13
$\begingroup$

SpaceX is known for preferring reusable parts whenever possible. Before the crew dragon incident, where a leaking check valved resulted in a plumbing explosion, it was determined that the check valve be replaced by a burst disk, which is not reusable. In fairness, the engines on the Crew Dragon are no longer going to be used for propulsive landing, and instead will only be used for emergencies, why didn't SpaceX decide to use 2-3 check valves instead? Doesn't this means that the Super Draco is no longer has multiple-restart capability?

Some basic statistics shows that using 2 or 3 already makes the probability of this even astronomically unlikely, and it would enable the possibility for engine re-ignition.

Improve this question
$\endgroup$
3
  • 4
    $\begingroup$ The aforementioned "basic statistics" assume that failures aren't correlated. If you can have an environmental or manufacturing factor that puts all the devices at risk at once, those statistics aren't necessarily valid. See also how the statistics showing that it was possible to create a AAA-rated tranche in a bucket of subprime mortgages turned out to be faulty. $\endgroup$
    – Charles Duffy
    Commented Oct 7, 2019 at 12:22
  • 1
    $\begingroup$ “aren’t necessarily valid” -> “are invalid” $\endgroup$
    – Reid
    Commented Oct 7, 2019 at 23:48
  • $\begingroup$ Related: space.stackexchange.com/q/35707 $\endgroup$
    – Machavity
    Commented Aug 3, 2020 at 18:08

1 Answer 1

Reset to default
28
$\begingroup$

It would be possible to design a check valve chain of high reliability of closing, but it would restrict flow and start to increase the risk of one valve sticking shut. You would probably also need pressure sensors between the valves to determine system state, so you end up with a large number of pipe to fitting joints that are potential leaks and a bunch of supporting electronics.

Burst disks have the advantage that they are reliable closed until they are supposed to be open, and allow very high flow rates in low fitting weight/volume when open. They are also easy to test/monitor, since they require very unusual circumstances to have slow leaks so you generally just need some form of basic integrity check with wire running through a loop rather than multiple pressure/flow sensors with related electronics.

So if your engine is only going to be used once, and possibly only for emergencies you want a system that has minimal weight, easy to monitor/test during prelaunch and flight and high reliability when you do need it, which may make burst disks a valid choice.

Improve this answer
$\endgroup$
6
  • 2
    $\begingroup$ The shuttle OMS helium pressurization system had parallel vapor isolation valves and a series/parallel quad check valve arrangement to prevent this kind of explosion without using burst disks. nasa.gov/centers/johnson/pdf/… p. 2.18-10 There weren't any extra pressure sensors for this part of the system. $\endgroup$
    – Organic Marble
    Commented Oct 7, 2019 at 0:39
  • 4
    $\begingroup$ @organic Marble, that must have been an interesting design and construction challenge for someone. Certainly possible, after all Spacex originally designed it that way originally. And quite possibly is 'better' with them - my supposition would be the change is more about being seen to move quickly towards human rating than the optimal long term design choice. $\endgroup$
    – GremlinWranger
    Commented Oct 7, 2019 at 1:08
  • 4
    $\begingroup$ @GremlinWranger: Isn’t it more likely that the check valves were an artifact of the original design which used the thrusters for propulsive landings? $\endgroup$
    – Michael
    Commented Oct 7, 2019 at 11:57
  • 2
    $\begingroup$ Also, the problem was a slug of oxidizer ended up in the line after ground handling, and got rammed into the check valve by the rapid pressurization. This could still potentially happen if the oxidizer got into the line connecting series check valves, even if it didn't get through the whole string. So the series redundancy might not actually be effective at preventing the problem, while the parallel redundancy would actually provide more opportunities for it to happen. $\endgroup$
    – Christopher James Huff
    Commented Oct 7, 2019 at 22:37
  • $\begingroup$ @ChristopherJamesHuff I'm not exactly sure what you're saying there, but the Shuttle OMS pods had a 30+ year record of not exploding due to oxidizer vapor migration. $\endgroup$
    – Organic Marble
    Commented Oct 9, 2019 at 3:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.