The radiation environment in underground workplaces of the LHC

Active dose-monitoring of workplaces is crucial in order to operate a high-energy particle accelerator safely. As the mixed radiation fields that are expected in the environment of the Large Hadron Collider (LHC) are very different from standard use-cases like in nuclear power plants, it is of highest importance to characterize and calibrate radiation monitoring equipment appropriately for their use in high energy mixed radiation fields. Due to their sensitivity to different particle types over a larger energy range high-pressure ionization chambers have already been used at CERN and they are foreseen to be included within the radiation monitoring system of the LHC. In the framework of this thesis a new method was developed which allows for appropriate field-specific calibration of these detectors using Monte Carlo simulations. Therefore, the application of common 238Pu-Be source based calibration in mixed radiation fields was studied and compared to more accurate field specific calibration based on FLUKA Monte Carlo calculations. Due to the fact that outside of lateral concrete shielding the radiation environment is dominated by high-energy neutrons the reliability of the FLUKA Monte Carlo code was benchmarked comparing measurements and simulations of the detector response to quasi-mono-energetic neutrons of 250 and 400 MeV. Finally, the determination of field-specific calibration coefficients was demonstrated for workplaces i n the cavern of the LHCb experiment. Setting up complex geometries for particle transport calculations requires a significant amount of time that can span from several days to weeks. The reason can be found in the fact that structures have to be defined by textual input of Combinatorial-Solid-Geometry equations. In order to simplify and streamline this workstep a solid modeler was developed which allows for interactive creation and three dimensional inspection of the geometrical setups. Consequently, the cumbersome and error prone process of manually typing in these Boolean equations was eliminated. Furthermore, a debugging module was developed to verify the logical validity of a geometry which is crucial as the used Monte Carlo particle transport code might stall or produce erroneous results in the presence of geometry errors. Therefore, a number of different debugging algorithms was developed including for the first time for this purpose the application of Quasi-Monte-Carlo methods to improve efficiency.