Hacker News Reader

Comment by kanox | Hacker News Reader

Not yet, not until re-entry. It's arguably the most dangerous part of the missions since launching on Falcon 9 has happened many times before.

Both Soyuz and the Space Shuttle had fatal accidents on re-entry.

View on HN

You're not wrong, but I think it's cool how much safer the generic capsule design is than something like the Shuttle. Really the Shuttle got a whole lot wrong—heat-intolerant aluminum structure and critical, sensitive machinery directly beneath a layer of tiles so fragile you can crush them in your hands (IIRC).

SpaceX's Starship will probably be a lot more tolerant to defects in the heat shield because of the relatively heat-tolerant steel underneath, but landing safely relies on the operation of massive aero surfaces and a relight of the engines with thrust vectoring. It'll be interesting to see how that goes.

This is at the stage where you cant really make assumptions about what may happen because you don't know what could go wrong. (A.K.A you don't know what you don't know).

I'm not claiming that OP or I are experts, but there's actually a pretty big body of knowledge for capsule re-entry. It's more 'known problems' than 'unknown unknowns'.

All you need to bring a traditional capsule design down to ~sea level terminal velocity safely is a blunt body with offset CG and the kind of circular ablative heat shield that's been shown over decades to be simple, safe and reliable. You don't even need active control unless you want more precision for hitting your landing target. The biggest risk, assuming your position and velocity were correct at interface, is the parachutes–which is why SpaceX took so much time testing and refining them.

Things can always go wrong, but we know enough about both classes of vehicle to be able to say that the Shuttle's design was in general terms both more complex and more risky.

This is slightly off topic, but the Shuttle was horribly unsafe on the way up as well as on the way down. Using solid rocket boosters with no means to get the crew safely off the stack until they stop burning is, for lack of a better word, demented. The Crew Dragon/Falcon 9 stack offers realistic abort modes from the moment fueling begins until the vehicle reaches its intended orbit.

The Shuttle was built to bring down soviet spy satellites and spacecraft, so in that regard, it was the best design made.

Huh interesting. Do you have sources for that information?

The main concern of the Soviets was that the Shuttle could swoop down and drop bombs without warning.

(Both Reagan and the Soviet top leadership didn't care about MAD, they cared about practical methods to stop incoming missiles. To get the arms limitations agreements signed, Reagan allowed the Soviets an exception to the anti-ABM ban to allow a ring of anti-ABM sites around Moscow, which exists until today. The "Star Wars" kinetic interception methods are also likely still active, though the laser methods probably never worked well.)

There was at least one mission to retrieve a satellite (not Soviet):

https://www.nytimes.com/1990/01/07/us/space-shuttle-poised-t...

I don't have access to classified military secrets, but you don't think they would build a vehicle that could return 15 tons of payload to earth to retrieve US satellites? US could just build and launch a new satellite easier and cheaper than bring one down to fix only to launch it again.

Obviously there is the movie about Salyut 7 that depicts the shuttle trying to capture the station. Salyut 3 had a machine gun attached to it.

Back in the 70s spy satellites used film cameras and used to return the film in re-entry capsules. There’s a _really_ good (fictional) move called Ice Station Zebra about a retrieval mission in the arctic that goes wrong.

Retrieving the satellite allows you to safely recover the film, and re-equip the satellite with new film and propellant. The latter is important because it’s useful for spy satellites to be able to change their orbits frequently to observe new targets or evade tracking.

Also bear in mind some spy satellites are huge and incredibly costly. The Hubble Space Telescope is based on a spy satellite series.

I read something about naval engagements with soviet ships. The US ships were trying to find something underwater before the soviets did while at the same time not alerting the soviet ships as to the location of what they were trying to find. It became clear they were both looking for one of these film canisters that maybe they weren't able to retrieve in mid-air. People today can't imagine the stakes during the cold war.

It was not built to retrieve Soviet satellites, despite an uncited claim in Wikipedia to the contrary. In fact, it probably couldn't. If nothing else, it'd be hard to fit a satellite designed for a different launch fairing into the payload bay of the shuttle and secure it well enough to be able to land with it. The shuttle was designed to bring cargo back, but only satellites that were specifically designed for this purpose.

What the Shuttle was designed to do in a military context is launch into a polar orbit from the coast of California, release a stealthy spy satellite, and return to Earth immediately after during the very same orbit (so about 90min from launch to landing). It therefore needed the very large wings to be able to glide back to Edwards Air Force Base in southern California, which would have rotated up to 1,500km away from the path of the orbiter. And because of the Pacific ocean, there'd be other landing sites available. This cross-range capability impacted every aspect of the design.

The idea was that big rockets which launch satellites are very easy to track, so it was hard to get a bird up without the enemy knowing its orbital path, and if the enemy knew the orbit they could evade overhead passes. So do a surprise launch on the shuttle, do a little orbit maneuvering while you are up there to maximize the enemy's uncertainty over its orbital plane, then immediately let go and land the shuttle--because otherwise they'd know where it was from tracking the shuttle.

I'm not sure if it was ever used this way, and the requirement is now obsolete anyway. You really can't hide a satellite from modern instruments, which have full global coverage, and its critical that orbits are known to avoid collisions in the much more crowded space up there.

Is the X-37b's orbit known? I dimly recall articles about observers catching glimpses of it places away from its predicted trajectory. Could it maybe be placing/observing/servicing satellites? Why's it a shuttle-style design?

I feel like it's not too outlandish to suspect that some military satellites might have a way to hide themselves from instruments, as well.

Known to the Russians and Chinese? Almost certainly.

The X-37B is a shuttle-like design because it was designed for NASA to be a crew escape vehicle for the ISS. The Air Force only took over the project when NASA cancelled it.

The idea that you could capture a satellite and return it was part of the mission requirements for the original shuttle design. If it was a 'friendly' or 'unfriendly' satellite was left unspecified.

It was also limited by the Shuttle's OMS in terms of delta-v so the orbits from which a satellite could be retrieved was constrained.

Unfriendly satellite can blow up or cause fire inside of shuttle, so it's bad idea to capture enemy satellites. It's like sending billion-worth mine capturer behind enemy lines, which can be destroyed with AK.

Don't see why you're downvoted for posting this. Certainly, satellites designed before the Space Shuttle would have to be very paranoidly designed to include a self-destruct mechanism, but it only takes a couple kilograms of explosives to strongly disincentivize kidnapping it.

If I was a military satellite designer, heaven forbid, I'd definitely include that in a world that included the possibility of stealing satellites. The capability, quietly communicated between the powers, would be enough to ensure that the satellite would never be stolen. And therefore, the billion-dollar kidnapping capability becoming useless.

Even without self-destruct, a satellite not programmed to be captured could simply fire some thrusters if it detects an orbit change, and that would likely make some big dammage inside the capturing shuttle.

As part of just normal station keeping it might do that.

That's typically ion thrusters, isn't it? That's very low mass at very high speed, and would probably take hours to even leave scorch mark. It would feel like a slow breeze.

Of course, military use could do something different, but that's more along the lines of "pack a kilogram of explosives in a pipe bomb".

Hydrazine thrusters more likely. Ion thrusters are novel.

Even without that inclusion you could still communicate it :) Save some money and more payload.

I agree with you that hostile satellite recovery would be fraught with challenges. That said, I expect there are people who 'war game' these scenarios all day and probably have various counters or techniques to prevent damage to the shuttle or its bay.

Seems like it came in handy fixing the Hubble multiple times, but what your talking about was probably overkill for what they ended up using it for.

Including features that are never used is bad design even if they don't compromise every other aspect of the design.

What mainly doomed the shuttle was the DoD requirement that the shuttle be able to pluck a satellite from a polar orbit in one orbit, making it necessary to include the massive wings so that it wouldn't be forced to land in the Pacific as the world rotated underneath it.

If they’d used those features it wouldn’t be public knowledge

Yes it would. There's no such thing as a secret satellite, it's impossible to hide them. General astronomy would notice that their's one less satellite.

> There's no such thing as a secret satellite

No known secret satellites :)

https://en.wikipedia.org/wiki/Misty_(satellite)

You absolutely could not conceal a polar shuttle launch. No such thing ever happened.

Perhaps because it existed, it never had to be used. Deterrence is a feature. An ounce of prevention, as they say.

Considering the complexity and requirements it is amazing that it was as successful as it was.

This is a bit of a partial myth. This idea was probably used when negotiating for funding in some quarters, but it was never a particularly plausible use of it, for at least two reasons. First, of course, the shuttle can only access extremely restricted orbits which most targets of interest would not be in. And second, countermeasures would be trivial if the Soviets believed such a threat to actually exist; just incorporate a few kilos of explosives into each device.

I'm reasonably sure this story was always more of a case of "how do we sell this thing" than "this is something that we can actually realistically do".

After reentry the next most dangerous part of the entire launch/return process is likely landing at sea and being rescued by helicopters or ships.

So I will be holding my metaphoric breath until the astronauts are delivered from SpaceX responsibility to NASA responsibility, which should be on land somewhere in Florida (too lazy to look it up sorry not sorry)

I don’t believe SpaceX has ever not tried and succeeded to return a Dragon sent to ISS back to Earth.

That is true, but it has only had one successful reentry (one out of one attempts) of a Dragon Crew capsule design (which, granted, is only slightly different than Dragon Cargo).

For clarity: I think "tried and not succeeded" is the intended phrasing.

I meant “not (tried and succeeded).” English is great.

Dragon has come back before (without crew) so this isn't a total unknown.

The Crew Dragon is very similar to the Cargo Dragon that has already had nearly 20 returns from orbit without incident, so there is already quite a bit of experience in the systems that will be used tomorrow.

Reply on news.ycombinator.com