Free-Flight Aerodynamics of Ellipsoidal Clusters in Hypersonic Flow

Abstract

Characterizing the separation behavior of meteoroids in the atmosphere is critical in risk assessment and mitigation. This work aims to better understand the separation behavior of meteoroid fragments by examining clusters of small spheres, arranged in initial ellipsoidal clusters of varying eccentricity. Hypersonic free-flight experiments were conducted at Mach 6 to study and quantify the separation behavior of these ellipsoidal clusters. Each experiment contained 52 equal sized plastic spheres packed within a 3D printed shell that would separate in a nondisruptive manner upon flow arrival. A machine learning algorithm was employed to initially identify and categorize spheres. Then, lone spheres were tracked using subpixel edge detection routines, whereas occluded spheres or sub-clusters were tracked with artificial image generation techniques. Streamwise and lateral sphere velocities were calculated by tracking each sphere position over subsequent frames, to quantify the late separation behavior. Comparisons are drawn between clusters of differing initial shapes, as well as prior data on perfectly spherical clusters, and a trend in collective lateral velocity is established.