X-Rays and AI Reveal How Spacecraft Heat Shields Burn in Real Time
Heat shields are arguably the most mission-critical components of any spacecraft designed to enter a planetary atmosphere. Without them, the violent physics of atmospheric reentry — generating temperatures exceeding 1,650°C (3,000°F) — would obliterate even the most robust spacecraft structures in seconds. Yet for decades, engineers and materials scientists have operated with a significant blind spot: they could examine heat shields before a mission and after one, but never during the fiery process itself. That has now changed, thanks to a landmark collaboration between NASA and the Advanced Light Source (ALS) at Lawrence Berkeley National Laboratory (LBNL), which has allowed researchers to watch, in real time and in three dimensions, exactly how a heat shield disintegrates under extreme conditions.
The implications of this breakthrough reach far beyond scientific curiosity. As humanity prepares to send crewed missions back to the Moon through the Artemis program and eventually on to Mars, understanding — and predicting — heat shield behavior with far greater precision may literally be the difference between mission success and catastrophic failure.
The Physics of Survival: What Is Ablation?
When a spacecraft plunges into a planetary atmosphere, it does not simply "slow down." It collides with atmospheric molecules at hypersonic velocities, often exceeding 28,000 kilometers per hour (17,500 mph), generating a shockwave of superheated plasma around the vehicle's leading surface. The resulting thermal load is immense — enough to vaporize most metals within seconds.
The primary defense against this inferno is a process known as ablation. Rather than merely reflecting or insulating against this heat, ablative materials are designed to absorb thermal energy and carry it away from the spacecraft by physically decomposing — charring, vaporizing, and shedding material from the shield's outer surface. This controlled self-destruction is counterintuitively elegant: by sacrificing the outer layers of the shield, the spacecraft's interior remains protected. The process has formed the backbone of spacecraft thermal protection since the earliest days of human spaceflight, from the Mercury capsules of the 1960s to the Orion crew module used on the Artemis program today.
"Nearly every major NASA ablative heat shield material has been studied with this technique at the Advanced Light Source, including those relevant to the Artemis and Mars entry missions." — Vishnu Oruganti, NASA Johnson Space Center
However, ablation carries inherent disadvantages that become increasingly consequential as mission complexity grows:
- Single-use limitation: Ablative heat shields are consumed during reentry and must typically be replaced before any subsequent mission, adding significant cost and logistics overhead.
- Difficult simulation: The multiphysics nature of ablation — involving simultaneous chemical decomposition, gas flow through porous structures, radiative heat transfer, and mechanical stress — makes accurate computational modeling extraordinarily complex.
- Unexpected behavior: The uncrewed Artemis I mission in 2022 revealed disturbing anomalies: unexpected charring patterns and uneven loss of ablative material on the Orion heat shield that were not predicted by any of the pre-flight simulations. While the capsule splashed down safely in the Pacific Ocean, the discrepancies raised urgent questions about the fidelity of current modeling tools.
- Limited diagnostic data: Without the ability to observe material behavior in situ, engineers have been unable to directly validate or improve their computational models against real degradation phenomena.
It is precisely this diagnostic gap that the new imaging system at LBNL was designed to bridge.
The Advanced Light Source: A Synchrotron as a Scientific Scalpel
The Advanced Light Source at Lawrence Berkeley National Laboratory is a synchrotron — a type of particle accelerator that generates extraordinarily intense beams of X-ray light by accelerating electrons to near-relativistic speeds around a circular storage ring. The X-rays produced are many orders of magnitude brighter than those generated by conventional laboratory sources, enabling imaging at resolutions and speeds that were previously impossible. Synchrotron light sources like the ALS have found wide application in fields from structural biology to materials science, but applying them to the study of actively ablating spacecraft materials represents a significant and creative new use of this technology.
For this research, scientists at LBNL designed a specialized environmental testing chamber capable of exposing material samples to temperatures of up to 900°C — the threshold at which heat shield materials begin their structural breakdown. The chamber was further engineered to allow researchers to adjust both the atmospheric pressure and gas composition inside, enabling partial simulation of the reentry environment. One notable limitation acknowledged by the team is that air velocity — a significant factor in real-world ablation — could not be replicated in the current chamber configuration, representing a target for future development.
The true power of the setup, however, lies not in the chamber itself but in what it is coupled to: a real-time X-ray micro-computed tomography (micro-CT) imaging system. Micro-CT works similarly to a medical CT scanner, but at vastly higher resolutions — capable of resolving features at the scale of individual micrometers (millionths of a meter). As material samples underwent thermal decomposition inside the chamber, the synchrotron X-ray beam was used to capture continuous, three-dimensional structural images of the material, revealing how its internal architecture evolved second by second under heat stress.
Two Materials, Two Very Different Failure Modes
The research team subjected two prominent NASA ablative materials — both variants of the Sprayed Lightweight Ablator (SLA) family — to this real-time X-ray scrutiny, revealing strikingly different degradation behaviors that have direct implications for spacecraft design and performance.
SLA-561V: The Cork-Based Ablator
SLA-561V incorporates ground cork as a natural organic filler within its composite matrix. Cork, derived from the bark of Quercus suber trees, has long been prized for its thermal and acoustic insulating properties. However, the micro-CT imaging revealed a critical behavior upon heating: the cork particles vaporize rapidly, creating a network of hollow voids within the material. This sudden porosity dramatically alters the mechanical and thermal properties of the shield as ablation progresses, and may contribute to the kind of uneven degradation seen on the Artemis I heat shield.
SLA-220: The Silicone-Elastomer Matrix
SLA-220, by contrast, uses a silicone-elastomer matrix as its structural backbone. Rather than forming discrete voids, this material degrades into a system of interconnected, branching micro-channels that permeate the shield structure. This network allows hot gases generated during ablation to vent rapidly and more uniformly throughout the material, suggesting a more controlled and potentially more predictable ablation process. Understanding these micro-channel dynamics could inform the design of next-generation ablative formulations optimized for specific mission profiles.
The differences between these two degradation modes are not merely academic. They have a profound effect on the thermal protection performance of the shield, the uniformity of material loss, the structural integrity of the shield throughout reentry, and the accuracy of the computational models used to design future missions. Without the ability to directly observe these phenomena in three dimensions and in real time, these distinctions would have remained hidden within post-mission residue analysis.
Bridging the Resolution Gap with Artificial Intelligence
Capturing real-time micro-CT images of an actively ablating material introduces a fundamental engineering trade-off. High-resolution scans require longer acquisition times, risking the loss of transient microstructural changes. Rapid scans, on the other hand, sacrifice the spatial detail necessary to fully characterize the material's evolving architecture. This is a classic tension in scientific imaging, and the research team resolved it through an innovative application of artificial intelligence.
Specifically, the team trained a Generative Adversarial Network (GAN) — a class of deep learning algorithm in which two neural networks compete against each other, one generating candidate images and the other evaluating their fidelity — to upscale intermediate-resolution micro-CT scans into high-resolution equivalents. This approach allowed the researchers to acquire rapid-fire scans during the most dynamic phases of ablation, then use the GAN to reconstruct the fine structural detail that would have required a much slower scan to capture directly. The result was a continuous, high-fidelity, four-dimensional record (three spatial dimensions plus time) of how each material warped, charred, and hollowed out under thermal loading.
This fusion of synchrotron imaging and machine learning represents a methodological advance that could find application well beyond aerospace engineering — from battery materials research to geological core analysis — wherever scientists need to observe the internal structural dynamics of materials under extreme conditions. Learn more about synchrotron applications at the U.S. Department of Energy's Office of Basic Energy Sciences.
What Comes Next: Updating the Models and Expanding the Scope
While the imaging and data collection achievements are substantial, the research team acknowledges that important work remains. The next critical step is to incorporate the new observational data into the computational fluid dynamics and thermochemical models that engineers actually use to design and certify heat shields. The Artemis I anomalies serve as a stark reminder that even extensively validated models can fail to capture real-world phenomena — and the micro-CT data now provides the ground truth necessary to close that gap.
The study also deliberately focused on NASA's own material portfolio. Notably absent from the analysis were the heat shield materials used on commercially operated spacecraft. SpaceX's Dragon capsule uses PICA-X (Phenolic-Impregnated Carbon Ablator - X), a proprietary derivative of NASA's PICA material, while Starship employs a reusable array of ceramic thermal protection tiles — a fundamentally different, non-ablative approach to thermal management. Both represent proprietary technologies, and their internal material architectures are not publicly disclosed. However, given that SpaceX currently accounts for the vast majority of orbital reentries globally, the potential value of applying LBNL's micro-CT methodology to these materials is obvious, and future collaboration — whether formal or commercial — seems a natural evolution.
Looking further ahead, future iterations of this experimental setup may address the current chamber's air velocity limitation, allowing researchers to simulate not only the thermal and chemical environment of reentry but also the aerodynamic shear forces that physically strip ablated material from the shield surface. Incorporating all three factors simultaneously would represent a major step toward truly predictive, physics-based heat shield design. For more on current NASA thermal protection research, visit the NASA Ames Entry Systems and Vehicle Development division, and for Artemis mission updates, explore the ESA Orion overview.
Why This Matters: Safety, Cost, and the Road to Mars
The stakes behind this research extend well beyond engineering elegance. As crewed missions to the Moon under Artemis become operational — with Artemis II scheduled to carry astronauts on a lunar flyby — and as mission planners look toward eventual crewed Mars missions, heat shield performance becomes a human safety issue of the highest order. A Mars entry presents even greater challenges than a lunar return: the thin Martian atmosphere requires a heat shield large enough to generate meaningful drag yet robust enough to survive the intense — if shorter — heating pulse of entry. The Perseverance rover's 2021 entry demonstrated the difficulty of this problem even for robotic missions.
Beyond safety, improved heat shield modeling has direct economic implications. More accurate simulations reduce the need for expensive, large-scale physical testing. Better materials understanding could enable the development of reusable or partially reusable ablative systems, reducing per-mission costs for both government and commercial operators. And for the emerging class of missions targeting the atmospheres of other solar system bodies — including possible future missions to Venus or the outer planets — robust, validated ablation physics will be an absolute prerequisite.
The real-time X-ray imaging work at LBNL represents, in essence, the scientific community finally gaining the ability to see a process it has relied upon for sixty years. What researchers discover in those illuminated microns of charring cork and branching silicone channels may well shape the spacecraft that carry humans to other worlds. Read the original research findings at the Advanced Light Source official news release.
