A Method for Selecting Software for Dynamic Event Analysis: II — The Taylor Anvil and Dynamic Brazilian Tests

Abstract

New nuclear power reactor designs will require resistance to a variety of possible malevolent attacks as well as traditional dynamic accident scenarios. The design/analysis team may be faced with a broad range of phenomena including air and ground blasts, high-velocity penetrators or shaped charges, and vehicle or aircraft impacts. With a host of software tools available to address these high-energy events, the analysis team must evaluate and select the software most appropriate for their particular set of problems. The accuracy of the selected software should then be validated with respect to the phenomena governing the interaction of the threat and structure. Several software codes are available for the study of blast, impact, and other shock phenomena. At the Idaho National Laboratory (INL), a study is underway to investigate the comparative characteristics of a group of shock and high-strain rate physics codes including ABAQUS, LS-DYNA, CTH, ALEGRA, ALE3D, AUTODYN, and RADIOSS. In part I of this report published in 2007, a series of five benchmark problems to exercise some important capabilities of the subject software was identified. The benchmark problems selected are a Taylor cylinder test, a split Hopkinson pressure bar test, a free air blast, the dynamic splitting tension (Brazilian) test, and projectile penetration of a concrete slab. Part II — this paper — reports the results of two of the benchmark problems: the Taylor cylinder and the dynamic Brazilian test. The Taylor cylinder test is a method to determine the dynamic yield properties of materials. The test specimen is a right circular cylinder which is impacted against a theoretically rigid target. The cylinder deforms upon impact, with the final shape depending upon the specimen density, the impact velocity, and the dynamic yield stress, in turn a function of strain and strain rate. The splitting tension test, or Brazilian test, is a method to measure the tensile strength of concrete using a cylindrical specimen. The specimen is loaded diametrically in compression, producing a fracture at the center of the specimen that propagates toward the loading points until the cylinder is split. To generate a dynamic load, different methods such as a drop-weight or a split Hopkinson pressure bar are employed. The Taylor anvil and dynamic Brazilian test analyses are presented, including discussion of the analysis approach for each of the five subject software packages and two vendor submittals; comparison of results both among the codes and to physical test results; and conclusions as to the applicability of the subject codes to these two problems. Studies of the remaining three benchmark problems and overall conclusions will be presented in future publications.