Calculation of transverse emittance from beam size measurements at three locations

[Ken]

Hi TraceWin users,
I am trying to find the input beam Twiss parameters and emittances given the rms sizes measured at the end of the line by three separated beam size detectors. This is the same input as in example 4 on page 88 in the manual, except that in this case the emittances do not have to be equal in each transversal plane, see structure file below. The problem is that after matching TraceWin does not seem to find reasonable solution in one plane:
X-X'
Emit [rms] = 0.2292 π.mm.mrad [ Norm. ]
Emit [100%] = 0.2292 π.mm.mrad [ Norm. ]
Beta = 11.2071 mm/π.mrad
Alpha = 12.7898
Y-Y'
Emit [rms] = 150800.0000 π.mm.mrad [ Norm. ]
Emit [100%] = 150800.0000 π.mm.mrad [ Norm. ]
Beta = 0.3749 mm/π.mrad
Alpha = 0.5619

Do you have any suggestions for me?

Thanks!,
Ken

===========================================================
; Structure file
; Proton @20 MeV, 10 mA
ADJUST_BEAM_CURRENT 99 1
ADJUST_BEAM_EMIT 99 1 1 0 ; Ex not equal Ey
ADJUST_BEAM_TWISS 99 1 1 1 1 0 0
DRIFT 0 100
DRIFT 100 100
QUAD 100 -15.18 20
DRIFT 200 100
QUAD 100 15.18 20
DRIFT 200 100
QUAD 100 -15.18 20
DRIFT 200 100
DIAG_SIZE 99 2.2 1.6
DRIFT 200 100
DIAG_SIZE 99 2.2 1.6
DRIFT 200 100
DIAG_SIZE 99 2.2 1.6
DRIFT 0 100
END

[Didier]

Dear Ken,

From my side it seems to work without any problem, now I don't have all the parameters of your project. Maybe you could attach the *.ini file here so I can understand why you are having difficulties.

Regards,

Didier

env.png (17.21 KiB) Viewed 2827 times

[Ken]

Dear Didier,
Please find the .ini file and the envelope plot attached.
Thanks,
Ken

[Didier]

Dear Ken,

I'm sorry, but I don't understand exactly what you are simulating. Using your file I still have the envelope below, very similar as my first answer and perfectly working. So, I have attached the complete project here (*.ini, *.dat, *.cal).

Regards,

Didier

env.png (17.21 KiB) Viewed 2822 times

emittance_1.zip

(8.55 KiB) Downloaded 207 times

[Ken]

Thanks Didier,
I tried your input files in a new folder and they worked just like you described. Then I compared them to my input files (in another folder) and could not find any differences, but the problem was still there, i.e. huge emittance in one plane. However, the problem went away when I deleted all other files except the .ini and .dat (structure files) and started over with the matching. So perhaps I had some left-over file in my original folder throwing off the matching routine?

Best regards,
Ken

[Didier]

Dear Ken,

The results are saved in the file (*.cal) to avoid redoing a calculation already done. So, if you want to redo a complete matching, you have to delete this result in the cal file.
I only see this reason to explain what you observe.

Regards,

Didie