kalvado wrote:benjjk wrote:Mostly wrote:
I’m not saying it’s likely. You’re right that it hasn’t happened yet. But I doubt you’d argue that a wing connected to a cracked fork can cope with less load before it fails. So the question is how much less. And you and I don’t know the answer to that.
Could bad turbulence cause a failure? Probably not. But maybe. Could a hard landing cause a failure, probably not. But maybe. The point is that we don’t know.
We do know that cracks in parts that weren’t engineered to crack can transmit stress to areas that may not have been engineered to handle that stress, so runway failures can happen.
Don’t get me wrong, I’m not suggesting that failure is likely, what I’m saying is that we don’t have enough data to make an assessment. And the track record isn’t great so far with the manufacturing consistency of the part, so that adds to overall uncertainty.
Quick edit: the accepted safety factor for commercial aviation is 1.5 to 2.5. Boeing’s corporate culture probably means they’re on the left side of that range. So figure we’re a 1/3 drop in structural integrity away from being at a safety factory of 1.0. That’s not outside the realm of possibility
The crack does reduce the margin of safety, which is why these particular aircraft are grounded - though ferry flights are permitted. But to suggest that means a wing could fall off tomorrow is, to borrow the Qantas press release, alarmist. Put it this way: after the MAX debacle I would expect the FAA to be taking a very conservative approach to this. But even they have said the inspections can wait until a convenient time (within limits obviously), and don't need to happen right away.
One more thing: I have been told that replacement of the pickle forks takes about 3000 man-hours. With a fix that expensive, "fly until it cracks" is the only realistic avenue here, unless it's a very high-time aircraft likely to see imminent cracking.
So how much time industry had before an accident if problem wasn't noticed? My impression is not that much.
Of course, we don't know how fast crack propagates and how bad consequencies can really be. Possibly - not very likely - this is a self-limiting process to begin with. From Boeing-FAA responce, though, this seems to be a bad case of propagating crack and possible bad outcome.
As far as I understand, once the crack is initiated - a countdown timer is started, and it is a matter of time before part is weakened to the point it cannot handle the load. AD assumes few thousand cycles before that happens; lets guess 5k - that is 3 years at 5 cycles a day.
Now there are 50 planes with cracking at various stages; no data on how badly or mildly cracked different frames are. Assuming more or less uniform distribution of crack progress, the worst case could be within 5000/50=100 cycles until a failure, that is about a month of flying, and expected crash rate would be on the order of 10/year.
Maybe I am missing something, but grounding older fleet doesn't sound that crazy with such probabilities. Apparently, FAA and Boeing decided that expedite inspections is a safe enough bet. One has to remember that those are just that - estimates and bets. Boeing just lost big time on "10 months before next event" one.
For me the answer is simple. When cracks are detected, the frame is grounded until the repair is done. So no flying with a frame with detected pickel fork cracks, that means, that the cracks are taken serious.
I expect now all frames being inspected when they reach 22,600 cycles, the inspection done before the frame reaches 23,599 cycles.
After the data out of the current inspections has been collected, a inspection regime has to be set, or all pickle forks to be replaced in a certain interval, moving it from a lifetime part to a limited time part.