Interactions of Particles with Momenta of 1–10 GeV in a Highly Granular Hadronic Calorimeter with Tungsten Absorbers

Linear electron-positron colliders are proposed to complement and extend the physics programme of the Large Hadron Collider at CERN. In order to satisfy the physics goal requirements at linear colliders, detector concepts based on the Particle Flow approach are developed. Central to this approach are a high resolution tracker and a highly granular calorimeter which provide excellent jet energy resolution and background separation. The Compact Linear Collider (CLIC) is an electron-positron collider under study, aiming at centre-of-mass energies up to 3TeV. For the barrel hadronic calorimeter of experiments at CLIC, a detector with tungsten absorber plates is considered, as it is able to contain shower jets while keeping the diameter of the surrounding solenoid magnet limited. A highly granular analogue hadron calorimeter with tungsten absorbers was built by the CALICE collaboration. This thesis presents the analysis of the low-momentum data (1 GeV $\leq$ p $\leq$ 10 GeV) recorded in 2010 at the CERN Proton Synchrotron (PS). The energy resolution is measured for electrons, pions and protons, and is compared with the performance of other calorimeters. In addition, comparisons of data with Monte Carlo are done for hadronic shower shapes, in order to validate Geant4 simulation models of the development of hadronic showers in tungsten. The electromagnetic resolution for a tungsten based HCAL is worse than for an iron based HCAL. This is expected due to the shorter radiation length of tungsten. The shower maximum tmax for electrons is between 3.9 $X_{0}$ and 5.6 $X_{0}$. The mean shower radius has a maximum value of 62 mm at 1 GeV and decreases to 26 mm at 6 GeV. The mean shower radius for protons is between 74 mm and 81 mm, while for the pions they are between 69 mm and 78 mm. Three physics lists were studied to validate the hadron interactions in tungsten: QGSP_BERT_HP, FTFP_BERT_HP and QGSP_BIC_HP. In general the Monte Carlo is within 8% agreement with the data.