sevenperforce

That seems odd to me as well. The booster lands approximately five miles downrange so that's a pretty vertical re-entry nominally. Maybe the abort kicked the capsule back toward the launch site which made it more vertical?

Again, these numbers aren't right; you need to factor in the additional engines that an expendable Starship Lite SSTO would need to get off the ground. But even if they were right, what of it? If SpaceX is willing to fly Starship upper stages expendable, then they would just put those expendable upper stages on top of the reusable Superheavy booster. Then they only need to throw away three vacuum engines and they get 175 tonnes to LEO.

Here's what that absolutely horrific monstrosity would look like, just in case you've got extra eye bleach you need to use:

We must always and forever lament that Orion Lite wasn't flying ISS missions this whole time. If it had been, we'd be accustomed to Orion reuse and adaptable Orion service modules. Things like upgrading the heat shield and doing an extended service module for cislunar missions would be natural evolutions rather than clean-slate challenges. But, alas, it was not to be. Orion is a little bloated for a capsule, but given its additional capabilities and carrying capacity, the bloat over the Apollo CM isn't THAT bad. It really just needs a meaningful service module. Lunar Gateway makes no sense absent Orion's current SM, but let's suppose we give Orion's new service module a total of ~2.1 km/s of dV. That's enough to shuttle between TLI and LLO with one stopover in NRHO (but not 2). Current injected lunar mass is 26.5 tonnes with 8.6 tonnes of propellant, giving it the measly 1,216 m/s it has right now. Multiplying the propellant mass by 2.3 (and adding on another ~2 tonnes for tank mass growth and associated margin) gives it 2,150 m/s and gives an injected TLI mass of just under 40 tonnes. Is there any commercial upper stage big enough to push 40 tonnes to TLI? You've got to have some reasonable T/W ratio at staging, after all. Here's what existing and near-future upper stages would do for a 40-tonne evolved Orion (note, some values estimated): Stage Stage m0 Stage mf Total thrust & Isp T/W ratio w/OrionX dV w/OrionX Centaur III DEC 23.3 mt 2.5 mt 198 kN, 451 s 0.32 1,762 m/s Centaur V 57.2 mt 5.2 mt 214 kN, 453.8 s 0.22 3,407 m/s New Glenn US 190 mt 15 mt 1,420 kN, 449 s 0.63 6,300 m/s Starship Lite 1240 mt 40 mt 7,825 kN, 378 s 0.62 10,160 m/s ICPS 34.2 mt 3.5 mt 110 kN, 465.5 s 0.15 2,438 m/s Ariane 6 ULPM 34.6 mt 3.6 mt 180 kN, 457 s 0.24 2,407 m/s Falcon 9/H US 111 mt 4.5 mt 934 kN, 348 s 0.63 4,170 m/s Any upper stage for an evolved Orion would need to do the same job as the S-IVB: completing orbital insertion and then providing the TLI burn. The requirement for TLI is 3.2 km/s, so that alone eliminates everything but Centaur V, New Glenn US, F9/H US, and of course Starship Lite. However, I'm skeptical that a staging T/W ratio of 0.22 will be sufficient, given the need for abort modes. For those four stages, this is the staging velocity you'll need: Centaur V: 7.6 km/s New Glenn US: 4.7 km/s Starship Lite: 0.8 km/s Falcon 9/H US: 6.8 km/s So, the question is whether there's a way to build a commercial frankenrocket capable of lofting any of the above configurations to those respective staging velocities. Note that Starship Lite is still just massively overpowered for this job. It would make more sense for SpaceX to build a Starship Ultralite with shorter tanks and only a single Raptor Vacuum if they were asked to send a 40-tonne payload to TLI in a single launch. There has already been some talk about a three-core Vulcan Heavy. It's horribly cursed, of course, but I would suspect that an expendable Vulcan core with two Falcon Heavy side boosters could surely loft the New Glenn upper stage to 4.7 km/s...possibly while preserving recovery of the FH boosters and perhaps even SMART reuse on Vulcan. The diameter change from Vulcan's 5.4 meters to New Glenn's upper stage 7 meters would be horrific, though.

The trouble is that SH+SS is just so capable that any reasonable analysis ends up reducing to "just use Starship". Comanifesting is fun and all, but part of the point of the whole way Artemis is envisioned is that comanifesting isn't really the point since you're sending other stuff separately, because neither Orion nor SLS have the capability. If political considerations were nonexistent and you were willing to field a lego rocket that is truly accursed, I wonder what capability you'd get if you slapped two Falcon Heavy side boosters onto a New Glenn core topped with a Centaur V upper stage. But Orion's SM is ultimately the problem.

Elon already said that a maximally slimmed-down Starship would only be around 40 tonnes. That's stretching the math. No, it would not, because your estimates for "reusability systems" aren't reasonable. There's already a reusability system for Starship: the heat shield and flaps and forward LOX tank. There's no re-entry configuration for Starship that would achieve meaningful payload at lower dry mass than the existing cargo Starship.

Definitely takes much longer, as @mikegarrison notes. Which is not ideal for life support, obviously. However, it is a way to get a free-return while also reducing the size of the insertion burn. I've done it a number of times in KSP when I have extra margin on the TLI stage but I still want a free-return. Obviously in KSP I could leave the TLI stage attached since you get unlimited restarts and infinite throttling, but it doesn't look as pretty. IRL, it could have been used for something like Artemis 1 to increase an uncrewed Orion's cislunar capabilities while also setting up for a re-entry test. With the high-apogee prograde free-return, you actually are still being gravity boosted into a more energetic orbit, but your trajectory is being altered so that you end up going back through Earth's atmosphere with that higher energy. You can also use this kind of a maneuver to do a more energetic Oberth maneuver at perigee for a deep space mission. The faster you're going when you start the final ejection burn, the more help Oberth will give you, but there's an obvious limit to how high your apogee can be. Accordingly, you could use this kind of trajectory to get a lunar gravity assist added to your ejection energy before your final ejection burn. Don't know whether this has ever been done in real life. Yep. The rotational direction doesn't matter but the orbital direction does, obviously. That's where the gravity assist (positive or negative) comes from. Apropos of nothing (somewhat), this is essentially the maneuver depicted in the xkcd comic that first introduced me to Kerbal Space Program:

You can also do a high-energy free-return where you circle the moon in a counterclockwise fashion but with a very high apogee, so you don't really get affected by the moon's gravity on the outward journey but you do a close pass on the return journey that will nudge you back into an Earth entry interface trajectory. This is useful if you want to take maximum advantage of the Oberth affect and you want to frontload some of that energy using your launch vehicle during the TLI burn and use less dV for a counterclockwise low lunar capture burn. But this sends you back into a free return with a much higher Earth entry velocity than you would get from a retrograde free-return, so your heat shield's mileage may vary.

Which is precisely why you would use those power packs with a propellant tank. You can simply handwave power density and specific energy, say you have this "uber" energy source, and move on from there.

No, that is not correct. The process of converting heat and pressure into velocity through a converging-diverging nozzle is a function of the total heat and pressure in the supercritical fluid in the chamber. Directionality has nothing to do with it. The nature of choke point at the throat is such that whatever is happening inside the combustion chamber is immaterial once the fluid passes the throat. Second, lasers are not magic. Even though the photons in the laser beam are parallel and coherent, that coherence and directionality is lost once their energy is absorbed by the fluid in the chamber. If you have something like metastable metallic hydrogen that you can decompose at will, then just put in smaller amounts continuously. Way easier than trying to do something with a laser. Getting a fluid to absorb photons from a laser will require that the exhaust be opaque to the laser wavelengths, which also may not be the case automatically. If you have a power source that has high specific energy, high energy density, and high specific power, then there is absolutely no reason to worry about pumping a laser gain medium to transfer that energy into the propellant. Just pump that energy directly into the propellant itself. (There is a sense in which the throat choke of a rocket engine is the sonic equivalent of the coherence gain process in a pumped laser gain medium but that is beyond the scope of this discussion.)

That is the ordinary way it works, yes. A converging-diverging throat creates a shock transition where the flow upstream is subsonic and the flow downstream is supersonic. Since pressure waves cannot propagate backwards in supersonic flow, this directly converts pressure into velocity downstream of the throat. For in-atmosphere operation, you don't want to make the downstream nozzle too large, or the pressure at the nozzle exit will drop lower than the atmospheric pressure, allowing atmospheric air to push its way inside the nozzle as the exhaust rushes out. In a near-vacuum you can make the nozzle as large as you want. Heating up the exhaust after it completely leaves the nozzle is useless. Heating up the exhaust while it is still inside the nozzle will increase thrust (also known as afterburning or reheat) but it will not do so as efficiently as if you added that heat inside the combustion chamber, before the shock at the choke point in the throat. Lasers are generally not considered to be a good way to reheat exhaust.

Some additional notes from wikipedia, LinkedIN, and elsewhere... It looks like the putative upper stage design is two turbopumps with fifteen propulsion chambers. Strange to have fifteen propulsion chambers; I would have expected an even number so that the engines could be shut off in pairs to reduce deep throttle requirements, but expander cycles tend to handle deep throttling pretty well, so there's that. It's also possible that they're simply starting small and only doing 15 active chambers but the final design has 30 active chambers. Based on some of their tweets it seems like both stages will be hydrolox, but I'm not sure. Diameter appears to be around the same as Falcon 9, maybe slightly smaller. Here's one cool photo with a notation about the actively-cooled heat shield: From the patent disclosure it sounds like the entire engine cluster may gimbal, which is odd because i figured it would be simpler to just use differential throttling for pointing and RCS for roll control.

If you want a spectacular green flame then you should go with pentaborane(9) and oxygen difluoride, and you'll get a whopping 361 seconds of specific impulse at sea level and 432 seconds of specific impulse in a vacuum, with a propellant mix bulk density 17% greater than kerolox, 46% greater than methalox, and 314% greater than hydrolox. Of course, your exhaust will be toxic and carcinogenic and your fuel has the unpleasant attribute of spontaneously bursting into flame upon exposure to air, but at least you never have to worry about an ignition system. If that's not enough bang for your buck, you can always go extra fancy and melt liquid lithium and then burn it with liquid fluorine and gaseous hydrogen. Don't ask me why the hydrogen has to be injected in gaseous form. It will get you 542 seconds but you have to come up with a way to effectively store and then mix a liquid fuel at 181°C with a liquid oxidizer at -189°C, so good luck with that. The amount of smoke and debris is concerning...I imagine there had to have been some damage to the test stand, which must not have been intentional. But I would think whatever this was, it was clearly intentionally risky.

Amusingly, this is exactly how the original creator of the OPT Spaceplane mod figured the 2.5m Nebula jet engine would work. Also keep in mind that for any ion-based propulsion system to work, you have to have a way for your net exhaust to be electrostatically neutral, or you end up with charge buildup on your vehicle. For ion thrusters, this is accomplished by firing a beam of electrons into the exhaust stream. It doesn't add any thrust but it keeps you electrostatically neutral: Also in an ion engine the actual acceleration is produced by an electrostatic field (between the positive and negative grids) while the magnetic field is what does most of the work to ionize, not the other way around. My background is physics, not pchem, but it could be possible to use exotic propellants and a multistage combustion chamber to generate two ionized exhaust streams rather than a single electrostatically-neutral exhaust stream. If you then had some outside source of electrical energy, you could use electrostatic fields or magnetic fields to accelerate those two exhaust streams and produce additional thrust. However -- and this is critical -- you must have some outside source of electrical energy. You cannot get the energy from your propellants. An engine which produces two ionized exhaust streams will always be less efficient than an engine using the same propellants which simply allows them to fully combust, because two ionized exhaust streams still have residual potential energy until they react with each other.

Well it looks like I'm pretty damn smart. Patent AU-2020398126-A1 Excerpts from the patent disclosures: And it looks like @tater is right about it being an aerospike. Here's the patent for that. The whole design is, frankly, pretty genius. It also explains more clearly why they are using hydrolox for their upper stage. This second stage is, in essence, an integrated expander cycle engine and actively-cooled heat shield. You will recall that an expander cycle engine like the RL-10 pumps liquid hydrogen around the nozzle at high pressure, superheating it, and then lets that superheated supercritical high-pressure hydrogen expand through a turbine. The turbine runs the turbopump which pressurizes the liquid hydrogen in the first place, and the turbine outlet runs into the combustion chamber directly. This is a very very efficient design, since all of the working fluid passes through the combustion chamber and is fully burned. However, it does have one major disadvantage: due to the square-cube law, the amount of nozzle area available for the heat exchanger won't increase as fast as the power needs of the turbine, putting a maximum possible size on such an engine. That's why the SLS EUS uses a cluster of four RL-10s rather than just one much larger expander-cycle hydrolox engine. The square-cube problem can be solved, however, by using an annular or linear aerospike nozzle, which provides ample heat exchanger area. What Stoke is proposing to do is pretty smart. They are using a single-turbine aerospike expander cycle hydrolox engine, but the aerospike nozzle surface is ALSO the actively-cooled aft heat shield. Thus, when the engine is running, it is operating like a normal expander cycle engine, but during re-entry, the heat from re-entry is used to operate the exact same turbine and pump used for the engine, but without any liquid oxygen flowing. As a result what is exhausted into the combustion chamber is merely gaseous hydrogen, which bleeds out through the nozzle. The hydrogen bleed produces thrust and creates a boundary layer of gas which further aids cooling, and part of the gas can be tapped off to maintain autogenous press, provide hot-gas RCS, and the like. Absolutely groundbreaking. Really solves SO many problems. It does mean you have to carry more liquid hydrogen, but a fluffy upper stage is better for re-entry anyway.

Upper stage is hydrolox which is a good choice for second stages but tough to work with for a startup. They look like bespoke designs to me. Also clearly very fuel-rich, perhaps for low-throttle setting. Feels very much like something I would design in KSP. Not that that's a bad thing. I wonder if they are going to have articulating shields of some kind for the engines or if they'll just re-enter raw. One possible solution would be to vent gas through the nozzles during re-entry at low pressure to create a boundary layer...

Gobble gobble, nom nom. Just an aside, but as usual, public reporting on spaceflight sucks SO bad. This, from EarthSky: Yes, "perhaps" Falcon 9 would struggle with lofting a 1.5 tonne payload.

The abort motor gimbals, yes. The retros definitely fired; you can see them kick up dust at touchdown. They have crush cores like child carseats. (sounds like the retro is compressed air, interesting) Now that I did not know. I wonder if they mean a pillow of gases (which could include a gas generator) or actual bottled air?

The reason rocket engines do not melt is because they carry the heat away with the exhaust through the propellant. But even that has it's limits... since if the energy injected into the propellant is so high that the mass flow rate is not enough to reduce the heat to a safe level... bad things happen. Unless you're dealing with an actual torch drive, the heat capacity of the propellant flow will always be far greater than the heat rejection needs of your engine.

They have a script.

You can always add thrust by adding energy to an exhaust stream. If you are getting that energy by taking it away from an earlier step in the process, it will make the whole engine less efficient. If you are getting that energy from a separate energy source then it will make the whole engine less efficient unless that separate energy source has better specific energy and energy density than the original engine. If the separate energy source has better specific energy and energy density than the original engine, just use that energy source to build an engine from scratch running only on that energy source. Once the easily ionizable gases are injected into the propellant flow (whether upstream or downstream of combustion), they will mix and will no longer be easily ionizable.

You keep using this number. It does not mean what you think it means. You are also arguing backwards. We know what SH+SS can take to LEO in reusable mode; pointing out that it could take a lot more payload into space in expendable mode does not make the payload of a reusable SSTO any closer to what SH+SS can do. That's also backwards. And they will definitely not manage to even get half the payload.

That's one of the things that jumped out to me. I would have expected the abort to trigger a second or two earlier, when the engine anomaly started.

Looks like the abort system triggered on an out-of-angle condition.

"just a 10% higher speed" is not nothing.