16 thoughts on “The Oceangate Disaster”

  1. Rickover had every weld X-rayed. Rush just glued his boat together and crossed his fingers.

    Over the coming weeks or months we’ll find out quite a lot about the likely failure sequence, based on the recovered sections.

    I’m curious about the viewport, which was being used at around three times the rated depth. If it just popped out as a result of the implosion, and didn’t take damage in the accident, then they could put it under 5500 psi in a pressure chamber and see how many cycles it had left in it until failure. I’m betting that number wouldn’t be very large.

    1. Some videos I’ve seen of the wreckage recovery show the titanium bow cover intact. I couldn’t tell the condition of the view port or if it was even present.

      1. They ran a lifting strap through the hole where the viewport is supposed to be, so it’s definitely not attached to the bow hemisphere.

        1. Hm. Must’ve missed that. Don’t know how I managed to do that. Bad eyes? Distracted from the TV by the cat?

    2. I wonder if David Pogue of CBS News is feeling extremely lucky or extremely stupid these days?

  2. In sum, this is a tragedy of epic proportions.
    While true, this is also a tragedy born of intentional decisions.

    Rickover’s Rules should be required reading in every engineering school in the country.

    I particularly love Rickover’s Rule #1.

    Auto manufacturers need to re-learn Rule #6. Esp. when supply chains get interrupted and buy vs repair is no longer an option.

    The submarine services had a bit of a reputation for insubordination. (See the movie Das Boot). But when lives depend upon you speaking up, you speak up.

    1. The ocean deep, like outer space will kill you in an instant. Not out of malice, but because of what they are and what you did wrong.

  3. …but basing the nuclear-power industry on submarine designs has been a mess.

    Quite true. It’s a case of trying to scale up that which got there first rather than best.

  4. I’m not a structures guy but I don’t recall ever seeing it being used for compression only loads. We always use it in tension like pressure vessels or bending where displacements are small, which they always are because carbon fiber is so stiff.

    1. I’ve never heard of filament wound structures being used to resist compressive stress, and can’t imagine that it is of any use for such load conditions. Supposedly, Rush used it to reduce weight to save on the expense of handling equipment, and for its advantages in fabrication. But even for its use in the kind of pressure vessels we think about (higher pressure inside than out), filament-winding requires exacting fiber layup geometry, especially when factors of safety are low. One can’t simply put a mins sign in front of pressure difference, and expect a filament-wound vessel to behave in the same manner.

      Even with the body of experience and knowledge gained in rocketry over the years, we have been bitten by surprises in subtle aspects of filament winding. When I was at TRW, we were working with Thiokol to develop the Castor 120 solid rocket motor for our fanciful commercial launch vehicle. One morning I got a phone call from Thiokol’s chief test engineer telling me that they had had a Castor 120 fail in hydrostatic proof test. It was so far outside of my experience band that it didn’t register with me at first; I must have automatically thought he meant “burst test.” But no, it was a proof test, and risked the loss of millions of dollars of investment by both companies. It took us a huge amount of effort to kind of find out what the problem was. It was where the forward skirt was bonded to the case, and the manner of transferring shear loads from the skirt to the case. It was, as I mentioned, subtle, and not everyone agreed that it was the actual problem. But Thiokol’s fixes worked, whether they were right or not.

      When I saw that Rush had bonded metallic structures to composite structures, the latter being unsuited to compressive loads, I just shook my head.

      The only structure I would trust in a submersible would be one made of a continuous, rather than composite, material. For weight saving, it is justifiable to look at non-metals. Alumina and silica have far higher compressive strength than any metal, and are lighter. I would trust them only in a spherical shape, and then only after extensive testing.

      The marine environment isn’t a place for amateurs.

      1. In the previous thread on this topic I commented in response to a good link to the topic of testing by George Turner here:

        http://www.transterrestrial.com/2023/06/22/the-titanic-submersible/#comment-517231

        I’m wondering why the recoil test as described needed to be destructive? Could you learn as much if not more from simple relaxation? In fact wouldn’t an oscillation between the modes give you some kind of an idea of the lifetime of a fiber under compression? I admit I’m not a mechanical, structural or chemical engineer. But I am curious to know.

        There also seem to be a lot of pieces here. Fiber, weaving, composition, glue even. Not to mention a titanium, composite interface!

  5. I was watching a youtube video with an amateur researcher testing the strength of JB Weld (a 2 part epoxy). He glued up some square tubing sections of steel, and put it in a small hydraulic press.

    The glued joints ultimately failed – no real surprise there. But as he was pumping the press, the glued bonds were cracking – loudly.

    Which may be what the Titan victims heard as their last sound

    1. A thought I just had:

      In one of the studies of compressive failure, a common mode was interlaminar separation. Basically, a big column of fibers fails in buckling when the layers split apart, so instead of a uniform material it becomes a stack of separate sheets, like squeezing a ream of paper by the edges.

      The separation forces would be enhanced if there was a force pulling inwards on the inside of the sub, or a force pulling outwards on the outside of the sub, such that the layers weren’t in pure compression aligned with their length. If you apply a side force on the center of a building column under heavy load, the column is more likely to buckle in compression.

      So I would suggest that however strong the pressure hull alone was under pressure, hanging something from the inside of it (glued to the inner part of the hull) that pulls inward would weaken it, as would gluing something to the outside that pulled outward. And sure enough, the hull has landing skids on the side and emergency ballast slung underneath. I don’t know if those forces would have contributed significantly, but they may have contributed.

Comments are closed.