Skip to content

Electrical engineering open-source software providing a user-friendly, unified, flexible simulation framework for the multiphysic design and optimization of electrical machines and drives

License

Notifications You must be signed in to change notification settings

Rutherfordio/pyleecan

 
 

Repository files navigation

Pyleecan

PyPI version License Code style: black

Presentation

PYLEECAN objective is to provide a user-friendly, unified, flexible simulation framework for the multiphysic design and optimization of electrical machines and drives.

It is meant to be used by researchers, R&D engineers and teachers in electrical engineering, both on standard topologies of electrical machines and on novel topologies (e.g. during a PhD work). Pyleecan is open source with a very permissive license (Apache V2 license).

The main objective of PYLEECAN is to boost reproduicible research and open-science in electrical engineering. For example, every PhD student should start with PYLEECAN. Instead of implementing your own scripts on your side (e.g. coupling Scilab or Matlab with Femm), you could benefit from the community support to help you in your research.

Getting Started

The procedure to install and use Pyleecan is detailed on pyleecan website

Origin and Current Status of the Project

EOMYS ENGINEERING initiated this open-source project in 2018 for the study of electric motors. The project is now backed by Green Forge Coop non profit organization, who also supports the development of Mosqito for sound quality and SciDataTool for efficient scientific data exploitation.

Main Models and Couplings:

  • PYLEECAN is fully coupled to FEMM to carry non-linear magnetostatic analysis including sliding band and symmetries. For now this coupling is available (only on Windows OS).
  • PYLEECAN includes an iron losses model (based on FEMM coupling output).
  • PYLEECAN includes an electrical model to solve the equivalent circuit of PMSM and SCIM machines.
  • PYLEECAN is coupled to GMSH 2D/3D finite element mesh generator to run third-party multiphysic solvers.
  • PYLEECAN is coupled to a multiobjective optimization library to carry design optimization of electrical machines.
  • PYLEECAN enables to define Parameter Sweep of variable speed simulations.

Main Topologies Features:

  • PYLEECAN includes a Graphical User Interface to define main 2D radial flux topologies parametrized geometries (SPMSM, IPMSM, SCIM, DFIM, WRSM, SRM, SynRM) including material library.
  • Possibility to import Slot or Hole from DXF files
  • User Defined Winding or automatic algorithm
  • Generic Geometry modeler to draw complex machines in the software coupled with PYLEECAN
  • Notches (Yoke and Bore) / Uneven Bore shape (Lamination without slot only) / Machine with more than 2 laminations

If you are interested by a topology or a specific model, you can open an issue or a discussion on this Github repository to talk about it. We will gladly explain how to develop it yourself or we will add it to the development list. We are always looking for experimental data and model validation based on the last scientific research work. Even if you don't have time to work on pyleecan, sharing your expertise will be valued by the community.

RoadMap

The mid/long term roadmap of the project is detailed here

Documentation / Website

All the information on the project are available at www.pyleecan.org. In particular, the media page gathers the publications, video and screenshots of the project.

Contact

You can contact us:

  • By opening an issue on Github (to request a feature, ask a question or report a bug) or starting a discussion
  • By sending an email at pyleecan(at)framalistes.org that redirect to all the maintainers.

You can follow us:

About

Electrical engineering open-source software providing a user-friendly, unified, flexible simulation framework for the multiphysic design and optimization of electrical machines and drives

Resources

License

Code of conduct

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Jupyter Notebook 80.5%
  • Python 19.5%