Re: new open-source electromagnetic field solver available (OpenWME)

[Jon Evans]

Hello Steffen,

Thanks for sharing this!

Fabien Corona started working on KiCad integration with FEM solvers but the work is still in early stages.

Perhaps there is some overlap in possible use cases.

It is not immediately clear to me: is your library meant to work from "real" geometry (i.e. a mesh generated from copper on a board) or on ideal geometry (i.e. a parametric definition of a microstrip)?

In other words, would the integration with KiCad happen in the PCB editor, to show a simulation of what the user has actually created, or in the PCB calculator, to provide calculations on "ideal microstrips" of a given geometry, and things like that?

In general, if we want to approach integrating this library into KiCad, there are two areas that would need to be taken care of:

1) How the actual features work would need to be discussed and agreed upon.  The best way here is to create a document (or GitLab issue) describing the use cases in as much detail as possible, for example with sketches of what a potential UI would look like, or other descriptions of how the features would be exposed to users.

2) For any new library, we need to decide whether to bring it in-tree or to use it as an external dependency.  If we do the latter, it needs to be packaged on all our supported platforms and distributions.  If we do the former, then it just has to compile under our supported toolchains on Linux, macOS, and Windows.  Your library itself does not appear to have any build system, so I guess right now it's designed to be included in-tree with a final application.  In this case, someone will have to develop a suitable CMake build configuration for the library so that it can be integrated with the rest of the KiCad build.

Best,

Jon

On Tue, Oct 31, 2023 at 8:23 AM Steffen Kuehn <steffe...@quantino-theory.de> wrote:

Dear KiCad Developer,

Recently I released a new open source EM solver:

https://github.com/StKuehn/OpenWME

If this would be integrated into KiCad, then one could calculate the
radiation characteristics of a PCB (EMC) or the self-inductance of wires
and microstrips. If there is interest, please let me know. I could help
to integrate it.

I am a scientist and have no commercial interests. I am working on a
simplification and modernization of classical electrodynamics. OpenWME
uses the newly discovered general symbolic solution of Maxwell's
equations for point charges. This eliminates the need to solve Maxwell's
equations numerically, which makes the solver extremely fast.

Best regards,
Steffen Kühn

[Seth Hillbrand]

Hello Steffen-

Thank you for your interest in KiCad.

We are certainly always interested in improving the speed and accuracy of simulations in KiCad.

That being said, KiCad is not the appropriate forum to test out new theories of electromagnetism.  We are not subject matter experts on where appropriate overlap between Maxwell and Weber's theories lie.  New theories will need to be appropriately vetted in the scientific literature before we look at using them to model physical effects in KiCad.

Best,

Seth

Seth Hillbrand

Lead Developer

+1-530-302-5483‬

Long Beach, CA

www.kipro-pcb.com    in...@kipro-pcb.com

--
You received this message because you are subscribed to the Google Groups "KiCad Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.
To view this discussion on the web visit https://groups.google.com/a/kicad.org/d/msgid/devlist/CA%2BqGbCAkhL28cx-%3DKaaJBakw6dZA03sh5%3DaGDd2NXsp5cDb88A%40mail.gmail.com.