The ATLAS Luminosity System

1
The ATLAS Luminosity System Per Grafstrom CERN
2
Luminosity Measurement WHY ?

• Cross sections for Standard processes
• t-tbar production
• W/Z production
• .
• Theoretically known to better than 10 will
  improve in the future
• New physics manifesting in deviation of ? x BR
  relative the Standard Model predictions
• Important precision measurements
• Higgs production ? x BR
• tan? measurement for MSSM Higgs
• .

3
Luminosity Measurement WHY ? (cont.)
Examples
Higgs coupling
tan? measurement
Systematic error dominated by luminosity (ATLAS
Physics TDR )
4
Luminosity Measurement Options

• Relative luminosity a DEDICATED luminosity
  monitor is needed
• See Jim Pinfolds talk on LUCID
• Absolute luminosity
• Goal
• measure L with ? 2-3 accuracy
• How
• LHC Machine parameters
• Use ZDC in heavy ion runs to understand machine
  parameters
• rates of well-calculable processese.g. QED, QCD
• optical theorem forward elastic rate total
  inelastic rate Use Roman Pots
• needs full ? coverage-ATLAS coverage limited
• Use ?tot measured by others (TOTEM)
• Combine machine luminosity with optical theorem
• luminosity from Coulomb Scattering Use
  Roman Pots
• ATLAS pursuing all options

5
L from LHC Machine Parameters

• Luminosity depends exclusively on beam
  parameters
• Luminosity accuracy limited by
• extrapolation of ?x, ?y (or ?, ?x, ?y) from
  measurements of beam profiles elsewhere to IP
  knowledge of optics,
• Precision in the measurement of the the bunch
  current
• beam-beam effects at IP, effect of crossing angle
  at IP,

6
Use ZDC in heavy ion runs to calibrate machine
instrumentation (Sebastian White)
7
IP1IP5 absorbers
TAN_at_140m
Exploded view
Luminosity Monitoring
8
Luminosity from other Physics Signals

• QED pp ? (p?)(p?)?p(????)p
• signal (µµ)-pair with ?(µ)lt2.5, pT(µ)? 5-6
  GeV, pT(µµ) ? 0
• small rate 1pb (0.01 Hz at L1034)
• clean backgrounds from DY, b, c- decays can be
  handled by appropriate offline cuts
• uncertainties µ trigger acceptance efficiency,
  
• ( A.Shamov V.Telnov, hep-ex/0207095)
• QCD W/Z ? leptons
• high rate W?l? 60 Hz at L1034 (e 20)
• current theory systematics PDF and parton
  cross sections ? 4
• gives relevant parton luminosity directly
• detection systematics
• trigger/acceptance/identification efficiency/
  backgrounds
• detailed study for ATLAS detector needed
• Both processes will be used

9
(No Transcript)
10
Roman Pots
11
The ATLAS Detector
Calorimetry
Tracking
R
?-chambers
Barrel
Diffraction/Proton Tagging Region
EndCap
RP
Tracking
ZDC/TAN
FCAL
TAS
LUCID
y
12
(No Transcript)
13

• Extension of ATLAS- A two stage process
• Short time scale
• Forward detector for Luminosity measurement
• Elastic scattering in the Coulomb region
• Longer time scale
• Gain experience in working close to the beam
• ? propose a diffractive physics program using
  additional detectors

14
Elastic scattering in the Coulomb region
From the fit we will get ?tot , ?, b and L
15
The total cross section
16
The ? parameter

• ? Re F(0)/Im F(0) linked to the total cross
  section via dispersion relations
• ? is sensitive to the total cross section beyond
  the energy at which ? is measured ? predictions
  of ?tot beyond LHC energies is possible
• Inversely Are dispersion relations still valid
  at LHC energies?

17
The b-parameter or the forward peak

• The b-parameter for lt llt .1 GeV2
• Old language shrinkage of the forward peak
• b(s) ? 2 ? log s ? the slope of the
  Pomeron trajectory ? ? 0.25 GeV2
• Not simple exponential - t-dependence of local
  slope
• Structure of small oscillations?

18
What else can we do?

• Coulomb region extremely challenging. All
  aspects of design optimized for this.
• Medium (0.1-1.0 GeV2) elastic scattering needs
  medium beta optics, low Lumi, short runs)
• Large t (1-10 GeV2) elastic scattering needs
  high Lumi, standard optics, and continuous runs.
  
• Studies needed
• Proton tagging to identify a diffractive
  interaction must be possible at some level . BUT
  t and ? acceptance and t and ? resolution need to
  be understood.
• Simulation and optics investigations needed to
  see if there is any physics reach for single and
  central diffraction using proton tagging.
• Signal and background rates have to be studied.
  Trigger set up?
• Many open questions

19
Requirements to reach the Coulomb region

• Required reach in t
• Requires
• small intrinsic beam angular spread at IP
• insensitive to transverse vertex smearing
• large effective lever arm Leff
• detectors close to the beam, at large distance
  from IP

Paralleltopoint focusing
20
Experimental Technique

• Independence of vertex position
• Limit on minimum tmin
• The main potential difficulties are all derived
  from the above
• Leff,y large ? detectors must be far away form
  the IP ? potential interference with machine
  hardware
• small tmin?
• ? large ? special optics
• small emittance
• small ns ? halo under control and the detector
  must be close to the beam

21
Roman Pot Locations
One Roman Pot Station per side on left and right
from IP1
Each RP station consists of two Roman Pot Units
separated by 3.4 m, centered at 240.0 m from IP1
22
Very high b (2625 m) optics

• Solution with following characteristics
• At the IP
• ?? 2625 m
• ? 610 ?m , ?? 0.23 ?rad
• At the detector
• ?y,d 119 m, ?y,d 126 ?m
• ?x,d 88 m, ?x,d 109 ?m
• (for ?N 1 ?m rad)
• Detector at 1.5 mm or 12?
• tmin 0.0004 GeV2
• Smooth path to injection optics exists
• All Quads are within limits
• Q4 is inverted w.r.t. standard optics!

Endorsed by LTC Compatible with TOTEM optics see
LEMIC minutes 9/12/2003
23
Emittance

• Emittance of 1106 m?rad needed to reach
  Coulomb region
• Nominal LHC emittance 3.75106 m?rad
• Emittances achieved during MDs in SPS
• Vertical plane 1.1106 m?rad and Horizontal
  plane 0.9106 m?rad for 7x1010 protons per
  bunch
• 0.6-0.7106 m?rad obtained for bunch intensities
  of 0.51010 protons per bunch
• However
• Preserve emittance into LHC means that injection
  errors must be controlled
• (synchrotron radiation damping might help us at
  LHC energy)
• emittance eN, number of protons/bunch Np , and
  collimator opening ns,coll (in units of s) are
  related via a resistive (collimator) wall
  instability limit criterion
• thus eN 1.5106 m for Np 1010, ns,coll 6
• ? Best parameter space from beam tuning sessions

24
Beam Halo limit on ns

• Beam halo is a serious concern for Roman Pot
  operation
• it determines the distance of closest approach
  dmin of (sensitive part of) detector
• ns dmin/sbeam 9 ns 15
• Expected halo rate (43 bunches, Np1010, eN 1.0
  µm rad, ns10) 6 kHz

25
Requirements for Roman Pot Detectors

• Dead space d0 at detectors edge near the beam
  d0 ? 100?m (full/flat efficiency away
  from edge)
• Detector resolution ?d 30 ?m
• Same ?d 10 ?m relative position accuracy
  between opposite detectors (e.g. partially
  overlapping detectors, )
• Radiation hardness 100 Gy/yr
• Operate with the induced EM pulse from
  circulating bunches (shielding, )
• Rate capability O(MHz) (40 MHz) time resolution
  ?t O(1 ns)
• Readout and trigger compatible with the
  experiment DAQ
• Other
• simplicity, cost
• extent of RD needed, time scale, manpower,
• issues of LHC safety and controls

26
Roman Pot Detectors

• Square scintillating fibers
• Kuraray 0.5 mm 0.5 mm fibers
• 10 layers per coordinate
• 50 µm offset between layers
• Main reason for choice
• Small dead space
• EM parasites not a problem

scintillator plate for triggering
y-measurement detector
x-measurement detector
halo inter calibration planes (well away from
beam)
27
Scint. fiber detector
expect Npe/hit 4.9 empirically 3 is more
likely
28
Detector Performance Simulations

• First simulation results
• strip positioning sfiber 20 µm
• light and photo-electron yield
• Npe ltdE/dxgt dfiber (dn?/dE) eA eT eC gR eQ ed

29
(No Transcript)
30
Fiber routing
Radius of curvature min 30mm Average fiber length
230mm Nb of fibers in X config. 71 x 10 x 2
1420 Total fiber length 330m
Square holes filled with glue (to be tested in
prototype)
31
Fiber housing installation
32
Integration
Free space PMs baseplate 106 40 137
9mm ? Separation at PMs level or hole in base
plate required
33
(No Transcript)
34
Simulated Elastic Scattering
Inner ring t -0.0007 GeV2
Outer ring t -0.0010 GeV2

• Reconstruct ?

35
Simulated dNel/dt and simple fit

• Event generation
• 5 M events generated corresponding to 90 hr at
  L? 1027 cm-2 s-1
• NO systematics on beam optics!
• Only 1 Roman Put unit/arm
• Simple fit
• range for fitting
• 0.00056 lt t lt 0.030 GeV2
• 4 M events measured for dN/dt

36
Conclusion

• ATLAS pursues a number of options for Absolute
  Luminosity Measurement
• Coulomb normalization
• W/Z rates
• production of muon pairs via double photon
  exchange
• elastic slope extrapolated to dN/dtt0 plus
  machine L
• elastic slope extrapolated to dN/dtt0 plus ?tot
  from TOTEM
• machine parameters alone
• Cross calibration from ZDC in Ion runs
• others
• The Coulomb Interference measurement is very
  challenging but seems within reach.
• Small angle elastic scattering will address old
  fashion physics in terms of ?tot , ? and b
• This experience of working close to the beam will
  prepare us for a Forward Physics Program
  with ATLAS in a possible future upgrade

37

• Back up slides

38
Summary on emittance and beam halo issues

• Looks feasible but no guarantees can be given
• However, if we dont reach the Coulomb region the
  effort is not in vain
• we can still
• Use ?tot as measured by TOTEM/CMS and get the
  luminosity by measuring elastic scattering in a
  moderate t-range( -t0.01 GeV2 ) and use the
  Optical theorem for the rest
• Use the luminosity measured by machine parameters
  and again via the Optical theorem get ?tot and
  all other cross sections relative to ?tot with a
  factor 2 better precision than from the machine
  parameters

39
Luminosity transfer 1027-1034 cm-2 sec-1

• Bunch to bunch resolution ? we can consider
  luminosity / bunch
• ? 2 x10-4 interactions per bunch to 20
  interactions/bunch
• ?
• Required dynamic range of the detector 20
• Required background ? lt 2 x10-4 interactions per
  bunch
• main background from beam-gas interactions
• Dynamic vacuum difficult to estimate but at low
  luminosity we will be close to the static vacuum.
  
• Assume static vacuum ? beam gas 10-7
  interactions /bunch/m
• We are in the process to perform MC calculation
  to see how much of this will affect LUCID

40

• Alternative photodetectors
• Geiger Mode Avalanche Photodiodes
• high gain, 106 - low QE 15
  (geometrically limited)
• low bias voltage 50 V - high dark count rate
  at R.T. 106 Hz
• very fast signal characteristics - not yet
  really commercialized
• very simple electronics
• small size
• GM-APD would become interesting if geometrical QE
  limitation can be overcome.
• QE gt 30 is claimed to be in reach.
• This would mean Npe 6-8 for our baseline
  configuration.

41
Cost Estimates Participants

• Participating institutes
• (as a subsystem, fully part of the ATLAS
  collaboration)
• University of Alberta
• CERN
• Ecole Polytechnique
• Institute of Physics Academy of Science, Czech
  Republic
• University of Manchester
• University of Montreal
• University of Texas
• University of Valencia
• SUNY Stony Brook