Honestly, the most difficult thing in spaceflight history seems to be the protection of the spacecraft in the last leg of returning to Earth.
It doesn’t matter from how far you’re coming home. From earth orbit, you’re dropping in at a little less than 17,000 MPH (28,000 KPH). Apollo astronauts returned from lunar missions at over 25,000 MPH (40,200 KPH). Hot is hot. And shielding is only half of the equation.
Designers of the first manned spacecraft had it easy. Their vehicles had only to return once.
The Complexity of Re-Entry
There’s no way I’m going to attempt to pretend I can summarize well, much less explain, the importance of returning to the atmosphere at the proper angles to the earth. But I will try a bit of discussion in a condensed way before leaving a film that’s far more useful.
Many meteors hit the earth’s atmosphere at a ballistic angle. This angle creates maximum heat and deceleration forces that typically destroy all but the largest space rocks before they come remotely close to the surface. Ballistic entry is also fatal for all but the most hardy unmanned spacecraft (and missile warheads) unless designed to do so. Humans can technically take 20G from a ballistic entry at less than earth-orbital speeds. Serious injury is probable (one Soyuz crew did it) but it is more likely that their spacecraft will be destroyed.
If you make the re-entry angle too shallow, the spacecraft will hit the atmosphere and skip off, unlikely to return.
Enter without shielding to resist or block the heat generated from high-speed friction with the atmosphere, even at the optimum angle for vehicle and crew safety, and there’s no hope at all.
My favorite old NASA films are from the Mission Planning and Analysis Division for Apollo. Their explanations of complex subjects were easy to digest despite trying to relate the most complex elements of orbital mechanics. Here’s an MPAD film on the calculations needed for the Apollo Command Module (and, later, other spacecraft like Orion) to enter at the right place and have enough control to survive a re-entry at speeds 1.5 times higher than dropping out of earth orbit.
Early spacecraft needed a simple one-time way to handle re-entry speeds and heat before 1981.
Ablative shielding works by wicking the heat away by simply burning up and away from the spacecraft body. But it’s not the burning material that helps protect a spacecraft. As burning occurs, secondary but critical reactions occur in ablative material. As the material heats, gases are formed that are “stuck” between the spacecraft and the burning material. This is known as a shock layer.
The shock layer forms a relatively cooler boundary of gases that work in concert with the burning away of material to further insulate the spacecraft exterior (technical term: pyrolysis). Complex calculations are needed to determine the optimum amount of shielding in the right place on every inch of the spacecraft based on the center of gravity (C/G) of the vehicle. No manned spacecraft didn’t just drop in with their shields completely perpendicular to their path. The spacecraft were often designed with a little more weight on one side. This shifted the center of gravity so that the spacecraft fell with the shield a little canted. This made an uneven flow of air, and one side of the spacecraft experienced a lifting force.
That lifting force allows the crew a little control of the spacecraft’s path as they fall. Leaving the spacecraft’s angle in a “lift up” position would allow them to fly farther through the atmosphere during re-entry. A “lift-down” adjustment would make the spacecraft drop steeper, which generally was bad. Drop too fast and deceleration forces greatly increase on both crew and spacecraft, threatening it with overheating and crushing the vehicle. For Apollo re-entry, the lifting control was critical because their speeds were greater than earth orbit velocity. Too much lift-down was bad, but too much lift-up (coming in “shallow”) and the Command Module could skip out of the upper atmosphere. While the Command Module couldn’t quite take advantage of that lifting control at all phases, it certainly helped in fine-adjusting the initial entry angle in case there was any minute error.
Weight was also a factor. The Apollo Command Module was covered tip to bottom with fiberglass honeycomb, forming over 370,000 individual holes to fill in the ablative resin called AVCOAT. But you couldn’t make the shielding too thick all over. The spacecraft would become overweight, lose the proper center of gravity and use more fuel for launch and less for cargo. If the resin was too thin, the shielding would be pointless, the C/G would again be off, and the ship would be heavily damaged if not destroyed on re-entry.
Here’s a film made by Avco Corporation (now Textron Systems), the contractor for all Apollo CM shielding. This film shows the entire process from start to finish. You’ll see that their work is done on Block I Apollo spacecraft (the type that ultimately never flew a crew) but the process was the same for Block II spacecraft except those were not painted blue-gray as the Block I ships.
You should know that the upcoming Orion spacecraft, very similar to that of the Apollo counterpart, will use a version of AVCOAT as well. It’s that dependable.
As said, the downside to ablative shields like AVCOAT is that they can (generally) only be used once. SpaceX is will push the envelope here using a NASA technology, an ablative material called PICA. They improved on the idea (very expensively) to create a variant called PICA-X. It’s still an ablative shield, but more durable. A spacecraft with PICA-X can handle dozens of earth-orbit re-entries before it requires replacement and can handle also lunar or Martian re-entries. There are others.
But ablative shields eventually require replacement. And they’re heavy. So when NASA envisioned the new, much larger Space Shuttle Orbiter, they had to think outside the box and create a lightweight heat shield that required little to no refurbishment between flights.
Enter the ceramic thermal protection system (TPS). Each tile was basically pure, lightweight silica foam. The Shuttle TPS were perfect insulators. One side could be 3,000 degrees F (1,648 C) and the other side of the tile would be room temperature. So, for the leading edges of the wings, the best stuff was a substance called reinforced carbon-carbon (RCC). It could be molded to the aerodynamic shape of the wing’s edge and take over 5,000 degrees F (2,760 C) of heat.
The downside of the Shuttle TPS is that the tiles and RCC were rather fragile. The tiles themselves chipped often to anything hitting them. In the worst case, the tile was destroyed and fell off, leaving the aluminum skin of the Orbiter exposed. Aluminum alloys might melt as low as 700 degrees F or up to 1200 degrees F (371 to 648 C). There was also the matter of size. You couldn’t slap a large slab of the material on the vehicle. The vehicle would flex a little and the manufacturing process couldn’t make large pieces. Thousands of tiles had to be made, custom fitted and individually glued, then later checked to ensure it would not come off.
If the Shuttle TPS seemed risky, it was. The fate of a safe re-entry system always rested in the design of the spacecraft.
America’s first orbital flight gave Mission Control its second heart-stopping moment (the first was Gus Grissom’s Mercury spacecraft sinking and his near-drowning in the flight before).
So here’s John Glenn in his Friendship 7 Mercury spacecraft, hoping to get up to 7 orbits. Around the second orbit, flight controllers get a message from telemetry suggesting that the heat shield of the spacecraft had come loose.
That’s right. I said loose. Landing tests of the spacecraft showed that the splashdown shock was too jarring to the astronaut.
So, after entry but before splashdown, the heat shield was designed to disconnect from the spacecraft, and a thick canvas-like diaphragm called a landing bag would deploy between spacecraft and heat shield, acting as a air cushion on splashdown.
The indicator on the ground said that the shield might have deployed the bag. Once the Mercury’s retrorocket pack, attached on the center of the heat shield, was fired and then jettisoned, ground controllers feared the shield would skew and Glenn had as much chance as a marshmallow above a campfire. So, the flight director instructed Glenn to keep the retrorocket pack on during entry, in hopes that it would keep the shield in place.
Thankfully, the reading on the shield was false; the shield hadn’t come loose, and all that came out of that event was Glenn enduring a rather more energetic re-entry of flaming debris past his window.
All of the Apollo Command Modules, manned and unmanned, returned safely. For the stricken Command Module of Apollo 13, it was no different.
But NASA controllers had to get the crew home with limited power and water. Early on, one option entertained was to jettison the useless, powerless Service Module (SM), rendered dead by the oxygen tank explosion that both drained breathing oxygen and gas used for the fuel cells, which combined oxygen and hydrogen to form water and electricity.
Taking the dead Service Module out of the equation would also make a lighter spacecraft, enabling the descent stage engine of the Lunar Module lifeboat Aquarius to make a faster trip. The jettison would also change the center of gravity of the docked Command Module and LM, making it easier for Jim Lovell to make course corrections with the reaction control thrusters.
But removing the SM would fully expose the thicker, business-end of the heat shield to the extreme cold and heat of shadow and sunlight in space. While the CM’s conical upper body was also covered with the same shielding (although covered also by aluminized strips of Kapton film for thermal control against the sun), the base had no such protection. NASA feared that that shield could crack, or, if the shield had somehow been damaged in the explosion, the limited passive thermal control that Aquarius could muster as it drifted home could make a bad problem worse.
In the end, the SM stayed on as insurance and Apollo 13 coasted slightly slower home to a safe landing.
STS-1, STS-27 and STS-107
With thousands of tiles on the Shuttles, NASA expected a few tiles to suffer minor damage. But what NASA continually underestimated throughout the program’s history was the potential damage of things that fell on the Orbiter during launch.
Ice could condense on the outside of the External Tank (ET), filled with liquid hydrogen and oxygen. While Saturn rockets would show slabs of this stuff falling off the rocket in an Apollo launch, the Orbiter was attached to the side of the ET, where falling ice could damage or destroy the TPS tiles. To eliminate this, heavy coats of orange insulating foam coated the ET. It was such a perfect insulator that it was said that one could put a piece of ice inside one of the tanks and it would take months for it to melt.
The maiden flight of the Space Shuttle, STS-1, began and ended with great potential for success and catastrophic defeat. There were several failures but among them: At launch, miscalculations of noise nearly broke the center body flap needed for re-entry and damaged portions of TPS all over the Shuttle. Then, some tiles were lost on the orbital maneuvering system pods. More tiles were damaged in critical areas: doors on the landing gear and fuel connectors. And the re-entry calculations were not right, which caused Columbia’s already damaged body flap to extend way farther than designed to compensate. OV-102 landed safely this first time.
Anything else around the STS rocket stack could also impinge on the TPS. The tiles were OK if the surface would get wet, but couldn’t handle rain impacting them at hundreds of miles or meters per second. The tile’s glue would be destroyed by any leaks of the Orbiter’s fuels and require cleanup of the toxic hypergolic fuels and tile replacement (this happened before STS-2 flew).
During the launch of STS-27 in late 1988, some of the thermal material used to protect the tops of of the left solid rocket boosters came off, causing a swath of damage along the left side of tiles on Shuttle Atlantis. The damage was inspected in orbit but, partially because of the low resolution of the cameras on the robotic arm and how the crew was forced to send the pictures (this was a secretive military payload), NASA didn’t notice that Atlantis had lost a tile near the nose of the Orbiter, and from a critical high temperature area.
The Orbiter did land safely. But on the ground, the metal where the missing tile had partially melted. Were it not for the steel of an antenna that happened to be in the same location, a melt-through would’ve have occurred and Atlantis would’ve burned up.
A similar event occurred years later for Shuttle Columbia in 2003, and the results were disastrous. This time, the left wing of the Orbiter was hit by the very foam that protected the Orbiter’s ET from allowing condensed ice to form on the tank that could hit the spacecraft. The foam hit in the worst possible place, the RCC edge, where heat would reach over 5,000 degrees F (3,000 C). On re-entry, hot gases entered the aluminum wing’s interior and melted the wing from within. STS-107 is the only crew to be killed during re-entry due to shielding failure (Soyuz 11’s crew were found dead on landing, but they died of asphyxiation when a valve inadvertently opened before re-entry that vented their cabin atmosphere to space).
Back to Conical Spacecraft
Humbled by the loss of two Shuttles and their crews, NASA has returned to conical spacecraft, even using the same kind of heat-shielding for the future Orion spacecraft. It was famously tested on the EFT-1 spaceflight in December 2014. Orion does use a mix of TPS tiling on its upper half but a more durable AVCOAT shield on the leading base.
Likewise, new private companies such as SpaceX, Blue Origin use or are expected to use conical spacecraft and ablative entry systems.
Thermal entry control technologies are continually in development. Wikipedia has a nice summary of them here.