Page 1 of 1

Off-momentum DA

Posted: 21 Jan 2019, 08:49
by SDastan
Hello Michael and all,
I want to track DA for off-momentum particles. I used the elegant example which is for off-momentum DA but I want to add some magnet errors, then I want to correct the orbit and close-orbit and finally track DA. I want to know should I increase the values like
closed_orbit_accuracy = 1e-9 or not?
I attached my .ele and .lte files and I really appreciate if you take a look at the files and help me to understand my problem, because for example when I track off-momentum (dP/p=0.04) particles I get zero result!!!
Thanks in advance,
Sara

Re: Off-momentum DA

Posted: 23 Jan 2019, 09:57
by michael_borland
Sara,

Sorry for the delayed reply. It took a while to put together an example of how to handle a case like this.

The basic problem is that when one adds errors to lattices with strong sextupoles, the tunes may change significantly or the closed orbit may not exist. It's necessary to start with the sextupoles turned off, then ramp them up to full strength while correcting the orbit. Once that is done, one can perform DA tracking. There is a Notebook file in the attached zip file that explains the process. The zip file also contains the results of scanning the momentum offset and computing the DA (4d and 6d methods) for 10 seeds.

Here's a sample from 6d tracking (i.e., with rf cavities and damping).
6dDAvsMomentumOffset.png
--Michael

Re: Off-momentum DA

Posted: 26 Jan 2019, 04:45
by SDastan
Dear Michael,
The example that you made is fantastic and solve the problem.
Thank you so much.
Sara

Re: Off-momentum DA

Posted: 03 Feb 2019, 03:55
by SDastan
The example which is provided for off-momentum DA is fantastic and the Notebook file is really helpful. I just want to add a point that if you want to run the last step in the Notebook file, I mean
postProcDA
you should run the initial steps, I mean
prepCase
run4dDA
run6dDA

for more than two different seed values.

Sara
michael_borland wrote:
23 Jan 2019, 09:57
Sara,

Sorry for the delayed reply. It took a while to put together an example of how to handle a case like this.

The basic problem is that when one adds errors to lattices with strong sextupoles, the tunes may change significantly or the closed orbit may not exist. It's necessary to start with the sextupoles turned off, then ramp them up to full strength while correcting the orbit. Once that is done, one can perform DA tracking. There is a Notebook file in the attached zip file that explains the process. The zip file also contains the results of scanning the momentum offset and computing the DA (4d and 6d methods) for 10 seeds.

Here's a sample from 6d tracking (i.e., with rf cavities and damping).
6dDAvsMomentumOffset.png

--Michael