Dear Duperrier and Dear Didier,
We are currently running simulations using both Toutatis integrated in TraceWin and Toutatis2_console_win64.exe, and are comparing the output results from the two.
We observe discrepancies in the results, for example, in the rms emittances and the beam transmission, as attached (.png). We would expect the two results to agree when the same initial conditions are used.
We would appreciate your clarification on the following points:
1. How is the mesh size determined or configured for Toutatis in TraceWin?
2. Are there any settings that require particular attention to obtain consistent results between Toutatis in TraceWin and the Toutatis2_console version? I have attached my simulation input files, and I would be grateful if you could review them.
3. In principle, should the results agree when the same initial conditions and accuracy settings are used? Or are discrepancies expected due to differences in implementation or coupling details between TraceWin and the console version?
Thank you very much for your time and support.
Discrepancies Between Toutatis in TraceWin and Toutatis2_console Results
Discrepancies Between Toutatis in TraceWin and Toutatis2_console Results
- Attachments
-
- comparisons.png (278.58 KiB) Viewed 92 times
-
- RFQ_324MHz.inp
- (54.56 KiB) Downloaded 5 times
-
- rfq.dat
- (36.88 KiB) Downloaded 4 times
-
- from_RFQexit.ini
- (43.77 KiB) Downloaded 4 times
-
RomualdDuperrier - Global moderators
- Posts: 83
- Joined: Fri 28 Aug 2020 14:59
- Location: Paris-Saclay
- Country:
France (fr)
Re: Discrepancies Between Toutatis in TraceWin and Toutatis2_console Results
Post
by RomualdDuperrier »
Hi,
I think I have found the main source of the discrepancy. In RFQ_324MHz.inp, you have both:
THE2TERMSFLAG
THE2TERMSFORMULATIONFLAG
The keyword THE2TERMSFLAG means that the code must generate a vane geometry with a modulation profile following the simplified 2 terms potential (sometimes called the ideal potential). But you also add the keyword THE2TERMSFORMULATIONFLAG, in this case the code will not use the finite difference Laplace solver and will use the ideal 2 terms potential instead. This leads to better transmission as no multipoles are considered.
In Tracewin, only the keyword TWOTERMS is used. This is equivalent to using THE2TERMSFLAG only with Toutatis. In this case, the Laplace solver is used and multipoles will be present. Transmission is degraded.
I recommend then that you delete THE2TERMSFORMULATIONFLAG in RFQ_324MHz.inp.
All the best.
Romuald
I think I have found the main source of the discrepancy. In RFQ_324MHz.inp, you have both:
THE2TERMSFLAG
THE2TERMSFORMULATIONFLAG
The keyword THE2TERMSFLAG means that the code must generate a vane geometry with a modulation profile following the simplified 2 terms potential (sometimes called the ideal potential). But you also add the keyword THE2TERMSFORMULATIONFLAG, in this case the code will not use the finite difference Laplace solver and will use the ideal 2 terms potential instead. This leads to better transmission as no multipoles are considered.
In Tracewin, only the keyword TWOTERMS is used. This is equivalent to using THE2TERMSFLAG only with Toutatis. In this case, the Laplace solver is used and multipoles will be present. Transmission is degraded.
I recommend then that you delete THE2TERMSFORMULATIONFLAG in RFQ_324MHz.inp.
All the best.
Romuald