sevenperforce

It doesn't. Steel was chosen because, among other properties, it becomes stronger at low temperatures. It has good thermal properties at high temperatures but structural properties are the important bit at low temperatures. There is one Starship base configuration. There are a number of different modifications and adaptations. Starships that will return to the surface of Earth get heat shields and flaps. Just like there is one base Soyuz configuration but some have boosters and others don't; some have third stages and others don't, etc.

We are talking about multi-pass aerobrake circularization by lunar Starship. Lunar Starship is never going to undergo re-entry. Why on earth would you think it would? Neither the enthalpy of formation nor the enthalpy of vaporization are relevant to structural weakening of a spacecraft skin due to heating. Ductile failure takes place long before you reach melting point.

IIRC they upgraded from niobium to RCC at some point in the last few years but I could be wrong. And yes, steel can handle higher temps than Al or AlLi or composites, which means we need less of a heat shield. But that's relevant to re-entry, not multi-pass aerobraking.

Indeed. But if the constructors of 1960s could use bare steel without asking for the overexpensive niobium, they would definitely do this. Since then, the steel unlikely has changed so much. If the designers in the 1960s were creating a vehicle which never had to undergo re-entry, they certainly would never have bothered with niobium.

A niobium heat shield for Starship would be vastly heavier than the tiles. The tiles are extremely lightweight. Also, heat management would still be a problem. The limiting factor for Starship is the temperature on the back of the heat shield, where it radiates to the steel. Niobium radiates and conducts heat very well, which makes it great for radiatively-cooled engine nozzles in a vacuum but not so great for a thin heat shield attached to steel.

And exactly zero of those projects (none of which were ever constructed, mind you) were intended to perform aerobrake circularization. The heat shields were for re-entry. This is, to date, the only vehicle to have ever performed multi-pass aerobrake circularization: You will note the conspicuous lack of a niobium heat shield anywhere.

I've never wished I lived in the Canary Islands before....

In the glorious early days of the spacenautics allof them required a niobium shield at some distance from the hull itself. Niobium was very expensive. If they could, they would never want it. It would be understandable if the Starrship was made of a post-industrial nanosilicarbon, not of same steel which needed niobium a half-century earlier... What on god's green earth are you on about? No vehicles have ever performed multi-pass aerobrake circularization into LEO. There has never been any niobium-shielded stainless steel re-entry vehicle. The only vehicle to perform multi-bass aerobrake circularization was the Mars Reconnaissance Orbiter, which did it at Mars, and quite famously had no heat shield at all. Well @tater it looks like you were right and I was wrong. I couldn't think of any reason that the OLT would need to have QDs to the individual engines, since the Raptor igniter is so small, but I forgot about the need for spin-up gas.

I think the challenge is not re-radiation of absorbed heat, but peak heating during the passes. The highest-velocity passes produce the highest dV savings but also have the highest peak heating.

Placement of the depot is anticipated to be LEO, right? It’s a little tricky, I think. It’s a shame to burn propellant to move a heat shield and flaps and such into lunar orbit when you don’t need them there. But a heat shield and flaps and such are the best way to get back into LEO. You want to spend your propellant in as efficient a way as possible. In the long-term, what if they put the fuel depot into an Earth-Moon cycler trajectory? Each tanker mission would need to boost to TLI in order to match the fuel depot’s trajectory, but then could return for EDL easily. The depot would perform the NRHO insertion, which it can do more efficiently than the tankers because it has less mass to contend with. And then once (nearly) empty, it could perform Earth Interface Injection, returning it to the cycler trajectory but never needing to brake back into LEO propulsively.

DART is set to impact at something like 6.6 km/s. Recent asteroid missions have shown them to be perhaps loosely bound conglomerates, so the result of the impact in terms of dv delivered might be expected to be whatever it is (cm/s?), the impact can also send a bunch of material out into a cloud of ejecta... should be interesting. Indeed. An inelastic collision may be simple enough…as long as it remains a two-body problem. But a three-body problem has only discrete solutions. And a 999,999-body problem? Well that’s a new bird entirely.

EDIT: I ran the numbers. An 84-tonne lunar Starship requires 5.79 km/s for a landing demonstration mission if it skips NRHO. That’s 333 tonnes of props. Lunar Starship’s LEO residuals plus a single tanker mission will get it to 321 tonnes of props, so if it can shave off some added mass (perhaps by only carrying a mock-up crew capsule and no cargo) it’s doable.

Moving this discussion with @tater to the correct thread…. Suppose we have an 84-tonne lunar Starship that leaves LEO, picks up crew from Orion in NRHO, goes to the lunar surface, returns crew to Orion, and then returns propulsively to GTO (that's the max burnout mass we can achieve without a tank stretch). If a Starship tanker (which we will notionally say has a dry mass of 85 tonnes and carries 30 tonnes of landing propellant) can deliver 150 tonnes of propellant to LEO, then it can deliver 33 tonnes of propellant to the lunar Starship (assuming a 2.27 km/s GTO burn at 380 s isp). This gives our lunar Starship, in turn, 1.2 km/s of dV, which it can use to lower its apogee significantly. The next tanker will only need to burn 1.07 km/s out of LEO to meet it, meaning it arrives with 86 tonnes of propellant, giving the lunar Starship enough propellant to reach LEO with 43 tonnes of remaining residuals. It will need just under eight tanker missions to refill its tanks for the next sortie. On its initial launch, it would have reached LEO with 171 tonnes of residuals, requiring less than 7 tanker missions to refill. So you need a total of 10 launches for the first mission (1 lunar Starship + 7 tanker missions for the outbound journey + 2 tanker missions for the return journey) and you need 10 launches for each subsequent mission (8 tanker missions for the outbound journey + 2 tanker missions for the return journey). On the other hand, if you are only going to make it a one-way trip (LEO to NRHO to the lunar surface to NRHO), that same 84-tonne lunar Starship will need only 7.8 km/s of dV, meaning it needs only 598 tonnes of propellant in LEO. Accounting for the 171 tonnes of residuals that lunar Starship will have when it reaches LEO on its initial launch, you only need 3 tanker missions and you'll have 127 m/s of extra margin...hopefully enough to get the now-derelict spacecraft clear of the Lunar Gateway. So that's 10 launches per mission for a reusable lunar Starship or 4 launches per mission for an expendable lunar Starship. This all assumes the minimum dry mass of a naked Starship is something like 45 tonnes (based on Elon saying that a naked Starship without SL Raptors is 40 tonnes) and that ~39 additional tonnes is enough mass budget for landing legs, paint, solar panels, landing engines, landing propellant, crew capsule, fairing, and cargo. A landing demonstration mission with no NRHO stopover and no lunar ascent (except perhaps a hop) could probably get away with only 1-2 tanker missions, though I haven’t run the numbers.

I still see propulsive return to LEO as a pretty long pole. If it weren’t for stupid Orion hanging out in NRHO, it would be easier, but the NHRO layovers make the whole trip a whopping 12.5 km/s. At 380 seconds for the RVacs (not accounting for any of the SL vac burns at lower efficiency), the required propellant fraction is 96.6%. Starship’s dry mass would need to drop to 42 tonnes to do that, with no payload at all. A tank stretch would only provide incremental improvements. A propulsive return to GTO, where Starship tankers could easily meet it and give it the additional props to return to LEO, is a little easier. You’d need 10.2 km/s, which comes to a propellant fraction of 93.5%, giving you an allowable dry mass of 84 tonnes, which is closer to something useful. Agreed.

Elon has claimed that it will only take 8 refueling launches max, citing 150 tonnes of residuals per launch and speculating that it may even take fewer launches if Lunar Starship is lightweight enough (lacking flaps and a heat shield) to make the trip without full tanks. Indeed, if they can get the dry mass of Lunar Starship down to 60 tonnes and if we assume 40 tonnes of crew cabin and payload, it can do the trip from LEO to NRHO to the lunar surface and back to NRHO starting in LEO with only 970 tonnes of residuals in its tanks. With that approach and 150 tonnes of residuals delivered per launch, it would only need 5-6 refueling launches, since it would reach LEO with 135 tonnes of residuals for its own launch. But nevertheless, I was citing the “12-15” number because that’s what Blue Origin was claiming and so that’s the argument they would have made to the court, whether true or no.

What would a New Armstrong configuration have looked like? The Blue Origin descent element for the National Team lander was planned to have used two BE-7 engines. But if they were proposing a single-element, integrated reusable lander, they might have gone in a different direction. The BE-3 can deep throttle to around 18%, and Bezos has talked about the closed dual expander BE-7 having similar capabilities. If we assume ad arguendo that the open-expander version of the BE-3U has similar deep-throttling ability, then it could conceivably throttle as low as 128 kN in a landing engine configuration. Padding it to around 135 kN to allow margin for hover, etc., then a single-element integrated lander using a single BE-3U as its landing engine would weigh in at around 83 tonnes at landing. For safety reasons, of course, you wouldn't want to be limited to only a single engine. But that's where Blue could borrow from the old Soviet LK lander design. The BE-7 produces 44.5 kN, so placing four of those around the BE-3U would give the lander a nice clean abort mode as well as alternative landing engines. The BE-3U engine bell is around 2.5 meters across and the BE-7 is less than a meter wide, so they could all fit easily in a quincunx within a 7-meter circle, with space for landing legs as well. A lander needs 2.6 km/s to get from NRHO to the lunar surface, and the same for the return trip. Assuming ~445 s of specific impulse for the open expander BE-3U, you'd need to leave NRHO weighing in at around 151 tonnes and you'd get back to NRHO after the mission with a burnout mass of around 45 tonnes. What do we know about the New Glenn upper stage? Well, estimates of its dry mass put it around 16-18 tonnes, and it is believed to carry around 175 tonnes of propellant. If "New Armstrong" was to be Lunar Starship knockoff based on the New Glenn upper stage, then it would be able to launch on the reusable New Glenn first stage, reaching LEO with around 17 tonnes of propellant residuals. After three New Glenn refueling launches with hydrolox propellant transfer, it would have 150 tonnes of propellant on board, enough to perform its own TLI and insertion at NRHO with 40 tonnes of residuals. With that approach, New Armstrong would need to take on about 66 tonnes of additional propellant at NRHO in order to get down to the lunar surface and back. As luck would have it, a New Glenn second stage which is fully refueled in LEO can perform its own TLI and insert at NRHO with 67 tonnes of residuals. That same New Glenn second stage would have reached LEO with 45 tonnes of residuals, requiring just three New Glenn refueling launches to be ready. Obviously there would need to be some dry mass added for maneuvering, docking, prop transfer, and the like. But I'm probably sandbagging how much propellant New Glenn can put in LEO with each launch, so I think it still closes just fine. So Blue Origin could have proposed a single-element "New Armstrong" integrated lander which would require only 8 New Glenn launches (rather than the 12-15 Superheavy launches required for Lunar Starship), based on the New Glenn upper stage. It would have about 28 tonnes for a crew cabin, additional structure, power generation, propellant management, landing legs, and lunar surface payload. And the hatch would have been 11 meters closer to the lunar surface than Starship's. Blue would argue that the 100-tonne+ capability of Lunar Starship is overkill and that cutting the number of required launches almost in half reduces complexity and schedule risk. If they did make an argument, it would definitely explain how the Court described their protest:

Agreed. The redacted portions seem interesting. There aren’t many redactions overall in the opinion. I wonder if they were hiding a “New Armstrong” architecture.

It’s hard to explain just how hilarious it is to see this damning of a statement in a memorandum opinion…. “Blue Origin has not submitted any contemporaneous documentary evidence to support its allegations of its alternative architecture.”

Some interesting excerpts: "The United States has moved to dismiss . . . under RCFC 12(b)(1) for lack of subject-matter jurisdiction. . . . The defendant argues that Blue Origin does not have standing to bring its challenge because Blue Origin did not have a substantial chance of award, even assuming its allegations were true, and Blue Origin's alternative proposal is speculative." "To establish standing under this court's bid-protest jurisdiction, a protestor must be an "interested party." The Federal Circuit has interpreted this term to require a protestor to have alleged facts, which if true, establish that it . . . possesses the requisite direct economic interest." "[T]he court must decide whether those alleged facts show the protestor was prejudiced by the alleged errors. . . . To establish prejudice for standing purposes . . . the protestor's complaint must show that there was a substantial chance it would have received the contract award but for the alleged error." "[E]ven assuming its allegations to be true, Blue Origin cannot establish that it would have had a substantial chance of award but for NASA's alleged errors." Sounds familiar. I seem to recall saying something about that: More from the opinion: "The Court finds that Blue Origin does not have standing because it did not have a substantial chance of award." "Considering the funding shortfall, even if the Court found that either NASA's evaluation of SpaceX was improper or SpaceX's proposal was unawardable, Blue Origin cannot show that it would be in a position for award. . . . NASA could not have awarded Blue Origin the contract at its proposed price or anything close to it." "Blue Origin alleges that SpaceX's proposal was not complaint with the terms of the solicitation and unawardable. In contrast, Blue Origin alleges that its own proposal was next in line for award because it was compliant and awardable. . . . This argument is fatally flawed because NASA determined that Blue Origin's submitted proposal also was not in compliance." The full memorandum opinion can be read here. Also, more details about Blue Origin's alleged, hypothetical, secret alternative plan: "Blue Origin asserts that it too would have proposed a single-element Integrated Lander, [redacted]. This alternative design would 'take full advantage of Blue Origin's [redacted].' The details of Blue Origin's alternative approach are unclear, as is the precise nature of what Blue Origin would have proposed; Blue Origin has not submitted any contemporaneous documentary evidence to support its allegations of its alternative architecture. "Despite Blue Origin's allegation that it has been [redacted], it would have had to start from scratch with NASA. Blue Origin's alternative proposal is purely speculative, including hypothetical pricing and hypothetical technical ratings. "Blue Origin is in the position of every disappointed bidder: Oh. That's what the agency wanted and liked best? If we had known, we would have instead submitted a proposal that resembled the successful offer, but we could have offered a better price and snazzier features and options. Blue Origin cannot use its speculative alternative proposal to establish that it would have had a substantial chance of award but for NASA's alleged evaluation errors." Now I want to know what the redacted part was. Were they suggesting a BE-3U lander architecture? Finally, a little more fun. "Even if it had standing and portions of its complaint were not waived, Blue Origin could not succeed on the merits. Bid protests are evaluated under the APA's standard of review. A court may grant relief only upon the finding that either 'the procurement official's decision lacked a rational basis' or 'the procurement procedure involved a violation of regulation or procedure.' "On the record before the Court . . . Blue Origin has not met its burden to show that NASA's award decision was arbitrary and capricious. The Court finds NASA's explanation to be 'a coherent and reasonable explanation' of its award decision."

I was going to do the math on this, casually, but it looks like someone already has: http://toughsf.blogspot.com/2021/10/nuclear-conversion-for-starship.html They go through the possibility of (a) adding a second bulkhead and using liquid hydrogen with NTRs in place of the RVacs, (b) the same as before, but filling the cargo space with liquid hydrogen too, (c) moving the common bulkhead to increase the methane proportion and pushing methane through simple NTRs, and (d) filling the whole thing (except header tanks) with methane and pushing that through advanced NTRs. Hydrogen is so fluffy and the mass ratio plunges so low that the hydrogen nuclear Starship can't even reach orbit. Filling the whole cargo space with hydrogen allows it to just barely reach orbit, without payload. Simple methane NTRs can approximately meet the performance of the current Starship. Advanced methane NTRs can approximately double the payload to LEO, but payloads beyond LEO are only slightly improved. They conclude that it's not worth it, and I'm inclined to agree.

Raptor already has such a deep throttle range, so the turbopumps must have a pretty large allowable speed range as well. And they run independently. So I don't see why you couldn't vary the mixture ratio already, like the F-1.

Wrap a duct around it. Raptor is a fairly decent engine for air augmentation. It has a high O/F ratio so its propellant is reasonably dense, but it has a high specific impulse so it can operate in AA mode up to a theoretical maximum of 3.7 km/s. It has a terrific T/W ratio which helps counteract the extra weight of the duct system. It's just a little oversized for air augmentation purposes. A mini-Raptor with a thrust closer to that of the Merlin 1D would work better.

The booster needs higher T/W more than it needs higher efficiency. Air-augmentation can increase static thrust a little but the thrust multiplication really only kicks in once you're high-subsonic or supersonic, and by that time Superheavy will be outside of half the atmosphere and it will have lost about a quarter of its liftoff weight. Also, you can't very well just add a duct system to Superheavy as it is. The length of an air augmentation duct needs to be on the order of its diameter in order to allow sufficient space for air-exhaust mixing and expansion. What you could do -- I suppose -- is create ram air inlets about halfway up the LOX tank with ducts that penetrate the LOX tank and open at outlets between the inner and outer ring of Raptors. Just like a boat-tail bullet, Superheavy has a region of low pressure at its base during flight, low pressure that will cause parasitic drag and suck plume recirculation up between the engines, robbing them of some of their exhaust efficiency. If you had a ram air duct pushing compressed air out there, it would mix with the engine plumes and increase efficiency, and also reduce parasitic drag. But I am guessing that the added weight and the structural issues in the LOX tank wouldn't be worth it. Yeah, the mass ratio is just so so different. I wonder what would happen if you moved the common bulkhead down and kept the SL Raptors in the middle but replaced the RVacs with methane-based NTRs. You would lost some mass ratio but would the methane nukes make up for it? A small spaceplane with air-augmented Raptors (preferably variable-mixture-ratio) would work for an SSTO. All the better if you carried extra methane and transpiration cooling, because if you did it properly the transpiration cooling system could double as a fuel injector for a shock combustion ramjet....

A definite taste of N1 with the engine-bell-hugging shield, although presumably without the doesn't-actually-work bits. I am sure @kerbiloid appreciates. I suppose some design choices are made because they are just good. I have no idea whether this applies to a Raptor engine, but I will mention that sometimes airplane engines operate close enough to the flameout margin that they turn the ignitors on in flight. It's much easier to restart a flame that partially goes out than to restart an engine after the flame is completely out. Raptor uses a blown-plasma augmented spark igniter. Well, two of them. They are quite small and screw directly into the side of the combustion chamber. There is no TEA-TEB being squirted up into the combustion chamber through the engine bell, as with the ground-started Merlin 1D. Plus, Elon says there will be additional close-out covers added to the outside of the engine skirt, so it wouldn't make sense to have ignition from the GSE. I think lacking the gimbal mount probably saves weight, but doesn't make any difference to the rest of the engine. The RBoosts have a different, simpler, higher-thrust set of guts because they don't throttle. The RVacs can definitely throttle. They have to be able to throttle in order to provide pitch and yaw by differential thrust. They have a very good oxygen-rich preburner so they should have no trouble scaling it up to make it an ORSC engine, and ORSC gives you a slightly denser fuel mix to boot. Perhaps using only a single preburner and turbopump would make the engine lighter. But then you lose three main advantages of FFSC: lack of ox/fuel hot seals, variable mixtures and efficiently-sized preburners, and gas-gas combustion efficiency.

And there's soft capture.