I'm a chemical engineer. I mainly work on CFD, with some brief experience in code development.
I think the idea to work on a free code is just great: we all need a free code with available sources in order to fully adapt it to our needs.
Now, some question:
1) Which is the actual development state of the code?
2) What are "next steps" in the code development?
3) What kind of help are you looking for?
Actually, the project is only just starting. So it is very easy for a developper to join the project since there is a lot to be done and all opinions are welcome concerning architecture.
So as far as the development of the code is concerned, at this point only mesh characteristics are implemented, ie reads an unstructured mesh, calculates all properties of the elements (volumes, surfaces, connectivities), etc...
Next steps will be the implementation of an architecture for equations and boundary conditions, as well as boundary conditions and post-processing architecture.
We are actually looking for developpers which have a good knowledge about C++ and hopefully software architecture.
But, someone with a numerical background (for matrix solvers, etc...) is also very welcome.
So please feel free to contact us if you need more information.
I've quite a good knowledge of C++, even if I never developed such a complex software.
I mostly worked on multiphase CFD calculations of chemical reactors, and I've a good mathematical background.
I'm a bit curious: what kind of architecture for equations and numerical methods would you like to implement? How do you think to manage boundary conditions: I mean, are you thinking to a FLUENT-like approach, or to more flexible B.C.?
P.S. I'm italian, so, in a future I could also translate if needed.
I'm really interested in your project, even if I don't have a lot of time for it due to my work. If you think I can be of some help, write me a message here, or e-mail me through sourceforge.net.
Well, as for the boundary conditions, we think of doing a really general implementation with a parser that would enable to specify boundary conditions values with formulas (possibly also with time dependance ). So only "one kind" of implementation for boundary condition specification (of course with all the possible BCs : inlet, outlet, walls, iso-flux,etc...) but very flexible.
As for the architecture, we plan to do a very simple but flexible and efficient thing considering equations and numerical methods. Since, we choose a co-located variable arrangement, we plan to derive only one version for the different convection schemes that would be available, and we plan to attach to a specified equation to be solved, the different numerical methods chosen : convection scheme, pressure solver if any.
As for time advancement, we plan to implement only a 2nd order semi-implicit adams-bashforth scheme for the moment : we will see in the future if there is the need to develop something else.
Parlo anche io italiano perche ho vissuto durante 4 anni a Milano in 1990-1994. Dunque, potremmo fare le traduzioni tutti i due !!! Ma penso che sara meglio farlo tu... Comunque, pensiamo prima di tutto fare un soft. veramente inglese. Poi se fra qualque tempo qualcuno ha voglia di fare una Graphical User-Interface, potremo vedere cosa si puo fare.
Anyway, since not everyone does speak italian, lets turn back to english !
See you soon for future possible developments.
An OpenFlower admin.
Thank you for all these information.
Mi fa piacere che parli italiano. Certamente prima pensiamo al software.
I completely agree with you on the boundary conditions and equations questions.
Just a note, why not a fully-implicit method for time stepping? It would implicitly grant stability of the solution.
I'm going to see the released code in these days (I'm on Easter holiday :-)).
Hi and thanks again :-)
Well, why not a fully-implicit method for time advancement ?
Because we had some experience about fully-implicit methods : they usually lead to high numerical diffusion and flow can not be fully turbulent (with a LES approach).
So we wanted to go for a semi-implicit method proposed by a research Lab from Stanford (CTR) that is energy preserving and adapted to co-located finite volume approaches.
An OpenFlower admin.
i want to do a simulation in which the gas excapes from the drums .These drums are placed in a container . how can i calc the conc of gas within the container at any particular region inside it.
Will this software help me do this
Well, the development of OpenFlower is really on its early stages !!! So there is no implementation of any equations yet and ... almost everything has to be done...
Therefore, what I can tell you is to come back in... say ... 6 months !
Sorry that we can't help you with your problem, but we will be pleased to provide you a full useable code in the near future.
Log in to post a comment.