Planetary impacts have been extensively studied using numerical simulations. However, simulations resolve length scales larger than 100 km while chemical transfers are efficient on length scales smaller than 1 cm. Thus, existing simulations cannot constrain the efficiency of chemical transfers during an impact.
The novelty of our approach was to develop analog experiments of planetary impacts, using the principles of dimensional analysis to scale down the geophysical system. Our experiments reproduce the large-scale flow observed in impact simulations but also produce turbulence and small-scale mixing, allowing estimation of chemical transfers. We developed two experimental setups to investigate deformation and mixing following planetary impacts.
First, we investigated the impact of a liquid volume (analog for the impactor) into a water pool (analog for the magma ocean) in order to estimate mixing by the impact itself (figure 2). We developed the first release mechanism able to produce large and spherical liquid impactors. Second, we investigated mixing in the aftermath of an impact, when the impactor core falls into the magma ocean (figure 3).
MAIN RESULTS:
Recent fluid dynamical models estimate mixing following an impact, but they entirely neglect the inertia of the impactor. Using our impact experiments (figure 2), we investigated the effect of the impact velocity on mixing between the impactor and the mantle silicates. We obtain scalings for mixing as a function of the impactor velocity and size. Applied to Earth formation, we predict full chemical equilibration for impactors much smaller than the proto-Earth but partial equilibration for giant impacts. We found that the degree of chemical equilibration is up to five times larger than in previous models. Our new scaling therefore affects the interpretation of geochemical data.
With our second experimental setup, we developed the first release mechanism able to inject a dense liquid volume into another immiscible liquid at high speeds (up to 3 m/s) (figure 3). This technical achievement allowed us to mimic the fall of the impactor core into the Earth’s mantle after an impact. We observed fragmentation of the released liquid into drops (figure 3). We obtained scaling laws for the size of the resulting fragments as a function of injection velocity and surface tension. Applied to Earth formation, our scalings predict that the impactor core fragments into sub-millimeter fragments in Earth’s mantle. Such fragments are small enough to allow for chemical equilibration of the entire impactor core.
Our impact experiments were then extended to investigate smaller planetary impacts, as these that formed the craters at the surface of the Earth, the Moon and other planets. In contrast with the giant impacts of Earth formation, a smaller impact does not melt entirely the target material. To model such impacts, we investigated the impact between yield-stress fluids. These are good analogs for solid rocks: they behave as solids under low stress but as liquid above a critical stress. We showed that our experiments reproduce the morphology of impact craters at the surface of planets. We obtained regime diagrams for the crater morphology as a function of the impact velocity, the yield-stress and the impactor size. Then, we characterized the final dispersion of the impactor into the target material. Applied to planetary impacts, these results will help to estimate the impactor size and the impact speed from geological observations.
