Dr Philip J Carter

Research: Crust stripping

Crusts are expected to contain a significant fraction of planet's incompatible elements (elements that are concentrated during magmatic processes because they more strongly partition into the melt than residual solid during partial melting), and it has previously been suggested that preferential removal of the outer crustal layers of planetary building blocks alter bulk compositions (e.g. O’Neill and Palme, 2008; Boujibar et al., 2015). Since planetesimals differentiate into distinct components (i.e. core, mantle and crust) early in the history of the solar system, collisional erosion is an obvious means of removing the chemically distinct outer layers of growing planets.

The modification of planetary compositions during growth has consequences for the long-standing paradigm that primitive meteorites, chondrites, represent the building blocks of planets. Chondritic ratios of cosmochemically refractory elements (elements with high condensation temperatures in a hydrogen-rich environment), such as the rare earth elements, have been taken to provide a robust bulk planetary reference. Yet, some doubt has been cast on the reliability of this assumption (e.g. Boyet and Carlson, 2005).

We carried out simulations using the smoothed hydrodynamics code GADGET-2, to investigate the fate of crusts during collisions between planetesimals and planetary embryos.

Figure 1: Erosive collision between planetesimals with radii of 250 km and 170 km (mass ratio 0.3), at an impact velocity of 1.9 km/s (4.5 v_esc) and impact parameter of 0.4. The numbers indicate the time in hours from the start of the simulation. The target’s core, mantle and crust are coloured orange, blue and cyan; the projectile’s are coloured yellow, purple and lilac. Only SPH particles below the equatorial xy-plane are shown. The view is centered on the largest remnant. More animations can be found here.

We found that crusts are preferentially lost during collisions, and developed a scaling law to describe the mass of crust retained by the largest remnant from the collision. (See Carter et al., 2018 for details).

Using this scaling law and the collision history extracted from N-body simulations of the growth of planetary embryos from Carter et al. 2015 (see here for details), we calculated how much crust is lost during accretion. Figure 3 shows that our results suggest one third to one half of the initial mass of crust can be lost, assuming it is not reaccreted. This loss of crust results in embryos with up to 30% less of the incompatible heat producing elements than the initial bodies from which they grew.