The Compressed Baryonic Matter (CBM) experiment at the future Facility for Antiproton and Ion Research (FAIR) complex will investigate the phase diagram of strongly interacting matter at high baryon density and moderate temperatures in A+A collisions. The beam energy will range from 2 up to 11 AGeV for the heaviest nuclei at the SIS 100 accelerator set-up. Due to their penetrating nature, electromagnetic probes are particularly interesting, as they deliver undistorted information from the fireball, unveiling properties of the created hot and dense matter.To cope with the CBM physics program, an efficient and clean electron identification and pion suppression (for momenta up to 8 GeV/c) will be provided by the CBM-RICH (Ring Imaging Cherenkov) detector. In the SIS 100 set-up, the RICH detector together with four layers of Transition Radiation Detector, should reach a combined pion suppression factor of 1000 to 5000 in a wide acceptance. The RICH detector will be made of a CO2 gaseous radiator, Multi-Anode Photo-Multipliers (H12700 MAPMTs from Hamamatsu) for photon detection and 80 trapezoidal glass mirror tiles, equally distributed in two half-spheres and used as focusing elements with spectral reflectivity down to the UV range.One of the technical challenges emerging, while interchanging the RICH with the MuCh (Muon Chambers) presumably on a yearly basis, is a rigid and stable mechanical design along with a mirror alignment monitoring system. The latter combines two methods used in the COMPASS and HERA-B experiments and was adapted to the CBMRICH detector geometry. In addition, a correction cycle was designed, which guarantees a proper operation of the detector even though mirrors are misaligned. These developments are the subject of the presented thesis.The first method to determine mirror rotations is the Continuous Line Alignment Monitoring (CLAM) method. It uses a dedicated equipment: cameras with good pixel resolution, LEDs as light source and a grid made of retroreflective material. The principle relies on the fact that the reflected grid on the mirrors will appear broken at mirror edges, if the neighbouring mirrors are misaligned with one another. This is the fast qualitative variant of the method.Furthermore, it can be used as a quantitative method to determine mirror rotations. In this case a previous laboratory calibration is prerequisite, which consists in measuring the pixel shift generated by given mirror rotations for a single mirror tile. This calibration has to be conducted on all mirrors to allow a complete measurement of mirror rotations The second technique adapted for the CBM-RICH detector uses data in software. The principle is to measure, for a high enough number of cumulated rings reflected on a single mirror tile, two quantities referred to as the Cherenkov distance and Cherenkov angle . In case of mirror misalignment, plotting the Cherenkov distance as a function of the Cherenkov angle , reveals a sinusoidal behaviour. After a fitting procedure, the extracted parameters can be related to the mirror rotations, allowing an accurate quantification of mirror rotations.The performances of this technique were investigated. For a horizontal rotation of the mirror tile, the technique works for misalignments ranging between 0.3 and 14 mrad. For a vertical rotation of the tile, the technique yields accurate results for misalignments ranging from 0.4 mrad up to 15 mrad. It is more reliable for tiles located in the centre of the mirror wall than for tiles in the outer region.A mirror correction cycle specifically designed for the CBM-RICH detector is introduced. It uses the mirror rotation information from the two presented methods to correct the track extrapolation on the MAPMT planes. Thus the distance between the extrapolated track and the reconstructed ring centre is reduced. It was added inside the CbmRoot framework and can run automatically in the reconstruction procedure.The correction cycle was tested in simulations. The mirror tiles were artificially misaligned on both their axes following a Gaussian distribution with a standard deviation of 1 mrad. The mean ring-track distance then amounts to 0.47 cm, while being 0.14 cm in the aligned case. Applying the correction cycle reduces this number to 0.17 cm, which is close to the ideal case. In addition, it corrects the matching efficiency of the CBM-RICH and the electron identification efficiency with respect to the misaligned case. Even with a standard deviation of 3 mrad, the results obtained remain within the specifications required by the technical design report of the CBM-RICH detector.
Verknüpfung zu Publikationen oder weiteren Datensätzen
