Reported by Adrian Oeftiger

Q by Stefan Sorge (?): if delta F contribution is large => some weight
factors w_i grow strongly for particles that go from large f0 regions
to small f0 regions, therefore induces non-physical plasma heating --
one needs to find a way around this!
A: you lose conservation properties in delta F codes that one usually
has in PIC codes, however, Sonnendruecker et al. found practical
solutions to the unphysical plasma heating issue, paper soon to be
published.

Q by Ji Qiang: important to know that delta F is to be handled as a
perturbation such that (f0 - delta f) / f0 is very small and f0 is the
general solution to the Vlasov equation. However, in general in (high
intensity) beam physics, f0 is not well known!
A: I agree, for instability codes delta F codes are not suitable
because you don't know your f0. However, for this particular case when
analysing halo or beam emittance growth studies and we know the
matched [unperturbed] solution already, then it is a fine solution.

Q by Frederik Kesting (?): Are collisions in 2.5D and 2D the same as
in 3D? How is this justified?
A: Particles have a finite size and we talk about interactions on a
grid. Particles are very long in the longitudinal direction and very
tiny in the transverse plane; furthermore synchroton motion is very
slow, a thousand times slower than the transverse motion such that
collisions mainly happen in the transverse and we assume this for our
model. Of course this does not hold true anymore for grids that are
the same in longitudinal as in transverse.

Q by Elena: What happens if M is equal to the number of macroparticles
N in your model, there is diffusion?
A: Then you have real Intra Beam Scattering.