Overlapping Elements in ImpactX

[Binabek]

What is the appropriate way to implement overlapping elements in Impact-X? In IOTA, the vertical injection kicker runs through two quads. For injection simulations should I implement this overlap as CFBend elements (this is a bit inaccurate since the kicker is not a magnetic dipole) for proper orbit adjustment? Or, should I implement a series of fractional strength Kickers elements over the physical kicker length including in the quad? Do the Kicker elements actually adjust the orbit? I recall in the past that rotated sbend elements did not necessarily change the reference particle orbit. I'm also not sure how this second approach might mess with the space charge slicing in the drifts/quads.

[cemitch99]

Thanks @Binabek. Handling overlapping elements (in a symplectic tracking code) is a delicate problem. Can you tell us a bit about the field of the kicker itself? The existing "kicker" element kicks the particles in the beam with respect to an unperturbed reference orbit--so you will see the effect of the kick on the beam centroid, but not the reference particle. I was going to suggest using CFBend. There is an open PR on symplectic integration through a user-defined potential, which would allow tracking through a superposition of fields.

[Binabek]

The physical kicker in question is a travelling wave strip-line type, see DOI:10.48550/arXiv.1607.00023 for a high level overview. In the case of the potential user defined fields, is it intended to impact the trajectory of the reference particle, or only the centroid?
Based on your response, it looks like my best course of action in the short term may be CFBends where the quads are and SBends for the drift regions between where the kicker spans.