Draft: Include thermal motion of gas

Problem: The convention when calculating collision frequencies and the velocity after an elastic collision is to assume that the relative velocity between electron and gas molecule is the velocity of the electron. For low E/N the velocity of electrons is so low that the velocity of the gas molecules can no longer be ignored.

Goals:

• Sample velocity of gas molecules following a Maxwell-Boltzmann distribution
• Take into account the velocity of the gas molecules when calculating the velocity of an electron after elastic collision
• Take into account the velocity of the gas molecules when calculating collision frequencies

Notes:

• Since this only becomes an issue at "low" E/N how should we switch between methods ?

Notable reference is Yousfi 1994: http://dx.doi.org/10.1103/PhysRevE.49.3264

An example of the difference that including gas temperature makes at low E/N can be seen in the reduced mobility (of CH4) calculated by BOLSIG+ (with and without taking gas T into account) and particle_swarm (not taking gas T into account)