MEIC Simulation and Reconstruction Group Report Tanja Horn 1 Tanja Horn, CUA Colloquium Tanja Horn, MEIC Simulation Group Report, EIC Detector Workshop

MEIC Simulation and Reconstruction Group Report

Tanja Horn

1Tanja Horn, CUA ColloquiumTanja Horn, MEIC Simulation Group Report, EIC Detector Workshop 2010

• Gluon and sea quark (transverse) imaging of the nucleon

• Nucleon Spin (G vs. ln(Q2), transverse momentum)

• Nuclei in QCD (gluons in nuclei, quark/gluon energy loss)

• QCD Vacuum and Hadron Structure and Creation

• Electroweak Physics

EIC@JLab Science SummaryEIC@JLab Science Summary

Energies s luminosity

EIC@Jlab Up to 11 x 60 150-2650 Few x 1034

Future option Up to 11 x 250 11000 1035

Tanja Horn, MEIC Simulation Group Report, EIC Detector Workshop 2010

2

Simulation WorkSimulation Work


• Nucleon spin

– FL(x, Q2)

– g1(x, Q2), ΔG/G, gp1-gn

1

– Gluon/quark transverse imaging (exclusive diffractive channels)

• Flavor Decomposition and transverse structure– Quark polarization Δq(x): Δu/u, Δd/d, Δs/s, Δu/u, Δd/d, Δs/s, Δu-Δd

– Transverse sea quark imaging (exclusive non-diffractive channels)

– Transverse parton momenta (Mulders, Sivers, Transversity)

• Nuclear Medium

– FLA(x, Q2)

– pT2

3

Activities of the Simulation Working Activities of the Simulation Working GroupGroup


• Event Generators (standardized format)– Rate predictions including simulations of the detector restrictions

– Input for detector design

o Momentum and angular distributions for various particles

• Fast MC– Input: resolution function

• GEANT MC– Based on the CLAS12 simulation package GEMC

• Event Reconstruction/Tracking (for GEANT data)

4

Exclusive Processes: EIC SimulationsExclusive Processes: EIC Simulations

Diffractive Channels

– Data experience from HERA, COMPASS

o γp(DVCS), ρ°p, φp, J/Ψp

– DVCS simulations [A. Sandacz 06/07]

Non-diffractive Channels

– New territory for a collider!

– Much more demanding in luminosity

– Feasibility studies: π+n, π°p, K+Λ

5Tanja Horn, MEIC Simulation Group Report, EIC Detector Workshop 2010

• J/Ψ rate estimate uses a model based on a photoproduction parameterization

– Branching ratio of 0.06 into e+e- (or μ+μ-) has been included in rate estimates

Models and rate estimates: Models and rate estimates: J/J/ΨΨ and and DVCSDVCS

[Sandacz, Hyde, Weiss ’06/07+]

• Gluon size directly probed by J/Ψ and φ production (Q2>10 GeV2)

– Require full t-distribution Fourier– Powerlike at |t|>1GeV2?

• Do singlet quarks and gluons have the same transverse distribution?

Gluon size from J/Ψ, singlet quark size from DVCS


6

Exclusive Generators: Exclusive Generators: ππ++/K/K++

e p π+ n

e p K+ Λ

[T. Horn with summer student: D. Cooper ’08]


7

Pushes for high luminosity ~1034 and lower and more symmetric energies

• Simulation for charged + production, assuming 100 days at a luminosity of 1034, with 5 on 50 GeV (s = 1000)

• Pion cross section models:– Ch. Weiss: Regge model– T. Horn: π+ empirical parameterization

• Kaon cross section model:– T. Horn: K+ empirical parameterization

based on DESY, Cornell, JLab data

• Do strange and non-strange sea quarks have the same spatial distribution?

Exclusive Generators: Exclusive Generators: φφ and and ρρ

8Tanja Horn, CUA Colloquium

• Simulation generates φ and ρ events w/:– Φ: parameterization of σL based on a

calculation by Kroll-Goloskokov including a branching ratio of 0.492 into K+K- in rate estimate

– ρ: parameterization by A. Sandacz based on NMC data, valid for Q2>1 GeV2

[Horn, Weiss 09]

Transverse spin

x

slower

quarks move faster

x

Asy

mm

etry

• Deformation of transverse distribution by transverse polarization of nucleon

– Helicity flip GPD E, cf. Pauli ff [M. Burkhardt]

• EIC: exclusive ρ and φ production with transversely polarized beam

– Excellent statistics at Q2>10 GeV2

– Transverse polarization natural for collider


8

4 on 250 GeV4 on 50 GeV

diff

ract

ive

DIS

• Both processes produce high-momentum mesons at small angles

• For exclusive reactions, this constitutes our background

Diffractive and SIDIS (TMDs)[W. Foreman 09]


9


[Horn 08+]recoil baryonsscattered electronsmesons

4 on

250

GeV

4 on

30

GeV

t/t ~ t/Ep

Θ~√t/Ep

PID challenging

very high momenta

electrons in central barrel, but p different

0.2° - 0.45°

0.2° - 2.5°

ep → e'π+n

Exclusive light meson kinematics

10


recoil baryonsscattered electronsmesons

no Q

2 cut

Q2 >

10

GeV

2

t-distribution unaffected

forward mesons: low Q2, high p

low-Q2 electrons in electron endcap

high-Q2 electrons in central barrel:

1-2 < p < 4 GeV

mesons in central barrel:

2 < p < 4 GeV

ep → e'π+n

Beam divergence

Θ~1/√β*

Low (J/Ψ) vs. high Q2 (light mesons) – 4 on 30 [Horn 08+]

11

DES at higher electron energies4 on 30 5 on 50 10 on 50

• With 12 GeV CEBAF, EIC@JLab has the option of using higher electron energies– DIRC no longer sufficient for π/K separation

• DIRC needs to be complemented by gas Cherenkov or replaced by dual radiator RICH to push the limit above 4 GeV

Mo

me

ntu

m (

Ge

V/c

)

Lab Scattering angle (deg) Lab Scattering angle (deg) Lab Scattering angle (deg)


12

Decaying particles: Decaying particles: π°π°

Tanja Horn, CUA Colloquium

[T. Horn with summer student: K. Henderson]

1° → 35mm / 2m

• Opening angle is small and requires fine calorimeter granularity

– JLab/BigCal: 38x38mm, H1 forward calorimeter: 35x35mm

Op

enin

g A

ng

le

(deg

)

π° Lab Angle (deg)

• Separating the π° decay photons is getting more difficult as the energy increases, but pion momenta are low at high Q2

Similar to π+, but photon opening angle places a constraint on the calorimetry

π° γγ • Monte Carlo generates π° events with subsequent decay

– Cross section model based on a parameterization of NMC data

π° Lab Angle (deg)

Op

enin

g A

ng

le

(deg

)

3 on 30

5 on 50

10 on 250


13


Decaying particles: Decaying particles: ρ°ρ°

14

T. Horn with summer student B. Pollack

e p ρ° p

ρ° π+π-

ρ°

14

Q2>10

Q2>10

Q2>10, x<0.1

Q2>10, x<0.1

15

• If DIRC is used, configuration on left side may need adjustment to make space for readout

Central detector layout

ions

electrons

EM

Cal

orim

eter

Had

ron

Cal

orim

eter

Muo

n D

etec

tor

EM

Cal

orim

eter

Solenoid yoke + Hadronic Calorimeter

Solenoid yoke + Muon DetectorTOF

HT

CC

RIC

H

LTCC / RICH

Tracking

5 m

• TOF (5-10 cm)

• DIRC (10 cm) ?


15

Focus on solenoid

Solenoid Fields - Overview

Experiment Central Field Length Inner Diameter

ZEUS 1.8 T 2.8 m 0.86 m

H1 1.2T 5.0 m 5.8 m

BABAR 1.5T 3.46 m 1.4 m

BELLE 1.5T 3.0 m 1.7 m

GlueX 2.0T 3.5 m 1.85 m

ATLAS 2.0T 5.3 m 2.44 m

CMS 4.0T 13.0 m 5.9 m

PANDA(*design) 2.0T 2.75m 1.62 m

CLAS12(*design) 5.0T 1.19 m 0.96 m

Conclusion: ~4-5 Tesla fields, with length scale ~ inner diameter scale o.k.


16

General Solenoid Field

D B

L

θ0

• BT = B sin θ (from v x B)

• pT = p sin θ

Initial solenoid: B=4T, L=5m, D=2.5m

• θ0 = tan-1(x/L)

• L’ = (L/2)/cosθ, θ<θ0

L’ = (x/2)/sinθ, θ>θ0


17

Formulas

..2cos

0136.0

3.0

1

p

plr

T

nL

z

B

4

720

'3.0

p

p

p2

nLB

r

T

• B=central field (T)

• σrφ=position resolution (m)

• L’=length of transverse path through field (m)

• N=number of measurements

• z = charge of particle

• L = total track length through detector (m)

• γ= angle of incidence w.r.t. normal of detector plane

• nr.l. = number of radiation lengths in detector

msc

intr

Assumptions:

• circular detectors around interaction point

• nr.l. = 0.03 (from Hall D CDC)


18

dp/p angular dependence

Can improve resolution at forward angles by offsetting IP

p = 50 GeV p = 5 GeV


19

Multiple scattering contribution


Multiple scattering contribution dominant at small angles (due to BT term in denominator) and small momenta


20

“Easier” Solenoid Field – 2T vs. 4T?

• Intrinsic contribution ~ 1/B• Multiple scattering contribution ~ 1/B


B=2T

B=4T


21

Include dipole field


As expected, substantially improves resolutions at small angles


22

GEMC (GEant4 MonteCarlo)


23

• EIC GEANT MC will be based on GEMC [M. Ungaro]

– Cross section model based on a parameterization of NMC data

• GEMC can retain a very high degree of commonality between its CLAS12 and EIC implementations

– Relies on external field map and geometry services implemented as queries to MySQL databases

– The code could be maintained for both applications.

SVT

CTOF

Beamline

DC

Run 27433

(October 2014)

Tilts, Displacements

RUN

• Geometry server is accessed through a series of PERL scripts– Further information:

https://eic.jlab.org/internal/images/f/f9/Gemc_overview_howto.pptx

EIC GEANT MC


24

• First iteration: understand rate distributions in various parts of the detector– Compare with simple background studies [J. Castilow 09]

Solenoid yoke + Hadronic Calorimeter

Solenoid yoke + Muon Detector

Tracking

5 m

• Implementation will be done in several steps with increasing levels of sophistication

– Cartoon serves as guideline for the implementation

EIC@JLab Simulation and Reconstruction Working Group

• The main goals of the simulation working group is to provide– Event generators for various processes

– Fast Monte Carlo to explore acceptance and resolution requirements

– GEANT4 based Monte Carlo to study detector resolution and acceptance

– Event Reconstruction for GEANT based MC

• GEANT4 MC will use the standard CLAS12 engine, GEMC.– Flexible design uses external, implementation independent field map and

detector geometry servers.

– Support for digitizations, etc provides data that can be used for full event reconstruction.

– Package will be maintained and developed together with CLAS12.


EIC@JLAB – further info

• EIC@JLAB webpage: http://eic.jlab.org– Overview and general information

• EIC@JLAB WIKI: https://eic.jlab.org/wiki– Ongoing project information

– Working groups

• EIC Collaboration meeting at Catholic University, 29-31 July, DC– Info: http://web.mit.edu/eicc/CUA10/index.html

• Weekly project meetings at JLab– Fridays at 9:30am in ARC724 or F324/25

• Series of Workshops at the INT, Seattle– Info: http://www.int.washington.edu/PROGRAMS/10-3/


Summary

• The EIC@JLab is well suited for taking JLab beyond 12 GeV

– Excellent tool to access nucleon/nuclear structure

• A medium energy collider is particularly appealing for measurements requiring transverse targets and/or good resolution and particle id (e.g., TMDs, GPDs).

– These processes benefit from high luminosity, excellent polarization, and more symmetric collision kinematics.

• Matches energy, luminosity, and detector/IR for deep exclusive and SIDIS processes

• Rapidly Expanding User Community


Documents

MEIC Simulation and Reconstruction Group Report Tanja Horn 1 Tanja Horn, CUA Colloquium Tanja Horn, MEIC Simulation Group Report, EIC Detector Workshop