3 ms·
The real offender here is the design of the seat... The design should be self-testing continuously. If any wire, microchip, or other component were to go bad
by londons_explore 4y ago
The real offender here is the design of the seat...
The design should be self-testing continuously. If any wire, microchip, or other component were to go bad inside the ejector seat, it should flash a big warning light on the dashboard saying "unsafe to fly".
In the case of an ejector seat sequencer, the electronics could run through the full ejection process, but with the power supply to the explosive detonators super low. Then another system detects that the right detonators are being activated at the right time. If they aren't... red light and buzzer.
If your design is good, cheap parts won't make it fail dangerously. Cheap parts will just lead to more time being repaired.
- Wowfunhappy 4y agoWhat if the electronics that run through the full ejection process are counterfeit?
- paulryanrogers 4y agoThen design it to fail safely, so if any one part in the sequence is out of line it a red warning light goes off.
- gilded-lilly 4y agoSounds dangerous. What if there’s a power surge?
- londons_explore 4y agoIt's the same design as done on car airbags... They test themselves all the time and will put a red light on the dashboard if the test ever fails...
- mjevans 4y agoA pre-flight readiness check could do the 'dry run' with the explosives redirected to a test rig instead. As an armchair engineer, I would imagine this check shortly prior to munitions load, which would also be the best time to validate the rest of the platform fitness.
- _djo_ 4y agoThat would be extremely expensive in terms of the rocket and cartridge usage and would subject the seat to a huge amount of stress, substantially reducing its lifespan. Even then, you can't be sure that the issue wouldn't only show up mid-flight for some reason, at different temperatures, or different vibration levels from wind resistance in a high speed ejection. Ejection seats are not designed to be used again and again, they're designed to be used exactly once and work flawlessly when that one moment comes. Rigid process control and verification in the manufacturing, storage, and installation steps is the only way to be certain it'll work. I get that you're just playing armchair engineer here, but you're ignoring nearly all the actual complexity and engineering challenges.
- mjevans 4y agoYou've clearly misunderstood, probably the common term 'dry run'. The test would be of the electronics _only_; not the explosives or any other single use component. Single use items are usually validated by randomly sampling from within the batch / production run.
- _djo_ 4y agoMy apologies, you mentioned the explosives being 'redirected', when I guess what you meant was that the explosive circuitry would be redirected. Even so, those checks aren't sufficient to handle counterfeit parts that might only fail under stress, rather than the safe conditions of a ground test. The best approach is what it has always been: Strict process control of the entire supply chain, rigorous component and subsystem testing under extreme conditions before delivery, deep analysis of failed components coming off the flight line, and regular inspections for anomalies.
- suzzer99 4y agoWhat about a manual parachute pull as a backup? Not sure if the pilot was conscious in this case but I assume so.
- _djo_ 4y agoMilitary systems do usually have extensive self-test functionality in place, but when it comes to things like ejection seats and canopy detonation circuits that absolutely must fire reliably the more logic you have in between the higher the risk that it won't work when needed. Nor is self-testing 100% reliable, as counterfeit chips might work well enough during a self-test sequence on the ground but fail when exposed to the extreme temperatures and vibrations of an ejection. It's not just about whether the part meets the spec, it's about meeting the spec in all conditions. That's why for things like this and similarly crucial things like munitions-related circuitry the preference is for circuits and components that are as simple as possible and to use process control to ensure that all components are reliable.