Performance of the CREAM calorimeter in accelerator beam test

Performance of the CREAM calorimeter in accelerator beam test

Y. S. Yoona, H. S. Ahnb, , M. G. Bagliesic, , G. Bigongiaric, O. Ganelb, J. H. Hand, H. J. Hyund, J. A. Jeond, T. G. Kangb, H. J. Kime, K. C. Kimb, J. K. Leed, M. H. Leeb, L. Lutzb, P. Maestroc, A. Malinineb, P. S. Marrocchesic, S. W. Namd, H. Parke, I. H. Parkd, N. H. Parkd, E. S. Seoa,b, R. Sinab, J. Wub, J. Yangd, R. Zeic and S. Y. Zinnb (a) Dept. of Physics, University of Maryland, College Park, MD 20742, USA (b) Inst. For Phys. Sci. and Tech., University of Maryland, College Park, MD 20742, USA (c) Dept. of Physics, University of Siena and INFN, Via Roma 56, 53100 Siena, Italy (d) Dept. of Physics, Ewha Womans University, Seoul, 120-750, Republic of Korea (e) Dept. of Physics, Kyungpook National University, Taegu, 702-701, Republic of Korea Presenter: Young Soo Yoon (ysy@umd.edu), usa-yoon-Y-abs3-he15-poster

The CREAM calorimeter, designed to measure the spectra of cosmic-ray nuclei from under 1 TeV to 1000 TeV, is a 20 radiation length (X0) deep sampling calorimeter. The calorimeter is comprised of 20 layers of tungsten interleaved with 20 layers of scintillating fiber ribbons, and is preceded by a pair of graphite interaction targets providing about 0.42 proton interaction lengths (λint). The calorimeter was placed in one of CERN’s SPS accelerator beams for calibration and testing. Beams of 150 GeV electrons were used for calibration, and a variety of electron, proton, and nuclear fragment beams were used to test the simulation model of the detector. In this paper we discuss the performance of the calorimeter in the electron beam and compare electron beam data with simulation results.

1. Introduction

The cosmic-ray spectrum follows a decreasing power law, with the incident flux dropping by roughly a factor of 50 for every ten-fold increase in the threshold energy. Thus, of all particles above 1 TeV, only 2% are above 10 TeV, and only 0.04% above 100 TeV. It is for this reason that the limitation on the energy reach of flight experiments measuring these particles is mostly statistical. The CREAM instrument was designed to directly measure the charge and energy of cosmic-ray nuclei, from H to Fe, for energies ranging from under 1 TeV to 1000 TeV [1]. CREAM incorporates both a calorimeter and a transition radiation detector (TRD) to measure particle energies (TRD for Z>3), and a timing-based charge detector (TCD) and a silicon charge detector (SCD) to measure particle charge or ID. The independent measurements allow verification and cross-calibration for large sub-samples of events collected. The calorimeter uses a sampling scheme with tungsten absorber to minimize the thickness required for a given depth in X0, allowing a greater collection power, or geometry factor (GF). The graphite targets, with their relatively low Z number (6 vs. 74 for tungsten), increase the fraction of nuclei interacting early enough to allow a reasonable reconstruction of their incident energy, while minimizing the weight, and having a minimal impact on the ‘age’ of the shower development regardless of the depth of graphite the shower secondaries must traverse before reaching the active scintillating fiber layers. The CREAM payload was flown for about 42 days suspended under a Long Duration Balloon (LDB) launched from McMurdo Station, Antarctica in December 2004.

2. CREAM calorimeter module

Figure 1 shows a schematic cross sectional view of various components in the CREAM calorimeter module. The calorimeter is made of 20 layers of one radiation length (X○) thick tungsten plates and fifty 1 cm wide

372

Y. S. Yoon et al.

fiber ribbons each made of nineteen 0.5 mm diameter BCF-12 scintillating fibers covering 0.25 m2. The scintillation light from the fiber ribbons is transmitted via a light mixer and a bundle of clear fibers to a set of
73-pixel hybrid photo diodes (HPDs). The HPDs combine a high channel count (73 each), low weight (~30 g), low power consumption (~0.8W each, including HV supply and front-end electronics), and have both high uniformity between pixels (RMS of ~5%), and a linear dynamic range of about 1:106. To cover both the low energy signals on the periphery of showers, and the high density energy signal at the core of the highest energy showers, a dynamic range of 1:200,000 was required. Since the front end electronics has a limited dynamic range (~1:1000 per channel), a means was needed to split the signal into several ranges, with some overlap for calibration purposes. The clear fibers carrying the signal from each scintillating fiber ribbon were thus divided into three different sub-bundles, with 37 thin fibers for the low energy range, 5 for the middle range, and 1 for the high range. These sub-bundles were glued into plastic cookies to position them against separate HPD pixels. In

…(Full text truncated)…