Closed simulkade closed 4 months ago
The creation of a GitHub organization around the finite volume method and your finite-volume tools would be a very nice initiative. I would suggest calling it something like FiniteVolumeTransportPhenomena
. The notion of "finite volume" should be in the name since it is central to the numerical approach. I think FVM is the more "physical" numerical method compared to finite differences and finite elements, and nicely related to the conservation of mass and energy. It is also close to the idea of control volumes in open systems thermodynamics.
The GH organization could eventually also host repositories with course materials and specific research models using FVM and your FV tools. My colleague and I have been thinking about putting our (introductory) heat and mass transfer FVM course material on-line, but that would still require significant editing. Unfortunately time is scarce, so let's consider that a long-term project.
I feel that your FVM tools are great for understanding how the finite-volume method works; one can see how the matrix equations are constructed and how the FV solution evolves. They are a stepping stone to something more powerful like FiPy and provide insight how "full-featured" numerical software works on the inside. Also, your tools can be used to construct simple transport models that can be efficiently calculated, e.g., as part of some curve fitting procedure. There are cases where "full-featured" numerical software is overkill and where your FV Tools provide exactly the features needed.
Many open FV codes are geared specifically towards fluid dynamics. Advection-diffusion-reaction FV tools (PyFVTool and FiPy, for example) are simpler to use for those frequent cases where the flow velocity field is already known.
In short, great idea! As far as I am concerned, there is no urgency. I am an occasional user with limited time, but happy to contribute when I can.
Thanks, @mhvwerts. I couldn't have summarized it better. I was also thinking about having some teaching/research materials in the organization; more specifically, reactive transport in porous media. I'm also struggling with limited time but a couple of upcoming projects -hopefully- will provide me some hours to spend on this work. BTW, your suggested name is excellent. I keep this issue open and come back to it next week.
I have created the organization now, and PyFVTool is now moved to its new home. I don't have any experiences with managing github organizations, so I keep the issue open until we feel that everything is working fine.
Cool! How do I apply for membership? Best wishes for the new year!
I invited you @mhvwerts but the invitation expired. I just invited you again as co-owner.
Let's think about a nice logo... a well-chosen volume element drawn with pencil on paper, perhaps.
Let's think about a nice logo... a well-chosen volume element drawn with pencil on paper, perhaps.
"Any logo is better than no logo", but do not hesitate to change the current organization logo!
Look's classic, in a good way 👍 .
I consider the logo as a reminder to us for one day writing that book chapter or tutorial review that will explain in a concise, direct, simple yet precise manner how the finite volume method works in PyFVTool. There is some material that I keep in a safe place, but which still requires significant editing. Long term project!
I consider the logo as a reminder to us for one day writing that book chapter or tutorial review that will explain in a concise, direct, simple yet precise manner how the finite volume method works in PyFVTool. There is some material that I keep in a safe place, but which still requires significant editing. Long term project!
Fantastic! You can always count me in :-)
Create an organization for
PyFVTool
,FVTool
, andJFVM.jl
? None of these tools are full featured FVM tools, so I am inclined to a name such asAdvectionDiffusionReaction
,NumericalTransportPhenomena
or a shorter version of them (AdvDiffReac, NumTP, etc.). What do you think @mhvwerts ?