Author index

Abstract

The structural response of liquid metal fast breeder reactor (LMFBR) subassemblies to local accidental events is of interest in assessing the safety of such systems. Problems to be resolved include failure propagation modes from pin to pin and from subassembly to subassembly. Factors which must be considered include: (a) the geometry of the structure, (b) uncertainty of the pressure-energy source, (c) uncertainty of materials properties under reactor operating conditions, and (d) the difficulty in performing in-pile or out-of-pile experiments which would simulate the above conditions. The main effort in evaluating the subassembly response has been centered around the development of appropriate analyses based on the finite element technique. Analysis has been extended to include not only the subassembly duct structure itself, but also the fluid environment, both within subassemblies and between them. These models and codes have been devised to cover a wide range of accident loading conditions, and can treat various materials as their properties become known. The effort described here is centered mainly around an experimental effort aimed at verifying, modifying or extending the models used in treating subassembly damage propagation.To verify the finite element codes under development, a series of out-of-pile room temperature experiments has been performed on LMFBR-type subassembly ducts under various loading conditions. The duct sections were instrumented to measure internal pressure, duct midflat strains and deflection of the mid-flat and corners. Since moderate deflections were expected, and effect influence on the radial deformation would occur over a relatively short length. Preliminary calculations and subsequent static and dynamic tests demonstrated that for the range of deformations expected in single subassembly prior to failure, a shortened duct section of only 30.48 cm in length was sufficient to provide a central section over which axially uniform conditions prevailed. As a result, with axial motion of the end plates constrained, the deformation over the uniform deflection range corresponds to two-dimensional, plane-strain conditions and a two-dimensional, finite element computer code could be applied. Tests were subsequently performed on several ducts made of type 316 stainless steel which were either annealed or 50% cold-worked. Material properties of the ducts used in the experiments were determined by testing samples obtained from each duct. Also, diamond point hardness measurements were obtained across the subassembly duct flats in order to establish that the material properties were uniform. Comparisons were then made between the code calculations and experimental results which demonstrated remarkable agreement, thus lending confidence to the code's ability to predict duct response, at least under quasi-static loading. Further preliminary work was performed on the dynamic response of hexcans to a pressure pulse designed to duplicate a postulated local event.