I have some Opinions about the titan submersible and maybe it will stop taking up so much brain power if i write it all out so here yall go. Almost all of this is about the engineering of the sub. Scroll on if you don’t care about this at all
So two important things about me: 1) i have a degree in materials science and engineering 2) i have an obsession with aviation disasters
So opinions that I feel I am qualified to talk about b/c of my degree
1) Carbon fiber - the choice of carbon fiber for the bulk of the hull for this was weird. Carbon fiber has a fairly high tensile strength and fairly low compressive strength. Tensile is when you are pulling something apart, compressive is when you a smushing something. Its tensile strength (how much stress it can withstand) is around 500,000 psi (3.5GPa) [this is dependent on other factors but lets go with it for now] it’s compressive strength is closer to 145,000 psi [again depends on a lot of factors, we are just going with this for the sake of my discussion]. It’s compressive strength is nothing to sneeze at, but it is much lower than its tensile strength.
The important part of a submersible is it’s compressive strength, the pressure from the ocean is going to be trying to collapse the submarine it and it needs to be able to withstand that. Can carbon fiber withstand the ocean? By the numbers I have found, technically yes. NOAA says that pressure increases about 1 atmosphere per 10 meters of water depth putting us at around 4400 psi for 3000m. The thing to know about pressure vessels is that they degrade every time they are pressurized and unpressured (a cycle). The pressurization can create defects in the material, and these defects often lead to a lower stress required for failure. The more cycles something undergoes, the more likely it is to fail. I am not sure how much research has been doing in cyclic loading of this specific carbon fiber under compression. I am dying to see the research data that went into this design, how much material testing did they do?
Third important thing about carbon fiber - the weave. Essentially, when using carbon fiber, you weave them together in different orientations to build the size/shape of composite you need. A typical weave would have the fiber all going in one direction, then the next layer would rotate the fiber 90 degrees in comparison to the first, then back to the original orientation. Or perhaps a 0/45/90 degree wave. This weave ensures that the material has similar properties when undergoing stress from all directions, not just one. If you have all the fibers going in the same direction, it will likely fail when being pulled any other way with little stress. And for a pressure vessel, you need to withstand stress in all directions. I heard from a friend they had their weave in the same direction, but I have not found sources confirming that. Frankly I am having trouble finding any sources about the specifics of the carbon fiber at all.
Finally, ceramics (like carbon fiber) tend to fail quickly with little warning. Metals tend to bend before the break. Carbon fiber and other ceramics won’t. It will get a crack that starts to grow and reach critical size and break with little or no sign visually. Nondestructive testing of ceramics is also harder than metals. We have some tests to find small cracks inside of metals, but using those on ceramics is not as well documented or are more expensive.
2) Mixing titanium and carbon fiber - why? Mixing materials is always something you need to be careful when doing. We want those materials to match some specific properties as best we can. One such property would be coefficient of thermal expansion. AKA how much does the material expand or shrink when exposed to hot or cold temperatures. If those materials do not have similar values, one will expand or shrink faster than the other, and causing the seal between the two to be weaker. The coefficient of thermal expansion for carbon fiber depends on the weave, but is much different than titanium. This means the seal could become an issue at low temperatures (such as occur down 3000m below the surface of the ocean)
3) The timeline - according to recent sources the design of this sub occurred in a very short period of time. Which is concerning. The amount of testing needed to design something like this should be many months, not a few weeks. This is experimental and it will carry people. You want to take your time to collect data, consider other materials or design choices.
4) Using the porthole that was only certified to 1300m for a sub that went to 3000m - i feel like I dont need to expand on why that was dumb :/
Now to the things I am not qualified for with a degree but I feel qualified to talk about because I am obsessed with aviation disasters and there's a lot of overlap here.
1) The remote controller - my problem with the video game controller is not its use, there is some good application for it, my problem is how it was the only way to control the sub at all. There was no redundancies in the sub so if something went wrong, there was little to be done.
2) The lack of oversite - I love innovation. I love seeing engineering trying new things and going to new places. I also love rules and regulations that stop people from doing really really dumb shit. And this sub was subject to none of them (?) at least not that I can tell. Which is awful. There needed to be someone else there making this company stop and breath. To not rush. To think about safety more. To consider plans B and C. And to ensure they were making good engineering choices. What was the factor of safety with this submersible? The fact that this thing could carry paying passengers w/o being certified by anyone is insane to me. No one looked at this and said: hmm, this seems dangerous lets do more testing b/c no one was required to.
The aviation industry is not and has not been perfect. But the aviation industry has overall, done a good job and looking at mistakes, disasters, incidents, and near collisions and changing how they do things. They engineered planes to be safer. They set up maintenance procedures that test parts to see if they have critically large defects. Clearly we need to think about doing some of the same here since damn, someone really dropped the ball on this.
Anyone I’m dying to see the materials testing data they did in advance of using this thing cause I want to know what they saw that they feel would make it safe.
(also if anyone wants resources for what I talked about or where i cited lmk, most of this I googled and pulled from scientific journals or newspapers today)