how can I use this grammar ?

[michael_borland]

Joy,

Can you post your .ele, .lte, and script files? It is hard to determine the problem otherwise.

Thanks--Michael

[Joytop]

Michael ,

I post them in the attachment. Many thanks.
You can run the files.

And I print out the .sig and .twi at the point of element of W10 :

ElementName betax betay alphax alphay
m m
W10 5.294596e+00 1.327473e+01 3.855053e+00 -7.775626e+00
ElementName s1 s2 s12 ex
m m m
W10 2.638371e-04 1.933819e-04 -4.814892e-08 1.687768e-08

betax_calculated = s1^2 /ex = 2.638371e-04^2/ 1.687768e-08 = 4.1244 < 5.294596e+00
alphax_calculated = -s12 /ex = 4.814892e-08 / 1.687768e-08 = 2.8528 < 3.855053e+00

Why ?

Joy

[Joytop]

Michael ,

The files are 10M , I can not attach them. So I email you all the files you need.
Please download the files from the email address : flood@aps.anl.gov<flood@aps.anl.gov>
Joy

[michael_borland]

Joy,

The problem with disagreement of the beta functions computed from the beam and from the twiss output file is that you are including collective effects (non-zero charge, plus CSR in the dipoles). If you set the charge to 0, you'll get good agreement except inside the chicane. You are also using CSRCSBEND elements, which results in some nonlinear dispersion. If you use first-order SBEN elements instead, you'll get good agreement inside the chicane as well.

The twiss calculation can't account for these effects, of course. So if you really have these effects at such a strong level, it will be quite difficult to get these calculations to match.

--Michael

[Joytop]

Dear Michael,

Thanks so much for your help.

Best wishes for your families,
Joy.