next up previous contents
Next: The tile calorimeter Up: The calorimeter Previous: The calorimeter

The liquid argon calorimeter

 For the electromagnetic calorimeter in ATLAS is chosen a liquid argon sampling calorimeter. Layers of lead/stainless steel and liquid argon are interspaced. The lead gives the shower development with its short radiation length and the secondary electrons create ionisation in the narrow gaps of liquid argon. An inductive signal from the ionisation electrons drifting in the electric field across the gas-gap is registered by copper electrodes.

Figure 4.5: The structure of the barrel accordion calorimeter. The presampler is in front of the accordion.
\epsfig {file=accordion.eps,width=\widefig}

To achieve a low capacitance of the detecting elements and thereby a fast signal the lead plates have an accordion shape as shown in fig. 4.5. At the same time this creates a fully homogeneous calorimeter in the $\varphi$ coordinate. In the central rapidity region there are four samplings:
A single thin layer of argon but no lead absorber in front. The purpose is to correct for the energy loss in the Inner Detector, solenoid and cryostat wall.
1st sampling
The first sampling has a depth of 4.3 radiation lengths. The readout is, as seen in fig. 4.5, in thin $\eta$ strips i.e. each strip has the size ($\Delta$$\eta$ x $\Delta$$\varphi$) = (0.0031 x 0.098) . This provides an excellent resolution in the $\eta$ coordinate for photon/ $\pi^{0}_{}$ separation. The $\varphi$ coordinate is not suited for this since converted photons will open up in the magnetic field and produce clusters with widths similar to $\pi^{0}_{}$ clusters.
2nd sampling
The majority of the energy is deposited in the 16 radiation lengths of the second sampling. Clusters with energy below 50 GeV are fully contained and the noise can be reduced by not adding the 3rd sampling. For the position measurement of the cluster the 2 coordinates are equally important resulting in square cells of size ($\Delta$$\eta$ x $\Delta$$\varphi$) = (0.0245 x 0.0245) .
3rd sampling
Only the highest energy electrons will reach this deep in the detector. The clusters are at this point wide and the cell size can be doubled in the $\eta$ direction without loss of resolution.

In the end-cap there is less material in front of the calorimeter and the presampler can be avoided. The end-cap EM calorimeters start at |$\eta$| = 1.5 and continue down to |$\eta$| = 3.2 but with an increased cell size above |$\eta$| = 2.5 . There is a crack with bad energy resolution where the end-cap and barrel calorimeters meet. A large effort has gone into reducing the size of the crack while still leaving space for cables and cooling for the Inner Detector.

The resolution of the EM calorimeter is

$\displaystyle{\frac{\Delta E}{E}}$ = $\displaystyle{\frac{a}{\sqrt{E}}}$ $\displaystyle\oplus$ $\displaystyle{\frac{b}{E}}$ $\displaystyle\oplus$ c, (75)

with energies measured in GeV. The sampling term a is defined by the number of lead/argon interfaces and is 8-11% depending on rapidity. Noise influences the resolution at the lowest energies through the term b which is of the order 400 MeV when running at high luminosity. The constant term affects the resolution for high energy clusters and is limited by the calibration of the global energy scale and local variations in this. It is hard to predict but is believed to stay below 0.7%[*].

To withstand the high radiation levels in the forward region the hadronic calorimeter is also of liquid argon type in the end-caps. The design is simpler than the EM calorimeter and has parallel copper plates as absorbers placed perpendicular to the beam.

The very forward hadronic calorimeter with a coverage down to |$\eta$| = 4.9 is made of copper/tungsten. The choice of copper/tungsten is necessary to limit the width and depth of the showers from high energy jets close to the beam pipe, and to keep the background level low in the surrounding calorimeters from particles spraying out from the forward region. The calorimeter is a metal matrix with cylindrical holes. The holes have rods inside with a slightly smaller radius allowing for a liquid argon gap of just 250 $\mu$ m. The small gap limits the sensitivity to pile-up effects which are large close to the beam pipe where energetic jets often hit the same area of the calorimeter.

next up previous contents
Next: The tile calorimeter Up: The calorimeter Previous: The calorimeter
Ulrik Egede