sevenperforce

There has been a lot of back and forth about what the lower control surfaces ought to be called. Wings? Flaps? Flaperons? Feathers? But accuracy is key. So why not call them what they really are? Plasma Deflector Shields.

It would depend on the RMS velocity of the gas in LEO. Any with escape velocity have a decent chance of escape. Let's see... the exhaust velocity of raptor is ~3.7km/s, and burns are generally tangential---but the wrong way---so that would in fact make them no longer have orbital velocity. Yeah, it seems like the exhaust products in LEO have a pretty strong chance of heading home. You don't need to hit earth escape velocity; just getting into or near orbit should be enough. I suspect, though I am not sure, that solar wind will be sufficient to disperse any CO2 molecules which reach a sufficient altitude. Whether that altitude is 80 km, 100 km, or 500 km is not known, but I'm sure they'll do the math if they need to. And of course any propellant used BLEO will count as negative carbon. Single-stage P2P seems very possible. And they are primarily testing skydiver mode here, after all, where CoM/CoP is everything.

Agreed. The outer mold line (OML) has a particular "default" center of pressure (CoP) and a baseline center of mass (CoM). You can adjust the CoM to a degree by putting in header tanks and other internals at various points. Your fins will have complete pitch authority as long as they are positioned opposite on the other side of the CoM from the default CoP. Fold them back, and the combined CoP moves above the CoM, causing pitch-up torque; fold them out, and the combined CoP moves below the CoM, causing pitch-down torque. Feather them halfway and you get a CoP aligned with your CoM and thus a stable entry. So as long as the fins are large enough, you can manage. One reason to have four fins (fore and aft) rather than just two is that it will damp dihedral torques for roll control. If you are stable in pitch because your CoP is resting on your CoM, then feathering back one of the two aft fins to correct roll will lift the CoP above your CoM and induce pitch-up. You can correct by feathering the opposite fin forward, but then you have a roll moment that is very far from your CoM and will thus tend to induce yaw. Having the canard-fins will allow roll authority that keeps everything located on the CoM. The likelihood of moving from the "chomper" payload model to a Shuttle-style cargo bay seems high, at least for the cargo variant. When I was trying to do builds to test the last version of Starship entry I kept having to put header tanks in the nose to trim the CoM, too.

Well, "storage" concerns are sort of the trick here. If you're going to be using a bunch of energy to capture CO2 and sequester it, fine, but you need a sequestration method that won't outgass or do any other stuff like that. A little more energy and you can crack the CO2 into fuel, which you can sell, but selling the fuel means you're no longer capturing the CO2. Carbon-neutral but not carbon-negative. The difference for SpaceX is that a significant amount of the CO2 they'd crack into methane would be potentially leaving the atmosphere altogether. Granted, some will fall back to Earth, but they can prove trivially that a great deal will not, and so for that portion that escapes, they are carbon-negative. So they can essentially have their carbon-cake and burn it too. My eyes may be deceiving me, but it looks like there could be a mounting point for forward canards on the left edge of that sucker.

If all that is being stuffed into the nose, I wonder if we see a move to a Skylon-style cargo bay/crew compartment rather than the pretty nose-mounted crew cabin. Elon mentioned going carbon-neutral on Earth, ultimately. This could be a super-lucrative move if more nations transition to a carbon tax. Companies can buy negative carbon from Elon less expensively than paying a carbon tax.

Indeed. My speculation has been rapidly invalidated. Calling them "fins" is really much simpler than "flaperon" or "stallwing" or whatever else we were coming up with. They are without question fins on the way up, after all. With the legs clearly not going through those fins, I think we have to assume the legs are mounted to the thrust structure. No other way they work.

If we end up with elevators on the tail, that's possible. But I would be worried about structural integrity in a horizontal landing situation. At most we might see a contingency gliding splashdown in the case of an engine problem.

I am certain they do not move up and down. The "rail-thing" is a propellant line. The wings are fixed-mounted to a rotating element at the base. If the nose is draggy enough, it ought to balance well enough. One of the things I encountered while testing a Starship mockup shortly after the last version reveal was that it was easier to just keep the forward canards fixed and "fly" on the rear ones as well, since they had so much more authority. You can fold them flush to pitch up, fold them all the way out to pitch down, and fold them differentially to roll. I would be very surprised if they didn't take advantage of working closer to the ground on this point. I'll wager that if they don't attach the canards to the fairing on the ground, they won't do forward canards at all.

Seeing this, I'm thinking there is a fair likelihood that the forward canards are not actually being planned at all, any more. Using only two control surfaces cuts the mechanical failure points in half. Also, note that these are actually substantially higher on the body than the last design, placing their center of pressure much closer to the nose. If we assume the "fairing" on the nose is going to be quite draggy (remember, despite cargo, that the tail of this sucker is heavy AF), then it could produce a center of pressure quite high even without forward canards. The center-mounted wings would be able to control pitch and roll quite well, leaving yaw to be handled with thrusters. Though that does present the question of what those earlier smaller control surfaces are. With some of the closeups, I think the leading edges are not so sharp as it seems from a distance. So this might be entry-ready as it is. Finally, keep in mind that we may see these come off and back on a few times if they are doing high-level fitting.

Another thought.... What if they take a page out of the Starhopper playbook and use 3-6 fixed landing legs? Stronger, fewer moving parts, no deployment fairings in the entry airstream, no JASON-3 incomplete deployment risk. The lower risk could permit a slightly smaller footprint.

So to recap, it looks like we have just two wings, tucked in a housing that also shields the external propellant flow lines (which is a smart choice, both for servicing and construction reasons and to save as much volume for internal props as possible). I'm curious as to the smaller control surfaces we saw before. Are those forward canards? Or are they body flaps mounted underneath the wings, or even articulating on the wings underneath? One of the big unknowns with EDL was the transition from "skydiver" flight to tail-sitting powered descent, with the very serious likelihood of a lawn-dart scenario. Having some control surfaces that act more like elevators would be very useful.

The elephant in the room seems to be reuse. Even assuming all the turbomachinery and combustion chambers work as intended, all we end up with is a very mass-efficient way to get into space, but not necessarily any way to get back. His image shows a biconic entry, and he certainly has the mass budget for a little shielding, but merely claiming "biconic entry" does not mean EDL is solved. In fact, that thing looks so much like a lawn dart that I'm very puzzled to know how he imagines he'd ever be able to land it. If orbital EDL was this easy, then ULA and SpaceX and Roscosmos and Arianespace would already all be re-using their upper stages. Even with Starship, where you have recovery margin for days, it has been a long hard track to get to a workable design. I would almost suggest that focusing on a Skylon-like approach is better -- build the engine as a monolithic whole and let the EDL issue be a vehicle problem. With a mass fraction (40%) fully eight times better than Skylon's (5-6%), he has all the margin in the world for a vehicle solution to EDL, but only if his engine is monolithic and can be nacelle-mounted.

I mean, you're right -- the flow choking point is very important. I don't see any reason to think it is not controlled/designed here. There may be a very slight divergence to ensure that the choke point is firmly inside the extensions, but those individual "nozzles" are more throat and less nozzle. Very accurate however that the flow is already supersonic (though wildly underexpanded) when it hits the main nozzle.

Cooling is not a problem for NTRs.

I am pretty sure the entire wings articulate. The gap looks like that. Also, the axis of rotation looks like it will be just past the chord edge, so that's another reason for the gap.

It's really hard to tell whether the fin currently being mounted will articulate or if it has a flap on the base or edge that articulates instead.

Maybe something like this?

Thanks! Similar to a LOX-afterburning rocket, where the liquid oxygen is injected in the nozzle rather than in the chamber, I think the ideal design for any airbreather is to have an air inlet with injection/combination below the nozzle throat. The mixing problem is immense for any airbreather, and this injects the air where the propellant flow is already supersonic but at its highest density and greatest turbulence. Much better than trying to inject in the combustion chamber, and completely obviates the need for trying to solve supersonic combustion.

Magnetic field lines or gravitational field lines? Magnetic field lines can absolutely be used to "push" against. That's how a magnetic tether works. Gravitational field lines are just a force gradient. They cannot be "pushed" against. A ship cannot push against itself. It's like trying to pick yourself up by pulling on your own suspenders.

Nothing wrong with active cooling. Ordinary bell nozzles need active cooling too. All operational bell nozzle rockets are either actively cooled or ablatively cooled (or, in the case of vacuum engine bell extensions, radiatively cooled). The easiest way to do this is to pipe some of the chilled propellant through channels in the engine bell before injecting into the turbopump or combustion chamber or whatever. Aerospikes, due to their inside-out configuration, are actually quite amenable to active cooling in this way. Precisely. One rarely-known fact about aerospikes is that the original term referred to a plug/truncated-spike nozzle, rather than a "pointy" spike nozzle. The "aero" in aerospike referred to the fact that gases would be entrained under the plug and provide a virtual spike extension, saving weight. Good questions. An ordinary bell-nozzle engine has a combustion chamber, a throat (where the flow converges and then diverges to transition from subsonic to supersonic), and an expansion nozzle. You need that converging-diverging shape at the throat in order to create the Mach shock that produces the supersonic exhaust transition. What you're seeing in the multichamber design is properly an extended throat coming off the combustion chambers. So it is a multichamber engine with 12 cylindrical chambers, 12 extended throats, and a single plug nozzle. I know the extended throats look like little nozzles because they are spread out, but it's really just a way to convert the cylindrical flow to a thin layer covering the actual nozzle, and the throats are not actually "diverging" in cross-sectional area until they open onto the nozzle. Our understanding of flow and combustion in cylindrical combustion chambers and spherical combustion chambers is much, much more advanced than our understanding of flow and combustion in a toroidal chamber or annular cylindrical chamber. So it's easier to do a multichamber design, usually. But you can certainly have a single chamber with multiple throats. You just lose some of the advantages like differential throttling. Yes, you can absolutely do this. The trouble is in the heating cycle. The upper bounds on conventional bipropellant rocket engines come from the chemical potential energies of the propellants, while the upper bounds on a nuclear-thermal rocket come from the melting point of the reactor. You want the reactor to heat up the propellant as much as possible, almost to its own melting point, but not beyond. Accordingly, you want the propellant to hit your reactor as cold as possible, and exit the reactor as hot as possible. It's an effectively infinite source of heat (unlike chemical potential energy); it's just got a maximum operating temperature. So it's not feasible to try and heat hydrolox exhaust after combustion. You'd be wasting the chemical potential energy since the reactor is going to get it to that temperature regardless. What you can do, however, is to use hydrogen or methane in a nuclear-thermal rocket, and then inject liquid oxygen after it passes through the throat. This way you are adding energy after the maximum reactor temperature has already been reached. The resulting combustion is not as efficient as combustion in a dedicated combustion chamber (subsonic combustion is always better than supersonic combustion), but it definitely adds hella thrust. Reactor can't possibly electrolyze water fast enough for these purposes. The problem with collecting oxygen from air in flight is that the air is REALLY really hot once you're moving at any appreciable speed, and so you end up needing to carry ridiculous amounts of liquid hydrogen to prechill the air, like SABRE. Additionally, there's a thrust problem. If you're trying to use propellant from the air, you have to collect it, which means speeding it up to your own speed. If you are flying at 2 km/s and collect one tonne of air every second, then burn it with hydrogen to push it out the back of the plane at 4.5 km/s, the actual effective thrust is as if you were only pushing that air out at 2.5 km/s. Once you accelerate to 3 km/s, then the effective exhaust velocity drops to 1.5 km/s. Once your airspeed exceeds your exhaust velocity, thrust becomes negative.

No material is heat-resistant enough to be the nozzle for any high-performance rocket engine. Not tungsten, not titanium alloy, not "diamond tungsten alloy" (which is impossible), not even tantalum halfnium carbide. Any operational rocket engine requires either ablative cooling or active cooling.

Wedge-shaped linear aerospikes can have thrust "vectoring" by differential throttling on the individual combustion chambers. On a linear aerospike like ARCA's (note, ARCA is a total scam, but their HAAS-2CA could be useful under different circumstances), there are eight chambers on either side of the aerospike. If they downthrottle the chambers on one side, they can pitch; if they downthrottle on opposite corners, they can roll. A combination of pitch and roll can produce whatever attitude control you need. If you do a toroidal aerospike, you have two options. You can use a multichamber toroidal aerospike, in which you can use differential throttling of the individual combustion chambers to effect pitch and yaw (though not roll): Or you can use an annular toroidal aerospike, which only has a single combustion chamber and accordingly cannot sustain any differential throttling: The annular toroidal aerospike is the most similar to a standard rocket engine, in terms of operation. It has a single combustion chamber and only one throttle setting. There's no reason you cannot gimbal the entire engine in the same way you would gimbal a de Laval engine. Your gimbal mechanism will need to be a bit beefier, however. The engines in KSP only show nozzle gimbal, but in reality some engines have gimbaled nozzles and some have gimbaled engines. I believe that the F-1 and SSME had full-engine gimbal while the Merlin 1D only has nozzle gimbal, but I could be wrong.

I always prefer fixed-component solutions. Like how the aerospike solves a problem that would otherwise require a moveable-geometry nozzle, simply by turning the problem inside-out. One of the challenges for any airbreather is the mixing problem. Whether you're just using air as working mass or for a portion of combustion, you have to have a way to mix the mechanically- or shock-compressed inlet air with your fuel (or your exhaust, in a working-mass-only approach). SABRE handles the problem, of course, by actually collecting, superchilling, and injecting the air into the engine. Anything that is pass-through, though, needs a way to ensure complete mixing. Linear aerospike engines (and some toroidal ones, though not of the annular variety) have multiple small combustion chambers which open onto the nozzle surface. I wonder if it would be possible to create an inlet between the chamber exhaust and the nozzle surface, so that the air inlet would enter perpendicular to the exhaust flow and thus mix very effectively.

Of course I would love to see someone try to build one. It just requires a LOT of moving parts with low tolerances under extreme conditions. You can have dry mass growth or slightly lower performance and not run into problems, true, but the tolerances and flight conditions are non-negotiable.

Really interesting! I've seen the rocket blade cycle approach before. It is a troubling starting point because I feel like the materials science challenges of trying to force rocket exhaust through rotating fan blade members seems immense. You'd almost definitely run into engineering tradeoffs that would reduce your specific impulse dramatically in the pure-rocket segment of flight. And speaking of the pure-rocket flight envelope, I think he's going to have trouble with the scramjet and film-cooled scramjet envelope. A non-combusting scramrocket is going to have slightly lower specific impulse but the drag penalty is lower and the workable envelope is wider. If he does a ramturborocket to ramjet transition and then goes to pure rocket shortly thereafter, he's going to have lower drag penalty and a lot simpler of a system. Doing that with drop tanks is probably the better approach. It would be interesting to reimagine Skylon with this engine, drop tanks, and methalox fuel.