I am a recent combustion masters graduate, now working in research and developement for a burner manufacturers. The company has aggreed to fund me to do a PhD to research into the problem of combustion resonance associated with multiport ceramic premixed burners.

I have met with many universities regarding this work to discuss possible PhD using a comercially available CFD package. I am also considering writing my own code for this specific application as i am concerned about not knowing the fudge facters involved in the commercial code(turbulance etc...), would this be a viable option in a three year PhD? What language is best for this type of code?, I know basic fortarn and pascal, i think these are a good foudation. Our company uses Pro E for design work, i think this is good for CFD geometries. Is that right?

The Universities I am considering are Cambridge, Leeds, Sheffield and UMIST. Can anyone tell me if any of these stand out above the rest in the world of CFD as this will aid my decision.

Any feed back at all will be greatly appreciated.

Thankyou

Ryan Brown

(1). Whether you will be able to get your PhD or not depends largely on your advisor's guidance and your personal academic contributions.(and other rules and regulations) So, your decision must depend on your discussions with the professors first. (2). The name of a school has nothing to do with your ability to get the PhD at all. (3). Combustion field is already difficult enough, I would suggest getting a less difficult topic, a more general subject, in case you need to change field later on. (putting yourself in a defensive position. writing your code is such an option. It would be a joke for a PhD candidate asking questions on how to use a commercial CFD code, especiall if they are from such well-known schools ......) (4). A PhD is not a new frozen pizza from a super-market. (just a joke)

My advice would be to look through the library and find people who have published what you consider good work in your area of interest. Once you have picked your place, you should have no problem getting a position (you're coming with your own money!). Remember that good reputation of the schhol is useful, but it is much more important to work with the good (best!) people.

Thanks for your reply. Do you mean that you suggest i wite my own code?

I have no choice over the topic, i am a research scientists for a burner manufacturer, and must develop CFD to model the stability of the flame.

I know the name of the school doesnt make it easier, i meant has any of them got a good reputation, thats all, more experience, facilities etc.

Thanks for the good advice..

(1). If the vendor of the commercial code you use decided not to support the code for some reasons, and if there was another young PhD who had written the code to solve the company's problem, what would be the consequence ? (2). It is essential that one must be able to write a code to solve his problem in order to claim that he is working in CFD field. (3). Using a commercial CFD code or codes is a different story. In your company's case, it is perhaps more realistic to find a CFD service company to work on the problem. (4). Writing a code is really not a big deal at all.

At the risk of sounding pessimistic (but not trying to be discouraging) I would say that it is practically impossible to do a PhD in only three years writing a CFD code for burner combustion from scratch. Give yourself 5 years of very very hard work - and don't expect to do it part-time.

You have to remember that even though your company will be paying for the studies, the bottom line (with honest universities) is that you have to contribute something new to the science of what you do. So, writing a CFD code will not give you a PhD but will put you at ground zero to do what you are actually expected to do (unless of course the topic is to develop a new methodology, and that usually means very simplified combustion, not the practical tool your company is expecting)

I think your company is expecting too much. Your first task is to lower _their_ expectations. Second task is to find a good professor (hopefully at a good university) and try to come to a reasonable agreement on what can be done for the duration of your stay. Note, more often than not people change direction during their PhD, so you can't expect a rock solid proposal of what you will work on.

On the technical note: It seems that you are interested in understading processes involved in flame instability. This to me implies studying physics - be it using analytic methods, experiments, or simplified or full-fledged CFD models. In this case, I believe that you really need to look at the details of a "simpler" problem of combustion in shear layers and/or jets, since this is where all the action is. Developing a full CFD code would have to be a secondary process where you'd need to add the bells and whistles of having a real geometry, radiation, soot formation, etc. This so-called simple problem will be useful to the company, will be PhD material, will require the least effort in learning the basics of CFD, and will hopefully lead to the development of nice simple models that you can later on implement in either a commercial code or one developed in-house.

Best of luck

Adrin Gharakhani

(1). In a university research environment, test setup and laboratory does not have to be re-invented by each PhD student. Same is true for the CFD student. (2). In a CFD research laboratory, there are always tons of codes on the system, created by the professors and students and from other sources. (3). That's why I said one must visit the professor and the lab, to see what's there. That is their expertie. (4). You don't go to a CFD laboratory to re-invent a CFD code,unless one is doing the new method development. (5). If one has to build a laboratory to do combustion instability experiment from nothing, it may take longer than 5 years. (6). From my point of view, there is nothing wrong with the company's goal. Combustion instability is a practical problem, it is a real problem. And companies are taking the parallel approach, both CFD and experimental approaches. And anytime a new combustor is designed, the stability of the new combustor is always a possible problem.

Good Afternoon Ryan,

Since your topic is pre-defined you may find it useful to go to your local university library and do a paper search for the last 5 years on your subject (try and limit to UK publications). Look at where the authors of relevant papers are based. You may find one university is doing a lot of work in your field. Then try and talk to researchers at that university.

Try to remember John's comments, while you are doing your work, the topic is likely to change at last a little. Your company may not realise this and they should understand this.

If the universities you have talked to think that you will be able to get a PhD using a commercial code, don't let other people tell you you can't. I used a commerical code for my PhD (and at a British University), although I don't have my viva for another 3 weeks!!

While commercial code vendors are unlikely to send you their code listings, they will be happy to give you all the information you want on the models you have used and the references that support them. They may also be able to help you find which UK universities are doing work in your field using their code. Several CFD codes have been developed in the UK and have UK offices (Fluent, Pheonics, CFX to name a few).

Wherever you end up, make sure there are several other researchers working in at least related fields (CFD and combustion in your case). Your sanity and blood pressure deserve it. I worked in too much isolation.

By the way, will you be doing a standard PhD or one of the new doctorates of engineering (EngD)?

All the best with your decisions and future work.

Althea

Dear Althea,

I will be doing a standard PhD, i dont know much about (EngD), whats the difference? Thankyou very much for your advice, im sure it will come in very useful.

Ryan

Good Morning Ryan,

I'm no expert on the EngD either, mine is a standard PhD.

The EngD is a relatively new initiative of an industrially based PhD. It involves collaboration and support from an industrial partner. I have read about it in the IMechE magazine, "Professional Engineering". You might find more information at their web site http://www.imeche.org.uk/

All the best

Althea

I would like get the orientation evolution equations of particles in a semi-dilute suspension. Brownian motion need not be considered. If any nearest nighbour coupling model is available I would like know more.

Regards Satheesh

What you are actually dealing with is a two phase flow of gas (or liquid) + particles. One way of solving the problem is to solve equations for each particles (you assume that you have a finite number of particles - 50,000 e.g.). The main force on the particles is the drag force and its exact form depends on the size, density and composition of the particles. This drag can also affect the fluid flow itself and might have to be taken into account when solving for the fluid flow (NS eqs). There might be additional forces such as gravity or interparticles forces, etc...

If your concern is the interaction between particles, then I cannot further help.

But if you need some more details on the two phase flow, I can give you some references.

Patrick

thank you for the response. I am dealing with flow of liquid plus particles. Flow may be simple shear flow. I want to get the orientation evolution equation of particles when particle-particle interaction is also considered. I have heard that some model with nearest nighbour coupling is available.

Satheesh