Satellite Fragmentation Modeling for Rapid Secondary Impact Mitigation Response

According to present trends, a drastic increase in the number of objects in low Earth orbit (LEO) has prompted risk assessments of short-term secondary impact. As a result, the Secondary Impact Predictor (SIP) was architected to provide a rapid warning system for high-risk satellites within hours after a detected collision event. With the progress mainly concluded on the validation and optimization of the predictor itself, attention has to be turned to the fragment distribution of the satellite breakup model on which the predictor relies. While the current NASA Standard Breakup Model (SBM) being used provides a continuous empirical distribution of fragments in mass- velocity space, the model is not sensitive to the satellite design or impactor parameters and does not provide the angular distribution of fragments relative to the impact direction. Therefore, the largest single improvement to the predictor is to improve the breakup model to be sensitive to all knowable parameters provided by the conjunction data messages (CDMs) which currently reports potential primary impacts up to 72 hours beforehand. With the use of the Collision Simulation Tool (CST) from the University of Padova CISAS laboratory, a series of hypervelocity impact simulations is performed to obtain fragment velocity distribution from impacts with the COSMO- SkyMed satellite with 1.1-10.6 cm aluminum spheres of various impact velocities (4.2-13.3 km/s) as well as various impact angles. Geometrical and statistical methods are employed to create a sensible upper bound and average of fragment density given the knowable and unknowable parameters, ready to be integrated into the existing SIP framework. This work demonstrates a first- step procedure to profile the fragmentation signature of any satellite model, especially applicable for mass-produced satellite constellations. In addition, an outline is proposed to implement a live advanced secondary impact predictor based on the existing conjunction data message and its API to create warning vectors for all satellites ahead of impact detection. An example of current LEO secondary impact assessment showed 225 satellites surpassing 1 in 10000 risks of millimeter-scale impacts within the first 3 hours after a breakup event using the worst-case model, with 1.2×10⁻⁴ expected impacts across all satellites using a more averaged model. These results can scale up drastically when accounting for future data-center constellations, longer evaluation, and sub- millimeter fragments, and thus, prompt further developments of the predictor and the breakup models. Calculations using 4 cores of Apple M1 CPU took an average of 880 seconds for a 3-hour assessment of the current 16,843 satellites LEO environment with acceptable truncation error.