There is much variation in the number of engines used in the liftoff stage of various launchers.

  • Saturn V had 5
  • Space shuttles had 5 (3 main + 2 SRB)
  • Falcon 9 has 9
  • Soyuz has 20
  • N1 had 30
  • Delta Heavy has 3
  • Long March 3B appears to have 4 main + 2 or 4 boosters
  • Ariane 5 had 1 main + 2 boosters

While the size and performance of these launchers varies, I can't see any straightforward correlation between those factors and the number of engines. The only correlation I can see is that Soviet/Russian launchers have a whole lot more and smaller engines.

What are the design tradeoffs when deciding to use many small engines vs few large engines?

Why does the Soviet/Russian calculus seem to be so different to everyone elses in this regard?

  • 1
    $\begingroup$ The Saturn V vs the N1 is because the Russians didn't have a high-output engine at the time. So in order to get the necessary thrust they used quantity > performance. The Soyuz is an older rocket, and I'd assume they figured if it isn't broke why fix it $\endgroup$ Commented Jul 29, 2015 at 20:55
  • 7
    $\begingroup$ Soyuz has technically only 5 booster stage engines (RD-108 main engine and one RD-107 per each of the 4 strap-on liquid fuel boosters, all LOX/Kerosene), but each engine uses four combustion chambers, thus 4 nozzles each. Note also that you're only counting booster stage (first stage with strap-on boosters) engines, but you consider the whole launch vehicle's performance later on. Not all of them would even have the same number of stages. $\endgroup$
    – TildalWave
    Commented Jul 29, 2015 at 21:06
  • $\begingroup$ If you want to be pedantic, F9 has 10 engines. $\endgroup$ Commented Jul 30, 2015 at 3:23
  • $\begingroup$ @EchoLogic Russell Borogove's edit clarified my original intent - I'm asking only about the liftoff stage. $\endgroup$ Commented Jul 30, 2015 at 17:29
  • $\begingroup$ There is a limit as to how much you can control engine thrust. Now I believe that in case of Falcon 9 one reason is precision thrust control. During 1-st stage re-entry and landing they only need three, resp. one engine. With one big it would not be possible. $\endgroup$ Commented Jan 23, 2018 at 12:18

2 Answers 2


The number of engines and strapons are a series of tradeoffs made based on available technology.

The Americans, planning a series of monstrous rockets, started on the huge F-1 engine. (1.5 to 1.8 million lbs of thrust). The only things the Russians ever had that came close was the RD-170 engine, with similar thrust, but needs 4 thrust chambers instead of the F-1's one. The RD-170 came several decades after the F-1 as well.

Combustion instability in large scale rockets like this is very hard to solve, and the Americans were possibly lucky that they succeeded. The Russians were not.

So when the time came to build a large booster, the Americans were able to get by with just 5 F-1s (Which is an astonishing amount of thrust, 7.5 million lbs of thrust is truly awe inspiring) whereas the Russians with the NK-15/NK-33 in the 380,000 lbs of thrust needed to use 30 of them to get the needed thrust.

There is a pragmatic explanation for each of those examples.

Strapon solids are usually a sign that the design grew during development. Usually the main engine size is set early in the design, and the vehicle grows over time. The way to mitigate this growth, since usually the main engine design is hard to scale up, is to add strap ons. Solid strapons are often available from the makers of ICBM's.

Strapons are great for being high in thrust, usually with pretty low ISP, so that they are mostly used to get the vehicle moving off the pad, and drop off early.

In terms of Soyuz, that is a bit of a misnomer, since the engines have multiple thrust chambers/bells, but count as a single engine. So the side boosters look like they have 4 bells, but is really a single engine each.

  • $\begingroup$ by chance do you know at what point in its development cycle STS gained SRB strapon boosters? $\endgroup$ Commented Jul 30, 2015 at 3:25
  • $\begingroup$ @EchoLogic - that's better asked as a separate question. $\endgroup$ Commented Jul 30, 2015 at 5:34
  • $\begingroup$ Skimming the wikipedia articles on shuttle development, it looks like the Air Force requirement for payload and cross-range ability landed about 1970-71, setting the size of the orbiter and killing Maxime Faget's "DC-3" proposal. SRBs were specified in the 1972 (?) RFP. en.wikipedia.org/wiki/Space_Shuttle_design_process#Final_design $\endgroup$ Commented Jul 30, 2015 at 14:50
  • $\begingroup$ SRBs are a cheap source of off-the-pad acceleration. $\endgroup$
    – Joshua
    Commented Dec 12, 2016 at 16:14
  • $\begingroup$ The development of the F-1 engines was started very early. The first stage engines should be ready before any flight tests of the rocket. $\endgroup$
    – Uwe
    Commented May 28, 2017 at 9:09

Everything's a tradeoff.

Designing and building small, low-thrust engines is generally easier than designing and building large, high-thrust engines, but the larger engines may be more efficient.

The more engines you have, the more likely one or more are to fail in any given launch.

Once you have more than 5 or so, you may be able to continue on if one fails (depending on which engine, and at what point in the ascent it happens).

The more engines you have, the harder it is to predict how they may interact; this is likely part of what contributed to the failure of the N1.

Given that N1 never flew successfully, and Soyuz is really a 4 booster + 1 sustainer first stage, the "Russians use more engines" assertion is misleading. Energia was 4+4, but flew only twice; Zenit uses a single 4-chamber engine; Proton uses 6 single-chamber engines.

Almost every engine and launcher in production has a complex development history; sometimes your requirements change and you wind up building a much larger launcher than you had expected to.

Engine development generally takes longer than the rest of launcher development (tankage, avionics, airframe, etc.), so you often don't have time to develop the optimal engine for a new launcher. During the height of the Cold War/Space Race, if your government made a request for proposals for a new launcher, you'd look around at the engines you had already developed or had experience with, divide the needed thrust at launch by the thrust of one of those engines, and that's how many engines you'd design for.

SpaceX wanted to use a common engine design for Falcon 1, Falcon 5 and Falcon 9, and use the same engine in the second stages of the 5 and 9. The upper stage naturally needs less thrust, so they designed the Merlin to drive a single-engine second stage, and payload goals set the number of engines in the first stage accordingly.

The Saturn program was originally going to be a larger family of launchers. The Saturn C-3 was to use 2 of the enormous F-1 engines, the C-4 would use 4, and the C-5 became the Saturn V with 5 of them.

  • 1
    $\begingroup$ Another possible reason for why Falcon 9 has so many engines in the first stage is that they want to use the very same engine on the second stage. (for simplicity and better economy of scale). It sounds plausible that with fewer engines on 1st stage, 2nd stage would have an unnecessarily powerful engine for its tank and end up too heavy. $\endgroup$
    – radex
    Commented Jul 30, 2015 at 21:40
  • $\begingroup$ Oh and there's also the fact that they use the first stage engine to land the rocket (which obviously makes the constraints different from other rockets): space.stackexchange.com/questions/7841/… $\endgroup$
    – radex
    Commented Jul 30, 2015 at 21:43
  • $\begingroup$ The Saturn V had engine-out capability on the first and second stage (en.wikipedia.org/wiki/Talk:Saturn_V/…). $\endgroup$
    – Hobbes
    Commented May 28, 2017 at 9:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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