5 ms·

NASA, SpaceX Share Data on Supersonic Retropropulsion

91 points by elektropionir 12y ago

clebio 12y ago
> That produced raw images in which the stage appeared 1 pixel wide and 10 pixels long, but subsequent enhancing by specialists at the Johns Hopkins University Applied Physics Laboratory improved the resolution dramatically. So, the final resolution was something like... 10x100 pixels? How do you enhance a 1 pixel image? I'm guessing the two aircraft, at two different altitudes, help ... you get 2 pixels instead of 1.
- 7952 12y ago
  Maybe they just resample the image to a higher resolution using interpolation. It will not add addittional information but it may make the image easier to analyse and understand. Sometimes an interpolated image will have less error from the actual phenomena than the low resolution version (in the same way that a compressed mp3 is a better reproduction than an uncompressed version of a similar file size).
- icegreentea 12y ago
  Probably some fun super-resolution (basically, frame stacking with stats!) work. It would trade away some time resolution for spatial resolution, but that could be what they needed.
- smoyer 12y ago
  It's (probably) an unstable video, so you take the data from each frame for some large fraction of a second and you can create an image that's somewhere between the total number of available pixels in all the frames and the number in a single frame. Note: While the spacecraft itself is moving at incredible speeds, the maneuvers themselves are slow enough that several (to many with a high speed camera) images can be combined.
- sp332 12y ago
  They know what the stage looks like, they just pasted a high-res photo on it :p
- phkahler 12y ago
  I had copied that exact text to clipboard to ask the same question. I would guess it's dependent on having video rather than a still but I really don't know.
- liquidise 12y ago
  Obligatory "uncrop" clip https://www.youtube.com/watch?v=JMIHNiR3CP8 https://www.youtube.com/watch?v=JMIHNiR3CP8
smoyer 12y ago
"Data-sharing deal will help SpaceX land Falcon 9 on Earth and NASA put humans on Mars" I find the sub-title a bit ironic as (right now) I'd bet that SpaceX astronauts will beat NASA astronauts to Mars.
- melling 12y ago
  Hard to say and kind of irrelevant. The ideal scenario is that they both make it. In fact, if China and Russia could do it, that'd be even better.
  - kivihiinlane 12y ago
    I'm curious, why is it better if Russia and China do it first?
    - hucker 12y ago
      Because then space exploration would be a political issue again, and we'd have another space race (hopefully)
    - smoyer 12y ago
      I believe @melling implied that it would be better if Russia and China could also get humans to Mars.
      - bhhaskin 12y ago
        The U.S. is just so far ahead in terms of space technology. I am sure one day other countries will catch up, but it is going to be awhile.
        angersock 12y ago
        It's a matter of will, not necessarily of tech. The US has shown it lacks the will to achieve its former dominance.
  - 12y ago
lscharen 12y ago
"The two midwave IR sensors—mounted in a nose pod on the WB-57 and internally on the P-3—were about 60 nm from the rocket when it reignited its engines for supersonic retropropulsion." They clearly mean Nautical Miles (nm) here, but I did a double-take on my first read through.
- worklogin 12y ago
  Nautical Mile is shortened to NM, or nml. -DDG
martinwastaken 12y ago
I enjoyed Elon's interview on TED. Just when it starts sounding like he's smoking a dream-pipe about recycling rockets, he shows a clip of a test landing. Early days, but it's amazing. I agree with @smoyer, the guys at SpaceX are ninjas, I wouldn't be surprised, because they're making insane advances in a relatively short time-frame. Can't wait to see what they do next.
BrandonMarc 12y ago
This video shows IR footage from one of the planes: https://www.youtube.com/watch?v=_UFjK_CFKgA https://www.youtube.com/watch?v=_UFjK_CFKgA
ChuckMcM 12y ago
SpaceX got a lot of help from NASA, especially in regards to pintle engines[1] and this seems like a great way to pay back that transfer of expertise by moving the ball forward. My assumption (clearly incorrect) was that descent could be managed subsonically as the vertical speed at apogee is 0 but of course that stage has been pushing the second stage into its position and so it travelling down range at several k/s. Presumably you could deploy some sort of upper drag system which would use atmospheric drag to reorient the stage but you still have the fact that its travelling super sonically when you light the engines. I would love to see the CFD simulation of a exhaust nozzle when going that fast backwards! And then you light it off! I really hope that at some point they figure out how to glue a camera on to that stage so we can see what that looks like. [1] http://blog.nss.org/?p=1900 http://blog.nss.org/?p=1900

91 points by elektropionir 12y ago

> That produced raw images in which the stage appeared 1 pixel wide and 10 pixels long, but subsequent enhancing by specialists at the Johns Hopkins University Applied Physics Laboratory improved the resolution dramatically. So, the final resolution was something like... 10x100 pixels? How do you enhance a 1 pixel image? I'm guessing the two aircraft, at two different altitudes, help ... you get 2 pixels instead of 1.

7952 12y ago
Maybe they just resample the image to a higher resolution using interpolation. It will not add addittional information but it may make the image easier to analyse and understand. Sometimes an interpolated image will have less error from the actual phenomena than the low resolution version (in the same way that a compressed mp3 is a better reproduction than an uncompressed version of a similar file size).
icegreentea 12y ago
Probably some fun super-resolution (basically, frame stacking with stats!) work. It would trade away some time resolution for spatial resolution, but that could be what they needed.
smoyer 12y ago
It's (probably) an unstable video, so you take the data from each frame for some large fraction of a second and you can create an image that's somewhere between the total number of available pixels in all the frames and the number in a single frame. Note: While the spacecraft itself is moving at incredible speeds, the maneuvers themselves are slow enough that several (to many with a high speed camera) images can be combined.
sp332 12y ago
They know what the stage looks like, they just pasted a high-res photo on it :p
phkahler 12y ago
I had copied that exact text to clipboard to ask the same question. I would guess it's dependent on having video rather than a still but I really don't know.
liquidise 12y ago
Obligatory "uncrop" clip https://www.youtube.com/watch?v=JMIHNiR3CP8 https://www.youtube.com/watch?v=JMIHNiR3CP8

smoyer 12y ago

"Data-sharing deal will help SpaceX land Falcon 9 on Earth and NASA put humans on Mars" I find the sub-title a bit ironic as (right now) I'd bet that SpaceX astronauts will beat NASA astronauts to Mars.

melling 12y ago
Hard to say and kind of irrelevant. The ideal scenario is that they both make it. In fact, if China and Russia could do it, that'd be even better.
- kivihiinlane 12y ago
  I'm curious, why is it better if Russia and China do it first?
  - hucker 12y ago
    Because then space exploration would be a political issue again, and we'd have another space race (hopefully)
  - smoyer 12y ago
    I believe @melling implied that it would be better if Russia and China could also get humans to Mars.
    - bhhaskin 12y ago
      The U.S. is just so far ahead in terms of space technology. I am sure one day other countries will catch up, but it is going to be awhile.
      - angersock 12y ago
        It's a matter of will, not necessarily of tech. The US has shown it lacks the will to achieve its former dominance.
- 12y ago

lscharen 12y ago

"The two midwave IR sensors—mounted in a nose pod on the WB-57 and internally on the P-3—were about 60 nm from the rocket when it reignited its engines for supersonic retropropulsion." They clearly mean Nautical Miles (nm) here, but I did a double-take on my first read through.

worklogin 12y ago
Nautical Mile is shortened to NM, or nml. -DDG

martinwastaken 12y ago

I enjoyed Elon's interview on TED. Just when it starts sounding like he's smoking a dream-pipe about recycling rockets, he shows a clip of a test landing. Early days, but it's amazing. I agree with @smoyer, the guys at SpaceX are ninjas, I wouldn't be surprised, because they're making insane advances in a relatively short time-frame. Can't wait to see what they do next.

BrandonMarc 12y ago

This video shows IR footage from one of the planes: https://www.youtube.com/watch?v=_UFjK_CFKgA https://www.youtube.com/watch?v=_UFjK_CFKgA

ChuckMcM 12y ago

SpaceX got a lot of help from NASA, especially in regards to pintle engines[1] and this seems like a great way to pay back that transfer of expertise by moving the ball forward. My assumption (clearly incorrect) was that descent could be managed subsonically as the vertical speed at apogee is 0 but of course that stage has been pushing the second stage into its position and so it travelling down range at several k/s. Presumably you could deploy some sort of upper drag system which would use atmospheric drag to reorient the stage but you still have the fact that its travelling super sonically when you light the engines. I would love to see the CFD simulation of a exhaust nozzle when going that fast backwards! And then you light it off! I really hope that at some point they figure out how to glue a camera on to that stage so we can see what that looks like. [1] http://blog.nss.org/?p=1900 http://blog.nss.org/?p=1900