Duration: 36 months
With the increasing reliance on orbiting satellites, the risks posed by their potential impact on Earth’s surface during re-entry has become a significant concern. Re-entering vehicles face extreme loads, leading to many propagating cracks, which then coalesce to form fragments in a process called dynamic fragmentation. Robust numerical models are direly needed to develop a fundamental understanding of such highly energetic breakup events. This research aims to promote the creation of an efficient, high-fidelity, physics-based FE software that can reliably predict the generation of debris. The focus will be on obtaining accurate statistical distributions of fragment sizes, shapes, and velocities to better inform the mitigation of risks during satellite re-entry. The LSMS group at EPFL has expertise in dynamic fragmentation and has developed high-performance algorithms for the insertion of cohesive cracks. Yet, this approach suffers from a mesh dependency of the crack paths and non-robust predictions of fragment shapes. Phase-field modelling of fracture belongs to another family of methods using diffuse crack approaches, resulting in mesh-independent crack paths. This method has been shown to lead to promising results, and its use in the context of dynamic fragmentation has yet to be explored. It could lead to drastic improvements in the predictive ability of numerical models, and is the main goal of the research. The deliverables are numerical data of fragment sizes for various structures and re-entry conditions. These will be compared to known semi-empirical models and experimental data, when available. Finally, the methods will be integrated in an open-source software for broader use by ESA scientists and the scientific community.
