Some preliminary report findings on the Falcon failure, indicating an upper stage liquid oxygen tank rupture, caused by a strut failure.
"When you’ve only seen success, you aren’t as afraid of failure," he said."
Ah, yes. Oddly enough, I think this was in a sense a major cause of the BP Deepwater Horizon disaster. A false sense of security stemming from too many years without a truly serious event.
Some preliminary report findings on the Falcon failure, indicating an upper stage liquid oxygen tank rupture, caused by a strut failure.
There is a very large "piece of something" coming off that rocket clearly shown in the video. It is a result of the mishap not the cause.
[EDIT] I just read more of the article and the "piece of something" is referring to is the capsule I believe.
Also this comment from Musk in that article:
Musk said certification of the faulty helium bottle strut was previously the responsibility of the manufacturer. Moving forward, SpaceX will test each strut individually This could increase the cost of each Falcon flight, but according to Musk, the financial burden might not be passed on to customers.
Aha. This is definitely in the anti-intuitive category :
Quote:
Inside the upper stage liquid oxygen tank, a steel strut about two feet long and an inch wide at its thickest point appears to have snapped, Musk said. "One of those struts appears to have failed, and as a result was unable to hold the helium bottle down. So the helium bottle was shot to the top of the tank at high speed."
So there is a bottle within a bottle as it were. The helium tank is normally sited within the lower/aft part of the oxygen tank. The helium tank has less dense contents than the surrounding liquid oxygen. Normally the tendency for the helium tank to rise ( go forward/up ) is balanced by the strut mentioned. This strut acts like a seafloor anchor holding down a buoyant craft. This will be true when the rocket is sitting on the ground in its normal 'up' position : one gee worth of acceleration times the density difference b/w helium and oxygen. But with the whole craft going upwards under rocket power that rises 3.2 times. The strut breaks, the helium tank goes up ( like a submarine rapidly rising to breach the ocean surface ), hits the inside top of the oxygen tank, the helium tank is breached, the helium spills and mixes into the liquid oxygen causing a rapid pressure rise in the oxygen tank ( too fast for relief mechanisms to kick in ), the oxygen tank now breaches, like a Coke bottle with way to much fizz it will continue to splurge the oxygen out, and it all goes to the crapper from there ....
Yup, I'd call that a new variant in the rocket industry failure mode roll call. AFAIK.
Cheers, Mike.
( edit ) The normal mode here is a controlled bleed of the helium from its tank into the surrounding liquid oxygen. The helium gas rises to the top and pushes down on the oxygen to facilitate its fast transfer via tubes & pumps to the second stage rocket engine. You want to keep the input pressure to the turbo well up to stop cavitation ( pumping of nothing ) and thus engine failure.
( edit ) So assuming this is confirmed as the cause then the fix is straight forward at least : make and test a stronger strut(s).
( edit ) And a ground hold-down test will not give that 3.2g ......
( edit ) Plus the above failure takes a much shorter time to happen than to describe ! :-)
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
So assuming this is confirmed as the cause then the fix is straight forward at least : make and test a stronger strut(s).
As a former design engineer, and also a former factory reliability manager, and a yield analyst, I'd say they will (should) spread a broader net.
First, regarding this failure itself: Was the stress on the strut improperly understood in setting the design requirements? If not, was it misdesigned to meet the requirement. If allegedly all that was right but this sample was mis-manufactured, how did that come to pass, and how did it escape notice?
As you say, fixing the strut is probably straightforward. But the really tough, and really essential part, is appropriately propagating the resulting searchlight beam of suspicion around the rest of the design and manufacturing. If you carry this too far, you will never launch again. But unless carried on with adequate diligence, there is a strong chance there is another lurking problem, lulling you to complacency by almost twenty serial successes, waiting to fail then next time you are two sigma on the bad side of natural manufacturing variation.
As a design engineering manager, I used to call this killing off the brothers, sisters, aunts, and uncles, and nth cousins mth removed of the problem one had just detected. It was a bit hard to motivate people to look--though once they found something, it got a bit easier.
...killing off the brothers, sisters, aunts, and uncles, and nth cousins mth removed of the problem one had just detected. It was a bit hard to motivate people to look--though once they found something, it got a bit easier.
Absolutely. Is it the problem indeed? That immediately ought shine a spotlight on the equivalent strut in the stage one tank for starters. In paediatrics we call this the 'paranoia works' strategy, as there are virtually no diagnoses which exclude any other.
Cheers, Mike.
( edit ) To be exact you want to test the strut under 3.x+ g and cryogenic temperatures and vibration and ....
( edit ) Assuming it matters, this is also an excellent illustration of the Einstein Equivalence Principle wrt the buoyancy mechanism.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Another hypothesis is longitudinal pogo : fore to aft oscillations in length ( compress & expand ). This was never completely solved for the Saturn series, sometimes there was ~ 10Hz +/- 5g pulsing felt by the astronauts. A real highly unpleasant bone shaker. There it was managed by partial throttling on ascent ( or early shut down of the engine in the center of the cluster ). It's confined to the liquid fuel designs as pressure waves ripple up and back within the tanks by reflection off the ends. That destroyed a few Atlases before they realised the effect and with the Titan series of ICBM's it got worse with power improvements, that is with higher specific impulse engines. So could this yield a peak gee load well above design for said strut ?
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Did you guys see the report yesterday about what actually caused the last crash? Apparently a support strut broke and a helium bottle crashed violently around the compartment causing the explosion.
Ah, yes. Oddly enough, I think this was in a sense a major cause of the BP Deepwater Horizon disaster. A false sense of security stemming from too many years without a truly serious event.
This started me thinking. I think there is a lot to learn from, and expecting of, failures. In fact we probably only really learn from failures. There have been many rocket failures in the past to learn from.
There is a message in (my interpretation) "Despite the catastrophic failure of Falcon - Dragon may have survived....if parachute deployment was included in the software"
It is not the first time a spacecraft has come to grief on ascent.
Apollo (and Soyez) LES used the re-entry parachutes for escapes for decades.
Fortunately this occurred before Dragon was crewed, but "forgetting to open the parachute" is as big a design fail as the strut.
Dragon will attempt to save itself in the event of a rocket explosion
Quote:
Space X is giving its Dragon spacecraft a software upgrade so it might be able to save itself in the event of a rocket explosion.
The decision, announced on Monday by company founder Elon Musk, is one of the results of an investigation under way into the explosion last month of a Falcon 9 rocket and the loss of its Dragon cargo capsule. It had been heading to the International Space Station.
....
While the rocket was destroyed in the explosion, the Dragon spacecraft survived and continued sending back telemetry data to Space X until it went under the horizon and contact was lost, Musk said. Shortly afterward, it smashed into the ocean and sank.
"If we had deployed the parachute, we believe Dragon would have survived," Musk said in a conference call with reporters.
Some preliminary report
)
Some preliminary report findings on the Falcon failure, indicating an upper stage liquid oxygen tank rupture, caused by a strut failure.
Interestingly, they thought, had they deployed re-entry parachutes on the Dragon, it would have survived.
RE: Some preliminary report
)
"When you’ve only seen success, you aren’t as afraid of failure," he said."
Ah, yes. Oddly enough, I think this was in a sense a major cause of the BP Deepwater Horizon disaster. A false sense of security stemming from too many years without a truly serious event.
RE: RE: Some preliminary
)
There is a very large "piece of something" coming off that rocket clearly shown in the video. It is a result of the mishap not the cause.
[EDIT] I just read more of the article and the "piece of something" is referring to is the capsule I believe.
Also this comment from Musk in that article:
Musk said certification of the faulty helium bottle strut was previously the responsibility of the manufacturer. Moving forward, SpaceX will test each strut individually This could increase the cost of each Falcon flight, but according to Musk, the financial burden might not be passed on to customers.
Aha. This is definitely in
)
Aha. This is definitely in the anti-intuitive category :
So there is a bottle within a bottle as it were. The helium tank is normally sited within the lower/aft part of the oxygen tank. The helium tank has less dense contents than the surrounding liquid oxygen. Normally the tendency for the helium tank to rise ( go forward/up ) is balanced by the strut mentioned. This strut acts like a seafloor anchor holding down a buoyant craft. This will be true when the rocket is sitting on the ground in its normal 'up' position : one gee worth of acceleration times the density difference b/w helium and oxygen. But with the whole craft going upwards under rocket power that rises 3.2 times. The strut breaks, the helium tank goes up ( like a submarine rapidly rising to breach the ocean surface ), hits the inside top of the oxygen tank, the helium tank is breached, the helium spills and mixes into the liquid oxygen causing a rapid pressure rise in the oxygen tank ( too fast for relief mechanisms to kick in ), the oxygen tank now breaches, like a Coke bottle with way to much fizz it will continue to splurge the oxygen out, and it all goes to the crapper from there ....
Yup, I'd call that a new variant in the rocket industry failure mode roll call. AFAIK.
Cheers, Mike.
( edit ) The normal mode here is a controlled bleed of the helium from its tank into the surrounding liquid oxygen. The helium gas rises to the top and pushes down on the oxygen to facilitate its fast transfer via tubes & pumps to the second stage rocket engine. You want to keep the input pressure to the turbo well up to stop cavitation ( pumping of nothing ) and thus engine failure.
( edit ) So assuming this is confirmed as the cause then the fix is straight forward at least : make and test a stronger strut(s).
( edit ) And a ground hold-down test will not give that 3.2g ......
( edit ) Plus the above failure takes a much shorter time to happen than to describe ! :-)
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
RE: So assuming this is
)
As a former design engineer, and also a former factory reliability manager, and a yield analyst, I'd say they will (should) spread a broader net.
First, regarding this failure itself: Was the stress on the strut improperly understood in setting the design requirements? If not, was it misdesigned to meet the requirement. If allegedly all that was right but this sample was mis-manufactured, how did that come to pass, and how did it escape notice?
As you say, fixing the strut is probably straightforward. But the really tough, and really essential part, is appropriately propagating the resulting searchlight beam of suspicion around the rest of the design and manufacturing. If you carry this too far, you will never launch again. But unless carried on with adequate diligence, there is a strong chance there is another lurking problem, lulling you to complacency by almost twenty serial successes, waiting to fail then next time you are two sigma on the bad side of natural manufacturing variation.
As a design engineering manager, I used to call this killing off the brothers, sisters, aunts, and uncles, and nth cousins mth removed of the problem one had just detected. It was a bit hard to motivate people to look--though once they found something, it got a bit easier.
RE: ...killing off the
)
Absolutely. Is it the problem indeed? That immediately ought shine a spotlight on the equivalent strut in the stage one tank for starters. In paediatrics we call this the 'paranoia works' strategy, as there are virtually no diagnoses which exclude any other.
Cheers, Mike.
( edit ) To be exact you want to test the strut under 3.x+ g and cryogenic temperatures and vibration and ....
( edit ) Assuming it matters, this is also an excellent illustration of the Einstein Equivalence Principle wrt the buoyancy mechanism.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Another hypothesis is
)
Another hypothesis is longitudinal pogo : fore to aft oscillations in length ( compress & expand ). This was never completely solved for the Saturn series, sometimes there was ~ 10Hz +/- 5g pulsing felt by the astronauts. A real highly unpleasant bone shaker. There it was managed by partial throttling on ascent ( or early shut down of the engine in the center of the cluster ). It's confined to the liquid fuel designs as pressure waves ripple up and back within the tanks by reflection off the ends. That destroyed a few Atlases before they realised the effect and with the Titan series of ICBM's it got worse with power improvements, that is with higher specific impulse engines. So could this yield a peak gee load well above design for said strut ?
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Did you guys see the report
)
Did you guys see the report yesterday about what actually caused the last crash? Apparently a support strut broke and a helium bottle crashed violently around the compartment causing the explosion.
http://money.cnn.com/2015/07/20/news/companies/elon-musk-spacex-mission-failure/index.html
RE: Ah, yes. Oddly
)
This started me thinking. I think there is a lot to learn from, and expecting of, failures. In fact we probably only really learn from failures. There have been many rocket failures in the past to learn from.
There is a message in (my interpretation) "Despite the catastrophic failure of Falcon - Dragon may have survived....if parachute deployment was included in the software"
It is not the first time a spacecraft has come to grief on ascent.
Apollo (and Soyez) LES used the re-entry parachutes for escapes for decades.
Fortunately this occurred before Dragon was crewed, but "forgetting to open the parachute" is as big a design fail as the strut.
Space X Dragon spacecraft to
)
Space X Dragon spacecraft to get survival software update
Dragon will attempt to save itself in the event of a rocket explosion