I got the exact same question a few years ago... And I still have no clear answer...

You have some papers comparing some models in different conditions with RANS approach but also LES, such as those:

http://www.sciencedirect.com/science/article/pii/S0016236108004304

http://papers.sae.org/2008-01-0962/

http://papers.sae.org/2009-24-0024/

And there is certainly more of them especially treatin the TAB model...

You can find a lot of paper having good predictions using all of these models. When I got confronted to the choise of the atomization model (KH-RT, Reitz & Diwakar , Huh, Hsiang-Faeth, Pilch & Erdman) in the first steps of my PhD, I first looked at the regimes taken into account by. For exemple, if you refer to the original paper of the Reitz & Diwakar model, it takes into account bag breakup and stripping breakup. You can look at this for every atomization model and rule out some according to your application.

The problem is you can find papers using a model in conditions different from there original application ( for exemple, a very high pressure application with the Reitz & Diwakar model) and fairly good predictions are obtained by authors... So it is tricky to eliminate them through the way presented in the first paragraph.

An becareful of the last problem of those models: contants tuning which, in my opinion, is the really tricky part.

To clearly answer you question: I think that the amount of papers using all of the atomization models can show that all of them can lead to fairly good results, so there is no better model. What I would do (and actually did in my PhD), is take two or three models, which are the most used in the litterature, and test them to see which would be the best in your application.

I would go for KH-RT.

please also read this paper: A breakup model for transient Diesel fuel sprays by Turner et al 2012 (Sazhin's group)

Majority of research have been done using KH-RT model. So as Ahmed said, I would also prefer for the same.

Hello,

there is no clear best model for break up. It depends on the compromize in cpu time and detailed results. You can find some discussion on the issue at :

L Le Moyne (2010) Trends in atomization theory International Journal of Spray and Combustion Dynamics 2: 1. 49-84.

F Dos Santos, L Le Moyne (2011) Spray Atomization Models in Engine Applications, from Correlations to Direct Numerical Simulations Oil & Gas Science and Technology , Vol. 66 (2011), No. 5, pp. 801-822.

hope it helps, good research

In my view it is KH-RT model