Mismatch of vertical chromaticity from twiss_output with other codes

[felix_armborst]

Hello,

We have encountered a mismatch in the vertical chromaticity for the BESSY II design lattice when comparing the results from elegant's twiss routine with those of other codes. Could you please help me get to the bottom of this problem? Here an overview of the obtained results (I have attached all used files):

Code: Select all

				nux 	nuy	dnux/dp	dnuy/dp
	Elegant			17.85 	6.74	1.18	1.396
	MAD-X			17.85	6.74	1.18	-1.05
	PTC (exact=true)	0.85	0.74	1.18	-1.05
	PTC (exact=false)	0.85	0.74	1.11	1.395

Thanks in advance and greetings from Berlin,
Felix Armborst

Version Info:
elegant 2019.1.1
MAD-X 5.05.02

[michael_borland]

Felix,

You need to switch to a more sophisticated edge model, which you can do using

Code: Select all

B : CSBEND, n_kicks="bkicks", L="L", ANGLE="angle", E1="hangle", E2="hangle", &
  edge_order=2, edge1_effects=2, edge2_effects=2, hgap=0.001, fse_correction=1

I've put HGAP=0.001 as a place holder. Ideally, you need to set HGAP and FINT based on measurements of your dipole fringe field.

--Michael

[felix_armborst]

Hi Michael,

thank you very much for your fast response! Using 2nd order edge effects indeed solves my problem. The hgap is 25mm and fint should be around 0.7 (Rogowski formed pole shoes). I unfortunately do not have any measurements of the fringe fields. Which setting of fse_correction and edge_order would you recommend for investigations on far off axis closed orbits (TRIBs)?

--Felix

[michael_borland]

Felix,

The settings shown in my previous reply will invoke the best model elegant has, so I'd stick with those for small-radius rings.

--Michael

[felix_armborst]

Michael,

Thank you again for your fast and very helpful response!

Best regards,
Felix Armborst