Bonner Sphere Spectrometry: main achievements from the 'CONRAD Computational Dosimetry Study'

摘要

In the framework of CONRAD-Work Package 4 on Computational Dosimetry a series of problemswas proposed concerning the usage of computer codes as supporting tools for irradiation facilities calibrationprocedures or detector characterization in the field of radiation protection. In this paper the results submitted bythe various laboratories for the Bonner Sphere Spectrometer response characterization problem (one of theproposed studies) are summarized. The aim of this problem is to study the response of a widespread neutronspectrometer exposed to the ISO standard neutron sources ~(241)Am-Be and ~(252)Cf.The Bonner Sphere Spectrometer (BSS) is a widely used system to evaluate the neutron spectrum through a seriesof different diameter moderating spheres (in this case polyethylene) including active or passive detectors. Anaccurate knowledge of the energy dependence of the response of each sphere is necessary to reconstruct thespectrum from the measurement (counts or activities) through an “unfolding” procedure.In the proposed comparison study, a typical BSS is proposed. The sphere diameters are: 0” (bare detector), 2”, 3”,5”, 8”, 10” and 12” , such configuration allows getting spectral information from thermal to 20 MeV neutrons. TheBSS were exposed to calibrated ISO sources at INFN and ENEA laboratories. The employed detector was a LiIscintillator.Participants to the intercomparison were asked to determine the response of a minimal set of BS exposed to aparallel neutron beam for ~(252)Cf and ~(241)Am-Be sources; simulate the irradiation of the 8” sphere to a parallelneutron beam for a set of given neutron energies and determine the neutron spectrum within the inner thinlenticular layer (0.1 mm thick) inside the Bonner sphere towards the source and perpendicular to the beam axis.The last point was important to check the correct modelling of the neutron slowing down within the spheres beforegetting to the scintillator, as it helps to clearly understand the “spectrum shifter” action of the polyethylenemoderator. Fourteen solutions were submitted and discussed during the CONRAD Workshop “UncertaintyAssessment in Computational Dosimetry: A comparison of approaches”, held in Bologna (Italy) from October 8thto 10th 2007.In general the majority of the solutions were fitting quite satisfactorily with the authors’ reference solutions. Someoutliers were nevertheless detected. The main causes of the deviations could be ascribed to an inappropriatesampling of the source both in space and energy, to the erroneous treatment of hydrogen and carbon elasticscattering cross-sections as “free gas” instead of taking into account the binding of the nuclei in the polyethylenemolecule, and some mistakes in the interpretation of the units in which to express the absolute response function.