17

So I went to the link in a recent question and discovered SpaceX turns its first stage around while it is going, what, several km/s in the upper atmosphere, restarts 3 of the engines and fires retrograde to slow the stage down and get it heading back towards the launchpad. Aside from suddenly realizing I had absolutely failed to appreciate the distances involved in this maneuver, I had a real 'Wait, what?' moment when thinking about turning around a mostly hollow tube at those speeds in even a thin atmosphere, aiming it precisely retrograde, and keeping it aimed that way as it slows down without losing control.

That sounds stupendously difficult. Is it? How do they manage it? (The link is to a good article in Aviation Week about the data NASA collected from the September CRS-4 launch during the retrograde burn, which the hope to use for design of propulsive deceleration on Mars.)

kim holder
  • 21,377
  • 9
  • 85
  • 188
  • I really thought that part of the challenge would be your exhaust gases heating up the air you're about to fly through... but perhaps that's insignificant compared to compression heating, and there seem to be enough challenges besides that. – craq Feb 14 '19 at 00:43

1 Answers1

18

  1. Having the business end of the rocket survive the dynamic pressure and heating of facing into the flow, where in general that part of the rocket is not designed to be aerodynamic. This is a challenge whether or not the engines are firing. Having the engines running can actually help a little here, rejecting flow into those nozzles, but they are not running all of the engines.

  2. Starting an engine with a supersonic flow impinging on the nozzle. This may or may not be a challenge, but you don't really know until you try. It is, to put it gently, problematic to simulate or test on the ground.

  3. Pointing the stick into the wind before you enter is not too difficult, but keeping it pointed into the wind is critical. Just a little bit off and the side forces on the fuselage can overwhelm the control authority of the gimbaled nozzles, flip the vehicle, and subject it to side forces that can break up the vehicle. You try to make the launch vehicle structure as light as possible, so it isn't designed to take full-on dynamic pressure loads from the side.

  4. Once in the supersonic flow, the aerodynamic effects of the messy business end of the rocket can be complicated, making control a bit of a challenge. You can have counterintuitive effects that redirect flow in unexpected directions at different angles of attack.

  5. Predicting the effect of the running engines on the drag is a challenge. The thrust plumes tend to reduce the drag, countering in part the intent of firing the engines to increase deceleration. With enough of a thrust to drag ratio, this is not a show stopper, but you need to be able to predict how large the effect is to know if you have enough fuel. This impact on drag also complicates what happens when you gimbal the engine, which is part of the challenge in #4. Again, high thrust to drag ratio can reduce the surprises here.

Mark Adler
  • 58,160
  • 3
  • 171
  • 251
  • 5
    Wow, now the achievement of SpaceX looks absolutely brilliant. – Quazi Irfan Oct 26 '14 at 05:33
  • 1
    @iamcreasy Thanks for the link. Watching that happen is pretty cool. Getting sufficient data from a launch to tweak this design looks like a huge challenge in itself. Even running an accurate virtual simulation strikes me as difficult. – kim holder Oct 26 '14 at 20:15
  • Surely it would be implemented in a manner similar to reverse-thrust on jet engines; redirecting the nozzles instead of a 180 on the air-frame ... ? – Everyone Oct 27 '14 at 13:00
  • 1
    No. It's a 180 on the airframe. – Mark Adler Oct 27 '14 at 15:56
  • The vid Iamcreasy linked to has the caption "Mars relevant retropropulsion regime" at around 1:50. This surprises me. I guess 70 km altitude would have density comparable to Mars atmosphere. But Mars entry velocity would be around 6 km/s, a good bit higher than what I imagine the Falcon 1st stage enters. – HopDavid Apr 28 '16 at 14:18
  • 1
    At Mars you don't fire the engines at entry. You let the atmosphere slow the vehicle down as much as possible, and wait until you're pretty close to the ground, and going much slower than at entry, to turn on the engines. – Mark Adler Apr 28 '16 at 15:42
  • 1
    "Starting an engine with a supersonic flow coming up the nozzle" - the supersonic flow isn't "coming up" the nozzle, as the chamber is otherwise sealed, meaning there is no exit. Instead, it will form a standing shock at the base of the engine bell, which the flow will spill around. – Skyler Feb 21 '18 at 14:52
  • @MarkAdler: "as much as possible" for a large vehicle on Mars is not as much as you'd like. Terminal velocity of a human (for example) is more or less the speed of sound, and the square-cube means bigger payloads will generally fall faster. – Steve Linton Aug 12 '18 at 16:32
  • @SteveLinton That's why for Mars, the landing and roving vehicles' EDL systems always deploy big parachutes before initiating the retropropulsion phase. – Tom Spilker Aug 12 '18 at 16:53
  • 2
    One minor advantage to the stick's geometry and mass properties: after separation the engines are the highest mass concentration of the whole assembly, this concentrates mass at the "tail", moving the CM aft and making the stick more or less aerodynamically stable engines-first. However, that is only mildly stable, not rock-solid, and without active control would allow large attitude excursions. – Tom Spilker Aug 12 '18 at 16:57
  • Sure, but parachutes are single-use (more or less) and a lot of mass. I think the point of the research into retro-propulsion is to get a more reusable solution (or land even bigger payloads). – Steve Linton Aug 12 '18 at 16:59
  • @SteveLinton: you can still get from ~7.5 km/s down to less than 1 km/s (http://spaceflight101.com/spx/wp-content/uploads/sites/113/2017/09/IAC2017-Musk-33.jpg). The struggle with Mars probes has been to get the landing vehicle subsonic so it can open parachutes, since the feasibility of supersonic retropropulsion was an unknown. The "Mars relevant retropropulsion regime" was relevant because it's similar to the supersonic flight conditions for a large craft doing a powered landing on Mars...you don't need to get subsonic first if supersonic retropropulsion works. – Christopher James Huff Aug 12 '18 at 19:01
  • @SteveLinton "as much as possible" may be more than you think. You can get a 6 km/s entry down to ~1 km/s with just a blunt body (no parachute). That is exactly what was planned for the Red Dragon concept, which wouldn't fire the engines until it was a (scary) few km above the deck. – Mark Adler Aug 12 '18 at 21:21
  • I think we're agreeing furiously. My point was that you can't get subsonic, so do need to explore this regime. I never meant to suggest that you have to brake from entry velocities. – Steve Linton Aug 12 '18 at 22:51