I think the idea is: implosion is so scary they work really hard to engineer it out (as long as you don't go beyond your max depth).

"The carbon fibre and titanium, there's a rule you don't do that. Well, I did." (Found the quote on WikiPedia.)

It seems insane to me. It's easy to be an armchair anything, but I don't understand why you would do such a thing. It's not like a submersible needs to be light weight, either.

Engineering a submarine is well above my pay grade, but some of the variables you have to balance are.

the walls have to be strong(thick) enough to withstand the pressure (this is obvious, I know)

the ratio of wall thickness to void(crew) space determines reserve buoyancy.

smaller submarines are stronger than bigger ones. but have a worse buoyancy ratio.

The end result is that as you design a submarine to go deeper it has to become smaller with thicker(heaver) walls until past a certain depth it goes negative buoyant and will sink. So you have to add additional buoyancy. not a trivial task given the pressures involved. the Trieste(first submersible to the challenger depth) was built like a blimp, with a tiny crew sphere hung below a massive tank of oil(they used oil because it is lighter than water and will not compress).

So a lighter material hull is very desirable, it makes the whole operation much easier. They probably should not have been testing it on commercial passengers however.

This is also the benefit of titanium, it is almost as strong as steel with half the weight.

It is insane, but not entirely pointless. A metal submersible is heavy and negatively buoyant, whereas you want it to be positively buoyant (then add removable ballast).

The way you add buoyancy is expensive fancy aerated concrete, or something like that, and a lot of it, since it's not that buoyant. But it is pressure resistant.

By making it out of CF, you save on that cost.

Not a worthwhile cost cutting exercise but at least there is some reason for it.

I believe because it was substantially cheaper to fabricate than a traditional metal hull.

I would think, with how advanced computer simulations are, you should be able to tell if something will implode or not. But maybe I'm wrong.

My understanding is that most commercial manned submarines don't actually go that deep. As with depth there isn't that many interesting things... And lot less light. So pressures they are dealing are quite reasonable. Which makes implosions pretty unlikely in normal operation. Common scenario is to operate in areas were sea bed is at depth where implosion is impossible.

It is, so it’s taken really seriously. The hulls are made from material with well characterised gradual failure modes - bending and deforming rather than sudden failure. This means metals with thoroughly inspected welds and joints to ensure no internal voids, and a process of gradually diving deeper to check the hull meets the design requirements.

Using materials that fail plastically and gradually increasing depth trials means the failure mode is hopefully deformation rather than complete failure, and will happen at the highest depth as possible so a quick surfacing can be achieved.

Submarines, along with space, are an area where innovative new methods need a lot of testing before you commit human life to it.

If I remember the early reporting correctly, this submarine had plastic deformation on every dive. So it was already failing, they just didn't do anything about it.