... So what you're saying, in essence, is that focusing solely on the N series rockets early on, developing two series within that family, the N1 and N2, and combining those two (the N2 "Block A" becoming the N1's "Block B") would avoid the problems that caused engine failures and pogoing on OTL N1, allowing for earlier development?
No. Earlier development, the product of a decision in the early 1960s to go with Korolev's plan for the N pair (and there might possibly even be an N-3, taking the V block of an N-1 (Cyrillic alphabet, V is the third letter) to make something Soyuz-launcher size would logically begin with a focus on developing the N2 first. I'd like to hope in TLs where the Russians do things better are not crazy ASB, but there is no reason here to think they do things a lot better procedurally than they did OTL. But an N2, particularly one not so ambitious as one made from the OTL stretched 95 tonne to LEO N1 Korolev promised later in the 1960s, but derived from a more modest but hence more realistically attainable early decade draft design, is much smaller than an N1, and a much less ambitious push. Mishin did all kinds of kludgy expedients on N1 OTL to get its weight down and its fuel supply up. He replaced metal helium pressurant tanks with plastic ones, he stripped out most of the telemetry(!) I don't know if you think of the N1 Blok A first stage as 24 or 30 engine--I think of it as 24 because the outer ring of that many engines is all that Korolev originally designed in, and the central hexagon of 6 more were not added until pretty late in the 60s--part of Mishin's desperate struggle to get the N1 up to the promised target of LEO and hence translunar capability.
So--if you have more relaxed targets for performance, for a smaller rocket (one basically in the same class as Proton) that only has, I forget which, either 4 or 8 of the engines that ultimately are going to go on the N1 first stage in much larger numbers, and you proceed in documented Soviet fashion to make items with indifferent quality control and hastily throw together a working rocket for testing rather than wait on the outcomes of years of ground testing the engine, I'm thinking that the first couple tests are going to be surefire failures, because we don't even have the engines tested in advanced mode, and they have their own teething and tantrums to go through before we even begin to isolate the unknown unknowns of the stage itself, or put better numbers on the known unknowns. But beyond basic engine reliability American practice would have had down much closer to pat, at the price of delaying installation in the first test article until years of such testing had elapsed already, and the sort of stage related shaking down American practice would be getting to after proving the engines were well known and the Soviets throw into one big exciting test, we at least here have a robust stage design that does not monkey around with basic standards--not yet anyway; they might finally get the N2 to launch, all stages operating OK, we might then find the payload is smaller than estimated. Perhaps then the designers would turn to expedients like plastic helium tanks--hopefully not ripping out the telemetry instruments. Or they might just accept the somewhat lower performance figures, and hope maybe to make up for it later with new editions of the rocket engines (at least 2, perhaps 3, generations of the engines meant for N1 were developed OTL) or some other future trick.
I'm saying that if the N series is go decision happened early in the 60s, and Korolev was immediately given the resources and authorization to get cracking on N2 ASAP, this series of test article N2 rockets might have flown, with the early ones blowing up at some stage or other, way before 1969. Perhaps '66? Maybe '67. By the beginning of 1968 at the latest (I guess--ask a real engineer or professional student of Soviet rocketry if that is fast, slow or right on the dot) the N2 itself is operational.
Nothing
avoids the problems that caused the N1 to crash OTL. What it might do is
expose those problems so someone could start working on fixing it. The pogo itself is said, to be a question of the installation on the Blok A. That won't be tested until a Blok A for N1 is ready to be tested. But other problems that perhaps fed into the OTL failures might show up much earlier and be solved.
Then the next round is to develop and test the Blok A. Perhaps, if there is no pressing target date for this one to be shaken down by, just maybe they borrow the von braun method of a test tank launch instead of a whole expensive N2 stack on top. At this point the engines should be shaken down already, given they are the same ones used on what is now the Blok B. (In a sense that means that when they take the final step, and mate the modified N2 atop the tested Blok A, the engines on B are a bit different and never flight tested before, because they will have different nozzles for superior vacuum performance. However even given Soviet OTL practices of cursory ground testing, they will have undergone some testing in their current form I think, and anyway the engine as a whole is the same core as the well tested one, just with a different nozzle.
Testing the Blok A is the first time that we take not 4 or 8 (whichever number B was supposed to have) of the first stage engines but 24 of them, mounted in a circle on the bottom of the spherical Blok A oxygen tank. The individual mounting of each Blok B engine on the N2 version would be quite similar to be sure, to a similar, though smaller, spherical oxygen tank. But with 24 instead of at the most 8 engines being fed by a kerosene manifold reaching around the oxygen tank somehow, perhaps we will encounter some kind of resonance phenomena? The stage should be static fired before a flight test, and I think this is one thing the Soviets took a lot of shortcuts on.
But again--the OTL tests were done with 30, not 24, engines, with an additional cluster of 6 in a hexagon in the center bottom of the sphere, and for those to be fed kerosene fuel they'd need either a pipe running through the LOX sphere (could be done, but I don't think it would be; the fuel would be running through a cryogenic volume and would have to be insulated making the pipe through the sphere thicker, and the whole idea violates the integrity of the sphere) or else an even more elaborate and lengthy reacharound pipe system is needed to supply this island of engines with fuel.
And I can't swear to it but I seem to recall all detailed discussion agreed these central added six engines were in fact involved in the anomalies and shutdown orders during at least one of the launches.
The kicker that finally gets me frothing mad about them is that they were installed to only burn for 30 seconds or so!
Getting rid of them will not guarantee other related problems will not need to be fixed. Clearly the KORD control program needed to be much more carefully written and the sensor hardware debugged. Though I seem to recall that on the second attempt, some vibration--IIRC this was in fact a surge following the sudden (but programmed) shutdown of the central engine--hammered the other 24 engines hydraulically through the fuel lines and this caused some of the turbomachinery to shatter, resulting in little bits of one engine being sucked in the fuel lines into the intakes of another one's turbopump. Clearly if stuff like that is happening, there are all sorts of unconsidered issues to be addressed. If this sequence is correct, then of course the KORD program would have no choice but to shut down all the engines anyway. That it was triggered by the shutdown of the central engines says two things--one, the central engines were a pesky pain, but two, the design engineers did not correctly think through the hydraulic consequences of their engine shutdown methods. Suppose a Blok A with 24 engines had passed several launch tests with no launch anomalies but then on the fourth launch, one engine fails suddenly--this does little harm since the manner of failure involves let us say a fuel line that starts to leak and the pressure drops gradually. And then the KORD computer does its job and orders the shutdown of the opposite engine. The boost is far along enough that losing that thrust, being down to 11/12 of nominal, is not a problem. Everything would be fine--but the programmed emergency shutdown of the opposite engine, which recall had nothing wrong with it itself, happens too abruptly, the hydraulic shock knocks out the two engines next to it, KORD shuts the opposites down (the ones flanking the original failed engine) and now we are down six engines, in two patches, 1/4 of the total--now the numbers are worrying, perhaps already past the launch failure red line anyway. Besides, thanks to the wrongly designed shutdown sequence, the cascade continues and I need not go on. The N1 certainly will not!
Actually I suspect N1 Blok A involved some programmed shudown of some of the 24 engines anyway, in lieu of designing them to throttle. Then again I believe they were designed to throttle somewhat because they were not gimbal mounted--yaw and pitch were to be accomplished by varying the thrust on opposite sides?
I'm not sure of that--and I think what I would do is design them to gimbal in one dimension, tangent to the circumference instead, and that would save having to design in any throttle at all. Having both gimbal and throttle would be best of course, and there would be considerable advantage to avoiding even single dimension gimbaling. Details, details! Which might be tweaked to be other than OTL by the way.
So to sum it up--OTL Korlev and Mishin played wait and hurry up, proposing what I think was a decent though ambitious design quite early in the decade, then sat around waiting for a green light they didn't get for years, at which point all of a sudden Korolev was on the spot to promise a single launch for a complete Moon landing and return mission, which demanded he stretch a decent if dusty design to the point of fingers crossed marginality, then rush to get it built ASAP all of a sudden, with a bunch of weight saving expedients thrown in at the last minute, and every component of it gets tested for the first time on the first all up test launch. No surprise that fails!
But if instead Korolev is working on the smaller N2 design, with no last minute rushed in weight saving kludges, everything proceeding methodically, then probably given their reluctance to test separate components for years in the American manner, or pay for all up static tests, they lose several N2, but this happens in 1965 or '66, and eventually they figure out the problems. This filters out
most but probably not all the problems that crashed N1 OTL, and in addition shows, via more launch failures, and solves upper stage problems in the V, G, and if tested D bloks as well that the program never got around to OTL. Say this process takes two years, and the shaking down of A blok takes a year and a half before they have complete confidence, and N2 testing starts in 1967 early--this means they have a functional N2, suitable for scheduling even crewed launches on, in early 1969. It is approximately Proton class but they have just missed the boat on beating Apollo 8 to get the first crew near the Moon. And in the summer, the Blok A is still buggy when Apollo 11 launches, so the Soviets must concede the Kennedy unilateral Moon challlenge at every step. But--in early 1971, they have an N1 that is not capable of 95 tons to orbit, but is capable of say 75. And they have a Proton class ker-lox launcher, and might even consider downsizing an N2 to form an "N3" using the V blok of an N1 as the first stage, with G blok as second and D blok as third if needed, for a Soyuz class launcher replacement.
At no point does any OTL known problem or God knows how many further gremlins were lurking in the works get wished away, and I assume above they do not adopt ATL superior QC either. The earlier start, more robust basic design going with more modest goals, and drawn out sequence of first N2 then N1 filter out most problems--in the sense of they expose them via catastrophic failure, and then they are solved--before it comes down to testing the massive Blok A stage.
In your opinion would the N1/N2 combo rocket still remain all kerosene, with multiple launches to LEO for assembling the Lunar Expeditionary Complex?
Yes and no. Yes to kerlox, for the first version of N1 ready in a 1969-72 time horizon, No to LEO assembly. Or at least, Earth Orbit Rendezvous is an option but not necessarily the best way to use multiple launch approaches. Korolev kidded himself into thinking they could do a kerosene engine TLI starting with 95 tonnes in LEO and the remnant headed for the moon, comprised of a Soyuz, a tiny little LK lander and a D blok ker lox stage that would first brake the combination into LLO and then brake the LK by itself down to near landing speeds as a crasher stage. The LK would land, the first and possibly last cosmonaut on the Moon gets out, plants a red flag, then gets back in and after not doing a lot more than that, just some small moonrock samples, gets in and launches back to the Soyuz, spacewalks back into it with the rock samples in hand, and they fly the Soyuz home to Earth. LOR like Apollo, like Apollo from a single launch--unlike Apollo, the mass margins are incredibly tight though.
Let us guess that while this is incredibly dangerous and marginal to do with 95 tonnes, it is not so crazy with more tonnes in LEO to start with--lets say, 120 would be just fine. Does it not follow then than if 120 tonnes can be evenly divided into 60, that two 60 tonne launches might also get it done? There is more than one way to combine two 60 tonne launches than docking them in LEO though. As it happened, if we assign a single 60 tonne launch for the LK alone, sending it unmanned from LEO to be braked into LLO around the Moon by a single Blok D devoted only to the LK, and then use a second 60 tonne LEO package to send a crewed Soyuz with its own D Blok stage to brake it and perhaps with leftover propellant from that, later assist in sending the Soyuz back before it switches to its internal hypergolic fuel, and rendezvous with the LK in LLO, we have two launch LOR with no EOR involved. Well, what if the first generation N1 is capable of putting not 60 but 75 tonnes into LEO? Now each of the two launches has 15 more tonnes of margin to play with. We can make the LK much bigger perhaps; with the Soyuz quite able to rely on internal hypergolic fuel to get out of LLO and back to Earth we can use ample residual fuel from its Blok D to add to the crasher braking function of the LK's own Blok D, either by maneuvering to dock the Soyuz Blok D in sequence ahead of the LK blok D, or perhaps by transferring propellant from the former to fill the latter. Thus a bigger, far less marginal, LK, perhaps one suited for two cosmonauts and several days stay, can be accommodated.
This still may not be good enough; the Americans just did Apollo which accomplished the same thing so the USSR is saying in effect "Me too! Second!" They can afford to go ahead and do that, but they might prefer to back off and wait until they can do something spectacularly better on the Moon.
But they may not have to wait until they have the hydrogen upper stages. What could be done with three launches? Maybe now is the time for EOR after all, instead of sending ships in sequence to hunt each other down around the Moon, assemble them in LEO. Maybe the crewed Soyuz goes independently on the third launch, but the first two involve sending two 75 ton "pieces" to a LEO assembly point, perhaps at an established Salyut type station, or near one--the cosmonauts there observe the docking of the two from a distance of some kilometers through telescopes perhaps. Say the first launch is nothing but a supersized Block G--doubled up on engines, and incorporating the whole 75 tonne mass in its tanks, with a docking port designed to go into the tail of the other one which is a supersized Block D plus a big LK, essentially, big enough for 3 cosmonauts for several weeks, say an entire Lunar day night cycle--a month! This goes up second, since LOX boil off can be controlled or even prevented in LEO, and once they dock, the assembly is rocketed off, not to LLO, but to Earth-moon Lagrange 1, between the Earth and the Moon. The Soyuz also goes there, which would save a lot of delta-V versus going to the Moon, and so it has extra fuel to help replenish the lander's giant Blok D crasher. The lander fuel supply is adequate to get, not the entire lander, but a small 3 crew flight capsule, back up to L1--the trip might take some days with minimal energy, to be sure. The descent to the Moon is direct from L1, as is the ascent back to it, no Lunar parking orbit. Because of the metastability of the L1 location the Soyuz can be left uncrewed while all three cosmonauts descend to the Moon.
I have not done the math on any of this and will not have time to until the weekend, but my point is, with 3x75 tonnes in LEO, one can really do quite a lot even with nothing better than kerosene.
But, of course, at that point the Soviets will surely feel the pressure to develop some sort of hydrogen engines themselves, with American designs so badly eclipsing theirs. To be sure, the Soviet ker-lox engines were quite good of their type, getting considerably better Isp than American ones would, by some 10 percent or so, though of course no Soviet single chamber rocket engine came to matching the F-1 for sheer thrust. Developing hydrogen versions of the G blok, and possibly V blok, should significantly improve the TLI mass ratios, and perhaps the LEO orbited mass will be raised too. Say this part takes until 1980 or so, they still are not dead in the water and can do something impressive with their vintage 1970 ker-lox version while developing the hydrogen stages. And really, it need not take all the way to 1980, that's just an outside worst case figure.
...I wonder if the L3M spacecraft would have enough delta-v to both land on the moon and return to Earth? That is, would a single-stage lander with the return capsule inside as above in the first post, be feasible?...
I think anything can be made feasible if you throw enough development money at it and stick to it, but I have to ask, why do it that way?
It is because of the long stay of course. LOR is far more efficient. Yes, it should be possible to launch a Soyuz landing capsule direct from the Lunar surface to a Trans-Earth trajectory that is nearly on target for an Earth atmospheric reentry. One had better not send the capsule all by itself, never mind that three or even two crew are cramped in a capsule designed for minimal habitability. Remember the Soyuz design includes a light but large Orbital Module the crew can enter from their return capsule, to give them room to live and breathe and not go crazy while in orbit. And it includes a service module, to provide for orbital delta V and course correction. If whatever stage launches a minimal return capsule back to Earth is adequate, the equivalent service module on the return leg can be pretty small, much smaller than the rather large standard Soyuz, as far as reaction mass and engines go--but it had better have that as midcourse corrections are just about certain to be needed. Also I believe the Soyuz service module like the Apollo one provided all sorts of utility support, air supply, electricity, etc to the crew capsule. So you can't get away with just launching a 2 or 3 tonne return capsule from the Lunar surface, you must have some tonnes of support systems too. I'm going to guess that getting the total down to 6 tonnes will be painfully spartan and risky. Let's say it really has to be 8 tonnes all up. Well, now, all of that mass must be landed on the Moon. And the actual Moon exploration and basing equipment is not included in it, that's all auxiliary. Oh yes, we must land enough propellant and engines to propel the 8 tonne return vehicle back to Earth. Will 25 tonnes landed be quite enough? I am not sure. But the design would look different from your pictures anyway.
Now if only LOR were possible! If there is a Soyuz, braked to LLO by a tailored blok D engine/tank that is now spent and has been crashed to the moon, its internal tanks are full, and has plenty of delta V to head back to Earth with plenty of course correction margin, and it has the extra habitable space of an Orbital Module so three crew will be fine for the time it takes.
Now instead of launching some 8 tonnes to get two or three cosmonauts home directly from the surface, we would only need based on Apollo experience, at the most 2 tonnes, plus propellant to launch it, but not all the way back to Earth, just to LLO and rendezvous with the Soyuz. Possibly a minimal life support capsule for 3 can be cut down way below 2 tonnes, so we are looking at an overall reduction of a system I believe would mass 16 tonnes at least to get crew home in an entry survivable capsule directly, down to no more than 4 tonnes tops, meaning if we were capable of landing the full 16 tonne direct return package, but we only need 4 for LOR, then we have freed up 12 tonnes of extra Moon exploration supplies and equipment, or we have reduced the mass of the package to be landed on the Moon by those 12 tonnes. Or more, I think I underestimate the minimal return requirements.
The catch is, of course, that leaving one's return ride to Earth in LLO is one thing when the stay on the surface is just a few days. But over time the return vehicle in orbit will drift into an orbit unsuitable for the surface vehicle to reach. Even if one knows that eventually the orbital vehicle will return to a useful inclination, there is no ability to abort and return any time; one has to wait for the return vehicle to drift back into a suitable window to be reached and to be aimed at a suitable TEI path.
But there is more than one way to approach this. For one thing, an equatorial mission can allow a parking orbit that aside from drift caused by mascons, will stay suitable for both purposes indefinitely.
For another, this is why I introduced L-1 into this. L-2 is more fashionable and might be what serious designers ultimately do settle on, but conceptually I think L-1 should get some attention. It is closer to Earth than the Moon rather than farther away. I believe a big advantage of L2 is that by swinging close past the moon and firing at perilune, one can reach it with especially little delta v--but the approach to L-1 should be pretty easy too. L1, between Earth and Moon, is a good point to channel communications from Nearside to Earth; an L-2 based mission that went to Near side would have to set up a relay station at L-1 anyway.
Staging the Soyuz, or perhaps some more ambitious developed surface to orbit vehicle, at L-1, the transfer from L-1 to the Nearside surface is more difficult, more energetic than from LLO. A vehicle to return to L1 from the surface would have to mass more than 4 tonnes I think, because for one thing the rocket delta V requirement is somewhat greater and for another, the time of transit is far longer. From surface to LLO is a matter of an hour or less; from surface to L1 a matter of days. The return capsule better be pretty habitable! In fact I envision it being the same thing as the major module they live in during their long stay. So it alone may mass say three tonnes, and the whole thing at a guess masses 9 or 10. Say 10 tonnes all up--that's still a lot less than 16, and the crew is a lot happier! From L1, having docked this habitation module to the Soyuz, a standard Soyuz might have plenty of delta V to shove not only itself but the whole hab module back to Earth--and if so, they keep its familiar comforts all the way back, until it is time to enter the entry capsule and abandon it and the service module, much like Apollo 13. With a mission model like that, perhaps the Soyuz they left Earth in had the habitation module instead of the standard Soyuz orbital module--it would weigh extra and require a more powerful emergency launch escape system but why not, the Soyuz goes up an N1!
Direct from Luna to Earth is possible but probably very dumb then. I think staging through L1 is the way to go. In fact they might build a Mir type station there. Or at L2 of course. Or both!