Lawrence Livermore’s 3D simulations test nuclear options for asteroid defense

Recently, researchers at Lawrence Livermore National Laboratory have begun using 3D simulations to see how a nuclear device might deflect or break apart an asteroid heading toward Earth. Using the Spheral code—a particle-based simulation tool designed to model how materials fracture—the team studied how X-rays from a nuclear explosion could heat, vaporize, and crack an asteroid’s surface and interior. Before applying Spheral to these complex 3D scenarios, the researchers validated it by comparing over 1,500 one-dimensional runs with LLNL’s established Ares deflection code, an earlier tool for simulating asteroid deflection, and found strong agreement between the two.

The simulations focused on an irregular asteroid about 160 meters across—a size that could pose a serious threat to a city, as noted in the study. A nuclear explosion just meters away could generate an intense burst of X-rays, heating the asteroid’s surface and causing material to vaporize. The test asteroid modeled in these simulations had roughly 30.8% porosity and weighed about 3.6 billion kilograms, making it a realistic stand-in for a potential city-level impactor, as the researchers note.

The team tested three disruption scenarios, adjusting the distance of the detonation and the asteroid’s assumed resistance to fracturing. Spheral’s ability to simulate material damage, fracture, and fragmentation made it ideal for this work. In the most detailed simulation, nearly 98% of the asteroid’s material lost its structural integrity, and about 97% of it moved faster than the asteroid’s escape velocity (the speed needed to break free from the asteroid’s gravity)—clear evidence that the body was breaking apart.

The simulations also revealed that the location of the explosion matters. A detonation farther from the asteroid’s surface deposited less total energy but spread the X-rays over a larger area, leading to more widespread damage. The best approach for nuclear mitigation depends on the asteroid’s size, shape, composition, and whether the goal is to nudge it off course or shatter it entirely. The most complex simulation required 1,680 processors running for 59 days, yet it only modeled 145 milliseconds of real-time asteroid behavior.

Researchers caution that longer-term tracking of asteroid fragments is essential to confirm no piece larger than about 10 meters remains and that debris scatters rather than clumps back together. Since the simulations only ran for fractions of a second, they cannot yet confirm whether fragments would disperse or later regroup. But the damage and motion observed in two scenarios strongly suggest that disruption was already underway.

The nuclear blast’s impact relies on intense X-rays rather than a shock wave, as these rays heat the asteroid’s surface, causing material to vaporize and fractures to spread inward. A nuclear explosion releases about 70% to 80% of its energy as X-rays, which cannot travel through a vacuum like a traditional shock wave. This energy deposits in a thin outer layer, vaporizing rock and transferring momentum to the remaining asteroid if the ejected material escapes its gravity.

The team verified Spheral’s accuracy by running 1,555 one-dimensional comparisons with the Ares code, achieving a standard deviation of just 0.8% in blow-off momentum results (the momentum of material ejected from the asteroid). Nearly 88% of these cases showed differences of 1% or less, confirming the tool’s reliability for modeling asteroid responses to nuclear energy.

The simulations included three distinct scenarios: two using fracture properties based on the Chelyabinsk meteorite with detonations at 10 meters and 25 meters above the surface, and a third using properties from the Aba Panu meteorite at 25 meters. Astrophysicist Isaiah Santistevan and his colleagues at LLNL modeled a 1-megaton nuclear device against an asteroid shaped like Bennu, positioning the explosion near the surface to ensure X-rays struck a broad area rather than a single point.

In the scenario with a 10-meter standoff using Chelyabinsk-like fracture properties, 98.2% of the asteroid’s material was fully damaged after 145 milliseconds, and about 97% of it moved faster than the asteroid’s escape velocity. Large sections of the asteroid shifted in opposite directions—roughly 22% moved one way at speeds exceeding twice the escape velocity, while about 23% moved the opposite way at similar speeds. This indicated the asteroid was actively breaking apart.

The 25-meter Chelyabinsk simulation absorbed only 30 kilotons of energy compared to 55.4 kilotons in the 10-meter case, yet after 68 milliseconds, it showed 92.6% of the material fully damaged—higher than the 78.1% damage seen in the nearer burst at the same time. Positioning the explosion farther away reduced the peak energy at any single point but spread the X-rays more evenly across the surface, leading to broader internal cracking. By 68 milliseconds, this scenario also produced a velocity change of 66.73 cm/s, or 8.46 times the asteroid’s escape velocity, further suggesting disruption was occurring.

When using Aba Panu meteorite properties at 25 meters, only 34.2% of the material was fully damaged after 34 milliseconds, compared to 60.2% for the Chelyabinsk model under the same conditions. While velocity changes were similar—around 53 cm/s—the Aba Panu asteroid’s material resisted fracturing more effectively, demonstrating how an asteroid’s composition directly affects the outcome. For true disruption, planners must ensure no fragment larger than about 10 meters remains and that debris spreads out rather than regrouping under gravity.

Because the longest simulation only reached 145 milliseconds, it cannot yet confirm whether any dangerous fragments would persist or later reassemble. However, the findings provide planetary defense experts with a more precise way to assess nuclear standoff explosions for large asteroids or situations with little warning. The distance of the blast cannot be chosen solely to maximize energy at one spot; a larger standoff may distribute X-rays more widely, creating deeper fractures even with less total energy deposited.

Future work will focus on improving simulation resolution, extending runtime, and incorporating more realistic asteroid structures, such as rubble-pile formations. Experiments at facilities like the National Ignition Facility, OMEGA laser, and Sandia’s Z machine will help refine models of X-ray heating and material ablation, reducing uncertainties before anyone makes any real-world nuclear mitigation decision.

The research was recently published in The Planetary Science Journal, with authors including Isaiah B. Santistevan, Mary T. Burkey, J. Michael Owen, Kathryn M. Kumamoto, Jason M. Pearl, and Robert A. Managan. This work is part of LLNL’s efforts in planetary defense, led by the Strategic Threat Analysis and Response directorate.