R56 with transverse SC

[Aamna_Khan]

Hello Michael,

I can get the disperion (etax) from .sig file as: s16/s6*s6 but I am confused how to compute the R56 with transverse SC kicks in elegant.

From my understanding,
1) R56 is matrix output as: ds/ddelta
2) R56 = Integral[etax/rho.ds]

It means R56 should also change with SC kicks. In my simple envelope code with SC kicks I got the attached results.

But I don't know how to get matrix element R56 correctly from tracking.

PS: when I'm using the option local_disperion=0 in twiss output then disperion is zero across the whole beamline.

Thanks,
Aamna

[michael_borland]

Aamna,

I'm not entirely sure what you want to determine. Two possibilities occur to me:
1. You want to determine how the centroid (i.e., average over the bunch) final arrival time varies with centroid initial momentum deviation, including some collective effects (in this case, transverse space charge).
2. You want to determine how the individual-particle arrival time varies with individual-particle initial momentum deviation, including some collective forces from a bunch.

Depending on which you want, the strategy in tracking would be different.

--Michael

[Aamna_Khan]

Hi Michael,

Sorry I didn't mentioned it clearly. Actually, I want to determine both possibilities mentioned by you for different disperion matching latter on. It would be very helpful if you can provide me some examples.

Thanks,
Aamna

[Aamna_Khan]

Hello Michael,

Sorry to bother you again. I got stuck in it. Could you please provide me any example?

Aamna

[michael_borland]

Aamna,

Can you provide your lattice file? That will make it easier to set up an example.

Thanks.

--Michael

[Aamna_Khan]

Here they are

Aamna

[michael_borland]

Aamna,

Attached you'll find a pair of examples that find Ri6 for the two cases: whole beam displacement and probe particle displacement. The results are nearly identical because the collective effects are very weak (Q=0.77e-19 C). If the charge is increased, the results appear to change only for the probe particle displacement case, which is probably to be expected for this simple model of space charge.

That reminds me: as the manual page for SCMULT states, the space-charge algorithm is design for weak space-charge kicks in storage rings, so I would validate it against another code (e.g., ASTRA) for your situation before trusting the results.

--Michael

[Aamna_Khan]

Thank you.

In real simulations Q is much higher (0.77e-12-0.77e-10 C)...I was doing some stuff with this example and you are right about the SC model I will benchmark the results against Impact or astra.

Aamna

[dmzhou]

Hi Aamna and Michael,

Thanks for the nice examples you shared in the previous posts. I am also looking at the transverse space charge (TSC) effects of a linac with beam energy about 10 MeV. I tried Aamna's example (test.lte and test.ele) and made minor changes:
Bunch charge Q =0.77e-19 => 0.77e-10
With this change, I can see significant change of beam sizes along the beam line (check output file test.sig). It means the linear TSC is turned on.
nonlinear =0 => 1 (in insert_sceffects)
With this change, I expected the nonlinear TSC is activated but I cannot see any difference when comparing the test.sig files.

I am using Mac version 2020.4.0. Any suggestions? Thanks.

Demin