Georgia engineer talks carbon fiber that failed in Titan submersible implosion

Hull of Titan submersible imploded undersea; expert says needed testing is 'not rocket science.'

[jorelldye]

Delamination is when layers of carbon fiber begin to unbond from eachoher due to stress. Steel returns back to its original shape. Damage to carbon fiber remains and accumulates. That's why the Titan could make multiple successful dives before imploding.

Also, carbon fiber holds INSIDE pressure well because the tensile strength of the fibers is incredibly strong. Outside pressure on carbon fiber will put more stress on the epoxy that the fibers are suspended in, which will necessarily promote delamination.

Imagine a barrel wrapped in ropes and glue. You can't stretch the barrel because the ropes hold tight, but if you squeeze the barrel, the ropes fall loose and the glue breaks.

[jonnyovideo]

James Cameron said his spherical titanium submersible compressed 3 inches in diameter at depth. It is safe to say that titanium and carbon fiber would compress at different ratios causing even more extreme pressure on the Titan joints that were GLUED into place. I'm impressed it didn't implode on the first dive.

[aurora_gypsy]

Carbon fiber is excellent to carry tensile stress, but not the best for compressive stress, and absolutely NOT the compressive stress within the fibers plane, which is a type of stress that promotes delamination. The two hemispheres of the Titan submersible would push against the CFRP cylinder with a force of thousands of tons, almost 20'000 t every time the Titan reached 4'000 meters depth. This led to a 210 MPa compressive stress within the cylinder wall, a value which is dangerously high for CFRP. Of course CFRP can carry much more in tensile stress, but not compressive stress within the fibers plane, this is something that NOT HAPPENS in airplanes, so if CFRP is safe for airplane wings, it doesn't mean that is suitable for submersible's hull because the stress type is different! For a typical carbon fiber in epoxy matrix I would not exceed 50 to 70 MPa of cyclic stress in that direction, so a submersible with that design should not dive deeper that 1'000, maximum 1'300 meters in my opinion. But I don't know what kind of CFRP the hull was made, I am doing assumptions basing on a commonly used CFRP type, maybe they adopted a fiber design specifically optimized to withstand compressive stress, who knows? In any case, a designer should know exactly WHEN the vessel collapses, so a destructive test should have been carried on before using it for carrying passengers: basically you dive the submersible deep down until it implodes. Then, if the submersible implodes e.g. at 5000 meters, you can assume that, in order to be safe respect to fatigue stress, you should not exceed 30% of the collapsing depth. I say this because you can run, of course, simulations, but in the simulation it's difficult to keep into account all the possible material defects, so a destructive testing campaign would give the most reliable result. And you SHOULD KNOW exactly when the failure occurs, it's a vital information! Then, I would also not use CFRP at all for that application, CFRP is anisotropic material, in that case stress comes from any direction and therefore an isotropic material, i.e. a metal, suits better. They made the hemispheres in titanium: so it would be a good idea to make the cylindrical section in titanium too. In this way you also avoid the mismatch in material's elastic modulus, which leads to additional stress in the junction area between CFRP and titanium, which is another possible source of delamination.
P.S. I'm an automotive engineer with 15 years of experience
Edit: I have found a video, you can watch it at the following link, where they compare the compression resistance of tubes made by different materials, including a CFRP. I don't know what type of CFRP and if comparable to the material used in Titan hull section, as details of materials used in the video are not given. Anyway, the CFRP tube collapsed at 2998 kg, that is equivalent to a pressure of 203 MPa for that pipe geometry. I wrote that Titan reached about 200 MPa of pure compressive axial stress every time they dived till the Titanic depth, so now you can realize how close they were to the compression limit of CFRP. Fatigue stress probably added delamination spots every time in a cumulative damage to the hull, so that section after ten dives was definitely not the same as the first dive. Considering these numbers, it's surprising they have been able to dive several times before the collapse of the hull

[famkis]

Dude, any engineering student can calculate in 5 minutes: Compressive stress in the carbon fiber polymer (CFRP) wall of the Titan submersible was about 200 newtons per millimeter squared at Titanic depth. This corresponds precisely to the stress level where CFRP typically fails under compression. Only under tensile load (which is not applicable underwater), CFRP can withstand about 10 times more. What's more, OceanGate used unidirectional (circumferential) CFRP. This means that in the longitudinal direction, the hull was only about half as strong, at around 100 newtons per millimeter squared. It's a miracle the ship even made it to Titanic a few times without imploding.

[grahamtaylor]

IKEA has drawers doing non-stop open/close cycles to prove their durability. Ocean Gate should have had a 2nd Titan doing simulated deep dives in a pressure chamber

[GenXer]

Human arrogance at its finest. If you want to experiment with your own life that’s one thing, but to be allowed to experiment with the lives of other’s is morally unconscionable. This submersible was a death trap waiting to happen. Even if the other passengers signed a waiver to acknowledge the risks, none of them fully understood just how poorly designed the submersible was. After listening to the experts in the field and gathering more information about the history of deep submersible diving, this situation was 100% avoidable. Stockton Rushes legacy will forever be tainted by his deliberate carelessness and refusal to do things the right way. There was a clear standard set in place for safety that had been tested and proven. The fact that Stockton knowingly went against that system by cutting corners at every opportunity he had to do the right thing is criminal. No one had to die and that is what makes this situation so tragic.

[kestutisbagusauskas]

After the Titan disaster I saw a YouTube video where they test the compression test of various material using a hydraulic press. The sample were basically 1 inch diameter tubes and about 2mmm thick from memory, and about two to three inches long. They were stood upright under the press head and the pressure recorded when they began to collapse. The interesting was that carbon fiber was one of the materials tested and it was only about twice as strong as aluminium! The strongest by far was stainless steel which suprisingly even outperformed titanium!!

[Mothball_man]

The only thing that kept that hull from imploding on earlier dives was the epoxy. There’s an old saying.. you can’t push a feather with a chain.

[richardpare]

One of the design issues not pointed out is that the CF was wound around the hull axially, meaning that the only thing resisting the longitudinal stresses was the epoxy.

[justaguyreal]

Big difference between keeping pressure in and pressure out. Huge difference.

[lalaLAX]

Interesting. I’ve heard experts in other interviews point out that carbon fiber performs excellently under tension (as with planes) but not always as well under compression (as with the submersible).

[vladimpaler]

Okay, but I want to point out that in airline applications the inside might be at surface pressure (say ~15psi) and the worst the outside can be is a vacuum or 0psi. This means it has to withstand at most 15psi outward pressure, pulling the fibers lengthwise. The Titan was heading towards 5672psi inwards pressure compressing the fiber. Use correct materials in the correct appliction.

[Ascgames]

As an engineer it beyond reasoning why someone would want to use carbon fiber to build a submarine when it is an unfit material to use under compression. Foolish pride and not listening to the experts to use the right materials was what got the CEO killed and unfortunately the passengers. Hope this will be a lesson to others to listen and get expert help always and only use certified crafts or equipment.

[luisramrod]

As an engineer student I love how this story has started a conversation on material science. 😊

[televison_is_dead]

Carbon fiber is the last thing I would decide to make something like this out of. It has awful compressive strength, especially when you're taking it thousands of feet underwater. How did this even get passed as a good idea

[imranbecks]

I don't trust the material on bicycles because they'll eventually give way. I definitely don't trust it to be used in a submarine. They'll break without warning if cracks isn't spotted early.

[granitfog]

Carbon fiber is really two things, the carbon fibers themselves and the resin that holds the fibers in position and gives them stiffness. Carbon fiber has a tensile strength of 3 - 7 Giga Pascals. The tensile strength of resin is 55 Mega Pascal to 2.7 Giga Pascals. So the tensile strength of resin is below that of the fibers. Under expansion, as in a tank filled with gass and pressure pushing out, the fibers are taking all the load and the weave helps the fibers stay together and the fibers strength against stretching prevents explosion. However, if the pressure is outside pushing in, it is the resin ( and not the fibers) that take the load to keep the fibers from bending. The strength of the fibers in this case is less important. 

The other issue is the viewing port was rated for a depth of 1800 meters, but the Titanic is located at about 3800 meters. So the implosion could also have been due to the viewing port itself.

[boeingnz]

From what I understand is that the material is better at inside presure not outside presure pushing in.

So it's perfect for going up things like planes and spacecrafts but not submarines.

[SoloRenegade]

Aerospace applications: tensile strength, carbon fiber

Deep Ocean applications: compressive strength, not carbon fiber

anyone working on a submarine should have these basic concepts understood.

[mygundidntdoit]

It's great for things like fishing rods, but not for subs. You need titanium for subs.