emittances in elegant

[astecpete]

Hi Michael,
I've had similar issues to this one for a while, but never asked!
I include a plot of ecnx, ecny and etax through an injection system. This comes from tracking a (non-gaussian) distribution through. I'm worried about the apparent emittance blowups in the y-plane, particularly in the third quadrupole (before the dispersive region). Can you suggest why this might be and/or tell me a bit about how elegant calculates emittances.

[michael_borland]

Pete,

My guess about the source of the vertical emittance growth is chromatic aberrations from the quadrupoles. The emittance growth scales like sigma_y^2*sigma_delta/f, where f is the focal length. So better control of the vertical beamsize at strong quads will help.

Elegant computes emittances as follows:
* The geometric emittance is e_g^2 = <x^2> <xp^2> - <x xp>^2
* The normalized emittance is e_n = e_g*<gamma>
* The corrected emittances are the same, except x -> x - delta*sigma_16/sigma_66 etc. Thus, the corrected emittance has dispersion-like correlations removed from the beam distribution. Put another way, it's the betatron emittance.

--Michael

[astecpete]

Thanks very much Michael - now you say it, it's obvious! The quadrupole has a k of -8 and betay is maximal there (but only 40m). Biggest thing is that i have a fairly large chirp there so sigma_delta is high.
I'll add another quad to control it better

Pete