Ring tracking with ciwggler element

[weijie]

Hi Michael,
We are doing damping ring tracking with the ciwggler element. But we have found that the bunched beam will get lost at the first pass and the paticle can not suvive. can you help us.
Best Regards
Weijie Fan

[michael_borland]

Weijie,

Thanks for posting this problem. It was interesting to analyze. (See attached for details.)

I see two issues with the design. First, with only one rf cavity in the ring, the optics is badly distorted by the strong energy variation introduced by the wigglers; I addressed this by putting an rf cavity after each wiggler.

Second, the second order chromaticity is very large, so the tunes cross the half-integer resonance at a momentum deviation of about 0.4%. However, because of the damping wigglers, the energy spread is 0.2%, so the quantum lifetime is very short. That causes the beam to leak away very quickly.

In analyzing this problem, I realized that I wasn't including the second-order chromatic effects of the CWIGGLER elements. I added that and it will appear in the next release, which should be out in the next week.

--Michael

[weijie]

Dear Micheal

Sorry for my late apply and thank you for your kind assistance!

We have realized that a single RF cavity is insufficient in compensating for the energy loss.
However, we have tried to decrease the number of the damping wigglers to 6 super peroid, but the paricle leak quickly also. Maybe the energy spread is still too large. In our second design, we incorporated distributed RF cavities and the dynamic apearture is still zero. Does the DA disappear with a relatively large beta function value (max to 40 m)?

Thank you for your efforts in incorporating second-order chromatic correction into the CWIGGLER elements, and I have gotten the latest version

Once again, we express our gratitude for your support.

Weijie Fan

[weijie]

However, we have tried to decrease the number of the damping wigglers to 6 super peroid, but the paricle leak quickly also. Maybe the energy spread is still too large. In our second design, we incorporated distributed RF cavities and the dynamic apearture is still zero.

Besides, we have also double checked DA by AT 2.0 (MATLAB version), and finally got DA solutions, as the picture shown. I am wondering why there are different results.
Thank you

[michael_borland]

Weijie,

I can't comment on why AT and elegant give different results in this case, since I don't use AT. We did a comparison of AT and elegant for the APS upgrade, and the results were very similar.

https://journals.aps.org/prab/abstract/ ... .22.114601

--Michael

[weijie]

Dear Michael

The results between AT and elegant are consistent for the ring without a damping wiggler. But when it comes to damping wigglers, elegant will show different results compared with AT. I don't know if there is any error in the damping wiggler design.

[michael_borland]

Weijie,

It is puzzling that they don't agree, then. The CWIGGLER element uses the code developed by Ying Wu for AT many years ago.

If you are just using sinusoidal mode, you could check the vertical tune using tracking and compare that to the twiss parameter calculation. They should agree closely.

--Michael