Hacker News new | past | comments | ask | show | jobs | submit login
Chrysler's Radical Space Shuttle Design (thedrive.com)
110 points by rbanffy 51 days ago | hide | past | favorite | 66 comments

> The aerospike engine isn't actually a new thing. Rocketdyne had been experimenting with them since the 1960s, and they're much more efficient than conventional rocket engines.

... kind of. Like they say, conventional bell-nozzle engines are optimized for a particular altitude, while aerospikes are not, and it's true that if you're going to use the same engine across a wide range of altitudes, aerospikes will be more efficient on average across that wide range than conventional engines will. At any particular altitude, though, a conventional engine optimized for that altitude will be more efficient than an aerospike -- they're "jack of all trades, master of none" engines. And it turns out, with contemporary multi-stage rockets, we can and do use different engine bells at different altitudes, which negates a lot of aerospikes' prospective benefits. Were single-stage-to-orbit vehicles ever to be a thing, aerospikes might have their moment, but it turns out SSTO configurations just aren't very good at lifting much mass to orbit, so that seems unlikely.

At least as a layperson, I really appreciated Everyday Astronaut's explanation for why aerospikes never caught on, and perhaps never will: https://youtu.be/D4SaofKCYwo

It digs in much deeper than the usual offhanded, easy explanations like the, "We're stuck with bell nozzles because they were invented first," that this article offers.

Agreed! It's also worth highlighting the bit where he asks Elon Musk: https://youtu.be/cIQ36Kt7UVg?t=382

    I've internally asked this question so many times,
    like, guys, shouldn't we maybe do an aerospike?
Musk focused on the high combustion efficiency of traditional combustion chambers, combined with the high efficiency of having two stages, as the reason why they won out. Jettisoning irrelevant fuel-containing mass as you go is great for efficiency and as a bonus it makes it easy to put sea-level-optimized nozzles on the bottom stage and vacuum-optimized nozzles on the top stage.

The issue isn't really single stage to orbit but rather single stage to vacuum. Every first stage rocket starts at sea level but ends flight in essentially vacuum pressure (over 30km). So there might one day be a role for aerospikes on stages that don't continue all the way to orbit.

Every time I tried running numbers the impression I got was raw performance for the first stage isn't really that important. And the cost vs size is aprox fixedcost + log(size). You can either increase ISP or make the rocket bigger. Increasing the throw of the first stage by 10% might only cost 3% more.

Indeed. The first stage drops off almost immediately, so it pays little to try to optimize it. For example, it makes little sense to use LH2 in a first stage; hydrocarbons are much denser and cheaper.

SSTOs have the advantage of enormously simplified operation - no staging, no vehicle assembly building, no stage mating, just land, fix whatever is broken, refuel and launch. If the payload fraction is smaller than that of a multistage rocket, it can compensate with cheaper logistics on the ground and less time between launches.

SSTOs trade-off is very unfavorable, which is why no one is seriously considering one anymore, for Earth's gravity well. In order to avoid solving the relative simple task of efficient stage mating and assembly, an engineering challenge you have to solve once per launch system and can amortize over the whole life of the system, you end up paying the cost of hauling into orbit a huge amount of unnecessary mass, every single launch. That's a very expensive trade-off.

To take the fanboy's pride as an example, how much time is spent taking a Falcon first stage from landing to re-launch? Days? Weeks? Months? If you are expecting to use your launch infrastructure the same way we use a large commercial airliner then this isn't good enough.

An SSTO could launch, orbit, return, refuel, and re-launch in the same day. For high-value cargo where the mass it not too much (c.f. humans) this might actually make sense.

The cost/benefit formula you're proposing works out most in SSTO's favor if you literally have just one one launch vehicle, but you're using it constantly. If you're allowed to have more than one launch vehicle, though, and can interleave their usage, then that gives you a lot more options for how to maintain throughput while controlling costs. Oftentimes the optimal solution to a multivariate problem ends up looking decidedly sub-optimal when you focus on just one variable at a time.

Also, YAGNI. Right now, it's not certain that we will ever see a time when trips to space are as commonplace as commercial flights. It's far enough off that, even if we take for the sake of argument that it does eventually happen, there's no particular reason to believe that the vehicles we use to do it will be chemical rockets.

> there's no particular reason to believe that the vehicles we use to do it will be chemical rockets.

That too. If we get sci-fi grade engines, anything can be SSTO.

It take 2 to 3 days to reach the ISS from the ground. Getting to orbit is only 8min of that. The rest is a slow approach of the station.

If you want to deliver cargo or people to a specific orbit and a specific place, say the ISS, you're not gonna be able to turnover a SSTO in a day. Even a week might not be enough for a quick return trip to and from the ISS. All this time your expensive ground to orbit capacity is idle.

A first stage can launch a second stage into orbit and return to launch site in less than 15min. How fast you can turn it around for relaunch is an engineering challenge, but there's no fundamental reason you couldn't do it in a day or even in an hour.

Note that two day approaches are not a real requirement, it's a choice that is made for operational flexibility. On Gemini 11 all the way back in 1966 NASA demonstrated docking a mere 94 minutes after launch (essentially one orbit). However, the launch window for that flight was two seconds, which is ... not ideal. By planning for a several day approach you give yourself a several hour launch window, which is much more workable.

There is another reason for the leisurely approach for manned spacecraft. Studies have shown that something like 75% of spacefarers suffer from Space Acclimatization Sickness, generally lasting 1-3 days before your body gets used to zero-gee and you are fine. A leisurely approach lets an astronaut deal with that phase NOT in the shared space station that spends decades in outer space, but in the capsule which will be coming back down in a few months.

SAS is a major limitation on space tourism, incidentally. You'll notice that there are very few plans between 15 minute suborbital hops and two weeks in space: that's because if even trained, physically fit astronauts take 1-3 days to adapt to space, no one is sure what a person who is not as trained will do, and if the word of mouth is "I went into space for three days and felt sick the entire time" that's not a good customer experience.

I guess that by the time we get to serious space tourism, we'll have a rotating station with centrifugal gravity.

But then you can no longer look out the window without puking.

For that you take the elevator to the observatory.

SpaceX claims Starship/Superheavy will be able to launch the first stage thirty times a day and second stage three times a day. I guess we'll see how that works out.

Making those things easier is a lot simpler than making SSTO work.

We thought that reusing the shuttle would be easy and that relanding boosters would be impossible and, yet, we were wrong both times.

Some things are impossible until they aren't.

No these are fundamentally different. SSTOs make things expensive because the rocket equation is unforgiving: every kilogram you haul into orbit needs fuel and propellant, which in turn makes the rocket bigger and heavier, which in turns requires more fuel and propellant, which in turns... So, you don't want to bring up anything unnecessary, like taking your first stage into orbit SSTO style.

Landing boosters is an incredible engineering feat. The rocket equation is just derived from the laws of physics, it's not something you can change with great engineering.

If the rocket equation is such an unforgiving bitch then why are you only using two stages to get to orbit? By your own logic you should be using four or five or even ten stages to get to orbit because otherwise you are hauling up too much unnecessary mass on each flight. What makes two such a magic number in this case? Is it perhaps possible that there are in fact real-world tradeoffs between efficiencies in the rocket equation and engineering parameters and that sometimes advances in technique or materials shifts the optimum point?

The more stages, the less favorable your ratio fuel+propellant / dry mass. When you add stages, you add engines, tanks, systems... that you then have to haul up to the next stage, making your lower stages heavier. Rocket equation again.

Sure it's possible that if you had super lightweight material for all the dry mass, 3 stages would make more sense.

You're right, there is a "real-world tradeoffs between efficiencies in the rocket equation and engineering parameters". This trade-off is unfavorable to SSTOs.

It's because the rocket equation really bites when you get far out at extreme mass ratios, as in an SSTO. Very small changes in mass of systems have a large effect on cargo, and perhaps keep you from even getting to orbit at all.

Viewed another way: if you had an SSTO, you could radically increase its payload to orbit by adding a really dumb first stage. Even a small delta-V from that stage would make the job of the "SSTO" part much easier. And SpaceX has shown that such a first stage can be straightforwardly made reusable, especially if the delta-V allows return to launch site recovery. This basically destroyed any remaining argument for SSTO.

The benefit from going from 2 to 3 stages is much less.

It typically takes two to get the third stage up there, and only the third stage stays in orbit. Alternative shuttle designs, BFR and New Glenn start at two but you wouldn't want to carry around all that extra weight on a second stage unless you had a very specific mission set (launch the ISS in a couple of missions or have a stage that goes and lands on Mars). Not optimized for typical payloads.

IIRC, a lunar probe launched a couple years back on a 5-stage solid rocket. That works well because, essentially, the engine is the fuel.

I didn't say it'd be easy, but that's why people continue developing new materials and engines. There is a point where it becomes practical.

Technically, making those things easier also makes SSTOs work since optimized multi stage orbits let us get to the moon where SSTOs are the default option to get off of the lunar surface. In a sense SSTOs are part of the current earth to moonbase to Mars strategy.

SSTO on the Moon or Mars is a lot easier.

And also already accomplished on the moon. The Apollo lunar module was SSTO during takeoff from the lunar surface.

And make a ton of sense since there be zero to none surface support for relaunching.

You'd probably still need to carry propellants, but yes, an Earth-capable SSTO should be able to hop around a lot on the Moon before needing to refuel.

Land-anywhere and launch-anywhere are also very attractive features for places with sparse support.

I wonder if SpaceX could build a demonstration landing pad, ferry some liquid methane and liquid oxygen to, say, Guam (which is kind of part of the US, making things a bit easier) and do a suborbital hop across the globe with the starship alone without the booster, refuel, and fly back to Florida.

Or, a second stage of a two stage earth-to LEO launcher. There's no reason to make the earth-to-orbit part single stage, just because reuse in space of that upper stage would be SSTO.

Ultimately SSTOs just don't work well on Earth given our currently available propellants. If we lived on Mars the equation would be different, but Earth's gravity well is just too deep for the engines we can make today.

It's not just about finding a new more energy dense but lighter weight propellant either, you need previously nonexistent alloys that can survive the intense heat, pressure, abrasive, and corrosive effects of a running rocket chamber.

One of the big challenges with aerospike engines is keeping them from melting. It's a difficult challenge for bell shaped rocket engines too, but aerospikes are more heavily impacted.

The most viable proposal for an SSTO launcher with significant payload appears to be the Reaction Engines Skylon. It doesn't use an aerospike.


I would not call it viable. The penalty imposed by using air breathing is considerable. Trade off studies of launchers with air breathing components almost invariably optimize to 100% rocket.

> SSTO configurations just aren't very good at lifting much mass to orbit, so that seems unlikely.

Depends what you mean. Performance wise they are not as efficient as multi stage expendable rockets, however they offer possibly the best operational efficiency of rocket system.

The venture star has a launch capability of 20 tons to LEO. Larger lifters like Boeing's proposed 'Big Onion' had a lift capability of 227 tons to LEO. Despite its large size, its fully reusable nature and simple design would have meant its operation would have been considerably cheaper than anything that has flown to date.

I suggest that it would be prudent to test at least one SSTO before the industry writes off the technology. Especially if we had one already 90% complete!

No need. Anything a SSTO can do, staging can do better. Staging will always produce higher payloads to orbit.

The VentureStar has a launch capability of 0 tons to LEO because the fuel tank was too heavy.

The Space Shuttle was considered a reasonably conservative design using a lot of existing technology. Despite this, it had cost overruns north of a billion dollars, and it was late by several years. Chrysler's more unconventional layout would've probably cost even more, and been delayed even longer.

This counterfactual is directly related to the mindset that led to the “conservative” option going upside down value-wise. Instead of viewing the difficulties as being opportunity, the nation chose to play it safe. It lost the attitude towards doing things “. . . not because they are easy, but because they are hard, because that goal will serve to organize and measure the best of our energies and skills, because that challenge is one that we are willing to accept, one we are unwilling to postpone, and one which we intend to win. . .”


I’d argue the Shuttle failed because it wasn’t conservative enough. What we needed was more cost effective versions of the Saturn rockets.

Instead they got infected with the SSTO fantasy, which can never work well. But it boxed them in a corner with the wrong fuel, wrong engines and forced-them to add super expensive SSRBs, which required putting the Shuttle on the side, which led to two disasters and a “reusable” launch system that had to be totally rebuilt between launches.

We’d still have Skylab if they had focused on manufacturing efficiencies for Saturn rockets instead.

In an alternate universe, the Saturn IB could have been evolved to something like the Falcon 9. The H1 engines are in a similar thrust class as the Merlin, although with lower performance.

I've thought about this as I've read the book Coming Home: Reentry and Recovery from Space. The fixation on a space plane probably set back the US space program by decades whereas a more boring evolutionary tree descended from the Saturn program would have been interesting.

I think that this is the right answer. Apollo had a very young, smart, energetic workforce. There are a limited number of these people in the country, and certain projects attract them. Apollo's ambition was very attractive to these people in a way that the shuttle wasn't, and SpaceX is. IMO SpaceX's workforce is the key to their extreme success, and they are able to get that workforce because of their ambition.

Shuttle failed because the question it was answering was "how can we employ engineers to help Nixon's reelection", not "how to we cheaply get to LEO".

This design was also radically unrealistic, due to the shape in particular. Fighting your own reentry decelerator on your way up is a ridiculously ineffective way to reach space, for one. Spike engines are also suboptimal in many ways, the cost isn't exactly the reason they aren't used.

Would it even reach orbit? The drag beyond Mach 1 must be huge, so the compression forces on the "saucer" would be huge too.

The path up is mostly vertical in the densest atmosphere, while the deceleration part is as horizontal as possible. It is fighting the saucer on its way up, but it's nowhere near as effective as it is when used correctly on the way down. I'm pretty sure the Boeing engineers ran the numbers on this one before proposing it to NASA.

There are plenty of paper launch systems that engineers have run numbers on showing they could work, that won’t work in practice. This and VentureStar are two examples.

In this case it’s not the shape as much as carrying way too much extra/dead mass, jet engines, aero spikes, and using a LH2 means much more massive tankage requirements.

In the CGI rendering of how it would have worked - I was surprised the whole thing made it into space.

I thought it would just be a booster like Falcon 9.

I still wonder about another radical idea: launching payloads (not people obviously) via a long ramp with some kind of magnetic accelerator and flinging them into orbit. I forget what this idea was called.

Mass accelerators aren’t useful because the atmosphere limits their final velocity. You’d need to have an enclosed vacuum tube with an exit north of 20,000 feet.

There have been proposals to build something just like that with a “plasma” door at the end that can snap open and closed to keep the atmosphere out of the tube. It still can’t get anywhere near orbital velocity, so the plan is to use it to accelerate a second stage that uses rocket engines after it is released to provide the rest of the orbital DeltaV.

The Problem is rockets are relatively fragile and the high gees and hitting an even thin atmosphere at Mach 6 would destroy them.

Look up “spin launch” for an even crazier idea that’s actually funded.

"Mass driver".

To get into orbit you need 2 pushes. One push upwards puts you at altitude, the other is sideways to make you fall around the Earth and not come back down.

If you have one big push, you escape the Earth's gravitational pull. If your push isn't big enough, you fall back to the ground.

The ramp needs to be very long or we'd need to build a very rugged spacecraft - enough of its brains and propulsion would need to survive the acceleration and the aerodynamic losses on the way up in order to circularize the orbit and not cross the ground on its way back.

Which brings a very scary failure more - being hit by the payload that is mostly dead, but coming back to the ground at near orbital speeds.

I think the hardening is pretty well understood - just fill any sensitive electronics etc. with resin and it's pretty much solved. But the heat problem certainly is not, and obviously humans, not being particularly amenable to being filled with protective resin, would be completely pulped without ridiculously long ramps.

I'm interested to know whether anything ever comes of this: https://www.wired.com/story/inside-spinlaunch-the-space-indu...

I'm sure I recall reading something on a forum somewhere from an anonymous former employee that claimed Spinlaunch was a mess and would never work.

This all reminds me of "The Moon is a Harsh Mistress"

I wouldn't worry about that "nearly orbital" failure mode: rockets are sometimes remotely "dissembled" during flight (boom!) in case of guidance failure. And even if it went totally dead, only if it passively orients itself for a safe re-entry would it not burn up on its way back.

Couldn't you park a "space truck" in orbit, with a normal rocket, and just have the payload be a dumb container. Then shoot the container into orbit using the mass drive and pick it up with the "truck".

It's still might be no good for scientific equipment, but for food and things of that nature might be okay.

This sounds like a lot like the concept called a "sky hook", you can even have it rotate and use it to fling stuff further into space.

Isaac Arthur has some nice videos about it on YouTube.

This is an amazing engine design!

Everyday Astronaut has a lot of material on aerospike engines.

I would question whether NASA didn't think they were still in the game or whether politics and pork-barrelling influenced whether they were in line for funding. Shame they didn't try and self-fund developing the engine.

Every Day austronaut did show though that even theoretical aerospike engines are not a good choice if you don't go single stage to orbit, and even with aerospike engines, an SSTO is worse than a multi-stage rocket, especially if it can be (partially) reused like the falcon.

The added weight of every aero spike engine ever made obviated its performance benefits, even for SSTO. SpaceX has always focused on super high thrust to weight ratio engines (Merlin and Raptor) and that (along with lighter tanks) has more than offset the lower ISP of their propellants.

Indeed. There are ways to get altitude compensation in more traditional bell nozzles.

> The YouTube channel Hazegrayart created an incredible CGI rendering of what Chrysler's craft would've looked like, and how it would've worked.

https://youtu.be/___JNGJog0A (4min, amazing channel BTW.)

Well, what about properly designing the Fiat, first?

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact