Because it can be hard to find much on the weekend...
NASA and Boeing Find Deeper Problems with Starliner
Not that they seem willing to put the launch off because of them.
Senior managers have said they intend to launch Starliner's Crewed Flight Test on June 1 at 12:25 PM EDT, as noted in the addendum to Wednesday the 22nd's post. Since the first scrubbed attempt to launch Starliner, there has been weeks of detailed analysis of a helium leak and a "design vulnerability" with the ship's propulsion system.
Extensive data reviews over the last two-and-a-half weeks settled on a likely cause of the leak, which officials described as small and stable. During these reviews, engineers also built confidence that even if the leak worsened, it would not add any unacceptable risk for the Starliner test flight to the International Space Station, officials said.
But engineers also found that an unlikely mix of technical failures in Starliner's propulsion system—representing 0.77 percent of all possible failure modes, according to Boeing's program manager—could prevent the spacecraft from conducting a deorbit burn at the end of the mission.
"As we studied the helium leak, we also looked across the rest of the propulsion system, just to make sure we didn't have any other things that we should be concerned about," said Steve Stich, manager of NASA's commercial crew program, which awarded a $4.2 billion contract to Boeing in 2014 for development of the Starliner spacecraft.
I love precision like "0.77% of all possible failure modes"! Such confidence! Such respect for significant digits!
Question for any of you with lots of "probability and statistics" course work:
how many trials would be needed to have high confidence it's more like 1%
chance instead of a 5% chance? How many to determine it's more like 1% than a
10% chance? I assume that if it's really a 5 or 10% chance given some large
number of trials, the failure would happen more often than if it was a 1%
chance. How many to distinguish a 0.77% chance from a 1% chance? I'm betting
on the order of tens thousands of trials. From a system
that's going to fly maybe 10 times. Maybe.
I found the entire article unsettling but while I'm not a "belt and suspenders" guy, I'm not a test pilot either. There's a Robert A Heinlein quote (I think it was RAH) about test pilots; something like they'd jump off a building with rubber soles if you gave them odds of surviving it.
Butch Wilmore and Suni Williams, still in pre-flight medical quarantine, will return to Florida on Tuesday to resume final launch preparations. NASA officials will call a flight test readiness review Wednesday, and if they give the green light, ULA will roll the Atlas V rocket and Starliner spacecraft back to the launch pad on Thursday, May 30.
Starship FT-4 Set for Wednesday, June 5 at 8:00 AM EDT
In a Friday announcement, SpaceX has set a June 5 launch date for its next Starship integrated test flight, with a focus on demonstrating the ability to bring both stages of the vehicle back intact. The mission details seem like what we covered on Tuesday, 5/21.
“The fourth flight test turns our focus from achieving orbit to demonstrating the ability to return and reuse Starship and Super Heavy,” the company stated. “The primary objectives will be executing a landing burn and soft splashdown in the Gulf of Mexico with the Super Heavy booster, and achieving a controlled entry of Starship.”
The company has made several upgrades to both the Super Heavy booster and the Starship itself, based on their failure analysis of Flight Test 3.
SpaceX said that Starship “began losing the ability to control its attitude” several minutes after engine shutdown, while coasting on a suborbital trajectory. Video of the flight showed the vehicle slowly rolling. That loss of attitude control led to an automated decision not to perform a planned engine relight.
“The lack of attitude control resulted in an off-nominal entry, with the ship seeing much larger than anticipated heating on both protected and unprotected areas,” SpaceX stated, with telemetry lost at an altitude of 65 kilometers.
The company believes the attitude control issue was caused by the clogging of valves in thrusters used for roll control. SpaceX has added more thrusters for redundancy while upgrading thruster hardware “for improved resilience to blockage.”
The booster fired 13 of its 33 Raptor engines after separation for a boostback burn, but six of the engines began shutting down, triggering an early termination of the boostback burn. Only two of 13 engines ignited for the final landing burn.
The engine failures, SpaceX concluded, were caused by filter blockage in liquid oxygen lines, keeping propellant from reaching the engines. SpaceX experienced similar problems on previous Starship launches and made changes to the engine design to try to prevent it. “Super Heavy boosters for Flight 4 and beyond will get additional hardware inside oxygen tanks to further improve propellant filtration capabilities,” SpaceX said, along with unspecified “additional hardware and software changes” to improve the reliability of Raptor engine startup.
Another interesting little change they've made is to jettison the interstage
adapter they added to allow igniting the upper stage before the booster
separates. They will still have that adapter in place but the way I read
that, they'll drop the interstage adapter once the booster separates.
Booster 11 and ship 29, the Flight Test 4 hardware, stacked for the May 15 WDR. Screen capture from SpaceX video posted on X.
Notwithstanding a few exceptions, it is a violation of regs for an aviator to launch with known defects.
ReplyDeleteBut as a friend with decades of experience with FAA and USAF told me, everything is waiveable.
I say the answer to your question is, infinitesible. I reckon if 1 is nominal, then 0.77 is a bit more than 3 of 4 chance not of full functionality but of nothing catastrophic.
Anyway, I suppose they weighed the chance of failure of vehicle to failure of future contracts. To show they really mean it, they'll launch with humans aboard.
NASA and FAA sign off on this.
To clarify: functionality = no cascading failures resultant from known detective part(s)
DeleteRunning NASA's statemen through the fewmetometer... "Oh, we found a bunch of small errors that we are maybe kind of pretty sure won't add up to a fatal collection of errors, so we're sticking with flying this dogturd."
ReplyDeleteYeah, if I was one of the two astronauts, I'd be getting strep throat or a sudden case of broken foot or something 'at the last minute' in order not to be near that conical death trap.
As to SpaceX ejecting the staging ring, very interesting. Hope they figure out how to retain it for future reusability, but it wouldn't surprise me if they did a cost analysis and figured it was cheaper to ditch it than try to keep it.
Of course, everything will change with Flights/Versions 2 and 3 and beyond. Very interesting to watch the evolution from Starhopper to now and, of course, beyond.
SpaceX already has plans for a different/new staging setup. Probably built-in from what I can see. Borrowing from the Soviets, they are...
DeleteDefinitely not a liquid gas tech, so what solids would be floating around in a liquid oxygen tank? The only thing I can think of is water ice. I would hope they would have got rid of that before tanking the rocket.
ReplyDeleteSince I live in a place with humidity like Starbase, if not worse, I suspect if it is ice that it's forming due to humidity freezing out of the air in the tank. That doesn't excuse allowing it to happen, but it may argue for a longer period of chilling the tanks to get that ice to freeze out and perhaps a way to get rid of it. I don't think it will sublimate out and frankly don't see a way to get rid of the ice.
DeleteIt might be a side effect of being the largest cryogenic tanks ever built. Maybe no engineers have seen that because there's not enough ice to be a concern in smaller tanks.
how many trials would be needed to have high confidence it's more like 1% chance instead of a 5% chance?
ReplyDeleteTrials? Plato's allegory of the cave is fiction. There does not exist a million alternate dimensions, each containing a spacecraft with a slightly different flaw, from which human-observed reality is a sampling. This cosmological idea is fiction, and the statistical theory built upon it is also fiction.
See also https://bayes.wustl.edu/etj/articles/random.pdf starting at pdf page 7 heading "BUT WHAT ABOUT QUANTUM THEORY?"
how many trials
ReplyDeleteWhat's actually happening is that humans have made a large pile of speculations, and extrapolations from past events assumed to be similar, and statistics tells us the best way to stack the assumptions. But it's all just a model inside a brain; none of it is forced to be coherent with physical reality. Take your brain's eye off of reality for even a moment and you get Socialism, Modern Monetary Theory, or gun control for ICBMs, aka NASA.
It's my understanding that Starliner was justified as being a backup to Dragon, in the event that those crazy kids at SpaceX couldn't make a reliable vehicle, like the big boys at Boeing counted be relied on accomplishing . ( Irony fully intended.)
ReplyDeleteSince Dragon had proven to be safe, effective and fliable, why are we still pouring big bucks into the Starliner money pit ?
Oh, wait, never mind . We need to keep giving Boeing big piles of taxpayers money so their DEI employees can keep making 737s with popout doors !
A minor correction. They thought Dragon would be the backup to Starliner. After all - they're Boeing, Space 1.0 before anyone knew Space 2.0 was coming, brought by SpaceX. Boeing's contract was about 50% bigger than what SpaceX got and the NASA administrator at the time, Charlie Bolden, said if Boeing hadn't been interested enough to bid, the commercial crew program might have been shut down by congress.
DeleteEverything else you say seems exactly right. Backups, like having Starliner around in case something weird happens, are pretty routine. Do you trust Boeing or Russia's Roscosmos more? At least Russia is actually flying.
At this point Boeing is into Starliner for over $1.5B (amount publicly acknowledged) of its own money. The NASA contract was fixed price. NASA is only incurring its own costs. It will be interesting to see Boeing's pricing for a Starliner flight assuming this one is successful.
DeleteIt has been awhile since there has been a catastrphic failure with fatalities in the rocket industry, even internationally. With the cadence increased, budgets constrained and pressure to produce increasing, it is to be expected corners will be cut.
ReplyDeleteI blame digital watches. There's that spurious precision everywhere now. "Earl Campbell averaged 4.283 yards per carry in his NFL career!' Yeah no.
ReplyDelete"There's a Robert A Heinlein quote (I think it was RAH) about test pilots; something like they'd jump off a building with rubber soles if you gave them odds of surviving it." The version I heard was if a combat briefing of 100 pilots stated "we expect 99 of you will die" all the pilots would look around thinking "those poor suckers."
ReplyDelete