Elegant users forum

Hi All,

For a simple bend system or linac, I want to impose FSE errors on the quadrupoles/dipoles and find the deviation of the on-energy orbit. At the moment, I am doing this by subtracting off the <Cx,Cy> (which is small but not non-zero) from the %s.cen output file for FSE=0 from the <Cx,Cy> parameter from the %s.cen output file for FSE=/=0 with the 'n_particles_per_bunch' parameter in 'bunched_beam' set to 1e5.

This is working but I would like to gain statistics on the deviation and run the simulation a few thousand times which would take a very long time. Is there another way or parameter to look at that could save time. Would <etax,etay> leakage from the %s.twi file be a good indicator of the on-energy orbit deviation (I could run this with n_particles_per_bunch=1 which would be much faster)?

Any advice would be greatly appreciated!

Unless your system has significant nonlinearities, I think using n_particles_per_bunch=1e5 is unnecessary. A single particle started at the centroid location should follow the centroid perfectly in a linear system.

Which part of the simulation is too time-consuming, the tracking or the post-tracking analysis?

--Michael