Selected questions and discussion:
AAn (on p.5 ): Is the field we see behind the bunch really generated by wall roughness only?
BB: Yes!
RI: Where does the regularity come from?
BB: Don't know. Perhaps only the main modes above some cutoff are shown (MP).
FLP (on p.6-8): Are these formulae also valid for ellipses/rectangles?
BB: No, they are only for a circular beampipe, which is the worst case (highest losses).
MP (on p.7 ): What's the difference between capital and small J?
BB: The small j are the roots of the Bessel function J.
FLP: How do these formulae compare with the ones given by Karl Bane and Roger Jones?
BB: I only did a rough comparison, the formulae I use are valid in a larger frequency domain.
FLP (on p.9) "Ar" should be Ag (silver).
FLP (on p.10/14): The effects from the two contributions do not seem to add up?
BB: They are not additive, but have to be combined using the formulae on p.6
MP: The effect on the energy spread is probably negligible (micro-bunching).
RI (on p.15): In the top graph at 1 mm, the wake function is almost the same, why then is the energy spread so large?
BB: I am not quite sure about the first two points, they need to be checked!
AO: Are these wake effects included in the ELEGANT code?
BB: I do not know.
FLP points out that the beam pipe corrugation can act as a grating emitting light back to the bunch (THz radiation). But the fact that it is a random pattern is to our favor.
BB: You can modify the formalism to arbitrary modulation, not just sinusoidal, as long as they are periodic.
MP: How do you evaluate the effect from the entire bunch compressor?
BB: It is a very simple estimation: I compute the wake function, then assume it is the same all along the bunch compressor and just add up the effect. I further assume maximal dispersion. Clearly an overstimation that should be checked with a tracking code!
FLP points out that the wall impedance is quite similar for Al and for Cu. But machining of Al is much easier than Cu!
BB: It looks like the dominant effect is due to the surface roughness rather than the wall impedance.
TG: Vacuum requirements will also play a role when choosing a material!
FLP: Why is our beam pipe rectangular?
AO: The beam grows transversally.
FLP: Then why not elliptical?
MP: That would be too complicated to manufacture.