RESUMEN
A multi-shell neutron spectrometer with indium foil detector (In-MuNS) was developed to evaluate intense neutron fields that are generated in medical accelerators. The response matrix of this new spectrometer was calculated from 1 meV to 100 MeV using MCNP5 v.1.6 with ENDF/B-VIII.0 nuclear data. An experiment with a252Cf source with known emission rate was performed to validate the computational model of the spectrometer. This included detailed modelling of the irradiation room to evaluate the room-scattered field. The contribution of scattered neutrons to the induced activity in the foil reached 30% for the smallest sphere configuration (diameter 5.0 cm). The quotient between the experimental and simulated foil activity remained satisfactorily constant (1.03±0.04) as the sphere diameter varied, demonstrating the validity of the simulation model. In-MuNS proved to be a portable and compact alternative to conventional Bonner spheres.
RESUMEN
A method for estimating source strength measurements of californium neutron sources is presented, based on the model of 252Cf, 250Cf, and 248Cm decay. This is combined with the Monte Carlo method (MCM) of propagating uncertainties. Californium sources were categorized into two types: Sort-A are those with most input quantities known while Sort-B are sources with only the mass at a certain reference date known. For Sort-A, the spread of all input quantities was estimated with Gaussian distribution and the deterministic 1st order GUM uncertainty propagation is applied to validate the MCM results. While, for Sort-B with inputs that have non-Gaussian distributions, only MCM is applied to evaluate uncertainties. Results show that for californium sources that are 25 y or older, a simple 252Cf decay correction is imprecise due to the contribution of 250Cf and 248Cm. The MCM was also shown to be a robust technique for uncertainty analysis that provides results for both Gaussian and non-Gaussian distributions. Moreover, the time-dependence of the contributors in the source strength and the corresponding uncertainties are presented. When exceedingly low uncertainties are not required, the calculation techniques presented in this work may serve as an alternative to actual measurements, which tend to be expensive.