Dear Didier:
I have a problem about the synchronous phase adjusted for two different sections. As the following linac(please see the attachment), there are two sections, a 324MHz spoke section and a 648MHz elliptical cavity section. For the elliptical cavity section ,why the synchronous phase of the second period is smaller than that of the first period? Generally, the synchronous phase of the first period should be the smallest one(2 times of the former period) and then improves by periods.
Thank you!
Regards,
Emily
ps: Because the forum does not recognize the file with the extension *.gen, so I changed the extension to *.txt.
about the synchronous phase in the transition period
Re: about the synchronous phase in the transition period
Dear Emily,
I fixed the problem about *.gen file extension in the forum.
Concerning your question, I think it is exactly the opposite of what you say. Doubling the RF frequency, you multiply by 2 the rms beam phase size (beam phase spread) with respect to RF frequency. So you roughly reduce the longitudinal acceptance by a factor of 2, then you have to double the synchronous phase to keep the longitudinal acceptance constant (roughly Synchronous phase / beam_phase_spread ) as requested.
Regards,
Didier
I fixed the problem about *.gen file extension in the forum.
Concerning your question, I think it is exactly the opposite of what you say. Doubling the RF frequency, you multiply by 2 the rms beam phase size (beam phase spread) with respect to RF frequency. So you roughly reduce the longitudinal acceptance by a factor of 2, then you have to double the synchronous phase to keep the longitudinal acceptance constant (roughly Synchronous phase / beam_phase_spread ) as requested.
Regards,
Didier
Re: about the synchronous phase in the transition period
Dear Didier:
Yes, you are right, at the first period of the elliptical cavity section, the synchronous phase should double to keep the longitudinal acceptance constant. However, the linac generated by the GenlinWin didn't obey this rule, as shown in the following picture. I don't know why.
Regards,
Emily
Yes, you are right, at the first period of the elliptical cavity section, the synchronous phase should double to keep the longitudinal acceptance constant. However, the linac generated by the GenlinWin didn't obey this rule, as shown in the following picture. I don't know why.
Regards,
Emily
- Attachments
-
- synchronous phase.png (28.43 KiB) Viewed 39049 times
-
- Linac.gen
- (138.16 KiB) Downloaded 2303 times
Re: about the synchronous phase in the transition period
Dear Didier:
I upgraded the code. However, there was still a problem, as shown in the following picture. The synchronous phase changed abruptly at the third period. The synchronous phases follow the rule to keep the constant longitudinal acceptance.
Regards,
Emily
I upgraded the code. However, there was still a problem, as shown in the following picture. The synchronous phase changed abruptly at the third period. The synchronous phases follow the rule to keep the constant longitudinal acceptance.
Regards,
Emily
- Attachments
-
- selection.png (12.49 KiB) Viewed 39033 times
-
- const_longitudinal_accept.png (23.84 KiB) Viewed 39033 times
Re: about the synchronous phase in the transition period
Dear Emily,
Many optimisations don't always work well together because they contradict or compensate for each other, so in your case, it's the continuity of the phase advance that is the problem. I have put a new version that improves this point a bit. Even so, there are still some points that are impossible to fix because they contradict each other.
So upgrade GenLinWin and for section#2 remove from "Phase advances" continuity with preceding section" or reduce "Maximum varation per period" form 4° to 1.5° (see picture)
In any case, I suggest you use the 'Const phase acceptance' choice rather than 'Const logitudinale acceptance'. The latter is clearly not as good as the first and is kept only to keep compatibility with the old designs generated by using it. Furthermore, the first one has no compatibility concerns with phase advance options.
Regards,
Didier
Many optimisations don't always work well together because they contradict or compensate for each other, so in your case, it's the continuity of the phase advance that is the problem. I have put a new version that improves this point a bit. Even so, there are still some points that are impossible to fix because they contradict each other.
So upgrade GenLinWin and for section#2 remove from "Phase advances" continuity with preceding section" or reduce "Maximum varation per period" form 4° to 1.5° (see picture)
In any case, I suggest you use the 'Const phase acceptance' choice rather than 'Const logitudinale acceptance'. The latter is clearly not as good as the first and is kept only to keep compatibility with the old designs generated by using it. Furthermore, the first one has no compatibility concerns with phase advance options.
Regards,
Didier
- phase.png (201.2 KiB) Viewed 39015 times