Analysis and characterization of perforated neutron detectors

Abstract

Perforated neutron detectors suffer the unfortunate effect that their efficiency is a strong function of the direction of neutron incidence. It is found, by Monte Carlo simulation of many perforation shapes, that sinusoidal-type perforations greatly reduce the variation of detector efficiency. Detectors with rod-type perforations are modeled using a hybrid transport method linking the MCNP transport code and a specialized ion-transport code to calculate the probability that a neutron is detected. Channel, chevron, and sinusoidal perforations are modeled using other customized transport codes. Detector efficiency calculations are performed for neutrons incident at various polar and azimuthal angles. It is discovered that the efficiency losses of the detectors result from the decreasing solid angle subtended by the detector from the source and streaming through the detector at specific azimuthal angles. Detectors achieving an efficiency in excess of 10% and having a relatively flat ± 1% angular dependence in all azimuthal angles and polar angles between 0 and 60 degrees are predicted. Efficiencies up to 25% are achievable at the loss of directional independence.

In addition to minimizing the directional dependence of the perforated detectors, the feasibility of developing a neutron detector for deployment in cargo containers to locate nuclear weapon pits is investigated using the MCNP transport code. The detector considered is a 7-mm diameter, 6LiF, rod-perforated detector surrounded in a cylinder of polyethylene. The optimum thicknesses of surrounding polyethylene, to maximize the response of the detector, is determined to be 10 cm of radial, 5 cm of front, and 5 cm of back polyethylene for end-on neutron incidence. Such a detector is predicted to produce a count rate between 12 and 15 cpm from a nuclear-weapon pit composed of 90% 239Pu and 10% 240Pu at a distance of 3 m. Side incidence is also considered, and the optimum moderator dimensions are 8 cm of radial, 10 cm of front, and 10 cm of back polyethylene that produce approximately the same count rate.