I think the idea is that external pressure just pushes the jointed parts more firmly together, but if there is a way for pressurised water to get to the end of the composite tube, it will find any small void or imperfection.
But if I was a betting man, I’d put money on there being something wrong in the glue. It looks like graphite loaded epoxy but they just buttered it onto the pieces. A small air bubble inside the resin, or a hair or spot of grwase or sweat that fell onto the surface could cause a lack of integrity at that spot. This might not manifest as material failure until its been through a few pressure/decompression cycles - each one compressing fluids into a tiny void, then when the outside pressure abates, the pressure trapped in the void forces the material apart microscopically
Thanks.
So where did they get the titanium end caps? That’s my thinking, yes it’s expensive, but they were already halfway there anyway. Just make the whole thing out of titanium and call it a day.
Bicycle frame manufacturers have a lot of experience dealing with carbon fiber in both tension and compression, and they’ve been working their layering patterns, resin formulae, and thickness profiles for a couple of decades now. Granted it’s not the same as a deep sea vessel. A bike frame is made up of a bunch of tubes with the forces mostly along those tubes (basically it’s a truss). The fork, top tube, seat tube, and seat stays are in compression, and the down tube and chain stays are in tension (mostly, things get weird when torsional stresses and standing vs sitting come into play). The forces aren’t trying to crush the frame tubes into themselves, hopefully. Bike frame builders were also some of the first to get really good at welding titanium in the requisite non-oxygen environment.