Comment by xoa | Hacker News Reader

Comment by xoa | Hacker News Reader

For those who haven't seen it, Ars had a fascinating incredibly in-depth "What If" article years ago exploring what might have happened if the damage to Columbia had been noticed in time and it had been determined that reentry was non-survivable so a rescue had to be attempted, "The audacious rescue plan that might have saved space shuttle Columbia":

https://arstechnica.com/science/2016/02/the-audacious-rescue...

It would have been barely on the brink of possible, but with huge risks of its own. Going through the exercise though reveals a lot of details of the shuttle, being in orbit, and our capabilities at the time. It's been discussed a few times over the years on HN, the biggest one being 8 years ago when it was first published [0], but seems timely along with this article.

I'm very glad though going forward that entire class of failure modes is gone for good and that particular possibility for heroics will not be needed. Side mounting the shuttle was always one of the worst design choices amongst many, many bad decisions that went into the STS. Also created one of the worst launch abort profiles of any system ever. Nobody will do that again though.

----

0: https://news.ycombinator.com/item?id=7305224

"Contrary to popular belief, the heat a spacecraft faces on reentry isn't generated by simple friction but rather by ram pressure—the fast-moving shuttle compresses the air in front of it, forming a massive shock zone in which air molecules ionize and break apart."

Finally! Even Arthur C Clarke wrote in one of his novels that it was friction.

I'm also amazed that the shuttle could not be landed by remote control because of a couple switches.

> I'm also amazed that the shuttle could not be landed by remote control because of a couple switches.

A deliberate, but dubious decision. Deploying the gear before it's time is guaranteed catastrophe-- they decided it was important to not leave this under software control.

An independent "arm" switch is a reasonable solution. I'd want to be able to bring the crew back if they were incapacitated, sick, etc. Or even just bring the shuttle back so what went wrong can be determined.

The Astronaut Office pushed strongly against an autoland capability for Shuttle.

Eh, this is a bit of a sophomoric “aha” and “ram pressure” really isn’t anything you’d find in modern aerodynamics describing atmospheric reentry or hypersonics in general.

“Friction” is fine. What’s actually going on is quite complicated and several things at once, not at all reflected by a “well actually it’s this not that”.

We could teach a little course on hypersonics, but meh.

Go on then, give us the info

In fact, the heating is produced by the gas going through a shock. This isn't heating by compression. Adiabatic compression is a reversible process; going through a shock is irreversible. Entropy is generated at the shock.

In more detail: at the shock, the conditions of the gas change over a distance on the order of a mean free path. This means that, locally, the gas is out of thermodynamic equilibrium, with fast molecules running into slow ones. The extremely high gradient of gas conditions causes extremely rapid relaxation to an equilibrium state, a process that very rapidly produces entropy.

As Mach number increases, the density jump across the shock approaches an asymptotic limit, but the heating increases without bound.

As you compress a gas, its temperature rises. The air at the leading edge of the machine is static, meaning its compressed. (This is how pitot tubes measure airspeed - they are simply pressure sensors.)

The temperature rise from the compression matches what one predicts from PV=NkT

A shock wave is unnecessary. It happens with subsonic flow, too. In fact, it's how refrigerators work.

(In ME101 Fluid Mechanics, the Prof showed how a re-entering satellite would melt even if the atmosphere was a frictionless gas.)

> The temperature rise from the compression matches what one predicts from PV=NkT

That's a null statement, since that holds for any (ideal) gas regardless of how it got into its current state.

If you mean, the temperature rise is the same as if it were adiabatically compressed to the density or pressure the gas has after the shock, then that is not true.

If you compress a gas, even an ideal frictionless gas, its temperature rises. Like I said, that's how refrigerators work. As for compression in flight, that's how pitot tubes measure airspeed. It has nothing to do with friction.

On a wing, the temperature rise is highest at the static point on the leading edge, where there is no motion of the air. The temperature declines from there back as the airflow speeds up. If friction was the source of the heat, you'd expect the temperature to be the lowest at the static point, and would increase towards the trailing edge.

BTW, how a diesel engine works is by compressing the air/fuel mixture until it heats up enough to ignite. A gas engine will sometimes "diesel" on after the ignition is turned off.

Yet, he has explained that this is not what happens, and why.

> It has nothing to do with friction.

The point I am making is that gas going through a shock is a fundamentally different process than a gas being adiabatically compressed. The former causes an increase in entropy; the latter ideally does not. While a shock isn't "friction" per se, like friction there is dissipation (and at entry speeds, quite a lot of dissipation).

Are you sure? Consider:

https://www.kimerius.com/app/download/5782475760/Calculation...

The paper does talk about heat from friction, but not at the stagnation point. It also talks about the heating of the surface being less than predicted by adiabatic compression, but that is the heating of the surface, not the heating of the air at the stagnation point.

That paper doesn't appear to have much to do with what we're talking about here. The transfer of heat to the vehicle from the hot shocked gas is another issue entirely.

Anyway, it's foundational supersonic fluid physics that shocks create entropy.

"Shocks are irreversible occurrences in supersonic flows (i.e. the entropy increases)." -- https://en.wikipedia.org/wiki/Shock_(fluid_dynamics)

Reentry with blunt bodies works because most of the dissipation is occurring at the shock, at some distance away from the vehicle, allowing most of the heat (that must be produced from the kinetic energy of the vehicle as it slows) to be carried away.

I have argued with @pfdietz about this before. In the zeal to make the point the majority of leading edge heating arises from the dissipative effects of shocks, they make statements that have to be very carefully parsed to be correct.

You are right. Isentropic compression will result in a higher temperature of the compressed fluid. This is undergraduate gasdynamics. In these supersonic flowfields with high shock strength though, the shock acts as a very dissipative compression mechanism. Behind the shock is a compressed fluid that is much hotter than it would be if compressed to the same pressure isentropically. Shocks are dissipative, like friction, but shocks and friction are not the same thing.

So unless you remove heat from a perfect compressible fluid being compressed (i.e., via a non-adiabatic process), it will always be hotter post-compression. It's just that the compression and heating that shocks effect causes much greater heating than the isentropic case.

In what we call "acreage" where the forward-facing area is limited (e.g., upper surfaces of wings and vehicle bodies), other effects, including turbulent skin friction heating, can dominate the local heating rate.

Thanks for the explanation. My undergraduate fluid mechanics class did get into shock waves, but not the temperature implications of them.

I've seen many offhand accounts that repair would have been impossible or nearly so. Yet a previous shuttle survive re-entry with a missing tile in part, it seems, due to a metal bracket being in place behind the missing tile and it did a good enough job.

So given that, would it have been impossible to space walk around the shuttle, identify missing/damaged tiles and shove some spare material in gaps? Could the shuttle crew have not carried epoxies designed for this purpose? Obviously it would be a repair that might not work, but at least give you a chance.

I think about this now as well when I see Starship tests that knock tiles off of it's heat shield just from vibrations.

A tile on the flat bottom of the orbiter had a much lower thermal loading than the carbon-carbon leading surfaces of the wings. That's why they used carbon-carbon there.

I wonder about this too. Instead of a rescue mission, which would have been extremely difficult and risky, as the article well shows, could NASA have come up with a repair procedure?

Maybe it was completely impossible, but the topic is not even discussed.

I wonder about the various kinds of "improvised repair" things speculated about in the CAIB report. (Jettison cargo to lower thermal loading, packing wing with ice and metal and smoothing geometry).

Columbia made it through almost the entire time of peak heating without a repair. This doesn't mean that a repair would be successful to get all the way to the ground, but it sure seems like it could be extended significantly longer.

Reply on news.ycombinator.com