I am starting a phD in CFD field. Can I join this project?
My major skills are:
Computer graphics - OpenGL,
C/C++, Fortran, Visual Basic.
My current and future interests:
Finite Volume Method (FVM),
Interface tracking - Volume of Fluid (VOF),
Just tell us which topic you are most interested in (numerics, post-processing... topics which are discussed in depth in a few messages in this forum) and we will add you to the developpers team if you think you can fit on any of these.
We have to tell you that the project is at its early stages : therefore it is a real opportunity to start with a flexible tool and to quickly be an expert (!) but also note that it is not fully functional !
See you soon,
I would like to work mainly on the solver: implementation of the SIMPLE algorithm to unstructured meshes.
I can contribute also to user interface and integration with meshers.
I will download your files and see what has been done.
Your interest in the solver part is greatly appreciated !
We hope that you will find the first developments we made at your convenience and that you will be able to easily go through what has been done.
I have downloaded the beta version and seen that it uses gmsh to build meshes. What do you think about using GiD? I think that it is easier to build more complicated meshes since you can import complex CAD surfaces. It was used in project ADFC.
I compiled the code using Visual C++ and saw that it mainly reads the mesh and calculates geometric properties such as volumes and areas which are needed for FVM.
GiD can be found at:
I tried ADFC not long ago, and downloaded academic version of the commercial software "GiD" that was need to run it. The academic version is very limited - it is not even possible to follow through the ADFC demo without reaching its limits.
I posted to ADFC discussion about the possibility of using an open source mesher (eg. gmsh) instead, and received a personal reply from someone offering me a crack (password generator) for GiD, presumably to remove the academic limitation.
I was very disappointed. So I think it is best not to uphold ADFC as an example for anything.
If GiD really does a better job then any open source alternatives, then how about we write a better mesher once openFlower has matured a little.
I agree with you Free Coder. I think we can find the human resources here at sourceforge to build a pre and post processor for openflower. I have some experience in that myself. What troubles me is heart of CFD which is the solver. There I think we need to join efforts to build a really good solver. Also I would like to exchange references of good papers and books we could study on this subject. I have been studying some already for my phD.
A good solver, with a rich documentation, can be found at www.mfix.org
MFIX is a fully functional CFD Fortran code applied to granular flows. It is not open source, but the code and the documentation can be freely downloaded and used.
I think it can be considered as a rich source of information for the numerical part of OpenFlower.
I have just took a look at MFIX website and at their numerical guide, and it appears to me that solve their problems in cartesian based grids. Therefore, we can not use their numerical techniques for the evaluation of the diffusive and convective fluxes in our unstructured approach.
However, I have not seen any information concerning Poisson solver : continuying to search !
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.