• ATLAS
• Hadronic calorimeter
• Laser calibration
• Photomultiplier
• TileCal

The ATLAS Tile Calorimeter (TileCal) is the central section of the hadronic calorimeter of the ATLAS experiment, which is taking data at the LHC. It provides important information for reconstruction of hadrons, jets and missing transverse energy. It is a sampling calorimeter made of scintillating tiles as active material and steel plates as absorbers. The light produced by the passage of particles is transmitted to readout photomultipliers (PMTs) by wavelength shifting fibres. The analog signals are added to form calorimeter trigger towers. The digitized signals are processed with an algorithm in the ReadOut Drivers (RODs) and propagated to the ATLAS ReadOut Systems.
To calibrate and to monitor the stability and performance of each part of the readout chain, a multi-stage calibration system is used. A system using a radioactive source floating inside the detector is used to calibrate the TileCal response and the full readout chain. A more fast calibration of the readout chain only, described in the thesis, is based on laser excitation of all readout PMTs. Laser light pulses, similar to those produced by ionizing particles, are transmitted simultaneously to all the TileCal PMTs through a light distribution system made of beam expanders and optical fibres.
The main purpose of laser calibration is to monitor the drift of the PMT responses in the time interval between two subsequent calibrations with the radioactive source (typically one month). An internal calibration system is used to monitor the laser source stability, while special algorithms are used to measure the drifts due to the optical transmission system only. The official algorithms, presently used in the TileCal calibration procedure, are based on the assumption that the corrections for optics drifts can be computed using the response from channels assumed to be stable in time. This assumption is no longer valid with the increasing peak luminosity delivered by LHC. All the PMT responses are now drifting due to the large amount of collected light.
The aim of this thesis is to present a new laser calibration procedure which is based on a statistical method to measure the PMT gain and its time evolution. The new method does not required any "a priori" assumption on the PMT response stability.
The new method is based on the statistical nature of photoelectron production and multiplication inside a PMT. Assuming that the electronic noise can be neglected, the two main contributions to the signal fluctuations are the Poisson statistics fluctuations associated to the photoelectron emission and multiplication and the intensity variation of the light source.
Validation studies of the proposed statistical method has been done in the Pisa-INFN laboratories.
In the new proposed calibration procedure the only assumption is that the gain evolution represents the variation with time of the PMT response. The assumption is based on the present knowledge about the response of PMTs receiving a light amount that generates average anode currents below 1 uA, as in the case of the TileCal PMTs reading out the cells of the outer calorimeter layer. The key feature of the new method is the measurement of all drifts due to the optics of the laser system which are not connected to the PMT response evolution itself. The drift caused by the optical transmission system is evaluated by comparing the PMT response variation and the PMT gain variation using only PMTs reading out the outer layer of the calorimeter. The response of all PMTs is then corrected for the drift caused by the optical transmission system. After correction, the PMT response evolution and the gain evolution overlap within the errors. The only exceptions are for the PMTs that read the scintillating counters placed in the gap between the electromagnetic and hadronic calorimeters (the most exposed devices read out by the PMTs). The observed difference between response and gain may include several effects like cathode quantum efficiency loss and PMT windows transparency degradation.
This is the first time that the statistical method is applied in the calorimeter calibration procedure and this algorithm has been proposed to become part of the official calibration tools.
The final part of the thesis is devoted to the study of the PMT response stability during a physics run. To verify the PMT stability, data collected during physics runs where laser pulses are sent to the PMTs in empty sections of the LHC bunch train are analysed. The use of the laser pulses during empty bunches in physics runs was originally introduced to monitor the timing of the PMT response. This is the first time that these data are used to monitor the stability of the PMT response.
After analysing all such data taken in 2015-2016 runs, it is found that the PMT response in physics runs is almost flat if the run has low peak luminosity; differently, in case of high peak luminosity the PMT time profile has a down-drift. So, if a run with high peak luminosity lasts, for example, 40 hours, the PMT response loss is about 0.3%.
The statistical method is also applied in this analysis to study the PMT gain evolution in time during a run. These studies will help to understand the PMT behaviour in the high-luminosity LHC environment where PMT response may vary inside a physics run.