The design was defective in many ways, all of which could cause implosion. One feature can be identified as most probable, where the titanium end bells joined the carbon fiber tube.
The deep sea is cold. Under temperature change, from surface to depth and back to surface, the two materials contracted and then expanded at different rates — with differing coefficients of expansion. When, at the factory, the end bells were attached with epoxy cement, all the parts were at factory ambient temperature. At any other temperature, there was stress between the parts.
- Each dive comprised a cycle of thermal and mechanical stress on the epoxy joints.
- After some number of dives, microscopic cracks began to grow in the joints.
- Cracks concentrate stress, which is why cracks usually grow.
- At some cycle, the epoxy failed in some region of the joint.
- An area of the cylinder then lost support from the end bell. This area of the cylinder also experienced the most flexure, so it was the most fatigued.
- Implosion of the cylinder began there, and spread longitudinally in just a few milliseconds.
One might expect that the end bell side of the joint was a flange, with a smaller diameter extending into the carbon fiber tube. Sometimes this is done to achieve a thermal shrink fit, so that the joint is always under compression, which makes it stronger. This was not done. The end bell began with a flat adapter ring that was simply glued to the end of the tube.
It is possible to join materials with different coefficients of expansion, by use of a slightly elastic gasket, a flange, and constant compression of the joint with tension bands. This was evidently too much bother for Stockton Rush. Even if this were done, it is likely the Titan would have eventually been claimed by snap buckling.
Which brings us back to safety culture: