Clarifications on cavities

[jarnovandewalle]

Dear Didier,

Some questions regarding the treatment of cavities and errors ... (+ probably a bug in the manual ... ?)

1/ The DIAG_DPHASE and DIAG_DENERGY refer to a Delta Energy and Delta centroid Phase (on p. 57,58 in the manual) but it's not explicit to what reference these differences refer. Assuming this would be the 'reference beam' in case of an error study ?
Additionally, the delta centroid phase, does that refer to the 'absolute phase' or the 'synch phase' in the cavity ?

2/ In the context of a study of our linac where some cavities are off (broken), I tried several methodologies ... (FYI: I'm using 3D EM fields for all cavities, so E and B fields)

A first method is to put the ke and kb factors of the 'broken cavities to zero 'manually' and than using the neighboring cavities to get the original energy back after a certain section. This works well, but has the disadvantage that any dipole after the linac assumes the energy out of the linac is the reference (which is not the case of course)

A second method, is to use the command "ERROR_CAV_NCPL_STAT 1 0 0 0 0 0 -100 0 0", which puts the ke factor to zero. But when no matching is done after this (so no DIAG_DPHASE and DIAG_DENERGY are put in the .dat file), the subsequent cavities do not accelerate properly anymore, although the phases for these cavities are defined as the phase 'when the generatrix particle enters the cavity' ... so they should still get accelerated (even though the absolute phase is -slightly- off ....)
Could you explain what actually happens in this case and whether there is a way around this ?

3/ In the ADJUST command, there is a mention that when nu is lower than zero, it 'allows to replace both command "ADJUST_DPHASE" and "ADJSUT_DFIELD"....' This is not entirely clear, since no further reference is made to ADJUST_DPHASE or ADJUST_DFIELD in the manual.

4/ I think there is a bug in the manual on p. 59 (DIAG_TWISS) in the 4th column
Wanted betZdp/p (mm/mrad) = 10 x time beZdp/p (mm/%)
This should be
Wanted betZdp/p (mm/mrad) = beZdp/p (mm/%) / 10 ?

Many thanks in advance for any advice on this!

With best regards
Jarno

[Didier]

Dear Jarno,

1- Yes, your hypothesis is correct, the differences are between the reference machine with no errors and the one with errors and the phase delta refers to the absolute phase.

2- You've clearly identified the limits of two possible solutions. Personally, I prefer the second. On the other hand, a second matching will be made, taking errors into account, if you have selected the right options on the main and error pages. But this last matching will not be saved as a reference machine in the dat file. But I can't really see what you've actually done, without access to your project files.

3- here, I don't agree with you and I think it's right.

Regards,

Didier