| Author |
Message |
    
4maitland
Member
Username: 4maitland
Post Number: 7
Registered: 05-2010
| | Posted on Wednesday, May 19, 2010 - 07:43 pm: | |
How does one go about doing a lengthy project when it is required and how does one go about using pdeflex to model that problem? i.e. It is easy to borrow from another's work and modify it for your own special needs, that is a gimmie. But, if you want to analyze something, say in solid state physics and you do not know the governing equations to use, what is a first step? I know this sounds like an elementary question, but it isn't meant to be. It is however, an honest question because when I work on something, I want it to be MY work not someone else's.
Jon |
4maitland
Member
Username: 4maitland
Post Number: 8
Registered: 05-2010
| | Posted on Wednesday, May 19, 2010 - 08:12 pm: | |
I do not work for a research company, nor am I a grad student (not yet anyhow). I am trying to get my feet wet using the software, so that in the near future I may excel as an FEM user. I have only had one semester of C programming, I'm sure that is enough to become a proficient and useful programmer.
By the way, MNelson, thank you for your insight and help on my sloppy program. I still have a question though, why the person who wrote it decided to use the "Q" in the div(grad(Q)/eps)=0?
I used, div(grad(D))=0 which basically the same is it not? i also observed that in the fieldmap.pde, Q though it is declared a variable, it is not initialized nor was V and yet throughout region 2,
natural(Q) == tangential(grad(Q))provides the field lines asked for. But when I substitutted into the same expression, using the "D", I had had something different than expected. So what is the Q/eps? Does this represent the gaussian surface of a pill box (B.C.) close to the interface of the figure and the medium? I know Etangential is continuous, so I guess I answered my own question, huh? Boy do I feel dumb!
Jon |
rgnelson
Moderator
Username: rgnelson
Post Number: 1367
Registered: 06-2003
| | Posted on Saturday, May 22, 2010 - 09:37 pm: | |
FlexPDE assumes that you know the PDE system that you want to solve. It will not tell you what is the governing system for some arbitrary application.
If you don't know what the governing system is, then you have to go to textbooks and journal articles to find out what mathematical analysis suits the needs of your model and present that system to FlexPDE.
|
|