Humor Random Comments Thread

[Thunder Chicken]

Briefing live now. Official statement confirms that it was a catastrophic casualty with loss of all hands.

 

[TheShuttleExperience]

a rapid implosion would probably be the the most merciful way this could have ended. The thought of being sealed in that can, anywhere in the sea, slowly dying with five people for days, hurts to think about.

Absolutely! Actually the news regarding debris felt like a slight relief to me. Better quick than long and paynful.

But it really raises questions regarding the design and the company. I doubt they continue to offer that "adventure"...

 

[Thunder Chicken]

They stated that they found the two titanium hemispheres capping both ends of the pressure vessel. From a structural standpoint, the cylindrical composite section would be a weak point as its primary buckling mode failure would be for a point on the wall to fail in tension causing the cylinder to flatten, with only the end hemispheres able to support the circular cross section. Spheres are far more resistant to buckling failures, all other things being equal.

Some poor young engineers on that team are getting a slow, horrible lesson in structural failure modes right now.

 

[Linguofreak]

They stated that they found the two titanium hemispheres capping both ends of the pressure vessel. From a structural standpoint, the cylindrical composite section would be a weak point as its primary buckling mode failure would be for a point on the wall to fail in tension causing the cylinder to flatten, with only the end hemispheres able to support the circular cross section. Spheres are far more resistant to buckling failures, all other things being equal.

"Fail in tension" meaning to actually tear, or just to deform without tearing?

In the former case, I wonder how much flattening you'd actually get before fluid dynamics through the hole dominated the event.

In the latter, I suppose the point of first water entry would be when the seal between the hemispheres and the cylinder separated as the cylinder flattened?

 

[Thunder Chicken]

I wonder if those engineers thought that since 6000 psi composite internally loaded pressure vessels work well, that a 6000 psi externally loaded pressure vessel would behave identically, and that was their complete analysis. But unfortunately, signs matter a lot.

Composites are not homogeneous materials and their properties depend on the interaction of the reinforcing fibers and the matrix as well as the mode of loading. Reinforcing fibers are very strong in tension and are meant to bear most of the tensile load instead of the polymer matrix. But composites generally are terrible in compression. In compression, those fibers can easily buckle and fail at stresses much lower than their tensile strength. This transfers the compressive load to the polymer matrix which isn't nearly as strong as the reinforcing materials, which can cause more strain and more fiber buckling and failure.

You technically can use composites in compressive applications, but to do so safely you need to limit compressive strain on both the fibers and the matrix, meaning you need need thicker sections to lower the applied stresses on the composite. Such applications don't really play to the strengths of composite materials as, in a best case scenario, the reinforcing fibers aren't taking much load and you basically are designing your part based on the properties of the low strength polymer matrix. For the weight, metals likely would be a better and less complicated design choice for compressive applications.

I hope they can recover something of the composite cylinder section and learn some lessons from it.

 

[Ripley]

... Things that work well on paper tend to get beaten to death when they go into the sea.

Indeed...

 

[TheShuttleExperience]

I wonder if those engineers thought that since 6000 psi composite internally loaded pressure vessels work well, that a 6000 psi externally loaded pressure vessel would behave identically, and that was their complete analysis. But unfortunately, signs matter a lot.

Composites are not homogeneous materials and their properties depend on the interaction of the reinforcing fibers and the matrix as well as the mode of loading. Reinforcing fibers are very strong in tension and are meant to bear most of the tensile load instead of the polymer matrix. But composites generally are terrible in compression. In compression, those fibers can easily buckle and fail at stresses much lower than their tensile strength. This transfers the compressive load to the polymer matrix which isn't nearly as strong as the reinforcing materials, which can cause more strain and more fiber buckling and failure.

You technically can use composites in compressive applications, but to do so safely you need to limit compressive strain on both the fibers and the matrix, meaning you need need thicker sections to lower the applied stresses on the composite. Such applications don't really play to the strengths of composite materials as, in a best case scenario, the reinforcing fibers aren't taking much load and you basically are designing your part based on the properties of the low strength polymer matrix. For the weight, metals likely would be a better and less complicated design choice for compressive applications.

I hope they can recover something of the composite cylinder section and learn some lessons from it.

I have no engineering background and numbers at all. But when I first learned that composite materials are involved for the hull, I was actually afraid It's only a feeling. But it just doesn't feel well. This thing was supposed to go down 4,000 meters and up again on quite a regular basis. The pressure and pressure changes acting on it is out of this world if you will. Also, I read that the port hole was only designed for 1,300 meters debth. But I don't know if this is true.

Seems the design/vehicle was anything but ready for such trips.

 

[MaxBuzz]

top most stylish underwater vehicles

 

[Thunder Chicken]

I have no engineering background and numbers at all. But when I first learned that composite materials are involved for the hull, I was actually afraid It's only a feeling. But it just doesn't feel well. This thing was supposed to go down 4,000 meters and up again on quite a regular basis. The pressure and pressure changes acting on it is out of this world if you will. Also, I read that the port hole was only designed for 1,300 meters debth. But I don't know if this is true.

Seems the design/vehicle was anything but ready for such trips.

I didn't know anything about the company or the dive until they went missing. When I heard that the pressure vessel had a cylinder of composite I physically cringed. It's not an impossible material to use or automatically wrong, but it's terrible in compression and designing with it needs some specialized understanding of the material.

They may have been motivated to use composites due to the small size of the vessel, and the desire to get five people into it instead of 2-3. Smaller submersibles will generally weigh more per unit of volume than larger submersibles of the same construction, and at some point a small submersible made of metal may not have net positive buoyancy even without ballast. Stuffing the weight of more people into the same displaced volume makes this problem even worse. The austerity of the interior with no seats or any other amenities may have also been weight-saving measures as well as to minimize cost.

However, that doesn't change the fact that composites are difficult to use safely in compression, and I'd wager a properly designed and safe hull using composites would require a thickness that would make the weight-savings negligible. I really think that a well-meaning but inexperienced engineer took composite properties, cranked them through a stress analysis, and got numbers that made net positive buoyancy possible, being completely unaware of the low-cycle fatigue buckling behavior of composites in compression.

I have heard the rumors about the window being only rated for 1300 m - I am unsure whether that is true or not. Practically they proof tested those parts of the hull on previous dives and the hemispherical sections of homogenous materials usually hold up well under repeated loading unless they were drastically under-designed. If any part of that sub were to fail under cyclic loading it would be the composite hull.

 

[nbcfrosty]

I was watching an interview of an expert online and he said that traditional hull materials like steel and titanium are more "compliant" than carbon fiber, and that the traditional materials provide certain integrity even during an implosion. What does that mean?

 

[Thunder Chicken]

In the former case, I wonder how much flattening you'd actually get before fluid dynamics through the hole dominated the event.

In the latter, I suppose the point of first water entry would be when the seal between the hemispheres and the cylinder separated as the cylinder flattened?

Buckling failures happen in the space of milliseconds. At that depth the pressure was just shy of 6000 pounds per square inch. The idea of fluid flowing and leaks and water entry and such is not even relevant. The hull was instantaneously crushed by a column of water over 2 miles high. The air in the sub at 14.7 psi collapsed to 1% of its original volume in that instant.

This is a video showing a cylindrical steel tanker buckling under atmospheric pressure of "only" 14 psi.

Now consider that the pressure on the Titan was over 400 times higher when the hull failed.

 

[Urwumpe]

Also a composite structure doesn't buckle like that. Its way more sudden and violent.

For comparison, Al-Li:

And this is the sound, their acoustic monitoring system was looking for. The crew likely heard it too during descend, I doubt it started without slow progressive failure at lower pressures.

 

[Linguofreak]

Buckling failures happen in the space of milliseconds. At that depth the pressure was just shy of 6000 pounds per square inch. The idea of fluid flowing and leaks and water entry and such is not even relevant. The hull was instantaneously crushed by a column of water over 2 miles high. The air in the sub at 14.7 psi collapsed to 1% of its original volume in that instant.

I'm aware of the timescale involved. But there is still a sequence of events over that timescale and the question of what the collapse looks like on a microsecond scale.

Fluid isn't going to exactly "flow" under these conditions, but whenever the failure produces the first opening in the hull, you're going to have an interface between 400 atm water and air at a considerably lower pressure with nothing in between, and that is not a static situation, so whether or not words like "flow" and "leak" apply (note that these were not words that I used), "fluid dynamics" and "water entry" are going to apply with extreme prejudice.

The timescale for the whole event is a few milliseconds, but I'm interested in the relative timescales of the deformation part of the failure, the tearing/cracking part of the failure, and whatever fluid dynamics occurs (including stuff like widening of the opening by erosion, which I imagine is significant). In the end, all pressures are equalized and the pressure vessel is debris, but what process dominates the event overall, and what stages are there to the event?

Also, what temperatures can we expect to develop?

 

[TheShuttleExperience]

There are reports now that the implosion was recorded and already known on sunday.

 

[Urwumpe]

I'm aware of the timescale involved. But there is still a sequence of events over that timescale and the question of what the collapse looks like on a microsecond scale.

The timescale can be best described as "By all means instantly". See above how composite structures fail. Its not like its first gets a bit wet and then more and more leaks open (and at this pressure, even tiny leaks can cut through humans and fill a small volume with water in minutes).

First you hear nothing special, the usual sound of the pressure vessel getting compressed just changes subtle and becomes sharper. Still no leak. But the first small fibers are torn. Now its the moment that you should surface ASAP and pray that the many layers and 5 cm thick hull are enough.

While the failure goes on, more and more fibers will break from the outside to the inside. All the crew inside notices about it, is the sound. The frequency at which the fibers fail increases exponentially, the integrity of the hull drops fast and then, all breaks in a few milliseconds. Especially if its also cold outside, the material will not be very elastic and simply scatter into small pieces when the energy is released.

The air bubble protected by the hull will simply collapses to a fraction of its volume, the persons inside not even have time to think about explosive compression. At these speeds and pressure differences, the crew is likely turned into jelly within a blink of the eye.

If you would have a 1 km (vented) tube, and fill it with water at that pressure of about 20 MPa, its completely filled within 1/6th of a second.

 

[GLS]

Interesting tidbits from "inside the community"

 

[GLS]

Look who also went down on that thing...

 

[jedidia]

I wonder if anybody will ever come up with a dsl that enables you to define trees in an actually convenient way in code...

 

[Thunder Chicken]

I was watching an interview of an expert online and he said that traditional hull materials like steel and titanium are more "compliant" than carbon fiber, and that the traditional materials provide certain integrity even during an implosion. What does that mean?

I don't know who this "expert" was, but if your vessel implodes it would make very little difference to you whether it was made of titanium or steel or carbon fiber. Failure is failure.

As for the material properties, metals generally are tougher than composites and have an elastic region where applied load will deform them, but when the load is removed the part will spring back to its original shape. If you take a paper clip and use it to hold a few sheets of paper together, it will spring back to its original shape when you remove them. If you load beyond that elastic region, metals will plastically deform - they don't break or fail necessarily, but if they are unloaded the part won't return to its original shape. If you take that same paper clip and pry it open to put it on too many sheets of paper it won't spring back. Many metals can take on quite a lot of plastic deformation before they fail completely, so they can give you some warning of impending failure if the parts are designed correctly.

Composites are very stiff and relatively brittle. They are strong until they suddenly aren't. Being stiff, they don't deform very much until they begin to fail, and once they begin to fail they tend to accelerate to failure exponentially fast.

 

[Thunder Chicken]

I also wonder if maybe the composite hull was actually designed appropriately for the pressure loads (it did make several dives, after all), but perhaps it took some seemingly insignificant impact while the vessel was being handled during launch or recovery that compromised it. The composite hull itself was covered by fairings, but maybe these fairings or the titanium dome took a hit during handling that transferred the impact force to the composite. It would not take much local damage to compromise the strength of that hull.