF for interface tracking which can i believe, both be studied in transient mode.

I would consider modelling of the two phase flow firstly neglecting heat transfer. If you model both together you will effectively double the amount of uncertainty in your results. I am not sure about the condensation part but I have seen the CFX's boiling model at work which may be able to help you.

One tip, the surface tension coeficient has a big effect at such a small scale and there are difficulties in defining what this value actually is. This also changes depending on what you clean your surface with and how rough the surface is.

Best of luck

Jas