From answers to How much time it takes for the capsules to descend from ISS back down to Earth? I understand it takes 19 hours from the undocking from ISS until it touches the land.

But what are the main stages of the descent and how much time it takes for every such stage?


1 Answer 1


We'll ignore the 18 hours spent in orbit between Endeavour's undocking and the actual beginning of descent; I don't know what the crew were doing during that period, but it wasn't really part of the return-to-Earth process.

I'm basing this answer on the times given Demo-2 return timeline posted on the NASA ISS blog on August 2, 2020, which as of this writing is the only crewed Crew Dragon reentry which has been flown. We can assume most Crew Dragon descents will be generally similar, though the exact time, altitude, and speed details may change.

We begin in orbit at around 400km altitude, flying horizontally at about 7700 m/s, circling the Earth every 93 minutes. If we don't do anything we will continue to circle the Earth for days or weeks before the traces of atmosphere up here slow us down enough to fall to Earth.

Crew Dragon is made up of two parts: the conical reentry capsule, pressurized, where the crew sits, and the cylindrical trunk, behind the capsule, which includes solar panels and heat radiators. The trunk is expendable; the heat shield which protects the spacecraft is on the crew capsule.

57 minutes before splashdown: Trunk separation

The first step in the reentry process is trunk separation -- we get rid of the trunk section of the spacecraft, because the aerodynamics of the trunk would cause the spacecraft to orient nose-forward when it enters the atmosphere, which is not the way we want to go. At this point, having detached the solar panels, the spacecraft is on battery power, and we're on a short time limit.

52 minutes before splashdown: Deorbit burn begins

In order to reenter the atmosphere, we have to slow down. We only need to cut about 100 m/s off our orbital velocity; this will cause our orbit to become elliptical, and the lowest part of the orbit will dip low enough into the atmosphere that air resistance will cause the spacecraft to slow further. To slow down we fire the small Draco rockets for about 12 minutes; firings of rocket engines are referred to as "burns", so this is the "deorbit burn".

40 minutes before splashdown: Deorbit burn complete

We've now slowed down enough that we are guaranteed to re-enter the atmosphere and land on this orbit, but it's going to be another half hour or so before re-entry really begins.

37 minutes before splashdown: Nosecone closes

I don't know what's up with the nosecone, exactly, but it closes up here. For some strange reason the timeline refers to "deploying" it rather than closing it. We have another open question about the nosecone.

16 minutes before splashdown: Crew Dragon maneuvers to attitude for re-entry

The heat shield is at the back of the crew capsule. Once the capsule is into the atmosphere, its center of gravity is positioned so that the capsule will remain heat-shield forward, tilted at a particular angle which gives the capsule just the right amount of lift to make the ride reasonably comfortable, but we need to make sure we're in that attitude before the aerodynamic forces kick in. We fire the small Draco thrusters at this point to turn the spacecraft.

13-15 minutes before splashdown: Entry

I don't know the exact timing of atmospheric entry for Crew Dragon, but for Apollo capsules it was about 15 minutes before splashdown. Apollo defined “entry” as the point at which air resistance was creating 0.05g of deceleration on the spacecraft, which was reached at around 120km altitude. Up to this point the spacecraft is still moving at over 7000 m/s. The g-force builds up rapidly from here, because the slowing spacecraft loses altitude faster, dropping into denser air, which produces more air resistance, which slows it more rapidly, and so on. The rapid compression of the air in front of the capsule creates a lot of heat; the blunt shape of the heat shield forces the superheated air to the sides to reduce heating of the skin of the spacecraft. The superheated air ionizes, creating electromagnetic interference which "blacks out" radio communications for a period of about 6 minutes.

4 minutes before splashdown: Drogue parachutes deploy

At this point, air resistance has bled off almost all the capsule's orbital speed; it's only going 150 m/s (350 mph), at an altitude of about 5.5km. Small parachutes called "drogues" are released here; they slow the capsule a little bit and help the main parachutes to release cleanly.

3 minutes before splashdown: Main parachutes deploy

The drogues have slowed the capsule down to about 53 m/s (119 mph), which is slow enough that the main chutes won’t be ripped apart. At 1.8km altitude, the 4 large parachutes begin to deploy.


Three minutes after that, the capsule hits the water at a relatively gentle 7 m/s (15 mph).

  • 1
    $\begingroup$ +1 for an excellent answer! About the first 18 hours drifting away from the ISS; it might be useful to allow for some small amount of phasing, systems checkout, making sure there will be daylight at the landing time and place, and making sure the exhaust products from the retro-propulsive burn don't affect any of the readings from the scientific instruments on the ISS, but those are just guesses. $\endgroup$
    – uhoh
    Aug 3, 2020 at 5:25
  • 1
    $\begingroup$ I have only heard them refer to "closing the nose cone", not to "deploying" it. In fact, that is the term they use for opening it, if I remember correctly, not closing it. As to "what's up with the nose cone": it protects the forward hatch, the docking system, and the GNC system. Which then poses the question: if there's so much sensitive stuff in there, why don't they close it earlier? It also houses the four forward bulkhead thrusters, which are the most efficient thrusters, and those are used for the departure burn. $\endgroup$ Aug 3, 2020 at 6:01
  • $\begingroup$ Great answer, thanks! Can I ask for the altitudes at the beginning of every stage? If possible, I would also ask for falling speed and inclination at every stage. Also the speed when touching the water - is it still 53 m / sec or its already slower than that? $\endgroup$
    – Joe Jobs
    Aug 3, 2020 at 8:26
  • $\begingroup$ @JörgWMittag I agree that “deploying” is a strange term to use, but it’s the term used in the descent timeline post from NASA yesterday. $\endgroup$ Aug 3, 2020 at 15:34
  • 1
    $\begingroup$ I noticed that after I wrote my comment. It is, however, the only place I have seen that. The term wasn't used once in the radio communications, nor in the commentary, nor in anything written by SpaceX. SpaceFlightNow also calls it "close". That blog post is literally the only instance of this usage in the entirety of Google's index. $\endgroup$ Aug 3, 2020 at 15:51

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.