orbit correction for a long beam line

Moderators: cyao, michael_borland

Post Reply
esmahmadi
Posts: 6
Joined: 09 Jun 2021, 08:46

orbit correction for a long beam line

Post by esmahmadi » 30 May 2022, 02:34

Dear Michael,
I am trying to correct the orbit for a ring with a large number of corrector magnets and BPMs (3324 correctors and BPMs).
The problem is that the correcting procedure stops at the stage of 'computing orbit response matrix' :
Adding CH#3324 (type HKICK) at 9.117189e+04 m to h plane steering
The steering parameter is name = KICK, value = 0.000000e+00
updating twiss parameters...27712 matrices recomputed for periodic Twiss parameter computation
statistics: ET: 00:01:57 CP: 116.93 BIO:0 DIO:0 PF:0 MEM:73637

done: ET: 00:01:58 CP: 117.53 BIO:0 DIO:0 PF:0 MEM:74858
computing orbit response matrix...
I attach the .ele file which I used for orbit correction.
Could you please give me an idea of what is the problem?
Best regards
Esmaeil
Attachments
run1.ele
(1.87 KiB) Downloaded 291 times

michael_borland
Posts: 1959
Joined: 19 May 2008, 09:33
Location: Argonne National Laboratory
Contact:

Re: orbit correction for a long beam line

Post by michael_borland » 27 Jul 2022, 20:06

Esmaeil,

Can you upload your lattice file?

--Michael

esmahmadi
Posts: 6
Joined: 09 Jun 2021, 08:46

Re: orbit correction for a long beam line

Post by esmahmadi » 31 Jul 2022, 06:36

Dear Michael,
Thanks for the reply.
I attached the lattice file and the .ele file which I used for orbit correction.
Two other points,
The elegant crashes when I am trying to run it in windows.
there is a big difference in vertical chromaticity (10 units) between MADX and elegant.
many thanks in advance,
Esmaeil
Attachments
lattice_2.lte
(1.57 MiB) Downloaded 248 times
error.ele
(1.61 KiB) Downloaded 252 times

michael_borland
Posts: 1959
Joined: 19 May 2008, 09:33
Location: Argonne National Laboratory
Contact:

Re: orbit correction for a long beam line

Post by michael_borland » 01 Aug 2022, 19:48

Esmaeil,

Your lattice does not close upon itself. I guess this is because of insufficient precision in converting from MAD-X to elegant format. elegant is unable to obtain a closed orbit even in the absence of errors.

You can use the floor_coordinate command to compute the the floor coordinates and see the closure error (about 0.1 mm in Z and 35 micron in X). I suggest fixing this problem before proceeding.

Also, you shouldn't use alter_elements to impart errors, particularly the way you are doing it because you are assigning the same error to every QF* and QD* magnet. That's going to change the lattice in a very different way to random errors.

--Michael

esmahmadi
Posts: 6
Joined: 09 Jun 2021, 08:46

Re: orbit correction for a long beam line

Post by esmahmadi » 02 Aug 2022, 09:13

Dear Michael,
many thanks for your help. I will try that.
Actually, the reason that I assigned the errors by using Dx, is that I need to study the effect of systematic horizontal displacement of all arc quadrupole magnets on the closed orbit.
Best regards
Esmaeil

Post Reply