The KARMEN Detector
Wenn Sie die deutsche Version
bevorzugen, klicken Sie
hier!
Figures for downloading
Central detector and inner anti counters
Inner passive shielding and shield counters
Neutrino-bunker
Veto counters
- This page describes the structure and the basic properties of
the KARMEN-detector, the anti counter systems inner anti, shield and
veto and the surrounding passive shieldings. For a better
understanding first the nomenclature and the position of the
different anticounters sides are introduced. The siedes of the
central detector and the anti-counters are called upstream,
downstream, left, right, bottom and top. Upstream refers to the side
towards the target, downstream to the opposite side. Expressed with
the points of the compass upstream is equivalent to north,
downstream to south, right to west and left to east.
Central Detector and Inner Anti Counters
- Due to the required good energy and time resultion and the
extremely low cross section of the neutrino reactions the
KARMEN-detector was constructed as a large volume liquid
scintillation calorimeter. The central part of the detector consists
of a 6 m high, 3.53 m deep and 3.20 m wide stainless steel tank
filled with 65000 l of a dedicated liquid scintillator developped at
the Forschungszentrum Karlsruhe. The liquid scintillator is made of
75% vol. parrafin, 25% vol. pseudocumol and an admixture 2 g/l of
the scintillator PMP (1-Phenyl-3-Mesityl-2-Pyrazolin). This mixture
provides a very high lightoutput of 7.2 photons/keV and a
attenuation length of 5 m at a wavelength of 425 nm wich perfectly
fits to the dimensions of the detector.
The central detector is segmented by double 1.5 mm thick lucite
sheets glued together and enclosing a thin air gap into a total of
512 central detector modules with a cross section of 18.1´17.7
cm2. This segmentation guides the isotropically emitted
scintillation light via total internal reflection very efficiently to
the module ends, where it is detected by a pair of 3 inch Valvo XP
3462 photomultipliers. The segmentation allows a strong concentration
of the produced light onto just two pairs of photomultipliers with
the advantage of a relatively high signal. Moreover this provides a
secure and simple position reconstruction. Perpendicular to the
module axis the position is determined by the module number, parallel
to the module axis the position is reconstructed from the measurement
of the time difference between the pulses registered by the
photomultipliers at each end of the module. In total 2048
photomultipliers are used for the read out wich are situated in
drilled holes of adjusted diameters in a 15 cm thick iron wall. The
photomultipliers a coupled optically to the scintillator through
glass windows and paraffin which circulates betweend the window and
the surrounding iron wall. The paraffin at the same time is used for
cooling and drains off the heat power of roughly 43 kJ per
photomultiplier and day. This allows to maintain a constant
temperature of 18.2±0.5 °C
in the central detector. Adjacent to the 16 columns and 32 rows of
central detector moduls there are one more row and column of in total
96 inner anti counters which only have half the cross section of the
central modules and are read out by only one photomultplier on each
end. These anti counters are used as the first stage of the anti
counter system and allow on four of six sides of the main detector to
identify and dismiss events intruding from outside like tracks of
cosmic ray muons and their successive decay products.
-
Optical fibers are connected to the module ends for calibration and
test purposes. Laser induced scintillation light can be fed into the
fibers to produce light flashes in the modules while the time and
amplitude is under perfect control. Among other things this laser
system is used for the calibration of the absolute time of the
events and to determine the time and energy resolution. The very
good energy resolution and the high portion of 96.5% active
scintillator in the total volume together contribute to a
spectroscopic quality in energy measurement.
- The air gaps of the optical segmentation of the central detector
except the outermost rows and columns contain paper coated with
Gd2O3. The gadolinium with an area density of
73.8 g Gd/m2 amounts to only 0.1% of the weight of the
central detector but contributes substantially to the detection of
thermal neutrons in the detector due to its extremely large cross
section of 49000 barns. During the capture of neutrons on gadolinium
a mean of 3 gammas with a sum energy of 8 MeV are emmitted resulting
in a visible signal in the detector. Beside this reaction the
neutron capture on the protons of the hydrogen with the emission of
2.2 MeV gammas is also used for the detection of neutrons.
Inner Passive Shielding and Shield Counters
- Outside the scintillator tank and the photomultiplier walls
follows the so called inner passive shielding of the
KARMEN-detector. This 18 cm thick and in total 180 tons heavy iron
shielding surrounds the detector on all six sides, provides
mechanical stability and protects the central detector from muon
induced background penetrating from the surrounding bunker. The
second stage of the anti counter system is mounted to this
shielding. It consits of in total 136 modules made of NE110 plastic
scintillator and is called shield detector. The shield counters are
3 cm thick, 30 cm wide and 2.4 to 3.1 m long. The counters are read
out at both ends with the help of 180°
light bending in order to achieve the most dense packing of the
counters which is possible. The light bending consits of a
reflecting aluminium foil glued to the module ends and a conical
light guide made of lucite which ends at a 2 inch EMI 9813 KB
photomultiplier. Before the KARMEN-upgrade the shield counters
surrounded the KARMEN detector on only five sides, during the
upgrade 8 additional veto modules have been installed below the
bottom of the shielding. The dense packing of the counters and the
good muon/gamma separation allows for an efficient detection of
muons entering the inner passive shielding. Thus the shield detector
provides an efficient suppression of muon induced background which
directly enters the central detector. Moreover background produced
outside the range of the shield counters, e.g. bremsstrahlung
produced by the michel electrons from the decay of stopped muons, is
strongly suppressed by the shielding and thus does also not enter
the central detector.
Neutrino-Bunker
-
All detector components described so far are surrounded by a 7000
ton heavy, 13.6 m long, 8.4 m wide and up to 10.6 m high iron
bunker, which completly eliminates the hadronic and electromagnetic
component of the cosmic radiation and reduces the muonic component
by about 60%. Moreover it serves as a shielding agains beam
correlated background caused by fast neutrons from the ISIS
spallation target. The inner clearance of the bunker is 10 m long,
4.20 m wide, 7.15 m high and is shut by a 600 ton door on the side
downstream of the target, which can be moved on rails. The bunker
may be entered throgh a maze at one side of the door for control and
maintenance. The bunker was constructed around a stiff cage more or
less like a house of cards mounting successive layers of iron slabs
up to 7.50 m high, 1.70 m wide and 18 cm thick wheighing up to 20
tons. After having set 2-3 wall layers the layers were secured by a
horzizontal roof slab connected to the walls by bolts. In total the
walls are build from 11 layers, each 18 cm thick, while the roof
consits of 17 layers which shorten above the 7. layer. Thus the
trapezoidal roof reaches a maximum thickness of more than 3 m. The
layers are numbered starting at the inside of the bunker.
- Only due to this layered construction was is possible to install
during 1996 in the framework of the KARMEN-upgrde to in stall an
additional, third stage of the anti counter system within the walls
and the roof of the neutrino-bunker. To achieve this roughly 2-3 of
the slabs had to taken away and pre-produced frames containing 2-3
veto counters were mounted to the remaining 5. wall layer on the
upstream, left and right side of the rest of the bunker. The space
available by leaving out the 6. wall layer amounts to only 14 cm.
Between the 6. and 7. wall layer a tight area of 4 cm thick
polyethylen sheets was mounted which are used to thermalize beam
correlated neutrons before they reach the veto counters. An
admixture of boron was not necessary, because the surrounding iron
itself captures the thermalized neutrons. The produced gammas are
elminiated and to not reach the veto counters in the 6. layer. After
partial reconstruction of the side walls a total of 30 veto counters
covering an area of 73.12 m² could be installed in
the space left between the 5. and 7. roof layer. The original 6.
layer converted to the 7. providing the same amount of shield at the
roof as before the upgrade.
- Around the main bunker additional shieldings were put into
strategic positions at the east and west side but mainly at the
north side towards the target to protect the main detector from fast
neutrons from the main target but also from the upstream
intermediate target for muon-spin-experiments (mSR-Target).
Those shieldings were in parts substantially strengthened and topped
up diring the KARMEN-upgrade. These shieldings now also protect the
veto roof counters against beam correlated background caused by fast
neutrons and by (n,g)-captures of
thermalized neutrons in the bunker. On top of the bunker a U-shaped
shielding was put into place in order to reduce the skyshine
neutron flux from above. At the same time this shielding also
reduces the muon induced countrate at the outer ends of the veto
modules.
Veto Counters
-
- The veto system including the added shield bottom consists of in
total 136 single veto modules made of 3.15 m, 3.75 m and 4 m long,
65 cm wide and 5 cm thick Bicron BC412 plastic scintillator. This
scintillator features a very good absolute light output of about 8.5
photons/keV deposited energy. The effective attenuation length,
averaged over the emission spectrum and the spectral quantum
efficiency is about 6 m. The picture above illustrates the
arrangement of the two groups of 2 inch Philips VALVO XP 2262
photomultipliers and the 180° light
bending at both narrow sides of each module. A 10 cm long and 6 cm
thick piece of scintillator covering the whole width and glued to
the scintillator bar is used as a light guide. The narrow sides of
the bar as well as the light guide are at an angle of 12 and 7
degrees respectively to the vertical axis. The light bending is
supported by highly reflective aluminium sheets (95%) at the end of
the modules. The photomultipliers were glued at the positions ±24.1
cm and ±7.6 cm relative to the
symmetry axis of the counters with the help of optical cement. This
cement has nearly the same index of refraction n = 1.54 as the
scintillator of n = 1.58 and thus minimizes losses by reflection at
the optical boundary between the counter, the light guide and the
photomultipliers. Aluminium mirrors were also attached in the space
between the photomultpliers in order to throw the escaping light
back into the light bending and thus to guide it partially to the
photomultipliers.
- The light transport in the scinitllator works similar to the
transport in the central detector modules via total internal
reflection at the optical boundary between the dense scintillator
medium and the optically thinner medium air. The angular limit for
total reflection is 39° relative to
the surface normal. The bars are completely wrapped in crumpled
aluminium foil supplementing the light transport. Crumpling the foil
guarantees that the foil does not cling to the scintillator and
prevents that total reflection is replaced by less efficient mirror
reflection. In additon to a mere protective purpose the foil
reflects emerging light back into the scintillator, where it can
again be transported by total internal reflection in case it hits
surface defect before. The raises the total light ouput of the
coutners substantially, this could be confirmed in measurments as
well as in Monte-Carlo simulations. The whole configuration of the
veto counters was optimzed to the highest possible lightoutput and
more important to a homogeneus detection efficiency for muons over
the total area of the counters. The light bending at the ends of the
counters does not only allow a compact arrangement of the veto
system but also homogenizes the muon detection at the ends of the
modules and guarantees a 100% efficient detection along the total
length of the modules. In contrast a read out by photomultipliers
glued directly to the bar would case blind spots between the
photomultpliers. The lower index of reflection of the
photomultiplier glass of n = 1.4 limits the angluar range of the
light reaching the photocathode. The lightbending has a very high
50% collection efficiency relative to the straight read out. The
large attenuation length of the scintillator and the excellent light
transport properties of the veto counters result in very flat
lightouput curves allowing together with the thickness of 5 cm an
excellent muon/gamma separation.
- The four photomultipliers at each side of a counter have a
single high voltage supply and the signals are passively added
directly at the modules. Thus it was necessary to obtain nearly
equal gains within the sets of photomultipliers over the range of
high voltages uses during operation. Therefore the gain of all 1100
photomultipliers delivered from Philips were measured and the
photomultipliers were matched together. The varation of the gains
within each set could thus be limited below 10%.
- The 3.75 m long roof and bottom counters and the 4 m long
downstream counters were covered betweend the light guides and on
the lower side of the modules with 4 cm thick boron polyethylen
sheets. The idea of this arrangement was to thermalize neutrons and
to absorb them with boron. For the installation of the 3.75 m long
counters of the upstream, left and right sides of the veto the space
was limited to only 14 cm. Thus the boron polyethylen could only be
attached between the light guides. All counters were wrapped and
welded into black and lighttight polyethylen foil. The counters of
the left, right, upstream and downstream sides were assembled to
units of 2 to 3 modules and attached to a single frame. Those frames
were installed at the bunker walls as a whole. At the downstream
side of the detector still within the bunker and downstream of the
boron-polyethylen shielding a 90 cm thick iron wall was build
covering the whole inner clearance of the bunker. At this wall
finally a total of 10 downstream modules were installed in 2- and
3-module frames. The wall is necessary to suppress neutrons which
are produced by muons in the downstream area of the bunker where the
muons cannot be detected by any veto counters. Therefore the iron
had to be installed between the counters and the central
detector. The attenuation length of 21.6 cm for neutrons results in
suppression down to 1% of the primary intensity after 1 m iron (90
cm downstream wall or 5 layers of slabs respectively in additon to
the walls of the inner passive shielding).
|
Side
|
Up
|
Down
|
Left
|
Right
|
Top
|
Bottom
|
|
Modules
|
2´11 |
10
|
3´11 |
3´11 |
2´15 |
8
|
|
Length
|
3.15 m
|
4.00 m
|
3.15 m
|
3.15 m
|
3.75 m
|
3.15 m
|
|
Total area
|
45.05 m2 |
26.00 m²
|
67.57 m²
|
67.57 m²
|
73.13 m²
|
16.38 m²
|
|
Frames
|
6´3er
|
2´3er
|
9´3er
|
9´3er
|
¾ |
¾ |
|
2´2er
|
2´2er
|
3´2er
|
3´2er
|
¾ |
¾ |
- During the assembly of the module frames and the installation at
the bunker it was very important to keep the gaps between the single
counters and the frames at their lowest possible value. This was
limited by the inevitable space for the support clamps and
tolerances due to the expected thermal expansion caused by
temperature varations in the vicinty of the counters. The table
above shows a compilation of the numbers of used veto modules and
frames at each veto side. The gaps between the modules and the
frames determine essentially the total efficiency of the veto
system. The veto top extends over the side walls of the veto in
order to throw a shadow over the gaps caused by the 3. and 4. roof
layer and the wall layers left and right to the veto downstream.
Although the total area coverage amounts to only 87.7% a muon
detection efficiency of 99.4% for muons entering the main detector
could be achieved.
To Neutrino Physics
