During the Apollo missions the Lunar lander was separated, turned around and docked during flight. And the command module was left in Lunar orbit during the surface mission. The Soviets planned to have the single cosmonaut to land on the surface do a space walk between the Soyuz-like orbital module and the lander.

Is it correct to say that both these mission designs were compromises made because the launchers were slightly too small for a direct launch from Earth to the Lunar surface? Or was it perhaps a deliberate choice after all, maybe in order to win time in the competitive space race?

Why weren't the launchers designed to be larger so that the Lunar missions could be simplified? How much larger would they have had to be, 10%? 20%? Were they maybe as large as it was feasible to build launchers back in the 1960s, or did they during the development process fail to achieve the initial specifications? Is SLS Block II designed to perform a direct launch of astronauts to the Moon without dockings and EVA's?

enter image description here Image from SpaceIsThePlace. A model of the two cosmonauts transferring a box with Moon rocks from the Lunar lander to Soyuz.

  • 22
    $\begingroup$ Larger... to simplify? I think you have a bit of a leap of thinking there :) $\endgroup$ – Luaan Dec 16 '16 at 16:01
  • 4
    $\begingroup$ Maybe the better question (which I've sort of wondered about ever since the event) is why they didn't simplify by doing 3 launches to Earth orbit. $\endgroup$ – jamesqf Dec 16 '16 at 18:18
  • 10
    $\begingroup$ @LocalFluff: No, it's not easy to make stuff larger, 'cause you run up against things like the square-cube law and the rocket equation. That is, if you make things bigger, you need to make them stronger, which means you need to make them heavier, which means you need more fuel to lift the heavier structure, which means you need still more fuel to lift the fuel.... $\endgroup$ – jamesqf Dec 17 '16 at 5:13
  • 6
    $\begingroup$ Landing on the moon is easy. Leaving Earth without exploding is hard. $\endgroup$ – Mazura Dec 17 '16 at 17:31
  • 3
    $\begingroup$ The basic problem with a direct landing approach is that you need to land on the moon with enough of everything to be able to get the astronauts back to earth. That means the moon lander had to be much larger then one that just had to get back into lunar orbit. Much, much larger as it would have to carry the fuel and life support for the return journey down to the moon and back up off it again, which would require lots of extra fuel. $\endgroup$ – Paul Smith Dec 21 '16 at 11:52

In order to use the direct ascent method of landing on the moon, which is where the entire vehicle descends and leaves the moon, you would need a rocket an order of magnitude bigger than the Saturn V, not just a bit bigger.

Here's an early comparison NASA made back before they decided to use Lunar Orbit Rendezvous:

Nova rocket comparison

The C1 became the Saturn I, the C-5 became the Saturn V, the Nova never got off the drawing board because it was going to be very expensive and hard to build considering that with LOR you could do it for much less.

Here's a good article on the history of the Nova rocket. Here's an excerpt:

Nova was NASA's ultimate launch vehicle, studied intently from 1959 to 1962. Originally conceived to allow a direct manned landing on the moon, in its final iteration it was to put a million-pound payload into low earth orbit to support manned Mars expeditions. It was abandoned in NASA advanced mission planning thereafter in favor of growth versions of the Saturn V.

As time went on the rocket grew and grew and grew until it was a true monster! See some of the concepts in the image below, they make the Saturn V puny in comparison.

Nova - Martin Marietta Advanced Designs

Basically, the program died because it wasn't needed, there was no plan for a manned mission to Mars, which was the only use case for it.

  • 7
    $\begingroup$ The concepts in your lower pic are for the post-Apollo Mars rockets which also were conceived under the Nova name. The C-8 Nova needed for direct ascent would have been 60%-65% larger than Saturn V, not an "order of magnitude" larger. $\endgroup$ – Russell Borogove Dec 16 '16 at 17:06
  • $\begingroup$ Any relationship to SeaDragon? $\endgroup$ – Sarah Bailey Dec 19 '16 at 17:45
  • $\begingroup$ @RussellBorogove Does SLS Block II, allow for any other mode for a Moon mission than an Apollo one? Could its bigger capacity allow for the Apollo lander NOT would have to be separated, turned around and dock in-flight? Allowing for fairing and support structures and redesign to have it docked upside down at launch. (Or maybe there's no reason to skip such turn-around maneuver, maybe there's almost no drawback to it?) $\endgroup$ – LocalFluff Dec 21 '16 at 13:11
  • 1
    $\begingroup$ SLS block II at 130 tons to LEO is not significantly more powerful than Saturn V at ~125 tons. The transposition, docking, and extraction maneuver, as discussed in an older answer, is basically forced by the Apollo launch escape system design, and in any case if you're planning LOR (the mass efficient strategy for lunar landing), you have to be confident in that docking maneuver anyway. $\endgroup$ – Russell Borogove Dec 21 '16 at 16:23
  • 2
    $\begingroup$ @sarahbailey No real relationship to Sea Dragon other than a million pound payload. Sea Dragon called for one enormous low-pressure engine on each stage, the Martin Nova concepts were arranging lots of little engines in a plug nozzle configuration, etc. $\endgroup$ – Russell Borogove Dec 21 '16 at 16:49

During the early part of the Apollo program, the "direct ascent" mode was favored and Lunar-orbit rendezvous (LOR) was considered far too complex. In fact, the specifications of the Apollo service module were set by the direct ascent plan: the SPS engine is sized to lift off from the moon, and the fuel tankage is sufficient for lunar ascent and return to Earth.

The Nova rocket needed for direct ascent would have been about 60% heavier than the Saturn V (~4750 tons versus ~2970 tons). Development of that large rocket was going to take a very long time. Kennedy's "before the decade was out" speech set a deadline for the program, and Nova would take too long. The project took another look at LOR, concluded it was the fastest and cheapest approach, and decided to commit to it. This shaped the Gemini program; if the lunar mission required rendezvous, NASA needed to learn how to do it.

Kennedy's deadline also effectively set a deadline for the Soviet space program, so it's likely that the same logic drove them to a rendezvous strategy. The great advantage of LOR is that the smaller the lander, the less fuel needs to be carried for lunar descent and ascent, and the Soviet LK was indeed tiny and compromised, carrying one man instead of two, requiring the crewman to EVA to board, and so on.

  • 2
    $\begingroup$ It would be worthwhile to mention the ideal rocket equation and explain why this can make a smallish increase in the size of the payload result in a largish increase in the size of the rocket. $\endgroup$ – David Hammen Dec 16 '16 at 15:59
  • $\begingroup$ Auhoo. This makes the SLS uselesser than... So it would only be able to repeat the Apollo concept with some small extras? It is like a shoemaker offering you two pairs of shoes which are half your size. If the SLS cannot improve upon Apollo's Lunar mission concept, then my last attempt of trying to like it has finally gone away. Even the launchers that the Russians and Chinese fantasize about aren't big enough for direct launch. If it has to be 65% larger than Saturn V. Suddenly the distance to the Moon increased in my mind. $\endgroup$ – LocalFluff Dec 16 '16 at 21:19
  • $\begingroup$ SLS simply isn't intended for single-launch moon landing. Should a moon landing budget magically appear, two-launch Earth orbit rendezvous is probably practical for an Altair-type mission. $\endgroup$ – Russell Borogove Dec 17 '16 at 0:03
  • 5
    $\begingroup$ ...and simply there's no point to carry all the fuel needed for Moon-Earth transfer down to Moon surface and back up. If you have the capacity, it's better to use it for more scientific payload and still leave the fuel in orbit. Be it in form of a 'command module', an extra propulsion module, or even just fuel tankage. Orbital rendezvous is currently a routine operation, so complexity isn't that much of a problem. But costs stemming from sheer size are more of a problem than in the 70s. $\endgroup$ – SF. Dec 17 '16 at 0:49
  • 1
    $\begingroup$ Ascent from lunar surface to LLO is about 1730m/s; LLO to Earth intercept is about 680m/s. From intercept, atmospheric braking does the rest. So ascent is the larger part but TEI is a very significant part of the budget. $\endgroup$ – Russell Borogove Dec 19 '16 at 19:05

On the Soviet side, Vladimir Chelomey's UR-700/LK-700 project would have used a direct ascent mission profile. The UR-700 was 76 meters tall and 17.5 meters wide at the base; it would have looked something like this:

enter image description here

enter image description here

(left-to-right: N-1, Saturn V, UR-700)

The UR-700 didn't have much design work done on it, but there were problems developing the high-thrust RD-270 engines powering the whole thing. It ended up overlapping too much with the already-existing N-1 project, and never received further development.

  • 5
    $\begingroup$ It's funny how much that looks like a Kerbal-Space-Program rocket design, with many cylindrical rockets connected together! $\endgroup$ – Peter Cordes Dec 16 '16 at 18:44
  • 1
    $\begingroup$ The strap-ons on the lower stage are derived from Proton (a.k.a. the UR-500) components; part of Chelomey's pitch for the UR-700 was that he could incrementally test parts of the UR-700 on smaller rockets. $\endgroup$ – DylanSp Dec 16 '16 at 18:57
  • 2
    $\begingroup$ @DylanSp: Also, move them in on railroads. $\endgroup$ – Joshua Dec 17 '16 at 4:07
  • 1
    $\begingroup$ @BlokeDownThePub That sounds like a good follow-up question. $\endgroup$ – DylanSp Dec 18 '16 at 1:32
  • 3
    $\begingroup$ @BlokeDownThePub Why do the Russians use these fence-like interstage fairings? $\endgroup$ – a CVn Dec 19 '16 at 20:12

Its not rocket science, well actually, it is rocket science. Rockets are not very efficient at all and the worst bit is that to carry anything useful, they have to carry the fuel to push the useful bit, then they have to carry the fuel to carry the fuel to push the useful bit and so on. Saturn V was about three thousand tons at takeoff, and was able to push 45 tons towards the moon, so only about 1.5% of the take-off weight left the earth towards the moon. To increase the moon lander to 50 tons, Saturn V would have had to increase to almost 4,000 tons. The Apollo 17 space capsule they picked up out of the pacific probable weighed not much more then three tons, giving a total efficiency of 0.1%.


There was another argument against the direct ascent mode. A very large and complex rocket should be landed and started on and from the moon surface. But rockets of this size would require on earth a launch pad, a start tower, connections to a control room, a count down with many steps and a lot of personnel. But the rocket motors of the service module and the lunar module used for lunar rendezvous mode could be built very simple and reliable. No turbo pumps, no gas generators necessary, just only compressed nitrogen from tanks for the pressurization of propellants. These motors could be ignited in zero gravity. The hypergolic fuel made unlimited reignition possible.


Compare the extra work to build a bigger rocket (and don't forget that costs have the ugly tendency to be exponentialised, thanks to our good friend, the Tsiolkovski equation) with a few hours for the docking. Rockets aren't cheap.

The main thing to consider is that you need to dock anyway - to get the lunar crew back from the Moon. So you're not saving on anything but those few hours; all the required docking equipment is still required with a bigger rocket. Unless you're considering having the landing module powerful enough to land, ascend and fly back to Earth and land there. In which case, refer to paragraph one - absurdly expensive.

The Soviets didn't use docking. But this wasn't because they thought docking is undesirable - it was because they simply couldn't do it. They didn't have the capability. So instead, their astronaut was expected to make a space walk, and they couldn't exploit the (remainder of the) landing module for the trip there and back again.

Note that even with all the modern technology we have now, we still use the same basic approach. The Space Shuttle was still a staged rocket, though the bit returning back to Earth was considerably larger than any earlier attempt. Modern private rockets experiment a lot with recovering the spent stages, but they still use the same basic staging approaches. This is because our engines are horribly inefficient, which in addition to having to carry fuel to carry fuel means that the costs of additional weight don't scale very well - we tend to use the smallest possible rocket for the task at hand. You use a big rocket to get out of atmospheric drag, a smaller rocket to make the rest of the trip (and velocity) to orbit, a smaller rocket yet to circularise your orbit or start an injection, even smaller rocket for the landing and even smaller for taking off again and going back home (do you think Mars is within our easy reach nowadays? For a single trip, kind of; add a return trip, and we're back to "oh boy").

  • 3
    $\begingroup$ The Nova direct ascent plan didn't require any docking maneuver; the command module with its crew of three would have landed as part of a two stage vehicle with nothing left in lunar orbit. $\endgroup$ – Russell Borogove Dec 16 '16 at 16:32
  • $\begingroup$ @RussellBorogove ... and the Nova was discarded because it just wasn't worth it. I don't see how that invalidates what I've said :) $\endgroup$ – Luaan Dec 16 '16 at 23:26
  • $\begingroup$ @Luaan: It invalidates the first sentence in your second paragraph. The Nova plan didn't "need to dock anyway". $\endgroup$ – TonyK Dec 16 '16 at 23:38
  • 1
    $\begingroup$ @TonyK "Unless you're considering..." If you want to avoid docking entirely, you need a ridiculously large rocket. In fact, the original Nova mission plan was not only too big for the planned Nova rocket, it would have required docking 15 separate launches in orbit using the Saturn design of the time. They considered nuclear engines to get the required delta-V (except for Earth ascent, of course) and other complex and untested technologies (like aerospikes) - not to mention that it was too big for the existing factories. I'm just pointing out the OP is severely underestimating the costs. $\endgroup$ – Luaan Dec 16 '16 at 23:52
  • $\begingroup$ What are you talking about? The Nova lunar mission would have been single launch, direct ascent, no docking, no nuke, using one rocket based on the Nova 8L concept. A single ~4800 ton launcher, ~180 tons to LEO, could land an Apollo CM/SM + lunar descent stage. astronautix.com/n/nova8l.html The 15-Saturn-I launch strategy was insane and didn't last past 1959. $\endgroup$ – Russell Borogove Dec 17 '16 at 17:17

During the Apollo missions the Lunar lander was separated, turned around and docked during flight.

This point hasn't really been addressed by any of the other answers.

Unlike separation to land the lander, then subsequent docking to get the crew back, there is a fairly simple reason for this manoeuvrer.

The command pod needs to be the top of the stack on launch, as there is no other way to have a working abort system. This means that you cannot pre-attach the lander on the top of the rocket. Likewise, it also cannot be dragged along behind, because the final stage rocket is behind the command capsule, and

  1. you can't transfer crew through the rocket
  2. firing the rocket engine at the lander is undesirable

That leaves the only viable option being to have the lander under (behind) the command capsule at launch, but move it to be in front once the abort system is no longer needed, which is the operation described in the Question.

A larger rocket would not allow this step to be removed.

  • $\begingroup$ If a much larger rocket would be used for a direct ascent from lunar surface no separate lander is used. The command pod would land on moon. $\endgroup$ – Uwe Dec 4 '18 at 16:07
  • 1
    $\begingroup$ @Uwe well, I was trying to touch upon the why so complicated part of the question, which I felt was under-represented by the other answers. They do the direct ascent part well enough that I don't need to repeat them. I wanted to make the point that scale alone isn't the cause of the complexity (though landing the capsule in one go would solve it). for general "why not a bigger rocket" thoughts, youtu.be/Sfc2Jg1gkKA answers that question quite well too. $\endgroup$ – Baldrickk Dec 4 '18 at 17:06

protected by Community Feb 26 '17 at 17:10

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.