Experiments were conducted by the Munitions Directorate at the Air Force Research Laboratory to investigate the fracture and fragmentation of two different metals due to explosive loading. The first metal, Eglin Steel 1 (ES-1), was a high strength steel alloy configured as a thin shell surrounding the explosive core. The second metal, Aero 224, was a tungsten alloy configured as a stack of rings around the explosive core. The two different configurations generated two different stress states, plane-strain and uniaxial stress. The radial expansion velocity of the ES-1 shell was recorded via a photonic Doppler velocimeter (PDV). Also, the fragments from the ES-1 shell test and Aero 224 ring test were soft captured in a water tank. Complimentary computational analysis was conducted at the Naval Surface Warfare Center Dahlgren Division. The Eulerian wave propagation code, CTH, was used to analyze the stress states of the different configurations and also investigate the use of statistical compensation on explosive fragmentation. The stress states were examined in the context of stress triaxiality where triaxiality is defined as the ratio of pressure to the Von Mises stress. From the computational analysis both the ES-1 shell test and Aero 224 ring test approached, but did not reach the ideal triaxial values for plane-strain and uniaxial stress. Lastly, parametric calculations were conducted in order to determine the effectiveness of using a statistically compensated Johnson Cook fracture model to simulate the non-homogeneous nature of the ES-1 and Aero 224. While using the model did result in different fragment distributions, all the resulting distributions were less accurate than the baseline homogeneous calculation. Scrutiny of the early time fragment formation in the statistically compensated calculations revealed a mesh bias which caused material failure on surfaces parallel to the Cartesian axes. This preferential fracture produced rarefaction waves which prohibited further fragmentation thus generating fragment distributions larger than those observed in the Aero 224 ring test. Potential solutions for this issue will be explored in the future.
