Download 1 - MICE

6 Detectors
6.1 Overall description, functionality and redundancy
The MICE detector system as sketched in Figure 3.1 is described in this section, element by
element. The driving design criteria are: i) robustness, in particular of the tracking detectors, to
potentially severe background conditions in the vicinity of RF cavities and ii) redundancy in
particle identification (PID) in order to keep contamination (e, ) below 1%.
Two spectrometers of very similar design, one upstream and one downstream of the cooling
section, measure the full set of six muon parameters. Each of them provides a high-resolution
measurement of the five parameters of the muon helix in a tracker embedded in a 4 T solenoid, as
well as a precise time measurement. In addition, muon/pion/electron identifiers (a t0 timing station
and a small Cherenkov) are situated in front of the upstream detector and muon-electron
identifiers (a larger Cherenkov and an electromagnetic calorimeter) are situated beyond the
downstream spectrometer.
6.2 Scintillators for timing, trigger and upstream PID
Three fast time-of-flight (TOF) stations equipped with fast scintillators are foreseen. The first two
stations (TOF 0 and TOF 1), upstream of the cooling section and separated by about 10 m, will
provide the basic trigger for the experiment, in coincidence with the ISIS clock. These two
stations have precise timing (around 70 ps) and will provide muon identification by TOF. The
second of these stations will also provide the muon timing (relative to the RF phase) necessary for
the measurement of the input longitudinal emittance. The coincidence with a third scintillator
station of similar nature (TOF 2), downstream of the second measuring station, will select
particles traversing the entire cooling section. The variation of emittance due to losses and decays
will thus be distinguishable from cooling. The TOF 2 station will also record the muon timing for
the measurement of the output longitudinal emittance. As discussed in [Janot01], a 70 ps
resolution provides both effective (99%) rejection of beam pions and adequate (5°) precision in
the measurement of the muon RF phase. Other design criteria are efficiency, redundancy and
quality of calibration. The design presented here satisfies these requirements.
The three TOF stations are 1212, 4040 and 4040 cm2 respectively. The two largest stations
(TOF 1, TOF 2) are equipped with 8 scintillator slabs (4062.5 cm3) to make a plane (Y), by
staggering and superimposing them at the edges for about 1 cm to allow cross-calibration with
impinging beam particles. Bicron BC-404 scintillator material (with 1.5 ns decay constant and
1.7 m attenuation length) is the most suitable choice. The smallest station (TOF 0) could be made
of two crossed planes (X-Y), each of two 1262.5cm3 slabs, using BC-420 plastic scintillator,
which is even faster than BC-404 but with a shorter attenuation length. Each slab is read out at
both ends by a fast photomultiplier through a Plexiglas light guide. The time-of-flight
measurement is achieved by combining leading-edge time measurements from a TDC with pulseheight information from an ADC.
TOF 0 will be equipped with Hamamatsu R4998 PMTs (0.7 ns rise time, 160 ps transit time
jitter) or equivalent. The fringe fields of the spectrometer solenoid have been estimated by a
Poisson-Superfish [Poiss] calculation to be as high as 1 T. The choice of the PMTs for TOF 1 and
TOF 2 is therefore a critical issue. One option is to use the same R4998 PMTs but with a multilayer mu-metal shield and a suitable design of the light guides. An alternative option is the
Hamamatsu R5505 which can operate beyond 1 T but with a reduced gain. Similar solutions were
successfully adopted in the BESS experiment BESS].
Studies with several scintillation counters equipped with both types of PMTs will be performed at
INFN Milano, INFN Padova, and in the free air bore of a large superconducting solenoid facility
LASA at the INFN LASA Laboratory in Milano. A test beam at the CERN PS is also foreseen
before a final decision is made. Funding for the initial phase of these tests has already been
granted by INFN. For the time inter-calibration of a single detector plane, cosmic rays will be
used with a dedicated set-up for the trigger (as done in the HARP experiment HARPTOF), or
with beam particles passing in the overlap region of two nearby counters. The time monitoring of
the system will be done with a laser-based system. Studies are under way to assess whether the
expensive laser system [HARPlaser] used in HARP can be refurbished, or whether a similar one
must be purchased.
The target of 70 ps resolution seems well within reach, as performance ranging between 50 and
90 ps intrinsic resolution has been published for TOF planes of similar dimensions [BESS,
NA49]. Assuming that the electronics and calibration system of the HARP TOF wall can be reused, a preliminary cost estimate for scintillator, PMTs and general reconfiguration would
amount to about 200 k€ (see Table 8.1 for details). A layout with crossed (X-Y) planes is also
under study for TOF 1 and TOF 2.
Table 6.1: Cost estimate for the TOF system (capital investment only).
Item
Detector system
40 PMTs + mu metal
Scintillator (BC-404 and BC-420), light guides
Mechanics
Calibration system
Laser
Optical system
Cosmic ray set-up
Electronics
QDCs
TDCs
Scalers, MT, NIM modules, delay boxes, splitters
Crates, HV system, discriminators
HV cables, signal cables
Total
aThese
bTotal
Cost (k€)
100
15
5
80a
15
5
10
10 a
20 a
30
10 a
300b
items may be recoverable at no cost from the HARP experiment.
would be reduced to 180 k€ if all indicated items are recovered from HARP.
6.3 Upstream mice Cherenkov detector for --e separation
The upstream Cherenkov detector , Figure 6.3.1, provides pion/muon/electron separation to
insure a clean muon beam for the MICE experiment. The purpose of the upstream Cherenkov is
to beat down backgrounds left over from the time of flight detector. The detector will use four
cells each 200mm across. Each cell will be 22mm thick, 20mm of C6F14 fluorocarbon liquid plus
a 2mm quartz window. C6F14 has been used by the SLAC SLD [Caval] and CERN DELPHI
[Annas] experiments as a Cherenkov radiator. Note that Cherenkov radiation produced in the
quartz window tends to be trapped by total internal reflection and does not reach the
photomultiplier tube. The DIRC detector at SLAC's BaBar experiment is based on trapping
Cherenkov light in quartz [4]. Four air light guides with 45 degree mirrors bring the light out to
four 200 mm photomultiplier tubes. The upstream Cherenkov will be located between the two
quadrupole triplets of the MICE beam line where stray magnetic fields are low. The magnets on
the ends of the triplets are labeled Q6 and Q7. Two layer shields of low carbon steel and mu
metal should suffice to protect the tubes. The total length of the device is 50 cm. The index of
refraction of C6F14 is about 1.25 and it has thresholds of 0.7, 140, and 190 MeV/c for electrons,
muons, and pions, respectively. Pulse height information is used to aid in discrimination. See
Figure 6.3.2 and [Aubert], [Bartlett], and [Crema].
PMT
UV
WINDOW
MIRROR
CL
C6F14
Figure 6.3.1: Schematic of Upstream Cherenkov Detector. A single quadrant is shown .
Up to 190 MeV/c only the muons produce light, so pions are completely rejected. See
Figure 2. The rejection should match the TOF numbers at 190 MeV/c. Above 190
MeV/c the pions start to produce some light, so one must start to reject particles that fall
between the muon and pion peaks. These leads to inefficiency but not contamination at
moderately higher momenta. With four PMTS, a MHz beam, and a 10 nanosecond gate,
one in 400 particles will overlap. These need to be rejected.
A prototype Cherenkov counter with C6F14 and a five inch RCA 8854 photomultiplier tube has
been built and used to detect cosmic ray muons.
For data acquisition, four ADC channels such as provided by the 10-bit LeCroy FERA
4300B and four high voltage channel are required. For slow controls, one channel for
triggering an LED light pulser, one channel for a temperature probe, one channel for a
humidity probe, and four channel to monitor the high voltages are required.
The estimated cost for the upstream Cherenkov is $96,000. This includes $66,000 of capital
cost and $30,000 for staff. The staff component consists of two people part time. Inflation of 3%
should be added if the schedule is substantially delayed. Contingency is not included in these
numbers. The estimated time for construction is 18 months once funding becomes available.
e Candidates 186MeV/c 1cm C6F14
0
20
40
100
60
80
Figure 6.3.2: Monte Carlo simulation of 186 GeV/c Pion-Muon-Electron
response (from left to right) for 1
Npe
cm of C6F14. The number of photoelectrons is recorded .
6.4 Tracker Solenoid
The tracker solenoid assembly consists of five superconducting coils that fit into a unique
cryostat. The five coils are assembled together so to give a single cold mass. Though the global
design (magnetic, mechanical and cryogenic) is still in progress, the degree of definition of the
system is advanced. The geometrical lay-out at room temperature of the coils is shown in Table 6.
Differently from the proposal, all coils have the same inner radius (255 mm). Other relevant
information is shown in Table 6.4.1.
The superconductor proposed is the same for all coils:
1. The conductor is a standard MRI magnet conductor with a copper-to-superconductor
ratio of eight;
2. Each conductor consists of 92 filaments that are 80 µm in diameter;
3. The dimensions of the insulated conductor are 1.65  2.40 mm, and the conductor is
rounded to prevent insulation cracking;
4. The design critical current for the conductor is 800 A at 4.2 K and 5.0 T
Table 6.4.1 Position of Tracker Solenoid Coils at Room Temperature
Coil Left Z
from center
of MICE
channel *
Coil Left Z
from left side
of tracker
cryostat
(mm)
(mm)
Matching
#1
3660
Matching
#2
Coil
Inner R
Coil
Length
Coil
Thickness
(mm)
(mm)
(mm)
132
255
202
50
+810
3910
382
255
202
73
+150
End #1
4261
733
255
120
116
+340
Center
4441
913
255
1260
50
+21
Magnet
Coil
Maximum
net axial
force (kN)
End #2
5761
2233
255
120
149
-1440
* The Z distance is defined as the distance from the center of the MICE channel. The position of coil at
minus z is symmetric.
Note: the coil current density at –Z, which is J(-Z) equals - J(Z).
Note:
The detector magnet shrinks around a point at Z = 4774 mm and R = 0 mm.
Fig. 6.4.1 shows the load lines of the five coils compared with the Ic(B) curve. The case shown
refers to a configuration with a current of 250 A is flowing in all coils. It appears clearly the
design choice to work with a large enthalpy margin, i.e. at a nominal current considerably smaller
than the critical current.
Table 6.4.2 Other characteristics of Tracker Solenoid Coils
Required
Conductor
length (m)
Inductance of
the single coils
(H)
Weigth
30
4500
3.9
170
84
44
6800
8.8
250
End #1
50
70
6900
8.9
255
Center
525
30
28000
46.2
1060
End #2
50
90
9400
14.9
340
-
-
55600
108
2300
Turns per
layer
No of layers
Matching #1
84
Matching #2
Magnet Coil
(kg)
Complete coil system
(including axial spacers
and flanges)
End#2
1000
End#1
Match#2
1500
Main Solenoid
Match#1
Current (A)
2000
500
0
3
3.5
4
4.5
5
5.5
Field (T)
Figure 6.4.1 The load lines for each coil (peak field in the winding) compared with the conductor critical current
curve.
The coils are supposed to be wound on a demountable bobbin. Each coil is axially contained by
two fiberglass epoxy (G10) thin flange and the remaining space between coils is filled with
aluminum alloy (5083) block rings. After the winding of each coil, a 2-3 mm thick strip in
aluminum alloy is wound onto the coil, so to form an external structure providing the hoop
strength. After the winding of all coils, the filling of axial space with aluminum and the banding
of Al-alloy strip, the whole system is epoxy impregnated under vacuum. When, after
impregnation, the winding bobbin is removed, we have a single solid system mechanically selfconsistent with no interfaces with mechanical structures. Figure 6.4.2(a) shows how the cold mass
shall appear.
The cold mass also includes two other components: the cooling circuit and the side support
flanges. The cooling circuit is made of aluminum pipes directly glued on the basic cold mass.
LHe is flowing inside the pipes being collected in two manifolds (a lower and an upper one). The
circulation is allowed by natural thermo-siphon. For the long center coil it is also foreseen to
include in the winding longitudinal copper strip (Figure 6.4.2.(a) shows these strips on the inner
diameter of the coil) to increase the longitudinal thermal conductance. The end #2 coil is much
thicker than other coils. In order to limit the radial thickness of the cryostat it has been decided to
include the cooling channels in the Al-alloy flange at the side of the coil (this choice allows to put
the two manifolds at a diameter comparable with the outer diameter of the end #2 coil (as also
Fig.6.4.2 shows). In order to support the cold mass inside the cryostat (essentially a vacuum
chamber), two Al-alloy 30 mm thick flanges have been put at the two sides. The flanges are
epoxy glued at the side, thought the gluing has no mechanical role because the 8 tie rods,
connected four by four to the flanges as shown in Figure 6.4.2.(b) are intended to work always in
tension.
Figure 6.4.2: The cold mass. A) The file coils (brown) sided by G10 plates (green) are separated by Al-alloy (blue) and
externally supported by an Al-alloy banding. B) The cooling circuit is directly glued onto the basic cold mass. The
figure shows the cooling pipes with two manifolds for thermo-siphon circulation. Two thick Al-alloy flanges are placed
at the sides, providing the cold attachments of the 8 fiber-glass epoxy tie-rods supporting the coil
As before mentioned, the cold mass is supported in the vacuum chamber through 8 tie-rods 830
mm in length. The tie-rods are one side attached to the two Al-alloy flange limiting axially the
cold mass and the warm side attached to the external cylindrical shell of the vacuum chamber.
The four axial rods placed on the top hold the weight of the cold mass (2300 kg). However one of
the main role of the tie-rods is to hold the axial forces (as high as 120 kN). Supposing to use
oriented fiberglass epoxy tie-rods working at a stress considerably lower than the tensile, say 200
MPa, we need rods of 15 mm diameter.
The cryostat has three main components: the thermal shield, the vacuum chamber and the
chimney hosting the proximity cryogenics and the current leads. The thermal shield surrounds
completely the cold mass. Consequently we have an inner shield, an outer shield (both of
cylindrical shape) and two cover flanges at the sides. The shield is made of pure aluminum plates
2-3 mm thick arranged on a light supporting structure. A simple cooling pipe system is coupled to
the thermal shield to allow its cooling to a temperature to be determined (in the 50-80 K range).
Fig. 6.4.3 shows a possible structure of the shield. The vacuum chamber is a stainless steel
vessel composed of several parts, as shown in general magnet view of Figure 6.4.4. The inner
cylindrical shell 4 mm thick and ID 400 mm; the outer shell 40 mm thick and 1080 mm OD,
including the 8 protrusions 28o angled hosting the tie-rods; 3) The four feet anchoring the coil to
the ground. The chimney includes the current leads (a couple per coil for a total of 10 leads) and
the proximity cryogenics. Fig. 6.4.4 shows a cooling system based on thermo-siphon method,
being the liquid helium at 4.2 K provided by an external system (dewar and/or cryoplant). In fact
the actual orientation is to involve cryo-coolers included in the cryostat. This reflects in a
modification of the turret, which would include three cryo-coolers and 10 current leads.
Figure 6.4.3: The thermal shield with its three parts (inner shell, outer shell and side covers. A cooling channel is
attached to the shield structures
Figure 6.4.4 General view of the tracker solenoid (the turret hosting the proximity cryogenics is not yet defined . A
solution with three cryo-coolers is under study)
The cooling system presently under study is based on three main concepts:



The cooling power is provided by cryo-coolers directly connected to the cold-mass
The cryo-coolers do not cool the coil and the thermal shields by conduction, but through
an exchange gas. This choice comes from the need to cool efficiently a long coil. The
cooling by thermal conduction starting from an extremity may result in unacceptable
temperature gradient inside the cold mass. The LHe coolant can better distribute the
cooling power.
With cryo-cooler off the coil cool-down shall be possible though direct feeding of
cryogens.

A possible scheme according these guidelines is shown in fig. 6.4.5. Two cryo-coolers work on
two stages: first one around 50 K with a cooling power of 60 W and second stage at 4.5 K with
1.5 W cooling power. These cryo-coolers indeed work as re-condensers. The third single stage
cryo-cooler provides 60 W of cooling power at 50 K. These cryo-cooler are currently available on
market. In total we have 180 W at 50 K and 3 W at 4.5 K. This power is necessary according to
the heat loads assumed in Table 6.4.3. One can see that at the temperature of 4.5 K we have
cooling more power than needed (and this extra-power will help in cooling-down the coil) whilst
at 50 K, we just have the required power with no margin. This is the main problem related to the
use of a limited and reasonable number of cryo-coolers. The high power at 50 K is mainly due to
the heat losses through the 5 pairs of copper (or brass) leads from 300 to 50 K.
With cryo-coolers ON, the shield temperature is minimum 50 K, the liquid nitrogen in the
reservoir (connected directly to I stage) is frozen. The shield is cooled by the conduction.
Considering that the most of heat dissipation is due to the current leads (150 W), which are
located close to the I stages of cry-coolers, the maximum temperature of the shield is found at the
side opposite to the turret and is around 80 K. The LHe reservoir is cooled by the two second
stages of cryocoolers. The LHe circulates in the coil cooling circuit through natural thermosiphon.
With cryo-coolers OFF, the LN2 shall be provided through the IN pipe (normally plugged). The
vent pipe of the LN2 reservoir allows the extraction of the gas. The LHe is provided directly
trough the IN pipe. During cool-down the cold valve connecting the LHe IN and the LHe
reservoir is closed to prevent a hydraulic shorten. The LHe or the (cold He gas) coming from an
external dewar pass through the coil cooling circuits and come to the LHe reservoir. The He gas
is extracted from a further pipe. Once the coil is cool and the reservoir if filled with LHe, the cold
valve is opened and the IN pipe is plugged. The thermo-siphon can operate. A third pipe (not
shown in the figure) is used for continuous refurbishing of the reservoir as the He gas evaporates
and leaves the system through the exaust pipe. The shield is cooled simply by filling the reservoir
of liquid nitrogen.
Figure 6.4.5 Schematic drawing of the turret with Cryo-coolers, Cryogens reservoirs and Current leads
Table 6.4.3: Heat loads to the tracker solenoid
Parameter
Heat Load at 4.5K
with shield at 50 K
(W)
Heat Load at 4.5K
with shield at 77 K
(W)
Heat Load at 50 K
(W)
5 pairs of current leads for
max current 300 A
0.250  5 = 1.50
0.4  5 =2.0
30  5 = 150
Tie rods (thermal intercepted
at 400 mm from cold mass)
0.05
0.1
0.500
-
-
0.05
Radiation
0.1
0.2
13
Other conduction losses
(pipes, cold valve, wires)
0.2
0.4
15
Total
1.85
2.7
178
Shield supports
Reference drawings
Front and side view
6.5 Tracker module
6.5.1 Overview
The MICE experiment (figure 6.5.1) requires that the emittance be measured as the muon beam
enters the cooling channel and again as it leaves. The emittance measurement will be
accomplished using two solenoidal spectrometers. The upstream and downstream spectrometers
will be identical in construction but will be installed such that the downstream spectrometer is a
copy of the upstream device rotated through 180º and translated in Z to the appropriate position.
Upstream spectrometer
Downstream spectrometer
Figure 6.5.1 Drawing of the MICE experiment showing the upstream and downstream spectrometers and the MICE
cooling channel.
The baseline spectrometer module consists of a 4 T superconducting solenoid of 40 cm bore
instrumented with five planar scintillating-fibre stations. Each station is composed of three
doublet layers laid out in a ‘u, v, w’ arrangement. The active area of the device is a circle of
diameter 30 cm. The fallback for the spectrometer instrumentation is a time projection chamber
(TPG) with gaseous electron multipler (GEM) readout.
6.5.2 Specification
The principle requirements that must be satisfied by the MICE tracking system are:

High efficiency reconstruction of muon tracks in the presence of background;

Adequate resolution in the reconstructed track parameters to allow a measurement of
emittance with an absolute precision of 1‰ to be made.
Simulations have shown that these goals can be achieved using the baseline five-station
scintillating-fibre tracker. A 3D engineering model of the fibre tracker is shown in figure 6.5.2.
Each station consists of three sets of fibre doublet layers mounted at 120º to one another. The
fibre doublet arrangement is illustrated in figure 6.5.2. The station-numbering scheme is defined
in figure 6.5.1 and the mechanical specification of the tracker is summarised in table 6.5.1.
Figure 6.5.2 Engineering model of the tracker module showing the 5-station scintillating fibre tracker installed in the
solenoid and the optical patch panel.
a)
b)
Figure 6.5.3 Detail of arrangement of fibres in doublet layer. (a) Cross-sectional view of fibre doublet. The dimensions
of the fibre and fibre spacing are indicated in m. The fibres indicated in red indicate the seven fibres ganged for
readout via a single clear fibre. (b) Layout of doublet layers in a station. The angle between the fibres in the doublet
layers is 120º.
Scintillating fibre tracker
Tracking volume
Spectrometer solenoid
Parameter
Scintillating fibre diameter
Primary dopant, para-terphenyl, concentration
Secondary dopant, 3HF, concentration
Fibre pitch
Estimated light yield per singlet (photo-electrons)
Number of scintillating fibres per optical readout
channel
Position resolution per plane
Views per station
Radiation length per station
Stations per spectrometer
Station separation: 1 – 2
Station separation: 2 – 3
Station separation: 3 – 4
Station separation: 4 – 5
Sensitive volume: length
Sensitive volume: diameter
Magnetic field in tracking volume
Field uniformity in tracking volume
Field stability
Bore diameter
Pressure in magnet bore
Value
350 m
1.25% (by weight)
0.25% (by weight)
427 m
8
7
470 m
3
0.35% X0
5
45
35
20
10
1,100 mm
300 mm
4T
1‰
400 mm
Vacuum
Table 6.5.1: Key parameters of the tracker module. The scintillating-fibre tracker parameters are followed by the
specification of the size of the tracking volume and the specification of the spectrometer solenoid.
The performance of the spectrometer has been simulated in G4MICE using the nominal input
beam defined in section ??. At the entrance of the upstream spectrometer, the nominal beam
momentum is 200 MeV/c with an emittance of 6.4 mm mrads. Figures 6.5.4a and 6.5.4b show the
distribution of transverse momemtum, pT versus longitudinal momentum, p Z measured in the
upstream and downstream spectrometers respectively. The robustness of the measurement has
been investigated by evaluating the migration from the generated transverse and longitudinal
momentum ( pTt , pZt ) to the measured values. In bins of the generated track parameters the mean
transverse and longitudinal momentum is calculated using generated and measured parameters in
turn. The migration for a particular bin is defined to be the vector in the pTt , pZt plane that joins
 p t , p t  to  p , p  . Figure 6.5.5.a and 6.5.5.b show the migration for the upstream
T
Z
Z 
 T
and downstream trackers respectively.
6.5.3
Definition of local coordinate system and number scheme for the tracker
stations – KL
6.5.4 6.5.4 Baseline: scintillating fibre tracker
6.5.4.1 Operating principle
The basis of charged particle tracking in MICE will be the production of scintillation light in
350 m double clad, doped polystyrene fibres. The concentration of the primary and secondary
dopants must be optimised to maximise the light yield while minimising the fibre-to-fibre optical
cross talk. The passage of a charged particle through the fibre causes energy to be transferred to
the primary dopant, para-terphenyl (pT). The peak of the scintillation light spectrum of pT is at a
wavelength of ~350 nm. The secondary dopant, 3-hydroxflavone (3HF), absorbs this light and reemits at a wavelength of ~525 nm. The concentration of primary dopant must be high enough that
sufficient primary light is generated but small enough to ensure that re-absorption of primary light
in the pT is small. The concentration of 3HF must be small enough to ensure negligible secondary
light attenuation along the length of the active fibre but large enough that the absorption length of
the primary light in the 3HF is small compared to the fibre diameter. The latter condition ensures
that fibre-to-fibre cross talk is eliminated. Measurements have shown that pT and 3HF
concentrations of 1.25% and 0.25% by weight respectively give sufficient primary light and an
attenuation length for absorption of the primary light in the 3HF of 25 m. A series of
measurements of scintillator properties as a function of primary and secondary dopants is planned
to optimise the dopant concentrations for the MICE fibre tracker. The baseline specification is
given in table 6.5.1.
Small-diameter fibres are required to reduce multiple scattering in the stations. However, reading
out each fibre leads to a large channel count and a significant electronics cost. The channel count
can be reduced significantly if 7 scintillating fibres are read out through a single clear-fibre
waveguide (see figure 6.5.6). Seven-fold ganging of scintillating fibres leads to a sensitive
element that is 1.63 mm across (see figure 6.5.3) and hence a resolution of 470 m. Simulation
has shown that this resolution is acceptable.
Clear fibre, of diameter 1.05 mm, will be used to transport the light from the stations to the patch
panel and from the patch panel to the photodetector. The longest clear-fibre run inside the magnet
bore (~1.5 m) will be matched to the shortest run from the patch panel to the photodetector
(~0.5 m). The total length of clear fibre will therefore be kept at or below ~3 m. The attenuation
length of the clear fibre has been measured to be 7.6 m. A total clear fibre length of 3 m therefore
corresponds to 40% of an attenuation length. It has been estimated that the reduction in light yield
by attenuation in the clear fibre is acceptable.
The Visible Light Photon Counter (VLPC) developed for use in the D0 experiment will be used.
The VLPC is a low band-gap light sensitive diode that is operated at 9 K to reduce thermal
excitation and is ideal for use in MICE because of its large quantum efficiency (85%) and high
gain (50,000). The device is also insensitive to the magnetic fields in the neighbourhood of the
MICE spectrometer solenoids and to the RF power radiated by the MICE cavities and associated
power supplies and RF-power distribution system. The latter was demonstrated in a dedicated
series of measurements in which the D0 VLPC test stand was exposed to levels of radiated RF
power several times in excess of those expected in MICE with no detrimental effect on
performance.
Figure 6.5.6 Detailed of seven-fold ganging. Seven scintillating 350 m fibres (shown in red) are readout through a
single 1.05 mm clear fibre (shown in black).
6.5.4.2 6.5.4.2 Mechanical design – ME/GB
6.5.4.2.1 Tracker overview – GB
6.5.4.2.2 Station – GB
6.5.4.2.2.1 Carbon-fibre station former – PC
6.5.4.2.2.2 Scintillating fibre ribbon
The scintillating fibres used in the prototype were 350 m diameter Kuraray multi-clad. They
used the standard p-terphenyl primary dopant. The prototype test was used to study the light yield
versus secondary dopant (3-hydroxflavone, 3HF) concentration. Fibre with 2500, 3500, and 5000
parts per million 3HF doping in this test. All fibres were first cut to length and then polished on
one end so that a vapour-deposited Al mirror could be applied. Although we have not measured
the quality of the mirrors on the fibres used in our test, the D0 experiment measured an average
reflectivity of approximately 90% for the fibres used in the D0 fibre tracker and we used the same
mirrroring procedure as was applied to the D0 fibre.
The ribbons were made following the technique developed for the D0 fibre tracker. A plastic
(Delrin) grooved mould was first fabricated (see Figure 6.5.7). The mould was measured on a
coordinate measuring machine and the mean groove pitch was determined to be 419 m. Our
target groove pitch was 420 micron (pitch/diameter = 1.2). A teflon release film (25 micron) was
first pressed into the mould with the aid of vacuum (pump-out holes were drilled into the grooves
in the mould). A tack adhesive was then sprayed on the teflon and the first layer of fibres was
placed in the mould. A circlular stop fabricated from a plastic sheet was placed over the mould in
order to form a ribbon with the proper circular active aperture. After the first layer of fibre was in
the mould the spray adhesive was applied to the fibre and the second layer of fibre (forming the
doublet) was placed on top of the first layer. A polyurethane adhesive was then spread over the
fibres and finally a 25 m mylar film was placed over the assembly. The assembly was then
clamped under pressure during an overnight adhesive cure. The resultant ribbon was removed
from the mould with the release film still attached. The final step in the ribbon fabrication was to
carefully remove the release film from the ribbon.
Figure 6.5.7 Schematic drawing of the fibre-double layer laid in the delrin mould.
6.5.4.2.2.3 Station assembly – GB
6.5.4.2.3 Optical connectors
6.5.4.2.3.1 MICE optical connectors at station – GB
6.5.4.2.3.2
MICE optical connector at patch panel – bulk-head connector – GB
6.5.4.2.3.3 D0 optical connectors at VLPC cryostat
The optical signal from the tracker is piped to the VLPC system via the fiber waveguides. The
waveguides terminate in the D0 warm-end optical connector shown in Figure 6.5.8. This is an
injection-molded part made of Delrin. Shown are the 128 holes for fibers, two holes (left/right)
for alignment pins and two holes (up/down) for threaded inserts. The typical optical throughput
for this connector interface is approximately 98%. Since MICE will use waveguide fiber of 1.05
mm diameter and the D0 cassette uses fiber of 0.965 mm diameter, we will use approximately
15% of the light due to this mis-match.
Figure 6.5.8 D0 warm-end optical connector
6.5.4.2.4 Clear-fibre light guides – MY
6.5.4.2.4.1
Light guides inside the tracker volume – MY
6.5.4.2.4.2 Light guides between optical patch-panel and VLPC cryostat – MY
6.5.4.2.5 Optical patch-panel and vacuum seal – GB
6.5.4.2.6 Tracker assembly and quality assurance procedures – ME
6.5.4.2.6.1 Tracker assembly – GB/PC
6.5.4.2.6.2 Quality assurance: ribbon manufacture – AB
6.5.4.2.6.3 Quality assurance: station assembly – ME
6.5.4.2.6.4 Quality assurance: tracker – GB/PC
6.5.4.2.7 Spectrometer solenoid, scintillating-fibre tracker integration – GB
6.5.4.2.8 Programme of prototype development and measurement of key parameters –
KL
6.5.4.3 Readout electronics – AB
6.5.4.3.1 Overview
MICE will use the D0 central fibre tracker (CFT) optical readout and electronics system. This
system has been operating reliably for the D0 experiment for almost 4 years now. The
photodetector is the visible light photon counter (VLPC) manufactured by Boeing. The VLPCs
operating at 9 K and thus require a cryogenic system. The VLPCs are packaged into a cassette
which contains 1024 channels. Two analog front-end boards (512 channel each) provide readout,
temperature control, and VLPC Bias.
6.5.4.3.2 VLPC system
The Visible Light Photon Counter (VLPC) is a cryogenically operated silicon-avalanche device.
The operation and development of has been discussed extensively in the literature [VLPC]. It is a
descendant of the Solid State Photomultiplier, an impurity band silicon avalanche photodetector.
It has undergone six design iterations, specified as HISTE I - HISTE VI. HISTE VI is the version
used in the DØ CFT. It is an eight element array in a 2 by 4 element geometry. Each pixel in the
array has a diameter of 1 mm. The HISTE VI operational parameters are given below:

Quantum Yield > 0.8

Gain > 40,000

Operating temperature 9K

Operating Bias 6-8V
6.5.4.3.2.1 VLPC cryostat
Since the VLPCs operate at cryogenic temperatures, a liquid Helium cryosystem is required. Our
current baseline for the VLPC cryo-system is to use Gifford-McMahon cryo-coolers to maintain
the 9K operating temperature for the VLPCs. The design work for this system has just started,
but we believe that commercial GM coolers are a cost-effective approach to the VLPC cryo
needs. If cryo-coolers are not used in MICE then the VLPCs will share the Helium refrigerator
with the solenoid magnets. In this case MICE will copy the cryostat design details from those
used in D0, but will build a 2 cassette cryostat instead of a 4 cassette cryostat. Four cryostats will
be used in MICE, each cryostat holding two cassettes (which will read out one half of one of the
trackers). Two cryogens are used in this system. Liquid Helium from the control dewar will
allow for VLPC operation at 9 K and liquid Nitrogen will cool an intermediate heat intercept in
the VLPC cassette in order to reduce the heat load to the liquid Helium. The cassette cold end
will sit in a stagnant gaseous Helium volume. Conduction through the gas will cool the VLPCs.
The temperature stability specification the VLPC cryostats (either GM cryo-coolers or liquid) is
 50 mK.
6.5.4.3.2.2 VLPC cassettes
The VLPC cassette contains 1024 channels of VLPC readout and is divided into 8
modules of 128 channels each which are interchangeable and repairable. This is illustrated in
Figures 6.5.9 and 6.5.10. Figure 6.5.9 shows the full cassette with readout boards attached.
Figure XXX shows the inner components of the cassette, with the readout boards and cassette
body removed. Both figures clearly show the 8-fold modularity of the cassette design.
Figure 6.5.9 The VLPC cassette with readout electronics board attached
Figure 6.5.10 Inside view of the VLPC cassette with cassette body removed.
Sitting directly over each VLPC pixel is an optical fiber which brings the light from the detector
to the VLPC chip. Each cassette module is comprised of an optical bundle assembly, a cold-end
electronics assembly, and an assembly of mounted VLPC hybrids. The cold-end assembly is
designed to be easily removable for repair without disturbing other modules due to the high cost
and delicate nature of this device. Another important design requirement of this cassette regards
the read-out electronics. Due to the nature of accelerator operations, the readout electronics
boards, the PC boards which act as interface to the data acquisition system, must be removable
and replaceable without removing a cassette from the cryostat. The readout electronics are
discussed in detail in the following section.
The cassette, for purposes of discussion, is broken down into several major components.
The cassette is distinguished as having a "cold end", that portion of the cassette which lies within
the cryostat, and a "warm end", the portion of the cassette which emerges from the cryostat and
is at room temperature. At the cold end, eight cold end assemblies each of 128 channels of
VLPC readout are hung from the feed-through by the optical bundles and are surrounded with a
copper cup at the cold end. Each cold-end assembly consists of sixteen 8 channel VLPC hybrid
assemblies, the "isotherm" or base upon which they sit, the heater resistors, a temperature
measurement resistor, cold end flex circuit connectors and the required springs, fasteners and
hardware. Running within the cassette body from top to bottom are eight 128 channel optical
bundle assemblies which accept light from the detector wave guides connected to the warm end
optical connectors at the top of the cassette and pipe the light to the VLPC's mounted at the cold
end (see Figure 6.5.10). The electronic read-out boards are located in rails which are mounted to
the warm end structure and are connected electrically with the cold end assemblies via kapton
flex circuits. In addition, the electronics boards are connected to a backplane card and backplane
support structure by card edge connectors and board mount rails. The flex circuits and read-out
boards are electrically and mechanically connected by a high density connector assembly.
The cassette body can also be broken down into cold end and warm end structures. The
cold end structure is broken down into several sub-assemblies: namely the "feed-through
assembly", the G-10 walls, the heat "intercept" assemblies and the cold end copper cup (see Fig.
XX). Along the length of the cold end, two heat intercepts are integral to the cold end cassette
structure. The first is the liquid nitrogen intercept (77 Kelvin) which serves to cut off the flow of
heat from the warm end. The second is the liquid helium intercept, with a name more historic
than functionally descriptive, which serves as an IR suppression device and terminating structure.
The warm end structure is mad of parallel aluminum plates spaced by spacer bars which form a
protective box for the optical bundles.
6.5.4.3.2.3 AFE boards
MICE will use the D0 AFEII boards in order to read out the VLPC system. This new electronics
board is currently under development by D0. It will include the following:
1. Front end preamplifier with 48 bin analog pipeline
2. Commercial 8bit 20 MHz flash ADC
3. Discriminator outputs for each channel
4. FPGA for each module of 32 channels
5. Temperature control circuitry for each cassette module (control and heater)
6. VLPC bias circuitry with voltage and current read back
Each board will service 512 channels or ½ of one cassette. MICE plans to also use the D0 VME
architecture for readout of the AFEs into VME. This is performed by the sequencer module. In
the case of D0, the readout is synchronized with the accelerator crossing clock. For MICE this
will not be the case. We will run asynchronously but triggered by beam particles.
6.5.4.3.3 VME system
6.5.4.3.3.1 Digitisation
6.5.4.3.3.2 Data transmission
6.5.4.4 Software – ME
6.5.4.4.1 Simulation
Details of the Monte Carlo simulation that was performed to study the fibre tracker’s performance
in a high-background environment are given in Section 9; a few salient points are given here. The
five-plane fibre tracker is very insensitive to photon backgrounds from the RF cavities and
maintains its excellent tracking performance at rates significantly higher (260 kHz/cm2) than that
expected in MICE based on current measurements. Representative results for the measured helix
radius are shown in Error! Reference source not found. for incoming (upper plot) and outgoing
(lower plot) muons. There is no cooling channel in this simulation (which corresponds to step III
in Figure 3.13) so both plots should be identical. We see that the distributions in (generated minus
reconstructed) radius are unaffected by background photons over a wide range in photon rate,
illustrating the robustness of this tracking approach.
6.5.4.4.2 Reconstruction
6.5.4.4.3 Sensitivity to background
6.5.4.5 Cost and schedule – KL/AB/YK/GB/ME
6.5.5 Alternative: time projection chamber with GEM readout – ER
ER to define level of detail and organisation of this section.
6.6 Downstream PID (e- separation)
In a small (~1%) fraction of events, a muon decays inside the cooling section or one of the
spectrometers. The resulting electrons bias the emittance measurement considerably and must be
rejected. Momentum distributions of electrons and muons arriving downstream of the second
spectrometer are shown in Figure 6.6.1.
Figure 6.6.1: Momentum distributions of muons and electrons downstream of the second spectrometer [Janot01].
Kinematics cuts can reject about 80% of decay electrons, but this rejection is not sufficient to
avoid a bias in the emittance measurement. Dedicated detectors are needed to separate electrons
from muons. The strategy is as follows:
a) positively identify muons by requiring low and longitudinally uniform energy deposition
in an electromagnetic calorimeter at the very end of the experiment
b) reject any residual background of electrons in the muon sample defined by the
calorimeter by means of a threshold Cherenkov detector immediately upstream of the
calorimeter
Geometrical features of these particle identifiers are determined by the angular and spatial
distributions of particles at the end of the experiment. The profiles in position and angle 50 cm
downstream of the solenoid, obtained from simulations, are shown in Figures 6.6.2 and 6.6.3. A
geometrical aperture of order 1 m is needed to take into account the defocusing of particles in the
stray magnetic field of the downstream solenoid.
Figure 6.6.2: Particle spatial distributions in
a plane perpendicular to the beam axis
[Janot01]
Figure 6.6.3: Particle angular distributions
in a plane perpendicular to the beam axis
[Janot01]
Geometrical features of these particle identifiers are determined by the angular and spatial
distributions of particles at the end of the experiment. The profiles in position and angle were
obtained from simulations, at realistic longitudinal positions of the downstream detectors along
the general MICE beam axis. These positions are sketched in Figure 6.6.4. The geometrical
aperture of all devices will match these profiles.
Figure 6.6.4. Longitudinal positions used in this note for the downstream detectors. Transverse sizes are drawn to
scale according to the latest dimensions. (Phone conf. PC73 on March 17, 2004). A 5-cm clearance is assumed
between detectors. The green element at left is the last correction coil whose center serves here as the origin of the zaxis.
THE x-y profiles at the three relevant depths FROM YAGMUR’s LAST
PRESENTATION COULD BE PUT HERE
6.6.1 The electromagnetic calorimeter
A state-of-the-art, high-resolution electromagnetic Pb-scintillating fibre calorimeter, of the type
built by KLOE [KLOE], is proposed. It offers adequate energy resolution to perform muon and
electron identification in the momentum range of interest for MICE.
6.6.1.1 Structure and layout
The proposed calorimeter, built by gluing 1-mm-diameter blue scintillating fibers between 0.3
mm thick grooved lead plates, has uniform and quite symmetric Pb–scintillating fiber structure,
with fiber spacing of 1.35 mm. When layers are superimposed, fibers are located at the vertices of
adjacent quasi-isosceles triangles, forming a homogeneous and compact structure with a fiber :
lead volume ratio of 2 : 1. With respect to the KLOE-standard mixture 1 : 1 , the choice is
dictated by the softer muon spectrum to be detected in MICE.
The resulting composite, with a density of ~ 3,7 g cm–3 , a Moliere radius of ~ 3, 5 cm and a
radiation length X0 of ~ 2,5 cm , gains considerable stiffness, and can be easily machined to the
shape required for the final assembly.
This ‘spaghetti’ design offers the possibility of fine sampling and results in optimal lateral
uniformity of the calorimeter. Fibers run mostly transversely to the particle trajectories, reducing
sampling fluctuations due to channelling, i.e., showers developing along the fibers’ direction, an
effect particularly important at the low energies of interest. Finally, the very small lead foil
thickness (~ 0.05 X0) results in a quasi-homogeneous structure and a high efficiency for
minimum ionizing particles and low energy electrons.
A calorimeter module of 72 x 72 cm2 of active area, 16 cm thick, consisting of ~ 180 lead and
fiber layers, can be built using the facilities of the Frascati (LNF) workshop formerly used for the
construction of the KLOE electromagnetic calorimeter [KLOE]. The lead and fiber planes are
perpendicular to the beam axis. The calorimeter is structured in four layers, each made of 18 cells
4 × 4 cm2 wide and 72 cm long. Its basic features are shown in Figures 8.1 and 8.2. (tbd)
6.6.1.2 Fabrication of components
The grooved lead foils are obtained by rolling 0.3 mm thick foils in a special shaping
machine.The drawing of the lead foil cross section where the fibers are glued is shown in Fig.
6.6.5.
Figure 6.6.5 Detail of Pb foil
Based on KLOE experience, the optical fibers will be Pol.Hi.Tech. type 044 [POL], emitting in
the blue-green region and meeting, at reasonable cost, the MICE technical specifications. Great
care will be taken to maximize the efficiency of the light collection system and to ensure uniform
photocathode illumination. KLOE experience suggests light guides consisting of a tapered mixing
part, where the quadrangular entrance face transforms smoothly to its inscribed circle, plus a
Winston cone concentrator [Welford], matching the area of the calorimeter element to the
sensitive area of the photocathode face. If it is possible to shield against the residual magnetic
field, a suitable readout choice, at both ends of each cell, is the 1-1/8 inch R1355 Hamamatsu
phototube. This tube, already used in the HARP experiment, has a transit-time spread less than 1
ns. In this case, the area concentration factor is ~ 2,65 , with a light collection efficiency of ~ 85
%. Figures 6.6.6 and 6.6.7 show the housing of voltage divider , photomultiplier and light guide
and the assembling with the fiber-lead module. (tbd)
Figure 6.6.6 Voltage divider, PMT, and Light Guide Housing
Figure 6.6.7 EM Cal Assembly
6.6.1.3 Front-end electronics
The deposited energy will be digitized by conventional ADCs, e.g., the VME CAEN 792, already
used in the HARP experiment. An unbiased cross-calibration of all calorimeter readout elements
will be obtained both from cosmic rays and test-beam muons. A rough measurement of the
impact point on the calorimeter will be derived from the ratio of pulse heights on two ends of the
fired cell. A more precise reconstruction of the impact point and an independent timing
information for trigger purpose can be achieved implementing a TDC readout of the first layer of
cells.
6.6.1.4 Energy and timing resolution, particle identification.
The fibres have a decay constant of ~ 2.5 ns, an attenuation length of ~ 3.5 m, and a yeld of about
80 photoelectrons per minimum ionizing particle (mip) crossing the center of one of the 40 × 40
cm2 calorimeter cells. Based on the response of the KLOE electromagnetic calorimeter [Adinolfi]
to electromagnetic showers of similar energies at the DAPHNE Φ-Factory, the expected energy
resolution is σE ~ 5% /)(GeVE, fully dominated by sampling fluctuations. The time resolution
was measured in KLOE, with modules 4 m long, and is also very good, σt ∼ 54ps /)(GeVE ,
giving a precise measurement of the impact point along the fiber. The KLOE calorimeter test
beam data [Antonelli] show that distributions of total energy are Gaussian, with almost no tails.
The response of the calorimeter to electrons and photons in the energy range 20–300 MeV is
linear, independent of incident angle [Antonelli]. The signal deposited by a minimum ionizing
particle in a single cell is equivalent to that of a 30 MeV electron or photon. In the momentum
range of interest for MICE, muons mostly punch through, whereas electrons leave basically all
their energy in the first two layers. In combination with the downstream Cherenkov, the
measurement of energy deposition and of its longitudinal profile in the calorimeter should
provide an electron rejection of 10–3. A preliminary simulation of the EmCal performance on
muon detection and muon - electron separation with different selection cuts has been performed.
The results , based on the beam profile and energy spectrum generated by G4MICE (version :…)
, are shown in Figures 6.6.8 and 6.6.9 (tbd). Cosmic ray tests are foreseen together with tests of
detector response to low energy electrons, hopefully at the Beam Test Facility in the DAPHNE
complex of LNF [TBF].
Figure 6.6.8 Muon detection efficiency at EM cal as a function of muon momentum
Figure 6.6.9 Electron-muon separation in EM Cal using baricenter method
6.6.1.5 Construction schedule
INFN technical manpower will mould the light guides, and design and fabricate the mechanical
supports for photomultipliers and patch panels. Some substantial refurbishing and tuning of the
KLOE tooling machinery will be needed. Lead foils grooving, construction of light guides,
production of fibers, gluing of layers of lead with fibers, machining of the assembled modules
and final gluing of light guides will be contracted to external firms. The completion of the
construction project will take one year after approval of funding. 8.6.6 Costs The expenditure
needed to build the MICE calorimeter – 16 cm thick – is estimated to about 120 kEuro plus 30
additional Keuro for ADC’s. This includes the purchase of lead foils, fibers, glue, light guides
and relative workings. The entire readout chain , except the ADC’s , worth about 150 additional
Keuro (photomultipliers & voltage dividers, mu-metal shieldings, cables, VME crate) and the
high voltage system, worth about 125 Keuro (main frames and power supply modules) will be
recuperated from HARP.
6.6.2 The Downstream Cerenkov
Further electron rejection will be provided by the downstream aerogel Cherenkov detector. In
principle, this detector should not be affected by background from the RF cavities, as X-ray
energies will rarely be high enough (above 2 MeV) to produce electrons that give Cherenkov
light in the aerogel. The RF noise at 201 MHz has a skin depth of 6 m in aluminium. But a
ferromagnetic shielding is however necessary to be protected it against the stray magnetic field.
6.6.2.1 Description and Performance
In the momentum range of interest to MICE, aerogel (1.01 < n < 1.06) appears to be the only
adequate radiator from which to build a threshold Cherenkov blind to the passage of muons. The
choice of the appropriate index of refraction for the radiator is governed by the relative light
yields of electrons and muons, and their respective detection efficiencies assuming a fixed
detection threshold. The goal is to maximize the response to electrons while minimizing possible
contributions from higher energy muons. The aerogel radiator proposed here has an index of
refraction n = 1.03 and a total thickness of 10 cm. The maximum transverse size of the radiator is
about 55 x 55 cm2. This geometrical aperture is deliberately chosen to cover the beam "spot" of
muons to within 3.
The upstream face of the aerogel radiator is supposed to be located at z= 500 mm from the last
correction coil and the downstream face of the Cherenkov vessel is at z= 1010 mm.
The transverse size has been evaluated on the basis of GEANT4 simulations. It is assumed that a
sufficient containment is reached in the 3 limit. In their present status, the simulations do not
take into account the presence of a thick magnetic shielding between the last correction coil and
the downstream detectors. This implies that the limits so deduced are "worst case" values. They
are given in Table 6.6.1 and sketched in Figure 6.6.10.
Detector
RMS (in cm)
TOF 2
10
Cherenkov 2
14
Calorimeter
33
Comment
at the upstream face
of the aerogel
Table 6.6.1 RMS beam sizes obtained by simulation (without shieldings)(Y. Torun, PC73).
We thus based the mechanical design of CKOV 2 on a transverse size larger than 3 = 42
cm. The longitudinal dimensions partly depend on the diameter of the chosen photomultipliers.
Figure 6.6.10 Muon distributions in a plane perpendicular to the beam axis [Y. Torun] Top row: x-, px- and x'distributions. Bottom row: y-, py- and y' distributions. All horizontal axes are in millimeters.
In principle, this detector should not be affected by background from the RF cavities, as
X-ray energies will rarely be high enough (above 2 MeV) to produce electrons that give
Cherenkov light in the aerogel. The RF noise at 201 MHz has a skin depth of 6 m in aluminium.
But a ferromagnetic shielding is however necessary to be protected it against the stray magnetic
field.
In the momentum range of interest to MICE, aerogel (1.01 < n < 1.06) appears to be the
only adequate radiator from which to build a threshold Cherenkov blind to the passage of muons.
The choice of the appropriate index of refraction for the radiator is governed by the relative light
yields of electrons and muons, and their respective detection efficiencies assuming a fixed
detection threshold. The goal is to maximize the response to electrons while minimizing possible
contributions from higher energy muons. The aerogel radiator proposed here has an index of
refraction n = 1.03 and a total thickness of 10 cm. The maximum transverse size of the radiator is
about 55 x 55 cm2. This geometrical aperture is deliberately chosen to cover the beam "spot" of
muons to within 3.
Text above was used elsewhere
6.6.2.2 The environment of the downstream detectors
Due to the large bore of the solenoids, it is obvious that there is a rather strong stray field
extending far away from the last coil and giving rise to residual fields unacceptable for some of
the downstream photomultipliers (Figure 6.6.11). Detailed studies have sought geometry with the
smallest shield while still keeping tolerable remaining fields (less than 2 kGauss) at the sensitive
parts of the downstream subdetectors.
Figure 6.6.11 Magnetic fields at the photocathodes of the downstream detectors without any shielding material. The
colored dots show the field intensities at the respective detector (photocathode) positions.
It was shown that an optimal solution (smallest transverse size, reasonable thickness) is a 15-cm
thick soft iron disk with a diameter of 150 cm. It has a circular opening 50-cm in diameter along
the beam axis and is located 20 cm downstream of the last superconducting coil.
The longitudinal positions of the downstream detectors and the approximate radial distances (to
beam axis) of the various photocathodes are taken from Figure 1 and summarized in Table 2.
Sub-detector
Longitudinal position
Radial distance to beam axis
z (cm)
r (cm)
TOF 2
45
26
CKOV 2
83
38
114
36
Calorimeter
Table 6.6.2 Positions of the photocathodes of the downstream subdetectors.
We then evaluate (z-r) maps of the stray magnetic field at these positions.
a) at the longitudinal position corresponding to TOF2 (z=45 cm), the remaining magnetic field is
less than 1000 Gauss (Figure 6.6.12).
Figure 6.6.12 The magnetic field as a function of the radial distance to the beam axis at the position of TOF2 (dashed
vertical line). The red(green) curve is obtained with(out) the MICE shielding.
b) at the longitudinal position corresponding to CKOV2 (z=83 cm), the remaining magnetic field
is less than 600 Gauss (Figure 6.6.13).
Figure 6.6.12. The magnetic field as a function of the radial distance to the beam axis at the position of CKOV2
(dashed vertical line). The red(green) curve is obtained with(out) the MICE shielding.
c) at the longitudinal position corresponding to CALO (z=114 cm), the remaining magnetic field
is less than 300 Gauss (Figure 6.6.13).
Figure 6.6.13 The magnetic field as a function of the radial distance to the beam axis at the position of CALO (dashed
vertical line). The red(green) curve is obtained with(out) the MICE shielding.
One general consequence is that the remaining fields at the various detector positions can still be
lowered with standard shields (typically 5 mm soft iron + 1-mm thick mumetal).
The gas-tight external box enclosing the optical system will be constructed from nickel-plated
(against rust) soft steel (for magnetic shielding). Using an cylindrical approximation for the shape
of this vessel, the remaining inside field was computed at the position of the photocathode (Figure
6.6.14).
Figure 6.6.14 Field lines inside an approximate Cerenkov vessel at the position of the photomultipliers.
Without the mumetal shield, the residual field is 30 Gauss at most. It is also seen that the field
lines are nearly parallel to the PM axis: their influence is thus even smaller on the photoelectron
collection efficiency.
Additional precautions (thin lead shielding and 10-m-thick copper cladding) may be necessary
to account for the possible X-ray and RF backgrounds.
6.6.2.3 Detailed description
6.6.2.3.1 Mechanical Layout
A perspective view of a partially assembled containment vessel is shown in figure 6.6.15.
Figure 6.6.15 Perspective view of the skeleton iron structure for the Cherenkov 2 vessel.
The structure is constructed by welding Nickel-plated soft steel sheets 5 to 15 mm thick.
The use of soft steel is needed as it shields the photo-detector from the stray field of the solenoid.
Whenever possible all parts should be welded and separately checked against leaks. Four
stainless steel "bridges" protrude upstream: they are intended to lean against the thick iron
shielding and to withstand the forces generated by the spectrometer and end coil. The inner
sidewalls of the box are covered with rectangular optical glass windows. These windows are
completely aluminized except in front of the photocathodes. The shallow volumes on all four
lateral sides are equipped with Winston cones. The sizes of these cones allow a limited freedom
to modify the overall acceptance of the instrument (green elements in Figure 6.6.16).
Figure 6.6.16 The Winston cones (green) and a couple of PM tubes at their position around the CKOV 2 box.
Next, the Winston cones and PM's are covered by a double layer shielding (5 mm soft + 1-mm
Armco layer) (Figure 6.6.17).
Figure 6.6.17 The final magnetic shielding in place (iron in grey color, Armco sheet in yellow).
The entrance and exit windows (not shown here) for the particles will be made with 7-mm thick
honeycomb panels with thin aluminium skins. Details of the assembly are shown in Figure
6.6.18. O-rings guarantee the gas tightness along the windows, the radiator box and along the
edges of the entrance and exit windows. Transverse and longitudinal views are shown in Figures
6.6.19 and 6.6.20, respectively.
Figure 6.6.18 Detailed drawing of a corner of the Cherenkov 2 vessel in the vicinity of the radiator box (at right) with
the particle entrance window (7 mm honeycomb). The edge of a Winston cone and part of a PM are visible at right.
The assembly is reasonably straightforward with four different pieces, prepared separately and
bolted together with O-rings at the appropriate places:
a. The radiator box is assembled, loaded and closed in the laboratory. Its
upstream face, made of 7-mm thick honeycomb, is also the entrance face
for the particles.
b. A "central" box containing the large tilted mirrors and their orientation
controls. This box also supports the glass windows. It is obtained by
welding flat sheets of steel.
c. Four identical assemblies containing the two PM tubes each, their
mumetal shielding and the HV circuitry.
d. The loaded radiator box is bolted to the vessel. Its downstream panel is
then removed.
e. The whole CKOV2 box is then closed by the particle output window (7mm thick honeycomb panel) and flushed with helium.
Figure 6.6.19 Transverse view of CKOV 2 (meridian cut). The green-hatched area at left is the aerogel radiator with
its safety protective window (thick blue line) for transportation only. The optical mirrors/windows are drawn in red.
The two lines at 45° to the beam axis are cuts through the mirrors reflecting light upwards or downwards (Dimensions
are given in millimeters).
Figure 6.6.20 "Longitudinal view of CKOV 2 (as seen by the incoming muons). Only half of the PM tubes and Winston
cones are shown. The blue-hatched lines in the center outline the lightweight (honeycomb) supports for the tilted
mirrors.
6.6.2.3.2 Aerogel radiator
The aerogel radiator has an index of refraction n=1.03. Stacking 250 tiles of aerogel of 11 x 11 x
10 cm³ define a maximum radiator size of 55 x 55 x 10 cm³. The manufacturer will start by June
2004 the delivery of a so-called improved "hydrophobic aerogel" less prone to degradation with
humidity. The thickness of the radiator is chosen on the basis of the small photoelectron yield
with low energy electrons. It is assumed that the aerogel wall will be assembled in the lab
(outside the beam area) and maintained in a gas tight container under helium. Upon completion of
the assembly of the experiment, this closed container is bolted as such in place to the Cherenkov
vessel. The downstream wall of the container can then be easily and rapidly removed (from the
downstream side) and the vessel closed with the tilted mirrors. This last operation takes only a
couple of minutes to avoid the degradation of the aerogel due to the ambient moisture while
filling the vessel with dry helium gas at atmospheric pressure..
6.6.2.3.3 Light collection walls and mirrors
For reasons of availability, cost and handling, the inner surfaces of the vessel potentially hit by
light are covered by optical glass plates (7 mm thick Schott B270), aluminized everywhere except
in front of the photomultipliers. Plane mirrors, tilted at 45°, reflect light at 90° to the optical
beam axis towards the photodetectors. The mirrors will be made from aluminized 3-mm thick
polycarbonate (Lexan) plastic sheets. In order to keep the symmetry of the device around the
beam axis, the four mirrors are assembled in a reflecting pyramid (visible as the yellow object in
Figure 6.6.21).
Figure 6.6.21 The 45° reflecting pyramid is attached to the downstream face. The mirrors are supported by a
lightweight honeycomb structure.
The reflectivity of the aluminized plastic sheets is taken to be about 90% (as used for the
Cherenkov detector of the HARP experiment). The reflectivity of the aluminized plastic sheets is
taken to be about 90% (as used for the Čerenkov detector of the HARP experiment). It is however
possible (A. Braem, private communication) to benefit from recent developments based on a
multilayer of Aluminium, Magnesium fluoride MgF2 and Hafnium oxide HfO2 to reach
reflectivities as high as 96% in the interesting visible domain (250 to 500 nm). (Figure 6.6.21).
Figure 6.6.21 Reflectivity of a multilayer Al + MgF2 + HfO2.
During the initial design phase of this project, we studied the performances of curved
(cylindrical) mirrors. The goal was obviously to improve the light collection efficiency. The
results were not better than for plane mirrors. The reasons come from the very large "object" size
and large divergence of the photons generated in the radiator. At the same time, the longitudinal
size of the detector has to be kept as small as possible. The optics looks then more to an
"illumination" system than to a "focusing/imaging" or "paraxial" optical system. The single
remaining design goal is thus to minimize the total number of reflections on the walls.
6.6.2.3.4 Photomultipliers
The system uses EMI 9356 KA 20-cm (8")-diameter photomultipliers, with a standard bi-alkali
photocathode. These very low noise PM's are those used by a former gas Cherenkov detector
built for the HARP experiment. This PMT has the advantage of a rather large gain, 3 x106, well
matched to the low light yield of aerogel radiators. The PM's are mounted in volumes separated
from the radiator section. It allows easy access to the PM's without disturbing the dry
environment of the aerogel and avoiding possible damages to the mirrors.
6.6.2.3.5 Estimated optical performances and detection efficiency.
The optical characterization of this updated setup has still to be performed and compared with the
initial proposal. As for the MICE proposal, it will be done by accurate 3D optical ray tracing,
taking into account realistic surface properties of the mirrors (spectral and angular reflectivity),
bulk scattering inside aerogel materials, transmittance of the window and the typical quantum
efficiency of standard photomultipliers. The overall light collection efficiency is expected to
reach about 80%. On the basis of the light collection and the photoelectron yield, the detection
efficiency for electrons versus their momentum could then be evaluated. This essential task
should start as soon as realistic simulations of the beam profiles for electrons and muons are
available, including the influence of the shieldings.
6.7 DAQ, trigger, on-line monitoring
6.7.1 Beam structure and trigger
The ISIS beam delivers a 1 ms spill of about 3000 pion bunches, each 100 ns long and separated
from its neighbours by 224 ns, with a repetition rate of few Hz. The RF power source can be
operated with a duty-factor of about 10–3. Using it in the most efficient way delivers a flat top of
850 s at 1 Hz, during which 2600 bunches will reach the experiment.
The basic enable signal to data-taking will be provided by the start of operation of the MICE RF
cavities (Start-Of-RF, or SORF). It will last until the End-Of-RF (EORF) signal, for a data-taking
gate of typically 500 s. (Because of the high level of multipacting during the RF transient, the
detectors must be insensitive during the rise- and fall-time of the RF pulse, so the data-taking gate
is shorter than the RF pulse.) This gate will be generated by the MICE trigger system.
The time structure of the beam is such that there is not enough time for particle-by-particle
readout. For this reason, all digitisers will be buffered during the spill and read out by the DAQ
system at the end of the data-taking gate. The storage capabilities of the digitisers may put limits
on the duration of this gate. ADC gates and TDC stops will be generated by simple coincidences
of TOF planes. The RF phase and characteristics of the RF pulse will be recorded at the same
time as the muon events.
For calibration and reference purposes, one or several SORF/EORF data-taking gates can be
generated, with beam but without RF, between the real SORF/EORF cycles.
The trigger system will be designed to provide enough flexibility to allow the detectors to run in
stand-alone mode for set-up, test and calibration purposes. Specific tests for the choice of a
suitable trigger receiver may be performed during the TPG test at CERN in 2003.
6.7.2 Working hypothesis: event rates and sizes





The maximum beam rate will be 1 particle per bunch. This is a limiting case, in which
there will often be more than one particle in a 100 ns time window.
One calibration (RF off) cycle will be taken per normal (RF on) cycle
The upstream spectrometer will have to measure out-of-acceptance and out-of-RF-phase
particles
For the purposes of evaluating the data rates, it has been assumed that only 1/4 of the
incoming particles will fall into the acceptance after the diffusing plates and will be able
to reach the downstream spectrometer.
While the Sci-Fi tracker is relatively insensitive to the x-ray background, the data size
from the TPG may change dramatically as the background conditions vary. For this
reason, the total data volume from the TPG has been evaluated for the case of 50%
occupancy over the whole detector.
6.7.3 Readout volumes and network structure
All the digitizers will be housed in VME crates, either 6U or 9U. As can be seen from Table 6.2,
no crate exceeds the maximum sustained rate attainable on a standard VME bus.
Table 6.2: Data volume for the MICE detectors.
Subsystem
Sci-Fi
TOF
e/-ID
# VME crates
30
1 front + 1 rear
1 rear (same crate as TOF?)
Data volume
[MB/s]
≤2
≤ 0.5
Negligible
Data volume/crate
[MB/s]
≤ 0.5
≤ 0.25
—
Each crate will be equipped with a VME processor, which is a single-board computer provided
with a VME interface and a network connection. The task of the processor will be to read data out
of the digitizers via the VME bus, and to pass them over the network to the Event Builders.
As is now commonplace, the MICE DAQ system will make use of a switched Ethernet network
for transferring data to the Event Builders and the Storage System, to synchronize and control the
DAQ processes running on several nodes and to perform on-line monitoring tasks.
To cope with the total rate of the experiment and to allow for redundancy, the DAQ will comprise
at least 2 Event-Builder (EVB) computers. Their role will be to assemble event fragments
produced by the many VME readout processors, make consistency checks, write assembled
events to the Storage System, and serve live data to the On-Line Monitoring System. In order to
minimize the cost of data storage, the possibility of building enough parallelism and computing
power into the EVB system to perform on-the-fly data compression will be considered and tested.
The most widespread storage solutions for scientific laboratories are rack-mount PC cases
holding several large-capacity, hot-swap E-IDE hard disks, connected to the PCI bus by
hardware-RAID controllers, with an available capacity of ~1 TB [Sanders].
The performance of such systems can be considered appropriate for the MICE data rate; however,
some care has to be taken since the on-line Storage System will not only provide space for storing
runs, but will also serve data for on-line analysis work. For this reason, the hardware will have to
be tested for concurrent sequential read/write access of large files before being considered
compliant with MICE rate requirements.
MICE will be able to cool, at most, 100 /s. To achieve a statistical precision of 10–3, one has to
take data for 1000 seconds. During this time the detector will produce up to 15 GB of data.
Storing a week of data, hundreds of runs, may require several TB. The availability of a
centralized storage system at RAL would be a great help by reducing the on-line storage system
complexity and eliminating the need for a high-performance backup system.
Several workstations will be available for monitoring data, either spilled on the fly from the Event
Builders, or read directly from the Storage System. The DAQ will have its own monitoring
library, useful for checking the main working parameters and performance of the detectors. In
addition, a proper design of the off-line software can offer a powerful tool to run the
reconstruction algorithms and a detector display on a sub-sample of events taken from the live
data stream.
The on-line cluster, including the VME processors, will use the Linux operating system. Since no
readout operations are foreseen during the spill, it is unlikely that the Linux real-time extension
will be needed on the VME processors; this matter will be subject to tests.
6.7.4 Cost
The following budget table (Table 8.5) is compiled under a few assumptions:



MICE will have to buy the VME crates
The cost of storage is evaluated at today’s prices, which will most likely drop in the next
year
The VME processors can be recovered from the HARP experiment at CERN, provided
tests show that they meet the MICE performance needs
Table 6.3: Cost estimate for the data acquisition system
Item
VME crates
VME processors
100 Mbit switches
Backbone Gbit switch
Event Builders
Mirrored disk buffers
Workstations + monitors
(control, monitoring, on-line analysis)
Number
10
10
2
1
Up to 4
51 TB
20
TOTAL
aNo
cost if recovered from the HARP experiment.
bReduces
to 220 k€ if VME processors available from HARP.
Cost (k€)
100
50a
5
10
15
40
50
270b
6.8 Analysis and Simulation
An essential goal of MICE is to measure the beam dynamics with unprecedented accuracy and
compare with simulations. Since particle detectors are interspersed with the beam manipulation
elements in the cooling channel (absorbers, solenoid magnets and rf cavities) and beam
preparation section (quadrupole magnets), it is essential to use a Monte Carlo framework in
which all are accurately modelled. To this end, the Geant4 toolkit [Agosti] from CERN was
selected as the most suitable base upon which to build the MICE software project which was
dubbed G4MICE. Geant4 is becoming the new standard for high energy physics experiments,
includes the functionality to support complex geometries as well as up-to-date models for physics
processes and is flexible and extensible. Although still in active development, G4MICE is
routinely used for studying the performance and optimizing the configuration of detectors in
MICE.
The analysis of MICE data will require specialized tools for creating "virtual bunches" in
software from a collection of single muons and exploring their properties in addition to the
pattern recognition, track reconstruction and particle identification tasks common in experimental
high energy physics. ROOT is another next generation tool [ROOT] currently used within
G4MICE for manipulation and analysis of simulation results.
6.8.1 Code Development Infrastructure
A central repository [IITWEB] of MICE software with strict version control has been set up early
in the project. Detailed documentation [DOXY] is automatically updated on the web to track
changes and the code is tested daily [CRONL] to flag problems. Regular releases [Softrel] are
made to consolidate fast paced development with the need for a stable simulation environment. In
addition to an email list, day-to-day development is coordinated through frequent phone meetings
and more in-depth discussion of global design issues are carried out at workshops [SM].
6.8.2 G4MICE Features
G4MICE already includes all major elements of the experiment in some detail. Further details
will be incorporated to replace some of the simplified geometry descriptions where appropriate
once the engineering designs converge. However care has been taken to accurately represent all
material in the beam path. Several hundred parameters [DataC] allow the user to adjust the details
of the simulation chain including the input beam, magnetic lattice, rf system, rf induced
backgrounds, absorbers, geometry, materials and electronics for detector systems, physics
processes and reconstruction and analysis algorithms. Magnets are modelled including the
materials and magnetic fields are calculated from the coil geometry and currents. Particles are
tracked with the time-dependent electromagnetic fields of rf cavities taken into account and all
relevant physics processes are included in tracking. Liquid hydrogen absorber vessels and
windows are modelled with the correct shape and thickness profiles. Geometry and materials are
implemented in detail for all the detectors.
The simulation chain includes inefficiencies and noise in the signal paths when digitizing hits to
form the raw output data and parameter values obtained from prototype tests are being fed back
into the model to improve realism. The full chain including reconstruction of tracks has been
implemented for the baseline scintillating fiber tracker. Detailed geometry and readout electronics
for the TPG option is also included in G4MICE. A fairly detailed rf background model has been
implemented to investigate the effect of backgrounds from field emission electrons originating in
rf cavities and will be using parameters measured from the commissioning of prototype cavities.
6.9 Integration
6.10 Safety
The detector systems in MICE present no extraordinary safety risks. The upstream and
downstream detectors are all based on PMT readout and only require standard PMT HV power
supplies and mains power for electronics. There are no hazardous materials in the detectors, no
flammable gas, and no systems that require high power or large drive currents. Standard
laboratory practices will be followed.
Since the fibre tracker is a totally passive device (at least within the tracking volume), its impact
on safety regarding operation of MICE is minimal. There are no active components in the
tracking volume. The only safety issue regards vacuum. Since the tracking volume will be
evacuated in order to reduce material presented to the muon beam, the optical feed-throughs and
the patch panel itself must be rated for pressure. A prototype patch panel will be fabricated and
tested for vacuum integrity and pressure in order to certify the basic design. At this point we do
not plan to test the final assemblies at pressure.
The fibre tracker readout system only requires standard line voltage (approximately 2 kW per
cryostat). The cyrogenic system is based on a closed-cycle refrigerator and therefore presents no
cryogen or oxygen deficiency hazard.
6.11 Conclusions
A complete set of detectors, capable of performing the measurement of emittance even in the
harsh background conditions of the experiment, has been identified. The resources and
competence exist in the collaboration to design, build and operate them. Given the crucial
importance of tracking, two options are presently under study. The set of crucial measurements to
be made has been identified, and is planned so as to allow a final decision on this issue in a time
frame that does not delay the rest of the experiment.
References
[Caval] M. Cavalli-Sforza et al. (SLD), IEEE Trans. Nucl. Sci. 37 (1990) 1132.
[Annas] E. G. Anassontzis et al. (DELPHI), Nucl. Instrum. Meth.A323 (1992) 351.
[Aubert] B. Aubert et al. (BaBar), Nucl.Instrum.Meth. A479 (2002) 1.
[Bartlett] D. Bartlett et al. (E691), Nucl. Instrum. Meth. A260 (1987) 55.
[Crema] L. Cremaldi and D. Summers, (2001)
http://www-mucool.fnal.gov/mcnotes/public/pdf/muc0221/muc0221.pdf.
[Ypsila] T. Ypsilantis and J. Seguinot, Nucl. Instrum. Meth. A343 (1994) 30.
[VLPC ] M.D. Petroff and M.G. Stapelbroek, IEEE Trans. Nucl. Sci. 36, No. 1(1989) 158. M.D. Petroff
and M.Atac, IEEE Trans. Nucl. Sci. 36, No. 1(1989) 163.
[Agosti] S. Agostinelli, et al., NIM A 506(2003), 250-303
http://geant4.web.cern.ch/geant4/
[ROOT] http://root.cern.ch/
[IITWEB] http://mice.iit.edu/cgi-bin/cvsweb.cgi
[DOXY] http://mice.iit.edu/software/doxygen/html/
[CRONL] http://mice.iit.edu/cgi-bin/cronlog
[Softrel] http://mice.iit.edu/software/release/release.html
[SM] http://mice.iit.edu/sm/sm.html
[DataC] http://mice.iit.edu/software/release/dataCards.html