Dispersion change in ver. 2.17.1.7 vs 2.16.6.1

[emo]

Dear Didier,

I opened a project in TraceWin 2.17.1.7, which I have created with an older version (2.16.6.1).
Apparently, the dispersion is calculated differently in the two versions (in envelope mode).
In the older version I had reached achromaticity whereas with the newer version this is not the case.
I have opened the same saved project and I get the same transverse envelopes but I get different results for dispersion.
It seems that the difference originates at the ELECTROSTA_ACC element.
Here are the two plots.
New TraceWin version:

Envelopes and dispersion - 2.17.1.7.png (41.38 KiB) Viewed 2431 times

Older TraceWin version:

Envelopes and dispersion - 2.16.6.1.png (40.7 KiB) Viewed 2431 times

The project files are attached too.

Could you tell me what is the reason and which version makes the correct calculation, please?
Thank you in advance.

Best regards,
Emil

[Didier]

Dear Emil,

Yes, you are right, corrections have been made recently on the ELECTROSTA_ACC element, in particular on the longitudinal matrix (see the manual). But for the moment they have not been checked yet, even if I am convinced that they were necessary. So it does have an impact on the dispersion.

Regards,

Didier

[emo]

Dear Didier,

Thank you for your answer.

Please, let me know when you know more on the subject and when you have checked the modifications.
I will be very grateful to know which of the two approaches gives a better estimation of the dispersion.
Thank you in advance.

Best regards,
Emil

[Didier]

Dear Emil,

If I add a small input energy error one the input beam and tune the line according to both ELECTROSTA_ACC element matrix, I got following results:
You can see that second simulation is much better (second matrix def.). Second point, if you compare tracking result wiht envelop result for a non CW beam, there are big divergences between both with the first matrix.

Regards,

Didier

first_def.png (17.88 KiB) Viewed 2422 times

second_def.png (17.35 KiB) Viewed 2422 times

[emo]

Dear Didier,

Thank you for the explanations.
Indeed, the second plot looks better.

The matching of the dispersion in envelope and multiparticle modes is another issue I have observed.
I was actually planning to ask you about it too.

Thank you.

Best regards,
Emil