Improvement of a numerical ballistic impact simulation of a 7.62mm armour-piercing bullet on an ultra-high hardness armour steel

Abstract

The main objective of this thesis was to get a step closer to the accurate and trustworthy simulation of impact simulations. The ultimate goal was and is to be able to predict if a plate can stop, or not, an impacting bullet. But first of all, one wants to be able to simulate correctly an impact and have residual velocities which are in accordance with experimental results and the theoretical model of Recht-Ipson. A brief literature review is done for all material and numerical models, as well as particular experimental observations that are necessary for good understanding. In a first part, the material of the plate (steel 600HB) has been calibrated using experimental data from different specimens with different triaxialities. The calibration has been done for the stress-strain curve, as well as for the failure. Mesh dependency has also been studied. The results were excellent (discrepancy of less than 5%) for the stress strain curve, but less precise for the failure. In fact, the initiation of failure could not be modelled accurately for the notched specimens. None of the tested failure models forced the notched specimens to fail in the middle, as it is observed experimentally. In a second part, the material of the projectile has been verified by means of a Taylor test. The failure modes were consistent with the literature. Furthermore, different tests have been conducted to choose the best definition for the contact between the different parts of the bullet. In the third part, using the chosen contact definition, a first simulation was launched using Lagrangian elements. The results were not really convincing. A lot of elements have been deleted, resulting into unrealistic results. Therefore, the projectile and the impact zone have been replaced by SPH. The results using this two different definitions are compared, concluding that the SPH version was a lot closer to experimental observations. Different failure laws have been tested for the plate. None of them was able to correctly predict a non-linear relation between the residual velocity and the impact velocity as it is observed experimentally. This can be improved when investigating the dependency of the plate's material and the strain rate.