The Artemis II heat shield, NASA agrees, is flawed.
The heat shield is the critical layer at the bottom of a spacecraft that protects it — and the astronauts inside — from searing temperatures upon reentering the Earth’s atmosphere. If the shield fails, the underlying metallic structure could melt, rupture and disintegrate.
And there is no backup, and no way for the astronauts to escape.
NASA officials, however, are confident that despite the known shortcomings of the heat shield, the four Artemis II astronauts will remain alive and comfortable as they arrive at Earth on Friday evening at a speed of nearly 24,000 mph, concluding a 10-day trip to the moon and back.
Extensive analysis and testing of the heat shield material “got us comfortable that we can undertake this mission with lots of margin to spare,” Jared Isaacman, the NASA administrator, said in an interview in January.
However, Charlie Camarda, a former NASA astronaut and an expert on heat shields, says NASA should never have launched Artemis II. The agency does not understand well enough the chances that the heat shield might fail, he says, and the mission, a success so far, could end with the deaths of the astronauts.
“I’m going to pray that nothing happens,” he said during an interview a few days before the launch of Artemis II.
His hunch is that there is a 95% chance that the astronauts will return safely. But that would mean a 1-in-20 odds of a disaster.
Compare that with the roughly 1-in-9-million chance that the International Air Transport Association calculates for dying in a commercial airline accident.
The crux of the disagreement lies in how much certainty is needed when a definitive, perfect answer is impossible.
During Artemis I, a flight without astronauts that circled the moon in 2022, the capsule, known as Orion, survived reentry. Had there been astronauts aboard, they would not have noticed anything amiss.
But when the capsule was pulled out of the ocean, the heat shield — the same design as the one on the Artemis II spacecraft — was unexpectedly pockmarked, with sizable chunks missing.
A couple of years of investigations followed. NASA officials said their analysis looked at what might happen under worst-case assumptions. Those findings, along with changes in the reentry path for the astronauts’ return to Earth during Artemis II, provide a significant safety margin, they have said.
The Artemis II crew is aware of the flight’s risks and how NASA has addressed them. “We have actually been there every single step of the way of the spacecraft being built,” Reid Wiseman, the commander of Artemis II, said in September.
Camarda counters that NASA still does not understand the basic physics of what happened during the Artemis I mission and thus cannot truly say what the worst-case scenario might be.
NASA officials have downplayed concerns about the heat shield.
During a news conference in January 2024, Amit Kshatriya, now the associate administrator of NASA, said the Artemis I heat shield had experienced “unexpected phenomena that we need to make sure we understand perfectly.” But, he said, it provided “very good performance from a thermal protection standpoint.”
Photographs of the Artemis I heat shield remained out of public view until they appeared in a report by the agency’s inspector general, an independent watchdog, in May 2024.
The heat shield is made of a material called Avcoat, similar to what was used during the Apollo program more than 50 years ago. By design, as it absorbs the heat of reentry, it gradually chars and burns off, preventing the heat from reaching the rest of the capsule.
In the investigation of the Artemis I heat shield, engineers concluded that inside some portions of the heat shield, gases built up, and the pressure created cracks, causing chunks of Avcoat to break off suddenly instead of burning slowly and steadily.
For future missions, the Avcoat formula has been modified to make it more porous to allow gases trapped inside to escape.
That left a conundrum of what to do with Artemis II.
For this mission, the heat shield, using the original formula, was already finished and attached to the Orion capsule. Replacing the shield or the entire capsule would have pushed the launch further into the future.
Instead, NASA engineers concluded that a steeper, shorter reentry trajectory would minimize the time during which the vehicle would experience high temperatures and help keep the astronauts safe.
Dan Rasky, a heat shield engineer who retired from NASA in December, shares Camarda’s objection to this decision.
“Let me just give you an analogy,” he said. “If you’re driving down the highway and if there’s pieces of one of your tires that start coming off, do you just keep driving and just hope it’s OK? Or do you pull over and change your tire because you’re worried about a blowout?”
The decision to fly Artemis II as is without changing the heat shield was “not prudent,” he said. “In fact, it’s reckless.”
If the Artemis II heat shield performs as well as the one used on Artemis I, the astronauts will splash down in the Pacific with no problem.
But the situation raises uncomfortable echoes of two of NASA’s worst days: Jan. 28, 1986, when the space shuttle Challenger broke up 73 seconds after launch, and Feb. 1, 2003, when the Columbia shuttle disintegrated on its return from orbit.
For both Challenger and Columbia, warning signs had occurred during earlier shuttle flights. But managers mistakenly drew comfort from the fact that the earlier missions had continued without issue, instead of acting with urgency to fix the problems that later led to the deaths of the Challenger and Columbia astronauts.
Now, the key question for Artemis II and its flawed heat shield: Could cracks form and spread at a catastrophic rate?
Calculating this possibility precisely is extraordinarily difficult.
Simulating the supersonic flow of air molecules around the bottom of a space capsule taxes the fastest of computers. For heat shields, other complex phenomena must be accounted for as well: the flow of heat created by the compression of air molecules and the difficult-to-predict process of how cracks form and propagate in Avcoat, sometimes suddenly.
“What I would do is I would take the time to stand down,” Camarda said. “I would put together a team to develop a real analysis capability,” incorporating all of the underlying physics.
That is not what NASA has done, said Danny Olivas, another former NASA astronaut and a scientist with expertise in the properties of materials.
“Charlie is 100% correct,” Olivas said. “We don’t have a physics-based model for this. It’s an impracticality and almost an impossibility because of the way this material behaves.”
But Olivas, who said he had initially had doubts about launching Artemis II with a known flaw, was ultimately reassured by the analysis that NASA did perform.
Unlike Camarda, Olivas was recruited by NASA to conduct an independent technical review of NASA’s investigation, and he recommended the creation of a larger panel of outside experts, which the space agency did.
Olivas said that NASA’s simulations assumed that if the temperature rose to a certain level inside one of the Avcoat blocks, it would crack, and that if it did crack, a layer of Avcoat would pop off the entire block. That would create a cavity where heating would accelerate and another layer would pop off.
Even with those assumptions, which Olivas described as “conservative,” repeated simulations examining a multitude of reentry variations found that enough of the heat shield would survive, and so would the capsule.
An additional analysis looked at what would happen if an entire block of Avcoat fell off. It concluded that a structure below the heat shield made of carbon fiber and titanium would keep the crew cabin intact through reentry.
He said NASA engineers were cooperative.
“What I will tell you is that every single time I talked to somebody, they had Columbia on the front of their mind,” he said. “They were thankful that I pushed them. They were thankful that I doubted them. They were thankful that I forced them to basically prove it.”
He said that he had also tried to make himself available to any dissenters who might have felt cowed by NASA leadership from raising their concerns, and that he had not heard from any.
That was very different from the culture he encountered in the aftermath of the loss of Columbia, Olivas said.
In January, Isaacman invited Olivas and Camarda to a day of technical presentations where NASA engineers explained their rationale for using the flawed heat shield.
Camarda was not convinced. “NASA definitely does not have the data to show that it’s safe,” he said. “I realized that they were using the same flawed thinking and crude analysis tools, similar to Columbia, similar to Challenger.”
For Olivas, however, that meeting closed a few remaining concerns, and he sent a text message to Wiseman, saying he was confident that NASA had done a good job of mitigating the risk for the Artemis II crew.
“I wouldn’t have told them that, out of respect for them or the family, if I didn’t feel that way,” Olivas said. “I would never have rubber-stamped it for NASA’s sake.”