Elegant users forum

Hi Michael,

I am trying to use a time reference for the phase in a linac- cavities structure instead of putting a defined phase in each cavity. The reason is that I do not have relativistic particles in the first linac structure . I am now using PHASE_REFERENCE=(n), but I am not sure I understand exactly how it is working.
For example, in the structure I’ve got: driftline1, linacA, dritfline2, linacB to begin with. Where linacA and B consist of 152 RFCAs and after 3 cavities the phase of the rf-singal have made one revolution. So, my questions are, does elegant start its phase reference at the first cavity in linacA? (or can it start at t=0 so that I can use the length of the first driftline to choose the input phase in the linacA )? If elegant is using the first cavity as phase reference, should I define a special phase or id there some kind of default value?

I hope you can help me with this !
Best regards,

Olivia

here is the ELE-file

Olivia,

Sorry for the long delay in replying.

The idea of having the PHASE_REFERENCE parameter is to allow tying together the phases of different elements. For example, imagine that you have a long accelerating structure with N cells, which you want to simulate using N RFCA elements. If the beam is not relativistic, the phase in each cell will be different. By default, elegant phases each cavity independently so that the beam sees the phase given by the PHASE parameter. If you want to instead force elegant to tie all the cavities to the same external clock, you give them all the same phase reference.

Looking at your lattice file, it seems you are doing this. One problem is that you don't assign PHASE values to the cavities, so they are set to the default PHASE=0 (zero crossing). Probably you want PHASE=90. Another problem is that the cell length (0.033m) is not quite right for the frequency (2985MHz). If I adjust both of these, the simulation seems to work as expected. See attached.

By the way, there's an element called TWLA (Traveling Wave Linear Accelerator) that might be more appropriate for the early part of your simulation.

--Michael