An improved criterion to minimize FE mesh-dependency in concrete structures under high strain rate conditions

In the context of rising demand for reliability and safety in concrete structures subjected to blast or impact loading, the behavior of concrete under high strain rate conditions is an important issue. Since concrete subjected to impact loading associated with high strain rates shows quite different material behavior from that under monotonically increasing quasi-static loads, several material models have been proposed and used to describe the high strain rate behavior of concrete under blast and impact loadings. In the process of modeling the high strain rate condition with these material models, mesh dependency in the used finite element (FE) is the key problem because simulation results under high strain-rate conditions are quite sensitive to the applied FE mesh size. This means that the accuracy of simulation results for concrete structures may be strongly dependent on the FE mesh size. This paper introduces an improved criterion that can minimize the mesh-dependency of simulation results on the basis of the fracture energy concept, and the HJC (Holmquist Johnson Cook), CSC (Continuous Surface Cap), and K&C (Karagozian & Case) models are examined to trace their relative sensitivity to the employed FE mesh size. Consistent with the purpose of a perforation test for a concrete plate under a projectile (bullet) impact, the residual velocities of a projectile after perforation are compared. Correlation studies between analytical results and associated parametric studies show that the variation of residual velocity with the used FE mesh size is substantially reduced and the accuracy of simulation results is improved by applying a unique failure strain value determined according to the proposed criterion. (C) 2015 Elsevier Ltd. All rights reserved.