Performance Studies of the SD Muon Detector for the Linear Collider C. Milstene -Arlington Workshop January 9-11,2003

1
Performance Studies of the SD Muon Detector for
the Linear ColliderC. Milstene -Arlington
Workshop January 9-11,2003

• The major issues
• m detection efficiency
• m p resolution
• p punchthrough
• are analyzed from within Jas
• The data samples sio files from SLAC
• The analysis is based on R. Markeloff software
• A direct comparison in the main parameters with
  Piccolo for Tesla.

2
Plan
1) SD versus Tesla Detectors. 2) The choice of
the energy points. 3) A set of typical and
Non-typical p events and their m
counterpart. 4) Choice of the parameters for m
Detection The m detection efficiency
stability upon changes of algorithms 5) The m
p resolution 6) The punchthrough p s
3
The Choice of the Energy points
From Calorimetry LC Overview Ray Frey, U. of
Oregon Chicago LCW, Jan 7, 2002
The majority of the particles are produced at
energies below 30 GeV arrow. m at and above 4
GeV reach MuDet. Very few p of 5 GeV do Part
are curling back in the magnetic field
4
4 GeV Muon run 1 event 2- 32 hits in the Muon
Barrel
5
4 GeV Muon- event 31 run 1- Fish Eye View- 32
hits in Muon Barrel
6
A 5 GeV p -gt m n Decay Event 206 Run 1- 33Hits
MuDet
7
Typical 10 GeV m -Event 9 Run 1-33 Hits in MuDet
8
A Typical 10 GeV m- Event 3 Run 1- with 33 Hits
in MuDet
9
A Typical 4 GeV Pi- event 15 Run1 y-FishEye
View no hits in Muon Detecto
10
Typical 5 GeV pi-Event 142 Run 1- gt100hits in
HDCal- no Hits MuDet
11
Typical 5 GeV pi- Event 21 Run 1-Curling back
effect of B5T
12
Typical 10 GeV punchthrough p-event 118-Run1- 6
hits MuDet
13
Punch Through 50 GeV pi- event 11 run 0-
y-EyeFish View-18 hits in Muon Detector
14
The SD and TESLA Muon Detectors
Comparison of the relevant parameters
SD
TESLA Outer_thick_plate
(Tesla only)
645cm Outer_Radius 660.5cm
585cm Inner_Radius
348.5cm
445cm
---------
-------- 312 cm
140cm The Unit
Fe 5cm
Fe 10cm
Gap 1.5cm RPC/gap Gap 4cm

48 Layers
10 Layers
80cm Fe16 planes 80cm
Fe10planes Inter. Length SD(EMHAD) 3.9
Lambda(SiW) Tesla 5.4 Lambda Prior to
MuDet
15
F,q,nlayer
The Algorithm
MU Coil 248cm
maxFbins(Tk-HDCal)4 max qbins(Tk-HDCal) 2
mcandidate
m
Track HDNLayers HDNHits HDHits MUNLayers MUNHits M
UHits
Minimum HDNHits 0 MUNHits 12
F,q,nlayer
HD 143cm
maxFbins(Tk-HDCal)3 max qbins(Tk-HDCal) 1
m list
EM 127cm
Track Extrapolated
Muon Package of R. Markeloff
SD Detector
16

Single Muon Efficiency F(Particle Momentum)
The Algorithm chosen are based upon the m
properties of penetration in correlation with
the lack of interactions and the track
continuity. The decays have been subtracted
following the procedure used in The source code
of MCPseudoParticle by M.Ronan/T. Johnson.

• The m Identification algorithm
• A charged track reconstructed
• Matching in HD within
• 3F bins 1qbins
• Matching in MuCal within
• DF 40mrd q20mrd
• More than 12 hits in Mucal

17
m efficency stability against algorithm variations
Cut1- DF 40mrd Dq40mrd- cut at 16
planes(80cm of steel). Cut2- DF 40mrd Dq20mrd-
cut at 12 planes. Cut3- DF 30mrd Dq10mrd-
cut at 12 planes. Cut4- DF 30mrd Dq10mrd-
cut at 8 planes.
18
Comparison with the m efficiency at Tesla

• m Efficiency for SD is represented
• on top of the plot of M.Piccolo for Tesla
• The algorithm used by M. Piccolo
• A m stub crossing at least 8/11 planes (80cm of
  steel)
• A stub is defined by angular hits consistency in
  MuCal The matching is within
• DF 40mrd Dq40mrd

19
Pion Response of the Muon System

• The response to p
• reported for 35000 events (Tesla)
• By M . Piccolo has been Reproduced The blue
  diamonds represent The SD
• Points for p after
• Normalization to account for the
• Difference in interaction length
• and statistics
• The Green stars
• Correspond to an
• Extra cut
• Requiring 5 planes with gt2 hits

20
Punch Through per Layer
The number of hits Per layer is shown For 3
points of Energy without applying the Extra-cut
of 2 hits or more in 5 layers.
21
Extra cuts were used to take into account the
multiplicity of hits in each Layer of the Muon
Detector. A cut on tracks having 5 layers with 2
or more hits within Df 40 , Dq40 mrd is shown
in the plot representing the p response of the m
system. It improves the separation p m by a
factor two at 50GeV and 20 GeV. Adding a
condition involving higher multiplicity per
layer, improves the separation by a factor 5
more. The separation at lower energies improved
very little. Those cuts did barely affect the m
efficiency within the error Bar , the m range
being of 1-2 hits per layer within the Df,Dq
range. A set of cuts of the kind could be
defined depending on the physics Channel Of
interest.
22
Conclusion
The m efficiency is stable through a large
variety of algorithms and comparable in both
detectors. The p response of the m system is
quite comparable as well. The p m separation
improves with cuts on the layer hits
Multiplicity. To the first order both SD and
Tesla have similar performances. And
improvements are anticipated by using other
calorimeter information.
23
From 6 GeV and up, the m of the decay will be
detected by MuDet being produced in the direction
of the original p and with a pgt 4GeV over almost
the whole range. (from G. Fisk)