"Six-dimensional" closed orbit

Moderators: cyao, michael_borland

Post Reply
Posts: 38
Joined: 28 Aug 2018, 09:46

"Six-dimensional" closed orbit

Post by GPenn » 13 Aug 2021, 16:33


I am doing comparisons between AT and elegant tracking, and now I want to compare the closed orbits. With rf turned on and with deterministic synchrotron radiation losses, AT has a function "findorbit6" which calculates a closed orbit that includes the variation in energy, and I think it also requires the time of flight to be an integer multiple of the RF period.

Is there an equivalent command in elegant? Does closed_orbit actually include the variation in energy but only reports the initial energy offset?
I can see how using rf_setup would in effect do a similar thing, by tuning the rf so the closed orbit sits on the nominal momentum. Is it expected that the single-turn variations in energy will not be significant enough to impact the closed orbit?

Finally, I have so far only seen values of 0 for the delta and lengthError parameters, even with errors, does that mean I'm missing something to force it to pay attention to longitudinal effects?

My typical function call is something like
output = %s.clo,
closed_orbit_accuracy = 1e-9,
closed_orbit_iterations = 500,
verbosity = 1,


Posts: 1815
Joined: 19 May 2008, 09:33
Location: Argonne National Laboratory

Re: "Six-dimensional" closed orbit

Post by michael_borland » 13 Aug 2021, 17:00


Elegant handles this via the "fixed_length" parameter in the &closed_orbit and &correct commands.

If fixed_length=1, elegant adjusts the initial momentum to ensure that the path length is the same as it would be for the ideal lattice. Hence, the effect of correctors in a dispersive location is included. With fixed_length=0, elegant assumes that the rf frequency is adjusted to remove the energy offset (hence, you'll see the delta parameter is 0 in the closed orbit output file).

Elegant does not tie these computations to the rf cavity. They can be done without a cavity and any cavity that's included is ignored for this purpose.

If you want synchrotron radiation included in these computations, use CSBEND, KQUAD, and KSEXT elements with SYNCH_RAD=1.


Posts: 38
Joined: 28 Aug 2018, 09:46

Re: "Six-dimensional" closed orbit

Post by GPenn » 13 Aug 2021, 17:53

Thanks Michael, that clarifies things a lot!


Post Reply