Optimization of Bonner sphere spectrometer design applying Bayesian approach.

Area of Science:

• Nuclear physics
• Radiation detection and measurement

Background:

• The Bonner Sphere Spectrometer (BSS) is a crucial tool for neutron spectrum unfolding.
• Accurate neutron dosimetry is essential for radiation protection and nuclear safety.
• Optimization of measurement systems is key to improving data reliability.

Purpose of the Study:

• To optimize the design of the Bonner Sphere Spectrometer (BSS) using a Bayesian approach.
• To determine optimal BSS configurations for two distinct measurement objectives: neutron fluence spectrum estimation and neutron ambient dose equivalent (H*(10)) determination.
• To evaluate the performance of the optimized BSS systems in the ISO 8529 neutron standard field.

Main Methods:

• Bayesian optimization was employed to refine the BSS design.
• Simulations were conducted using the ISO 8529 standard neutron field (241Am-Be source).
• Two optimization scenarios were addressed: maximizing spectral accuracy and minimizing dose equivalent uncertainty.

Main Results:

• An optimal BSS system with up to six different diameters was proposed for achieving low uncertainties in neutron fluence spectrum estimation.
• For neutron ambient dose equivalent (H*(10)) estimation with a standard uncertainty below 10%, only two optimal BSS measurements are required.
• The study demonstrates the effectiveness of the Bayesian approach in tailoring BSS configurations for specific dosimetry needs.

Conclusions:

• The Bayesian optimization approach effectively tailors Bonner Sphere Spectrometer configurations for specific neutron measurement goals.
• Different BSS designs are optimal depending on whether the primary objective is detailed spectral information or dose equivalent estimation.
• The findings offer practical guidance for optimizing neutron measurement strategies in various applications.