I have some hope that rotating detonation engines will make aerospikes viable. But I don’t even see them mentioned in this paper.
The idea with the constantly moving flame front is that it spreads the heat out. The limitation with aerospikes is getting enough coolant through the spike. Bells are simpler to cool, which as I understand more than makes up for them needing more cooling.
I think they were mentioned briefly. Aerospikes can work with rdes potentially if the certain versions catch on, but at the end of the day the heat fluxes are even worse for the detonation based engines. The main reason aerospikes don’t make sense is that you adding more area that gets the highest amount of heat flux and your plumbing and cooling jackets becomes a nightmare.
The estimates are that a fully functional RDE will be up to 25% more efficient than a standard bell/deflagration design.
Efficiency means less heat per kg of mass to orbit. And with the rocket equation, that’s a lot less engine per kg of payload.
Efficiency means either less heat, or intensely more incentive to solve the heat problem even with unobtanium. You could make the damned things out of molybdenum, which has 8 times the thermal conductivity of titanium, and a higher melting point.
I wonder if aerospikes were popular as an idea in the J-2 engine ~Apollo era since USA only had gas generator engines (and expander RL-10).
You can't get to very high chamber pressures with those, and then maybe aerospike was a way to work around the limitations.
Then XLR-129 and SSME came along with staged combustion cycle and you can just have higher pressure engines. They can both run at sea level and still have a decent efficiency in vacuum.
The linear aerospike for X-33 was kind of a neat tech demo and fit in with the whole shape of the vehicle and composites, non-tubular tanks and large base area. Maybe too many new things at once in retrospect.
Doesn't seem like a front rotating around the spike would gain that much "spreading out" over a continuous front. At the end of the day, its a spike that narrows to a very small point.
RDEs are far less tested than aerospikes. (In part because you can build a dipshit aerospike in your garage. I don’t know anyone who has made an RDE at home.)
The abstract brings up SSTOs, but has there been anything in recent invention that will make them anything other than the white whale people have been chasing since forever?
Source: worked at a startup that took over the patents for the X-33 next gen shuttle and VentureStar SSTO (aerospike design!)
The Columbia disaster really set back SSTO appetite. Probably the whole reason we got the patents, truly.
SSTOs are, like everything else going to orbit, delimited by weight.
If you are going to make the fuel tanks internal to the vehicle and not something that falls off and sheds their weight mid-flight, you have to get vehicle weight to the absolute minimum. Losing weight has second order effects because it means you now have to carry less fuel so you now have a smaller fuel tank which means the tank weighs less which means you get to carry less fuel… etc.
The key, IMO, is material science advancements, specifically around plastics and composites. Very efficient engine design is matters too, but if you can just bring less mass up with you you can start to approach an achievable fuel weight.
It’s a hard job, you need plastics that can handle orbital temperature cycling (+300 to -300 F every 30 mins), atomic oxygen (nasty corrosion), UV with no atmospheric protection, FST for crew exposure…
Exotic metal alloys can get you around some of these problems, but they can be difficult and expensive to work with. Same issue with high-performance polymers. No free lunches here.
With 3D printing of metals and high-performance composites, you can probably remove additional weight so there’s some light in that tunnel.
But all in all it’s very hard to get out of the gravity well with your fuel in tow and survive the extremes of space. My belief is the first vehicle to pull it off will look like a Swiss cheese of voids and lattices from printing / honeycombs and be made almost entirely out of plastic and carbon fiber.
The 1990s were a lost decade for reusable space flight because instead of chasing incremental improvements to the Space Shuttle (an orbiter with reusable tiles that could be turned around in days, not months) or something like the Falcoln 9 or the fly-back version of Saturn V that O'Neill's students drew in 1979, it was all about SSTO.
SSTO is just marginally possible, if it is possible you need exotic materials and engines and you're never going to get a good payload fraction and adding wings, horizontal takeoff, horizontal landing and such just makes it worse. The one good thing about it is that you get closer to "aircraft-like operations" because in principle you can inspect it, refill it, and relaunch it -- whereas something like the STS or Falcoln 9 or Starship will require stacking up multiple parts for each launch.
My guess is aerospikes are making a comeback though because of interest in hypersonic weapons system. I could also see them being useful for the second stage of something like Starship which mostly operates at high altitudes but has to land at low altitudes. There are a lot of other technical problems, like the thermal management system, which really have to be solved before worrying about that optimization.
Currently the Starship upper stage simply has two different sets of bell nozzles: Three engines with nozzles for atmospheric pressure, and three for vacuum. I wonder how inefficient this really is compared to having just aerospike nozzles.
That's the same as the genesis of the question I asked above. SSTOs are a concept, but given their complete lack of market share, I assume as a non-aerospace engineer that there are valid reasons smart people have not been able to design a competitive one yet.
Similarly, I assume there are valid reasons SpaceX has chosen not to use aerospike Raptors, especially given their well-earned reputation for innovating things everyone else swore couldn't be done. If even they haven't been able to make it work, that's a strong data point as to the state of the art.
I'd argue that the brilliance of SpaceX is the opposite. They stick to technology and markets that are proven and use technically conservative approaches. Falcon 9 is about relentless improvement in small ways, not bold new ideas -- unless you count not getting caught up in the politics and psychology of bold new ideas as a bold new idea.
Sure, they talk about Mars, and in-space refueling seems radical, but they've yet to succeed at doing anything radical... yet.
Rumor has it they were struggling with the payload fraction w/ the first generation of Starship and they switched to a second generation that struggles with blowing up. A big advantage of the two-stage architecture is that you can develop the two stages independently. Presumably they will eventually get Starship to orbit and bring it home, they will have plenty of time to improve it get the payload fraction up just as they did with F9.
I don't think that's true, at least it wasn't conceptually radical. People have noticed the cost of "throwing away" the lower stages for ages, and many approaches have been thought of how not to do that. Take the (partly) renewable SSRBs of the space shuttle program for example, which came down by parachute. Landing a rocket on its tail is also quite an old idea. NASA had several demonstrators demonstrating the concept in flight.
SpaceX took a lot of ideas which had been individually proven before, and then put in the work to perfect them and integrate them in a production ready spacecraft. That is important work and good engineering, but not radical. An aerospike had literally never been flown to orbit at that time (I think still not), so it would have been a way worse fit for the SpaceX method of developing the Falcon 9.
A reusable lower stage with powered landing also had never been flown to orbit at that time. And in contrast to aerospike engines, which had been tested before on the ground [1], you can't do ground testing with rocket stage landings.
I think SpaceX didn't try to develop aerospike nozzles because the advantages probably aren't that large compared to the mixed nozzle design they are currently using. They also reused the same ceramic heat shield material developed for the space shuttle instead of developing something new.
Compare that to the cancelled "VentureStar": It would have used both linear aerospike engines and a new metallic thermal protection system (TPS) instead of a ceramic one. I remember an interview where Musk answered the question of why they aren't doing aerospikes or metallic heat shields etc, that there are many ways to skin a cat. They are only doing one thing that they think will work, which is not necessarily the best possible solution, but potentially faster or cheaper to develop.
Radical in terms of economics, but also radical in its incrementalism.
Falcon 9 was a highly competitive rocket without reuse. If they didn't get reuse to work it would have been a successful project. Reuse of the first stage was a huge cost optimization that put it in a class by itself -- but they they did it radically reused risks.
Contrast that to the X-33 which would have required a large number of new technologies to all work to fly at all.
Fixed-cost pricing was also a radical innovation because it drove SpaceX to do everything it could to lower costs. It was known for a long time that reusing (only) the first stage was a good path to lower costs, the SpaceX business model rewarded them for doing it.
SpaceX is highly technically innovative but it's been so successful because technical innovation has been centered around cost reduction and practicality, not chasing high performance for the sake of high performance.
The SpaceX model might need change to get to Mars because of latency. You can launch a Starship to LEO, have it blow up, and launch another one in a few weeks. If a Starship fails to land on Mars, however, you have to wait another two and a half years to try again. Similarly, SpaceX runs everything by remote control from mission control which is great in LEO but to stick a landing on Mars you need something that flies autonomously.
> Similarly, SpaceX runs everything by remote control from mission control which is great in LEO but to stick a landing on Mars you need something that flies autonomously.
I don't believe the stage landings are remote controlled. I've seen several times where they lost contact with the craft but it landed safely.
It would also be a weird choice because radio connections are way too unreliable to be a single point of failure.
As is landing rockets on the launch tower (or as SpaceX would say: catching them). And I might be wrong, but I believe they are the first to use a crane on the launch tower to stack the rocket. Usually you do that before you roll it out to the pad. They were also the first to fly a full-flow staged-combustion engine. Maybe that one was less radical because prototypes have been around for 60 years, but SpaceX were still the first to actually fly one
This. In my very uninformed opinion the only way we'll get useful SSTO is if we can get a meaningful amount of oxygen from the atmosphere rather than carrying it up in heavy tanks. The failure of Reaction Engines with their SABRE engine is disappointing on this front.
It sounds good at the one sentence level. When you need to write more about the topic, the problem is that oxygen makes up only about 20% of the air. So you have need to accelerate all of this N2 that gives you nothing in energy and the result is a much lower Isp (specific impulse is the thrust per massflow, and all of that N2 is not adding anything to your thrust and increasing your massflow). And you need to be able to pull in enough air to get enough oxygen to drive your engine, so you need very large structures to move all of this unnecessary nitrogen around.
It is possible that only needing one tank rather than two can make up for the dramatic loss of Isp we see from an air-breathing engine and the air-handling structure, but no one has yet managed to demonstrate that, and the general consensus runs against it. I recall reading that HOTOL (https://en.wikipedia.org/wiki/British_Aerospace_HOTOL) calculations were actually driven by an extremely light structure estimate rather than the airbreathing engine, to the point where if you plugged a rocket engine in they would actually get more payload to space as a SSTO, because those aggressively light structure estimates were doing all of the work.
SpaceX is very close to demonstrating an architecture that ameliorates almost all of the drawbacks of two stage to orbit architectures. The tyranny of the rocket equation ensures that while a SSTO carrying all of it's oxygen is possible, it's never going to be able to carry enough mass to be useful.
Therefore nobody is ever going to invest the tens of billions required to develop a rocket based SSTO.
If somebody develops an engine that makes air breathing most of the way to orbit feasible, this has a chance of competing a Starship style architecture.
For the reasons you espoused, this is highly unlikely. However "highly unlikely" is more likely than "never".
Jet engines have on the order of 10x the specific impulse of a chemical rocket.
Atmospheric density reduces exponentially with altitude, which implies that you would need to go exponentially faster to maintain mass flow into your engines and lift over your wings. The truth is that breathing air only gets you a third of the way to space, at best, so you have to have a rocket, and now you're battling that complexity. If your space plane doesn't breathe air, it probably is just better to punch your way out the way conventional rockets do.
Of course, the rocket equation is logarithmic, so reducing the amount of mass you loft gives you an exponential gain. This is true for all propulsion systems to an extent (different constants) but getting into space is the hardest propulsion problem we face. A space plane may or may not be better in this regard (it's been a while since I've looked into that kind of thing, so no opinion either way) but imo the inherent complexity is enough on its own to kill the idea.
Only because traditionally the airplane industry measures specific impulse on just fuel flow, completely ignoring the oxidizer and atmospheric nitrogen. If you calculate like for like, including the air, jet airplanes have significantly worse Isp than a rocket engine.
They get this fictitious specific impulse by scaling the effective exhaust velocity by a scaling parameter to account for the fact that the exhaust mass flow consists of extra mass (air), in addition to what is carried onboard (fuel). This specific impulse is still used for comparing jet engines based on efficiency. Another use for it is to calculate space mission requirements in launchers utilizing air-breathing engines in their first stage (as your parent commenter points out). Though such vehicles don't exist yet, there are concepts being pursued. Some of them use a scramjet and others are more elaborate like the (cancelled?) SABRE engine. So those Isps are not completely meaningless.
The general idea is that you can get much better results in terms of deltav if you can find at least part of the reaction mass from elsewhere without carrying it onboard. Even inert nitrogen is useful as a reaction mass. Another way to get a good result is to use separate sources of reaction mass and energy. Then use that energy to accelerate the reaction mass as much as possible, so that you get a decent deltav by the time you exhaust the reaction mass. This is what ion and plasma thrusters do.
There are air-breathing rockets, some of the oldest were ultimately canceled soviet projects for road-mobile ICBMs (canceled for reasons AFAIK unrelated to air breathing concept), and the recent Meteor air-to-air missile
Aren't rockets more powerful (as in energy/time) than rocket engines in that they are getting compressed/liquified oxygen out of a tank as opposed to taking the comparably tiny amount that passes into the intake of an engine?
There are two performance parameters for a rocket/jet engine. The first is thrust and the second is specific impulse. You are thinking about thrust. The others in the tread are talking about specific impulse. Thrust is important for some stages, especially the early stage booster engines (as opposed to later stage sustainer engines). As a simple example, any space rocket will need a first stage with an enormous thrust so that it can lift itself, the subsequent stages and the payload off the launch pad. Additionally, the rocket has to finish its initial vertical climb as fast as possible. Otherwise the propellants will be wasted in just lifting off (this is called gravity loss). That will also require a high initial thrust.
However, the requirement of the high thrust disappears once you finish the vertical climb. There's no danger of falling back to ground once you reach orbit. What you need at this stage instead, is to add velocity (deltav) to the craft to change its orbit/trajectory. This can be done even at very low thrust, because you have all the time you need. The limiting factor now is that you have only a finite amount of propellant onboard. You want to add as much deltav as possible before it runs out. A high thrust doesn't help because the engine will simply consume the propellant faster and exhaust it before you get the required deltav. This is where specific impulse comes into play. The maximum deltav you can get is proportional to the specific impulse of the engine (see rocket equation for details). As you can imagine, high specific impulse is critical for space missions requiring high deltav, like the New Horizons spacecraft that imaged Pluto or the Parker solar probe (interestingly, getting to the sun is harder than escaping the solar system). Rockets/jets with low thrust and high specific impulse are called sustainers.
The general trend seen is that specific impulse drops off as thrust increases. For example, the space shuttle solid booster has Tmax = 15 MN, Isp = 268s, and space shuttle orbiter cryogenic engine RS25 has Tmax = 2.28 MN, Isp = 452s. Meanwhile, the NEXT xenon ion thruster used in the DART mission has Tmax = 236 mN and Isp = 4200s. Note that the thrust has changed from Mega newtons to milli newtons. You would hardly recognize it if the ion engine thrusted against your body.
doesn't scale well. The amount of air entering is proportional to square - cross-section - while the mass of rocket is cubic. While scramjet/turbojet/air-augmentation, say as a separate detachable stage, can be pretty efficient for smaller rocket, anything making significant improvement for say Starship would looks like a fat monster cross-section-wise with tremendous hardware cost and weight loosing outright to the straight option of adding additional tanks and rocket engines.
Wrt. aerospike engine - sounds nice, yet hardware wise it is heavier than the classic engine, and just look at that large number of pieces - just all those small mini-engines - it is made of and compare to Raptor 3. And for the optimal expansion - i'm waiting somebody will add a dynamically adjusting telescopic kind of end section to the classic bell nozzle.
A napkin to illustrate. Lets say you add a Raptor and 80 tons of fuel plus oxygen for it. That will give you 100 seconds of excess impulse of at least 160 tons (240 ton of thrust minus 80 tons) at the beginning to 240 tons at the end, so roughly 100 seconds of 200 tons. To get 200 tons thrust you'd need 20 fighter turbojet engines capable of at least Mach 3 - that is cost, complexity and weight dwarfing that one Raptor engine.
For scramjet, assuming we got a decent one, napkin is about the same. The best, my favorite, is air-augmented - scram-compress the air and channel it on the outside of the hot bell nozzles of the already working rocket engines - unfortunately the scaling mentioned above comes into play for meaningfully sized rockets though it has worked great for small ones.
That really depends on how fast you can cruise. High speed scramjets above mach 15 will make space missions possible. The craft will be at the sufficient height and just enough speed, so that a rocket engine won't have to add too much deltav. Scramjets are still in their infancy. There are already developments on for variable-geometry multi-mode ramjets for this purpose.
PS: I have seen early-stage (but successfully tested) scramjets being developed for this purpose.
I don't think any of the considerations you mention there change my conclusion.
We have to ask: what exactly is a scramjet vehicle delivering? It's enabling the use of air instead of liquid oxygen. But how valuable is this? LOX is the second cheapest industrial liquid after water. The fuel part of a rocket propellant combination typically dominates the propellant cost. If a scramjet launcher uses more fuel (especially hydrogen) than a rocket vehicle would, it will end up increasing propellant cost per unit payload to orbit. It will also likely increase propellant volume per unit payload to orbit, especially if LH2 is used (LH2 being just 5% of the density of LOX).
All scramjet launchers need a rocket to reach stable orbit (since a scramjet cannot produce thrust at apogee to circularize above the atmosphere. So one can ask, what the tradeoff between the delta-V this rocket provides and that of the scramjets? From what I've heard, all such trade studies end up optimizing to 100% rocket and 0% scramjet.
As the other commenter already pointed out, it's the mass that's the limiting factor here, not the cost. The key idea here is that rockets and jets need two things - a reaction mass and energy. Scramjets and other air breathing engines don't just take oxygen from the atmosphere. They derive much of the reaction mass also from it. Even the inert nitrogen absorbs heat from combustion and acts as reaction mass. The primary purpose of the fuel onboard is to provide the energy. It's contribution as reaction mass is only secondary (note that this is for air breathing engines). This is very evident in the case of turbofan engines, where much of the thrust is contributed by the uncombusted air from the fan.
A scramjet stage will be very light compared to an equivalent rocket stage, since it carries only the energy source (fuel) and not the full reaction mass. If this scramjet stage is able to impart a velocity close to the orbital velocity by the time it reaches the upper atmosphere, the subsequent rocket stage will have much less work to do to get it into orbit. And that translates to much less propellants (including oxidizer) and much less mass in the upper stage. It's not necessary to collect oxygen from the atmosphere to see an advantage.
Obviously, the raising of the perigee at apogee is going to need this rocket engine again. There are no launcher concepts that depend purely on scramjets.
What is the point of saving mass here? LOX is cheap, so it isn't the cost of the LOX. Does saving LOX make the vehicle cheaper? No... it increases the quantity of fuel needed, which (particularly if it's LH2) makes the empty vehicle much larger and more massive. This is doubly bad, since every last gram of that empty mass is taken to orbit, unlike the mass of LOX.
Minimizing fueled mass of the vehicle is a stupid thing to do. It's optimizing the wrong metric.
Scramjets also suffer from bad thrust/mass and thrust/$ ratios compared to rocket engines.
Overall scramjet launch vehicles are an example of pyrrhic engineering: even if one could make such a vehicle "work", no one would want it.
I'm looking at the issue deeply, not using the superficial reasoning you seem to be engaging in. This reminds me of "mark logic" where if you are promised pie in the sky you commit yourself to believing it even if it's a scam.
Scramjet launchers (and air breathing launchers in general) have been looked at extensively, for decades. They don't make any sense. The math just doesn't work. Here's an insightful post from the Old Usenet that illustrates the difficulty:
LOX halfway to orbit is significantly more expensive than the same LOX delivered to the launchpad.
Its not the cost, its the mass you're trying to reduce.
So far, the engineering challenges have made it unfeasible, but its not a surprise that people look at the hundred tons of LOX on a rocket and imagine exchanging it for payload (or aircraft style re-usability).
A gram of oxygen that you carried to orbit is more valuable than a gram of oxygen you collect at that location: oxygen that you carried is moving at the same velocity as you.
Why make an SSTO when you can make a TSTO? First stage recovery is a solved problem and will always greatly relax the engineering problems over making a SSTO.
The argument for SSTOs was that staging was too scary. But experience since then shows this argument was bogus. Staging can be made highly reliable.
Even a small amount of delta V provided by a first stage makes the job of the "almost SSTO" second stage much easier. And a low delta V first stage can be rugged, with high high safety factors, and is easy to recover at the launch site.
Put another way: if you have an SSTO, its payload increases dramatically if you stack it on a very low performance recoverable first stage.
I don't see any way SSTOs are going to be preferable to TSTOs, especially if the SSTO has to use hydrogen to get off the ground.
Against the cost of a lower payload fraction. Say you get 4% with TSTO, 2% with SSTO. And if you run into trouble with the design, the SSTO payload moves towards 0%. Your operational savings need to compensate for a doubling of cost per unit mass payload, possibly much more.
And even though full reusability is much easier with two stages, and has not been achieved yet.
The article has a link (citation 17) to a site selling toy models of that vehicle as an image source, I can find one (fourth image in the gallery) there where it sorta looks like the wing is missing because the wing is black against a black background but it's not the same image shown in the paper:
The name "Venturestar" is properly rendered in that image but "NASA" and "Lockheed Martin" are thoroughly mangled the way I'd expect text to be mangled in an AI image. The image from the toy site could have been used as as reference image to create the image in the paper one way or another.
Yes, if you look close, the paper is replete with error-filled generative reproductions of existing illustrations in the citations; including Fig. 6 (MC Escher struts), Fig. 7 (sprouting greeble tubes), and Fig. 8 (actuators replaced by tubes connected to mystery manifolds).
Even Fig. 2 shows the spike geometry magically changing, which is not addressed in the text and seems like an error carried over from the original illustration in the cited source.
Casts serious doubt on the credibility of the rest of the work.
I have some hope that rotating detonation engines will make aerospikes viable. But I don’t even see them mentioned in this paper.
The idea with the constantly moving flame front is that it spreads the heat out. The limitation with aerospikes is getting enough coolant through the spike. Bells are simpler to cool, which as I understand more than makes up for them needing more cooling.
I think they were mentioned briefly. Aerospikes can work with rdes potentially if the certain versions catch on, but at the end of the day the heat fluxes are even worse for the detonation based engines. The main reason aerospikes don’t make sense is that you adding more area that gets the highest amount of heat flux and your plumbing and cooling jackets becomes a nightmare.
The estimates are that a fully functional RDE will be up to 25% more efficient than a standard bell/deflagration design.
Efficiency means less heat per kg of mass to orbit. And with the rocket equation, that’s a lot less engine per kg of payload.
Efficiency means either less heat, or intensely more incentive to solve the heat problem even with unobtanium. You could make the damned things out of molybdenum, which has 8 times the thermal conductivity of titanium, and a higher melting point.
I wonder if aerospikes were popular as an idea in the J-2 engine ~Apollo era since USA only had gas generator engines (and expander RL-10).
You can't get to very high chamber pressures with those, and then maybe aerospike was a way to work around the limitations.
Then XLR-129 and SSME came along with staged combustion cycle and you can just have higher pressure engines. They can both run at sea level and still have a decent efficiency in vacuum.
The linear aerospike for X-33 was kind of a neat tech demo and fit in with the whole shape of the vehicle and composites, non-tubular tanks and large base area. Maybe too many new things at once in retrospect.
Doesn't seem like a front rotating around the spike would gain that much "spreading out" over a continuous front. At the end of the day, its a spike that narrows to a very small point.
RDEs are far less tested than aerospikes. (In part because you can build a dipshit aerospike in your garage. I don’t know anyone who has made an RDE at home.)
I know one idiot on YouTube who has made an RDE at home. He was smart enough to leave the room while it fired.
Tomatoes are disgusting.
I know Integza tried, but did he ever confirm his engine was doing a rotating detonation?
He said near the end of the video 2 years ago, "I'm not sure if I got a rotating detonation wave": https://www.youtube.com/watch?v=fRMMSyCcTDI
The abstract brings up SSTOs, but has there been anything in recent invention that will make them anything other than the white whale people have been chasing since forever?
Source: worked at a startup that took over the patents for the X-33 next gen shuttle and VentureStar SSTO (aerospike design!)
The Columbia disaster really set back SSTO appetite. Probably the whole reason we got the patents, truly.
SSTOs are, like everything else going to orbit, delimited by weight.
If you are going to make the fuel tanks internal to the vehicle and not something that falls off and sheds their weight mid-flight, you have to get vehicle weight to the absolute minimum. Losing weight has second order effects because it means you now have to carry less fuel so you now have a smaller fuel tank which means the tank weighs less which means you get to carry less fuel… etc.
The key, IMO, is material science advancements, specifically around plastics and composites. Very efficient engine design is matters too, but if you can just bring less mass up with you you can start to approach an achievable fuel weight.
It’s a hard job, you need plastics that can handle orbital temperature cycling (+300 to -300 F every 30 mins), atomic oxygen (nasty corrosion), UV with no atmospheric protection, FST for crew exposure…
Exotic metal alloys can get you around some of these problems, but they can be difficult and expensive to work with. Same issue with high-performance polymers. No free lunches here.
With 3D printing of metals and high-performance composites, you can probably remove additional weight so there’s some light in that tunnel.
But all in all it’s very hard to get out of the gravity well with your fuel in tow and survive the extremes of space. My belief is the first vehicle to pull it off will look like a Swiss cheese of voids and lattices from printing / honeycombs and be made almost entirely out of plastic and carbon fiber.
The 1990s were a lost decade for reusable space flight because instead of chasing incremental improvements to the Space Shuttle (an orbiter with reusable tiles that could be turned around in days, not months) or something like the Falcoln 9 or the fly-back version of Saturn V that O'Neill's students drew in 1979, it was all about SSTO.
SSTO is just marginally possible, if it is possible you need exotic materials and engines and you're never going to get a good payload fraction and adding wings, horizontal takeoff, horizontal landing and such just makes it worse. The one good thing about it is that you get closer to "aircraft-like operations" because in principle you can inspect it, refill it, and relaunch it -- whereas something like the STS or Falcoln 9 or Starship will require stacking up multiple parts for each launch.
My guess is aerospikes are making a comeback though because of interest in hypersonic weapons system. I could also see them being useful for the second stage of something like Starship which mostly operates at high altitudes but has to land at low altitudes. There are a lot of other technical problems, like the thermal management system, which really have to be solved before worrying about that optimization.
Currently the Starship upper stage simply has two different sets of bell nozzles: Three engines with nozzles for atmospheric pressure, and three for vacuum. I wonder how inefficient this really is compared to having just aerospike nozzles.
That's the same as the genesis of the question I asked above. SSTOs are a concept, but given their complete lack of market share, I assume as a non-aerospace engineer that there are valid reasons smart people have not been able to design a competitive one yet.
Similarly, I assume there are valid reasons SpaceX has chosen not to use aerospike Raptors, especially given their well-earned reputation for innovating things everyone else swore couldn't be done. If even they haven't been able to make it work, that's a strong data point as to the state of the art.
I'd argue that the brilliance of SpaceX is the opposite. They stick to technology and markets that are proven and use technically conservative approaches. Falcon 9 is about relentless improvement in small ways, not bold new ideas -- unless you count not getting caught up in the politics and psychology of bold new ideas as a bold new idea.
Sure, they talk about Mars, and in-space refueling seems radical, but they've yet to succeed at doing anything radical... yet.
Rumor has it they were struggling with the payload fraction w/ the first generation of Starship and they switched to a second generation that struggles with blowing up. A big advantage of the two-stage architecture is that you can develop the two stages independently. Presumably they will eventually get Starship to orbit and bring it home, they will have plenty of time to improve it get the payload fraction up just as they did with F9.
Landing and re-using their Falcon first stages was pretty radical though.
I don't think that's true, at least it wasn't conceptually radical. People have noticed the cost of "throwing away" the lower stages for ages, and many approaches have been thought of how not to do that. Take the (partly) renewable SSRBs of the space shuttle program for example, which came down by parachute. Landing a rocket on its tail is also quite an old idea. NASA had several demonstrators demonstrating the concept in flight.
SpaceX took a lot of ideas which had been individually proven before, and then put in the work to perfect them and integrate them in a production ready spacecraft. That is important work and good engineering, but not radical. An aerospike had literally never been flown to orbit at that time (I think still not), so it would have been a way worse fit for the SpaceX method of developing the Falcon 9.
A reusable lower stage with powered landing also had never been flown to orbit at that time. And in contrast to aerospike engines, which had been tested before on the ground [1], you can't do ground testing with rocket stage landings.
I think SpaceX didn't try to develop aerospike nozzles because the advantages probably aren't that large compared to the mixed nozzle design they are currently using. They also reused the same ceramic heat shield material developed for the space shuttle instead of developing something new.
Compare that to the cancelled "VentureStar": It would have used both linear aerospike engines and a new metallic thermal protection system (TPS) instead of a ceramic one. I remember an interview where Musk answered the question of why they aren't doing aerospikes or metallic heat shields etc, that there are many ways to skin a cat. They are only doing one thing that they think will work, which is not necessarily the best possible solution, but potentially faster or cheaper to develop.
[1] https://youtube.com/watch?v=FcW9kUUTfxY
Radical in terms of economics, but also radical in its incrementalism.
Falcon 9 was a highly competitive rocket without reuse. If they didn't get reuse to work it would have been a successful project. Reuse of the first stage was a huge cost optimization that put it in a class by itself -- but they they did it radically reused risks.
Contrast that to the X-33 which would have required a large number of new technologies to all work to fly at all.
Fixed-cost pricing was also a radical innovation because it drove SpaceX to do everything it could to lower costs. It was known for a long time that reusing (only) the first stage was a good path to lower costs, the SpaceX business model rewarded them for doing it.
SpaceX is highly technically innovative but it's been so successful because technical innovation has been centered around cost reduction and practicality, not chasing high performance for the sake of high performance.
The SpaceX model might need change to get to Mars because of latency. You can launch a Starship to LEO, have it blow up, and launch another one in a few weeks. If a Starship fails to land on Mars, however, you have to wait another two and a half years to try again. Similarly, SpaceX runs everything by remote control from mission control which is great in LEO but to stick a landing on Mars you need something that flies autonomously.
> Similarly, SpaceX runs everything by remote control from mission control which is great in LEO but to stick a landing on Mars you need something that flies autonomously.
I don't believe the stage landings are remote controlled. I've seen several times where they lost contact with the craft but it landed safely.
It would also be a weird choice because radio connections are way too unreliable to be a single point of failure.
They likely use control algorithms like Model Predictive Control which can do a lot on its own.
As is landing rockets on the launch tower (or as SpaceX would say: catching them). And I might be wrong, but I believe they are the first to use a crane on the launch tower to stack the rocket. Usually you do that before you roll it out to the pad. They were also the first to fly a full-flow staged-combustion engine. Maybe that one was less radical because prototypes have been around for 60 years, but SpaceX were still the first to actually fly one
> SSTO is just marginally possible, if it is possible
Looking at the specs it would appear the first stage of a Falcon 9 plus a nosecone could get itself to orbit with no cargo. Barely.
There has been some progress on scramjet propulsion.
This. In my very uninformed opinion the only way we'll get useful SSTO is if we can get a meaningful amount of oxygen from the atmosphere rather than carrying it up in heavy tanks. The failure of Reaction Engines with their SABRE engine is disappointing on this front.
It sounds good at the one sentence level. When you need to write more about the topic, the problem is that oxygen makes up only about 20% of the air. So you have need to accelerate all of this N2 that gives you nothing in energy and the result is a much lower Isp (specific impulse is the thrust per massflow, and all of that N2 is not adding anything to your thrust and increasing your massflow). And you need to be able to pull in enough air to get enough oxygen to drive your engine, so you need very large structures to move all of this unnecessary nitrogen around.
It is possible that only needing one tank rather than two can make up for the dramatic loss of Isp we see from an air-breathing engine and the air-handling structure, but no one has yet managed to demonstrate that, and the general consensus runs against it. I recall reading that HOTOL (https://en.wikipedia.org/wiki/British_Aerospace_HOTOL) calculations were actually driven by an extremely light structure estimate rather than the airbreathing engine, to the point where if you plugged a rocket engine in they would actually get more payload to space as a SSTO, because those aggressively light structure estimates were doing all of the work.
SpaceX is very close to demonstrating an architecture that ameliorates almost all of the drawbacks of two stage to orbit architectures. The tyranny of the rocket equation ensures that while a SSTO carrying all of it's oxygen is possible, it's never going to be able to carry enough mass to be useful.
Therefore nobody is ever going to invest the tens of billions required to develop a rocket based SSTO.
If somebody develops an engine that makes air breathing most of the way to orbit feasible, this has a chance of competing a Starship style architecture.
For the reasons you espoused, this is highly unlikely. However "highly unlikely" is more likely than "never".
Jet engines have on the order of 10x the specific impulse of a chemical rocket.
Atmospheric density reduces exponentially with altitude, which implies that you would need to go exponentially faster to maintain mass flow into your engines and lift over your wings. The truth is that breathing air only gets you a third of the way to space, at best, so you have to have a rocket, and now you're battling that complexity. If your space plane doesn't breathe air, it probably is just better to punch your way out the way conventional rockets do.
Of course, the rocket equation is logarithmic, so reducing the amount of mass you loft gives you an exponential gain. This is true for all propulsion systems to an extent (different constants) but getting into space is the hardest propulsion problem we face. A space plane may or may not be better in this regard (it's been a while since I've looked into that kind of thing, so no opinion either way) but imo the inherent complexity is enough on its own to kill the idea.
Only because traditionally the airplane industry measures specific impulse on just fuel flow, completely ignoring the oxidizer and atmospheric nitrogen. If you calculate like for like, including the air, jet airplanes have significantly worse Isp than a rocket engine.
They get this fictitious specific impulse by scaling the effective exhaust velocity by a scaling parameter to account for the fact that the exhaust mass flow consists of extra mass (air), in addition to what is carried onboard (fuel). This specific impulse is still used for comparing jet engines based on efficiency. Another use for it is to calculate space mission requirements in launchers utilizing air-breathing engines in their first stage (as your parent commenter points out). Though such vehicles don't exist yet, there are concepts being pursued. Some of them use a scramjet and others are more elaborate like the (cancelled?) SABRE engine. So those Isps are not completely meaningless.
The general idea is that you can get much better results in terms of deltav if you can find at least part of the reaction mass from elsewhere without carrying it onboard. Even inert nitrogen is useful as a reaction mass. Another way to get a good result is to use separate sources of reaction mass and energy. Then use that energy to accelerate the reaction mass as much as possible, so that you get a decent deltav by the time you exhaust the reaction mass. This is what ion and plasma thrusters do.
There are air-breathing rockets, some of the oldest were ultimately canceled soviet projects for road-mobile ICBMs (canceled for reasons AFAIK unrelated to air breathing concept), and the recent Meteor air-to-air missile
Aren't rockets more powerful (as in energy/time) than rocket engines in that they are getting compressed/liquified oxygen out of a tank as opposed to taking the comparably tiny amount that passes into the intake of an engine?
There are two performance parameters for a rocket/jet engine. The first is thrust and the second is specific impulse. You are thinking about thrust. The others in the tread are talking about specific impulse. Thrust is important for some stages, especially the early stage booster engines (as opposed to later stage sustainer engines). As a simple example, any space rocket will need a first stage with an enormous thrust so that it can lift itself, the subsequent stages and the payload off the launch pad. Additionally, the rocket has to finish its initial vertical climb as fast as possible. Otherwise the propellants will be wasted in just lifting off (this is called gravity loss). That will also require a high initial thrust.
However, the requirement of the high thrust disappears once you finish the vertical climb. There's no danger of falling back to ground once you reach orbit. What you need at this stage instead, is to add velocity (deltav) to the craft to change its orbit/trajectory. This can be done even at very low thrust, because you have all the time you need. The limiting factor now is that you have only a finite amount of propellant onboard. You want to add as much deltav as possible before it runs out. A high thrust doesn't help because the engine will simply consume the propellant faster and exhaust it before you get the required deltav. This is where specific impulse comes into play. The maximum deltav you can get is proportional to the specific impulse of the engine (see rocket equation for details). As you can imagine, high specific impulse is critical for space missions requiring high deltav, like the New Horizons spacecraft that imaged Pluto or the Parker solar probe (interestingly, getting to the sun is harder than escaping the solar system). Rockets/jets with low thrust and high specific impulse are called sustainers.
The general trend seen is that specific impulse drops off as thrust increases. For example, the space shuttle solid booster has Tmax = 15 MN, Isp = 268s, and space shuttle orbiter cryogenic engine RS25 has Tmax = 2.28 MN, Isp = 452s. Meanwhile, the NEXT xenon ion thruster used in the DART mission has Tmax = 236 mN and Isp = 4200s. Note that the thrust has changed from Mega newtons to milli newtons. You would hardly recognize it if the ion engine thrusted against your body.
doesn't scale well. The amount of air entering is proportional to square - cross-section - while the mass of rocket is cubic. While scramjet/turbojet/air-augmentation, say as a separate detachable stage, can be pretty efficient for smaller rocket, anything making significant improvement for say Starship would looks like a fat monster cross-section-wise with tremendous hardware cost and weight loosing outright to the straight option of adding additional tanks and rocket engines.
Wrt. aerospike engine - sounds nice, yet hardware wise it is heavier than the classic engine, and just look at that large number of pieces - just all those small mini-engines - it is made of and compare to Raptor 3. And for the optimal expansion - i'm waiting somebody will add a dynamically adjusting telescopic kind of end section to the classic bell nozzle.
A napkin to illustrate. Lets say you add a Raptor and 80 tons of fuel plus oxygen for it. That will give you 100 seconds of excess impulse of at least 160 tons (240 ton of thrust minus 80 tons) at the beginning to 240 tons at the end, so roughly 100 seconds of 200 tons. To get 200 tons thrust you'd need 20 fighter turbojet engines capable of at least Mach 3 - that is cost, complexity and weight dwarfing that one Raptor engine.
For scramjet, assuming we got a decent one, napkin is about the same. The best, my favorite, is air-augmented - scram-compress the air and channel it on the outside of the hot bell nozzles of the already working rocket engines - unfortunately the scaling mentioned above comes into play for meaningfully sized rockets though it has worked great for small ones.
There's been progress on scramjets for cruise missions. For acceleration missions, like launchers, scramjets make no sense at all.
That really depends on how fast you can cruise. High speed scramjets above mach 15 will make space missions possible. The craft will be at the sufficient height and just enough speed, so that a rocket engine won't have to add too much deltav. Scramjets are still in their infancy. There are already developments on for variable-geometry multi-mode ramjets for this purpose.
PS: I have seen early-stage (but successfully tested) scramjets being developed for this purpose.
I don't think any of the considerations you mention there change my conclusion.
We have to ask: what exactly is a scramjet vehicle delivering? It's enabling the use of air instead of liquid oxygen. But how valuable is this? LOX is the second cheapest industrial liquid after water. The fuel part of a rocket propellant combination typically dominates the propellant cost. If a scramjet launcher uses more fuel (especially hydrogen) than a rocket vehicle would, it will end up increasing propellant cost per unit payload to orbit. It will also likely increase propellant volume per unit payload to orbit, especially if LH2 is used (LH2 being just 5% of the density of LOX).
All scramjet launchers need a rocket to reach stable orbit (since a scramjet cannot produce thrust at apogee to circularize above the atmosphere. So one can ask, what the tradeoff between the delta-V this rocket provides and that of the scramjets? From what I've heard, all such trade studies end up optimizing to 100% rocket and 0% scramjet.
As the other commenter already pointed out, it's the mass that's the limiting factor here, not the cost. The key idea here is that rockets and jets need two things - a reaction mass and energy. Scramjets and other air breathing engines don't just take oxygen from the atmosphere. They derive much of the reaction mass also from it. Even the inert nitrogen absorbs heat from combustion and acts as reaction mass. The primary purpose of the fuel onboard is to provide the energy. It's contribution as reaction mass is only secondary (note that this is for air breathing engines). This is very evident in the case of turbofan engines, where much of the thrust is contributed by the uncombusted air from the fan.
A scramjet stage will be very light compared to an equivalent rocket stage, since it carries only the energy source (fuel) and not the full reaction mass. If this scramjet stage is able to impart a velocity close to the orbital velocity by the time it reaches the upper atmosphere, the subsequent rocket stage will have much less work to do to get it into orbit. And that translates to much less propellants (including oxidizer) and much less mass in the upper stage. It's not necessary to collect oxygen from the atmosphere to see an advantage.
Obviously, the raising of the perigee at apogee is going to need this rocket engine again. There are no launcher concepts that depend purely on scramjets.
What is the point of saving mass here? LOX is cheap, so it isn't the cost of the LOX. Does saving LOX make the vehicle cheaper? No... it increases the quantity of fuel needed, which (particularly if it's LH2) makes the empty vehicle much larger and more massive. This is doubly bad, since every last gram of that empty mass is taken to orbit, unlike the mass of LOX.
Minimizing fueled mass of the vehicle is a stupid thing to do. It's optimizing the wrong metric.
Scramjets also suffer from bad thrust/mass and thrust/$ ratios compared to rocket engines.
Overall scramjet launch vehicles are an example of pyrrhic engineering: even if one could make such a vehicle "work", no one would want it.
The point here is how it alleviates the problem of the tyranny of the rocket equation. But you seem to have made up your mind.
I'm looking at the issue deeply, not using the superficial reasoning you seem to be engaging in. This reminds me of "mark logic" where if you are promised pie in the sky you commit yourself to believing it even if it's a scam.
Scramjet launchers (and air breathing launchers in general) have been looked at extensively, for decades. They don't make any sense. The math just doesn't work. Here's an insightful post from the Old Usenet that illustrates the difficulty:
https://web.archive.org/web/20220131115058/http://www.island...
When reading that, remember that LOX is 20x the density of LH2 (or 17x the density of "slush hydrogen").
Another useful article by Henry Spencer on the subject:
"Rockets, not air-breathing planes, will be tomorrow's spaceships"
(New Scientist, 2009)
https://web.archive.org/web/20101017031109/https://www.newsc...
LOX halfway to orbit is significantly more expensive than the same LOX delivered to the launchpad.
Its not the cost, its the mass you're trying to reduce. So far, the engineering challenges have made it unfeasible, but its not a surprise that people look at the hundred tons of LOX on a rocket and imagine exchanging it for payload (or aircraft style re-usability).
A gram of oxygen that you carried to orbit is more valuable than a gram of oxygen you collect at that location: oxygen that you carried is moving at the same velocity as you.
Why make an SSTO when you can make a TSTO? First stage recovery is a solved problem and will always greatly relax the engineering problems over making a SSTO.
Because of course SSTO has benefits over TSTO, simplicity of operation being one example.
The argument for SSTOs was that staging was too scary. But experience since then shows this argument was bogus. Staging can be made highly reliable.
Even a small amount of delta V provided by a first stage makes the job of the "almost SSTO" second stage much easier. And a low delta V first stage can be rugged, with high high safety factors, and is easy to recover at the launch site.
Put another way: if you have an SSTO, its payload increases dramatically if you stack it on a very low performance recoverable first stage.
I don't see any way SSTOs are going to be preferable to TSTOs, especially if the SSTO has to use hydrogen to get off the ground.
Everybody may have their own criteria of preferences, so the same solution with fail for one and succeed for another.
Sure, you can cook up whatever BS criteria you want, but when it hits the market SSTO doesn't make sense.
You're so sure you proved yourself right, huh?
I've been exposed to this argument for more than 40 years, so, yeah, I'm confident my opinion is a reality based one.
Against the cost of a lower payload fraction. Say you get 4% with TSTO, 2% with SSTO. And if you run into trouble with the design, the SSTO payload moves towards 0%. Your operational savings need to compensate for a doubling of cost per unit mass payload, possibly much more.
And even though full reusability is much easier with two stages, and has not been achieved yet.
video on the rectangular one https://www.youtube.com/watch?v=FcW9kUUTfxY
I'm not sure if this one counts but recent https://www.youtube.com/watch?v=UShD03eG9IU
Is that an AI generated image of the Venture Star? It's missing portside wings..
Gotta be, the skunk logo is an approximation of the real one.
The article has a link (citation 17) to a site selling toy models of that vehicle as an image source, I can find one (fourth image in the gallery) there where it sorta looks like the wing is missing because the wing is black against a black background but it's not the same image shown in the paper:
https://fantastic-plastic.com/lockheed-martin-x-33-venturest...
The name "Venturestar" is properly rendered in that image but "NASA" and "Lockheed Martin" are thoroughly mangled the way I'd expect text to be mangled in an AI image. The image from the toy site could have been used as as reference image to create the image in the paper one way or another.
Yes, if you look close, the paper is replete with error-filled generative reproductions of existing illustrations in the citations; including Fig. 6 (MC Escher struts), Fig. 7 (sprouting greeble tubes), and Fig. 8 (actuators replaced by tubes connected to mystery manifolds).
Even Fig. 2 shows the spike geometry magically changing, which is not addressed in the text and seems like an error carried over from the original illustration in the cited source.
Casts serious doubt on the credibility of the rest of the work.
I thought this would be about the key-value store, Aerospike.
Pretty neat, my dad worked on that X-33 program at Lockheed.
cool contribution from a top Colombian university researchers