When falling back to Earth, why does the Falcon 9 first stage perform a burn at about 60 km (37 mi) altitude during re-entry instead of performing a longer burn closer to the drone ship? Would it have been damaged by aerodynamic heating without that braking?

  • 1
    $\begingroup$ Related, maybe not exact dupe $\endgroup$ Commented Feb 23 at 11:42
  • 1
    $\begingroup$ It's interesting to look at the speed of the booster during the launch. You will see that when coming back velocity starts building up, they do the reentry burn which slows it down significantly. Right after the reentry burn usually you see velocity increasing again for a little bit but after a few seconds it stops and starts decreasing because the booster is already reaching an altitude where there's enough air to have significant friction... so the timing is very precise to slow it down as much as possible before it hits the "thick" part of the atmosphere. $\endgroup$
    – Bakuriu
    Commented Feb 24 at 13:32

2 Answers 2


A returning Falcon 9 first stage will perform at least 2 of 3 possible burns.

If it is performing a return to launch site (RTLS) operation and landing at LZ-1/2/4 landing pads (LZ 1/2 are in Florida, LZ-4 is in California) then it does a boost back burn, to start heading back to the pad, since by stage separation it is already quite a bit downrange.

This is not performed when it is landing on a SpaceX Autonomous Spaceport Drone Ship (ASDS) barge downrange.

The 60 Km height burn you refer to is the re-entry burn. They need to slow down, so that when they hit the thicker part of the atmosphere they are going slow enough to survive the temperatures from friction.

After that they use the body of the booster to 'fly' and generate air friction and slow down, until the landing burn where they scrub off all remaining velocity, usually in the last kilometer of height so they land at 0 altitude and 0 velocity.

  • 2
    $\begingroup$ They used to sometimes do a brief boostback burn even when landing on the drone ship as a way to decrease the downrange landing distance. Presumably this was to reduce the time and cost of recovery operations and also get the booster back to port quicker. I'm not sure why they still don't do this on at least some missions, maybe it's because they need to be farther downrange anyway to recover the fairings? Although only the support ships are needed for that not the barge. There is an existing question here but I didn't see any definitive answers. $\endgroup$ Commented Feb 23 at 14:53
  • 2
    $\begingroup$ @StevePemberton: A possible explanation is that F9 has become so powerful that they can do RTLS for cases where they couldn't before (e.g. Dragon missions since last year). OTOH, F9 has become so dominant that payloads are sized specifically to extract maximum performance (most obviously Starlink). This would reduce the "middle ground" of "not enough margins for RTLS, but not fully using the power either". And, of course, Starlink just massively skews every possible statistic. But I vaguely remember a "partial boostback" just last year. $\endgroup$ Commented Feb 23 at 17:09
  • $\begingroup$ As you mentioned, I don't think the fairings are the reason. SpaceX just recently had one support ship pick up four fairing halves from two launches, so that is an easy way to improve cadence. For Starlink, SpaceX has received an FCC permit to choose between downrange and RTLS trajectories on short notice – all they need to do is unload a couple of satellites. $\endgroup$ Commented Feb 23 at 17:12
  • $\begingroup$ @JörgWMittag - okay thanks, if there was one last year then I guess it's sort of like attempting to recover Falcon Heavy core boosters, rare enough that's it's easy to mistakenly conclude that it isn't done anymore. $\endgroup$ Commented Feb 23 at 20:25
  • 2
    $\begingroup$ @dlu: the FCC. For every rocket launch, you need a permit to operate radios with a certain frequency and certain power at a certain place pointed in a certain direction. In particular, there is a big difference between operating the landing radar in the middle of the ocean or near not only one of the busiest airspaces in the US but also the busiest cruise ship terminal in the world. $\endgroup$ Commented Feb 24 at 11:08

WARNING: I will use a lot of simplifications, metaphors, and analogies. If you are a physicist, rocket scientist, or aerodynamicist, please skip this answer :-D

This is called supersonic retropropulsion, which roughly means "firing your rocket backwards into a gas stream while flying at supersonic speed".

While firing a rocket backwards obviously slows it down, this is not the only, and maybe not even the primary, reason for doing a supersonic retropropulsive burn.

When flying supersonic through the atmosphere, the gas in front of the vehicle "cannot get out of the way" fast enough. That is essentially what "supersonic" means: one way to think about the speed of sound in a gas is the speed at which gas molecules can move. When flying supersonic, you are flying faster than the gas in front of you can move, so you compress it and a shock wave forms in front of the vehicle.

According to the laws of thermodynamics, compressing a gas heats it up. This is called adiabatic compression and is the same process – just in reverse – that makes a spray can get cold. One of the ways to think about the temperature of a gas is "the average motion of the molecules". If you compress the gas, you have more molecules in the same space, which means you have more motion in the same space – in other words, a higher temperature.

At the hypersonic velocities that orbital boosters like Falcon 9 reenter at, this temperature is so high that the gas even turns into a plasma.

What supersonic retropropulsion does, is that the exhaust gases form kind of a "heat shield". The exhaust gases "push away" the plasma from the rocket. Somewhat ironically, the rocket is flying through its own flames because those flames are still cooler than the gas they would otherwise fly through. This is sometimes jokingly referred to as "fighting fire with fire", although it should more precisely be referred to as "fighting plasma with pressure".

Of course, slowing down helps as well since the temperature (roughly) goes up with the cube(!!!) of the velocity.

Here is an infrared video captured by NASA of the reentry phase of the CRS-4 mission:

So, in conclusion, there are two effects at play here:

  1. During the supersonic retropropulsive burn, the exhaust gases act as a heat shield.
  2. After the burn, the lower velocity reduces the heat flux.

As always, this is an engineering compromise: the most propellant-efficient way would be to do a single, hard suicide burn with all engines at the last possible moment. But propellant is not the only factor to be optimized. Mass needs to be considered as well, specifically in this case you would need a lot more heat shielding on the booster. If this heat shield weighs more than the propellant you save, your tradeoff is bad.

Another factor to consider, which is unique to SpaceX, is the refurbishment time of the heat shield. While these numbers are not public, there are indications that SpaceX has brought down the refurbishment time for a Falcon 9 booster to 5 days. Which means you cannot use an ablative heat shield, for one.

I was surprised there is no single source with a simple explanation of supersonic retropropulsion, so here's two search links instead:

  • $\begingroup$ I wonder how much would the stage heat up without the retropropulsion burn (in ° Celsius) and what effects this would have on it. $\endgroup$ Commented Feb 23 at 13:13
  • $\begingroup$ It would be interesting to compare how much of the heat reduction is from speed reduction compared to the heat reduction from pushing the plasma away. I would think the effects of the slowdown itself is pretty significant, even during the burn. Looking at yesterday's Starlink launch for example the entry burn started at just under 65 km altitude at 7900 km/h, by the end of the 22 second burn the booster was at 45 km and had slowed to 5,100 km/h. After the burn ended the velocity slowly climbed to 5,250 then began decreasing as it passed through 35 km. $\endgroup$ Commented Feb 23 at 14:39
  • $\begingroup$ One more factor: The reentry burn also reduces the mass of the rocket (by expending fuel), which in turn reduces its kinetic energy (a bit) and changes its aerodynamic properties. I'm sure that for this reason, there's a maximum amount of fuel onboard beyond which safe booster reentry is not possible. I have no idea if that maximum is low enough to ever be relevant in planning; maybe it's only relevant when trying to recover the core booster in a Falcon Heavy flight? $\endgroup$ Commented Feb 23 at 17:14
  • $\begingroup$ @CharlesStaats - your comment reminded me that very early on the reason given for the entry burn by Elon or maybe SpaceX commentators, I don't remember which, was to reduce stresses on the booster when it hits the atmosphere. I don't remember heat being mentioned, that explanation seemed to come later. Not that anything has changed, but perhaps the "stresses" reason originally given was just an oversimplification and now more details are now coming to light, either from SpaceX or from people figuring things out from the details that we do have. $\endgroup$ Commented Feb 23 at 17:48

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.