Stories
Slash Boxes
Comments

SoylentNews is people

posted by on Monday February 06 2017, @11:36AM   Printer-friendly
from the nobody-goes-to-mars-until-we-say-so dept.

SpaceX is no stranger to delays. The private space firm headed by Elon Musk has pushed back is launch schedule several times in the last few years after rockets have been lost. Now, the Government Accountability Office (GAO) says there may be an issue with the Falcon 9 rocket that delays the expected launch of the first manned mission in 2018.

The report from the GAO (just a preliminary release for now) cites issues with the turboblades used in Falcon 9 rockets. These are the components that move fuel from the tanks to engines. The blades apparently have a tendency to develop cracks, which could cause catastrophic failure if they develop or worsen during a launch.

According to NASA acting administrator Robert Lightfoot (who also has an amazing name) says the agency and SpaceX have been aware of the issue for months (or possibly years). NASA expressed concern to SpaceX that the turboblade cracks presented too great a risk to launch manned missions. Cracks have been found in the turboblades as recently as September 2016.

SpaceX says it has been conducting extensive testing on the Falcon 9 rocket and believes it to be safe. It has made changes to the design of the turboblades in an effort to mitigate the cracking issues. Although, the company may still undertake a full redesign of the blades depending on the upcoming GAO report. If that happens, the manned launch will almost certainly be delayed.

Source:

https://www.extremetech.com/extreme/243883-problems-falcon-9-design-delay-manned-missions


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 4, Interesting) by VLM on Monday February 06 2017, @02:28PM

    by VLM (445) Subscriber Badge on Monday February 06 2017, @02:28PM (#463425)

    (just a preliminary release for now)

    The only primary data source seems to be one management dude talking to one WSJ reporter, and then at least 100 clickbait sites all quoting the paywalled WSJ story.

    I wouldn't say that's outright fake news but it is kinda smelly. A managerial political appointee could entirely accidentally get the complete story from the engineers correct, or more likely not, or most likely of all only get it partially correct based on partial out of date data.

    The only thing I know for certain having tried to find the official report is the most recent GAO report containing the keyword spacex is from late last year GAO-17-100 which is a discussion of how to fit spacecraft training into the FAA regulatory framework. The story is undocumented at this time.

    The problem I see with man rating as it stands is in aerospace you tend to need to build that in from the start, you can't add it like a checkbox after its done. The falcon platform already has issues with tank within a tank design such that they lost one vehicle when a strut snapped and another when they played games with the exact order and temperature of launch pad testing of tank filling and emptying (picky picky). The government tolerates a lot of BS with unmanned one way missions as long as you're not a huge threat to randos on the ground but man rating has to be baked into the cake initially, you can not just patch failure modes until you feel lucky. The market has always been somewhat bifurcated because man rated design is more expensive because the stakes are a little higher.

    The point of the above paragraph is the most likely failure mode for spacex failing to get permission to launch dudes is going to be a core design issue like "this tank within tank mounting system ain't happening" and that can't be simply patched around. A rather unlikely failure mode for the overall licensing procedure is "ya gotta xray inspect 100% of turbine blades on a man rated vehicle" and the problem with that is ... no problem at all, you just do your extra QA/QC work and go flying.

    Starting Score:    1  point
    Moderation   +2  
       Interesting=2, Total=2
    Extra 'Interesting' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   4  
  • (Score: 2) by DannyB on Monday February 06 2017, @03:09PM

    by DannyB (5839) Subscriber Badge on Monday February 06 2017, @03:09PM (#463442) Journal

    The falcon platform already has issues with tank within a tank design such that [1] they lost one vehicle when a strut snapped and [2] another when they played games with the exact order and temperature of launch pad testing of tank filling and emptying (picky picky).

    As for problem 1, wasn't the strut not actually built to withstand the stress it was supposed to withstand? Or do I mis-remember? Didn't SpaceX switch manufacturers for that strut? SpaceX tested a bunch of struts and found that quite a few failed at a much lower stress than they were supposed to?

    As for 2, if the problem is well understood, then isn't it safe to change the fuel / oxidizer loading procedure to avoid the problem? Sort of like if you change the procedure for working on electrical equipment to require electricians to use insulated gloves? Or any other normal, sane precautions for a problem that becomes well understood? Wear proper safety gear around noxious chemicals, etc. Don't ban the noxious chemicals. Just fix your procedures.

    --
    People today are educated enough to repeat what they are taught but not to question what they are taught.
    • (Score: 2) by VLM on Monday February 06 2017, @05:10PM

      by VLM (445) Subscriber Badge on Monday February 06 2017, @05:10PM (#463516)

      Both situations fail the intrinsic safety goals you'd like to have in a man rated vehicle and would be really difficult to patch around, need to do that at design time.

      Intrinsically safe means you eliminate entire categories of failure mode by implementing a safer design. Lets say people keep dying replacing ceiling lights. You could make ever safer ladders. Or airbags underneath ladders. If you want it intrinsically safe, wouldn't it make a heck of a lot more sense to install LEDs that will theoretically last the life of the house? Or even better yet eliminate the high light source so you don't need ladders to change the light? Its hard to fall off a ladder you never put there because there's no light bulb to replace because there's no light bulb installed there.

      The intrinsic safety failure of problem 1 is they suspended one tank inside another and if that suspension snaps the tank bowling balls out the bottom of the big tank and you suddenly lose the vehicle. You make that intrinsically safe by merging the tanks with a common bulkhead or mounting the little tank on a bulkhead or somehow firmly attaching the little tank to the bottom of the big tank so if it breaks loose at least it won't "bowling ball" out the bottom of the vehicle causing an explosion. You might lose the mission because the engines shut off early but at least the astronauts would live, unlike going out in a giant explosion. Sure the tank only self destructs 1 in 50 flights and that maybe makes economic sense when launching unmanned GPS satellites, but it doesn't make sense for a manned launcher. It has to be designed in at the start, its a major redesign to move those tanks around.

      The intrinsic safety failure of problem 2 is carbon and carbon fiber plus liq or solid O2 equals death and there's no way to work around it. Charcoal and liq O2 was used quite a bit as a "safe" mining explosive a few decades ago but its too unpredictable. Sure they cooled the tank too deeply maybe freezing O2 and a pressure change caused a sudden crack in the solid O2 that was enough friction to ignite the carbon fiber and blow the vehicle to pieces. So ... how abouts making helium tanks out of something less flammable than carbon fiber, I donno titanium or an aluminum alloy? Even bulk magnesium is harder to ignite than carbon fiber. They think they've reduced the odds of another explosion by changing ground procedures but no one really understands cryogenic O2 mixed with carbon. Its like walking around with a balloon full of welding gas, there's things you can do to help with static electricity and improve your odds but its inherently reckless. Again, sure it'll explode 1 in 50 times but thats OK for unmanned and not OK for manned. It might be slightly easier to redesign to use non-flammable tanks inside a liq O2 tank, but it'll throw off weight and balance and lower performance. A block of steel is inherently safer inside a liq O2 tank than a piece of charcoal. There's no non-cheaty way to make the bar of steel more likely to catch fire than the piece of charcoal.

      You see intrinsically safe in a lot of medical hardware. Oh 12 volts would kill a patient? Lets power the thing off 5 volts then no body can get electrocuted no matter how they F it up. Oh 10 mA will kill the patient, well, lets put a resistor in series such that if you shorted the power supply thru the resistor thru a zero ohm "patient" that it would flow less than 10 mA. Lightning from a thunderstorm could electrocute the patient? Well we COULD F around with lightning protection but wouldn't running off a battery be infinitely safer? Lithium batteries explode at sterilization temperatures so our pacemaker will use a different battery technology. etc.

      • (Score: 2) by DannyB on Monday February 06 2017, @05:17PM

        by DannyB (5839) Subscriber Badge on Monday February 06 2017, @05:17PM (#463522) Journal

        I see your point about problem 2.

        I also see your point about problem 1 as a design issue. My point about problem 1 was that no matter what design you use, if someone makes inferior parts that don't meet specs, this is not a design problem. You could try to address poor quality manufacturing (or sabotage) in the design. But I think that is the wrong place to address either of those.

        So if you didn't have struts fail due to poor quality or sabotage, then problem 1 isn't really a problem. But I take your point about trying to design so that the strut isn't even an issue.

        --
        People today are educated enough to repeat what they are taught but not to question what they are taught.
        • (Score: 2) by VLM on Monday February 06 2017, @07:34PM

          by VLM (445) Subscriber Badge on Monday February 06 2017, @07:34PM (#463585)

          But I take your point about trying to design so that the strut isn't even an issue.

          You could even take the helium tank/strut problem further. So the problem is if the mounting fails it bowling balls thru the bottom of the ship and blows it up. Well, if you're going to replace one big tank inside the O2 tank with a dangerous strut, why not put like 3 small tanks right on top of the engines and get rid of all that high pressure gas line (which is heavy) and install three one-way valves and if a tank broke off the vehicle not only would it not smash thru the bottom but if it happened late enough in the flight and the one-way valve worked, it might not even affect the flight. Oh maybe booster recovery would literally run out of (pressurized) gas, but that wouldn't kill any astronauts. Inherently safe by design.

          The nuke people like inherently safe stuff too. Go ahead, slowly overheat it, the structure expands and it falls out of criticality. Go ahead, cut the wires to the control rods, the rods are attached to the actuators by giant electromagnets so cut the wire and they fall into place shutting down the reactor if you take an axe to the control wires.

      • (Score: 2) by choose another one on Monday February 06 2017, @08:31PM

        by choose another one (515) Subscriber Badge on Monday February 06 2017, @08:31PM (#463624)

        > Sure the tank only self destructs 1 in 50 flights and that maybe makes economic sense when launching unmanned GPS satellites, but it doesn't make sense for a manned launcher.

        What number makes sense for a manned launcher? 1 in 65 fatality rate perhaps? 1.5% of flights?
        Would a 1 in 50 flight fail rate with a launch abort be better that 1.5% with no abort option? Which would you rather ride on?

        The shuttle took the baked-in-from-the-start approach to man rating and chose to use SRBs (which at the time had around 1 in 50 fail rate), chose segmented SRBs vs. single piece SRBs (why have a joint that can fail if you can build in one piece?), chose to use manned vehicle heat shielding that was fragile and then put the manned vehicle not on the top of the stack (like every manned rocket before or since, out of harms way as much as possible), but instead put it _behind_ most of the main tank, which was covered in an insulation jacket that was sprayed on rather than manufactured and was known to shed fragments.

        The shuttle had multiple intrinsic safety failures _despite_ having baked in man-rating from the start, and despite this had no launch abort / escape option, Falcon9 will also probably (inevitably?) have multiple intrinsic safety failures, but it will have have a launch abort / escape option, and the kinks will be ironed out first with unmanned flights.

        The COPV tanks are safe as long as the struts are manufactured to spec and the launch temps are "correct", except no one really knows what "correct" is. Sounds bad - except it is exactly the same as the O-Rings on the shuttle, not intrinsically safe, vehicle could have been designed without it, safe if used at the right temps but no one knew what those were in advance. The big difference is that the O-Rings were a man-rated-baked-in design and were tested at new launch temps with a crew on board, SpaceX only tested new launch temps with a satellite on board.

        Which is the best approach? Only time will tell, but time has not been kind to NASA's shuttle approach. NASA's approach also carried a cost premium of 10x (maybe 100x if SpaceX does get reusable). Was it worth it?

        • (Score: 2) by VLM on Monday February 06 2017, @09:21PM

          by VLM (445) Subscriber Badge on Monday February 06 2017, @09:21PM (#463671)

          The shuttle had multiple intrinsic safety failures _despite_ having baked in man-rating from the start

          They baked in political deals from the start. Hard to have a SRB failure if you don't have SRB although you don't get the Utah vote without SRBs so we gotta have SRBs even if they kill people.

          The insulation deaths were the same way. The ET spray on foam is environmentally negligible but they couldn't use the good stuff because it looks bad to be anti-environment on TV, so instead they killed a bunch of people by using stuff that is more environmental but falls off and cracks heatshields. Sure we littered parts of the shuttle and parts of the astronauts all over Texas but thank god we didn't use environmentally unfriendly insulation, that's what really matters in a human-life space program, not mere human lives.

          You have to balance SpaceX is doing questionable things vs NASA just plain old sucks.

  • (Score: 2) by PiMuNu on Monday February 06 2017, @03:30PM

    by PiMuNu (3823) on Monday February 06 2017, @03:30PM (#463452)

    > you can not just patch failure modes until you feel lucky

    I thought that was how NASA did it in the 50s/60s? Another era maybe...

    • (Score: 2) by DannyB on Monday February 06 2017, @04:45PM

      by DannyB (5839) Subscriber Badge on Monday February 06 2017, @04:45PM (#463500) Journal

      So NASA is where Microsoft learned it from?

      --
      People today are educated enough to repeat what they are taught but not to question what they are taught.
    • (Score: 2) by mrchew1982 on Monday February 06 2017, @04:46PM

      by mrchew1982 (3565) on Monday February 06 2017, @04:46PM (#463502)

      Yup. Boeing and Lockheed et. al. Would like you to believe that their designs have always been safe and never had any failures when the truth actually is that the failures were so long ago (or on differently named launcher) that no one remembers. If you look back at their history (and all of the histories of the companies that they bought out or merged with) you'll see plenty of failures where they kept iterating on a design until they got it right. The new rockets that they build are actually slight modifications to old, blown up several times on the road to perfection, proven designs.

      What ULA definitely doesn't want anyone to realize is that all of their R&D (and blown up rockets) was done on a cost plus contract with taxpayer money at the height of the cold war...

      • (Score: 2) by gman003 on Monday February 06 2017, @10:55PM

        by gman003 (4155) on Monday February 06 2017, @10:55PM (#463761)

        There was a point, in the early days of Atlas, where Convair (later sold to McDonnell Douglas, then to Lockheed Martin, then spun off into ULA) was promising the government it could bring reliability up to 60%. While extra precautions were taken with the manned versions, no ground-up redesigns were done before Mercury-Atlas launches started, and the Mercury flights had an estimated success chance of 90%. They flew anyways - because mission failure doesn't have to result in loss-of-crew, and they just built the best launch abort system they could. Two out of the nine Mercury-Atlas flights failed - both unmanned test flights, and both would have been survivable if they had been crewed. All four manned flights succeeded.

        Parts of that design continue to be used - the Centaur stage used on some later Atlas flights is still used, with only minor changes, on the Atlas V, and it is planned to be used even with the next-generation Vulcan rocket. The Atlas V is going through the process of man-rating as well, since Boeing's entry in the Commercial Crew program will fly on Atlas V. Atlas V has enough weird heritage (the main engine is a Soviet design drawn from the *boosters* of Energia) that if SpaceX is getting shut down for cutting corners that ought to be round anyways, ULA should get shut down for the same reasons.

  • (Score: 0) by Anonymous Coward on Monday February 06 2017, @04:05PM

    by Anonymous Coward on Monday February 06 2017, @04:05PM (#463472)

    "The problem I see with man rating as it stands is in aerospace you tend to need to build that in from the start"

    Cheap access to space means the vehicle just has to work.
    Reusable means it has to work over and over. (Yet to be seen.)
    That may be a higher bar than 'man rating'.

    Just because X is nimble does not mean that they are less reliable than current aerospace.
    NASA was nimble once before a variety of factors fixed it.

    • (Score: 0) by Anonymous Coward on Monday February 06 2017, @04:38PM

      by Anonymous Coward on Monday February 06 2017, @04:38PM (#463493)

      NASA was nimble, and pretty damned lucky. The issue that caused the deaths of the astronauts on the launchpad for Apollo 1 (pressurizing the capsule with pure oxygen on the launchpad) was standard operating procedure for the Mercury and Gemini missions. They were just damn lucky they didn't have a fire on those programs. "Nimble" can very often mean "sloppy". It achieved its goal of beating the Russians to the Moon, but that doesn't mean that it is the model a reasonable program should follow as a matter of course.

      • (Score: 2) by DannyB on Monday February 06 2017, @04:54PM

        by DannyB (5839) Subscriber Badge on Monday February 06 2017, @04:54PM (#463508) Journal

        Shortcuts should not be taken. Corners should not be cut. Low cost can still be a goal. Reusability can still be a goal. That said, it is still a dangerous business. No matter what you do, there will be an LOC accident sooner or later. Strive for a high level of safety. But don't cripple the program.

        I'm sure aviation was much more dangerous in its early days. As the magic of electricity reached more and more homes, we learned a lot more about safety. Try looking the original electrical wiring practices of a house built in the 1930's.

        --
        People today are educated enough to repeat what they are taught but not to question what they are taught.
      • (Score: 0) by Anonymous Coward on Monday February 06 2017, @10:00PM

        by Anonymous Coward on Monday February 06 2017, @10:00PM (#463716)

        What a twisted answer.

        Nimble doesn't lead to unsafe.
        Nimble should lead to being able to look at the data, think, and quickly adjust to something better.

        NASA today is about following rules even if they lead to something worse.
        This is the definition of safe for a bureaucrat.
        That is, what ever most likely to get me to retirement.