Buncher cavities in version 2.23.2.1 not working

[NataliaM]

Hi,

Yesterday I was trying to simulate the MEBT section of ESS and I noticed that the buncher cavities were completely defocusing in the longitudinal direction (I only tested envelope). If you set the RF amplitude to negative values it looked the dynamics was restored.
I rolled back to version 2.23.1.6 and it works fine (using same lattice and input twiss) so my guess is that there is something funny on the news version.

Best,
Natalia

[Didier]

Dear Natalia,

I took a design from the MEBT+DTL of your machine that I had in 2022 and bellow, this is the envelope that I get with the latest version of TraceWin without changing anything in the initial project.
I'm sorry, but I don't know what you're talking about. Please send me a very simplified example illustrating your problem so that I can help you.

MEBT.png (18.93 KiB) Viewed 3259 times

Regards,

Didier

[NataliaM]

This is exactly what I was running yesterday in the latest version that did not work. I have no idea what was the issues but I only managed to get it back on track after I uninstall and rolled back.

[Didier]

Hi,

Ok thank you for the example, but once again I get the previous plot!
Could you show me the plot you have that you think is wrong?

Regards,

Didier

[NataliaM]

I would get this plot constantly. I even erase everything and started fresh in another folder. This the MEBT+DTL envelopes.

[NataliaM]

I thought was an issue with the version since the only way I managed to get it working was to roll back to an older one. Maybe there was some setting that got stuck or something with my computer in particular when I did the update, but since in your case it runs fine then it will be hard to tell.

[Didier]

Yes, it's complicated to help you. Having said that, I'm going to focus on the first rebuncher, which obviously works the wrong way round. I'll try to reduce the problem as much as possible and understand where its field value comes from.

Regards,

Didier

[ryoichimiyamoto]

Dear Didier,

I also looked at this problem and found that the issue seems only for macos. I summarize what I found in the following.

- I tested recent versions for linux and macos. (I don't have access to a windows machine.) There was no issue for any of the linux versions. For macos versions, it was fine until 2.23.1.6. The issue was there for 2.23.2.0. I didn't have versions between 1.6 and 2.0.

- In the recent macos versions, for the "GAP" element, the sign of the phase seems ignored. For example, the transfer matrix for -90 deg phase was identical to that of 90 degree case.

This is all. Thanks for your support.

[Didier]

Dear Ryoichi,

Very surprising but unfortunately true.
That's fixed, thank you for pointing out the problem.

Regards,

Didier