Elegant vs MAD

11
‘Elegant’ code was used for my previous ATF2-EXT- line simulations. It has a ‘drawback’, which can’t calculate multipoles matrix (e.g., in the coupling correction); it has equivalent solution but is only limited to n3, thus higher order multipoles have to be off, e.g., for the coupling correction. ‘Elegant’ tracking is ok for only QM7 multipoles with either vertical or horizontal shift because multipoles effects dominates by the sextupoles n=3 in this case. Results show that no obvious vertical emittance growth observed after perfect corrections. But for both QM7 and SepA with, both vertical and horizontal offsets, n>3 multipoles also have significant contribution, and thus ‘elegant’ has no solution. But MAD does calculate multipoles matrix n 9, and can process corrections for orbit, dispersion, coupling and Twiss. Elegant vs MAD F. Zhou 07/10/07 Slide 1

description

Elegant vs MAD. - PowerPoint PPT Presentation

Transcript of Elegant vs MAD

Page 1: Elegant vs MAD

• ‘Elegant’ code was used for my previous ATF2-EXT-line simulations. It has a ‘drawback’, which can’t calculate multipoles matrix (e.g., in the coupling correction); it has equivalent solution but is only limited to n3, thus higher order multipoles have to be off, e.g., for the coupling correction.

• ‘Elegant’ tracking is ok for only QM7 multipoles with either vertical or horizontal shift because multipoles effects dominates by the sextupoles n=3 in this case. Results show that no obvious vertical emittance growth observed after perfect corrections.

• But for both QM7 and SepA with, both vertical and horizontal offsets, n>3 multipoles also have significant contribution, and thus ‘elegant’ has no solution.

• But MAD does calculate multipoles matrix n 9, and can process corrections for orbit, dispersion, coupling and Twiss.

Elegant vs MAD

F. Zhou 07/10/07Slide 1

Page 2: Elegant vs MAD

MAD tracking procedures

• Orbit corrections

• Dispersion corrections

• Coupling corrections

• Twiss matching

• Observe beam at the end of EXT line

F. Zhou 07/10/07Slide 2

Page 3: Elegant vs MAD

Nonlinear field

0.01 0.005 0 0.005 0.01

0

0.2

0.4

0.60.65

0.12−

Re BB( ) 10 4−⋅

Re Fit XY( )( ) 10 4−⋅

Im BB( ) 10 4−⋅

Im Fit XY( )( ) 10 4−⋅

6.5 10 3−×8.5− 10 3−× Re XY( ) 10 2−⋅

Ideal position Interesting area

F. Zhou 07/10/07

0.01 0.005 0 0.005 0.01

0.2

0.22

0.24

0.25

0.185

Re BB( ) 104−⋅

Re Fit XY( )( ) 10 4−⋅

7.059 10 3−×7.642− 10 3−× Re XY( ) 10 2−⋅

0.01 0.005 0 0.005 0.01

0.01

0

0.01

0.015

0.015−

Im BB( ) 10 4−⋅

Im Fit XY( )( ) 10 4−⋅

7.059 10 3−×7.642− 10 3−× Re XY( ) 10 2−⋅x (x - 0.0855) (m) x (x - 0.0855) (m)

Ideal position Ideal

position

QM7

SepA SepA

Slide 3

Page 4: Elegant vs MAD

MAD tracking • QM7 multipoles only, either horizontal shift of 4mm or

vertical shift of 0.6mm: no obvious vertical emittance growth observed after all corrections.

• QM7+SepA, either horizontal shift of 4mm or vertical shift of 0.6mm: no obvious vertical emittance growth observed after all corrections.

5 0 51

0

11

1−

newi 3,

0newi 3,

55− newi 2, 0newi 2,,

5 0 51

0

11

1−

newi 3,

0newi 3,

55− newi 2, 0newi 2,,

QM7 only QM7+SepA

y’

y y

y’

Green: perfect beam Blue: test beam

Green: perfect beam Blue: test beam

F. Zhou 07/10/07Slide 4

Page 5: Elegant vs MAD

10 5 0 5 102

0

22

2−

newi 3,

0newi 3,

1010− newi 2, 0newi 2,,

MAD tracking (con’t) • Only QM7 multipoles with both horizontal shift of 4mm

and vertical shift of 0.6mm: no vertical emittance growth observed after all corrections.

• QM7+SepA with both horizontal shift of 4mm and vertical shift of 0.6mm: significant vertical and horizontal emittance growths observed after all corrections.

QM7 only QM7+SepA

5 0 51

0

11

1−

newi 3,

0newi 3,

55− newi 2, 0newi 2,,

y’

y

Green: perfect beam Blue: test beam

y

y’

Green: perfect beam Blue: test beam

F. Zhou 07/10/07Slide 5

Page 6: Elegant vs MAD

10 5 0 5 102

0

2

10 5 0 5 10

0

4

4−

newi 1,

0newi 1,

1010− newi 0, 0newi 0,,10 5 0 5 10

2

0

22

2−

newi 3,

0newi 3,

1010− newi 2, 0newi 2,,5 0 5

10

0

1010

10−

newi 2,

0newi 2,

55− newi 0, 0newi 0,,x

yy’x’

x y

X=4mm and y =0.6mm

10 5 0 5 102

0

2

10 5 0 5 10

0

4

4−

newi 1,

0newi 1,

1010− newi 0, 0newi 0,,

10 5 0 5 102

0

22

2−

newi 3,

0newi 3,

1010− newi 2, 0newi 2,,

5 0 510

0

1010

10−

newi 2,

0newi 2,

55− newi 0, 0newi 0,,x

y

y

y’

x

x’

X=2.6mm and y =0.6mm

Position-dependency

Slide 6 07/10/07F. Zhou

Page 7: Elegant vs MAD

Energy spread-dependency

10 5 0 5 102

0

22

2−

newi 3,

0newi 3,

1010− newi 2, 0newi 2,,

10 5 0 5 102

0

22

2−

newi 3,

0newi 3,

1010− newi 2, 0newi 2,,

E=0.13%

QM7+SepA

y

y’

E=0.08%

QM7+SepA

Green: perfect beam Blue: test beam

y

y’

F. Zhou 07/10/07Slide 7

Page 8: Elegant vs MAD

Orbit corrections using existing correctors

F. Zhou 07/10/07Slide 8

Page 9: Elegant vs MAD

Dispersion correction

F. Zhou 07/10/07Slide 9

Page 10: Elegant vs MAD

Coupling correction, and Twiss matching

• Off-diagonal R-matrix <10^-5

• Twiss matched using EXT last 4 quads

5 0 510

0

1010

10−

newi 2,

0newi 2,

55− newi 0, 0newi 0,,x

y

F. Zhou 07/10/07Slide 10

Page 11: Elegant vs MAD

Summary• MAD can include multipoles n 9, and proceed all

corrections – orbit, dispersion, coupling and Twiss match.• When beam is trapped in either horizontal or vertical

off-center position at QM7 and SepA, no obvious vertical emittance growth observed after all corrections; same results as in elegant code.

• When beam is trapped in both horizontal and vertical off-center position (e.g., x=4mm and y=0.6mm) at QM7 and of course transversely offset at SepA, both horizontal and vertical emittance growths are obviously observed even after all corrections.

• Emittance growth slightly depends on the initial energy spread.

• All simulations are done on ATF2 EXT line; next is to simulate the existing ATF EXT line since both the EXTs are not fully identical.

F. Zhou 07/10/07Slide 11