S.A. Ricketts, “Investigation of Cadmium Lining Effects in Neutron Detector Shroud”, M.S. Thesis, Nuclear Engineering, Texas A&M University, College Station, TX (2021).
The Center for Exascale Radiation Transport (CERT) project experiments, which employ graphite configurations with a rage of complexities, are used as calibration and validation benchmarks for Texas A&M University’s (TAMU) Parallel Deterministic Transport (PDT) code. A layer of hand cut cadmium lining was added to boron trifluoride (BF3) thermal neutron detector shrouds in order to reduce counts from neutrons that do not enter through the shroud’s opening. The cadmium faceplate along the front of each detector caused inconsistent count rate ratios, Rb/c, between two detectors at symmetric measurement locations a and b on the least complex configuration. Data analysis methods that assumed detector symmetry incorrectly inferred a humidity change between measurement periods that required a 1.62 ± 0.12% asymmetric correction factor. Asymmetric cadmium faceplate openings in tandem with the curved YZ flux profile of the pile were believed to have caused a difference in detector response in one of these detectors. Following a Monte Carlo N-Particle Transport Code (MCNP) sensitivity investigation, it was predicted that a hand cut modification to enlarge the faceplate opening would remove this detector response asymmetry. This was shown to be false, as following the cadmium faceplate manipulation, the results indicated once more a false humidity correction needed of 1.22 ± 0.23%. Rotating the detectors 180 degrees and repeating previous experiments indicated that the asymmetric properties of the cadmium faceplate vary in sensitivity depending on the gradient of the flux profile. Experiments without the cadmium faceplate, then rotated 180 degrees, provided a consistent Rb/c between experiments and removed the false humidity correction. Overall, the small asymmetries in the faceplate were sensitive to the XZ flux profile of the pile and this detector configuration is unreliable as calibration and validation benchmarks. For the current experimental setup, a symmetric detector configuration is needed in both the ±X and ±Y direction to avoid having detector response differences when changing measurement locations.