Physical properties of asteroid Dimorphos as derived from the DART impact

The Double Asteroid Redirection Test (DART) was designed to demonstrate kinetic impact as a planetary defense technique by altering the orbit of a target asteroid. The impact on Dimorphos was highly effective, reducing the moon’s orbital period around Didymos—originally 11 h 55 min—by 33 ± 1 min. Early post-impact imaging by the LICIACube LUKE instrument revealed filamentary ejecta streams extending kilometers from the impact site, and sustained brightening of the system was observed for weeks by telescopes. The observed orbital period change implies a momentum transfer efficiency β that exceeded the incident spacecraft momentum, with β estimated to range from ~2.2 to 4.9 depending on the (as yet unmeasured) mass of Dimorphos. β is defined as the ratio of the target’s post-impact momentum increment to the impactor momentum in the direction of net ejecta momentum, and depends strongly on impact conditions and target material properties (strength, porosity, bulk density, and surface structure). Dimorphos’s mass, bulk density, and surface/subsurface properties were not directly measured by DART, motivating this study. This work numerically simulates the DART impact over a broad range of plausible surface and interior configurations to infer Dimorphos’s material properties by matching the observed momentum enhancement, ejecta cone geometry, and ejected mass.

Prior theoretical and experimental studies show that momentum enhancement β in kinetic impacts depends on target strength, porosity, internal friction, bulk density, and impact geometry. Scaling analyses and code benchmarks (for example, Holsapple & Housen 2012; Raducan et al. 2019; Stickle et al. 2022; Luther et al. 2022) underpin the expected trends. The Bern SPH code used here has been validated against laboratory experiments and applied to asteroid impacts, including Hayabusa2’s Small Carry-on Impactor experiment at Ryugu. Meteorite analogs and spectral data suggest L/LL ordinary chondrites for Dimorphos’s boulders, with grain densities ~3,200–3,600 kg m−3 and microporosities ~8–10%, guiding porosity and strength assumptions. Observations from HST and LICIACube constrain ejecta cone opening angles (~125–135°) and ejecta mass (>1.3–2.2 × 10^7 kg), providing additional benchmarks for model validation.

Impact simulations employed the Bern SPH shock physics code to model DART’s collision with Dimorphos-like rubble-pile targets. The target shape and size were fixed to match Dimorphos (ellipsoid 177 × 174 × 116 m; volume 0.00181 km^3). Targets comprised spherical boulders (>2.5 m diameter) embedded in a granular matrix filling voids; smaller fragments were treated as matrix. Boulder distributions were generated via gravitational collapse N-body simulations (pkdgrav) to reproduce observed boulder SFDs; near-surface particles were selectively removed to match reported topography. Material models: Both boulders and matrix used Tillotson EOS for basalt with grain densities 3,200 or 3,500 kg m−3. Boulders had fixed 10% porosity and average tensile strength ~10 MPa. The matrix porosity (macroporosity + microporosity) spanned 35–65%. Porosity compaction followed a P–α model with parameters informed by terrestrial analog crush curves, bounded between sands and lunar regolith strengths. Strength and friction: The matrix used a pressure-dependent strength (Lundborg) with cohesion Y0 varied from 0 to 500 Pa and damaged internal friction coefficient f from 0.4 to 0.7; cohesion weakened with strain. Relations between f and angle of repose informed plausible bounds. Boulder tensile properties were held constant across runs. Projectile representation: The DART spacecraft was approximated as an under-dense aluminum sphere (radius ~0.52 m, ρ ~1,000 kg m−3) with the spacecraft’s mass (579.4 kg) and impact speed (~6.145 km s−1). This approximation reproduces penetration depth and is acceptable because late-time, low-speed excavation dominates crater growth and ejecta relevant to β. Late-stage approach and runtime: Simulations resolved the first hour post-impact, transitioning to a low-speed integration scheme after shock passage to allow long timescales. Dimorphos’s rotation and Didymos’s gravity were neglected due to short modeled durations. Most production runs used ~5.6 × 10^6 SPH particles; resolution tests up to ~14.5 × 10^6 showed modest differences (few percent) in high-speed ejecta mass and ~6% in cumulative momentum below v/U < 10^3. Parameter space: Fixed parameters included target size/volume, boulder tensile strength (10 MPa), and boulder porosity (10%). Varied parameters were boulder packing (0–50 vol%), grain density (3,200 or 3,500 kg m−3), matrix porosity (35–65%), matrix cohesion (0–500 Pa), and internal friction (0.4–0.7). Impact conditions (location, angle ~17° from normal) matched DART. β calculation: Momentum enhancement was computed via two independent methods: (1) summing momentum of ejecta exceeding escape speed (including Dimorphos gravity) and (2) tracking the asteroid center-of-mass velocity after ejecta reaccumulation (excluding Didymos gravity). Discrepancies between methods provided β uncertainties. Observational comparisons: Simulated ejecta cone opening angles and morphology were compared with LICIACube and HST measurements (opening angle ~125–139°, ejecta cone shadows at T ~178 s), and with observed ejecta directions. Sensitivity tests examined impact location/topography, boulder packing, cohesion, friction, and grain density effects on ejecta and β.

- Momentum enhancement and bulk density: For cohesionless targets, simulations indicate β is relatively insensitive to boulder packing up to ~30 vol%, but drops significantly above ~30–40 vol% due to boulder armoring/interlocking. Matching the observed β implies Dimorphos’s bulk density is less than ~2,400 kg m−3, favoring a more porous, rubble-pile structure. - Surface/subsurface structure: The surface and shallow subsurface likely have a low volume fraction of large boulders (≤ ~40 vol%), consistent with DART imagery and derived boulder SFDs. Macroporosity estimates from observed SFDs are ~34–38%. - Cohesion constraints: Multiple combinations of cohesion, friction, and bulk density can yield the observed deflection, but plausible geological bounds (f ≥ 0.4) constrain cohesion. For f = 0.4 and matrix porosity ~45%, surface cohesion is likely < ~50 Pa; the best-matching scenarios indicate cohesion less than a few pascals. - Ejecta cone morphology: Simulations reproduce wide ejecta cone opening angles consistent with observations. Cohesionless targets produce a wider late-time cone (~140° at 1 m s−1), while cohesive targets (Y0 ~500 Pa) yield narrower cones (~120°). The presence of cone shadows at T ~178 s in LICIACube data implies continued crater growth and low-speed ejecta release, favoring low-cohesion surfaces. - β bounds: For a cohesionless surface, the upper bound on β (~3.6 for plausible densities and porosities) constrains Dimorphos’s bulk density and porosity; increasing bulk density reduces escaping ejecta momentum and β. - Global deformation: The DART impact likely induced global-scale deformation and resurfacing rather than forming a sharply defined crater, consistent with large displaced mass fractions below escape speed and target reshaping in simulations. - Formation implications: Material properties (low cohesion, high porosity, limited fines) support origin via rotational or impact-induced mass shedding from Didymos with subsequent reaccumulation, where fine particles are preferentially lost by radiation pressure. Dimorphos likely differs in cohesion from the faster-spinning primary, Didymos, which may require higher cohesion (tens of pascals) to maintain stability. - Observational consistency: Simulated ejecta momentum direction and cone geometry are consistent with LICIACube/HST constraints within uncertainties, though observed cone segments may not perfectly trace the total momentum vector of the full ejecta plume.

By matching the observed orbital period change and ejecta morphology, the simulations constrain Dimorphos’s physical properties directly relevant to momentum transfer efficiency in kinetic impact deflection. The findings indicate that low cohesion and high porosity enhance β by enabling extensive crater growth and abundant low-speed ejecta that can escape Dimorphos’s low gravity, thereby amplifying momentum transfer. Conversely, high boulder packing fractions (> ~40%) suppress excavation efficiency and reduce β. The inferred low bulk density (< ~2,400 kg m−3) and low cohesion (likely < a few pascals) suggest a weak, rubble-pile structure consistent with prior in situ observations of small asteroids like Ryugu and Bennu. These properties, and the likely paucity of fine-grained regolith within Dimorphos, support a formation scenario via mass shedding and reaccumulation from Didymos, where fines are preferentially lost. The contrast with Didymos’s required higher cohesion to maintain rapid rotation underscores heterogeneity within binary systems. Operationally, the results validate kinetic impact as an effective planetary defense approach while highlighting sensitivity to surface cohesion, porosity, and boulder distribution. The predicted global deformation and resurfacing imply Hera may encounter a reshaped morphology rather than a simple crater, offering critical tests of these inferences. Overall, the work refines scaling from full-scale experiments toward reliable mitigation modeling by isolating key material parameters that govern β and ejecta dynamics.

This study uses validated shock-physics simulations to infer Dimorphos’s properties from the DART impact outcome. Best-matching scenarios indicate a weak, rubble-pile secondary with bulk density below ~2,400 kg m−3, low surface/subsurface boulder packing (≤ ~40 vol%), and very low cohesion (likely below a few pascals). These characteristics explain the observed momentum enhancement, wide ejecta cone, and global deformation, and point to formation via mass shedding and reaccumulation from Didymos. The results advance understanding of momentum transfer physics in small-body impacts and inform design margins for future kinetic deflection missions by quantifying sensitivity to cohesion, porosity, and surface structure. Upcoming measurements by ESA’s Hera mission—particularly Dimorphos’s mass, detailed shape, surface texture, and interior—will critically test and refine these inferences. Future work should incorporate full binary gravity, target rotation over longer timescales, higher resolution to better capture fast ejecta, improved projectile geometry modeling, and exploration of alternative interior architectures beyond the rubble-pile assumptions used here.

- Dynamics and gravity: Simulations neglected Didymos’s gravity and Dimorphos’s rotation due to short modeled timescales, which may affect long-term ejecta trajectories and reaccumulation. - Resolution: Primary runs used ~5.6 × 10^6 SPH particles; resolution tests indicate a few percent uncertainty in high-speed ejecta mass and ~6% in cumulative momentum at lower speeds. - Projectile simplification: The spacecraft was approximated as an under-dense sphere; detailed geometry affects only very fast, early ejecta but could influence fine features of the plume. - Material model assumptions: Porosity compaction parameters and strength/friction values are informed by terrestrial analogs; actual surface crush curves and cohesion are uncertain. Boulder tensile strength was fixed. - Target structure: Only rubble-pile configurations with surface-like interiors were considered; components smaller than 2.5 m were treated as matrix. Alternative deep-interior structures were not explored. - Non-uniqueness: Multiple combinations of cohesion, friction, and porosity can match β; bounds rely on reasonable geologic limits for friction and observed ejecta geometry.