Nuke tests leave asteroid fragments fate still open

A 1-megaton nuclear device detonated metres from a 160-metre asteroid can leave more than 90 percent of the body damaged and much of it flying faster than escape speed, according to new 3D simulations from Lawrence Livermore National Laboratory. The work, led by astrophysicist Isaiah Santistevan and published 14 August 2026 in The Planetary Science Journal, treats the blast as a last-resort planetary-defence tool when warning time is too short for gentler methods.

The catch sits inside the result itself. The longest run captured only 145 milliseconds of physics yet required 59 days on 1,680 processors, so the team could not track whether the pieces keep flying apart or whether gravity pulls some of them back into a still-dangerous clump.

X-rays do the work, not the blast wave

Space is a vacuum. A conventional shock wave has almost nothing to push against, even at a standoff of 10 metres. Roughly 70 to 80 percent of a nuclear explosion’s energy emerges as X-rays. Those photons hit a thin surface layer, heat it violently, and drive vaporization and ablation.

• Vapor expands and, if it exceeds local escape speed, carries momentum away from the asteroid.
• The sudden energy dump launches a strong shock into the interior that cracks and damages rock.
• Height of burst controls both the energy delivered and the surface area illuminated, giving operators a lever that kinetic impactors lack.

Santistevan’s group implemented an earlier X-ray energy deposition model for asteroids developed at LLNL inside the open-source smoothed-particle hydrodynamics code Spheral. That library lets the team skip full radiation-transport costs and focus on the hydrodynamic response.

Three Bennu-shaped runs and a counter-intuitive standoff

The simulated target was a homogeneous 160-metre forsterite body whose overall shape and porosity drew on asteroid Bennu, visited by NASA’s OSIRIS-REx. Fracture behaviour came from laboratory data on the Chelyabinsk meteorite (the 2013 Russian airburst) and the Aba Panu meteorite that fell in Nigeria in 2018. Yield was fixed at 1 megaton. Height of burst and damage-model parameters varied across three cases.

At 10 metres the body ends the run almost completely damaged, with large volumes of material racing in opposite directions. That pattern suggests the asteroid is being pulled apart. Moving the device out to 25 metres reduced the energy that reached the surface yet spread the X-rays over a broader area. At the 68-millisecond mark the more distant burst had already fully damaged a larger fraction of the rock.

The paper’s own summary is cautious. In Nuclear Mitigation of Hypothetical Asteroid Threats the authors write that disruption looks highly likely for two of the three scenarios on the basis of damage extent, material directionality and imparted velocity change. They stop short of claiming “robust disruption,” the technical term for a fragment field in which no piece larger than roughly 10 metres remains a hazard.

Based on the extent of damage, the directionality of the material motion, and the [velocity change] that we impart on these asteroids, we suggest that disruption is highly likely for two of the three scenarios.

The researchers add that such simulations help inform the viability of nuclear options and support emergency-response planning.

DART already proved a softer push works

Nuclear devices sit at the high-energy end of the planetary-defence toolkit. The low-energy end has already been flight-tested. On 26 September 2022 NASA’s Double Asteroid Redirection Test spacecraft struck the 160-metre-class moonlet Dimorphos at roughly 6.6 km/s. Ground telescopes later showed the impact altered Dimorphos orbit by 32 minutes, shortening an 11-hour-55-minute period to 11 hours 23 minutes. That change exceeded the mission’s success threshold by more than a factor of 25. Ejecta recoil amplified the momentum transfer well beyond the spacecraft’s own mass.

Stats snapshot

• DART impact speed: ~22 530 km/h
• Orbit period change: 32 ± 2 minutes
• Success floor: 73 seconds
• Momentum boost factor from ejecta: roughly 2-5 times the spacecraft alone

DART works when years of warning allow a slow nudge. A city-killer discovered only months out, or one too massive for practical kinetic impactors, pushes planners toward higher specific energy. Nuclear devices deliver that energy density. The new Spheral runs map the near-threshold region where a 1-megaton standoff may or may not finish the job cleanly.

The compute wall stops the story early

Every useful number in the paper carries a computational asterisk. Resolving centimetre-scale X-ray deposition on a body hundreds of metres across forces extreme aspect-ratio particles and huge point counts. The longest simulation therefore ends at 145 milliseconds. That is long enough to watch damage saturate and bidirectional velocities appear. It is far too short to watch fragments separate over hours or days, to measure the final size distribution, or to test whether mutual gravity re-assembles a hazardous remnant.

What we know

• Early-time damage fractions exceed 90 percent in the strongest cases.
• Large volumes exceed escape speed and move in opposing directions.
• Farther standoffs can damage more surface area even with lower total fluence.

What stays unconfirmed

• Whether any fragment larger than ~10 m survives.
• Whether gravity can pull damaged material back together.
• Exact Δv and long-term trajectories relative to Earth.

Public conversation around planetary defence has long mixed movie imagery with treaty caution. The simulations sharpen the physics side of that discussion while leaving the hazard side open. Decision-makers still lack a full end-to-end verification that the pieces will miss the planet.

Treaty language and the short-warning problem

The 1967 Outer Space Treaty prohibits stationing nuclear weapons in outer space or placing them in orbit or on celestial bodies. A one-time intercept that flies a device past Earth and detonates it near a threatening asteroid is widely viewed as a different legal category, especially under existential necessity, yet the text itself does not spell out an explicit planetary-defence exception. Any real mission would therefore require rapid international consultation as well as technical readiness.

Recent near-Earth object discoveries with short warning arcs keep the nuclear option on the planning board. Kinetic impactors and ion-beam concepts remain preferred when time allows. The LLNL results supply quantitative damage maps that emergency planners can fold into those contingency trees without claiming the problem is solved.

What the next simulations must answer

Santistevan and colleagues note that definitive checks against common robust-disruption metrics need far longer run times. Higher-resolution or hybrid codes, better damage models that include macroporosity and boulders, and coupling to orbital propagators would close the fragment-hazard gap. Laboratory X-ray shots at facilities such as Sandia’s Z machine already supply ablation data that feed the same models. The 2023 energy-deposition library and the 2026 Spheral disruption runs form successive steps on that ladder.

For now the ironic result stands: a standoff megaton blast can shred a football-field-sized asteroid in milliseconds of simulated time, yet the same physics that makes the early damage look decisive also makes the full safety case too expensive to finish. Nuclear planetary defence therefore remains a carefully informed last resort, not a ready button.