Author Topic: SpaceX will almost certainly have another failure within the next 3 years.  (Read 93696 times)

Online steveleach

  • Senior Member
  • *****
  • Posts: 2248
  • Liked: 2821
  • Likes Given: 975
Iím going to risk re-opening this thread Ö will lock if we canít stay on-topic and be respectful.

Since this thread was started in December 2016, SpaceX has flown nearly 300 (297?) consecutive successful Falcon (F9+FH) orbital missions.

Any thoughts on the key factors that have enabled this?

Clearly the rocket designs are sound. Re-use, particularly recovery and inspection of flown boosters, is important. But SpaceX have also had to ramp S2 & Mvac production, as well as significantly reduce pad turn around times and sustain record breaking and still increasing launch cadence that is currently about 3 launches per week.

In short, why hasnít SpaceX had another Falcon failure?
I think it is simple: SpaceX have a culture of continuous improvement, and don't subscribe to "if it ain't broke..."

Given it is effectively impossible to maintain a perfect zero-improvement/zero-degradation balance in the face of constant external change, if you aren't improving then you are - almost by definition - degrading.

This applies just as much to their manufacturing and quality assurance processes as to the products themselves.

Offline edkyle99

  • Expert
  • Senior Member
  • *****
  • Posts: 15462
    • Space Launch Report
  • Liked: 8669
  • Likes Given: 1369
Since this thread was started in December 2016, SpaceX has flown nearly 300 (297?) consecutive successful Falcon (F9+FH) orbital missions.

Any thoughts on the key factors that have enabled this?
As of April 7, 2024 (after Starlink 8-1) SpaceX has had 290 consecutive Falcon 9 orbital successes since the AMOS 6 static test failure and 299 consecutive Falcon 9 orbital launch successes since CRS-17 (June 28, 2015).  They also had one suborbital Falcon 9 success.  And 9 Falcon Heavy successes.  They've lost about 11 boosters (~3.5%) during landing or recovery attempts since 2016, which is its own form of failure now perhaps.

Why the successes?  Merlin 1D.  Good design.  Testing.  And Merlin 1D.

 - Ed Kyle
« Last Edit: 04/07/2024 02:42 pm by edkyle99 »

Online steveleach

  • Senior Member
  • *****
  • Posts: 2248
  • Liked: 2821
  • Likes Given: 975
Since this thread was started in December 2016, SpaceX has flown nearly 300 (297?) consecutive successful Falcon (F9+FH) orbital missions.

Any thoughts on the key factors that have enabled this?
As of April 6, 2024 (after Starlink 6-47) SpaceX has had 289 consecutive Falcon 9 orbital successes since the AMOS 6 static test failure and 298 consecutive Falcon 9 orbital launch successes since CRS-17 (June 28, 2015).  They also had one suborbital Falcon 9 success.  And 9 Falcon Heavy successes.  They've lost about 11 boosters (~3.5%) during landing or recovery attempts since 2016, which is its own form of failure now perhaps.

Why the successes?  Merlin 1D.  Good design.  Testing.  And Merlin 1D.

 - Ed Kyle
Yeah, they were really fortunate to be given Merlin 1D.

Offline rsdavis9

Maybe the software does a better job of self test before launch. Monitor more hardware for something out of family. I could imagine software so complete that I as an idiot human could not screw it up.
With ELV best efficiency was the paradigm. The new paradigm is reusable, good enough, and commonality of design.
Same engines. Design once. Same vehicle. Design once. Reusable. Build once.

Offline edkyle99

  • Expert
  • Senior Member
  • *****
  • Posts: 15462
    • Space Launch Report
  • Liked: 8669
  • Likes Given: 1369
Yeah, they were really fortunate to be given Merlin 1D.
"given"?

 - Ed Kyle

Offline abaddon

  • Senior Member
  • *****
  • Posts: 3091
  • Liked: 3968
  • Likes Given: 5394
Yeah, they were really fortunate to be given Merlin 1D.
"given"?

 - Ed Kyle
Best not to engage, you didnít say anything wrong.

On topic, other launch providers have had impressive records of success, e.g. Atlas.  What makes the SpaceX record to date so impressive is volume and speed, but those have some benefits to reliability (as long as you donít scale too fast too far) as well as booster reuse helping reliability as well.

Super impressive, in any case.

Offline RedLineTrain

  • Senior Member
  • *****
  • Posts: 2529
  • Liked: 2457
  • Likes Given: 10387
I think that it's mostly rocket design, with a side of desire to iterate.

The Merlin 1Ds are so reliable because SpaceX has built, tested, and launched so many of them.  SpaceX has been able to gain experience with them 5-10x more quickly than other rocket engine builders have with their engines.  I don't believe that engine-out capability has been exercised with the 1Ds, but that is an additional benefit.

This is also why it seems likely that Raptor will have a wonderful career ahead of it.  It should be an iron law of rocketry that more engines is better.  Preferably, more engines on both stages.  Starship will have 42 engines on it and it seems likely that Starship eventually will be considered a good design.  So the upper bound of this desirability is unknown.
« Last Edit: 04/07/2024 04:22 pm by RedLineTrain »

Online steveleach

  • Senior Member
  • *****
  • Posts: 2248
  • Liked: 2821
  • Likes Given: 975
Yeah, they were really fortunate to be given Merlin 1D.
"given"?

 - Ed Kyle
Sarcasm, but you knew that.

Your post suggests that the success of Falcon was due to the Merlin engine, but given that engine was developed by SpaceX it would be more accurate to say that the culture and engineering practices employed by SpaceX result in a very reliable launch vehicle, and the very reliable engine is just part of that (admittedly one of the most critical).

Offline king1999

  • Full Member
  • ****
  • Posts: 441
  • F-Niner Fan
  • Atlanta, GA
  • Liked: 306
  • Likes Given: 1284
Since this thread was started in December 2016, SpaceX has flown nearly 300 (297?) consecutive successful Falcon (F9+FH) orbital missions.

Any thoughts on the key factors that have enabled this?
As of April 7, 2024 (after Starlink 8-1) SpaceX has had 290 consecutive Falcon 9 orbital successes since the AMOS 6 static test failure and 299 consecutive Falcon 9 orbital launch successes since CRS-17 (June 28, 2015).  They also had one suborbital Falcon 9 success.  And 9 Falcon Heavy successes.  They've lost about 11 boosters (~3.5%) during landing or recovery attempts since 2016, which is its own form of failure now perhaps.

Why the successes?  Merlin 1D.  Good design.  Testing.  And Merlin 1D.

 - Ed Kyle

It is Elon. During one of his interviews, Tom Mueller mentioned that Elon gave some important inputs for M1D's design which Tom was initially skeptical or even resistive, but turned out to be key to M1D's success.
« Last Edit: 04/09/2024 02:30 am by king1999 »

Online meekGee

  • Senior Member
  • *****
  • Posts: 14332
  • N. California
  • Liked: 14201
  • Likes Given: 1398
Since this thread was started in December 2016, SpaceX has flown nearly 300 (297?) consecutive successful Falcon (F9+FH) orbital missions.

Any thoughts on the key factors that have enabled this?
As of April 7, 2024 (after Starlink 8-1) SpaceX has had 290 consecutive Falcon 9 orbital successes since the AMOS 6 static test failure and 299 consecutive Falcon 9 orbital launch successes since CRS-17 (June 28, 2015).  They also had one suborbital Falcon 9 success.  And 9 Falcon Heavy successes.  They've lost about 11 boosters (~3.5%) during landing or recovery attempts since 2016, which is its own form of failure now perhaps.

Why the successes?  Merlin 1D.  Good design.  Testing.  And Merlin 1D.

 - Ed Kyle
3.5% failure on recovery suggests to a casual reader that those failures are somehow part of a statistical distribution and have something to do with the reliability of recovery.

You know that it doesn't, and yet you keep making those statements.  The failures occurred during the development phase, and are clustered at the beginning.

Data is only as impartial as the person collating it.
ABCD - Always Be Counting Down

Offline envy887

  • Senior Member
  • *****
  • Posts: 8150
  • Liked: 6816
  • Likes Given: 2965
Since this thread was started in December 2016, SpaceX has flown nearly 300 (297?) consecutive successful Falcon (F9+FH) orbital missions.

Any thoughts on the key factors that have enabled this?
As of April 7, 2024 (after Starlink 8-1) SpaceX has had 290 consecutive Falcon 9 orbital successes since the AMOS 6 static test failure and 299 consecutive Falcon 9 orbital launch successes since CRS-17 (June 28, 2015).  They also had one suborbital Falcon 9 success.  And 9 Falcon Heavy successes.  They've lost about 11 boosters (~3.5%) during landing or recovery attempts since 2016, which is its own form of failure now perhaps.

Why the successes?  Merlin 1D.  Good design.  Testing.  And Merlin 1D.

 - Ed Kyle
3.5% failure on recovery suggests to a casual reader that those failures are somehow part of a statistical distribution and have something to do with the reliability of recovery.

You know that it doesn't, and yet you keep making those statements.  The failures occurred during the development phase, and are clustered at the beginning.

Data is only as impartial as the person collating it.

To put a finer point on it, if the true success rate is only 96.5%, then the probability of actually getting the ~209 consecutive successful landings which SpaceX has demonstrated is only 0.05%.

To have a 50/50 shot at 209 consecutive successes, the true probability of success on each attempt needs to be 99.7%.
« Last Edit: 04/09/2024 01:03 pm by envy887 »

Offline AmigaClone

3.5% failure on recovery suggests to a casual reader that those failures are somehow part of a statistical distribution and have something to do with the reliability of recovery.

You know that it doesn't, and yet you keep making those statements.  The failures occurred during the development phase, and are clustered at the beginning.

Data is only as impartial as the person collating it.

We were only talking about post-2016, so I did not include the first half-dozen or so landing failures that occurred during that early development phase. 

3.5% is just to show the relative losses over that time since the end of 2016.  That is a good number, IMO, statistics or not.  My post was showing how highly reliable Falcon 9 has been and I get jumped on.  Sheesh. 

 - Ed Kyle

List of boosters lost (as opposed to expended) since 1 January 2017. Except where noted, the loss was caused by a failed Falcon 9 Block 5 droneship landing attempt.

B1033.1 - Failed FH droneship landing (6 February 2018).
B1048.5 - Failed droneship landing (18 March 2020).
B1050.1 - Only failed RLTS landing attempt (5 December 2018).
B1055.1 - Only successful FH core landing - lost during transit to Port Canaveral (11 April 2019).
B1056.4 - Failed droneship landing (17 February 2020).
B1057.1 - Failed FH core droneship landing (25 June 2019).
B1058.19 - Successful droneship landing - lost during transit to Port Canaveral (23 December 2023).
B1059.6 - Failed droneship landing (16 February 2021)

Online meekGee

  • Senior Member
  • *****
  • Posts: 14332
  • N. California
  • Liked: 14201
  • Likes Given: 1398
3.5% failure on recovery suggests to a casual reader that those failures are somehow part of a statistical distribution and have something to do with the reliability of recovery.

You know that it doesn't, and yet you keep making those statements.  The failures occurred during the development phase, and are clustered at the beginning.

Data is only as impartial as the person collating it.

We were only talking about post-2016, so I did not include the first half-dozen or so landing failures that occurred during that early development phase. 

 - Ed Kyle
I'd be happy to stand corrected, but how many of the 3.5% occurred, say, during the first 20% of the operational history?

Because the data you choose should be usable to reflect the reliability of the system we're discussing, which was the whole point, right?
ABCD - Always Be Counting Down

Offline abaddon

  • Senior Member
  • *****
  • Posts: 3091
  • Liked: 3968
  • Likes Given: 5394
3.5% failure on recovery suggests to a casual reader that those failures are somehow part of a statistical distribution and have something to do with the reliability of recovery.

You know that it doesn't, and yet you keep making those statements.  The failures occurred during the development phase, and are clustered at the beginning.

Data is only as impartial as the person collating it.

We were only talking about post-2016, so I did not include the first half-dozen or so landing failures that occurred during that early development phase. 

3.5% is just to show the relative losses over that time since the end of 2016.  That is a good number, IMO, statistics or not.  My post was showing how highly reliable Falcon 9 has been and I get jumped on.  Sheesh. 

 - Ed Kyle

List of boosters lost (as opposed to expended) since 1 January 2017. Except where noted, the loss was caused by a failed Falcon 9 Block 5 droneship landing attempt.

B1033.1 - Failed FH droneship landing (6 February 2018).
B1048.5 - Failed droneship landing (18 March 2020).
B1050.1 - Only failed RLTS landing attempt (5 December 2018).
B1055.1 - Only successful FH core landing - lost during transit to Port Canaveral (11 April 2019).
B1056.4 - Failed droneship landing (17 February 2020).
B1057.1 - Failed FH core droneship landing (25 June 2019).
B1058.19 - Successful droneship landing - lost during transit to Port Canaveral (23 December 2023).
B1059.6 - Failed droneship landing (16 February 2021)
A few additional notes on the most recent landing failures, as per Wiki:
B1056.4 - Failed to land due to incorrect wind data
B1048.5 - Premature booster engine shutdown on ascent due to residual cleaning fluid; mission success due to booster engine redundancy
B1059.6 - Engine early shut down on ascent due to hole in heat shield cover; mission success due to booster engine redundancy

The last two are particularly relevant for this thread as they relate to booster resiliency in completing the primary mission to losing an engine prematurely on ascent, both on Starlink flights which are among the heaviest payloads carried by the launcher.
« Last Edit: 04/09/2024 02:45 pm by abaddon »

Offline RedLineTrain

  • Senior Member
  • *****
  • Posts: 2529
  • Liked: 2457
  • Likes Given: 10387
Is it true that the missions of B1048.5 and B1059.6 were saved by engine out capability?  I was unaware.

Offline ZachS09

  • Space Savant
  • Senior Member
  • *****
  • Posts: 8453
  • Roanoke, TX
  • Liked: 2377
  • Likes Given: 2077
Is it true that the missions of B1048.5 and B1059.6 were saved by engine out capability?  I was unaware.

Methinks that if an engine goes out and the first stage has to burn longer on only eight engines, then thereís no way to make the recovery work at all.
Liftoff for St. Jude's! Go Dragon, Go Falcon, Godspeed Inspiration4!

Offline AmigaClone

Is it true that the missions of B1048.5 and B1059.6 were saved by engine out capability?  I was unaware.

Methinks that if an engine goes out and the first stage has to burn longer on only eight engines, then thereís no way to make the recovery work at all.

I suspect that the viability of even attempting a landing after an engine out situation might depend on the timing and location of the engine that failed. If the engine that goes out does so only a few seconds early and that is not one of the engines that would be used in the planned boostback, reentry, and landing burns then it might still be possible to attempt a landing.

I agree that in most cases it likely would not be viable to even make the landing attempt - if that was even possible.

Offline Barley

  • Full Member
  • ****
  • Posts: 1049
  • Liked: 715
  • Likes Given: 390

I suspect that the viability of even attempting a landing after an engine out situation might depend on the timing and location of the engine that failed. If the engine that goes out does so only a few seconds early and that is not one of the engines that would be used in the planned boostback, reentry, and landing burns then it might still be possible to attempt a landing.
I agree that in most cases it likely would not be viable to even make the landing attempt - if that was even possible.
What is meant by "make the landing attempt"?  If the software determines that landing is impossible the first backup plan would be a "crash" in the hazard area.  In many cases this would not be distinguishable from a failed landing attempt by an outside observer without access to detailed telemetry or access to the flight software.  I'm also not sure the software would make the land/crash decision before the last few seconds, so it may not be clear even with full access.

Offline LouScheffer

  • Senior Member
  • *****
  • Posts: 3432
  • Liked: 6205
  • Likes Given: 870
I suspect that the viability of even attempting a landing after an engine out situation might depend on the timing and location of the engine that failed. If the engine that goes out does so only a few seconds early and that is not one of the engines that would be used in the planned boostback, reentry, and landing burns then it might still be possible to attempt a landing.

I agree that in most cases it likely would not be viable to even make the landing attempt - if that was even possible.
I don't think this is correct.  The strategy (i'd think) would be to make a shorter entry burn (as long as possible while leaving fuel for landing), and take somewhat higher re-entry heating.

SpaceX tried this a few months ago, where they loaded the F9 with an extra StarLink, then accepted a shorter entry burn.  It landed OK, but they have not tried the higher payload again, so maybe the F9 got over-toasted.  But it DID land OK.

Online daedalus1

  • Full Member
  • ****
  • Posts: 941
  • uk
  • Liked: 487
  • Likes Given: 0
Isn't all this discussion pointless? The question was set in 2016 so the end of the discussion should have been 2019.

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1