• Publisher-DOI

Activation of high-intensity heavy-ion accelerators due to beam losses is a serious issue for accelerator parts like collimators, magnets, beam-lines, fragment separator components, etc. The quantification of the induced activity and the prediction of the radiation levels in an accelerator and its environment may have a positive impact on the decreasing of radiation hazards during “hands on” maintenance as well as on the optimisation of the machine operation. Modern Monte Carlo simulation codes are often used at accelerator facilities for different scientific applications, including the preparation of experimental setup or even for a complete substitution of experiments exclusively by numerical calculations. In the presented work the benchmarking of the FLUKA code was based on a comparison of simulated and experimental results gained by gamma spectroscopic analysis of irradiated aluminium targets. The field of interest was the evaluation of simulation precision in the range of low energies, where new physical models are implemented. The Boltzmann Master Equation (BME) model is presented in the latest version of FLUKA 2011, and its calculation possibilities are expanded by the nucleus nucleus interactions of heavy ions with energies below 100 MeV/u. The previous version of FLUKA 2008 uses the Relativistic Quantum Molecular Dynamics (RQMD) model which simulates transportation of heavy ions with energies lower than 100 MeV/u without considering their further interactions. Different aluminium targets were irradiated by 124Xe48+, 238U73+ and 238U89+ ion with energies of 300, 200 and 125 MeV/u, respectively. Using the so called stacked foils geometry was essential for the determination of induced activity as a function of the depth, known as depth profiles of residual activity. Acquiring gamma spectra of each foil at different cooling times guaranteed a reliable nuclide inventory of all accumulated nuclides. The depth profile collection of about 100 nuclides was a valuable source of experimental data for further verification of simulated results. FLUKA 2011 performs very precise estimation of the range of primary particles. It was shown that the discrepancy was only 0.5% for 238U ions with an energy of 200 MeV/u. The imprecision of the simulated results caused by imperfect transition between the RQMD and the BME models is demonstrated on the depth profiles of 7Be and 22Na target nuclei fragments. Other inaccuracies of FLUKA 2011 calculations were observed in the depth profiles of intermediate Z nuclides (e.g. 46Sc, 48V, 52Mn, 54Mn, 56Co, 58Co, 65Zn), where an artificial increase of the induced activity close to the range of primary particles was predicted, but it was not confirmed by experimental data. Additional FLUKA calculations following the production rates of the problematic nuclides introduced an origin of this impreciseness, where inexplicable production increases appear at certain energies. All documented deficiencies were shared with the team of the FLUKA developers for their application in future code updates. The second part of this dissertation is dedicated to a real life application of the FLUKA 2011 version and a demonstration of the physical model development. The beam-loss criteria of 1 W/m for proton accelerators, known in a radiation safety community, are based on the empirical experiences; their equivalent for heavy ion machines was determined by using the FLUKA 2008. The set of Monte Carlo simulations served to establish the scaling law which allows a comparison of the amount of induced activity by any heavy ion beam and the proton beam, regardless of their mass number or energy. Tolerable beam losses, defined within this approach, are 5 and 60 W/m for uranium ions with energies of 1 GeV/u and 200 MeV/u, respectively. These energies are also the limits of applicability of FLUKA 2008 in this field. The fact that the criterion was calculated by this code version, which is considered as obsolete nowadays, as well as the fact that it covers only the situation of the first 100 days of a machine operation, was a great opportunity to validate the tolerable activation limits with FLUKA 2011. Additionally, an expansion of a criterion down to an energy of 25 MeV/u and simulation of a long term operation of 20 years followed by a cooling time of 20 years was performed. As it was expected, the Boltzmann Master Equation model included in FLUKA 2011 influenced the final results of induced activity in constructing materials of accelerators. The beam loss criteria became stricter for heavy ion machines already for short term operation. The limits do not change at an energy of 1 GeV/u, where the beam loss value of 5 W/m remained for uranium ions, but the tolerable losses of uranium ions at an energy of 200 MeV/u decreased to a level of 40 W/m as compared to values predicted by FLUKA 2008. Long term simulations revealed that an accumulation of the nuclides with longer half life leads to an even stricter criterion. The beam losses of 4 W/m at 1 GeV/u are tolerable for uranium ions, but the importance of evaluation of the criterion by FLUKA 2011 is more visible at an energy of 200 MeV/u, where the tolerable beam losses are only 30 W/m, which is exactly half of the limits determined by FLUKA 2008. A complex set of calculations proved that the scaling law of the induced activity is not reliable for the heavy ion beams with energies below 150 MeV/u and therefore, the use of the tolerable beam loss criteria is not recommended anymore below this energy.