Page 1 of 1

insert_sceffects in linacs

Posted: 26 Oct 2020, 00:44
by nakamura
Hello Michael,

We are using insert_sceffects for transport lines or linacs to estimate TSC effects.
In the Elegant manual, this command is not designed for simulation in linacs.
But in fact I could get a simulation result of cERL using insert_sceffects.
I would like to know if this result is correct approximately at least.
Thank you in advance.

Norio

Re: insert_sceffects in linacs

Posted: 28 Oct 2020, 13:58
by michael_borland
Norio,

I've had reports of this feature being used with decent results for linacs or few-pass systems. I would think it is not too bad for smooth (e.g., gaussian) bunches. For complex longitudinal phase space such as might be found in beam subjected to significant bunch compression, I would consider it very unreliable.

--Michael

Re: insert_sceffects in linacs

Posted: 28 Oct 2020, 19:23
by dmzhou
Hi Michael,

I am working together with Nakamura-san on the transverse space effects in cERL using ELEGANT. Could you please check my post as follows?
viewtopic.php?f=17&t=875

What I found:
1) I implemented the LSC and TSC model of ELEGANT (SCMULT) into a SAD script. But with TSC my simulation results are always significantly different from that of ELEGANT. Without TSC, I could find good agreement.
2) For a linac, likely the nonlinear TSC model cannot be activated in ELEGANT simulations. I tried to look into the source code (insertSCeffects.cc) but I could not find where the function complexErf is defined.

We are also running GPT for benchmark. We may report to you our findings later after we better understand ELEGANT simulations.

Thanks.
Demin

Re: insert_sceffects in linacs

Posted: 28 Oct 2020, 23:37
by nakamura
Michael,

Thank you very much for your quick reply.
Your answer may be a good news for us because we would like to use ELEGANT
for low energy accelerators such as the recirculation loop of cERL if possible.
LSC effects are most important and TSC effects are also significant there.
We also uses GPT for the recirculation loop as well as the injector, but it needs
much more time for simulations. ELEGANT is more suitable for survey or preliminary
survey of many parameters.
My coworker, Zhou-san, is interested in comparing Elegant, SAD and GPT
and will develop these codes if necessary.
I would like to thank you if you could answer his questions.

Norio