Hi,

I am currently studying for a masters degree by research. I am basically using Fluents CFD software to model the flow thru' a pump. While interesting, it is not what I would like to continue to PhD. I am more interested in understanding the software itself. I would like to write CFD/Numerical analysis software. For this reason I hope to more into that field as soon as I finish my masters. I would like to work towards a PhD in this area. I have a degree in Mechanical Engineering, with FLuids as my major. I have however, very little experience of hands-on programming. I studied CFD thru' Verstegs book while at college, but that is it. Has anyone any ideas on whether it would be possible to make this move? What level of Programming and mathematical numerical method knowledge would be required before starting on such a PhD?? Thank you all, Jack.

(1). I think, in your case, you already have some idea about the cfd code as a product. I mean a commercial product. This definitely will help you later on. (2). I think the most important part is the desire to do something. Fortran and BASIC are very simple and easy to learn languages. Very few engineers actually took the courses in Fortran or BASIC. The same is true for the c language. It is very similar to Fortran. It shouldn't be a problem at all. (3). The newer generation of C++ oop language is somewhat difficult at the begining, but currently it is not widely used in CFD. And in C++ , you don't have to write everything in oop style as JAVA(which is oop 100%), you still can write in the old structure, procedural style codes. (4). But in the numerical analysis side, you definitely need to study the basic numerical analysis, the partial differential equations , and these are covered in graduate school. There are also books available with a lot of sample of Fortran codes in the numerical analysis. So, you will learn the programming by following the samples in the book. (5). I would say, it is not very difficult to study cfd and programming in cfd. (6). But if you are interested in inventing new solution algorithms, or new turbulence models, then it is a different story. Those things are hard to invent. But in cfd, there are many different application fields available, 3-D geometry and mesh generation are still the biggest problem waiting to be solved. Not to mention the specific technical problems waiting to be solved. (7). The hard part is, non of these are push button yet. cfd requires a lot interaction with the computer system, keybord and mouse. (8). The basic fluid and thermodynamics problems remain the same, but only the tools have been changed.

Thank you for your input John. I have studied partial differential equations and some numerical analysis. I expect this is not likely to be too difficult to extend. Also, I have a basic understanding of C, as I covered it in college for one semester. I am currently reading through my C programming book and my calculus books, so I hope to be ready to start working on simple, introductory problems soon. I think I would be most interested in working on new solution algorithms, as I imagine this would be extremly challenging. Do you ( And anyone else!) know what universities in Europe and The US would be the best places to enquire? I am from Ireland, so it would be nice to see some other parts of the world! Thankds again. Jack.

(1). Surprise! Your not only have the desire to learn but also are well prepared for it. (2). Start looking around from this forum, I mean the resources sections. (3). If the school has a cfd program, it is likely that you can find it there, sound reasonable? Let the finger do the walking (I mean using the mouse) (4). From there, you can use the e-mail and start your search . And you can always come back here to ask specific questions.

Hi, Jack,

Look at

http://www.ses.soton.ac.uk/PG_Admiss...olarships.html

and then follow the link to

http://www.afm.ses.soton.ac.uk

Not too far from Ireland but who knows?

Rgds, Sergei