With any normal liquid fuelled rocket engine it is possible to throttle the thrust up and down, but in Kerbal space program it is possible to bring it all the way down to around a percent thrust. Is this possible with normal rocket engines?

  • $\begingroup$ Related: space.stackexchange.com/a/28095/195 $\endgroup$ Dec 5, 2022 at 6:44
  • $\begingroup$ Just FYI, you can actually throttle way below 1% in KSP by first changing the "thrust limiter" setting to whatever you want the new 100% to be. This lets you do RCS-like burns with your big main engine and other funny stuff like that. Obviously there's nothing even remotely realistic about this. $\endgroup$
    – TooTea
    Dec 5, 2022 at 9:05
  • 3
    $\begingroup$ @TooTea The realistic counterpart is to just vent cold propellants through the engine. The unrealistic part is that KSP scales fuel usage linearly with the thrust, without any loss of Isp. In reality throttling the engine massively reduces specific impulse - lower chamber pressure, lower exhaust speed, throttling the engine is very wasteful. That's why IRL it's much better to extinguish some of the engines and run the remaining ones near max thrust instead of throttling them all, while in KSP half throttle is half thrust and half the fuel usage. $\endgroup$
    – SF.
    Dec 5, 2022 at 13:57
  • $\begingroup$ The Realism Overhaul and Real Fuels mods for KSP change this behavior so that engines have real-life throttleability (sometimes none for early engines). $\endgroup$
    – lamont
    Dec 6, 2022 at 6:12

2 Answers 2


It is not usually possible to throttle a rocket engine that deeply for a range of reasons and especially for liquid propellent rockets. As an example the merlin engine of the Falcon 9 booster provides so much thrust and the booster has so little mass as it approaches the landing pad, that it is very hard to get it landed in one piece at all because even at minimum throttle there is too much thrust (it can't hover).

The solution adopted by SpaceX, rather than attempt to adjust the engine for deeper throttling, was exact timing. As the booster descends with the engine firing it is losing speed. The landing is timed so that at 0 altitude it is at zero velocity and the engine cuts out.

If the engine cut out any sooner the booster would crash into the landing pad under the force of gravity any later and the booter would take off again.

  • $\begingroup$ So, you are saying the minimum throttle thrust from a single Merlin exceeds the dry weight of Falcon9 first stage? With Falcon9, is there a "ground cushion" effect like fixed and rotary wing lift? Can Falcon9 be "throttled" by progressive shutdown of engines? $\endgroup$
    – Woody
    Dec 5, 2022 at 16:46
  • 1
    $\begingroup$ @Woody: Correct. Even with a single Merlin at minimum throttle settings, the thrust-to-weight ration is greater than 1. Hence, the hover slam / suicide burn approach to landing. (It is also the most fuel-efficient, but I assume especially at the beginning, SpaceX would have loved to have some more leeway in timing the landing burns.) $\endgroup$ Dec 5, 2022 at 17:00
  • $\begingroup$ @JörgWMittag ... Right you are. Dry mass 1st stage Falcon9=14Tonnes. Max thrust Merlin=100Tonnes $\endgroup$
    – Woody
    Dec 5, 2022 at 18:24
  • $\begingroup$ In fact, now that they have the timing down, SpaceX does a 3-engine landing burn rather than a 1-engine; this shortens the final deceleration phase, saving fuel, leaving more margin for larger payloads. $\endgroup$ Dec 5, 2022 at 21:34
  • $\begingroup$ @RussellBorogove, when did they make that change? It's been a month or two since I last watched a Falcon 9 launch, but the commentary was still describing single-engine landing burns then. $\endgroup$
    – Mark
    Dec 7, 2022 at 4:08

Typical rocket engines can't throttle that low, but it's possible.

The widest throttle range I know of is the sustainer engine of the Lance tactical missile, at 350:1 (yes, really), meaning it can be throttled to about 0.3% of full power.

The Space Shuttle main engine could throttle down to 67% or so, and the Falcon 9's first-stage Merlins are claimed by some to throttle down to 40%.

Widely-throttleable engines tend to run into problems with stable combustion and exhaust flow, so it tends to be easier to design engines with little or no throttling capability.

  • $\begingroup$ Interesting - what makes Lance missile able for such throtte-down? I could find only that it's liquid-fueled... $\endgroup$
    – Heopps
    Dec 5, 2022 at 12:46
  • $\begingroup$ @Heopps: Using a hypergolic fuel/oxidizer mix (UDMH + IRFNA) probably helps, since you don't have to worry about the engine going out (or guttering and then reigniting explosively) if you throttle it down too much — as long as there's fuel and oxidizer getting mixed together, they'll burn. I doubt this is the full answer, but it's probably part of it. $\endgroup$ Dec 5, 2022 at 14:49
  • 1
    $\begingroup$ Also, the Lance missile had a funny design with concentric(!) sustainer and booster engines. I wonder if that 350:1 figure is really for the sustainer alone, or if it's rather the ratio of sustainer + booster at full throttle vs. sustainer only at minimum throttle. $\endgroup$ Dec 5, 2022 at 14:52
  • 3
    $\begingroup$ The 350:1 range was for the sustainer alone. The booster engine was about 8x as powerful as the sustainer at full throttle. Apparently the booster would cut off when it was on a proper trajectory and the sustainer would continually adjust to cancel out drag, which would vary significantly with altitude; today there would be more sophisticated ways to do that guidance. $\endgroup$ Dec 5, 2022 at 21:26
  • 1
    $\begingroup$ Sutton says it's a "moving pintle injector", i.e. the same throttling strategy as Merlin, with a 15% loss of specific impulse at the low end of throttle. It's got a very short nozzle, which helps exhaust flow stability at the expense of specific impulse. $\endgroup$ Dec 5, 2022 at 21:29

Your Answer

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

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