Collisional evolution is a key process in planetesimal formation and decimeter bodies play a key role in the different models. However, the outcome of collisions between two dusty decimeter bodies has never been studied experimentally. Therefore, we carried out microgravity collision experiments in the Bremen drop tower. The agglomerates consist of quartz with irregularly shaped micrometer-sized grains and the mean volume filling factor is 0.437 ± 0.004. The aggregates are cylindrical with 12 cm in height and 12 cm in diameter, and typical masses are 1.5 kg. These are the largest and most massive dust aggregates studied in collisions to date. We observed rebound and fragmentation but no sticking in the velocity range between 0.8 and 25.7 cm s −1 . The critical fragmentation velocity for split up of an aggregate is 16.2 ± 0.4 cm s −1 . At lower velocities the aggregates bounce off each other. In this velocity range, the coefficient of restitution decreases with increasing collision velocity from 0.8 to 0.3. While the aggregates are very weak, the critical specific kinetic energy for fragmentation Q μ=1 is a factor of six larger than expected. Collisions of large bodies in protoplanetary disks are supposed to be much faster and the generation of smaller fragments is likely. In planetary rings, collision velocities are of the order of a few cm s −1 and are thereby in the same range investigated in these experiments. The coefficient of restitution of dust agglomerates and regolith-covered ice particles, which are common in planetary rings, are similar.

The collision dynamics of dusty bodies are crucial for planetesimal formation. Decimeter agglomerates are especially important in the different formation models. Therefore, in continuation of our experiments on mutual decimeter collisions, we investigate collisions of centimeter onto decimeter dust agglomerates in a small drop tower under vacuum conditions (p ∼ < 5·10 −1 mbar) at a mean collision velocity of 6.68±0.67 m s −1 . We use quartz dust with irregularly shaped micrometer grains. Centimeter projectiles with different diameters, masses and heights are used, their typical volume filling factor is Φ p,m = 0.466 ± 0.02. The decimeter agglomerates have a mass of about 1.5 kg, a diameter and height of 12 cm and a mean filling factor of Φ t,m = 0.44 ± 0.004. At lower collision energies only the projectile gets destroyed and mass is transferred to the target. The accretion efficiency decreases with increasing obliquity and increasing difference in filling factor, if the projectile is more compact than the target. The accretion efficiency increases with increasing collision energy for collision energies under a certain threshold. Beyond this threshold at 298 ± 25 mJ catastrophic disruption of the target can be observed. This corresponds to a critical fragmentation strength Q * = 190 ± 16 mJ kg −1 , which is a factor of four larger than expected. Analyses of the projectile fragments show a power law size distribution with average exponent of −3.8 ± 0.3. The mass distributions suggest that the fraction of smallest fragments increases for higher collision energies. This is interesting for impacts of small particle on large target bodies within protoplanetary disks, as smaller fragments couple better to the surrounding gas and re-accretion by gas drag is more likely.Subject headings: planets and satellites: formation -protoplanetary disks 1 johannes.deckers@uni-due.de

We present collision experiments of centimetre projectiles on to decimetre targets, both made up of solid ice, at velocities of 15 m s −1 to 45 m s −1 at an average temperature of T avg = 255.8 ± 0.7 K. In these collisions the centimetre body gets disrupted and part of it sticks to the target. This behaviour can be observed up to an upper threshold, that depends on the projectile size, beyond which there is no mass transfer. In collisions of small particles, as produced by the disruption of the centimetre projectiles, we also find mass transfer to the target. In this way the larger body can gain mass, although the efficiency of the initial mass transfer is rather low. These collision results can be applied to planetesimal formation near the snowline, where evaporation and condensation is expected to produce solid ice. In free fall collisions at velocities up to about 7 m s −1 , we investigated the threshold to fragmentation and coefficient of restitution of centimetre ice spheres.