[PPT] - Challenges for Polarimetry at the ILC Spin Tracking Studies Moritz PowerPoint Presentation

SLIDE 1

Challenges for Polarimetry at the ILC

Spin Tracking Studies Moritz Beckmann, Jenny List

DESY - FLC

LCWS 2012, Arlington, TX, USA October 25, 2012

SLIDE 2

Introduction: Polarimetry at the ILC

Two laser Compton polarimeters per beam in the beam

delivery system (BDS)

150 m ~1 650 m upstream polarimeter IP downstream polarimeter

Polarimeters measure with 0.25 % systematic uncertainty

(goal)

What happens between polarimeter and IP?

Moritz Beckmann (DESY) LCWS Oct 25, 2012 2/ 17

SLIDE 3

Introduction: Polarimetry at the ILC

Two laser Compton polarimeters per beam in the beam

delivery system (BDS)

150 m ~1 650 m upstream polarimeter IP downstream polarimeter

Polarimeters measure with 0.25 % systematic uncertainty

(goal)

What happens between polarimeter and IP?
In addition: calibration with average polarization from

collision data (up to 0.1 %)

Must understand spin diffusion/depolarization to 0.1 %

Moritz Beckmann (DESY) LCWS Oct 25, 2012 2/ 17

SLIDE 4

Introduction: Simulation Framework

UP IP DP Particle / spin tracking along the BDS Bmad Beam-beam collision GP++/CAIN Data analysis ROOT Polarimeter simulation LCPolMC Polarimeter simulation LCPolMC

UP/DP: up-/downstream polarimeter

Framework could be used with different input also for other machines, e. g. CLIC

Moritz Beckmann (DESY) LCWS Oct 25, 2012 3/ 17

SLIDE 5

Introduction: Principles of Spin Propagation

Spin propagation in electromagnetic fields

is described by T-BMT equation (semi- classical)

Approximation (

B⊥ only) for illustration: spin precession θspin =

g−2

2

· E

m + 1

≈568

·θorbit

B

Moritz Beckmann (DESY) LCWS Oct 25, 2012 4/ 17

SLIDE 6

Introduction: Principles of Spin Propagation

Spin propagation in electromagnetic fields

is described by T-BMT equation (semi- classical)

Approximation (

B⊥ only) for illustration: spin precession θspin =

g−2

2

· E

m + 1

≈568

·θorbit

B

Polarization vector

P =   Px Py Pz   with polarization

P
Moritz Beckmann (DESY)

LCWS Oct 25, 2012 4/ 17

SLIDE 7

Introduction: ILC Beam Delivery System

Latest available beamline design (SB2009 Nov10 lattice)

Moritz Beckmann (DESY) LCWS Oct 25, 2012 5/ 17

SLIDE 8

Spin Propagation through BDS (Idealized Lattice)

distance s along BDS [m]

500 1000 1500 2000 2500 3000 3500

polarization

0.8
0.6
0.4
0.2

0.2 0.4 0.6 0.8

UP IP DP

e

z

P | P |

UP/DP: up-/downstream polarimeter

1000 runs with random bunches, 10 000 sim. particles each
Drawn: median ± 1σ
Perfect magnet alignment, no collision effects

Moritz Beckmann (DESY) LCWS Oct 25, 2012 6/ 17

SLIDE 9

Spin Fan-Out

distance s along BDS [m]

3400 3450 3500 3550

polarization

0.7992 0.7994 0.7996 0.7998 0.8

]

3

relative change [10

1
0.5

IP DP

e

z

P | P |

Only minor spin fan-out in quadrupoles

Moritz Beckmann (DESY) LCWS Oct 25, 2012 7/ 17

SLIDE 10

Spin Fan-Out

distance s along BDS [m]

3400 3450 3500 3550

polarization

0.7992 0.7994 0.7996 0.7998 0.8

]

3

relative change [10

1
0.5

IP DP

e

z

P | P |

Only minor spin fan-out in quadrupoles

P P' P

Moritz Beckmann (DESY) LCWS Oct 25, 2012 7/ 17

SLIDE 11

Collision Effects

Simulation of Collision Effects (GP++):

T-BMT precession: deflection from colliding bunch

(∼ 10−4 rad)

Sokolov-Ternov: spin flip by emission of beamstrahlung

Moritz Beckmann (DESY) LCWS Oct 25, 2012 8/ 17

SLIDE 12

Collision Effects: Energy Loss

Energy loss by beamstrahlung:
0.6
0.4
0.2

# of entries

1 10

2

10

3

10

4

10

5

10

6

10

7

10

particle energy [GeV]

100 150 200 250

before collision after collision

Spin precession ∝ E

⇒ Spin fan-out due to energy spread

Moritz Beckmann (DESY) LCWS Oct 25, 2012 9/ 17

SLIDE 13

Collision Effects: Energy Loss vs. Laser-Spot

Laser-spot size at Compton IP only ∼ 0.1 − 1 mm
chicane ⇒ dispersion (black: reference particle)
Without collision: 0.124 % beam energy spread

Entire beam within laser-spot

0.005

0.005

0.15
0.1
0.05

0.05 0.1 0.15

3

10 ×

1 10

2

10

3

10

4

10

5

10

no collision

particle energy [GeV]

249 250 251

vertical particle position y [mm]

0.1

0.1 Moritz Beckmann (DESY) LCWS Oct 25, 2012 10/ 17

SLIDE 14

Collision Effects: Energy Loss vs. Laser-Spot

Laser-spot size at Compton IP only ∼ 0.1 − 1 mm
chicane ⇒ dispersion (black: reference particle)
After collision: Off-energy particles evade laser-spot
Downstream polarimeter needs detailed investigation

(energy and polarization correlated!)

0.4
0.3
0.2
0.1
0.02
0.015
0.01
0.005

0.005 0.01 0.015 0.02

1 10

2

10

3

10

4

10

5

10

6

10

after collision

particle energy [GeV]

150 200 250

vertical particle position y [mm]

20
10

10 20 Moritz Beckmann (DESY) LCWS Oct 25, 2012 11/ 17

SLIDE 15

Collision Effects: Spin Propagation

Collisions, but still perfect alignment
Crossing angle 14 mrad, bunches crabbed

distance s along BDS [m]

3400 3450 3500 3550

z

long. polarization P

0.77 0.78 0.79 0.8

]

3

relative change [10

40
30
20
10

IP DP

e

no collision lumi-weighted after collision measurable

Much stronger spin fan-out
Polarization within 0.1 mm laser-spot different: “measureable”

Moritz Beckmann (DESY) LCWS Oct 25, 2012 12/ 17

SLIDE 16

Collision Effects: Spin Propagation

10 20 30

z

longitudinal polarization P

0.77 0.78 0.79 0.8

]

3

relative change [10

40
30
20
10

no coll. full lumi

UP IP before collision IP lumi-weighted IP after collision DP DP measurable (r=0.1, 0.2, 0.5, 1 mm)

Moritz Beckmann (DESY) LCWS Oct 25, 2012 13/ 17

SLIDE 17

Collision Effects: Spin Propagation

10 20 30

z

longitudinal polarization P

0.77 0.78 0.79 0.8

]

3

relative change [10

40
30
20
10

no coll. full lumi

UP IP before collision IP lumi-weighted IP after collision DP DP measurable (r=0.1, 0.2, 0.5, 1 mm)

2.5% ≈ 0.3% ≈

What does the measurement tell us about the

polarization at the IP?? ∆Pz ∼ 2.5 %

Can we trust the simulation to calculate back?

More details to come: detector magnets, misalignments

Uncertainty in DP laser-spot size/position

⇒ ∆Pz = O(0.1 %)

Moritz Beckmann (DESY) LCWS Oct 25, 2012 14/ 17

SLIDE 18

Collision Effects: Spin Propagation

10 20 30

z

longitudinal polarization P

0.77 0.78 0.79 0.8

]

3

relative change [10

40
30
20
10

no coll. low lumi full lumi

UP IP before collision IP lumi-weighted IP after collision DP DP measurable (r=0.1, 0.2, 0.5, 1 mm)

Low luminosity sample (switched off bunch crabbing):

Collision effects and also their consequences reduced
Downstream measurement less affected by collision effects

and less dependent on laser-spot size/position

Moritz Beckmann (DESY) LCWS Oct 25, 2012 15/ 17

SLIDE 19

Conclusion

A spin tracking framework for high energy linear colliders

including collision effects has been set up

ILC: understanding of polarization to permille-level required
Precision goals for upstream measurement seem achievable

Moritz Beckmann (DESY) LCWS Oct 25, 2012 16/ 17

SLIDE 20

Conclusion

A spin tracking framework for high energy linear colliders

including collision effects has been set up

ILC: understanding of polarization to permille-level required
Precision goals for upstream measurement seem achievable
Downstream polarimeter struggles fiercely with collision

effects:

High-precision simulation including all effects required at

high luminosities to obtain polarization at IP from data

Measurement highly sensitive to size/position of laser-spot
Idea: determine lumi-weighted polarization rather/also from

upstream polarimeter and luminosity measurement?

Moritz Beckmann (DESY) LCWS Oct 25, 2012 16/ 17

SLIDE 21

Conclusion

A spin tracking framework for high energy linear colliders

including collision effects has been set up

ILC: understanding of polarization to permille-level required
Precision goals for upstream measurement seem achievable
Downstream polarimeter struggles fiercely with collision

effects:

High-precision simulation including all effects required at

high luminosities to obtain polarization at IP from data

Measurement highly sensitive to size/position of laser-spot
Idea: determine lumi-weighted polarization rather/also from

upstream polarimeter and luminosity measurement?

Downstream polarimeter needed nevertheless:
Measure depolarization without collision effects / calibrate UP
Measure additional depolarization at low luminosities to test

simulations

Moritz Beckmann (DESY) LCWS Oct 25, 2012 16/ 17

SLIDE 22

Thanks for your attention!

Thanks for support and useful discussions to:

David Sagan (Cornell U.)
Deepa Angal-Kalinin (Daresbury Lab.)
Anthony Hartin, Mathias Vogt, Nick Walker (DESY)
Andrei Seryi (JAI)
Kenneth Moffeit, Yuri Nosochkov, Michael Woods (SLAC)
Jeff Smith (formerly SLAC)
und many others...

Moritz Beckmann (DESY) LCWS Oct 25, 2012 17/ 17

SLIDE 23

Backup slides

Moritz Beckmann (DESY) LCWS Oct 25, 2012 18/ 17

SLIDE 24

Compton Polarimeters: Principles

Compton scattering with polarized laser:

∼ 1500 electrons per bunch

Measure energy spectrum of scattered electrons
Energy distribution → spatial distribution
Cherenkov gas detector counts electrons per channel

Moritz Beckmann (DESY) LCWS Oct 25, 2012 19/ 17

SLIDE 25

Compton Polarimeters: Principles

50 100 150 200 250 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

σ Compton [mbarn / GeV]

Compton−scattered Electron Energy [GeV] λ Pe = +1

(same)

λ Pe = −1

(opposite)

σCompton depends on polarization (laser × beam)
Measure asymmetry and compare to analyzing power

(predicted asymmetry for 100 % polarization)

Moritz Beckmann (DESY) LCWS Oct 25, 2012 20/ 17

SLIDE 26

Compton Polarimeters: Systematic Errors

Goal: relative systematic error on measurement < 0.25 % (SLC polarimeter: 0.5 %)

Detector linearity: contribution of ∼ 0.1 − 0.2 % (goal)

Prototype tests ongoing . . .

Laser polarization: ∼ 0.1 %

Moritz Beckmann (DESY) LCWS Oct 25, 2012 21/ 17

SLIDE 27

Compton Polarimeters: Systematic Errors

Goal: relative systematic error on measurement < 0.25 % (SLC polarimeter: 0.5 %)

Detector linearity: contribution of ∼ 0.1 − 0.2 % (goal)

Prototype tests ongoing . . .

Laser polarization: ∼ 0.1 %
Analyzing power: ∼ 0.1 % (UP: , DP: ?)
Detector alignment: can be determined from data ()

0.5 mm precision sufficient

Alignment of magnets negligible compared to detector

Field inhomogeneities? to be investigated

Disrupted electron beam at downstream polarimeter:
Dependence on laser-spot size and position: ??
Beam energy spread no concern for small laser-spot sizes

thanks to dispersion

Moritz Beckmann (DESY) LCWS Oct 25, 2012 21/ 17

SLIDE 28

Misalignments

Every element is shifted/rotated randomly in/about all

directions/axes

Gaussian-distributed random numbers, σ = 10 µm/µrad
Static and time-dependent misalignments

Moritz Beckmann (DESY) LCWS Oct 25, 2012 22/ 17

SLIDE 29

Misalignments

Every element is shifted/rotated randomly in/about all

directions/axes

Gaussian-distributed random numbers, σ = 10 µm/µrad
Static and time-dependent misalignments
Simplified orbit correction with kicker magnets and fast

feedback at IP

Moritz Beckmann (DESY) LCWS Oct 25, 2012 22/ 17

SLIDE 30

Misalignments: Correction with Kicker Magnets BPM kicker magnet beam design orbit

∼ 40 kicker magnets and many more Beam Position Monitors

spread over BDS

Calculate required kicks from measurements (SVD)
Automatic correction of spin alignment as well?

Moritz Beckmann (DESY) LCWS Oct 25, 2012 23/ 17

SLIDE 31

Misalignments: Orbit Correction Strategy

Strategy here:

Interested in effects of kicks on polarization,

not in sophisticated correction algorithm

Get orbit corrected somehow with kickers such that
beam does no go lost
approximations (small coordinates) still hold

Moritz Beckmann (DESY) LCWS Oct 25, 2012 24/ 17

SLIDE 32

Misalignments: Orbit Correction Strategy

Strategy here:

Interested in effects of kicks on polarization,

not in sophisticated correction algorithm

Get orbit corrected somehow with kickers such that
beam does no go lost
approximations (small coordinates) still hold
Fake correction at IP: shift and rotate bunch coordinates to

0.1 σ precision (goal), adjust beam size

Moritz Beckmann (DESY) LCWS Oct 25, 2012 24/ 17

SLIDE 33

Misalignments: Spin Propagation

Misalignments reduce luminosity ⇒ less collision effects
Measured polarization depends on laser-spot size and position

10 20 30 40

z

longitudinal polarization P

0.77 0.78 0.79 0.8

]

3

relative change [10

40
30
20
10

ideal FFB µ 5 µ 10

e

UP IP before collision IP lumi-weighted IP after collision DP DP measurable (r=0.1, 0.5, 1 mm)

Moritz Beckmann (DESY) LCWS Oct 25, 2012 25/ 17

SLIDE 34

Collision Effects: Energy Loss vs. Laser-Spot

Laser-spot size at Compton IP only ∼ 100 µm − 1 mm
chicane ⇒ dispersion (black: reference particle)
After collision, bunch crabbing switched off
0.4
0.3
0.2
0.1
0.02
0.015
0.01
0.005

0.005 0.01 0.015 0.02

1 10

2

10

3

10

4

10

5

10

6

10

after collision

particle energy [GeV]

150 200 250

vertical particle position y [mm]

20
10

10 20 Moritz Beckmann (DESY) LCWS Oct 25, 2012 26/ 17

SLIDE 35

Collision Effects: Spin Propagation (Polarization)

Total polarization affected likewise
Polarization decrease in chicanes: fan-out due to energy

spread

distance s along BDS [m]

3400 3450 3500 3550

| P total polarization |

0.77 0.78 0.79 0.8

]

3

relative change [10

40
30
20
10

IP DP

e

Moritz Beckmann (DESY) LCWS Oct 25, 2012 27/ 17

SLIDE 36

Collision Effects: Spin Propagation (Positron Beam)

10 20 30

z

longitudinal polarization P

0.29 0.295 0.3

]

3

relative change [10

40
30
20
10

no coll. low lumi full lumi

UP IP before collision IP lumi-weighted IP after collision DP DP measurable (r=0.1, 0.2, 0.5, 1 mm)

Moritz Beckmann (DESY) LCWS Oct 25, 2012 28/ 17

SLIDE 37

Collision & Misalignments: Downstream Polarimeter Measurement

0.4
0.3
0.2
0.1
0.02
0.015
0.01
0.005

0.005 0.01 0.015 0.02

1 10

2

10

3

10

4

10

5

10

after collision

particle energy [GeV]

150 200 250

vertical particle position y [mm]

20
10

10 20

Moritz Beckmann (DESY) LCWS Oct 25, 2012 29/ 17

SLIDE 38

Luminosity

Design values 1.8(1.5) · 1038 m−2s−1 (without waist shift)
Need to improve tuning of grid parameters in GP++
Does not change statement of this talk (effects might just

get stronger for higher luminosities)

2000 3000 4000 5000 ×

# of entries

20 40 60 80 100 120 140 160 180 200 0.1 0.2 0.3 (a) 0.008 ± : 0.165 σ ± µ 12 14 16 18

33

10 × 20 40 60 80 100 120 140 160 180 200

]

1

s

2

m ⋅

38

luminosity [10

e

+

e

0.8 0.9 1 1.1 1.2 (b) 0.010 ± : 0.862 σ ± µ 0.012 ± : 0.833 σ ± µ

CC on CC off CC on, no WS

Moritz Beckmann (DESY) LCWS Oct 25, 2012 30/ 17

SLIDE 39

Polarization correction by angle measurement?

Detector solenoid and anti-DID
θr: angular spread within bunch
Solenoid field invalidates “B⊥ only” approximation
Still sharp value for b (ϑpol = b · ϑbunch) due to ideal

conditions (no misalignments)

3410 3412 3414 0.05 0.1 0.15 10 ×

rad] µ angle [

50 100 150

entrance solenoid IP

r

θ

bunch

ϑ b

3410 3412 3414

b [1]

200 400 600

distance to IP [m]

4
3
2
1

γ b=a

Moritz Beckmann (DESY) LCWS Oct 25, 2012 31/ 17

SLIDE 40

Polarization correction by angle measurement?

This plot without detector magnets
Small misalignments (2µm / 2µrad) make correction for

incident angle impossible, since there is no more simple correlation between angles of bunch and polarization vector

“Steps” due to correction kickers with zero length

3410 3412 3414 0.02 0.04 0.06 10 ×

rad] µ angle [

20 40 60

entrance solenoid IP

r

θ

bunch

ϑ b

3410 3412 3414

b [1]

1000 2000 3000 4000 5000

distance to IP [m]

4
3
2
1

γ b=a

Moritz Beckmann (DESY) LCWS Oct 25, 2012 32/ 17

SLIDE 41

Polarization

Here: longitudinal polarization Pz

(along beam axis)

Pz = pR − pL ∈ [−1, +1]
Beam with 90% R (and thus 10% L)

→ 80% longitudinal polarization

More general: polarization vector
P =

  Px Py Pz   with polarization

P
Moritz Beckmann (DESY)

LCWS Oct 25, 2012 33/ 17