Spike engines aren't that great not because of cost, but because their supposed benefit is largely eaten by added weight and cooling complexity in most designs, so the entire thing isn't worth it, usually. Meanwhile, things like nozzle extensions can provide 80% of their benefits for 20% effort. The article is light on information on whether they solved typical issues, and what exactly they traded for the efficiency.
A favorite of mine is Thrust Augmented Nozzle or TAN, where, at low altitudes, extra propellant is injected into the nozzle (not the chamber) to prevent flow separation and to increase thrust, albeit at lower specific impulse. Think of it as a bit of an afterburner for a rocket engine.
>A favorite of mine is Thrust Augmented Nozzle or TAN, where, at low altitudes, extra propellant is injected into the nozzle (not the chamber)
and that gets close to my favorite - air-augmenting regular rocket engine. Very simple and bumps up your thrust and ISP up to 2-3x http://www.astronautix.com/g/gnom.html
A similar, but much more useful thing for the future seems to be LANTR, a.k.a. the only practical nuclear rocket engine we know of. At the very least it doesn't waste all oxygen from water mined on other bodies of our solar system like traditional pure-hydrogen NTRs do. And its average impulse density of propellant is way higher, of course.
This is solving a different issue. Aerospike engines deal with the fact that rocket engines need to be different in an atmosphere vs in vacuum. Nuclear thermal rockets (NTRs) are not really envisioned for being used in the atmosphere, the risk of a blowup is too high. I consider them a wonderful technology which could really open up new possibilities, but we first need to nail down the challenge of affordably getting to orbit.
Except LANTR solves a very similar problem, which is the requirement for different performance parameters at the beginning and at the end of a rocket flight (high thrust and propellant density in first stage, low thrust and high Isp in the second stage) in a single stage vehicle. And I mentioned it because its way of operation is very similar to the thing describe above, i.e., injecting something extra into the nozzle, so it reminded me of it. Not because I consider it comparable to using aerospikes on Earth (I'm not convinced that's useful at all) but because this is one of the few such variable Isp options that appears to provide any useful benefits at least somewhere (even if it's not on Earth).
The problem with NTRs in the atmosphere isn't the risk of blowup, it's that their thrust/weight ratios are just terrible. And in the atmosphere, it's early in the launch, so Isp is not important.
I don't think it necessary true that NTRs have terrible trust/weight. Depending on what your 'fuel'. You potentially lose ISP of course, but your Thrust would go up.
It also depends on the engine configuration, you could do much better in terms of weight and density.
Thrust of a rocket is a function of nozzle throat and exit areas and thrust chamber pressure, and is (ideally) independent of the density of the propellant.
Engines used on a first stage typically have high thrust, but that's because they're designed with large throats, not because of the propellant they use.
A dense propellant does allow the pumps to be smaller. But I don't think that's going to help much in a NTR, since the reactor itself is heavy. The problem with an NTR is the difficulty of getting high mass flow past very thin fuel elements (the thinness necessary for high heat transfer) without destruction of the fuel elements by that flow. The thin fuel elements also ensure some fission products will escape and be carried off in the propellant.
I had to look that up. This was the nicest explanation I found, but if you recommend a different one, please post. I also found stuff on Kerbal, but I'm not clear how real that is.
Umm, could you help educate us. How do you know/state this so confidently?
The way I see it, there are roughly infinite previously non-constructible parts that by definition we as a society couldn't test. Now that we can construct these new parts, shouldn't we test them and then use them if they are better/cheaper?
There have been rocket engines built by stacking carefully made thin metal slices (with holes cut for coolant channels) and then using diffusion bonding to join them into a single monolith. I could see 3D printing being used for this instead, and perhaps more cheaply.
I see, you just don’t believe that additive manufacturing is a truly new paradigm and that there are parts that can be constructed now that were previously just non-constructible.
Why? Even if a additive manufacturing is slower than traditional manufacturing, 3d printers can be mass-produced (unlike assembly lines and skilled engineers).
I assume that all of the support gear (turbopumps, etc.) are not 3D printed. I'd be super impressed if the fuel and oxidizer injectors were printed.
Anyway, there's a bunch of stuff they're not including in this statement; it's like "we can print your car's engine block" but glosses over the additional whirly-sparky stuff that turns a lump of metal -- even if it is in the right shape -- into a functional engine.
Indeed, people outside of USA (and France, Germany, and a few others that dub everything) have lived with this their entire lives. Even today my wife and I watch with subtitles as it's not always easy to hear all of the dialogue over the other noises in movies/shows.
From what little i've seen of this guy; he appears to have no actual technical understand or depth to anything he reports on and basically just echos pop science and press releases. Scott Manley, Curious Mark and Periscope archival footage comes to mind for high quality content to learn about rocketry and related technology.
He actually interviews people who know what they are talking about and does the research. He has good contacts in the industry, better then most people. The video actually contains lots of information you can't find anywhere else. He spends a lot of time validating his videos himself with industry expert and then has a huge patreon community that goes threw everything and tries to find errors.
He is an educator he doesn't have to be able to design his own engine to make good educational videos.
And Scott Manley is also not an engineer and Scott Manley actually used Tims graphics in his own videos.
Right now he is working on a 1h 30min video on Russian engines that is more detailed then anything out-there and he spend 2 years researching it.
I think that’s over-stating it. His content is deliberately designed to be accessible to everyday people, so it necessarily needs to be limited in depth and detail. Having heard him on podcasts and other settings, his knowledge seems plenty in depth to me (an amateur enthusiast who wants lots of detail) for that role. He’s not an engineer, and that’s fine.
In Tim's interview with Elon Musk, Elon mentioned[1] low combustion efficiency as one of the major disadvantages of aerospike engines compared to the more traditional bell nozzles used on Raptor.
I wonder if Pangea has done anything in an attempt to solve that problem.
>> Elon mentioned low combustion efficiency as one of the major disadvantages of aerospike engines
If you haven't seen the SpaceX presentation on their combustion CFD software, it's a must see. They have state of the art simulation capability in this area - better than any commercial offerings.
I don’t understand your comment. What does Raptor have to do with CFD method developments? What is the frontier? What does it mean to run out of frontiers? Shouldn’t there be more resources rather than fewer?
His target audience is not the highly technical, but the public at large - including children. He's a space ambassador in that he wants to get the current and upcoming generation excited and learning about space.
With that said, he has been able to get in-situ interviews with rocket company CEOs including his amazing 2 hour almost single take video tour with Elon Musk [0] where I was able see Elon comfortable in speaking much more deeply, candidly, and enjoyably than with any interviewer/reporter prior. Tim Dodd is no professional engineer but he really knows his stuff. His personality helps bridge the gap between the highly technical and the layman.
Well, after spending eight hours working demanding jobs, be it physical or mentally, everyday people usually like to spend their free time with something a little more relaxing, and for most people that isn't reading documentation or reading highky technical engineering papers but rather watching a highly enthusiastic guy try and share his joy of rocket science.
That’s a good point. I’m not an engineer yet I prefer there be more description and less personality shining through; but you make a good point about target audience. And as you say he prepares for the material and does well to understand the experts.
He actually researches history, has contacts with many important people in areospace and actually managed to all the CEO from all the companies that had reached Orbit about the topic, Musk, Bruno, Beck. Since that video he also talked to it with the CEO of Firefly (that used to develop a spike and was propulsion engineer at SpaceX).
If you think you are to good to hear from those people then maybe you should get of your high horse.
Just because its not designed for only engineers doesn't mean it shouldn't be recommended. Find me a better video or article about areospikes.
'the line of messianic descent was defined by the French word Sangréal.... In English translation, the definition, Sangréal, became "San Gréal", as in "San" Francisco. When written more fully it was written "Saint Grail", "Saint", of course, relating to "Holy"; and by a natural linguistic process came the more romantically familiar name, "Holy Grail".'
That's just bad etymology. The word/expression Saint Graal in Old French. Saint was translated into English directly as Holy, while Graal was adopted as is morphing into grail. It seems the etymology for graal is likely ultimately Ancient Greek krater, a type of wine bowl.
The 'sang real' (royal bloodline) nonsense was a much later invention (15th century).
Honorable mention: an aerospike engine has been flown by Arca space.
Well, kind off.. it's a spike, but they're using water steam as propellent, I'm guessing to "solve" the cooling problem. So not sure if it counts (and how successful their rocket will be).
Ah yes, ARCA Space. That company founded by a guy who faced 13 counts of fraud, 5 counts of embezzlement, and one count of forgery. I'm not surprised that company has so many "buzzword" technologies that it's supposedly developing. What better way to scam... err I mean... provide opportunities for bold investors.
LH2/LOX engines (and most others) typically are operated a bit fuel rich. At stoichiometric fuel/oxidizer ratio, the flame is so hot the mixture has significant dissociation, so it loses some of the chemical energy. It's advantageous to win back some of that by adding extra hydrogen, reducing the temperature slightly but also significantly reducing the average molecular weight of the gases.
No it hasn't. Please stop ever linking to Arca space.
That is not a company, its a fraud. They defraud investors. They are not a real company.
I recommend ignoring them and warning people against them.
The most recent company who was seriously developing a areospike was Firefly. That was for a pressure fed rocket. They have since dropped that, and in the most recent interview between Everyday Astronaut and the CEO of Firefly he addresses why they did that.
I don't see that as a reasonable constraint - nuclear propulsion does not need combustion, thermo electric does not need combustion, and there's also things like cold gas thrusters, laser ablation etc. etc. etc.
But it's not combustion. I guess you meant something else. Now you label they approach; do you think laying boundary line this way looks obviously correct? Say, nuclear rocket engines heat up gases - also without chemical reaction; resistor jets also have reasonable history in rocketry - why ARCA's approach is wrong?
I understand atmospheric pressure has an impact on the exhaust shape, but why aren't the nozzles that feed onto the aerospike affected just like a traditional engine?
I.e. in their illustration on the page, the two nozzles next to the spike itself?
They are affected by air pressure, but it doesn’t actually hurt, it actually helps! because of the way an aerospike engine works it actually relies on that pressure change behaviour on the “outside” edge. The expansion of the exhaust on the atmospheric side of the aerospike is actually part of how it obtains its efficiency…. That expansion pushes the expanding exhaust plume back towards the nozzle wall in proportion to the normal expansion with outside pressure and that drives the physics that make aerospike nozzles work at all altitudes. It’s a clever feedback loop if you want a computer analogy, but it’s really just a clever arrangement that maximises system efficiency in that lovely “continuous” way that mechanical/physical systems work.
It seems that the nozzles feeding onto the aerospike are sufficiently small to operate safely at sea level air pressure. That kind of nozzle can operate safely at lower air pressure as well, but is inefficient. The aerospike gives you back your efficiency at lower atmospheric pressures, so you end up with something that's both safe and efficient at all altitudes, relative to a traditional bell nozzle.
At least that's my understanding - I just like to learn about rockets.
You may want to check out the video / text from everyday astronaut pointed to elsewhere in this post, which goes through a much deeper research and explanation on the topic.
The TLDR as I remember it, is physical nozzles on the aerospike don't work the same way as they do on a traditional rocket. It's more like like just an exit from the combustion chamber.
The appeal of an aerospike is by having lots of exhaust port aligned around a spike, is it creates a sort of virtual nozzle that is the right size for the atmospheric pressure. And as such acts sort of like a nozzle that changes shape and size as the rocket goes through the thinner and thinner atmosphere.
This in theory gains efficiency over a rocket that has a set shape for it's nozzle, and isn't always operating at optimal efficiency due to the size of the nozzle.
* Not an expert on these things, just going off of memory
Beyond all the theoretical implementation complexities of space elevators, we don't even know of a single material with the required tensile strength. Graphene was thought to be strong enough, but recent research found practical graphene had defects that significantly reduces its strength below that required for a space elevator.
Space elevators aren't scrapped, just on the shelf until we even have a material that let's us think they might be feasible. That said, Lunar space elevators could be done with lower-strength materials, like Kevlar, and there are some people working on them.
For Earth orbit, I think the centrifuge concept (e.g. SpinLaunch) has some promise. Still some huge implementation hurdles, but could be a huge step forward to put all of the energy for the first stage on the ground.
Space elevators don't work on the moon. While gravity is weaker, the moon's rotation period is also much lower. To build a space elevator, its center of mass must be in a stationary orbit over the body. A geostationary orbit around the moon (selenostationary) would be about 88,000 km. The moon's Hill sphere, the region where its gravity dominates and thus things can stay in stable orbit around it, is only about 58,000 km. Basically a lunar space elevator would be so tall that Earth's gravity would yoink it off the moon.
You could potentially build a skyhook on the moon, which a space elevator is merely a special case of, but you lose a lot of the advantages of a space elevator - namely you still need something to blast off the surface to reach the skyhook, and you have to carefully time it because everything's moving at extremely high speed. Saves a lot of fuel though.
The most likely place for a space elevator is clearly Mars. At the Mars Society conference there was a talk where somebody went deep into space elevators and Mars is the most likely place to make one anytime soon.
Its far easier then on earth.
One the moon you might just use a mass driver instead, at least for bulk materials. And if not that a reusable rocket works just fine.
What if you position the counterweight of the space elevator at the L1 lagrange point of Earth/Moon? Since the moon is tidally locked, it should always stay about the same spot relative to the lunar surface.
But why not just build a mass driver, then? It's a somewhat large horizontal structure, compared to a space elevator which is an absolutely humongous vertical structure. If you can reach 2 km/s on magnetic rails, you don't need anything else to launch from the Moon -- or even land on it.
Because unless you reach escape velocity, you still need to adjust your trajectory to insert into orbit, or else you're just going to crash back on the surface.
If you use a rocket that's fine you just point in a different direction and burn, but if you're trying to launch, say, a few trillion tons of stuff, you need something a tad more efficient
It's not like you need no maneuvering when you use a space elevator. At some point you have to detach from it. You'll always need course corrections afterwards.
Spinlaunch is potentially okay for cargo, assuming they can overcome limitations imposed by atmospheric drag, and scale up sufficiently to make it practical; but the G forces involved make it impossible to use for human spaceflight.
Is something like spinlaunch (centrifugal force) even better than a hydrogen gun (linear acceleration) from the perspective of g forces? For example Quicklaunch (6 km/s speed, 1100 m barrel) would require ~1700g of acceleration. Spinlaunch appears to be significantly more demanding while providing even smaller initial velocity. Even if it's just cargo applications, surely there's still a difference between 1000g and 10000g.
The main advantages of spin launch's approach over a more conventional mass driver are that it takes up less space (Quicklaunch's proposed driver is over 1km long), and that it requires less peak power to get up to speed (Spinlaunch accelerates over a period of one and half hours).
One could argue that Quicklaunch doesn't practically require much peak power either since you have means to store hydrogen and oxygen in tanks over time if you generate them in trickles. The fact that firing it generates high peak power is no more relevant for practicality than the fact that firing a handgun generates high peak power. It's just combustion.
As for size, honestly, that doesn't seem to be much of a problem. You still have exclusion zones, and I'd argue that the exclusion zone around a spinning device like Spinlaunch proposes would have to be fairly large in all directions. Quicklaunch doesn't have a failure mode where the payload goes sideways at its full speed. Even if Quicklaunch were on land (which it wasn't planned to be), it would still probably require a smaller exclusion area.
> The fact that firing it generates high peak power is no more relevant for practicality than the fact that firing a handgun generates high peak power. It's just combustion.
This is not a trivial engineering problem. And in the case of a light gas gun, not only do you need to dump all this energy into a driver to compress the light gas extremely quickly, you also then must bring that driver to a halt extremely quickly without destroying everything.
I'm not saying that a space gun is simple. But I'm not convinced that it's that much more complicated than a hypersonic centrifuge that provides smaller benefits to begin with.
BTW did Quicklaunch even have a driver/piston? I'm not quite sure that it had.
Yeah, quicklaunch actually uses methane combustion to drive the piston, the hydrogen is just the working gas. The whole advantage of light gas guns is that since you don't need your working fluid to do the combustion, you can use gasses with much higher speeds of sound which can thus produce much higher velocities, such as superheated pure hydrogen.
I am in total agreement though that spinlaunch is a much worse system though.
I checked on the Quicklaunch design (https://vimeo.com/29822477) and the only solid moving mass in the gun is the projectile. So apparently you don't need to "bring [...] driver/piston to a halt extremely quickly" since there is none. They did propose to recover most of the gas by closing the muzzle just after the projectile exit, though. That might be complicated.
There is a piston. At 2:07 they show a close up of the first stage which they are referring to as the "pump tube", that larger diameter section, where the piston is driven up to speed, pumping the hydrogen into the second stage barrel at high pressure.
There is no piston. Where do you see one? Did you even watch the video? The whole launch sequence is shown there. There's a pressurized hot hydrogen tank that empties into the barrel. There's no "piston [...] driven up to speed, pumping the hydrogen". The hydrogen flows into the barrel on its own because it has high static pressure at the moment of firing. The presenter literally explains it in detail.
SpinLaunch is easily the dumbest thing I've seen this year, and I got to become aware of NFTs this year. It's going to end up in the long list of grifts designed to bleed investors.
Curious about your reasoning. I immediately wondered what type of cargo would be suitable for this sort of launch system. I came up with a fairly short list. Fuel/water/perhaps food for interplanetary spacecraft destined for Luna or Mars. I imagine one could design some types of cubesats to withstand the tremendous acceleration forces. That was about it.
Is your skepticism based on cargo suitability or other factors.
I don't think that even Rh(null) blood, unicorn tears or HP color ink liquids are commercially viable to launch at prices around 10-50k$ per liter. And that if that whole assembly will work at all at prices they are advertising today (1/20 of comparable commercial rockets, but it will have very small payload).
PS: I don't have numbers, but I'm guessing that F9 hitchhike Transporter missions cost less per kilogram, compared to the imaginary numbers for non-existing full scale Spinlauch.
Who the fuck cares about cargo before trying to figure out if any second stage rocket that isn't a rock can reasonably survive the extended application of high G-forces
They have some data that indicates most off the shelf electronics survive or need very minor modifications to withstand the G forces. That's better than no data at all but they do have a ways to go.
Some classification considers 3 types of cargo - bulk materials (metals, fuel); complex electronics; humans. The majority of weight is in the 1st type. Unless first two types have complex mechanical structure, where loads can't be distributed into supports - which is an important issue - the first two groups are candidates.
IT make no sense to build a dedicated launch system and non-reusable rocket to deliver a few 100kg max into orbit.
A Starship launches 150 tons reusable. That far cheaper and less work then building 500 Upper-Stages and shoot 500 of those rockets with a SpinLaunch system.
There is really no comparison in cost.
Maybe they could make the upper stage reusable but that would reduce the payload to and even more pathetic amount.
This system simply can not compete against reusable rockets.
But this launch system still needs the projectile itself to be a fairly complex rocket capable of delivering several km/s of delta V and of steering the payload onto a desired orbit. That still implies electronics and mechanical systems onboard even if you're carrying only bulk cargo.
Electronic chips themselves are rather good to withstand acceleration - can be packaged well. Mechanical systems for simple rockets could also be, well, simple - e.g. first orbital launcher of France used pressure-fed engine.
First, you're making a giant spinning thing that contains a ridiculous amount of energy. If the release of the projectile is off by a millisecond, instead of flying through the outlet it's instead flying into the wall of the launcher where it will deliver all of its kinetic energy in the form of an explosion with the energy of about a half ton of TNT, which sounds bad but really isn't compared to the arm its exploding next to that would release the energy of a small nuke if it gets damaged. This isn't a failure mode that can be monitored and avoided; eventually you're going to have a component fail or a software glitch and before the system can even register that something is wrong the launcher will be a crater.
You're launching bulk material into orbit where no one cares if the occasional launch fails so long as it's cheap, and you're wasting that on a launch system where the most minor failure results in not just the loss of the launch vehicle but the entire infrastructure for launching.
But even if everything works exactly as intended there are still issues. Your giant centrifuge spins up in a vacuum because at those speeds air resistance would be extremely damaging. Unfortunately, after you release the payload, it breaches the seal on the outlet and now you have a giant inrush of air into your vacuum chamber. Unfortunately, the giant arm is still spinning at 8000 kph. The surface of the arm is going to ablate as it moves through sea level air at hypersonic speeds, and its going to generate massive shock waves which are going to reverberate in the chamber. All those precision components for releasing your payload with extreme precision are going to be exposed to these hellish conditions. You're going to need extensive repairs or replacements after every launch.
You're doing all this and you still need a launch vehicle with its own rocket engine and propellant, flight control surfaces and surface protection for its own hypersonic journey through the lower atmosphere. Everything needs to survive ridiculous g forces. All this to deliver a few kilograms of low value cargo?
These problems don't go away as you refine the technology, they are fundamental. You will always need precision release mechanisms to avoid catastrophic failure, you will always be exposed to hypersonic conditions, you will always experience ridiculous g forces, you will always need rockets for orbital insertion, you will always be restricted to low value cargo.
It's an interesting engineering problem; they might learn some cool lessons along the way, maybe some valuable patents will come out of it, but there is no hope for developing a practical space launch method competitive with existing methods.
> that would release the energy of a small nuke if it gets damaged
All we need to do now is containerise it for easy transportation across borders! Err..
My first thought on seeing the prototype was whether they've taken a map and drawn an arc in line with the rotation to see the areas that might be impacted when this thing RUDs. Perhaps they'll put it on a turntable?
Nonetheless I'd suggest it's probably safest to build this thing in a concrete pit such that failure results in a big hole, and not hot and spicy hypersonic plasma flung across the continental USA.
Skepticism? I'm not "skeptical" about SpinLaunch. I'm incredulous about SpinLaunch. Skeptical implies that I think there might be a chance, if certain criticisms are addressed. No, I am certain there is no chance SpinLaunch will end up being a thing.
First of all, putting aside the several red-flags for why this looks like a scam, and assuming it's actually do-able, it's still not going to happen. The global space industry is built on rockets. At best, stuff like this--even if it does work--gets a "cute project, kid" pat on the back and then ignored. When was the last time any fundamental redesign of an existing technology take over an industry? It just doesn't happen.
And I think everyone involved probably knows that. You're not going to compete against Google as a scrappy startup trying to make a search engine. You're not going to compete against SpaceX as a scrappy startup trying to make alternative launch systems. Somebody, somewhere, has to understand that. So they have to be in on the scam.
Now, as for the actual idea, there are several problems. First and foremost is: energy is energy. If this thing fails, it blows up just as bad as a chemical rocket on the platform.
At least chemical rockets are based on decades-old materials science. This spinning arm malarky expects us to believe that they can support 10,000x their payload on the arm and be able to release it on a hair trigger? I suppose next they're going to tell us the arm is made of carbon nanotubes or some other unobtanium.
So they want to spin the object in a vacuum. How are they going to seal the spin chamber in such a way that they can generate a significant vacuum while also allowing the payload to escape? They show a paper or some other thin membrane door over the escape hatch that the payload punches through. OK, that means that door needs to be able to support 13.75 pounds per square inch of atmosphere. A 2m x 2m door needs to be able to hold up over 40 tons of atmosphere. The payload needs to punch through a door that is holding up over 40 tons of atmosphere. That payload needs to punch into a 40 ton column of atmosphere. At ~5,000 mph?!
So what I expect to happen is that the payload hits the column of air, creates a mach shockwave that destroys all of the windows in a 5 mile radius, while the rush of air into the chamber and clapping back around the tunnel of vaccum the payload creates blinding spike of plasma (not unlike lightning), that ends up destroying the launch chamber.
Oh, yes, there are "challenges" to "figure out". From their launch command center that was clearly designed for aesthetics more than functionality. Lots of problem solving gonna happen there.
You may or may not have a point, but your entire "argument" is the "argument from incredulity" fallacy and not any actual reasoning.
For example: arguing about "paper" strength to resist air pressure vs vacuum and throwing around numbers like "40 tons" isn't based on any actual materials science. The burst pressure strength for pretty standard office paper is 250 - 300 kPa. Atmospheric pressure at sea level is 101 kPa. So right off the bat, you've multiplied a bunch of numbers together and come up with the wrong answer: the paper I buy from the stationary store is 2-3 times strong enough to resist atmospheric pressure against a vacuum.
Now of course, at a suitably large dimensionality, we have to worry about fiber strength loading etc. but this has so many solutions it's absurd - i.e. 2 sheets of paper with fibers perpendicular for strength, a plastic backer for air permeability, and then weaving strengthening fibers into a grid - remembering that the burst strength is quite different to the resistance to piercing forces (i.e. kevlar will stop a bullet but not a knife).
A similar but somewhat more physicaly sound idea is the Slingatron: a launch track several kilometers long in a spiral shape, on stilts that rotate in synchronization. The g-forces are much more reasonable because of the increasing track diameter.
At about 10,000Gs of acceleration (as per their claims¹ ~2200m/s in a 100m vacuum chamber, e.g. a=v²/r ≈ 97,000m/s² or roughly 10,000Gs) any human passenger would be liquified and form a nice smooth film on the inner surfaces of the vehicle.
Space elevators are still pretty far out there. And if we ever do end up building one, we're going to need pretty damn good rockets to do that.
To reach maximum theoretical performance, a rocket engine nozzle needs to expand the exhaust to be of equal pressure as the ambient air. Being too underexpanded can actually destroy the engine, and even well before that being over- or underexpanded saps efficiency, so you can improve performance by specializing for the pressure you target. But of course, as a rocket ascends, pressure falls. This means that the expansion ratios of traditional engines are compromises over the pressure range they are expected to operate in.
Aerospike engines use a neat hack to make a "virtual nozzle", where the pressure of outside air is used to push on the exhaust stream. This makes it slightly less efficient than a traditional de Laval nozzle that is specialized for the exact pressure, but it can maintain that not-perfect but high level of efficiency for the whole ascent, from atmospheric to vacuum.
When everyone was trying to build single stage to orbit vehicles, aerospikes sounded very promising as they would allow a single engine to be used from the launch pad to vacuum with reasonable efficiency. However, now that first stages are routinely returning to the launch site and landing on their own, SSTOs are not nearly as attractive, and with a two-stage architecture, you want to give the second stage a proper vacuum engine, and then the first stage won't lose that much if it's optimized for near-sealevel conditions. I'm not sure aerospikes make that much sense anymore.
> From someone who knows little about this -- what is the advantage of this kind of engine?
Take a traditional bell nozzle engine. The shape of the bell is designed to redirect the exhaust in the correct direction, but one of the design parameters for the shape is the ambient air pressure. If you design it to work at a specific altitude, it will be less efficient at other altitudes.
The aerospike is more efficient over a broader range of altitudes. (That is, if you take the average efficiency over a wider range of altitudes, the aerospike wins. If you pick one altitude or a narrow range of altitudes, the traditional nozzle wins.)
Piggybacking off this too, if you look at rocket specifications, they’ll often mention the same motor on different stages with a (VAC) behind it. This denotes the different bell shape required to make the motor efficient in vacuum as compared to sea level.
The issue with an aero spike is that the benefits just aren’t there for a multi stage rocket. We do multi stage rockets to balance the needs of high thrust at the launch pad with the need to minimize our final non-payload weight in orbit. Varying the bell design to match flight profile naturally plays well with this approach. If you’re going to ditch parts of your rocket for weight saving during ascent, you might as well tune each stage for the altitude it will actually work at. Motors that need to work at all altitudes are pretty rare, the only ones I can think of are SSTOs in theory, and the space shuttle main engines.
If we could get a falcon 9 with aero spikes it would theoretically be an improvement, but not a huge one. All the gains would be in the edges of various stages where say, the stage 1 motors are flying above their designed altitude. The efficiency gains are there, but they might be completely offset by increased weight, cost, cooling concerns, etc.
My naive understanding: The appeal of a space elevator is that all you need to get to space is the energy to climb the elevator. But if Starship is able to hit its "fully and rapidly reusable" goals, it will have basically achieved the same thing: You can go to space on Starship for not much more than the cost of fuel to get there.
The next question is which version is more energy efficient. Again, my very naive understanding is that they seem to be kind of similar. The space elevator is less efficient than you might think, because you can't just run electrical wires up and down it. The weight would be a problem, and resistance losses over such a long cable would be high. Instead, you might send power through the air with a big laser, but that also comes with efficiency losses, in the same ballpark as rocket engines.
And of course, Starship has the obvious advantage that it seems like it actually might work with current technology. I think this comparison is an interesting way to highlight what a big deal it will be, if Starship does work.
I agree that space elevators are very unlikely in the next few centuries.
But arguing that Starship achieves basically the same thing seems wrong to me.
Starship is subject to the rocket equation. That means all but a few percent of its launch mass is rocket fuel.
Space hooks of all sorts require some transfer of energy, but the idea is it's on the order of magnitude of the actual potential energy gained.
Edit: It strikes me as implausible that we'll get unobtainium with sufficient material strength to build a space elevator without accompanying superconductors run through the elevator structure, but when I ran the numbers, current lithium ion batteries are about 1/60th of the energy density of mass moved from the highest equatorial point on earth to geosync orbit.
Not so very different from the rocket equation, you're right, if we decide there's no way to convey energy through the space elevator.
We haven't. We just can't produce enough carbon nanotubes cheaply to build a space elevator. Amongst other technical issues.
One limiting factor of a rocket engine is the exhaust cone. This big dome-shaped piece of internally cooled structure tries to make the burning of fuel most efficient by controlling the shape of the exhaust reaction.
In atmosphere you need a different size than in orbit to burn optimal, that's one reason why staging is done and the second stage is much different in cone size.
Since in aerospike the direction and shape of the exhaust gasses is different, air pressure is used as a "dynamic cone" making single-stage to orbit" rockets much more feasable. I'm not sure if we want that at all, though.
The space elevators problems include not only finding a good cable material, but also solving the problem of collision with satellites, so I'm not holding my breath.
Aerospike nozzles are spike nozzles where the spike is cut short and "replaced by air". They are shorter and usually lighter than equivalent bell-shaped nozzles, but they have bigger surface area in the critical section - most heat-loaded part of the engine - so cooling them is harder. For small engines the problem is cooling, and for big engines there is a problem of area where to put those engines (for large rockets length and weight of nozzle isn't a problem, but cross-section of the rocket, where the engines have to be installed, is), so aerospikes have different trade-off than bell-shaped nozzles.
They'd be most useful in single-stage-to-orbit craft, but, if they can increase efficiency of reusable boosters and orbit-to-ground propulsive landings, that's still great - an aerospike-based Starship wouldn't need both vacuum and sea-level engines and could use the same engines in space as it would use to land.
> an aerospike-based Starship wouldn't need both vacuum and sea-level engines and could use the same engines in space as it would use to land.
In vacuum, an aerospike will not be as efficient as a maximally expanded traditional nozzle. If you could conjure an aerospike raptor from thin air, you'd still probably want the vacuum engines there, but you'd of course want to replace the sea-level ones with the spikes for higher efficiency immediately after stage sep (when all 6 are firing).
I agree, but I don't think its worth it. At sea-level you rather have a sea-level optimized engine then a spike as I understand. So you would actually lose landing efficiency, and that is the most important place to save energy.
I'm not even sure if Starship will light all 6 engines on separation. Its a question of gravity loses vs lower ISP of sea-level engine. Maybe they turn them on for a initial boost and then turn the center 3 engines off.
In principle the SSTO is better for reuse because you can refuel a single vehicle and send it out again, compared to a Starship-class vehicle which has two stages that need to be stacked each time.
The trouble is that SSTO is on the margin of the possible and mastering reuse of a 2 stage system is a certain win whereas the SSTO is risky. You have to perfect quite a few technologies such as the aerospike engine, very lightweight tanks and thermal management system and get them to all work together and really be able to reuse them without tearing it down each time like the old STS.
For SSTO to be worth for a Starship-Superheavy class vehicle, the aerospike engines would need to be as efficient as the Merlin ones at the same weight and the removal of the Starship engines would need to account for added thermal protection of the lower part.
The napkin math seems workable, but rockets require a lot more than napkin math.
Having said that, Shuttles were almost SSTOs in the sense that the fuel tank could be placed in orbit (and reused as pressurized habitation space, with some extra work). Maybe some SLS-derived work can add some very large volumes able to dock into a future space station.
> Shuttles were almost SSTOs in the sense that the fuel tank could be placed in orbit (and reused as pressurized habitation space, with some extra work).
That is super cool! I gotta dig into that.. pity they never did that. It would be neat to have an ISS module that was a re-purposed shuttle external fuel tank.
The Shuttle still dropped the two solid rocket boosters.
Still it performed better than Tsiolkovsky and everyone else thought a chemical rocket could perform because running rich with extra H2 lets you get a better ISP since it lowers the molecular weight of the exhaust even if it does lower the temperature a little.
SSTO is not great, but when you add full reusability to the mix, it becomes attractive. You spend more in propellant but makes operations so much easier.
I would say it like when you have the best ever basketball team with all Hall of Famers on it. And then team that has never done anything drafts a rookie that might be good (nobody knows) but at least that rookie looks cool.
