how does Cwiggler radiation works?

[Zeus]

Dear Elegant coders,
We are trying to implement a general wiggler with radiation in AT. We are very happy with Elegant's Cwiggler but we also routinely use AT. We started looking at the way AT includes radiation in dipoles and tried to include it in Wu’s code. After many unsuccessful tests, since it is open source (I think this is very nice), we decided to take a look at your code.
Regarding your code, there is one basic thing that I do not understand, maybe you could explain it. Why before applying the radiation kick, there is a momentum coordinate change:
X[1] -= ax;
X[3] -= ay;
This may be a too basic question, maybe we are lacking some basic teaching about how to include radiation in the integrators, could you point out any paper or publication where this issue is addressed?
By the way, it just occurred to me that maybe you know if there already exists a version of the code with radiation for AT.
Sorry if this is not the appropriate place to rise all these questions,
Best regards,
Zeus

[michael_borland]

Zeus,

Before applying the radiation kicks, we remove the vector potential part of the canonical momenta. That allowed using some existing code in which the radiation affects the particle slopes, without incorrectly affecting the entirety of the canonical momenta. After applying the kicks, we add the vector potential contribution back. Strictly speaking, this isn't quite right as we assume in applying the kicks that the coordinates are x' and y', even though they aren't exactly. However, I think it is close enough.

I think the ESRF group (Boaz Nash) or SSRL group (Xiaobiao Huang) may have a version of AT with radiation effects.

--Michael

[Zeus]

Michel,
Thanks a lot for your answer.Truth be told, I always struggle with Hamiltonian mechanics. We asked Boaz before he left the ESRF, as far as I know, he had no time implement that, we'll ask Xiaobiao!
Zeus