ergodicio / adept

Automatic-Differentiation-Enabled Plasma Transport in JAX
MIT License
27 stars 3 forks source link

Vlasov1D update - Ampere solver / Wave solver / spline edfdv #27

Closed joglekara closed 8 months ago

joglekara commented 8 months ago

This PR was an attempt to get a working Maxwell solver for Vlasov2D2V but did not quite make it there. To develop this, an Ampere solver was needed. It was added as a 1D Vlasov pusher and it works fine there but is unstable in the 2D2V case.

To see if the problem was due to the velocity push, a cubic spline interpolator was added and tested in 1D. Adding it in 2D remains to be done.

There are also some small changes to the units file