Calculations to compare different ways of modelling the plate geometry cells of the Zebra fast critical assembly, MZA

The core region cells of the Zebra fast critical assembly MZA comprise 14 plates in a square steel tube, with 12 cells being stacked axially to form the core section of the assembly. The cells can be modelled in different levels of detail, ranging from a three-dimensional representation in which the core (The word core is used to describe both the region of a plate containing the main material, such as plutonium, UO2 or sodium, and the region of the assembly containing fissile material cells.) and canning regions of the plates and the void gaps between the edges of the plates and the steel tube, and between tubes, are represented. Simplified models include a three-dimensional representation in which the void regions are combined with the tube material. A further simplified three-dimensional model, called the MURAL model, represents the core regions of the plates but homogenises the canning, tube material and void regions. Two types of one-dimensional slab geometry model are found in the literature, one in which the materials are homogenised within each of the three axial slab regions of a canned plate (plate core and upper and lower canning regions) and a further simplified version in which the plate is modelled as a single region, the compositions being averaged over the whole thickness of the plate, comprising the plate core material, the canning and the tube material. MONK Monte Carlo calculations have been made for each of these models, and also for the fully homogenised cells, and the k-effective values, core sodium void reactivities and reaction rate ratios are compared.