Titanic tourist sub missing

[smartcooky]

Not necessarily. Thunderf00t was only speculating and he seems to be somewhat wrong about how the end caps are attached to the carbon fibre tube.

The tube was actually inserted into a slot in the titanium ring. The epoxy resin was only there to stop the end caps from falling off when on the surface and maybe to keep the joint water tight. At depth, the water pressure kept everything together.

My theory which is mine is that, if the carbon tube tended to shrink more under compression than the titanium end cap, then the carbon inside the slot would be under tension caused by the ring preventing it from shrinking, but outside the slot it would be under compression. That doesn't seem like it would be a good thing.

I'm not an engineer (neither is Thunderf00t), so the above should be taken as the speculation that it is.

Well, I am, and I can tell you its not an unreasonable suggestion.

Pound-for-pound, titanium is the strongest and least flexible metal on the planet - tungsten is the outright strongest. Now in theory carbon fibre has the potential to be as much three times stronger than titanium... that is why SpaceX initially considered using it for BFR (Starship) construction. However, once the alignment of the stressors departs from the alignment of the fibres, its strength drops dramatically - by the time the stressor alignment is 45° from the fibre alignment, carbon fibre has become about three times weaker than titanium.

I'm not sure what grade of titanium OceanGate used to build this thing, but to put some figures on the above...

Density (gm/cm²)
Grade 9 Titanium - 4.5
Carbon Fibre - 1.6

Ultimate Tensile Strength (MPa)
Grade 9 Titanium - 530
Carbon Fibre - 600 @ 0° and 90°, 110 @ 45°

Young Modulus* (GPa)
Grade 9 Titanium - 100
Carbon Fibre - 70 @ 0° and 90°, 17 @ 45°

* elasticity

To summarize, If what you say is correct, that the carbon fibre hull slotted into the titanium pressure dome (and if the carbon fibre hull was manufactured as a wound CPV) then that could be a weak point. If, as you suggested, the carbon fibre compressed more than the titanium, then in the area around that joint, the stressor alignment could start to be more misaligned with the direction of the fibres. As the pressure pushes down at right angles to the fibres.....

1. a small "ring" area just inboard (behind) the slot, flexes inwards.
2. the pressure is now pushing at an angle to the direction of the fibres at that point
3. this makes the hull weaker at that point.
4. goto 1

-- Repeat until the strength around that ring is so badly compromised that it collapses.

 

[Puppycow]

Slotting the tube into a groove in the titanium ring would make it a more robust joint than simply joining two flat surfaces with epoxy glue, but I think smartcooky is right that you would still get weakening at the section where tube meets the ring slot. It gets compressed more toward the middle of the tube than it is allowed to compress due to the titanium ring holding the ends, and the strain would be greatest along the line where the tube meets the ring slot.

 

[Puppycow]

Could NDT (non-destructive testing) have tipped them off?

Alleged Lack of NDT on Titan Submersible ‘Stuns’ Many in Industry

 

[Darat]

Mmm, you can't. Obviously, they must also be joined together somehow, otherwise they would come apart when the craft was out of the water, or even floating in the water. However, the joining method needs not be able to withstand the high pressure deep down, because that pressure will press them hard together.

Hans

I may be misunderstanding you or the concept as I would have thought it did have to be able withstand the highest pressures? If for example you used a glue that glue can't degrade under the highest of pressures else the craft would potentially start to leak, and even fall apart as you surfaced as the "pressure" joint loosens due to decreased pressure and the "glue" joint takes over the maintaining of the watertight seal and structural integrity of the craft?

 

[Shepherd]

I wouldn't have trusted this design no matter how well it was constructed, but it didn't appear to have been constructed well, either. Virgin Oceanic, of course, totally scrapped their plans with the one on which OceanGate's was based after acquiring it and conducting their tests.

Another issue that I don't think I've seen mentioned, btw, is that titanium is apparently a highly reactive metal that forms an oxide coating on its surface when exposed to air like, I believe, aluminum does. To get an adhesive bond with it, the surface needs to be specially prepared either through abrasion, or by etching with chemicals or lasers.

I don't know if any of this was actually done with the construction of the Titan, but after seeing the video of them wiping down the titanium ring with a rag and slathering on epoxy by hand after apparently mixing it with what looked like a paint stirrer, I kind of have to wonder.

Maybe someone knows?

 

[smartcooky]

Yes, I think that is right. Thus, conical windows, hatches, and such function because the pressure only pushes them tighter, without distorting them or the hul.
Hans

Nope, that's not how hydrostatic pressure works on submarines. Pressure is exerted from inwards from all directions on the outside of the hull, not just in one direction.

This is a very simplified illustration of the pressure on a submarine when it is diving. The length of the red arrows indicate the pressure - longer for higher pressure, shorter for lower pressure

During a dive the pressure is higher on the bottom-side, reducing to lower on the topside, and there is a pressure gradient from top to bottom. The faster the submarine is diving, the higher the pressure on the bottom-side, and lower the pressure on the topside and therefore, the steeper greater the pressure gradient.

If the submarine is also in forward motion, the pressure on the bow is increased; on the tern it is decreased. Again, the faster it is going, the greater the pressure gradient between bow and stern.

If the submarine is not diving and not in motion, the pressure is the same from all directions (apart from a very shallow gradient due to the bottom-side being a little deeper than the topside.
‎
‎

 

[MRC_Hans]

I may be misunderstanding you or the concept as I would have thought it did have to be able withstand the highest pressures? If for example you used a glue that glue can't degrade under the highest of pressures else the craft would potentially start to leak, and even fall apart as you surfaced as the "pressure" joint loosens due to decreased pressure and the "glue" joint takes over the maintaining of the watertight seal and structural integrity of the craft?

Of course, the joint must be water-tight, but that could be obtained by a gasket (as in the case of the hatch that any sane deep-water craft is designed with). And some retaining force will also be useful, but the point is that whatever it is, it does not need to hold anything like the full pressure at max depth.

Hans

 

[MRC_Hans]

Nope, that's not how hydrostatic pressure works on submarines. Pressure is exerted from inwards from all directions on the outside of the hull, not just in one direction.

[qimg]https://www.dropbox.com/s/5sn93436jvtrvrz/Pressure-hull.jpg?raw=1[/qimg]

This is a very simplified illustration of the pressure on a submarine when it is diving. The length of the red arrows indicate the pressure - longer for higher pressure, shorter for lower pressure

During a dive the pressure is higher on the bottom-side, reducing to lower on the topside, and there is a pressure gradient from top to bottom. The faster the submarine is diving, the higher the pressure on the bottom-side, and lower the pressure on the topside and therefore, the steeper greater the pressure gradient.

If the submarine is also in forward motion, the pressure on the bow is increased; on the tern it is decreased. Again, the faster it is going, the greater the pressure gradient between bow and stern.

If the submarine is not diving and not in motion, the pressure is the same from all directions (apart from a very shallow gradient due to the bottom-side being a little deeper than the topside.
‎
‎

Well, I can't see how this contradicts my point. Of course, if the craft is moving very fast, there can be zones around the hull that experience a negative pressure, at least at moderate depths, but that is not relevant for submersibles.

Hans

 

[PartSkeptic]

My point may have been dealt with already but I do not see it being raised in various YouTube videos.


What about the temperature differential?


At 4000 meter I think the temperature is about 4degC. If they start out at 25 degC then the faster they go down, the more there is a temperature differential across the carbon fiber.

A colder outer layer means that it is trying to compress the inner layer. This added force can be substantial.

It also means that the carbon fiber is shrinking with temperature and then expanding again. Each of these cycles is bound to cause micro-stresses and delamination.

In about 1974 I had some 11kV current transformers explode when commissioning a substation. Cheap alternative to the normal ceramic because they were made of epoxy. Due to delays they were not powered up for a period of 6 months. This let the cold nights and hot days heat and cool the units. They developed micro-cracks that then let in moisture. This moisture caused the catastrophic failure when energized.

The other problem is that the titanium would contract faster because the temperature would be more uniform. Then there is the difference in thermal coefficients. Did they even consider this in the design? Certainly did not test it.

The sub apparently went down in 90 minutes instead of 150 minutes. This would have made the problems much worse.

 

[PartSkeptic]

Another factor that I do not see raised in YouTube videos is the vastly increased pressure on the join.

The titanium end caps are being pressed towards the carbon fiber shell. Area versus pressure gives a huge number in terms of force. But that force is now distributed only on the area of the ends of the carbon fiber tube. The pressure multiplication gives massive pressure on the "glue". No way it could survive being pressed out.

I worked with problems encapsulating electronic circuit boards in epoxy resin. It became clear that the epoxy was not being stored in a a temperature and humidity controlled space, and then applied and allowed to set in the same area. It also turned out there was a reaction with slight contamination in the solder flux.

The mixing and the application of the epoxy in the construction of the sub was reckless and had no quality control. I am surprised the sub survived even one dive.

Even winding the carbon fiber tube has no quality control. Surely one does this in a vacuum chamber to eliminate any air voids?

 

[PartSkeptic]

One last comment. I worked in a electrical coal fired power station. There was enormous pressure in the pipes and equipment.

To join them there were bolts almost side by side around the flange. A red-hot heating rod was put in a hole in the center to expand the bolts. They were then tightened by two men using a 3-meter bar on the spanner.

We were instructed that if we heard a whistling sound indicating a leak, we were not to run our hand over the pipes to feel for the leak. One man did this with a hole less than the diameter of a human hair and it neatly sliced all his fingers off like butter.

 

[jeremyp]

To summarize, If what you say is correct, that the carbon fibre hull slotted into the titanium pressure dome

Not quite. There's a titanium ring with a slot in it into which the carbon fibre tube is seated. The titanium hemisphere is then bolted on to the ring. I think this is effectively the door.

Here's a short video showing the construction. At 1:52 there is a shot of the underside of the ring showing the slot. It is not as deep as I had thought. And the "lips" of the slot are themselves not very thick (see attached screen shot).

(and if the carbon fibre hull was manufactured as a wound CPV) then that could be a weak point. If, as you suggested, the carbon fibre compressed more than the titanium, then in the area around that joint, the stressor alignment could start to be more misaligned with the direction of the fibres. As the pressure pushes down at right angles to the fibres.....

1. a small "ring" area just inboard (behind) the slot, flexes inwards.
2. the pressure is now pushing at an angle to the direction of the fibres at that point
3. this makes the hull weaker at that point.
4. goto 1

-- Repeat until the strength around that ring is so badly compromised that it collapses.

The same video also shows them constructing the tube. It seems like they are just windoing the carbon fibre on there in one direction only, but that might just be the way the film is edited.

 

[Myriad]

I was only speculating back when the debris was first discovered, but it looks like I was on the right track:

I'd worry most about some unforeseen cumulative change as a result of repeated differential responses to the duty cycle of pressures and temperatures. Some kind of "creep," speaking very generally. Imagine, for instance, a crack forming inside the wall of the carbon fiber tube, parallel to the curve of the wall, near where it's bonded (no idea how they're actually attached together) to one of the end caps, caused by differential expansion and contraction during changes of pressure and/or temperature. (It's hard to explain where I mean. Shove a movie ticket into one end of a full roll of toilet paper, between the paper layers, a half inch from the outer circumference. That's the place I mean.) The tube is spec'ed to resist radial pressure or tension across or along the whole tube, but perhaps unexpected tension across the thickness of one wall had an unexpected effect that wasn't tested for, and wouldn't be visible without removing the end caps.

Probably not specifically that, but something of that nature.

More tests at well above the expected pressure should have been done, in between manned excursions. Obviously, in hindsight.

 

[jeremyp]

I was only speculating back when the debris was first discovered, but it looks like I was on the right track:

Titan had made thirteen dives to Titanic already. If there wasn't some progressive degradation, it would be difficult to explain why it didn't fail until the fourteenth trip.

 

[Myriad]

Titan had made thirteen dives to Titanic already. If there wasn't some progressive degradation, it would be difficult to explain why it didn't fail until the fourteenth trip.

I'd be very interested to learn more about the sharp sounds that were heard during previous dives, which might very well have been cracks forming or expanding in the carbon fiber tube. In what stages of the dives did they occur? How frequently? What did the operators think had caused them?

Obviously such cracks couldn't have been through the thickness of the tube, or it would have failed immediately. The direction of the stress from differential expansion and contraction of the end caps would, it seems to me (without running some computer model), tend to cause cracks inside the wall of the tube, parallel to the fibers. High external pressure would tend to squeeze such cracks closed. They wouldn't have had a large impact on the overall resistance of the tube to compression, until some number and/or size of them had accumulated.

Another possibility or possible contributing factor is the failure of the adhesive in tension. Where would that happen? There were three adhered surfaces between the titanium coupling rings and the tube. The annular surface on the flat end of the tube (perpendicular to the main axis of the craft), the inner cylindrical surface between the inner rim of the titanium ring and the inner wall of the tube, and the outer cylindrical surface between the outer rim of the titanium ring and the outer wall of the tube. The latter (outer) adhesive surface would be in tension, when high external pressure causes the diameter of the tube to contract (or tend to contract) more than the diameter of the coupling rings and end caps. And the annular surface would also therefore experience shear. Again, any resulting failure would most likely be incremental and cumulative, because the inner cylindrical adhesive surface would continue to be compressed at depth, and no leak through the glued joint could occur until that part of the adhesive was breached.

 

[jeremyp]

I'd be very interested to learn more about the sharp sounds that were heard during previous dives, which might very well have been cracks forming or expanding in the carbon fiber tube. In what stages of the dives did they occur? How frequently? What did the operators think had caused them?

I found a video that was an interview with another submersible operator who went on the second dive (the sub's second dive overall, not its second dive to the Titanic). I linked it in an earlier post on this thread, if you want to see it.

Anyway, He claims it was making loud noises all the way down and all the way back until about 300 metres. He characterised them as being like small calibre pistol shots.

He declined to go on any further dives.

 

[Myriad]

I found a video that was an interview with another submersible operator who went on the second dive (the sub's second dive overall, not its second dive to the Titanic). I linked it in an earlier post on this thread, if you want to see it.

Anyway, He claims it was making loud noises all the way down and all the way back until about 300 metres. He characterised them as being like small calibre pistol shots.

He declined to go on any further dives.

Thanks, I did see that one. I wasn't sure if it was doing that on every dive, though. (And I forgot the pistol-shot description, which is quite disturbing). The "what the **** did the operators think was happening?" question really bothers me! Ships of all kinds have "creaked" and "groaned" under operating stresses since time immemorial, but sharp sounds suggest sudden displacement of parts relative to other parts, which should not have been happening in any part of a submersible under pressure.

My suspicion is that cracks inside the tube wall, running in (cylindrically curved) planes parallel to the carbon fibers, forming or working open, made the pistol shot sounds. And that the operators knew that. But because the overall strength of the tube had some supposed large "safety" margin, they didn't think it would lead to anything worse.

 

[bruto]

One last comment. I worked in a electrical coal fired power station. There was enormous pressure in the pipes and equipment.

To join them there were bolts almost side by side around the flange. A red-hot heating rod was put in a hole in the center to expand the bolts. They were then tightened by two men using a 3-meter bar on the spanner.

We were instructed that if we heard a whistling sound indicating a leak, we were not to run our hand over the pipes to feel for the leak. One man did this with a hole less than the diameter of a human hair and it neatly sliced all his fingers off like butter.

I've seen similar warnings for such things as fuel injectors, which won't cut your fingers off but will tattoo you in ways you may not have intended.

Again in a somewhat less drastic manner the stretching of bolts is also something one runs into sometimes in automotive applications. I used to do some work on Mercedes, and their head bolts were designed to be tightened to a specific torque and then exceeded, not by specified torque but by degrees of rotation. Bolts thus stretched were not supposed to be reused.

 

[MRC_Hans]

I've seen similar warnings for such things as fuel injectors, which won't cut your fingers off but will tattoo you in ways you may not have intended.

Again in a somewhat less drastic manner the stretching of bolts is also something one runs into sometimes in automotive applications. I used to do some work on Mercedes, and their head bolts were designed to be tightened to a specific torque and then exceeded, not by specified torque but by degrees of rotation. Bolts thus stretched were not supposed to be reused.

Ah, the old "As far as it will go and then one more turn."

Hans

 

[bruto]

Ah, the old "As far as it will go and then one more turn."

Hans

Sort of, except that these were actually special bolts with an engineered neck, quality controlled alloys, etc., designed to produce just the right spring at the specified torque and rotation. Imagine PS's monster boiler bolts reimagined by Dr. Strangelove's eccentric cousin who lives in the basement and makes scale model spaceships.