sevenperforce

Yikes. I don't like the way that sounds.

Yeah, Vulcan is in the vein of Atlas V, and Atlas V really was optimized for high-energy missions, and high-energy missions really make upper stage reuse challenging. Just look at Starship -- it can deliver 100-150 tonnes to LEO but I believe it is only slightly better than FH for GTO. I'm not sure that it can even do GEO missions at all. If Starship is successful then I could see the growth of transporter ferry spacecraft that pick up individual comsats from LEO, take them to GEO, then aerobrake back down to LEO to refuel and repeat.

Not too much technical data but there are a few interesting things. Spoilered to make for easier viewing.

Elon says it's a grid fin actuator. Trying again on Saturday.

Thanks! I hit him with a couple of technical questions.

Ah, good catch. That might well be the gimbal mount, and the mounting brackets on the sides of the engine bell could simply be for the gimbal actuators. That exaggerated lip between the nozzle and nozzle extension is kind of weird, though. This is ORSC so I'm not sure why there would be any boundary; it's not like they need an exhaust injection manifold a la F-1 or MVac. Maybe that's part of the regenerative cooling manifold and they're cooling part of the system with LOX and part of the system with CH4? If I was putting hot oxygen anywhere I would put it as far down the nozzle as possible so that there wouldn't be as much corrosion or risk of failure at burn-through. Now you've also got me wanting to speculate about what kind of ORSC they're doing, whether they are doing a single-shaft turbine with a helium purge or some other design, whether they have boost pumps, and so forth. It would be cool to do, say, an expander-cycle boost pump. There's not any visible detail on that image. I wonder if they are going to try test firing it from pressurized tanks without the preburner? Is the Centaur on Atlas V the same regardless of position (under or not under the fairing)? The fair it when the loads are too high (more SRBs), so it's clearly load driven at some level. Neutron doesn't face that issue. It might be possible to make an even lighter Centaur for lower mass payloads with high C3 (if said Centaur was faired)—but this being "old space" probably not worth the effort. Centaur uses the same tanks whether or not they use the fairing. And the fairing isn't just for bigger loads; the Atlas V 501 (5-meter fairing, no SRBs) has been launched a total of seven times. It does like like the 4-meter fairing has never been launched with more than three SRBs, though, so maybe that's where the aero loads get dicey for Centaur. But the comparison is really something. Centaur has a dry mass of 2,316 kg, which less the 190-kg RL10C-1 comes to 2,126 kg. That, for an external cylindrical volume of 77.15 cubic meters and a total prop load of 20.83 tonnes. Pixel counting is always rough, of course, but based on a 5-meter outer diameter for the Neutron second stage I'm estimating an external cylindrical volume of 119 cubic meters. Even if the more rotund design means lower volumetric utilization -- let's say around 105 cubic meters -- that's 36% more propellant volume than Centaur and 3.02 times more propellant mass given the higher bulk density of methalox over hydrolox. Looking at 63 tonnes of propellant for a 300-400 kg tank is just utterly shocking...that's, like, a propellant mass fraction a little over 99%. Trying to sanity check this. Assuming typical T/W ratio on the order of 150 for the 890 kN vacuum version, I would expect the weight of a single vacuum-optimized Archimedes engine to be around 600 kg, which still leaves us with a 98.4% stage propellant fraction. They advertise 13 tonnes to LEO and 1.5 tonnes to either Mars or Venus. You need about 3.9 km/s to get a transfer orbit to Venus out of LEO which is lower than typical transfer orbit requirements for Mars so let's go with that. Back-calculating with the expected 365 seconds of specific impulse gives us 6.1 km/s of Δv on the ascent stage with a 13-tonne payload, meaning that the first stage has very little work to do other than getting out of the atmosphere. Back-calculating for a 1.5-tonne payload gives us 11.6 km/s of Δv, which is 5.5 km/s of extra: higher than needed for the Venus transfer, but not entirely out of the ballpark. And we're working with dicey numbers here anyway because we don't really know what a Mars or Venus transfer looks like, or whether that's even reusable.

I'm trying to get my kids' elementary school to stream the launch, but Twitter is blocked by their server. Is it going to be anywhere on YouTube or anywhere else?

Yeah that's truly incredible. If they can really get a 5-meter methalox tank in under 400ish kg then...just wow.

Oh, I see your point -- adjusting to where the booster is, not trying to push the booster to the center. Yeah, that's workable.

Yeah, same. The RTLS landing of the last F9 was waaaay off-bullseye when it came down. I know that before they were talking about giving Superheavy those ten-tonne gas-gas methox thrusters for translation during hoverslam and that seemed like a workable plan but AFAIK they have scrapped the thrusters entirely. Yeah, but I don't think the chopsticks are capable of, like, "pinching" the booster into place. And catching Starship with the chopsticks seems like a REALLY long pole.

you are actually incorrect in that statement. N2O4 is just NO2 in liquid form. considering its NO2's melting point is above stp conditions, when removed from a cryogenic thermos or not actively being cooled, it will boil off into NO2 gas. Holy necro batman! But anyway this is sort of true. N2O4 is a different molecule than NO2; it's a dimer of NO2, two -NO2 groups bonded together. But the N-N covalent bond in N2O4 is a very weak bond because the paired oxygen atoms on either side of the dumbbell are holding onto the electron shells tightly, so you only need a small temperature transition to break or form the bond. Red fuming nitric acid, a mix of nitric acid, N2O4, and water, has a lower freezing point than pure N2O4 (as noted above four years ago by @IncongruousGoat) which is great. But it will continue to "fume" toxic NO2 continuously, whereas pure N2O4 will produce very little NO2 gas as long as it is kept below ~12°F. So you can load the propellant tanks at low temperature, tamping down fume production, and then you're good as long as you aren't trying to launch at significantly lower temperatures where N2O4 will freeze.

The top of the engine looks like the SM-6A Service Module from the Making History DLC: Maybe they plan on putting the turbopumps on those attachment nodes? /s In more serious review, it looks like there are gimbal mounting points on the nozzle extension and the nozzle extension seam is fluted, so this could be a true throat-gimbaled engine! Those are fairly rare. I don't think those bits up top are a gimbal, although they might be.

I wonder what capture/collection system you'd actually want to use for that. Titanium's melting point is around 3000°F which is lower than typical cislunar-return re-entry heating, but let's suppose that the lunar manufacturing was such that ablative losses were negligible. What would the ideal collection system look like? Shoot the titanium slag on a trajectory such that it lands in specified desert target zones, then come in after the fact with a bulldozer and scoop up all the sand on the surface and filter out the chunks of metal?

Really fascinating stuff:

Yeah, that's what it looks like. Hence "allegedly". Apparently Willis has gotten good intel previously. Comparing the two images, it is obvious that they are from the same heritage, but if it is a photoshop of the original then it is a VERY good photoshop.

New HLS renders, allegedly (they look pretty legit as these things go). From David Willis on Twitter.

I don’t believe so, certainly not with Skylab and pretty sure on the ISS. Wasn’t much need, since they all had crew tunnels, and seems very high risk. Just looked it up -- there were plenty of spacewalks to assemble ISS, but the first that would have allowed an EVA between vehicles was STS-104 when the Quest airlock was actually installed. Although there were three spacewalks during this mission, two from the Atlantis airlock and one from the Quest airlock, nobody ever left one airlock and entered through another. Maybe it has never happened! Apollo 15 required an EVA to retrieve a recording device from the Apollo service module, but this was after the lunar module had been detached. For Skylab 2 (the first crewed mission to Skylab) the crew had to EVA, climb onto Skylab, and manually disassemble part of the Skylab docking ring in order to get the ring to function for hard capture, so this was at least a transfer between vehicles, even if they didn't enter Skylab via EVA. But yeah, more interesting to see instances where EVA transfer was planned and integral, like in the N1 lunar lander. No, I'm talking about the 1L/Vostok-7 spacecraft (not to be confused with the cancelled Vostok-7 mission, the Vostok-7 I refer to is of the same category of designations as the Vostok-3KA, the official name of the original Vostok). Wow, cool! I love how I'm continually learning new stuff here. Very Kerbal. But that begs the question, why not just launch in the 1L? Not the first time the Soviets launched with no abort ability… They may have had no concerns about launch abort, but the mission profile called for sequential crew-guided assembly of the propulsion modules with the circumlunar crew vehicle attached last: The IL vehicle was added last, so a separate crew vehicle was required.

Wait Vostok-derived lunar flyby?? Probably talking about the early Soyuz-derived lunar flyby using the Soyuz-A-B-V design, where two crewed Soyuz capsules would go up to meet a Soyuz-based tug that would have already been refueled by several subsequent R-7 missions. The two capsules would have had some EVA transfers while assembling the whole circumlunar stack. I'm guessing @SunlitZelkova means EVA transfer for a lunar mission architecture. Weren't there at least some EVA transfers during Skylab and during the construction of the ISS? There was an EVA transfer proposed (in hindsight) as a rescue mission for Columbia, but obviously that wasn't operational.

I think there's likely a top port for fuel transfer, then the side port for crew transfer See, the trick is that they don their spacesuits and swim through the liquid hydrogen. They needed the spacesuits anyway! The best part is no part! Brilliant! If they had a four-engine cluster at the center instead of a five-engine cluster, they could do "sweep" the surface on touchdown by vectoring two engines inward and two engines outward, then reversing. I suppose they could do the same thing with the five-engine cluster since the central engine is going to be doing the work of dusting off the center anyway. Still, it does make a true surface-level egress impossible because you need vertical space for the engines (just less vertical space than a single engine of comparable thrust).

I suppose cranes are easier to manage on the moon thanks to the low gravity. But still, geez. I'm guessing there are some fairly straightforward maths that relate the amount of nozzle exit area under a lander to its carrying capacity....

Any indication of how the cargo offload would actually work? Where does the cargo go, anyway? For a one-way hydrolox moon vehicle, I think it makes a lot of sense to have the fluffy hydrogen tank on top, the LOX tanks on the sides, and 4-6 smaller engines mounted on or around the LOX tanks to straddle a large central cargo bay at ground level. Something like this: But of course NASA is in love with a single central thrust path, which I suppose makes sense for structural reasons as well as heritage.

Good examples, but arguably those are things which would be one-way launches, which obviates the "up and down" element needed for the whole manufacturing economy of scale to operate. One possibility would be higher-capacity microchips, or even quantum computing chips. Moore's Law is reaching its limits and will soon be obsolete (some are saying it already is). Once we have lower-cost access to space, I'm sure that some of these tech companies will be itching to experiment with ways that microgravity can increase the density or function of integrated circuit chips. With crystal growth in microgravity, perhaps we could see more true monolithic 3DIC chips, where transistors are "grown" on a three-dimensional lattice to achieve exponentially greater processing speed with a smaller footprint and reduced power consumption. That could be one of the enabling technologies that pushes manufacturing offworld, since the demand for higher transistor speeds scales in a way that niche/bespoke things like organ manufacturing do not. Another potential avenue could be some sort of yet-unimagined solar powered fuel production system. We have a lot of technologies that depend on liquid-based fossil fuels and internal combustion engines, and there appear to be limits to battery density. What if we could send up a few hundred tonnes of waste plastic with some sort of solar-powered bacteria or bacteria-fungal symbiote that, once exposed to sunlight and microgravity, could "eat" through the plastic and produce hydrocarbons or some other hydrogen-dense fuel at scale, creating its own lattice structure to continually absorb more and more sunlight?

They are really maximizing the utilization of space by having the big LH2 tank up top. If you recall from the Artemis thread, the volumetric utilization for the old Altair lander design was awful, on the order of 19%. Doing this really big lightweight LH2 tank up top takes maximum advantage of the square-cube law so that you can fit a LOT of liquid hydrogen into a relatively small volume.

Oh, certainly. Obviously none of these things can be done, not at this point. I was just musing about what omelets could have been made if we had been willing to break a few eggs. The comanifested volume is one of the big issues. The Altair lander would have been 10 meters wide and 15 meters high. Of course the descent stage would have been hydrolox which drove up volume, but even with a denser propellant choice I don't see how you can fit a lander underneath Orion on SLS. Revisiting this quickly just for general edification. The Altair lander was a two-stage design with a hypergolic ascent stage using an AJ-10, a cryogenic stage using a pressure-fed deep-throttling RL-10 for lunar orbit insertion and lunar surface descent, and a separate airlock and cargo capability on the descent stage: Total mass launched to TLI would have been 45.9 tonnes, plus the 26.5 tonnes of Orion. The pressure-fed version of the RL10 was built and test-fired as the CECE demonstrator and was capable of throttling down to 8% with a max specific impulse of 445 seconds. To brake the entire stack into low lunar orbit, Altair would have needed to burn 13.5 tonnes of hydrolox. Orion would have detached, leaving the weight of the sortie vehicle at 32.4 tonnes; it would need to burn another 11.3 tonnes of hydrolox to get down to the lunar surface. So in total, Altair would have had a propellant capacity of roughly 24.8 tonnes of hydrolox. Schematically, the Altair lander had a central CECE RL10 with a height of 1.53 meters and a diameter of roughly 1 meter (allowing for gimbal); it also shrouded the ascent vehicle's AJ-10 engine. Thus the hydrolox all would have been in the octagonal outer envelope surrounding these two engines: The RL10 has a mixture ratio of 5.88:1, meaning that those 24.8 tonnes of hydrolox include 3.6 tonnes of liquid hydrogen, occupying 50.71 cubic meters, and 21.2 tonnes of LOX, occupying 18.58 cubic meters. The portion of the vehicle containing the tanks can be approximated as a right toroidal cylinder with an external diameter of 10 meters, an internal diameter of 3.33 meters, and a height on the order of 5.2 meters (based on some rough pixel counting). In total that's a "tank carrying volume" of 363 cubic meters, giving a volumetric efficiency of ~19.1% (not unexpected given how fluffy hydrogen is and how suboptimally the tanks need to be arranged here). So if we somehow had a beefed-up SLS capable of launching an arbitrary payload to TLI, what could we fit in the 10-meter-high, 8.4-meter-wide space under Orion on top of the EUS? Let's start by making this a slimmed-down sortie lander, assuming pre-emplaced surface assets. First let's get a better grip on Altair. NASA's assumptions about an Artemis ascent stage assume 9-12 tonnes, including extra props to get to NRHO, and we know Altair was bloated, so let's take Altair's ascent vehicle (with props) as 9.5 tonnes. This means the dry mass of the descent stage was 11.6 tonnes. Subtract the airlock and we save around 1.5 tonnes, dropping the descent stage to 10.1 tonnes dry mass. Remove the 350-kg engine and the expected 500 kg of unpressurized payload to the surface, and we get ~9.2 tonnes for bare structure, landing legs, and tankage weight. So the structural-and-tankage ratio here is a disappointing 2.7:1. These tanks are roughly the same diameter as the tanks on Centaur (which boasts a propellant fraction of 10:1) and held 20% more props, but these would also have been pressure-fed tanks with five times as many hemispheric bulkheads, so I'd ballpark the tankage weight on Altair at 2.5x that of Centaur or 5.2 tonnes. This suggests that the load-bearing structure, RCS propellant, RCS thrusters, radiators, landing legs, and so forth all come in at 4 tonnes. What would an Altair built for SLS look like, then? Well, the minimal-mass ascent vehicle would come in at 9 tonnes...except that we only need to go to LLO, not all the way to NRHO, so we only need 1,870 m/s instead of 2,600 m/s. At 316 seconds of specific impulse, a 9-tonne ascent vehicle would burn 5.12 tonnes of propellant to develop 2,600 m/s of Δv, whereas we will only need ~3.23 tonnes of props, meaning our ascent vehicle need only have a wet mass of 7.11 tonnes. Let's borrow @Exoscientist's idea and use the old discontinued Standard-sized Cygnus as our pressure vessel OML, making it 5.75 meters high including the engine (3.65 meters without). Let's ditch any extensible solar panels and have it run on batteries and fixed panels after detaching from the descent module. Since we are using hypergols, we can fit our 2.2 tonnes of dinitrogen tetroxide in a pair of 1.2-meter external spherical tanks placed as far forward as possible, and fit our 1.3 tonnes of hydrazine (let's take extra for monoprop RCS) in four 0.9-meter spherical tanks clustered underneath: This gives us a good estimate for the absolute maximum amount of volume (purple) we can allocate to a lander and descent stage. The cylindrical region below the engine bell has a volume of approximately 188 cubic meters, the toroidal region surrounding the engine bell has a volume of approximately 95 cubic meters, and the available volume on either side of the ascent vehicle (leaving the front and back open for both thrust balance and egress) is roughly 21 cubic meters, for a total available structural volume of 304 cubic meters. Of course the descent vehicle needs engines. To conserve space (and perhaps utilize a little of the wasted volume around the top of the EUS tank), let's use four BE-7 engines (two would do it, but we need redundancy). With gimbal allowance, each of those takes up a height of 80" and a diameter of 47.4", but since we can let them hang down slightly around the top of the EUS tank we only end up losing about one cubic meter of volume for each engine, bringing us to ~300 cubic meters of structural volume. Setting volumetric efficiency at 19.1% as before, this gives us 57.3 cubic meters of usable tank volume. Bulk density of hydrolox is on the order of 0.365 g/cc, so that's just under 21 tonnes of propellant. Tankage weight will be a little better than Altair since we aren't using pressure-fed engines -- let's say 3.7 tonnes. If the load-bearing structure of Altair was 4 tonnes for 363 cubic meters of structural space, then we can ballpark the load-structure here at 3.3 tonnes. All four engines together probably come to something like 300 kg. So we have a descent stage with 21 tonnes of propellant and a dry mass of 7.3 tonnes, and an ascent stage with a total wet mass of 7.1 tonnes. So let's throw this to TLI on a magically-beefed-up SLS and see what happens. With this approach, Orion doesn't need quite as much propellant since it will be braked into LLO by the lower stage. Orion ordinarily carries 8.6 tonnes of props, but we can drain about 2.5 tonnes out and still have enough for the return voyage (there really won't be a meaningful decrease in dry mass with this approach) to make it lighter. Total injected mass to the moon is 59.4 tonnes. Ballparking the BE-7 at 449 seconds specific impulse, the lower stage burns through 11.1 tonnes of hydrolox to brake itself and Orion into low lunar orbit. Orion detaches, leaving the lander at a wet mass of 24.3 tonnes with just 9.9 tonnes of hydrolox remaining. Fortunately, that gives it 2.3 km/s of Δv, which is sufficient margin for reaching the lunar surface. We can probably shave a total of 2.5 tonnes of hydrolox off and still make it work, reducing our lunar injection mass to 56.4 tonnes. Cramped, but doable...if it was possible to make an SLS Block 2+ with 10-15 tonnes more capability than SLS Block 2, which is already never going to happen.

Oh, certainly. Obviously none of these things can be done, not at this point. I was just musing about what omelets could have been made if we had been willing to break a few eggs. The comanifested volume is one of the big issues. The Altair lander would have been 10 meters wide and 15 meters high. Of course the descent stage would have been hydrolox which drove up volume, but even with a denser propellant choice I don't see how you can fit a lander underneath Orion on SLS.