Slashdot Mirror

These vehicles typically use glass spheres (containing air) for buoyancy. We also experimented with a ceramic "foam" - basically millions of tiny hollow glass beads glued together and molded to fit into unused portions of the vehicle. They're more reliable, but don't provide as much floatation for a given volume. Any equipment and electronics you have on board has to go inside the larger glass spheres, so you always have some of those aboard.

The founder of Benthos gave us a lecture on the tech. When one of them implodes, the energy released is the ambient pressure at depth times the volume. This is why you can't simulate an implosion in a tank - the moment the implosion begins, the pressure drops, and the energy release stops But at depth, the water simply fills in any lost volume with more water at the same pressure, and the energy release continues until the entire volume of air is crushed. The smaller sphere in the link (13 inch or 33 cm diameter) has an air volume of about 15 liters. At 9000 meters, the pressure is 90.57 MPa. So its implosion releases 1.36 MJ - about as much energy as 2/3rds of a stick of dynamite. The glass spheres which implode basically revert back to sand. You can imagine how much more energetic it is if something as large as a submarine implodes.

It did NOT break up at altitude. Something rendered the aircraft uncontrollable. A loss of hydraulic pressure or power does this for a 777.

A loss of hydraulic pressure or power does not do this for a 777. It has a RAT (ram air turbine) which pops out in such cases. Basically a big propeller which gets turned by the wind as the plane glides at 500 mph and generates enough power rudimentary electronics (including radio) and hydraulic pressure. That's what happened with the Gimli Glider - a 767 mistakenly loaded with insufficient fuel (the original boneheaded imperial vs metric conversion foul-up before the Mars Climate Orbiter). which basically turned into a 100 ton glider when it ran out of fuel mid-flight. The RAT popped out and allowed the crew to control the plane to a safe landing. (Which of course means if this did happen on MH370, the search area needs to be much larger than where they're currently looking).

Hydraulic failure usually involves structural damage which compromises all the hydraulic lines. Most commercial aircraft have 3 independent hydraulic systems; some have 4. If there's damage which severs lines in all of those systems, the plane can "bleed" hydraulic fluid until there's not enough left to control the flight surfaces. I believe the 777 used a hybrid fly-by-wire + hydraulic system though, where pilot commands are transmitted to the flight surfaces by wire, and a hydraulic pump there moves the flight surface. So severing the hydraulic lines may have killed one control surface, but not all. (Severing the wires OTOH...)

Anyway, I'm skeptical that it broke up at altitude too. That usually generates a lot of floating debris (papers, luggage, clothing, bodies, etc.) scattered over a wide enough area that the crash area is quickly located. The pingers should be firing away so it's just a matter of one of the search boats traveling within a few miles from the plane's resting location. (KAL007 wasn't located because the Soviets knew from their radar tracks where it went down, and set up decoy pingers far away to get the U.S. and South Korea to search the wrong location).

A part (half?) of the sphere will have to be removed to allow people/things in and out (unlike "ecospheres") it can't be seamlessly sealed. Isn't that the most likely place of failure?

Benthos already makes glass floatation spheres (they provide buoyancy for deep-water submersibles and platforms). They're cut so the two halves fit perfectly (or as near to perfect as mechanically possible with current technology). It's precise enough that even though the cut appears flat to the eye, the two halves are keyed - you can't offset the angle at which they mate or there will be microscopic gaps/ridges. Also in their tests, they found that a glass-on-glass mating worked best. Rubber or a gasket is useless because under extreme pressure it gets deformed and squeezed out. If it doesn't deform uniformly (maybe one section of the rubber is harder), it's actually worse than not having a gasket.

The hard part is keeping the two halves mated at 1 atmosphere (which is why you see tape on the middle in the picture). Underwater, and especially at depth, the water pressure alone is enough to hold the two halves tightly together. Friction increases with the normal force, so at depth there's a helluva lot of friction preventing the halves from sliding relative to each other. The glass would probably shatter before any impact would cause the two halves to slide apart.

In 2003, I was on a small team using a similar WHOI system called REMUS to take surveys of ports and waterways, looking for mines. We had been training with the system, mainly in Southern California and when Iraq started up, they decided that they wanted to try the new technology there. We did, and it was successful. http://www.joetalbot.net/pages/030401-N-3783H-075A.htm In the course of our training, we managed to get the things stuck, beached and lost several times despite a system that would take it to the surface when it's little PC-104 mind was blown. Remember, there are things down there (plants, critters, caves, rocks, ships and junk) that are unknown to us before these things are launched (that's often why we launch them, you know? To take surveys?). These things navigate underwater using a small network of buoy "transponders" the respond to pings from the unit. The slight response delay (caused by the water and very predictable) tells the unit how far it is from the buoy, who's locations are known to the units (windows) programming software. These things are pretty cool and useful, but they're also kind of primitive. We used to attach a "pinger" to the units so that when they got stuck, we could recover them with divers. A pinger is a small capsule that contains a device that periodically emits a 20-70 khz "ping" that is easily picked up by a hand held, very directional receiver carried by a diver. http://www.benthos.com/undersea-pingers-locators-product-overview.asp Given what the system is worth, I would imagine that it has a pinger for location as well.

• A descent weight, used to make the craft negatively buoyant for the initial descent, then dropped to leave it neutrally buoyant.
• An ascent weight, dropped at the end of the mission to make the neutrally buoyant craft positively buoyant.
• A bladder which can be pumped with oil from a reservoir tank to fine-tune buoyancy.

Air doesn't work because of the enormous pressure involved. A 3000 psi scuba tank could only inflate a balloon down to about 2000 meters. Below that, the water pressure is greater than that inside the tank, and opening the valve would result in water forcing the balloon into the tank, rather than air inflating the balloon. A 10000 psi high pressure tank would work at 5000 meters, but would only result in about a 30% increase in volume, meaning you'd need a very big tank to be able to raise the entire craft in a catastrophic failure. Furthermore, the air would expand as the craft rose, risking rupturing the balloon. That's why the buoyancy control uses an oil bladder - oil is relatively incompressible.

Dropping the ascent weight helps raise the craft at the end of a mission. But usually they're relatively lightweight so you can attach them manually. The 17-inch glass spheres typically used to house equipment provides over 50 pounds of buoyancy. The failure of one of these spheres at a depth of 3000 meters (~4500 psi) would release (4500 psi) * 4/3 * pi * (8.5 inches)^3 = 1.3 MJ of energy. A stick of dynamite is about 2.1 MJ, so losing one sphere is pretty much guaranteed to cause all the other spheres to fail. If the remainder of the craft somehow survived all that energy release, the loss in buoyancy would overwhelm what buoyancy you'd get by dropping the ascent weight.

With 2 beacons with angle (and depth) information it could determine its position without knowing the speed of sound in water (time * speed = distance), since the distance between the two beacons is known.

If additional beacons with known positions where placed then with 3 beacons the sub could also triangulate its position without angle information and without sending a ping.

Systems like this already exist:
http://citeseer.ist.psu.edu/184601.html

Using a GPS underwater exists too:
http://www.patentstorm.us/patents/6807127.html
http://www.longbeachdive.com/und-water-gps-product s.htm

Add a modem
Sonar communication (sonar modems):
http://www.benthos.com/acoustic-telesonar-modems-u ndersea-sub-sea.asp

And a directional receiver (antenna)... to determine angle...
(just needs two fast enough receivers + processing a known distance apart
(think ears). probably also not new technology .

Essentially its a slight variation on sonar triangulation
It goes back to basic trig...

This isn't that new and it sounds more like patent trolling to me...

But that's just based on a few minutes of googling and some back of the envelope trig,
not the opinion of a qualified expert... and I might be wrong...