http://arxiv.org/abs/1711.09766
In some laboratory and most astrophysical situations plasma wakefield acceleration of electrons is one dimensional, i.e. variation transverse to the beam’s motion can be ignored. Thus, one dimensional (1D), particle-in-cell (PIC), fully electromagnetic simulations of electron plasma wake field acceleration are conducted in order to study the differences in electron plasma wake field acceleration in MeV versus GeV and linear versus blowout regimes. First, we show that caution needs to be taken when using fluid simulations, as PIC simulations prove that an approximation for an electron bunch not to evolve in time for few hundred plasma periods only applies when it is sufficiently relativistic. Our 1D PIC simulations establish that injecting driving and trailing electron bunches into plasmas with $n_0=5 \times 10^{22}$ m$^{-3}$, results in electric fields of about $- 10^{10}$ V/m, if the bunch density is one third as that of plasma (linear regime), and about $- 10^{11}$ V/m if the driving bunch density is $2.5 n_0$ (blowout regime). We study the differences in the plasma wake created and what is an optimal position of the trailing electron bunch. Starting from initial 36 MeV trailing bunch with $n_b=0.3 n_0$ its acceleration to 85 MeV is readily possible within 200 plasma periods. Beyond this time the approximation for a driving electron bunch not to evolve in time becomes invalid. Starting from initial 20 GeV trailing bunch with $n_b=0.3 n_0$ its acceleration to 21 GeV occurs within 2000 plasma periods. When driving bunch density is increased to $n_b=2.5 n_0$, starting from initial 20 GeV trailing bunch with $n_b=n_0$ its acceleration to 24 GeV occurs within 2000 plasma periods and plasma wake size is much larger. In this case, optimally there should be approximately $(90-100) c/\omega_{pe}$ distance between trailing and driving electron bunches.
D. Tsiklauri
Tue, 28 Nov 17
32/82
Comments: submitted for publication, in peer review