tuneshift with energy: tracking versus using twissparameters

Moderators: cyao, michael_borland

Post Reply
nadolski
Posts: 22
Joined: 20 Mar 2013, 05:03
Location: Synchrotron SOLEIL, Gif-sur-Yvette, France
Contact:

tuneshift with energy: tracking versus using twissparameters

Post by nadolski » 12 Jan 2022, 09:21

Dear All,

When I compute the tune shift with energy by tracking I am not able to reproduce the results obtained using the variation of the twiss parameter with energy

I have joined two scripts:
  • offMomentumTwiss.tgz: Twiss parameters with energy (from Michael)
    runScan
    collateScan
    nux_vs_energy.png
    nuy_versus_energy.png
  • by tracking, I got very different curves. The half integer resonance is crossed around 5% (2.5% for Twiss)
    runTuneshiftwithenergy
    tuneshift_bytracking.png
I suspect something wrong in the script making the computation by tracking.

Thank you for your guidance.

Laurent.
Last edited by nadolski on 12 Jan 2022, 09:30, edited 1 time in total.

nadolski
Posts: 22
Joined: 20 Mar 2013, 05:03
Location: Synchrotron SOLEIL, Gif-sur-Yvette, France
Contact:

Re: tuneshift with energy: tracking versus using twissparameters

Post by nadolski » 12 Jan 2022, 09:28

Somehow, I was not able to join the script in the previous post.

by tracking
tuneshiftbytracking.tgz
(581.83 KiB) Downloaded 66 times
using Michael script for the variation of the Twiss parameter with energy.
tuneshiftbytracking.tgz
(581.83 KiB) Downloaded 66 times
Attachments
offMomentumTwiss.tgz
(19.92 KiB) Downloaded 54 times

michael_borland
Posts: 1772
Joined: 19 May 2008, 09:33
Location: Argonne National Laboratory
Contact:

Re: tuneshift with energy: tracking versus using twissparameters

Post by michael_borland » 13 Jan 2022, 19:50

Laurent,

I suggest two changes for the tracking runs: first, use the sixth-order integrator by setting INTEGRATION_ORDER=6 on the KOCT, KSEXT, KQUAD, and CSBEND elements. Second, increase the number of slices by about a factor of 2.

--Michael

nadolski
Posts: 22
Joined: 20 Mar 2013, 05:03
Location: Synchrotron SOLEIL, Gif-sur-Yvette, France
Contact:

Re: tuneshift with energy: tracking versus using twissparameters

Post by nadolski » 14 Jan 2022, 01:10

Dear Michael,

Thank you for the suggestion.
I already did some studies playing with the number of slices and the integrator order.
There is no major effect. The converge is slightly improved but at the margin.
Frequency map are very similar.

I post for you information the tune shift by tracking with N_SLICES=10 and INTEGRATION_ORDER=6.
tuneshift_6thorderIntegrator.png
Question: the tracking is done around the closed orbit.
!track particles whether to center the beam transverse coordinates on the close orbit !before tracking
&track center_on_orbit=1 &end
Are you sure that the option is working properly?

Laurent.

nadolski
Posts: 22
Joined: 20 Mar 2013, 05:03
Location: Synchrotron SOLEIL, Gif-sur-Yvette, France
Contact:

Re: tuneshift with energy: tracking versus using twissparameters

Post by nadolski » 14 Jan 2022, 01:41

Finally I have got the point, looking directly at the phase space.
The RF cavity was defined twice.
The tracking was performed in 6D with a cavity instead of 4D.

michael_borland
Posts: 1772
Joined: 19 May 2008, 09:33
Location: Argonne National Laboratory
Contact:

Re: tuneshift with energy: tracking versus using twissparameters

Post by michael_borland » 14 Jan 2022, 09:55

Laurent,

Yes, that would do it. Sorry I missed it.

--Michael

Post Reply