muse simulations nsf - rutgers physics & astronomyransome/muse/talks/muse... · nsf review...

NSF Review Meeting, NSF, 03/24/14

Steffen Strauch*University of South Carolina

Geant4 Simulationsfor MUSE

*This work is supported in parts by the U.S. National Science Foundation: NSF PHY-1205782

beamCherenkov

GEM3

GEM2

GEM13 planes

of scintillation fibers

2 planes ofstraw chambers

beam-linemonitor

2 walls ofscatteredparticle

scintillators

veto

Outline

1. Geant4 physics model and simulation verification

2. Simulation of detector and beam properties

3. Preliminary study of analysis topics

‣ Particle tracking and scattering-angle reconstruction

‣ Møller/Bhabha scattering

‣ Muon decay in flight

2

Physics: test of ep - scattering cross section

3

(deg)θScattering Angle 0 45 90 135 180

/sr)

2 (

cmΩ

/dσC

ross

Sec

tion

d

-3310

-3210

-3110

-3010

-2910

-2810

-2710

-2610 = 150 MeV/ceGeant4, p

MottDiracElastic ep scattering

ep elastic cross section reasonably well described in Geant4 at 150 MeV/c and 210 MeV/c.

2 cm lH2 target

150 MeV/c electrons

6 x 109 events

FTFP_BERT

1

10

210

310

410

510

610

(deg)θScattering Angle 0 45 90 135

(MeV

/c)

e'M

omen

tum

p

0

50

100

150

= 150 MeV/cep ee→Moller / ee

ep→Mott / ep

Momentum-angle correlation approximated for forward angle elastic ep scattering.

Physics: Mott and Møller scattering

4

Mott

Møller

Pulse-height distribution: electrons and muons

ADC0 1000 2000 3000 4000

Cou

nts

1

10

210

310

410

, 115 MeV/c+Beam: e

Run: #934Data (geoadc)

g4PSI

ADC0 1000 2000 3000 4000

Cou

nts

1

10

210

310

410

, 153 MeV/c+Beam: e


g4PSI

Data (Summer 2013):

‣ 5 cm x 5 cm x 50 cm scintillator with 2 PMTs.

Geant4 simulation

‣ Universal energy scale of the MC data for all runs and all particle types.

‣ Histograms normalized for equal number of entries.

‣ Adjustment of beam parameters in simulation to fit pulse-height spectrum. ADC

0 1000 2000 3000 4000

Cou

nts

1

10

210

310

, 153 MeV/c+µBeam:


g4PSI

5

115 MeV/c 153 MeV/c

153 MeV/c

0

20

40

60

310×

SC Wall Front Paddle5 10 15

SC W

all B

ack

Padd

le

10

20

2 * Nα - 1N-5 0 5

Cou

nts

per I

ncid

ent P

artic

le

-510

-410

-310

-210

-110

1

Maximum-signal scintillatorsClosest above-threshold scintillators

-Particle: e

= 115 MeV/cin

p

= 2.0 MeVthE

Scattered particle scintillator - paddle correlation

• Front- and back-wall hits are highly correlated if the event originated in the target.

• Imposing a hardware coincidence with the expected correlation does reduce accidental background.

• Measured correlation will help validate the simulation.

6

Detector efficiency of TOF wall scintillators

• Efficiency of scattered particle scintillators is ≳ 99%

• Small (< 1%) effects which will be corrected

- Annihilation of positrons in the front wall

- Multiple scattering of low-momentum muons (with directional cut)

7

Particle Scattering Angle (deg)0 50 100

Effic

ienc

y

0.94

0.96

0.98

1.00

+Particle: e = 115 MeV/c

inp

= 2.0 MeVthE

one plane only

two-plane coincidence

directional cut

Particle Scattering Angle (deg)0 50 100

Effic

ienc

y

0.94

0.96

0.98

1.00

-µParticle: = 115 MeV/c

inp

= 2.0 MeVthE

one plane only

two-plane coincidence

directional cut

Multiplicity of scintillator-paddle hits

• In most cases one or two bars trigger per scintillator wall.

• Hit multiplicities can be used to validate the simulation.

8

)th

Number of Paddles (E > E0 5 10

Cou

nts

per I

ncid

ent P

artic

le

-410

-310

-210

-110

1-µParticle:

= 115 MeV/cin

p = 2.0 MeVthE

Front SC wallBack SC wall

)th

Number of Paddles (E > E0 5 10

Cou

nts

per I

ncid

ent P

artic

le

-410

-310

-210

-110

1-µParticle:

= 210 MeV/cin

p = 2.0 MeVthE

Front SC wallBack SC wall

Typically low multiplicity Highest multiplicity

Veto detector essential to reduce trigger rate

9

Trigger-rate reduction after veto:

‣ electrons ≈ 1/3‣ muons ≈ 1/2

Veto

GEM3

Target

16 cm

elec

tron

bea

m, p

= 153

MeV

/c

1

10

210

310

410

x-position at z = -17 cm (cm)-40 -20 0 20 40

y-po

sitio

n at

z =

-17

cm (

cm)

-40

-20

0

20

40

1

10

210

310

410

x-position at z = -17 cm (cm)-40 -20 0 20 40

y-po

sitio

n at

z =

-17

cm (

cm)

-40

-20

0

20

40

Beam

Beam-momentum distributions are an important input to radiative corrections of the scattering cross section.

Simulation helps to understand the beam momentum distribution

10

0)/p

0Relative particle momentum change (p - p

-0.20 -0.15 -0.10 -0.05 -0.00 0.05

Cou

nts

10

210

310 p = 115 MeV/c Target entrance-e Target exit-e

0)/p

0Relative particle momentum change (p - p

-0.20 -0.15 -0.10 -0.05 -0.00 0.05

Cou

nts

1

10

210

310 p = 115 MeV/c Target entrance-R Target exit-R

Electron beamMuon beam

Incident electron beam with momentum spread, Δp/p = ±1.5%

p(e,e’) scattering eventinside the target

153 MeV/celectron

Ionization andBremsstrahlung

in the thick rearscintillator wall

11

Particle-track reconstruction using hit information from

‣ hits in GEM detectors (incident) and

‣ straw chambers (scattered)

Tracking of simulated events

Vertex Resolution 𝝈 ≲ 6 mm

12X (cm)ΔVertex

-2 -1 0 1 2

Cou

nts

0

50

100

0.00 cm± = 0.00 µ 0.00 cm± = 0.10 σ

Beam-µ

p = 210 MeV/c

Z (cm)ΔVertex -2 -1 0 1 2

Cou

nts

0

20

40

0.01 cm± = -0.04 µ 0.01 cm± = 0.30 σ

Beam-µ

p = 210 MeV/c

Z (cm)ΔVertex -2 -1 0 1 2

Cou

nts

0

20

40

60 0.01 cm± = -0.05 µ 0.01 cm± = 0.61 σ

Beam-µ

p = 115 MeV/c

X (cm)ΔVertex -2 -1 0 1 2

Cou

nts

0

50

100

150 0.00 cm± = 0.00 µ 0.00 cm± = 0.24 σ

Beam-µ

p = 115 MeV/c

Reaction z-Vertex (cm)-20 0 20

Cou

nts

0

100

200

300 Beam-µ

p = 153 MeV/c

Vertex reconstruction of muon events

13

scatteringchamber exit

targetscatteringchamber entrance

GEM3detector

Reconstruction of muon scattering vertex is of sufficiently high resolution to cleanly separate scattering events within the target from other sources.

Reaction z-Vertex (cm)-20 0 20

Cou

nts

0

100

200 Beam-e

p = 153 MeV/c

Vertex reconstruction for electrons

14

scatteringchamber exit

targetscatteringchamber entrance

GEM3detector

Radiation processes affect vertex reconstruction of electrons. Small background can be studied and corrected for with empty target measurements and simulations.

Sufficient resolution of the scattering angle reconstruction

Reconstruction of the scattering angle for simulated events with reconstructed reaction vertex within the target volume and 𝜽 > 20˚

15

(mrad)θΔScattering Angle -50 0 50

Cou

nts

0

50

0.25 mrad± = 0.38 µ 0.17 mrad± = 12.34 σ

Beam-µ

p = 153 MeV/c

(mrad)θΔScattering Angle -50 0 50

Cou

nts

0

50

0.27 mrad± = 0.33 µ 0.19 mrad± = 11.76 σ

Beam-e

p = 153 MeV/c

Reconstruction of the scattering angle is unbiased and of sufficient resolution.

Δθ = θreconstructed −θsimulated

Møller scattering background

Signatures

‣ Scattered Møller electron forward peaked

‣ Scattered electron has low momentum

‣ Forward going high-momentum beam electron

Suppression

‣ Beam-monitor scintillator as Møller Veto

‣ Scintillator Wall bar combination

16

ee ￫ ee scattering eventinside the target

Møller electron

Forward goinghigh-momentum

electron

Beam-monitorscintillator

as Møller Veto

153 MeV/celectron

Scattered Particle Momentum (MeV/c)0 50 100 150

Cou

nts

0

500

1000

Beam-e

p = 153 MeV/c

Electron momentum distribution dominated by Møller events

17

Møller events

Mott eventsScattering events

of interest

All events have a reaction vertex within the target volume.

Experiment does not measure the momentum of the scattered particle, E(SC2) > 5 MeV.


Cou

nts

0

500

1000

Beam-e

p = 153 MeV/c

Full suppression of Møller events with beam monitor veto

18

Møller eventscompletely suppressed

with a veto signalfrom the downstream

beam monitor

Mott eventsScattering events

of interest

All events have a reaction vertex within the target volume. Events marked in red do not have a hit in the downstream beam monitor.

Experiment does not measure the momentum of the scattered particle, E(SC2) > 5 MeV.


Cou

nts

1

10

210

310 Beam-e

p = 153 MeV/c

19

Møller events

Mott eventsRadiative tail

Simulation gives input to calculations of radiative corrections

simulation determines detection threshold, which is an input to the calculations of radiative corrections

Muon decay in flight

20

Suppression of background from muon decay‣ Target vertex cut‣ Time of flight

153 MeV/c muon beam

electron

µ− → e− +νe +νµ

muon decayvertex

neutrinos

tSC

tUSBS

USBS

SC

Δt = tSC −

SCc

⎛⎝⎜

⎞⎠⎟ − tUSBS −

USBSβµc

⎛

⎝⎜⎞

⎠⎟

Vertex-time difference from path lengths and measured times

assuming electronafter muon decay, βe = 1

(ns)BC - tSCVertex Time Difference t0 1 2

Cou

nts

0

50

100

150TEST153 MeV/c scatteringµ

decayµ


Cou

nts

0

100


decayµ


Cou

nts

0

50

100


decayµ

Muon decay in flight — vertex-time differenceDetection of muon decay by path-length corrected time-of-flight:

‣ time-of-flight walls (tSC, 𝝈 = 50 ps)

‣ upstream beam Cherenkov (tBC, 𝝈 = 100 ps)

21

6𝝈

4𝝈 2.2𝝈most of theseevents are μ

scattering events

High fraction of scattering events are detectable without significant background contribution

22

Vertex-time difference cut on scattering event distributionVertex-time difference cut on scattering event distribution

μ momentum -σ < tSC - tBC -3σ < tSC - tBC

115 MeV/c Signal 85(2)%Background 0.0(0.0)%

Signal 100(2)%Background 0.0(0.0)%


Signal 100(2)%“Background” 1.6(0.2)%


Signal 100(4)%Background 63(1)%

• The time-of-flight technique can reduce the muon-decay in flight background to < 4% prior to any correction of subtraction.

• In practice, vary cut to optimize final uncertainties.

Direct measurement of the muon decay in flight background

In situ measurement of muon decay-in-flight background from events upstream & downstream of the target.

23

target z40 mm20 mm 20 mm

Muon decay flagged in simulation Muon decay distribution measured


Cou

nts

0

50

100


decayµ


Cou

nts

0

50

100

150TEST153 MeV/ctotalbackground

Summary

• Key physics processes are included in the simulation code.

• Simulation of detector components helped with the optimization of the experimental setup.

• Particle vertex and scattering-angel reconstruction is sufficient.

• Møller/Bhabha scattering are efficiently suppressed with an event-veto from the beam-line monitor detector.

• Muon decay-in-flight background can be removed with time-of-flight measurements, direct measurements of the background up- and downstream of the target, or through empty target measurements.

24

muse simulations nsf - rutgers physics & astronomyransome/muse/talks/muse... · nsf review...

Documents