Fragmentation models based on void coalescence

Understanding the fragmentation process in solids and liquids, and predicting the fragment size distributions, are important for a number of applications that include meteorite impacts on space structures, armor/antiarmor applications, and debris generation in pulse power facilities. The fragmentation process is one in which microscopic cracks, voids, or shear bands are nucleated at discrete sites and thereafter grow to coalescence. Mesomechanical models of this process are fairly mature for failure via shear banding or cleavage cracking, but not for voids in liquids or plastically deforming solids (1). Even when mesomechanical models are available, they are typically challenging to install in ever-evolving hydrocodes, and simpler engineering models that predict average fragment size as a function of a few material properties and the strain rate have come into wide use (2). We assess the expected accuracies of such engineering models by comparing the predictions of two "equally plausible'' simple models for the fragmentation of liquids and plastically deforming solids. The two models agree closely for liquids, but differ by about a factor of ten for plastically deforming solids. We conclude that such engineering models are suitable for scoping problems and predicting trends, but mesomechanical models will be required for more accurate predictions.