About the news of missing Titan sub

[Vanadium 50]

All fibres in that layup are running in the plane of the cylindrical surface, none are radial.

There is a cabin in the way.

The radial forces are transferred to the collar of the hemispheres and to some degree the hoop fibers.

And epoxy is strong stuff as well.

 

[Baluncore]

The fibres used should have the same temperature coefficient, young's modulus, and speed of sound as the epoxy filler.

There is a cabin in the way.

The radial fibres should cross the wall, stitching the layers together, to prevent delamination.

And epoxy is strong stuff as well.

Then why use the carbon fibre in a compressive orientation. It would be better to avoid the fibre which will progressively delaminate if the temperature coefficient, modulus of elasticity, or speed of sound are different to the epoxy.

 

[Vanadium 50]

It may not look it, but tensile strength is critical. Compression is not the whole story.

The problem is not that the hull might evenly shrink under pressure, killing the passengers. The problem is that it might buckle. That has compression on one side and tension on the other.

This is why the US Navy uses high tensile strength steel for its submarines. They have not lost a submarine in over 50 years, even after driving at least one into a mountain.

 

[jrmichler]

I used a unidirectional carbon fiber rod in an application where we needed light weight, high stiffness, and Euler buckling for overload. Since a design review raised concern about it breaking when buckling, I ran a test. A 15" length of rod needed to buckle sideways about 3", and it did not fail until bent far enough that the free ends almost touched. It failed in tension. The failure was a sudden brittle fracture with a loud noise and pieces flying across the room. The test was repeated several times, and each failure was on the tension side. The tensile failure was interesting because the specifications for the rod call for it to be stronger in tension than compression:

The rated compressive strength of the rod is almost two orders of magnitude larger than the compressive strength of epoxy. I believe the compressive force are supported by the fibers, while the epoxy prevents local buckling of the fibers.

Since we needed to convince a number of people, we made a video of the test. I knew that it would shatter, so was wearing goggles and heavy leather gloves. I did not realize that I was standing under a bright light that made me look like some sort of demented mad scientist on the video. Several years later, that video was still being shown to new engineers, apparently to warn them about the guy in that office over there.

 

[gmax137]

I did not realize that I was standing under a bright light that made me look like some sort of demented mad scientist on the video.

Any way you can share the video, or at least a still frame?

 

[phinds]

Oh, come on. Now that you've shared that it happened, you HAVE to show us so that we can warn the newbies about the monster in our midst.

 

[Baluncore]

The problem is that it might buckle. That has compression on one side and tension on the other.

The tension that prevents buckling under external pressure is circumferential, on the inside of the cylinder.

That inner layer has been pre-compressed and shortened by hydrostatic pressure from outside the vessel, which is not helping to increase tension in the inner layer to prevent inward buckling.

When tension is applied to the inner hoop, that inner surface would tend to be pulled away from the concave neutral layer. I would expect any inner tension to result in a further crushing of the outer hoop layers that are then under greater load, with a delamination of the inner layers, and a failure to prevent the buckling.

Since any inner layer hoop tension must be countered by outer layer hoop compression, I see no advantage in relying on tension to prevent inward buckling of external pressure vessels.