Jump to content

Blue Origin Thread (merged)


Aethon

Recommended Posts

4 hours ago, Mitchz95 said:

With an impact like that, so far off to the side, I'm kinda surprised the barge didn't capsize.

Large flat barges are *VERY* hard to capsize - especially with something as trivial (compared to the barge's displacement) as an F9 stage hitting it at fairly modest velocity.

Link to comment
Share on other sites

7 hours ago, StrandedonEarth said:

But they need to if they're ever going to catch up on their backlog. That accident really set them back because I imagine they had to rebuild all their rockets, since they use that suspect strut all over the place. Once they have an inventory of ready-to-fly rockets again, they should be able to pick up the pace, especially if they use Vandy more and once Boca Chica comes online.

I would argue not, unless another disaster happens, as they'd had a lot of time to rebuild the rockets, and LC-39A is going to be operational in about a year, taking significant pressure off LC-40.

Link to comment
Share on other sites

The falcon heavy was promised to take flight in 2014, then pushed back to second quarter of 2015, and now 2016. I suspect they are struggling with the rocket launch schedule so i see no reason as to why they can meet all events as indicated in the timeline.

Link to comment
Share on other sites

I don't expect them to make the schedule either. That's...ambitious. However, I'd love for them to prove us all wrong, and it wouldn't be the first time Elon Musk has defied expectations over his deadlines.

Link to comment
Share on other sites

1 hour ago, Jovus said:

I don't expect them to make the schedule either. That's...ambitious. However, I'd love for them to prove us all wrong, and it wouldn't be the first time Elon Musk has defied expectations over his deadlines.

Indeed. But another accident and spaceX would be screwed over.

Link to comment
Share on other sites

13 hours ago, KerbonautInTraining said:

Just to be sure, still no landing footage right?

At this point I suspect we're not going to see it. The booster apparently came down very fast given that it punched a hole in the barge deck so maybe there's only a few frames of video anyway.  

On the other hand SpaceX is both very stingy with information and very image conscious.  The previous video of failed barge landings at least showed a progression from near miss to almost but not quite landing before falling over and exploding to landing safely and then falling over and exploding.  Whacking the barge at high speed does not fit into that narrative so no video (even though they said ahead of time that they did not expect success because of the degree of difficulty of the attempt).

Link to comment
Share on other sites

1 hour ago, KerBlammo said:

At this point I suspect we're not going to see it. The booster apparently came down very fast given that it punched a hole in the barge deck so maybe there's only a few frames of video anyway.  

On the other hand SpaceX is both very stingy with information and very image conscious.  The previous video of failed barge landings at least showed a progression from near miss to almost but not quite landing before falling over and exploding to landing safely and then falling over and exploding.  Whacking the barge at high speed does not fit into that narrative so no video (even though they said ahead of time that they did not expect success because of the degree of difficulty of the attempt).

In all fairness, they did say that they expected the landing to fail.

Link to comment
Share on other sites

34 minutes ago, Rdivine said:

In all fairness, they did say that they expected the landing to fail.

Which meant that it was a pretty stupid idea to put the drone barge on course with a landing attempt that was expected to fail. Sometimes, it seems that SpaceX has a Kerbal development process.

You learn from your failures. To a point. There is also a point beyond which your failures cost more money.

Most aerospace companies do this sort of testing in computer simulations. Only when the design is frozen, and when they are 100% confident, do they actually build a working prototype. You don't see Boeing or Airbus blowing stuff up to see if it works.

Edited by Nibb31
Link to comment
Share on other sites

14 minutes ago, Nibb31 said:

Which meant that it was a pretty stupid idea to put the drone barge on course with a landing attempt that was expected to fail. Sometimes, it seems that SpaceX has a Kerbal development process.

You learn from your failures. To a point. There is also a point beyond which your failures cost more money.

Most aerospace companies do this sort of testing in computer simulations. Only when the design is frozen, and when they are 100% confident, do they actually build a working prototype. You don't see Boeing or Airbus blowing stuff up to see if it works.

I'm sure Boeing simulated the heck out of the Dreamliner too. Didn't stop problems showing up once the design had been frozen. Likewise, why bother with test flying new planes if the almighty computer says 'yes this will work'.

I'm pretty sure (without a shred of evidence to support this mind) that SpaceX would have simulated that last landing. If it didn't have a hope of working then they wouldn't have bothered. As it was, they very nearly pulled it off, so I'm guessing that the simulations were too close to call and they decided to run the test and get the extra data.

Link to comment
Share on other sites

20 minutes ago, Nibb31 said:

Which meant that it was a pretty stupid idea to put the drone barge on course with a landing attempt that was expected to fail. Sometimes, it seems that SpaceX has a Kerbal development process.

You learn from your failures. To a point. There is also a point beyond which your failures cost more money.

Most aerospace companies do this sort of testing in computer simulations. Only when the design is frozen, and when they are 100% confident, do they actually build a working prototype. You don't see Boeing or Airbus blowing stuff up to see if it works.

Yes, but it there are too many factors involved in landing a rocket back down on Earth, so the only chance to gather good data is to use chances like this to *try* to land a first stage. Crashing a first stage on a drone ship can yield valuable data so that they can actually land a first stage on a similar trajectory on future missions. If they don't try, they will never know if it truly works, or not.

Link to comment
Share on other sites

20 minutes ago, Nibb31 said:

Which meant that it was a pretty stupid idea to put the drone barge on course with a landing attempt that was expected to fail. Sometimes, it seems that SpaceX has a Kerbal development process.

Barges are pretty cheap.  Falcon9 landing practice costs $60M.   Slapping a few inches of steel over the hole they just put in "Of Course I Still Love You" won't sound cheap, but it won't be visible in the $60M launch cost of CRS-8.  As far as I know, they've already retired one barge.  It doesn't seem to be a concern to them.

Link to comment
Share on other sites

asdfasdf

49 minutes ago, Rdivine said:

Yes, but it there are too many factors involved in landing a rocket back down on Earth, so the only chance to gather good data is to use chances like this to *try* to land a first stage. Crashing a first stage on a drone ship can yield valuable data so that they can actually land a first stage on a similar trajectory on future missions. If they don't try, they will never know if it truly works, or not.

This. I think this sort of distance and landing profile is exactly what they'll have to pull off to land the core of the Falcon Heavy, albeit with more fuel available to slow down. I get the impression this landing was a bit of an un-dress rehearsal for just that. AFAIK, SpaceX has yet to reveal what actually went wrong, too (too much velocity/not enough fuel, mechanical error, computer error, etc), so there's quite alot of speculation being thrown around right now.

Link to comment
Share on other sites

1 hour ago, Nibb31 said:

Which meant that it was a pretty stupid idea to put the drone barge on course with a landing attempt that was expected to fail. Sometimes, it seems that SpaceX has a Kerbal development process.

I'm no SpaceX, but is a hole in a barge really worth worrying about if it gives you critical test data from a 10s of millions of dollars orbital rocket booster?

There's only so much a simulator does and can show you. And yes, even Airbus and Boeing DO blow stuff up to see if it works.  

Only difference here is that SpaceX was going to lose the extremely expensive piece of equipment anyway.

Edited by Lukaszenko
Link to comment
Share on other sites

1 hour ago, Lukaszenko said:

I'm no SpaceX, but is a hole in a barge really worth worrying about if it gives you critical test data from a 10s of millions of dollars orbital rocket booster?

There's only so much a simulator does and can show you. And yes, even Airbus and Boeing DO blow stuff up to see if it works.  

Only difference here is that SpaceX was going to lose the extremely expensive piece of equipment anyway.

i think that's part of normal jet engine certification process - they have to fully demonstrate that a broken blade can be contained inside the engine's hull. (if such a debris isn't contained, it could lead to much more severe damage to the plane - there's no Launch Escape System on a plane, so the hability to manage various critical damage has to be demonstrated to ensure safety of the passengers, while on a rocket you mostly have to prove the reliability of the escape systems (and those are demonstrated too).

if we end up building a SSTO for space tourists without a launch escape system, the kind of failure tests made on planes are going to be performed on such spacecraft too :)

 

 

 

Edited by sgt_flyer
Link to comment
Share on other sites

13 hours ago, DerekL1963 said:

Large flat barges are *VERY* hard to capsize - especially with something as trivial (compared to the barge's displacement) as an F9 stage hitting it at fairly modest velocity.

And when you punch through steel plates, you distribute the loads over a longer time.

Link to comment
Share on other sites

seems SpaceX showed their Dragon 2 touchscreen UI at GDC (for docking / RCS control mode)

i hope there's some kind of autopilot for normal operations (or that the pilot will have a better training than the guy who flew the demo ;)). especially if there's no tactile feedback, the kind you could have with at least digital sticks (analog sticks would require some kind of RCS impulse frequency control) - with those digital sticks, you could at least feel the clicking point, even with gloves.

https://www.reddit.com/r/spacex/comments/4avhuw/dragon_2_software_demo_gdc/

 

Edited by sgt_flyer
Link to comment
Share on other sites

15 minutes ago, sgt_flyer said:

seems SpaceX showed their Dragon 2 touchscreen UI at GDC (for docking / RCS control mode)

i hope there's some kind of autopilot for normal operations (or that the pilot will have a better training than the guy who flew the demo ;)). especially if there's no tactile feedback, the kind you could have with at least digital sticks (analog sticks would require some kind of RCS impulse frequency control) - with those digital sticks, you could at least feel the clicking point, even with gloves.

https://www.reddit.com/r/spacex/comments/4avhuw/dragon_2_software_demo_gdc/

 

It's DPAI!  Oh, and does SpaceX use magnetic docking ports too?

Link to comment
Share on other sites

The difference between those tests and what SpaceX is doing is that those are mandatory certification tests, and although there is a lot of breakage and fire, the tests are actually successful.

SpaceX's method is entirely different. They favor rapid prototyping instead of software modelling techniques.

Link to comment
Share on other sites

18 hours ago, Nibb31 said:

Which meant that it was a pretty stupid idea to put the drone barge on course with a landing attempt that was expected to fail. Sometimes, it seems that SpaceX has a Kerbal development process.

You learn from your failures. To a point. There is also a point beyond which your failures cost more money.

Most aerospace companies do this sort of testing in computer simulations. Only when the design is frozen, and when they are 100% confident, do they actually build a working prototype. You don't see Boeing or Airbus blowing stuff up to see if it works.

The testing is pretty cheap its not like they can use the rocket afterward if they don't test. They have to deploy the barge, do some welding if stage crashes in it.
At least they get some data and experience, unlike building passenger planes we don't have much data or experience with powered landings of rocket stages.
 

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...