hi,

Could anyone tell me as to which of the universities in the US are really into CFD work.I need to make a decision on the choice of the university for graduate studies by september.All suggestions are invited.

thanking you Thampy.

Hi there,

I can give you a couple of names and places of CFD people using Spectral Methods. These are mostly people from Applied Mathematics. Mainly:

Brown University, the group of David Gottlieb

Princeton University, the Group of Steven A. Orszag, Karniadakis, etc..

I know that CalTech has also some excelent people, in all fields of CFD (not especially using spectral methods).

Take a few Journals like J. of Fluid Mechanics, J. of Computational PHysics, etc.. and check who are the big guys there (in the articles as well as in the Editorial staff, etc..).

I hope this helps, PG.

(1). Since you are able to get on Internet, you should be able to visit the home pages of ME or AE departments. (2). From there, you can find out more information about their specialized area of CFD or even e-mail to the department head or professors. (3). CFD covers a very wide range of fields and problems, so you need to do some research to find out the area you are interested in first. This also depends on your background. (4). Actually, if you have access to a library, you can read some CFD related journals and I think, you should be able to identify the school and authors easily. (5). So, this is one thing which you have to do some thinking.

I would recommend you to look at the list of top 50 schools in the country in engineering and see which area you would like to work on.

Karniadakis has long left Princeton. He's at Brown now.

Vitually all good (and even not so good) universities offer CFD programs these days.

As was suggested by others, it is important for you to decide what you want to do in the field of CFD. Do you want to get involved in incompressible or compressible flow? Combustion or purely fluids or fluid-structure interaction, newtonian or non-newtonian fluids, algorithm development or application to phyiscal problems, turbulence or LES modeling or DNS studies, etc.?

An additional factor you have to consider is whether you will get financial assisstance during your studies. As a student you cannot afford to worry about housing/food and simultaneously do good research. Some universities will accept a student only if there is funding available, others will take them without the availability of funding. So this, as a matter of practicality, is just as important as (if not more) the CFD field you wish to focus on.

As a rule, competition is extremely tough with excellent universities and the chances of getting guaranteed funding could be low. Since each application form you send out will cost you money, you need to focus on a maximum of 5 universities that you think you will have a good chance of entering.

To do so, you need to talk to the particular professor to get a feel for availabilities. If (s)he doesn't have money for this year, you're wasting your time/money sending an application to that university.

Adrin Gharakhani

hi all,

First of all I would like to thank all of for having responded to my query here.I'm looking for universities which are into using distributed and parallel computing to solve cfd related problems.In fact I'm doing my undergrad project in that area.I ve heard that the ERC under Mississippi state University has good facilities for cfd work.kindly comment.

Thanking you all Thampy.

If what you're looking for is parallel CFD then you should orient your search in the direction of these institutions who do have Massivelly parallel machines. There are two that I know (and many that I don't):

again CalTech: they have Cray YMP2E/232 and a HP/Convex SPP2000, both of them with 512 processors I think. Both computers are in fact CalTech and JPL computers (California Institute of Technology and the Jet Propulsion Laboratory). I know of people who have been writing hydro code in parallel there and there are few good groups, ranging from Planetary Atmospheres modeling, Oceans to Engineering.

And there is the National Center for Supercomputing Applications on the Campus of the University of Illinois, Urbana-Champaign.

Many other institutions do have access to Supercomputers even if these are in National Labs.

And by the way, CalTech (and Southern California indeed) is a nice place to be...

(1). I have just visited the site briefly. (2). I think it is a nice place to get started. I also recognized an old colleague of mine, who is now the director of the CFD program there. And it looks like that they are also working on a wide range of projects. (3). California, too expensive (I have been living in orange county California for many years.), Illinois, too cold, so Mississippi just right. It is about the hurricane season in these southern states now (June through October). Bring your umbrella when you decide to come.

Hi,

I agree with the structured analysis based on the type of problems that one could solve (incomp, comp., subsonic, transonic, combustion, etc.) BUT how does one, who has never done any work or research in the field make this type of decision apriori???? (this is assuming the typical student who has just finished a BSc.)

Another criterion which you could use which is a bit closer to the types of things that undergrads ar ABLE to evaluate is to base it on THE TYPES OF METHODS AND PROBLEM SOLVING TECHNIQUES THAT THEY ENJOY!

For example, if one is really pragmatic enginner, has strong physical intuition and not overly mathematical then they should look for people and groups, papers and texts like: Patankar, Spalding, Peric', International Journal of Numerical Heat Transfer, etc. They should stay away from JFM, the math, applied math, and CS departments. In this analsis, the specific problem to work on is almost totally irrelevant! The people in this group will give you a practial problem and problem solving approach!

Good Luck.............................................D uane

Mississippi (sp?) state is actually a good school for CFD - they usually have good amount of funding available. (it would be one of the schools at the top of my list _for_ CFD)

I agree with John Chien, california is too expensive to be a student in (again, unless you have funding guaranteed for you ahead of time). Caltech does not take students unless they have funding. It's a small school and they usually have 2-3 openings a year for CFD in the aero/astro dept. (if that). So you will have very high competition to deal with.

As for parallel CFD, years ago when parallel programing was becoming popular, I would say that it would have been a good idea to do research in the field. I would, however, advise you (if I may) not to focus on this topic for a PhD program. Nowadays, you would be "expected" to know parallel programming like you would fortran and C.

As for computing facilities, the days when universities would have to have their own computing power have long gone. Thanks to the internet and supercomputing at national labs what you need to worry about is to get funding for the project, once you have that you can pretty much guarantee access to national lab facilities.

Adrin Gharakhani

Lot of schools are now-a-days working on parallel computation in CFD. First as a MS student, you need to more or less make a judgement about yourself. Say for example, whether you are a very good student, outstanding etc. Places like JPL at Caltech, MIT, UCLA, Stanford etc. are hard to get in because of their stringent requirements and competition.