Right,

Hello out there, I wish to model sprays for various kinds of liquids (from perfume to fuel) at pressure ranging from about 2 bar to around 100 bar. After having found out that a commercial code I was using simply cannot produce sprays at these pressures, I turned to KIVA3V R2, being told by various people and some litterature that it was one of the best codes available, albeit based on oldish methods. Then someone came along and told me, to cut a long story short, that kiva was not that good and I should look for another code.

Ok so now I have two different viewpoints : very good and not very good.

My question to you, wise people of the ether, is : 1) how would you rate K3 R2 for those who know it. 2) does anyone know of any code, preferably VERY low cost or free, that does the same or better, preferably with GUI and open source code.

Pretty tall order, and rather rudimentary questions, I know, but then is there anything else .

Anyway, in the hope that anyone can supply some sort of answer, I shall be forever more eternally grateful.

Thanks a lot.

Spray and spray combustion, especially these occured in the internal combustion engines, is a very complicated phenonema (just think of options when spray hitting the wall). Therefore I am not sure you are able to find a low cost or free software for that. As far as I am aware (with my very limited knowledge about KIVA), KIVA can only be used for research purpose because of its limitation in dealing with geometry - if you have a geometry which can be describled by block-structured mesh, KIVA will be a very good choice since it has been used very extensively in research community. Among three major commercial CFD codes listed in the Forum, only STAR-CD has this capability and it's widely used in the automobile industry - a good source would be SAE papers. Another code with engine simulation capability is FIRE from AVL.

Thanks for the response. My geometry is bery basic : a cylinder. So there is no problem in using the rough interface (i.e.: none) kiva gives me. As for the other codes you mention, STAR is the one I was refering to : its spray capabilities only cater for automibile-industry pressure ranges. And that, by far, goes beyond our pressure ranges. So, no go. I hear that kiva is also used extensively in industry too. As for FIRE, I have not heard about it one way or the other.

First, thanks for using STAR-CD. Current spray models in STAR-CD (evaporation, automisation etc) are empirical correlations based on the experimental data catering for the current needs of the automibile industry. For the pressure range you mentioned (2~100bar), I believe this is still the focus of extensive research. A possible source for these models are FEV consulting company based at Aachen and University of Wisconsim in USA. Imperial College is also actively involved in these research through collaborative project. However STAR-CD is designed in such a way that many new models can be easily plugged into it through its user coding. Therefore if you have any model in mind, we are more than happy to advise you how to link it with STAR-CD.

I have used FIRE for fuel sprays and for spray painting. It worked quite well in both cases. In the spray painting case I added a subroutine to the code for columb forces as the spray droplets approach a slightly charged wall (just for curiousity); adding additional physics like this is possible with many of the commercial codes (via "user defined functions").

I have yet to see any code that can predict the droplet size distribution that would be created by atomizing a liquid. To my knowledge, this is beyond the current state of the art. Thus all the available codes require that you know the initial size and velocity distribution of the spray at its source.

Free spray code?? Well that would only be a research code. For that I would expect no validation effort; if it did have significant validation work, then I would expect you would have to pay for the result.