High-velocity dust collisions: forming planetesimals in a fragmentation cascade with final accretion

In laboratory experiments we determine the mass gain and loss in central collisions between centimetre- to decimetre-size SiO₂ dust targets and submillimetre- to centimetre-size SiO₂ dust projectiles of varying mass, size, shape and at different collision velocities up to ~56.5 ms^-1. Dust projectiles much larger than 1 mm lead to a small amount of erosion of the target but decimetre targets do not break up. Collisions produce ejecta, which are smaller than the incoming projectile. Projectiles smaller than 1 mm are accreted by a target even at the highest collision velocities. This implies that net accretion of decimetre and larger bodies is possible. Independent of the original size of a considered projectile, after several collisions, all fragments will be of submillimetre size which might then be (re)accreted in the next collision with a larger body. The experimental data suggest that collisional growth through fragmentation and reaccretion is a viable mechanism to form planetesimals.