STS-27 was able survive reentry without a tile because the was over a metal component with sufficient thermal inertia to serve as a heat sink without melting.

Would the steel frame of Starship be able to do the same, or is the skin too thin to transmit heat away from the hole quickly enough without melting?

If yes, is there any way it could handle the mechanical strain of what's probably going to be a mess of shocks & turbulent hypersonic airflow?

  • 1
    $\begingroup$ I guess we might find out on Monday $\endgroup$ Apr 15, 2023 at 19:27
  • $\begingroup$ The first sentence needs clarification: " ... because the was over a metal component ...". $\endgroup$
    – Fred
    Apr 16, 2023 at 11:57
  • $\begingroup$ More accurately "The shuttle wing damage was over a steel antenna instead of the aluminum skin, steel being better able to withstand the heat of re-entry". $\endgroup$ Apr 21, 2023 at 23:43

1 Answer 1


I think to answer this question, we need to define the term "endure" first. We could define three different failure modes:

  1. The vehicle breaks up during reentry,
  2. The vehicle survives reentry, but is permanently taken out of service afterwards,
  3. The vehicle survives reentry, is temporarily taken out of service, refurbished, and re-flight-qualified, but this additional servicing causes the overall system to not be rapidly reusable and thus not economically viable relative to competing systems.

If any of these failure modes were to occur, then we could say that the vehicle failed to endure the loss of a heat shield tile. Although, in the third failure mode, it depends on how frequently tiles are lost in practice, and how much inspection and refurbishment work must be done to reduce the tile loss rate.

There is also the question of whether the white padding material between the tile and the skin will be able to protect the steel. It might be Coated Nomex Felt with a temperature rating of 213°C.

Flame-Resistant Felt - Nomex® (also called meta-aramid felt) is a high-temperature felt that protects against extreme heat.

enter image description here

But, without precise knowledge of what this material is, how much of it there is, and what its properties are, it is not possible to determine whether it will remain intact and protect the steel through reentry. However, as your question focuses whether the skin is too thin to transmit heat away from the hole quickly enough without melting, so let's assume for this question that the white material does not hold up long and the steel ends up being directly exposed to the hypersonic airstream after a short while.

Finally, there are lots of tiles and it seems obvious that the answer will depend on which tile. Tiles at the fringes of the protected area are less critical. Let's assume that we are considering only the more important non-fringe tiles.

Let's now examine some of the data that we have on heating during vehicle reentry...

On the third Space Shuttle flight, the Kuiper Airborne Observatory flew under the orbiter as it reentered and captured an image of it's belly.

Heating on the Space Shuttle during reentry (ref)

As you can see from the image above, temperatures range from 2000 to 3000 °C.

Stainless steel's melting temperature is between 1,400 to 1,530 °C, depending on the specific consistency of the alloy in question. (ref)

The Space Shuttle endured these temperatures for around 13 minutes.

Reentry Altitude versus Time (ref)

A second source says...

Vehicles zooming through the atmosphere at Mach 20 experience temperatures in excess of 3,500 degrees Fahrenheit (1,927 degrees Celsius), DARPA officials said — hot enough to melt stainless steel.

The size of a thermal protection tile is around the size of a dinner plate, and the thickness of the stainless steel skin is about 4mm.

The 4mm thickness will not provide enough thermal mass to prevent the metal from melting during the 13 minutes of heating that will occur during reentry. However, if there were mass in addition to the skin, such as components of the frame, this may help as it did in the case of STS-27. However, note that STS-27 actually experienced more damage than STS-107 (the Columbia disaster), but just not in as critical a location. Atlantis did not really "endure" the loss of its thermal protection tiles, even though it technically "survived" reentry.

The second question we need to consider is whether the heat would be conducted away from the center of the unprotected region to the edges fast enough to prevent the center melting. Again, anyone who is familiar with heating or welding steel will understand intuitively that stainless steel has insufficient thermal conductivity to conduct heat away fast enough.

enter image description here

Next, there is the question of whether the steel could be kept cool by the fuel or oxidizer that is left in the tanks. If a tile were lost on a surface that did not have a propellant tank behind it, the answer would be, of course, no. If there was a tank behind the lost tile, that tank would be almost empty, and so it is unlikely that the vapor remaining inside the tanks would be able to provide significant convective cooling. But if there did happen to be liquid propellant on the other side of the skin, that propellant would probably prevent the steel from melting.

Finally, it is important to note that well before the skin reaches the melting point of stainless steel, it will reach its ignition point. For a reentry on Earth, oxygen in the air will begin to react with the metal causing it to burn, and this reaction will generate additional heat.

To summarize, the answer to your question as defined above is that, for most of the heat shield tiles, it is unlikely that the stainless steel of Starship will be able to endure the loss of the tile during reentry.

But you don't have to take my word for it. Elon Musk himself said, "...if there's any crack in the armor, it's toast.". (ref)

  • 1
    $\begingroup$ Many are predicting that the loss of one tile on Starship would likely not be catastrophic. I don’t think Elon Musk has weighed in publicly, only expressing concern about a zipper effect if one tile falls off leading to the loss of others. However I don't know that anyone has convincingly rebutted the points that you are making. We should find out soon, because it is highly likely that one or more tiles will fall off during the first reentry attempts. If these flight tests confirm your prediction that the loss of even one tile will be catastrophic that will be a huge challenge for the program. $\endgroup$ Feb 2 at 17:31
  • $\begingroup$ The STS-27 damage was larger than Columbia only if counting the hundreds of damaged STS-27 tiles which were chipped and lost some of their borosilicate glass coating (what made white tiles black). I never heard it mentioned that the chipped tiles posed an extreme risk to the vehicle. Not saying that the risk was zero, the coating was there for a reason which is why it was used in the higher temperature critical areas (including the entire underside). But all of the catastrophic predictions seemed to be only about the hypothetical situation where the missing tile had been located elsewhere. $\endgroup$ Feb 2 at 17:33
  • $\begingroup$ I generally don't find that the predictions of the "many" are very accurate when it comes to the Starship program. While I do think that the tile loss problem might ultimately be solvable, after doing the research, I do not think the unprotected stainless steel behind a lost tile will endure reentry. We learned some hard lessons about the loss of tiles during the the shuttle program. We should not underestimate the challenge nor repeat the same (or similar) mistakes. $\endgroup$
    – phil1008
    Feb 2 at 22:40
  • $\begingroup$ It was breached RCC material on Columbia not tiles coming loose but that doesn't change your point about hard lessons. But catastrophe would have likely also occurred if a critical tile had vibrated off of a Shuttle, thus requiring incredibly labor intensive tile application, inspection and maintenance. SpaceX seems to be trying to avoid this by experimenting with simpler methods. If your predictions turn out to be correct on the test flights then they would probably use transpiration cooling in areas where needed rather than employ labor intensive methods to avoid occasional tiles popping off $\endgroup$ Feb 2 at 23:20
  • 1
    $\begingroup$ The whole Artemis/Starship thing is really out on a limb as is. To accomplish a landing this decade (2026 is ridiculous) they can't rely on SpaceX's goal of superfast ship turnaround to be a real thing that soon. Even just turning around the launch pads fast enough will be a challenge. More likely SpaceX will concentrate on its other goal of fast, low cost production, and build a bunch of expendable second stages with no fins, tiles, or header tanks. All built and tested prior to the beginning of the launch campaign. That's the only reliable way to meet the required launch cadence for Artemis. $\endgroup$ Feb 3 at 6:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.