Business, Space, Technology and Society

The SpaceX Launch

61 Comments

You can follow live here. Weather is looking good half an hour before liftoff, with no technical issues.

[Update after the launch]

Perfect primary mission, but they landed on the ship too hard to survive, according to Elon.

[Update almost an hour post launch]

[Update a little over an hour after the launch]

Thoughts from (SpaceX investor) Steve Jurvetson.

[Update a while later]

Lee Billings has the story over at SciAm.

[Update a while more later]

And here‘s the Space News report.

61 thoughts on “The SpaceX Launch”

  1. Since there isn’t live video from the landing barge, I’m assuming that Elon or someone else will at least be tweeting the success/failure of the landing attempt?

    Based on the timelines I’ve seen, liftoff at 4:10 EDT, touchdown should be around 4:19+/- EDT, so we should know by 4:25 EDT if recovery was successful?

    I’m almost as anxious for today’s launch as I was the day of (and time of) the SpaceShipOne X-Prize winning flight. I’m glad it’s a much shorter timeline from takeoff to success/failure for this one. 🙂

    1. “Now, did they stick the landing?”

      Very good question. It seems that the coverage is all about the CRS-6 mission, rather than the ancillary testing that was tacked on at the end.

      I would be curious to see poll results of how many people logged in to watch the launch of the resupply mission vs. how many people logged in to see how the landing went.

      I’m in the latter category. SpaceX has made launches semi-routine at this point, the landings are the interesting part.

      UPDATE:
      @elonmusk:
      “Ascent successful. Dragon enroute to Space Station. Rocket landed on droneship, but too hard for survival.”

  4. Elon Musk @elonmusk · 1m 1 minute ago
    Ascent successful. Dragon enroute to Space Station. Rocket landed on droneship, but too hard for survival.
    0 retweets 0 favorites
    Reply Retweet Favorite
    More

      1. Indeed. Bummer that they had another hard landing, but hopefully it was more upright than the first/most recent landing attempt.

  5. Congrats again to SpaceX.
    I’m excited to hear how the landing went. No word about S1 after they announced it was making its landing burn at +8:09.

    S2 and S1 separated at 4:14pm on 4/14

    1. Watching the Hawthorne control center crew a minute or so after that announcement, two guys at the same time put their hands behind their heads, stretched back exactly as they might if they had just witnessed the crash. My thought at the time.

  6. Ah well. Maybe next time.

    I hope they post another sweet vid of S1 coming in hard. If they can’t be entirely successful at least they get to see a sweet crash landing.

  7. At this point, a good landing is almost incidental. What they really need is for one of the nine engines to have survived the impact intact, and hopefully on the deck before sea water starts corroding. Data!

  9. Hardly unexpected. From what I understand their margins are kind of low and this is something no one has done before after all. They just need to keep at it. Perhaps it will require the larger Falcon 9 redesign to have enough margin for it to become possible.

    I would still like to see the crash video though. Those are always fun.

  10. Engineers will tell you that they learn more from failure than from success, but those are expensive lessons. In a sense, though, the test cost them almost nothing. They were going to throw away that stage anyway. I hope they captured plenty of data and video to help analyze what went wrong.

  11. Elon’s tweets mentioned that video won’t be available until the drone returns to base. Not sure why that couldn’t be uplinked in the interim; even at 56kbps, they might still get the video back home before the drone ship.

    It also sounds like they stuck the landing, but then tipped over due to lateral momentum.

    Longer landing legs in the future?

    Is there anything else that could reduce lateral motion once they arrive at the drone ship? The trajectory seems to indicate that it might be very unfeasible to bleed off horizontal velocity during descent. At what lateral speeds did they conduct their grasshopper (and other non-orbital) tests?

    1. I wouldn’t be surprised if the link is with a satellite phone or something similar. The bandwidth and probably the transmission costs could be atrocious.

      Grasshopper is a low altitude test vehicle so it is never going to be a faithful replica of the flight conditions. I think SpaceX has limitations due to permits with the FAA which restrict the altitude of the tests they can perform.

  12. Man looking at the picture it (barely) hit the target circle. That was really close even if they couldn’t cut down the lateral velocity properly. I think on the last time it also had too much lateral velocity when it came crashing down. I guess this time they should at least have some pieces to examine, perhaps flight data, which they can use to further improve their design. Good stuff.

    It does make one consider if it is possible to do VTVL using base first reentry like proposed in the 1960s though. If something like this happens you might need to turn the vehicle around in flight to do some lateral deceleration and that could require a completely different vehicle design, perhaps one which can flip over, like the DC-Y. It is kind of a bummer if it can’t be easily done.

    1. Well, last time, the attitude was an issue, too. The booster wasn’t upright when it touched down and things “went a little… caca…”

      So my question now is, “did it land too hard, or was there too much lateral translation?” Or, I suppose, was it both? Did a harder-than-anticipated landing damage the legs and keep them from holding up the rocket?

      Maybe just toss up some tall nets around the perimeter and let it fall over after touchdown?

    2. From the look of the picture, it’s aiming for the target circle, and probably still moving laterally toward it. It presumably just didn’t cancel enough of that velocity before it touched the deck.

      We’ll know for sure when we see the video. Either way, it looks like they’ve nailed the ‘getting back to the platform’ part and just need to work through the stopping part.

  13. The American flag in the landing photo is being held straight out by the wind. That can be hard for a tail-sitter to cope with.

    Also, shortly after separation you could see the first stage firing its thrusters. 🙂

  15. SpaceX just release chase-plane video of the landing attempt. I hate to kibbitz (since I wasn’t there), but I think their flight-control software needs work! The stage is swinging wildly off-vertical on short final, just before touchdown. There should be a virtual safety “bubble” around the drone ship where the rocket doesn’t enter unless it’s in a near-vertical and stable configuration. In this case, could they have throttled-up the Merlin and slowed the decent until the rocket’s pitch/yaw was better? I don’t know, maybe there wasn’t enough propellant left to delay the touchdown. But for the flight control software to let the rocket continue descending that close to the landing pad with that much pitch/yaw doesn’t bode well at all.

    1. That was my thought, too. Perhaps they don’t have enough fuel for another couple of seconds’ flight, but it did seem to be trying to do the rocket equivalent of that stuntman who slides cars sideways into a parking space.

    2. Obviously the software needs work but “doesn’t bode well at all” sounds like trolling rather than intelligent comment.

      1. Yes, looks more like fine tuning than a major redesign. The sad part is that it would probably have worked on land, since they wouldn’t have to try so hard to hit a precise target… they could just give it a larger area to land in.

    3. Harking back to the earlier Grasshopper tests, the flight regimen there was a whole lot less stressful and the video of those tests show a lot less speed of translations both x-y and z than these two (so-far) operational first stage recovery attempts. I think Paul may be onto something here. Wondering if we see a few more failures if we’ll start to see more activity at Spaceport America with the F9Dev2. One has to be very careful with simulations that one doesn’t get conned into believing their own BS, er I mean relying too heavily on one’s assumptions.

      1. David, It certainly appeared to me that the video from yesterday shows the F9 approaching the drone ship much faster than the old Grasshopper videos showed their descent. I’m still wondering (a SWAG) if the controllers had a minimum-fuel emergency on their hands during short final. But I doubt we’ll hear anything for at least a few days, until SpaceX can analyze the data and get a complete answer.

  16. Looks like some of my early attempts at the old “Lunar Lander” game back in the 70s. It seemed to be coming down okay until the last minute, maybe something failed and it didn’t have enough time to correct fully.

    Still, going CRASH BOOM to CRUNCH is progress.

  17. Just noticed that the steering grid fins were swinging pretty wildly at touchdown – should have been programmed to be more or less inert at that low speed. Perhaps the algorithm depended too much on them at the end of flight.

    SpaceX has a ton of material to work with after this landing. Hopefully that’s literally a ton or more -we’ll see if any wreckage remained onboard.

    1. Elon Musk didn’t say specifically what shape the stage was in, perhaps it is just lying on its side on the deck (damaged of course), but maybe still mostly intact.

    2. OK, for the record those were not the grid fins. That was the plume from a reaction thruster. Much clearer on the extended video. And detonating after falling over would indicate all the good bits likely got blown over the side. Next time!

  18. Today was a great day, a new term was coined, Falcon Punch.

    A friend plops down on the couch and knocks it into your living room wall. “Hey bro, don’t falcon punch the sofa.”

    You flop into the driver seat of your car causing it to shake. “Wow, I really falcon punched that one.”

    “I’m so tired, I am going to falcon punch the first chair I see.”

    “What are the kids doing?”

    “Oh they are out falcon punching those leaves you raked up earlier.”

  19. I’m sure the engineers working on this have thought of it, but what would be the effect of the position of GPS receiver on the F9? When I look at the center of rotation of the desceding F9, which should be pretty close the the CG, the F9 is going in a pretty straight line. Is the receiver is in the top or bottom or center of the F9 Stage 1? Wouldn’t it induce oscillations like what is in the video if it was in the top or bottom? The force inputs from the engine, & the inputs for the grid fins look like they are not convergent on a solution, as if it was chasing the solution. A situation of positive feedback? The solution has not reached a 1st order critically dampened condition. It seems like you would want position information from a location near the CG, and accelerometers at the top and bottom that limit inputs. (pitch in this case I think ) coming down the z-axis to the barge.

  20. The Sci Am article seemed to miss a lot. What was tweeted by Elon and what I have seen in video is that it did not hit ‘too hard’. It had an excess velocity vector in the x-y plane. z looked fine and right on target initially. We’ll see more later on I am sure. Th author of the article seems to have entirely missed what Elon said. Also… I have to say, Scott, Scott… what happened to you?

  21. My take… they need a minor correction in the control laws. Might have been higher or varying winds; might have been the barge rising and falling, might have been minor issues with the control inputs. Whatever it was, they now have the data and they will use it as inputs to modify the controls for next time. Feedback control systems can be a bitch for really dynamic systems.

  22. I think the deck of the barge is going to have a nice Merlin 1D blowtorch scar over the entire traverse of the deck? That rocket really makes the barge look small! Looked like the rocket blast started over the water, and rotated over most of the deck before oscillating back towards the center. This gives graphical reinforcement to the barge being unmanned. I wouldn’t want to be on it if I was inside a reinforced steel bunker below deck.!

  23. Jonathan McDowell is reporting “according to tweets by Elon Musk, valve stiction caused control problems in the final burn and the stage crash-landed”.

    1. Elon later deleted the tweet (a reply to John Carmack) about valve stiction; maybe they aren’t sure yet. There definitely doesn’t appear to be enough control authority in that final descent.

    2. Well that’s interesting. Possibly good news, or possibly serious news, since those valves get used in the ascent part of the launch as well. Changing them risks adding problems during the $$$ part of launches.

  24. I was very pleased with today’s events. First and most importantly, there was yet another flawless launch that sent Dragon to the ISS. Falcon 9 is now 17 for 17. That’s worth celebrating. Their last few launches have gone very smoothly, with few technical problems and little drama. They’re working out the bugs and getting the hang of it.

    Back in the late 50s and early 60s, the U.S. and the Soviet Union were still trying to get the hang of rocket launches. There were several failures for every successful launch.

    Today launches are pretty reliable, and failures are comparatively rare. There have been thousands of space launches since 1957, but only two attempts to soft-land a booster after an orbital launch. This one was a marked improvement over the first one in January. SpaceX just has to tweak a few things, and I’m confident that they will succeed before long.

    1. Hear, hear. And for 2015 it’s four launches in four months, with a fifth scheduled in just 9 days, compared to six launches in all of 2014. Shotwell’s prediction of 50 Falcon 9 launches before the first crewed launch in 2017 still seems ambitious, but they’re picking up the pace.

  25. The only tests I’ve seen on land show the stage taking off…rising slowly to some alt., hovering for a bit then slowly landing.

    While this is a good and necessary test, did they ever loft the thing up at such an alt that they can throttle down the engine to almost nothing, have the booster fall back and build up lots of speed as the booster falls back, and then try to land land?

    They may have done that – perhaps I just missed the video.

    But if not it would seem to me to be a necessary test.

    1. Completely pointless. There is no way to predict the rocket’s lateral velocity at touchdown – it’s supposed to be zero. There are a lot of big thrusters on the barge to keep it on station for that very reason.

      1. But it might make sense to have the barge moving with the wind so there’s less coming across the deck, while reducing the cross-wind component on the final descent.

  27. The landing vid is awesome. They nearly nailed it.

    Just a little oversteering at the last… too much angular momentum for that poor little RCS thruster on top to compensate for. It tried hard though.

    1. Are they using RCS, or are they using the guide fins and a gimbaled engine to control attitude?

      Elon (and others) said in the past that the rocket remains at rather high velocity until touchdown, which is why the guide fins remain usable until just before touchdown. Elon has also tweeted since yesterday that the flight profile isn’t programmed to hover before landing, though he didn’t say if they have the performance to do so, even if they wanted to (due to fuel requirements, I’d imagine).

      That said, even if they don’t NEED to hover right before landing, is there any benefit to increasing thrust during descent to scrub off a little more speed? Or is there even the capability to do so?

      I seem to recall that the Falcon9 burns three engines during descent, and only a single engine during the final landing sequence. I have to wonder aloud what combination, if any, might make for a softer landing.

      I would imagine that the throttle capability on the engine has a LOT to do with this, as there probably isn’t enough low-end throttle to be able to reliably use two or three engines during the final landing, but maybe they need to burn all three a little longer before switching over to the single engine?

      1. Never mind, I see the RCS blasts at the tip of the stage just before touchdown in the Vine video.

        All of my other questions still linger, like smoke clouds from my pipe while I sit comfortably in my armchair, playing next-day quarterback.

      2. I think their problem is that they can’t throttle deep enough to hover, final T/W is greater than one. They have to swoop in and get it right the first time.

        1. Shouldn’t T/W always have been >1 (or it wouldn’t have taken off in the first place)? I could see it being far greater T/W than the throttle control can handle since its primary purpose is to control throttle at a much lower T/W under the rest of the stack and more fuel, but they appear to halt the vertical velocity of S1 just fine with the thrust control they have.

          Hopefully a mostly drained tank doesn’t change the XY-axes of the CoG on the 1st stage. Thrusting slightly out of line would cause extra rotation that would be hard for RCS thrusters to compensate for at such a high T/W.

          I hope some of the pad mounted camera videos can show us what the engine end of the S1 was doing. If the lateral velocity was great enough for it to skid across the deck a bit, causing the tip-over, maybe some grippy pads on the landing leg’s feet would help. 😀

          2. Ah, I see what you were saying now, they can’t throttle down to a T/W <=1. Gotta stop vertical velocity at the same instant it hits the concrete and immediately cut engines or it will just take off again.

          3. Ya know, I think they burned off enough velocity that they could land this thing on a giant pillow without taking damage.

            Landing on tarmac is hardcore and all, but you’re going to have to lay it over eventually to attach new stages and payloads.

          4. I think they burned off enough velocity that they could land this thing on a giant pillow without taking damage.

            At one point, back in the 90s, Kistler were talking about dropping their proposed SSTO into a net when it returned to Earth. But I believe that was a much smaller vehicle.

  28. Any news on damage to the ASDS?
    I can’t read the instructions if the page is torn and scorched.

  30. I was expecting the rocket equivalent of a ‘flare’ before landing. Seems they’re just trying to slam that thing in on the bounce.

    1. No point in “flaring” when they were already at max thrust for landing. It’s a software challenge to cut the main engines off just in time for a soft touchdown.

Comments are closed.