Like I said, it was incompetently designed/built/qualified/etc.. My point is that even in spite of this, it survived several dives. So it had excess margin even well past the point where any respectable operation would have stopped using the hull. That’s good! It means that if you qualify a hull to X dives without damage, it’s still very likely to survive even if your analysis is wrong and it does start taking damage before dive X.
I’ve not watched the docs. How was this sound transmitted? Does this mean that they heard the implosion through the water and then the air, or was there some sort of sound transmission doo-hickey involved, relaying the sound from down below?
Or they got very, very lucky.
That’s a problem with extrapolating statistics from a single sample. Did the design have excess margin or did they happen to flip heads 10 times in a row and wrongly assume the coin was always going to come out that way until it didn’t?
Assuming the former is just plain bad engineering practice. If rich but stupid people are willing to throw their lives or money away to provide more statistics, more power to them. But no engineer worth their salt should draw such a conclusion until they do get more data.
Calvin’s Dad? Is that you? 
I think the business case, which was always thin, would be completely unworkable if the hull was a consumable. How many rich shipwreck fanatics are there?
We know how to make strong subs. For our limited use cases, they seem to be fine.
It’s obviously both. Rush ignored the many, many warning signs that the craft was in a state that was not safe. Nevertheless, it’s good to know that the craft would not catastrophically fail the instant one of those warning signs presented itself.
I mean, he’s not wrong. Design your hull properly, test to failure, then pick some point well short of that to be your design lifetime, and develop tests to verify that your data hold up under actual use. Maybe you wouldn’t build a bridge that way, but for a hull that is only expected to have some finite lifetime, it’s not a bad idea…
That may well be true. On the other hand, if the business itself was run as incompetently as their engineering department, maybe the business case isn’t as bad as it seems. Could be that they were wasting money on worthless things that stole from the rest of the budget.
My point is that it categorically cannot have been both.
Did the design have excess margin or did they happen to build one sample that was 15 standard deviations beyond the mean? If the design would normally produce fridges that last 5 years on average, did they luck upon one that is still running 20 years later on their first try?
It’s not obvious at all that the design happened to be better than expected and simultaneously this one sample was an extreme outlier. That contention seems to be based more on confirmation bias rather than engineering practice or principle.
I think you’re still misunderstanding my point, so let me try an analogy.
Suppose there’s some incompetent operator of a twin-engine aircraft. Right from the start, one engine is having obvious problems. A few flights later, the engine fails completely, and yet the operator still makes flights, apparently under the delusion that they can flip the coin forever. But a few flights after that the other engine fails and they crash and die.
The operator clearly should not have flown the plane at all in even a partially degraded state. And yet, the fact that they did make multiple flights even with a failed engine is evidence that the plane had some margin left. No one should ever depend on that margin, because without understanding it you can’t know when it will actually fail. And yet it is there, and gives confidence that if you did experience an unexpected failure, it would probably not be catastrophic.
One thing about the design: how roomy it was. The large cylinder really made deep diving accessible to tall people that can’t sit unmoving for 8 hours. Like me. Traditional deep submersibles are small spheres with teeny tiny windows. For a very good reason. But they also are claustrophobic.
Me? If I had his money, I’d spring for a 12 foot diameter titanium sphere. You only need one, and you have the benefit of proven tech.
People underestimate the need for high strength-to-weight ratio materials. Could you make a deep-diving sub out of normal steel? Sure, but then the crew portion would be tiny. Worse, it would no longer be buoyant. So you’d then need large amounts of buoyancy material, like gasoline balloons (the Trieste) or syntactic foam (Cameron’s Deepsea Challenger). Which you can do, but it massively increases the size of the vessel compared to the crew volume.
You’re possibly right about the titanium sphere, though it wouldn’t be cheap.
I mentioned this upthread too, like a Geiger counter going off as you approach the radioactive source. I fear that those 5 guys in there may have heard a lot of Snap! Crackle! Pop! for perhaps a few seconds before it imploded. Not a good feeling.
But at least the end came mercifully quickly for them. When I die, if I have a choice of methods I might select that one. It happens so quickly that the brain cannot register what is happening. But I’ll take it without the Snap! Crackle! Pop!
It came over some speaker, on some frequency, but I don’t know what it was monitoring. It was described by the USCG investigation team as like a door slamming.
Did it? I thought it looked like it came through the hull of the support vessel.
I believe so. The Coast Guard officer introducing the recording to other members of the committee, in any case, says
You will hear a noise that is external to the ship—or external to the room, I should say—and you’ll see their reaction to the noise [emphasis mine]
Without identifying it as “a noise over the loudspeaker” or “a noise from the hydrophone” or anything.
Going back to the back and forth about margin involving @Dr.Strangelove & @Great_Antibob …
What we learned is that some popping that occurred with an unknown spatial distribution didn’t fail that particular hull on the first pop. Nor the second. Nor the third … But eventually there was enough that the hull failed catastrophically.
@Great_Antibob said in effect that we have no idea if this hull would’ve been one of the best of a production run of 100, or one of the worst. I agree with that point, and with their comments about its significance.
What I will add is that we also have no idea if the distribution of delams and fiber failures that occurred on earlier dives was especially fortuitous or especially not. From the final voyage data it seems as if they eventually got a cascade failure going. Which cascade lasted just a few seconds before The End Was Nigh.
Failure cascades are like that, and especially in structural materials. Crack propagation in metals works substantially the same way.
Were they just lucky that voyages 1, 2, and 3s delams happened to be spread around the hull? Perhaps if they’d occurred closer together the hull might’ve cascade failed at 500m on dive #1. We just don’t know. And given the pulverized condition of the wreckage, we won’t ever know. Which of course doesn’t affect Oceangate’s nonexistent future production run.
But does say that anyone in the future wanting to engineer, not just cobble together, a CF submarine hull has gained zero useable data from this event.
My bottom line:
Uncontrolled experiments aren’t really true “experiments” yielding reliable “data”; they’re anecdotes yielding anecdata. Sometimes that’s better than nothing. Other times it simply is nothing. IMO this is one of those latter times.
I know sound propagates better through water than in air, but I’m surprised that the sound of the implosion could be heard on the surface.
I’m not. That submarine had a pretty decent interior volume. When that much water collapses the shockwave must be massive. Think about how loud a kettle can get just before it boils - and those are microscopic bubbles collapsing.
This is why I have a bit of trouble with Nightcrawler’s (from the X-Men) bamf when he teleports. It’s supposed to be from the air surrounding him collapsing. It should be a deafening thunderclap.
It doesn’t just collapse - the air bubble contracts and re-expands. Air isn’t infinitely compressible, so it compresses to a large extent and rebounds and re-contracts and rebounds again and again losing energy each time until it dissipates altogether.
Those later expansions and contractions probably wouldn’t be audible (and would occur several times a second so they’d blur with the main bang anyway to any human being) but any recording gear on the surface has a good chance of detecting them as separate events.
Yeah, I’ve actually worked on submarine sonar (as an enlisted sonarman and then in engineering) for most of my adult life and I’ve had a few email discussions with Bruce Rule (who was critical to the analysis of the loss of SCORPION.)
They’re “bubble-pulse” trains and they oscillate at a frequency that’s dependent on the collapsing volume. It’s interesting because they can actually estimate the collapsed volume (and energy of the implosion) from this frequency alone.
I’m unconvinced that you could actually hear the collapse event through the hull at that depth and with so small a collapsing volume, but it isn’t impossible. I’ve no real interest in watching any of these documentaries though–just from reading about them they sound mostly sensationalistic and don’t seem to add much to the narrative.
(Truth be told, one of the greatest casualties of this tragedy has been my sanity. As mentioned, I’ve worked with this stuff for most of my adult life but after the Titan collapse I’ve run into many “experts” who believe they have an understanding of these topics after five minutes of Googling.)
I beg to differ. Stockton Rush has proved that a cylindrical CF hull can go to 4,000m deep. It can work. IANAE but if that hull was reinforced by titanium rings it might be certifiable for some number of dives.
That said, the failure behavior of a CF hull is not well understood (for a deep submersible). What hasn’t been mentioned recently is the pressure test of the scaled down model of Titan that failed during pressure testing. In the documentary (I forget which one) the failure was shown to be sudden and immediate. The documentaries did not share any data on its behavior up to when it failed.