CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Computational modelling of Sprays (

Abdul Sattar April 9, 2000 20:48

Computational modelling of Sprays

I am a final year undergraduate student studying Aerospace Engineering at UMIST (England U.K) and as a partial fulfilment of my degree I am carrying out research in Computational Modelling of Sprays. I have found out there two distinct methods of modelling gas-liquid two-phase flows of the type where there exist discrete liquid droplets in a carrier gas that have been developed. Both involve the coupling of separate equations for the two phases. The first is the Eulerian-Eulerian method and the second is the Eulerian-Lagrangian scheme. I would be very grateful if you could provide me with an insight into both these methods as I have not found much literature into this subject, or recommend me to any information that could be helpful. I would greatly appreciate any help.

r.sureshkumar April 10, 2000 01:43

Re: Computational modelling of Sprays
Hi, There are many work's available on spray modeling; provided you look into the right journal's. For ex: Int J. Multiphase Flows, each issue almost has one work atleast, on Eulerian and Lagrangian model.

Visit Imperial College site, (Prof. Gosman and Colleagues), you will find many reported work's (provided online).

For an understanding of both the methods, ie continuous droplet mmodeling (CDM - Eul.-Eul) and discrete droplet modeling (DDM -Eul.-Lag.), I suggest you go thro' W.A.SIRIGNANO, "THE FORMULATION OF SPRAY COMBUSTION MODELS: RESOLUTION COMPARED TO DROPLET SPACING", J. HEAT TRANSFER,V108,1986.

For DDM - Plenty of works are available ex:search for work's by C.T.CROWE, A.D. Gosman. For CDM - Search during 60's and 70's, for work's by, F.A.Williams and Lockwood.

If you can find any work giving comprehensively (See the book by 'Kuo' on combustion), all the methods and published very recently, you can inform me/the-forum!.

If you have problems still, I believe there are plenty of people in this forum, to make you think (laterally or thro' provocation or by completely ignoring your cry for help!), and therby can help you, to help yourself.

All the best and welcome to "MULTIPHASE RESEARCH".


Jin Wook LEE April 11, 2000 01:10

Re: Computational modelling of Sprays
E-L method can be applicable and can give you good result when volume fraction of discrete phase(liquid, in your case) is fairly low, for example below 0.05(maximum value might be about 0.1).

In addition, within my experience, when mass loading(discrete-mass/gas-mass) exceeds 1.0, there was severe problem of numerical convergency. So, I think that maximum allowable mass loading of discerte phase is about 1.0 or slightly higher.

For high mass loading or high volume fraction of liquid, E-E should be used.

In general, for spray modleing, volume fraction of liquid is less than 0.01 in gas-liquid two phase flow. I am sure that your case is belong to this range(below 0.01). For air-water spray problem at room temperature. mass loading of 1.0 means volume fraction of about 0.0013.

So I recommend E-L method for your spray analysis. But, do not skip to check mass loading and/or volume fraction of liquid before your simulation.

Sincerely, Jinwook LEE

r.sureshkumar April 11, 2000 05:26

Re: Computational modelling of Sprays
Sattar, you have asked where to locate the Journal, i am sorry, i don't know your library. However, there is a recent (1991!) paper by Crowe i don't have it!. I think that should also do. As far as F.A.Williams papers, It's available in the Journal, "Combustion and Flames", 1959 to 62 series. One paper in each issue was the frequency. Good that you have visited the Imperial College site, however, don't get stuck up with net alone. For, even if you locate the journal sites it's difficult to get the back issues. Most of them have issues from 90's some even worse from 95' onwards. Library helps a lot, so have a second thought. All the best. I hope you have a good one.

LEE i agree with you on the convergence problem, this is especially true when we consider momentum effects of droplets on gas phase. However, underrelaxation of the source terms helps a lot (to a good extent!). Again as far as convergence, i have put up a clarification earlier. Hope you can help.

i.e. When, we normalize the residuals by inlet flux, should'nt we also add the mass/etc flux due to sprays also. B'cause let's say for mass loading fraction of 1.0 (discrete/continuous), then the criteria we impose (ex. 0.1% or so) is actually 50% more strict, (i.e. 0.05%) b'cause of addition in inlet mass/etc due to droplets.

Thanks for the helps.

All times are GMT -4. The time now is 00:55.