Analytical Evaluation of Concrete Penetration Modeling Techniques

Abstract

The resistance of concrete targets to penetration of high speed projectiles is a topic of high value in the national security and catastrophic design fields. Many methods have been developed to effectively analyze these types of problems. Currently there are a number of numerical codes and constitutive models used to analyze concrete impact and penetration with new methods developed continuously. This research evaluates the accuracy of four analysis codes and five concrete constitutive models. Two Lagrangian analysis programs, EPIC and LS-DYNA, as well as an Eulerian code, CTH, are compared in this work. A developmental version of the Material Point Method is also evaluated in order to study the effectiveness of Arbitrary Lagrangian Eulerian (ALE) modeling methods for concrete impact and penetration. The concrete models evaluated in this research include Holmquist Johnson Cook, Brittle Failure Kinetics, Osborn, Karagozian and Case, and Drucker-Prager. The modeling programs and constitutive models are evaluated by comparing simulation results to a series of concrete impact and penetration experiments. The experimental test data, provided by Sandia National Laboratories, comprises concrete targets of two compressive strengths (3.3 and 5.7 ksi) and two thicknesses and projectiles of two nose shapes. Extensive material testing of the experimental concrete is used to calibrate the constitutive models in each analysis package. An additional parametric study investigates the influence of experimental variables on the most promising analytical configuration. Observations from this research show that the EPIC and LS-DYNA analysis codes are currently best suited for concrete impact and penetration problems. Both codes contain features which allow for realistic modeling and produce accurate results for the experimental impact tests. Recommendations for improving analysis methods specific to concrete impact and penetration are presented.