Disruptive collisions have been regarded as an important process for planet formation, while non-disruptive, small-scale collisions (hereafter called erosive collisions) have been underestimated or neglected by many studies. However, recent studies have suggested that erosive collisions are also important to the growth of planets, because they are much more frequent than disruptive collisions. Although the thresholds of the specific impact energy for disruptive collisions (Q_RD^*) have been investigated well, there is no reliable model for erosive collisions. In this study, we systematically carried out impact simulations of gravity-dominated planetesimals for a wide range of specific impact energy (Q_R) from disruptive collisions (Q_R ~ Q_RD^*) to erosive ones (Q_R << Q_RD^*) using the smoothed particle hydrodynamics method. We found that the ejected mass normalized by the total mass (M_ej/M_tot) depends on the numerical resolution, the target radius (R_tar) and the impact velocity (v_imp), as well as on Q_R, but that it can be nicely scaled by Q_RD^* for the parameter ranges investigated (R_tar = 30-300 km, v_imp = 2-5 km/s). This means that M_ej/M_tot depends only on Q_R/Q_RD^* in these parameter ranges. We confirmed that the collision outcomes for much less erosive collisions (Q_R < 0.01 Q_RD^*) converge to the results of an impact onto a planar target for various impact angles and that M_ej/M_tot = C * QR/QRD* holds. For disruptive collisions (Q_R ~ Q_RD^*), the curvature of the target has a significant effect on Mej/Mtot. We also examined the angle-averaged value of M_ej/M_tot and found that the numerically obtained relation between angle-averaged M_ej/M_tot and Q_R/Q_RD^* is very similar to the cases for 45-degree impacts. We proposed a new erosion model based on our numerical simulations for future research on planet formation with collisional erosion.
Download