Reentry will push Artemis II to about 11 kilometres per second. That's the risky bit.

Why the final 20 minutes matter

Everyone talks about the trip to the Moon—the launch, the translunar injection, and the burn that puts the spacecraft into lunar orbit. But for NASA engineers, managers and the astronauts themselves the trickiest bit happens much closer to home: the return through Earth's atmosphere. That final sequence lasts less than 20 minutes and it will subject the Orion capsule to conditions the agency hasn't tackled with a crew in half a century.

On reentry for Artemis II the capsule will be moving at roughly 11 kilometres per second — roughly 32 times the speed of sound. That's almost twice the reentry speed of a typical return from the International Space Station. At those velocities the interface between vehicle and air becomes extreme.

As the spacecraft ploughs into the upper atmosphere it will form a plasma sheath, a glowing cocoon of ionised gas. Temperatures at the outside of that cocoon can spike to about 2,700 degrees Celsius (4,900 degrees Fahrenheit).

For about six minutes the capsule will probably lose radio contact with mission control and the crew will have to ride out deceleration forces of nearly 4 g's.

And all of that stress rests on a single piece of hardware: the heat shield. Orion's shield uses Avcoat, a material engineered to burn away in a controlled manner so that heat and energy are soaked up instead of being transmitted to the crew cabin. The shield is the literal barrier between astronauts and a white-hot exterior.

Lessons from Artemis I

The current concern isn't hypothetical. NASA already tested Orion on a return from lunar distance during Artemis I, an uncrewed flight. That mission was meant to trial systems and, crucially, to check how the Avcoat heat shield behaved under actual reentry conditions.

The results were unsettling enough to force attention. The Avcoat didn't ablate evenly. Instead of the near-uniform, predictable wear engineers had modeled, the material burned away unevenly and shed more mass than expected. Erosion patterns on the shield didn't match the computer predictions used to plan the capsule's protective envelope.

Thing is, that mismatch matters for several reasons. For one thing, models drive the safety margins. If the real behaviour of the shield diverges from those models, the margin for error shrinks. For another, uneven ablation can create hot spots or change the capsule's aerodynamic balance during the critical deceleration window.

Mastheads

What NASA has to get right before astronauts fly

The Artemis II crew, the first to come back from lunar orbit in about 50 years, face more than just the symbolic weight of history. They're depending on a material and a reentry profile that have already produced surprises. The authorities running the programme will need to be confident that whatever caused the unexpected erosion on Artemis I has been identified and fixed, or at least mitigated.

The timeframe for reentry is tight. From the initial contact with the upper atmosphere to splashdown takes under 20 minutes, and much of the life-or-death action happens in a window only a few minutes long. Communications blackout is part of the deal; for roughly six minutes the capsule and the ground will be blind to one another. During that span the heat shield must work exactly as planned while the crew endure nearly 4 g's of deceleration.

Validating the thermal protection is obviously central to the program. Avcoat is supposed to ablate — to burn away in a controlled, predictable fashion. When it doesn't, engineers have to trace why. Was the material batch inconsistent? Did manufacturing processes introduce defects? Were the models missing physics that only show up at lunar-return velocities? The Artemis I findings mean those questions are now priorities.

Engineering, modelling and risk

Orion's heat shield performance ties directly into the mission's risk calculus. The capsule will be exposed to temperatures that would melt most metals; the Avcoat needs to erode in a way that keeps the underlying structure and the crew compartment safe. That controlled burn is as much about chemistry as it's about aerothermodynamics.

Models and wind-tunnel tests only give an approximation of what really happens. Artemis I provided a rare end-to-end data point at lunar-return speeds and it showed deviation from expectation. Engineers will want repeatable evidence that their fixes work before they saddle astronauts with that risk. The margin between 'works in the lab' and 'works with people aboard at 11 km/s' is the margin NASA now has to close.

Operational implications for Artemis II

Right now, the plan for Artemis II is to carry four astronauts around the Moon and bring them back to Earth inside Orion. If NASA can't satisfactorily explain the erosion patterns from Artemis I and demonstrate a reliable repair or mitigation, the agency will be facing tough choices: delay the crewed flight until more testing is done, change reentry profiles, or alter hardware.

Any of those options affect timelines, costs and astronaut readiness. Delays can cascade into training shifts and scheduling headaches. Changes to the reentry flight profile could reduce stress on the heat shield but might shift other risks elsewhere. Hardware swaps mean additional manufacturing, inspection and verification. The agency's managers will be weighing those trade-offs closely.

Public perception and a high-stakes return

There is also the public side. Artemis II isn't just another mission; it's a visible proof that NASA can safely carry humans beyond low Earth orbit again. The first crewed return to lunar distance in 50 years will draw attention worldwide. If the heat shield showed any sign of acting unpredictably, it would make front-page news. and rightly so.

Still, it's worth remembering the technical teams built Orion and Avcoat to handle extreme conditions. The Artemis I data didn't mean failure, but it did mean surprises. Engineers now have real data to work with rather than just simulations. That's valuable, even if the findings force extra work.

And practical work is what will follow. The teams responsible for the craft and its thermal protection will pore over the Artemis I telemetry, the recovered heat-shield sections and whatever imaging and sensor data exist from reentry. They won't be the first engineers to chase down a mismatch between models and reality; aerospace history is full of similar iterations where an initial flight revealed gaps that got closed before crews flew.

What's at stake

Artemis II's success depends on a short period of brutal physics. The spacecraft will face a plasma sheath, 2,700 °C at the exterior, a six-minute communications blackout and nearly 4 g's of deceleration — all while the heat shield burns away to keep the cabin cool. The erosion seen on Artemis I showed that the real world can surprise even the best-laid models.

Bottom line: reentry is the technical choke point for the Artemis programme. If NASA nails the heat-shield behaviour and proves predictability at lunar-return speeds, the path to carrying humans again beyond low Earth orbit is clear. If not, the agency will need more testing, more fixes, or a rethink of how it brings crews home.

Related Articles

Artemis I's Avcoat heat shield burned unevenly and shed more material than engineers expected, producing erosion patterns that didn't match models.

This article was created with AI assistance.