Cyclotron velocity scale: negative velocity and time units for a negatively charged bulk species #825

Units.derive_units (src/struphy/physics/physics.py) sets the velocity unit for velocity_scale="cyclotron" with the signed cyclotron frequency of the bulk species:

self._v = Z_bulk * con.e * self.B / (A_bulk * con.mH) * self.x
...
self._t = self.x / self.v

For a negatively charged bulk species (e.g. electrons, Z_bulk = -1) the velocity unit and the time unit t = x / v are therefore negative. Everything derived from them inherits the sign: the current-density unit j = e n v, the time grid in physical units (Output.time_scale), and the equation parameter epsilon = 1 / (Omega_c * t) is +1 for both signs of the charge (the sign of the force then sits in the units instead of in epsilon).

Models with velocity_scale="cyclotron" currently use positive bulk species, so this does not show up in the examples; electron runs set the charge sign through epsilon=-1 with the default charge number.

Possible fix: use abs(Z_bulk) in the velocity unit, so that v and t are always positive and epsilon = 1 / (Omega_c t) carries the sign of the charge (as for the other velocity scales). This changes the normalization of cyclotron-scale models with negative bulk charge, so it needs a check of those models and of the docs (doc/sections/subsections/models-normalization.rst).

Found while setting up the cyclotron normalization of VlasovMaxwellQuasiNeutral (#821).

🤖 Generated with Claude Code