Lawrence Livermore National Laboratory (LLNL) on October 13, 2026, announced the results of 3D simulations examining how X-rays from a nuclear explosion affect an asteroid. The team modeled a hypothetical asteroid about 160 meters in diameter under three sets of conditions, and two of the cases showed signs that the body was breaking apart. However, even the longest simulation could be followed only up to 145 milliseconds after the explosion, so it remains unconfirmed whether the debris would break into pieces small enough, and spread widely enough, to pose no danger to Earth. The work offers a detailed look at how nuclear fragmentation of an asteroid begins. It does not demonstrate that a collision with Earth could be reliably avoided.
The study was led by LLNL's Isaiah B. Santistevan. The paper he co-authored with Mary T. Burkey and others, "Nuclear Mitigation of Hypothetical Asteroid Threats in Smoothed Particle Hydrodynamics," was published in Volume 7 of The Planetary Science Journal on August 14, 2026 (article number 196, DOI: 10.3847/PSJ/ae8a49). The accepted manuscript, made public by the U.S. Department of Energy's Office of Scientific and Technical Information (OSTI), contains calculation conditions and comparison data that LLNL's announcement does not cover in detail.
Recreating how nuclear X-rays break an asteroid apart in 3D
The research team added a model of X-ray energy deposition on an asteroid's surface to Spheral, a fluid simulation code developed at LLNL. The scenario assumes a nuclear device detonated at a standoff distance from the asteroid's surface. A thin surface layer absorbs the X-rays and heats rapidly, and the vaporized material blows off into space. The resulting recoil changes the asteroid's velocity, which is the basic principle of nuclear "deflection" to avoid an Earth impact. In addition, if the shock wave traveling from the surface into the interior inflicts heavy damage on the rock, the asteroid itself could potentially be fragmented.
Deflection changes an asteroid's velocity slightly while keeping the body largely intact, moving it onto a trajectory that misses Earth. In 2022, NASA's DART mission demonstrated a kinetic method by crashing a spacecraft into the asteroid Dimorphos to alter its orbit. A nuclear approach can deliver far more energy than a spacecraft impact, so it is being studied as an option when an object is large or warning time is short. When fragmenting an asteroid, it is important not merely to break the rock but to disperse the pieces widely enough that they do not threaten Earth.
What distinguishes this study is that it calculates internal damage, crack propagation, and the resulting motion of material inside the asteroid in three dimensions. The team first ran more than 1,500 one-dimensional simulations and compared them with results from an existing code, Ares. After confirming that the two largely agreed across a variety of conditions, they moved on to three 3D fragmentation simulations. The figure of more than 1,500 refers to code-to-code comparisons, not to the number of experiments in which an asteroid was actually destroyed.
The 3D simulations assumed an irregularly shaped asteroid with a sphere-equivalent diameter of about 160 meters. The outer shape was based on a model of the real asteroid Bennu, but the interior was assumed to be a single uniform rock with the properties of forsterite (magnesium-rich olivine). Porosity was set at 0.308 and bulk density at 2.228 g/cm³, so the model does not reproduce Bennu's actual internal structure. The explosive yield of the nuclear device was assumed to be 1 megaton of TNT equivalent (1,000 kilotons). Under these conditions, the researchers examined how detonation position and the fragility of the rock change the way fragmentation proceeds.
Moving the blast farther away reduced energy but increased damage
The simulations compared detonations at 10 meters and 25 meters from the asteroid's surface. Moving the detonation farther away reduces the total energy the asteroid receives. Yet when damage was compared at the same point in time, the 25-meter detonation unexpectedly produced a larger fraction of damaged material.
Comparing the two cases that used the same damage model, based on Tables 3 and 4 of the paper, gives the following:
| Case | Detonation distance from surface | Energy deposited in asteroid (TNT equivalent) | Mass fraction fully damaged at 68 ms |
|---|---|---|---|
| C10 | 10 m | 55.4 kilotons | 78.1% |
| C25 | 25 m | 30.0 kilotons | 92.6% |
In C10, with a 10-meter detonation distance, the energy deposited in the body was equivalent to 55.4 kilotons, compared with 30.0 kilotons in C25 at 25 meters. Even so, 68 milliseconds after the explosion, the fraction of fully damaged material reached 92.6% in C25 versus 78.1% in C10. Both cases used the same damage model, which is based on the Chelyabinsk meteorite that exploded over Russia in 2013. Note that "fully damaged" here refers to a state in the numerical model in which the rock has largely lost its ability to hold together as a single body. It does not mean the rock has broken into pieces of a safe size.
The researchers explain that when the X-rays are delivered from 25 meters away, the energy is spread over a wider area of the surface, which likely contributed to the greater spread of damage. Concentrating strong energy on a small area and delivering weaker energy over a broad area produce different shock effects inside the asteroid. However, this result does not mean that moving the detonation farther away is always better for fragmentation. The appropriate distance depends on the asteroid's size, shape, and composition, and also on whether the goal is to change its orbit or to break it up.
Assumptions about how easily the rock breaks also strongly affected the results. With the detonation distance fixed at 25 meters, the mass fraction fully damaged after 34 milliseconds was 60.2% in C25, which used the Chelyabinsk-based model, but only 34.2% in AP25, which used the Aba Panu meteorite. Even with the same device yield and asteroid diameter, how readily cracks form and spread can greatly change how fragmentation proceeds. These are nonetheless numerical results based on assumed material properties and irradiation conditions, and it has not been confirmed by experiment that the same phenomena would occur in a real asteroid.
Fifty-nine days of computing reproduced just 145 milliseconds
A major challenge in this study was the enormous computing time required. For C10, the longest run, 1,680 computing cores were used for 59 days, yet the simulation reproduced only up to 145 milliseconds after the explosion. Even so, major changes occurred inside the asteroid in that brief time. At the end of the calculation, 98.2% of the mass had reached full damage, and about 97% of the material was moving faster than the escape velocity of the asteroid's own gravity. The escape velocity in this model is 7.89 cm/s, and material was also seen moving in opposite directions.
In LLNL's announcement, Santistevan cited as a key result that damage spread widely through the asteroid and that material was moving at speeds that allow it to escape gravity. Such damage and material motion are strong indicators that fragmentation is under way, he said. But this is an assessment that fragmentation has begun, not evidence that the body has been completely disassembled or that all the fragments have dispersed safely.
To judge that an asteroid has been fragmented and dispersed enough to remove the threat to Earth, it would be necessary to confirm that the remaining fragments are roughly 10 meters or smaller and that they will not re-gather under their mutual gravity. This calculation does not reveal what size of fragments would ultimately form or in which directions they would travel. Even if most of the asteroid's material exceeds escape velocity, that alone does not guarantee that an Earth impact would be avoided. Depending on fragment size, speed, and direction of dispersal, some danger to Earth could remain.
According to the paper, determining whether the debris would disperse sufficiently or re-accumulate under gravity would require calculations covering minutes to hours after the energy is delivered. There is a very large gap between that and the 145-millisecond maximum tracked here. In other words, although the early stage in which damage spreads rapidly inside the asteroid can now be reproduced in detail in 3D, studying what ultimately happens to the material after fragmentation will require improvements in both computing power and methods.
Better accuracy alone will not resolve the open questions
Simulation resolution also needs to be kept in mind when evaluating these results. According to the team, low-resolution calculations may underestimate the peak pressure of the shock wave and thus underestimate the damage inside the rock. On the other hand, if fine cracks cannot be sufficiently reproduced, the fraction of fully damaged material may be overestimated. Higher resolution therefore does not necessarily lead to a more favorable assessment of nuclear fragmentation, and it must be checked which effect dominates.
The paper also points out features thought to stem from the numerical method, such as damage patterns that align with the arrangement of the computational particles. Such patterns do not necessarily reflect the way cracks would actually spread inside real rock. The 3D calculations could not secure sufficient resolution because of computational load, so how accurately they capture the progress of damage and fragmentation remains a matter for future verification. The researchers say that higher-resolution calculations are needed to examine shock propagation and crack growth in detail and to see how much the results change.
The asteroid model also assumes a uniform, single-rock interior, but some real asteroids have a structure known as a "rubble pile," a loose aggregation of many rock fragments. In such bodies, the contact between blocks and internal voids affect how shocks propagate, so results could differ from those found here. In addition to making the calculations faster and higher in resolution, the team plans to extend the model to asteroids with more realistic internal structures and compositions. It also plans comparisons with small-scale experiments, but independent experimental verification of these three cases has not been completed.
This study is an important step toward examining nuclear fragmentation of an asteroid in detail in 3D, from energy deposition on the surface through internal damage to the motion of material. However, to judge the effectiveness of a nuclear approach as a planetary defense measure, it must be confirmed not only that an asteroid begins to break apart, but that no dangerous-sized fragments remain and that the debris disperses safely over a long period. If long-duration simulations that incorporate more realistic internal structures and experimental verification advance, more solid grounds should emerge for judging which asteroids, and with how much warning time, a nuclear approach could be considered.