The Titanic Submersible

It has ties to space exploration.

There are a lot of parallels with space tourism here. It is a high-risk, high-cost activity for wealthy individuals, that is currently unregulated. It is a visit to a harsh environment (worse than space in some ways), where life support is required, and finite. I hope that this incident doesn’t result in clamoring for regulation, of either submersibles, or space adventures. In the case of the submersible, it’s not clear who would regulate something that’s taking place in international waters.

[Update a few minutes later]

[Noon update]

One of the people on the boat is an adventure addict.

Well, that’s the sort of people who do this sort of thing.

[Wednesday-morning update]

It’s looking more and more like this was negligence, and the people who did it were daredevils.

And then there’s this.


[Noon update]

The company was warned twice of potentially catastrophic consequences. It will be ironic if Rush is killed by his own decision, but it shows his misplaced confidence in it.

[Thursday-afternoon update]

“They likely didn’t know what hit them.”


[Update a while later]

Thoughts from James Cameron:

73 thoughts on “The Titanic Submersible”

  1. What I find most disturbing is the inability to contact the submersible. In today’s day and age I find that simple fact to be the most disturbing and a probable indicator that there are no survivors.

    One would think that a survivable mechanical failure that would prevent the submersible from surfacing would require that they settle on the sea floor as close to the Titanic wreckage as possible. They supposedly have a few days worth of air aboard according to a radio report I heard. Their exact path over the wreckage with time hacks at each point along the way would be known to the support vessel over the site.

    High risk tourism is exactly that. You might as well make climbing Mount Everest illegal as well. It takes an annual toll too. Over the years well in excess of these types of events. There are still frozen bodies along the path near the summit.

    1. The company was warned twice of potentially catastrophic consequences. It will be ironic if Rush is killed by his own decision, but it shows his misplaced confidence in it.
      The problem with you folks here is that you are missing the fact that the company needs to make money and hold to a schedule. You need to remove your engineering hat and put on your management shroud!

      1. “You need to remove your engineering hat and put on your management shroud!”

        Does a management shroud look anything like a funeral shroud?

    2. If you listen to the whole vide put out by SubBrief, at the end he discusses the lack of communication with the submersible. It was a feature, not a bug. See the CEO got annoyed with the constant status checks from the mothership that happened every 90 minutes. He didn’t want to be disturbed while exploring the ocean depths.

      I get it, but if you are selling tickets to customers, you should show some respect for their safety. Alas, it seems they all died suddenly and communication wouldn’t have saved them. So the lack of comms decision just wasted time for others.

  2. “There are still frozen bodies along the path near the summit.”

    I heard it said the the slopes of Mt. Everest are littered with the bodies of highly motivated people.

    Also recently came across an article discussing how much trash and debris (and not just the bodies) is littering the routes, and what would need to be done to start cleaning it up. It appears nobody takes any of that equipment back for the deposit, or to use again.

    Like taking private jets to climate conferences, it seems that going to the summit is a way for the rich to flaunt their disposable wealth through conspicuous consumption.

  3. The legal doctrine of “tesseram tuam emisti, casum tuum accipis”* would seem to apply, both to the Titanic sub disaster and potential space tourism,

    *you buy your ticket, you take your chance”.

  4. Some are saying contact was lost during descent, which to me suggest a catastrophic leak or implosion. Almost all small submersibles have droppable ballast that will quickly return an non-flooded sub to the surface, even if all other systems fail.

    If it didn’t reach the wreck then it couldn’t be snagged on the wreck and couldn’t have run into part of the wreck and been catastrophically damaged in a collision.

    And it rules out suicide by a tourist who is obsessed with the Titanic and who wants to lie there forever with the long-dead passengers, which is a possibility given that almost anyone who pays $250K to visit the Titanic is obsessed with the Titanic.

    One thing that hasn’t been ruled out is a giant squid attack on the way down. Maybe a kraken got them.

    1. “Almost all small submersibles have droppable ballast that will quickly return an non-flooded sub to the surface, even if all other systems fail.”

      Not this one, though. I guess such a mechanism wouldn’t have been inspirational.

  5. Maybe some orcas did it?

    When I was a marine machinery mechanic, my life oscillated between acrophobia out on the scaffolding, not wanting to fall in the drydock (only forty feet, but still…) and being in a tight spot inside the boat and getting stuck, as sometimes happened.

    If these people don’t turn up alive (and traumatized), I hope it was an implosion, which would be over in no more than a few tens of seconds, tops. Otherwise, suffocating with your pals at the bottom of the sea for a few days…? Remember the Kursk.

    1. “I hope it was an implosion, which would be over in no more than a few tens of seconds, tops.”

      I said pretty much the same thing to my wife this evening when she asked if I had heard about this. I just hope they find the vessel this calendar year, for the sake of the families.

        1. Well, it would depend on many variables. Submarine implosions are not as depicted in TV and movies. And to be careful here, this stuff is not from what I learned working for DoD. I checked it against easily accessible public info, primarily a standard treatise on explosive forming, and descriptions written by my distant kinsman, Otis Barton (designed and built the original bathysphere):

          5,000 psi (less than 300 bar) is not enough to get microsecond responses, even in the unlikely event of an instantaneous implosion. You get down to 1K microseconds at around 40,000 psi, down to a single microsecond at 4,000,000 psi.

          That said, the likelihood of an instantaneous implosion is low. More likely, you’d hear a loud GRONK! You’d have time to scream with agony, and then the rising pressure would kill you. At least a couple of seconds. During an unmanned test of the Bathysphere, it leaked and came back up filled with pressurized water. Otis Barton felt that the spray from the leak would be forceful enough to cause fatal injury, like being machine gunned.

          1. I did like the depiction in the movie “The Abyss” whereby the protagonist (Ed Harris) dons an experimental super deep water diving suit which forces him to “breathe” highly oxygenated water. As he descends deeper and deeper the robotic camera he holds in his hand eventually implodes.

  6. *you buy your ticket, you take your chance”

    aka “play stupid games, win stupid prizes”

    1. People who want power will use whatever event to create their narrative. They might even make something up. It all depends on how much that power is worth to them. There is no money or power angle with this event.

      There is, and will be, lots of money and power in space.

      This incident is more of an oddity. The public is only interested because of the novelty. They think the people are weird but feel some sympathy for them. There isn’t a money and power angle to motivate government to get involved and no pressing public peril.

  7. If the Canadian sono-buoy information is correct, it would argue against an implosion. From what I’ve read, I doubt that they were down all the way to the Titanic before comms were lost. To me, this suggests a major system failure while descending, possibly a power and/or power control failure. Now if this was a well designed submersible, there would have been an emergency ability release of ballast so that the submersible would automatically surface. If, however, the release mechanism needed power (dumb), they would continue to sink until they bottomed out. In that case, they are doomed unless an ROV can snag them with a cable that can haul them up. I wouldn’t put any money on that possibility.

    1. They were headed to almost 4000 meter depth, in a submersible design tested to around 1400 meter depth.


  8. “You’d have time to scream with agony, and then the rising pressure would kill you. At least a couple of seconds.”

    Adiabatic compression of the air. Huge temperature rise.

    1. I did some back of envelope calcs using the ideal gas law – at 8000 depth the temp rise would be on the order of 149,000 degrees F.

      Fast, too.

      1. What you’re not taking into account is the heat capacity of a human. Let’s say there are 10 cubic meters of air, massing 10 kg. There are 400 kg of humans (mostly water). Water has 6 times the heat capacity of the air.

        So the air is compressed to 400 atmospheres, all else being equal it would rise to 400 x 300 K, or 120,000 K. But with the mass of the humans, the equilibrium temperature would be roughly 1/240 of that, or 500 K or 440 F.

        Of course, the system would not be in equilibrium – at least three major factors intervene:
        1) The human actually takes a long time to heat up – you get a skin burn, not cooked through
        2) The air around the human cools, and does not mix quickly with the hot air – this protects the human further
        3) The reason the air has compressed is that another 10,000 kg of water has replaced it. This annihilates any heating before it even starts, as the water will not appreciably warm.

        Mind you, the human is not alive to notice it!

  9. I find the recent articles mocking or scoffing at the use of a game-style controller to be annoying at best and misleading at worst.

    For something like this submersible, a known quantity piece of COTS hardware with reliability reports and a bog standard interface is a much better choice than some custom designed physical control system or other custom fly-by-wire scheme with proprietary hardware. Why reinvent the thing?

    I also find it ironic that the owner was on there and that he may have ignored warning signs. Better that he not have to deal with the fallout of completely upending three families besides his own.

    That said, I’d probably want to be slightly more aware of the relative karma scores of everyone I was climbing on board with if I were to do something like that myself. No point in tempting fate if one of the others on my adventure is teed up for a righteous smiting…

  10. A pinhole leak at 3,000 PSI, 200 Bar will easily take off fingers within several inches. A slightly bigger leak will take the hand, arm or neck. The subsequent infection and necrosis from the injected material generally requires a high amputation, it that’s an option. Fatalities are common. It’s why you never run your hand along a pressurized hydraulic line, most especially when you suspect a leak.

    One of those old guys that were so uninspirational would have told them. was that there are two rules for this sort of submersible. First: There will be a ballast weight big enough to insure positive buoyancy under all circumstances. Said weight will be released through a mechanism as simple and foolproof as possible, most especially not mediated through a wireless game controller or any other electronic actuator. One exception is a weight held by an electromagnet that is released when power is interrupted.

    Second: Absolutely nothing is ever to get between the sub and the surface.

    Anything made from composite is, by definition, a first article. The only assurance is rigid adherence to fabrication procedures validated by full scale destructive testing coupled with the very best NDT that can be done. See SpaceX and all the money they spent on composites and ended up with stainless steel. This was not the ethos of these clowns. RIP

  11. I think the real tragedy is that nobody put up $500K to buy a trip to the Titanic for Harry and Meghan.

    1. The one thing everyone seems to be missing is the possibility it hit a piece of semi-submersed ice prior to descent.

      1. Alright, already. The irony of this craft involved in a disaster on its maiden voyage has already been noted.

        1. I like Green Mountain Morning Blend first thing. It always puts me in a happy mood prior to browsing.

    1. The first time I heard the company name, I thought it was a joke. Apparently people will still drink the KoolAid and pay the price.

      1. Well it does fit in well with the cable news cycle.

        Now for the latest on the OceanGate scandal we turn to our intrepid reporter Richard Wannabee…

  12. Contact was lost only two hours after submersion. I think an electrical fire took out their comms and the fumes likely took the passengers out thereafter.

        1. Thanks for that link. Still reading through it. Where do I send the coupon for a discount on K-Cups? 😀

    1. The part probably most of interest is that it requires testing to 4x the design pressure for 30 mins; 1x design pressure for 80,000 hours; and 1,000 design pressure cycles of 8 hours each.

      I gotta admit, 9 years minimum testing time is a little extreme…

      1. I think that’s to certify a window design that’s outside the range of what’s already been thoroughly tested and certified using those rigorous test methods.

        As a side note, such a window or viewport must also be inspected every year, though this is likely impossible on the ISS. I realized that was a huge issue when looking over the standard O’Neill type space colonies. Under the standard, the entire population would just spend their lives doing window inspections. ^_^

        That’s when I realized that ;lighting the interior using energy from high-efficiency solar cells or concentrating collectors and heat-cycle engine to power high-efficiency LEDs or sulfur bulbs was going to be far simpler and cheaper.

      2. The design pressure was at least 400 atm…so 1600 atm for 30 min?

        I don’t think that’s gonna work. And aircraft (human occupied pressure vessels) get certified in a lot less than 9 year.

        1. The pressure extreme and testing are just for the viewports, and the 9 year certification is for a new geometry that’s outside those already tested enough to produce the design formulas for them. So there’s a set of “standard” viewport types that you can just do by the handbook. Flat circles, conical shapes, ovals, bulging circles (fish eye?), a couple types of hemispheres, etc.

          The reason for the extreme testing is that Plexiglass and Lexan undergo slow plastic deformation under pressure, so they have to be really robust and subject to a tremendous number of full cycles to see if the plastic deformation and fatigue will eventually compromise them.

          The accepted design formulas are for windows where they tested a large number of windows of the same basic design and different dimensions to determine their limits, so there are charts and graphs for them based on thickness, diameter, and any particular edge geometries and mounting methods. The ASME book is quite interesting, at least to engineering geeks.

          There are quite a lot of stories about how Rush decided to go with his own window standards, claiming he had some kind of monitoring method that was superior to the accepted design limits. I’m betting he was wrong.

          If his window was indeed only certifiable down to 4200 feet and he took it to 12,400 feet, he was overstressing it by a factor of three compared to the formulas. As the testing procedure shows, the window should survive those kind of overpressures for quite a few cycles. And then it won’t.

          1. When the Trieste made its descent to the Challenger Deep in 1960, Lt. Don Walsh and Jacques Piccard were shaken by the cracking of the outer pane of the sub’s plexiglas window at a depth of about 9,000 meters. They kept going to 10,911 meters. That’s the definition of guts, in my opinion. I’d have dropped the ballast instantly.

    2. Why not just copy the Trieste? It went almost three times this deep over 60 years ago. Not a lot of original engineering going on here.
      And why not build it out of glass? Isn’t glass a liquid and therefore sealable against extreme pressures?

      1. Glass is actually a pretty good material for subs except that it’s very brittle, and marine engineering does not like brittleness.

        Concrete is another very good material for subs because concrete is cheap, easy to work with, and works well under compression. At one point the Navy was worried that one of our adversaries might field concrete subs. But it never caught on. I suspect brittleness and potential voids were considered a fundamental problem.

      2. That big tank above the crew sphere was mostly full of gasoline for buoyancy, it was open to the ocean at the bottom and floated on top. Water is one of the least compressible liquids while hydrocarbons are much springier. The gasoline could, in 1960, be contained in a thin walled tank exposed to the ocean and provided buoyancy without the weight and cost of providing a void. I couldn’t begin to count all the laws that sort of arraignment would violate today. The compression of the gasoline helped compensate for the increase in density of the water in a rather complicated relationship that was a big part of the genius of the design. The size of the Trieste was very much bigger than modern designs and would have required a much bigger support ship.

    3. Thanks for that link, George. Back in the early 1980s, I wanted to start a tourist submarine business. At the time, there had been only one such business, ever. We contacted the Coast Guard to find out what certifications might be required, and were informed (in writing) that there were none, and that there were no regulations pertaining to tourist submarines. Further, they encouraged us to pursue the business. I was never able to find a standard like this, but had plenty of pressure vessel design engineers at my disposal.

      The financing never materialized, unfortunately. Five years later, I took a submarine tour off the coast of Georgetown, Grand Cayman Island. It wasn’t as impressive as I had imagined, but part of that was surely location. Many of my dive trips have been in much more spectacular settings.

      At any rate, it’s interesting how much regulatory infrastructure has appeared around submarine tours. None of it is mandatory, it appears, and because of that I’m surprised that anyone would insure any part of OceanGate. I had a hell of a time finding an insurer for a reusable launch vehicle, back in my KST days, and that wasn’t for a passenger (or “spaceflight participant”) carrying vehicle. It was before the FAA released the first “final” rule regulating RLV flight, and part of the insurance problem was lack of a regulatory regime.

      1. I think the marine insurance industry has published some basic standards for submarine hulls, with dimensional formulas, though it’s been a few years since I dug into it.

        My interest in seriously building a manned submersible was sparked by a college housemate who wanted to make a sub out of an unused LP gas tank that was on his farm. The hull would’ve been too thin, and require internal bulkheads, but I dug into the problem to the extent that I could at the time. Many years later I revisited the idea of building a manned submersible many times, and used to frequent (which for some reason seems to be offline right now, though I don’t know how long it’s been down).

        One book they had which I read many times was
        Manned Submersibles by R. Frank Busby (1976) (PDF)

        700+ pages of submarine goodness.

        It didn’t have design formulas, but combine it with the ASME PVHO book and the hull design formulas from the shipping industry (which is online somewhere), and the really difficult design questions are easily answered.

        One thing noted was that by far the #1 reason that a homebuilt or hobby sub project fails is the inability for most people to accurately shape large compound curves out of reasonably thick sections of steel, which is generally going to require hot working.

        I looked into fiberglass construction and couldn’t help but conclude that only a fool would trust a hand-laid composite at depth.

        1. “I looked into fiberglass construction and couldn’t help but conclude that only a fool would trust a hand-laid composite at depth.”

          Same here. I’m actually contemplating a commercial unmanned submersible project, and am seriously considering marine concrete for the pressure vessel. For a human occupied vessel, it would be metal or nothing…with nothing being a non-starter, so I guess it would be metal.

          If you are amenable, please ask Rand for my email address. I’d very much like to correspond directly with you.

    1. I loved how entire instrument panels would go up in flames on that show. To be dowsed by CO2 fire extinguishers. Then flip a switch and back in business!

    2. Also note they had huge watertight doors that sealed that section off from the control room. But then you couldn’t get access to the Flying Sub once that had been done so…

      Where’s my Ritalin?

    3. And nearly got destroyed every week (often by a squid) but it was designed to be quickly and easily repaired by CPO Sharkey. (And cannot believe I remembered his name)

      1. Here’s your reward for that great recall:

        James Doohan, who played Lawrence Tobin in Hot Line (1964) and Hail to the Chief (1964), was offered the part of Chief Sharkey, but turned it down, because that same week he was offered and accepted the role of Chief Engineer Montgomery Scott of the Starship Enterprise on Star Trek: The Original Series (1966), which became his definitive role. Terry Becker accepted the role of Sharkey.

        1. They must have had him lined up for a couple of years before filming started since Voyage was on in 1964 and ST started 1966.

          Still, one of those things. I think Doohan made the right decision.

          After all, he’s the only guy in a red shirt that survived.

      2. Doctor, I’m confused!

        The “Chief” character on Voyage to the Bottom of the Sea had the same name of “Sharkey” as the Don Rickles character on a much later TV series, the sitcom CPO Sharkey?

        The sitcom, of course, was a vehicle for stand-up comic Rickles brand of sarcastic put-down and insult comedy, which fits in with the popular stereotype of the senior enlisted man on a Navy vessel being the salt-water counterpart to an Army drill instructor.

        Only in the sitcom, CPO Sharkey’s edgy exterior hid a heart of gold and serious efforts to protect the men serving under him. “Chief”, as how I recall the character was addressed on Voyage, was played straight. His main interactions were with the bridge crew, warning them about the limits of the machinery much in the same style as Scotty on Star Trek.

        1. Now that you mention it, yes – they had the same names. Synchronicity, or something

        2. Yep. Not to be confused. Two distinct Chief Sharkeys.

          TV’s VttBotS’s Chief Sharkey: Chief Francis Ethelbert Sharkey, but you had better just call him “Chief”.

          Walter Pidgeon played the original Adm. Nelson in the premier movie. With a climate change plot, way ahead of its time for 1961.

  13. “For something like this submersible, a known quantity piece of COTS hardware with reliability reports and a bog standard interface is a much better choice than some custom designed physical control system or other custom fly-by-wire scheme with proprietary hardware. Why reinvent the thing?”

    Agree. It is like prejudice against using R/C model airplane servos in your own design autopilot for your homebuilt airplane. The things have had 60 + years of development and have been made and used in their millions and various grades of quality power etc are readily available.

    1. I would think technology is getting close to the point where some of those sophisticated sex bots could be upgraded to act as an autopilot.

  14. My question is: Do the signed legal waivers of future liability-related litigation, signed by each ‘participant’ before the activity, prevent their surviving family members from liability-related litigation in the future?

    I’m thinking the families are going to litigate the company out of business before the business will be able to pay the judgements.

  15. Corporations are known to grant their officers and shareholders indemnity from any legal consequences of the actions of the corporation.

    I’m wondering how bullet-proof “low-tier indemnification” can protect from higher-tier judicial interpretation of ‘Indemnification’

    1. At this point, what resources does the company have? A failed design for a submersible? Wreckage on the bottom of the sea?

      Likely the families could sue – but recover?

      1. It appears from photos on the interwebs the company owns a series of submersibles. What value they’d have at a liquidation auction is anybody’s guess.

        1. I’m sure people will be flocking to ride on submersible that were bought used from OceanGate at fire sale prices. ^_^

          This morning I’ve been reading papers on the compressive strength of unidirectional carbon fiber composites. The strength perpendicular to the fibers is only 50 or 60% the strength aligned with the fibers, but almost all the research is aimed at the aligned strength.

          The axial (perpendicular to fiber) stress would be about half the hoop stress, so I’m not making any bets on whether a catastrophic implosion of the pressure hull as a whole (as opposed to the viewport blowing in) would’ve been axial or radial.

          Epoxy also weakens and softens with temperature, raising another failure possibility if they used some part of the pressure hull as essentially a heat sink for some big motor controller, creating a weak spot that would be undergoing slow plastic deformation.

          The more I read about how composites fail in compression, the more negatively I view their use in this application.

          1. Is it possible to spin fibers along both axis in a dual-hull design? Just curious. But yeah, I get the feeling that with composites the failure modes might be more severe than what you might experience with steel. Haven’t done any research, just my gut instinct.

          2. All kinds of winding and weaving methods are used for various applications.

            Science Direct: A review of the structural factors which control compression in carbon fibres and their composites

            As you wade through that, it looks like very small factors can have a big influence on the compressive strength and failure mode. No matter how much quality control went into the manufacturing, I’d want to test the finished article very thoroughly in a pressure chamber to verify that one of the numerous small factors didn’t result in a much lower than anticipated compressive strength. And that still won’t fully account for creep over time.

            And there are some major differences from forming the structure out of steel. If you were rolling or hammering steel, some tiny surface irregularity won’t matter after subsequent surface machining operations. But with unidirectional composites, a piece of Scotch tape on the form would make two tiny bends in the fibers, creating bands where a kink band failure might show up (scroll down to figure 24).

            I would want a really fat design margin with such a structure, or the vast resources of Boeing or Lockheed Martin to make sure the hull met specifications. But from numerous sources, Rush was convinced he could work right at the edge, and even a bit over it, in what in comparison is pretty much a garage operation.

          3. Seems odd to me to assume Young’s modulus to be the same for tensile vs compressive stressors. I found the recoil method to be very interesting but wonder why a destructive cut was necessary vs just allowing compression via relaxation? In fact you could do all sorts of mechanically induced stressors that way including the possibility of inducing standing waves in the fibers depending on their length and the frequency of oscillation. Might give better figures for endurance.

            Speaking of which. What is the generally accepted practice for gauging when a fiber composite is near it’s ‘breaking point’ Microfractures at the surface? I don’t think you’d want to bore into the material to obtain samples unless it’s already considered end of life and destructive testing is ok. But that defeats the purpose of finding a gauge. Esp. for compressive stress. Can you take an xray and see the internal fiber structure? With steel it seems to much easier from electromagnetic methods to even photo elastic stress optics?

            I’m not sure how much of these measurement techniques were in practice at OceanGate (whatta name) if any. I’m getting strong Space Shuttle vibes here. “We’ve flown (dived) in all these extremes before all without catastrophic outcomes. There’s plenty of demonstrated margin. Why would this flight/dive be any different? Why needless testing?”

  16. “I’m struck by the similarity of the Titanic disaster itself, where the captain was repeatedly warned about ice ahead of his ship and yet he steamed at full speed into an ice field.”

    That’s an odd statement, coming from someone with such vast knowledge about Titanic. It was common practice – and White Star Line policy – to steam at full speed even with ice warnings, relying on lookouts to spot major icebergs in time to avoid them. Unfortunately, the lookouts on Titanic had left their binoculars in Southampton, and had to rely on unaided vision. But the ship would have survived a head-on collision with the iceberg. (It is sometimes said that Murdoch’s mistake was ordering the porting around maneuver, and reversing the engines. Recent analysis has shown that neither action influenced the ship’s course in time to change the way it hit the iceberg.)

Comments are closed.