Criticality experiments: analysis, evaluation, and programs. 8. Prompt Neutron Decay Constants in Uranium Diluted with Matrix Material Systems

Rossi-Alpha measurements were performed on uranium diluted with matrix material systems to determine the prompt neutron decay constants. These constants represent an eigenvalue characteristic of these particular critical assemblies, which can be experimentally measured by the Rossi-Alpha or pulse neutron source techniques and calculated by a deterministic or Monte Carlo method. In the measurements presented in this paper, highly enriched foils diluted in various X/235U ratios with polyethylene and SiO2, and polyethylene and aluminum were assembled to a high multiplication, and the prompt neutron decay constants were obtained by the Rossi-Alpha technique. The uranium diluted with matrix material experiments were fueled with highly enriched uranium foils. The average dimensions of the bare foils were 22.86 cm squared and 0.00762 cm thick. The foils were laminated with plastic sheets to reduce the amount of airborne contamination. Each foil weighed ∼70 g. The diluent material consisted of SiO2, or 6061 aluminum plates, which were embedded into polyethylene plates. The SiO2 and aluminum plates were 22.86 cm square and 0.64 cm thick. The polyethylene plates were 39.12 cm square and 1.91 cm thick. Each polyethylene plate had a central recess whose dimensions were 22.86 cm by 22.86 cm by 0.64 cm deep and was used to accommodate the SiO2, or aluminum plates as well as the uranium foils. There were eight 39.12-cm-squared by 2.54-cm-thick high density polyethylene plates that form the top and bottom reflectors (four at the top and four at the bottom). Also, one of the polyethylene plates located in the center of the assembly had holes drilled in a radial direction to accommodate neutron detectors. Four 3He detectors were placed in this plate. The 3He detectors were 1.27 cm in diameter and ∼15 cm long. Rossi-Alpha measurements were performed at several subcritical separations for both experiments. The data were collected with a type I time analyzer (PATRM). This time analyzer has the capability of time tagging each arrival pulse from each of the neutron detectors and sorting them one by one into a time window containing 100 bins. The stored data can be analyzed using different time windows without having to retake the data. The alphas for both experiments at different subcritical separations were obtained from least squares fits to the functional forms f(t) = A exp(-αt) + C and f(t) = A exp(-α1t) + B exp(-α2t) + C. The prompt neutron decay constants at delayed critical were obtained by plotting the alphas at a particular subcritical separation as a function of the inverse count rate and extrapolating linearly to an inverse count rate of zero (delayed critical). Table I shows the delayed critical prompt neutron decay constants for the uranium/SiO2/polyethylene, and uranium/ aluminum/polyethylene experiments. It is important to note that for both experiments the least-squares fit to 'one exponential' was not the perfect fit because the system was operated at a high multiplication (high count rates) and the detection system experienced some saturation. However, when we removed the neutron source and counted for longer periods of time at these high multiplications, the fit approached a one-exponential function. In addition, for the 'two-exponential' function, the ratio of the amplitude A of the fundamental-mode exponential to the second exponential amplitude B tends to increase as any of the systems approaches delayed critical. Thus, the second exponential diminishes with respect to the fundamental-mode exponential, whose decay constant is very similar to the one-exponential fit decay constant (see Table I). Based on the fact that both alphas are statistically the same, we can say that the neutron lifetime for both systems is approximately the same. Finally, the temperature of the experiments was kept within 1 deg. C for each run, which would yield essentially the same alpha if we consider that the temperature coefficient for these experiments is on the order of $0.03/deg. C. (authors)