As I understand it, a rocket should produce its maximum thrust on liftoff. As time progresses, the mass of the rocket will decrease, but since the engines produce the same thrust, the rocket will accelerate faster. So, all rocket engines on the first stage (which aren't blocked by other stages) should ignite on lift off, which will reduce the mass of the vehicle quicker, giving more acceleration. This seems true for most rockets.

For the GSLV MK3 by Indian Space Research Organisation (ISRO), it seems different:

GSLV MK3 Staging Timing

The GSLV MK3 ignites its core liquid engines (L110) approximately 120 seconds after liftoff. As described on the launch video, the exact timing depends on RTC (Real Time Decision). This requires the S200 Solid strapons to lift the full L110 mass for up to 120 seconds.

What factors dictate this kind of staging sequence? Why are the L110s not ignited on liftoff? Is max Q the reason?

  • 8
    $\begingroup$ It's worth noting that, depending on the grain (that is, the cross-sectional layout) of the solid rocket propellants, the solid rocket motors may be designed for very high initial acceleration that tapers off over time, instead of staying constant or increasing like you might expect. This allows using the low-efficiency, high-thrust SRMs early in the flight, and switching over to liquids once you're high enough for their higher efficiency to really shine. $\endgroup$
    – CBHacking
    Commented Sep 13, 2016 at 18:12

3 Answers 3


Besides limiting aerodynamic stress and drag losses as you and Antzi mention, using the core engine only at high altitude means the engine can be optimized for low-pressure use by putting a larger nozzle on it. This optimizes expansion of the exhaust, and in the case of GSLV MkIII contributes to a ~6% increase in specific impulse over the sea-level version of the engine.

Note that GSLV III isn't unique in doing this; the Titan III and IV launchers also launched using only SRMs, igniting the liquid-fueled core stage as the solids burned out.

  • $\begingroup$ impact of nozzle design is seems to be true, even though they are using same vikas engines, they seems to modify for MK3(Vikas-X). en.wikipedia.org/wiki/Vikas_(rocket_engine) $\endgroup$ Commented Sep 13, 2016 at 10:34
  • 7
    $\begingroup$ This is the correct answer. I would like to add, that specifically the performance is increased, because they can use a nozzle so large that it would not work at ground level. The flow would detach, leading to poor performance and enormous side loads. They considered doing this for Ariane 5, but instead opted to run the main engine for 7 seconds before igniting the boosters, so they can shut the engine down again and abort launch, if it does not function nominally. $\endgroup$ Commented Sep 13, 2016 at 14:46
  • $\begingroup$ I mean, they considered going with the same approach as GSLV. I did not mean to say they considered running an engine with detached nozzle flow. A lot of work is also being done on dual bell nozzles. I would not be surprised if we will see them launching with a decade. $\endgroup$ Commented Sep 13, 2016 at 15:00

Your assumption that we need max thrust at takeoff is partially wrong. Although right at takeoff you do want max thrust, it might be counterproductive short after:

  1. Your rocket and payload are Max G rated. You can't have an arbitrary high acceleration
  2. Atmospheric drag is higher at lower altitude and increase with the square of speed.

If you go too fast too quickly you'll lose all your speed to drag.

As quoted from wikipedia:

In flight, as the thrust from the S200 boosters begins to tail off, the decline in acceleration is sensed by the rocket’s onboard sensors and the twin Vikas engines on the ‘L110’ liquid propellant core stage are then ignited. Before the S200s separate and fall away from the rocket, the solid boosters as well as the Vikas engines operate together for a short period of time

As it might be technically challenging to stop then restart the engine in flight, or throttle them too much, the liquid engine is used only as a relay for the solid boosters.

I couldn't find the acceleration curve to support my claim; sorry for that.

  • $\begingroup$ what-if.xkcd.com/24 about halfway down is an acc. curve $\endgroup$
    – Tim
    Commented Sep 14, 2016 at 5:34
  • $\begingroup$ @Tim we'd need that of that specific rockets tho :) $\endgroup$
    – Antzi
    Commented Sep 14, 2016 at 5:37
  • $\begingroup$ Sure, but that is a reasonable estimate of your average backyard rocket :) $\endgroup$
    – Tim
    Commented Sep 14, 2016 at 13:13
  • $\begingroup$ "your rocket and payload are max G rated" - and humans are particularly bad at withstanding acceleration. what-if.xkcd.com/116 $\endgroup$ Commented Sep 14, 2016 at 17:58
  • $\begingroup$ @Antzi Recently posted the GSLV acceleration profile plots on another thread. Here is the profile I had created. (Reverse engineered from launch telecast screenshots - github.com/ravi4ram/Launcher-Profile ) [![enter image description here](i.sstatic.net/VaDPy.png )](i.sstatic.net/VaDPy.png )   Wikipedia quote, that you had mentioned looks to fit on the acceleration profile plot. L110 is fired when the acceleration starts to show decreasing trend. $\endgroup$
    – ravi_ram
    Commented Jul 29, 2020 at 13:20

Any excess energy, potential and kinetic, the boosters carry after separation is wasted. Therefore, you want to separate them at the lowest possible altitude and velocity, saving the core rocket's fuel for when it must no longer drag the boosters with it.

This holds true for any stage separation and it's the reason we use staged rockets at all.

It contradicts the objective to produce maximum thrust at takeoff, so any choice of staging is an engineering trade-off.

One way to achieve both is asparagus staging, where all engines ignite immediatly, but the core engine is fueled by the boosters' tanks, also saving on it's own fuel. This, however, comes at the price of additional complexity and weight and has consequently never been done in real life. The falcon heavy, scheduled for 2017, is the first rocket to use it. As @DylanSp pointed out, crossfeed has been cancelled.

  • $\begingroup$ First statement you have pointed out actually support maximum thurst at take off. Fuel cross-feed is seems to be the optimal idea, but makes things much complex. $\endgroup$ Commented Sep 14, 2016 at 2:31
  • 1
    $\begingroup$ @JithinJose No, it doesn't support it. Assume our boosters burn for 50 seconds. When the core stage fires in parallel to them, they still separate after 50 seconds, but by then the rocket is higher and faster. This means some of the rockets $\Delta v$ has been shared with the boosters and is now lost. $\endgroup$
    – Rainer P.
    Commented Sep 14, 2016 at 8:02
  • $\begingroup$ Lol'd at this link to the Kerbal Space Program wiki page. $\endgroup$ Commented Sep 14, 2016 at 10:55
  • $\begingroup$ Cross-feed's been cancelled on the FH. $\endgroup$
    – DylanSp
    Commented Sep 14, 2016 at 12:38
  • $\begingroup$ This is the only answer that mentions the most important reason, i.e. staging! By throwing away the dead weight of the SRM casings around the same time that you ignite the core, the (higher-Isp) energy of the core stage isn't wasted on lifting the SRM cases. The SRMs really are effectively just a 0th stage in this case. $\endgroup$ Commented Sep 15, 2016 at 23:16

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.