Why did the Titanic tour submarine implode?

Maybe the same forces that caused the Unsinkable Titanic to sink were at play, if they got too close and collided with part of the Titanic wreckage? Maybe even a bump would have been enough to momentarily exceed the structural design parameters at a particular spot causing an instant cascading implosion failure? Operator error, maneuvering malfunction, or perhaps unexpected currents, etc. Could have been at play?

There was a report at one point that "signal banging" against the hull wall could be heard at repeated 30 minute intervals, so it could be there was a loss of maneuverability which might have allowed a collision with something? Maybe using Morse code could have been and asset if it works under those extremely deep open conditions, and if anyone on board knew how to use it?

 

There are several full ocean depth test chambers large enough for full sized vehicles. It is one of those engineering quirks that works for you sometimes and against you on others. As we used to say referring to the DSV 4 SEA CLIFF 20K upgrade; "NASA has it easy, they only have to hold 10 psi IN, we have to hold 9000 psi OUT".

Really; a full ocean depth test chamber is just a big balloon... and compared to a pressure hull, very flimsy and light as it's biggest structural load is actually supporting the weight of all the water. During WWII the Germans even constructed a barge that allowed them to pressure test whole U-Boats. Because there are so few, and so specialized, and the risk of damage to the facility growing exponentially with test pressure; they are costly to use and have very full schedules, with time often going to those with the deepest pockets.

During the 1960's, 70's, and 80's the US Navy's DSSP used these test facilities to not only pressure test new equipment as it was developed, but to conduct extensive testing, including cycling to failure, of different candidate pressure vessel and pressure hull materials to failure (and in the case of Titanium spheres, the pressure facility failed in fatigue first). This is why certain materials are prohibited from manned systems, the collapse of even a small pressure vessel next to a pressure hull can cause a catastrophic failure of the main hull. So while today wound CF is used as an internal pressure vessel, it is prohibited for use as an external pressure vessel given the inability to accurately predict failure. I would suggest the texts Manned Submersibles by R. Frank Busby and SNAME's Submersible Vehicle Systems Design for a full discussion of pressure vessel selection.

 

Weight. Weight is your enemy in deep submergence. Notice how little syntactic buoyancy foam TITAN had compared to most other metallic pressure hulled vehicles.

 

There actually is a incident when investigating USS THRESHER that the DSV 1 TRIESTE actually came down on a piece of wreckage, barely missing damage to the float. Cameras were later fitted to prevent this from happening again.

 

And that's the issue.

Fatigue has 3 stages to it,
i) crack initiation
ii) crack propagation
iii) failure

And that would be i) - crack initiation.
In other words, unlike most isotropic materials that have movement in dislocations along slip bands, to 'initiate' the crack, having a small void - air bubble - is exactly that.
The stage i) can be "built-in" from the outset.

Thus the only way to mitigate is extensive testing and extremely strict QA procedures to eliminate this possibility, and of course extensive means of verification.

 

They now think that the implosion was on Sunday during the initial decent. The “banging” noises that were heard appear to be an unrelated natural sound still to be identified.

 

In this CNN article
OceanGate marketed its vessel as safe. But experts say it used materials that 'simply didn't work' | CNN https://edition.cnn.com/2023/06/23/us/oceangate-submersible-titanic-safety-invs/index.html
They mention how the carbon fibre 'increased the vessel's buoyancy' - wow, amazing!
'In online videos, Stockton Rush explained the Titan’s unconventional design, which he said included carbon fiber to increase the vessel’s buoyancy. It “hasn’t been used in a crewed submersible ever before,” he said in a video last year.'
The article is also rather scathing about the safety aspects of the submersible.
“At some point, safety just is pure waste,” Stockton Rush said. “I mean, if you just want to be safe, don’t get out of bed. Don’t get in your car. Don’t do anything.”

 

Stockton used the Russian roulette method. If you put a gun in your head, pulled the trigger, and the gun did not go off, it must be safe. After several dives and it did not implode, he concluded it must be perfectly safe.

 

As I used to tell my operators, "...the last event is only tenuously connected to the present one..."

 

True. The cylinder is CF, the dome is glass, held together by titanium structure, bolted together. One gave way first. When?

At descent where it is nearing freezing temperature where it is supposed to operate?

Or probably the internal stress created by differing material with different CTE? Remember the temp at sea level (start of launch) is presumed 21 degree C standard and the temp it is supposed to operate is at near freezing temp.

 

How deep a depth can they simulate? E.g., could they handle the depth the OceanGate submarine had to sustain, + a reasonable safety margin?

What would they probably have charged to do that type of (repeated) testing?

It is interesting to note that the Trieste bathyscaphe, which went down to the bottom of the deepest known part of the ocean, was tested manned, in a set of dives to increasing depth. During one of those dives, the window cracked. The crew chose to keep going down. That suggests that the U.S. Navy, which sponsored the deepest Trieste dives, and which had a Navy Lieutenant on board for them, didn't always do what some here would consider "adequate safety testing" before putting people on board.

And it would be putting it mildly to say that during the race to space, NASA didn't do adequate safety testing, and some early mission astronauts died. Likewise, multiple NASA & U.S. military aircraft test pilots have died testing aircraft. I presume the same is true of other countries who developed spacecraft and high performance military aircraft.

Of course the OceanGate sub is different. Their passengers were sort of tourists, and they weren't diving to record depths, or meeting national security goals - but they signed waivers indicating that they knew they were doing something risky. That is roughly akin to the waivers people participating in dangerous sports (e.g., whitewater rafting, downhill skiing, rock climbing), which tend to kill people every year, are often expected to sign.

BTW, according to news reports, OceanGate claimed they designed their submarine in partnership with Boeing, the University of Washington, and NASA, but all three have denied this. Interesting. It strikes me that overstating or fabricating the roles of those organizations wouldn't help OceanGate in court trials. It makes sense to only say things they can prove, or no one will believe them.

 

This aspect of the post I have excerpted it from may be key.I believe that a filament wound tube would contain fewer flaws than a component made using any other laminating technique.There are good reasons why it wouldn't make a reliable external pressure vessel and I can't think of any other method of producing a CF tube that wouldn't introduce a lack of perfect uniformity at numerous points.A well engineered layup specification for an autoclave cured piece would have to include multiple phases and even with warm debulks,there is still the certainty that not every piece will be perfectly adjacent to the next piece or that no dirt will have been transferred by the laminators moving around the workshop.Even if the workshop is maintained to clean room standards.So that is one level of potential flaws before we get to the next phase of the build,when structure is added to the tube.There may have been all manner of hard spots introduced,even without considering the titanium rings at the ends.A few diving cycles may have been a great way to introduce micro cracks around any of them.I expect there is a reasonable chance that an ROV will be collecting a few pieces for examination in due course and the results of close examination will be published.There are five people who won't get the chance to acquaint themselves with the facts,sadly.

 

Here us the USNA facility brochure. TITAN's pressure hull would fit in Tank A which is equivalent to 27,000 FSW.
https://www.usna.edu/NewsCenter/sites/NAOE/images/DOTF_Brochure_003.pdf

Not my functional area so I don't know the exact cost, but you had to schedule well in advance.

You are sorely misinformed. The viewport never cracked. What cracked was the thin plexiglass diver monitoring window on the free-flooding access tunnel due to thermal and compression movement between the float and the sphere at depth. This window allowed the divers to ensure the access tunnel was pumped down before opening the sphere hatch. Even though the window cracked, it leaked very little (and there is always duct tape). I'm sure you will be shocked, just shocked, to know that the Krupp sphere that went to Challenger Depth was made of 3 pieces and was not welded or bolted together....

 

Year 1994, @Aerospatiale, Toulouse, France.

It looks like Beechcraft have also experienced some great difficultie to achieve the security level with composite materials. No surprise. Aeronautical standards are very demanding in term of risk insurance. Exactly the same situation for AIRBUS, developping the A350 XWB, and the A380. For them, the issue with composite lied in being capable of ensuring a reproducible and constant material quality, and -curiously- managing the interface between composite parts and metallic parts. Talking about transportation, in France, the first production car that uses composite was built by PEUGEOT, for their 205 model. With caution, they used it only for door trims, given them no structural role. Good for them, as they had some issues with their glued bounds.

FRP have more strenght in tension than in compression, it is true. But don't be mistaken, this is not exact to say that the resin handles the compressive loads in such materials. Due to the adherence between fibers and matrices, it's the whole compound which takes the load. In relation with the porosity rate of your material, FRP will brake in compression either due to fiber break, either due to resin micro-buckling. The resin itself has a very low compressive strenght.

Adding to your rightfull remarks, I may precise that, at macroscopic level, variation in strenght of FRP are due to variations in : Fiber to resin content ratio, material porosity, fiber alignement. And unfortunately, it's extremely hard, even impossible in some cases, to measure their respective values without destructive testing. Especially with the supposed 137mm thicknesses of the outside shell. In this case, you have to make samples in different locations, put thin slices under microscope, and burn the rest of the sample. Then repair the sampled zones. And this is just one solution that I've practice, in carbon unidirectionnal stocks up to 250mm. There are other local technics, or design approaches, in complement with QA loops, that can assure the safety of laminates in extreme usage. But it seems also that this sad accident could be the result of the work of a company with a very poor culturen know-how, in term of security.

For instance, it seems that OceanGate just ignore warnings from their director of marine operations and from one engineer, that has been put out of the game and sued to shut is mouth, when he reported that the integrity of interface between the pressure chamber and the shell couldn't be checked, making the submarine at risk. Other several warnings have been also delibarately ignored.
OceanGate Was Warned of Potential for ‘Catastrophic’ Problems With Titanic Mission https://www.nytimes.com/2023/06/20/us/oceangate-titanic-missing-submersible.html