Development of modelling design tool for harpoon for active space debris removal

The level of debris in Earth orbit presents an increasing risk. Active debris removal, involving capture and deorbit of larger items, has been identified as important in controlling future debris population growth. One concept for capturing debris is a harpoon, with low post-penetration residual velocity being a key design requirement. Aluminium sandwich panels are common structural components on typical spacecraft and consequently a probable target structure. An ideal harpooning event is characterised with complete penetration of the sandwich panel and low harpoon residual (post penetration) velocity. This paper presents development of a numerical modelling tool for a harpoon design for active orbital debris removal. The ability to predict the ballistic limit of an aluminium sandwich panel with a honeycomb core representative of a large satellite structural element being the primary requirement for the modelling tool. The modelling approach was validated against the experimental results from normal and oblique impact tests performed for ogive and flat nose projectiles over a velocity range between 50 m/s to 120 m/s. The modelling was based on explicit finite element method, using the commercial LS-DYNA code. An element failure criterion was used to approximate material damage and failure. Sensitivity studies were performed to investigate the influence of model key features on the projectile exit velocity. The features with the strongest influence were the face sheet material model and the projectile-panel friction model. The projectile exist velocities observed in the experiments were used to select the final parameters used. The use of an element deletion criterion that incorporated the influence of stress triaxiality improved the agreement for the plate deformation behaviour between the numerical and experimental results. For the ogive nose projectile, the exit velocity agreed well with the experiments for the range of impact velocities and angles considered. For the flat nose projectile, the model overestimated the exit velocity. In the case of the normal impact this was due to the model's inability to capture the honeycomb crushing ahead of the projectile.