Pressure vessel fracture tests and core barrel integrity

A cylindrical vessel-model (proportion model to original vessel 1 to 5.5) and a burst protection device is used. The tests shall demonstrate that the core-vessel-model will remain intact after bursting of the pressure vessel model. The fracture mode of the pressure vessel is to be a longitudinal fracture. In addition to a qualitative appraisal of the results the real loads on the core-vessel model during the pressure-vessel model bursting are measured. The analytical computer programs for the original core-vessel design are applied to the vessel-model geometry. All dynamic strains measured on the core barrel model during bursting showed elastic behavior. These findings demonstrate that the core barrel load imposed by a longitudinal fracture of the pressure vessel will not cause any damage of the core barrel. The burst tests prove that a fracture starting from a longitudinal fault will be arrested within a small distance in the full cross-section of the pressure vessel. This holds true both for the case that the material is in a pronounced brittle condition and for the normal ductile condition. This ductile behavior of the A 533 grade B reactor steel and its striking fracture arrest behavior was shown under a definitely higher nominal circumferential stress (223 N/mm2) than under operating conditions (196 N/mm2). The fracture surface shows pronounced ductile fractographic patterns which still prevailed when the steel was quenched in water without subsequent tempering. This demonstrates that it is not possible to realize the boundary conditions of the analytical computations with respect to an infinitive velocity of crack propagation. (orig./HP)