Henry Spencer wrote:

Right up till they got to their 30 engined N-1 Moon rocket it worked,
then it didn't work.



There was nothing intrinsically wrong with the N-1 design;

I disagree, I think it had too many motors to be reliable, and that the
decision to mount the kerosene tanks on top of the Lox tanks was
fundamentally flawed for the following reasons:
1.) One of the basic ideas that the N-1 was based upon was that of
having the ability to lose a motor or two during ascent, and have the
KORD system shut down the opposing motor to keep things balanced in
regards to thrust from the ring of motors so far from the rocket's
centerline. Good idea, but it relies upon a benign shutdown of the motor
before it suffers a catastrophic failure and damages the motors near it;
since the maximum number of motors that could be shut down was four (two
malfunctioning motors and two opposing motors shut down to compensate
for the asymmetrical thrust), and a catastrophically failing motor was
liable to damage the motors on either side of it...causing three motors
to shut down, and thereby exceeding the total of four shutdown motors
maximum- as three motors on the opposing side of the vehicle also shut
down- what you've effectively got is a rocket that will fail if it has
even one motor undergoes a destructive failure rather than a benign
shutdown, and of course this scenario actually happened during the
second launch attempt.
Further, the large number of motors that need to be manufactured for
even a modest number of launches (the Soviets made a total of around
fourteen N-1's, or at least had that many at some stage of construction
although they only launched four- that's a total of 420 motors for the
first stages alone, and a total of 588 if you count the basically
similar second (8) and third (4) stage motors) means that you can't
afford to do strenuous tests on the individual motors if you want to get
up to a reasonable launch rate, and must minimize the man-hours spent on
making each motor. Neither of those choices is going to add to the
overall reliability of an individual motor.

2.) Whereas the stacked-spheres approach to the propellant tankage was a
seemingly clever use of the different volumes need for the Lox and
kerosene, it inevitably ends up with the larger diameter Lox tank sphere
in the bottom position in all three stages of the basic booster; and
with that decision comes a real problem- if you put the Lox tank on top
(like in Atlas and Saturn V) you can put an insulated Lox feed pipe
through the interior of the kerosene tank...the pipe will begin to chill
when Lox flow starts at the beginning of the engine ignition process,
but the cold won't have much chance to migrate through the insulation
into the kerosene during the short time that the motors will be burning,
and the thermal effect on the feed pipe will cause it to shrink in
diameter from the cold; so that it might pull itself free from some of
its insulation at worst.
Now, consider what happens if you try the same thing with the Lox tank
on the bottom, like in the N-1; now you have a insulated kerosene_ feed
pipe going through the interior of the Lox tank. During the fueling
process this pipe is going to be slowly immersed in cryogenic Lox, and
it will start to supercool even if it's insulated, due to the amount of
time the propellant loading process takes. Then comes the moment of
propellant feed starting to the motors in preparation for ignition...
two things happen... first, warm kerosene comes in contact with the cold
metal of the kerosene feed pipe that is traversing the Lox tank, causing
thermal shock and rapid expansion of the pipe inside of its insulation
blanket, this could easily result in cracking of the pipe in much the
same way that sticking a hot dinner plate into cold water can. Second,
as the kerosene transmits its heat into the feed pipe, the feed pipe
transmits its cold into the kerosene, which begins to gel, choking off
the fuel flow, and causing chunks of gelled kerosene to break free and
enter the engine turbopumps, with disastrous consequences.
So on the N-1, the fuel had to flow around the exterior of the Lox tank
rather than through it; and as soon as you do that, the amount of
plumbing needed goes right through the roof. The Lox belongs on top of,
not under, the fuel supply.
I tried to figure out a way to feed a large number of motors while
minimizing plumbing weight, and assuming you're going to arrange them in
a circle, like the outer twenty-four on the N-1 were, then the only
thing I could come up with was have the propellants feed into two
toroidal pipes- one, carrying the heavier propellant (kerosene in the
case of the N-1; Lox in the case of Lox/LH2) sets just inside the ring
of motors; the other sits on top of the motors, and the motors and their
turbopumps are "plugged into" these toroidal pipes at their tops and
sides. The toroidal feed pipes are themselves connected to their
respective propellant supply tank by a single feed pipe.


it simply was
an overly-rushed and cash-starved development program, e.g. with *no*
ground test of the full first-stage propulsion system, and the
politicians' patience ran out before debugging was finished.



Note that when it came time to do a new heavy lift booster, Energia was
a far different design with far fewer motors.



The Saturn V, with somewhat more time and a lot more money, still had
serious bugs as late as Apollo 13. By luck -- quite a bit of luck in the
case of Apollo 13's second stage -- none was catastrophic.


I think that the Saturn V's good luck was partially due to it's good
engineering and straightforward and tough design; and that the N-1's bad
luck (it never even got to the end of first sage burn in four launch
attempts) was due to bad engineering and an overly complex design.





The big feed lines for a handful of large engines aren't light either.
There *is* extra plumbing mass with a many-engine cluster, but the
difference comes from second-order effects.


In the case of N-1 it was from the need to get those kerosene lines all
the way around the big Lox tanks; each of those pipes is around fifty
feet in total length from the base of the kerosene tank and the motor it
goes to, and you are dealing with thirty of them- for around 1,500 feet
total of kerosene feed pipe in the N-1's first stage.

Some of us don't think four launches per day is at all unrealistic --
although it might have required a somewhat different vehicle design --
viewed from a clean-sheet-of-paper perspective, rather than from the
stifling trap that we've gotten ourselves into in the last few decades.



Remember- four launches a day; first and second stages recovered at sea;
and a twenty shuttle fleet. In short; you have to launch one, recover
the two stages, check everything for condition, restack it, add payload,
move it to the pad, fuel and relaunch it inside of five days. On any
given day, you have four in operation, and sixteen being recovered,
refurbished, and readied for launch. That's a hell of a big
infrastructure- even with something far simpler that the Shuttle. We
were awestruck when the Soviets managed to put up two Vostoks in a 24
hour period in 1962; and the Vostok booster was a far smaller and less
complex thing than these rockets would be. We're talking about launching
things the size of Saturn V's every six hours, day in and day out, and I
think that would be almost impossible nowadays- much less in the time
frame we are discussing particularly given the rocket's reusable aspect,
and all the effort that implies in regards to recovery and re-flight
certifying the first and second stages.



Bell shares a fundamental error with most of today's Old Guard rocketry
establishment: he thinks today's incredible stupidities are laws of
nature, that the Emperor couldn't *possibly* really be standing there with
no clothes on.


Well, would you care to define a design using the technology available
in the 1945-1960 time period that would allow the building of a space
station and manned moonship? It wasn't like we were sitting on our
thumbs during that period in regards to rocketry; we were moving from
tactical missiles to IRBMs to ICBMs at a pretty good clip, as were the
Soviets. The only thing we weren't doing was pushing for manned
spaceflight immediately, as I assumed that both sides realized that the
technology was advancing at such a rate, that what would be very
expensive and difficult in 1950 would be far easier by 1955, and easier
yet by 1960.

A lot of Armstrong's problems were because he was improvising a landing in
unfamiliar terrain. And *that* was the result of, to put it bluntly, a
mistake he made earlier:


Of course, heading for the Moon with a manned expedition before you have
had a good look at it via robotic landers and photographic probes like
Surveyor and Lunar Orbiter means that the whole place won't be known in
detail until the first ship is on landing approach...and the robotic
probes and landers require fairly sophisticated electronics to perform
their missions, electronics that weren't in existence in the time frame
we are discussing.

he had his attention inside the LM looking at
the computer alarms, when he should have left those to Aldrin and kept his
mind on navigation and his eyes on the surface. Nowadays, this is a
standard lesson all pilots learn -- you must divide responsibilities, it's
a lethal mistake to have *everyone* preoccupied with troubleshooting and
nobody flying the damn plane -- but it wasn't part of the gospel then.
And the Apollo 11 crew probably wasn't all that well integrated, simply
due to shortage of training time, so Armstrong may not have been all that
confident in Aldrin.

With more eyes watching -- say, a dedicated navigator -- this would have
been much less of a problem. And with no computer, you don't have
computer alarms. :-) The only big question is whether you abort the first
landing attempt when you realize you're coming down well off course, or
try to correct. (If you abort, the *next* ship has the proper correction
cranked in ahead of time.) Probably you try to correct, because the
navigator catches the growing discrepancy early.


I'd be very concerned about your descent speed in a situation where you
can only rely on your radar return and eyes during landing. With no way
to know the scale of the surface features you are looking at, a horizon
that is far closer than on Earth, and a lack of any haziness of objects
in the distance due to atmosphere, I could see the crew of the moonship
ending up hovering at several hundred feet above the surface and using
up all their fuel thinking that they are going to touch down at any
moment, or worse hitting the surface at high speed thinking that they
are still several hundred feet up and have time to brake their descent.





Not a trivial issue, although eased considerably by assembling the lander
in space so you're not wrestling with packaging constraints too. And the
LM landing gear turned out to be drastically overbuilt.


That's because Grumman built it; the thing was designed to slam down on
a carrier deck at 100 mph, you know. :-)


What matters is not so much how big the lander is, as how well it responds
to the controls. Large size and poor control authority don't *have* to go
together.


That means big RCS motors though, and with those comes a structure
strong enough to tolerate their thrust, and the higher propellant
consumption that would go with them.

The problem was more that it was going to be as *tall* as an Atlas, and
that the constraints of launch from Earth meant that the crew were going
to be at the top.


No one ever did figure out a good solution to that conundrum, did they?


Which made for very awkward problems of adequate view
for the landing, the issue that finally sank EOR. This is much less of
a concern with a space-assembled vehicle.


Yeah, you could put the crew at the bottom, the fuel on top of them and
four outward angled rocket motors at the top. By gimbling them and
varying their thrust, you wouldn't even need an RCS system. Not that any
of those 1950's designs I've seen ever _did_ it that way of course...the
poor astronauts always face a long trip down to the lunar surface from
around a hundred feet up or so, even on the space assembled landers.




and we couldn't get
one of those to take off reliably, much less have it gently touch down
without exploding.



A problem that had very little to do with its mass.


But to work in a reasonably sized package (something smaller than a
Saturn 1), a lunar lander has to be fairly lightly built in regards to
its mass ratio, like the LM was. A lot of propellant weight in a lightly
built spacecraft structure isn't going to like a hard landing.





Getting a heat shield to take those reentry heats would have been a real
problem given the state of technology at the time, and Titanium and
Inconel metallurgy for spacecraft structure wasn't nearly as finessed as
it later became...



You don't really need high-temperature structures if your thermal
protection is good. And suitable steels make quite good high-temperature
structures -- more heat-resistant than titanium -- although they're rather
heavy.

But an adequate heatshield would indeed have been a problem; von Braun's
ideas on that aspect were naive, in hindsight.


Remember, these were the guys who put wooden wings on the A4b, and then
were surprised when one came off during reentry. Prior to 1950, our
knowledge of the upper atmosphere wasn't all that good, and we didn't
get a good handle on what a space vehicle encounters during reentry at
high speed until around 1955.

The difficulty is not so
much materials technology -- notably, there would be nothing very
difficult about making an ablative heatshield, even with WW2 technology,
so long as you weren't too worried about how much it weighed -- as the
insight that reentry bodies should be *blunt*. It took quite a while for
people to realize that; it wasn't obvious.


I'm picturing a CM sized copper heat sink heatshield to take the heat of
lunar reentry...okay we get the LM and SM up into orbit on one Saturn
V....then a second Saturn V is launched with the CM on board, after the
two S-IVB stages accelerate the separate vehicles onto a lunar
trajectory, the CM docks to the LM, and then to the SM... :-)






Assuming somebody hits on a suitable approach to the heatshield problem,
say five years for the first orbital scout flights, another five for a
heavy ferry and construction start on a station, and five more to finish
the station, fly a scout mission around the Moon, and gear up to attempt a
landing. Maybe 1960.


That's about what I thought if you threw a tremendous amount of
resources at it, and got some lucky breaks. But the cost of such a
crash program would have made Apollo look cheap.



You might be able to cut a few years off that if it's a crash program from
the start, aimed at a lunar landing soonest rather than systematic
progress while building infrastructure. That would mean (as Pat says)
tolerating both failures and loss of life, and a certain amount of
conspicuously wasted money. And perhaps a bit of luck.


This would be like making the space program the equivalent of the
Manhattan Project- absolutely ASAP, no matter what the cost. It wouldn't
have been a rational approach to the problem, and I doubt if it would
have been politically possible.



There has been quite a bit of development work on automated docking in US
labs; what is lacking is funding for flight tests, and that is closely
tied to a lack of any real requirement for it (given that all US station
flights are manned anyway).


It would have allowed us to assemble stations in orbit before a crew was
launched to them, or enlarge them without needing further manned
launches or involved EVAs. Mir made Skylab look clumsy; but having the
Saturn V's lifting capacity along with a means of docking Skylab sized
modules in orbit automatically would have given us a great station by
1980 at a comparatively knock-down cost.



Anyway, with a von Braun approach, both the ferry and the station's tug
are manned, so that's not an issue. The problem is that you have to
rethink both the station design and the moonship design to be *modular*,
so that you are plugging modules together rather than riveting girders
together. That may also require one more rev of the ferry design, to give
it a larger cargo hold -- not necessarily more cargo mass, you can outfit
the modules from within once they're connected up, but more cargo volume
so you can launch a reasonable module shell in one piece.


Of course, one of the reasons that von Braun had everything manned, was
that they didn't have sophisticated electronics to do the comparatively
mundane tasks that he had people doing instead...he points out that a
enemy missile with a sophisticated enough guidance system to intercept
his space station would be too heavy to work...and even if it could be
built, it's going to encounter the station's _cannons_...which oddly
enough, don't seem to be on those cutaways of the station that are in
Colliers.




Even in WW2, if I've got the dates straight, people understood that solar
flares produced radiation; the neutrons from upper-atmosphere particle
hits at high latitudes are detectable on the ground.


I don't think they knew the severity of the radiation till they
encountered the Van Allen belts and the space beyond them though.



Mind you, the crash-program timing I noted above is unfortunate, in that
it may put your first lunar expeditions during the nasty solar maximum
of the late 50s.

The belts would be discovered by early orbital scout flights, cosmic
radiation being recognized even then as an area of concern. In the real
world, while nobody (well, except for Nick Christofilos and a handful of
other people acquainted with his highly-classified work) was expecting
trapped-radiation belts, cosmic rays had been known since 1911 and their
intensity outside the atmosphere was a serious unknown... as witness the
fact that the major science instrument on the first US satellite was a
cosmic-ray detector.

Some of the scout pilots might get rather high doses while finding out the
extent of the problem, mind you.


I'm still waiting for the ferry rocket to malfunction during launch and
having it fall on one of the other ferry rockets on it's pad and then
all four scheduled for that day's launches go up in one big poisonous
fireball. :-)

Pat