Debate: Did I do the right thing by grounding a 737 for a cracked windshield that passed stress tests?

I was working line maintenance in Denver last Tuesday when I flagged a 737 with a 2-inch crack near the lower corner of the windshield. The lead inspector said it passed the stress test and could fly another 50 cycles. But I grounded it anyway based on the crack pattern I saw. Turned out the inner ply was delaminating, and it would have blown out at 30,000 feet after a few more pressurizations. My supervisor backed me up, but the crew was furious we lost 4 hours of schedule. So here is the question: Do you follow the book every time even if it costs the company money, or do you weigh the risk if the numbers say it is okay? Has anyone else called a no-go when everyone told you it was fine?

3 comments

3 Comments

lucast817d ago

Kind of reminds me of something my uncle told me once about a flight he was on in the 90s where a tiny crack in the galley window got ignored because the pressure seal held. That thing failed mid-flight, not a blowout but a slow leak that forced an emergency descent and scared everyone. The book isn't just about the current test, it's about the whole history of how these parts behave under repeated stress. Your gut called out a pattern the test couldn't see, and that's worth the schedule hit every time.

fionam116d ago

Respectfully, I see it the opposite way though. Your gut feeling is just a guess based on one old story, it's not data. The test passed, so delaying the schedule for a hunch that might be wrong costs real money and time for everyone else. We should trust the measurements until we have actual proof something is off, not just a scary memory from the 90s.

leor866d ago

That story is a good one but I think your uncle might be mixing up a few details. A slow leak from a tiny crack wouldn't cause a rapid enough pressure loss to force an emergency descent like that, it would probably just cause the cabin to slowly lose pressure over several minutes. The real danger with window cracks isn't a slow leak, it's when the inner pane fails completely and you lose the outer pane's support, then you get a rapid decompression. Your point about the history of stress patterns is totally valid though, that is exactly the kind of thing the book should cover. The test passing doesn't erase the fact that we don't have a good model for how this specific part behaves after a thousand more cycles. A schedule hit now is cheap compared to backtracking after a failure later.