Analysis of Fissile Materials by Cyclic Activation of Delayed Neutrons

A method for analyzing nuclear material for different fissile nuclides by cyclic activation has been developed and tested experimentally with samples of ²³⁵U and ²³⁹Pu, singly and in combination. The method of analysis is based on the differences in the abundances and half-lives of delayed neutron groups between the various fissile nuclides. The steady-state delayed neutron response to periodic activation is independent of activation cycle period at short periods, decreases exponentially with period at long periods, and has a break point, or knee, where the response changes from constant to exponential, or from one exponential to another, with greater slope for each characteristic emitter present. The activation cycle periods at which these break points occur, the slopes of the exponential fall-off or response with cycle period between break points, and the absolute magnitude of the response at any cycle period are all functions of the effective half-lives and abundances of the delayed neutron precursors activated, so, consequently, the characteristic delayed neutron response as a function of activation cycle period is different for the various fissile species. In the experiment, cyclic activation was accomplished by moving the samples containing fissile material cyclically through a thermal-neutron beam from the Georgia Tech Research Reactor, out of the beam and through a delayed neutron detector, and back through the beam, etc. The delayed neutron response was recorded at activation cycle periods ranging from 0.1 to 100 s for samples containing varying amounts of ²³⁵U and ²³⁹Pu. Deviations in the responses of the samples containing both ²³⁵U and ²³⁹Pu from the response of standards containing only ²³⁵U or ²³⁹Pu were determined to infer the ²³⁵U- to-²³⁹Pu ratio. After the ratio of the two fissile nuclides present was obtained, the delayed neutron response at short cycle periods was used to estimate the mass of each fissile nuclide present in the sample. For samples containing about a gram of fissile material, accuracies on the order of 2% for ²³⁵U and 4% for ²³⁹Pu could be obtained for 1.5-h experiment run times when the fissile nuclides were present in about equal portions. Accuracies were dependent on the ²³⁵U-to-²³⁹Pu ratio and on the total mass of fissile materials present.