I’m mostly putting this up as a way to keep a record of my own own speculation on the sequence of events that happened with the Titan Submersible when it imploded. I expect the final report is going to be months, or even years away.
To be absolutely clear, I have no special knowledge, or even qualifications here, everything I have to say is purely speculative based on the bits and pieces that have come to be known in the public domain through news media.
At some 4000-5000 psi, there’s a ton of pressure on the hull of the Titan. While the pressure is evenly distributed, it only takes a very small flaw in any one part of the hull to allow ingress. The points we know of where that’s most likely on the Titan were:
The interfaces between the carbon fibre tube and the titanium rings that the end caps connect to.
Any points where power and control wiring harnesses enter the hull
Additionally, any flaw in the carbon fibre matrix resulting from repeated compression / decompression sequences could become a point of ingress for water into the craft.
First, I am going to go to the obvious conclusion that the ultimate point of failure was the carbon fibre tube. I don’t think this is particularly controversial.
I will assert that a flaw in the glue binding the titanium ring to the carbon fibre tube allowed one or more small streams of high pressure water to enter the carbon fibre matrix of the hull. This likely started several dives before the failure, and small amounts of water had already begun to disrupt the polymer structures that hold together the carbon fibre tube.
Over the course of the dive, the submersible began to take on more water, and by the time they realized they were in trouble, the effective mass of the craft had increased substantially with the addition of water which was weakening the carbon fibre matrix. Once that reached a critical point, probably when the “acoustic monitoring” of the hull started issuing alarms, it was too late. Any actions taken were “too little, too late”, as water had taken hold of the craft and was introducing the inevitable flaws that would result in the implosion.
All of the other criticisms of the design and implementation of Titan aside, once water started to make ingress into the hull, it was doomed. How could this have been mitigated? Some kind of scan of the hull after each dive might have helped identify the emerging failure. Are monitoring systems like the acoustic monitor that they dreamed up adequate? Clearly not. Current technologies on that front are clearly inadequate in terms of detecting problems early enough to be able to take corrective action.
Please remember, this isn’t definitive, and the final report could come along and prove my speculation here completely wrong. I’m writing this so I don’t lose track of my current thoughts, and so that we can see what the analysis shows when the professionals doing the analysis are finished with it.
[PS - July 14]: A couple of additional thoughts regarding the Titan submersible.
The more we see about the way this thing was constructed, and the risks and shortcuts taken in so many areas, the development of the Titan strikes me as a failure of fundamental engineering principles that stem from the Tacoma Narrows bridge collapse.
It was well known that carbon fibre vessels do not do well under compressive load, yet the developers of Titan chose to use that material without doing any kind of prototyping and destructive testing. (Yeah - costs money - I know - too bad)
Mixing different materials in a design intended for a high compression load environment has enormous risks associated with it. Again, destructive testing would have been more than justified - and feels like basic materials science stuff that should have been done.
"Acoustic Monitoring System" - sounds like a good idea, but I suspect it was nowhere near sensitive enough to provide anything more than a "you're fucked" alarm. Again, a lot of prototyping would be necessary to make this viable. I would also think that sensors embedded throughout the carbon fibre matrix to detect water ingress would be absolutely necessary. I'm not sure how feasible such a design is - I know the technology around the sensors, but I lack the knowledge of carbon fibre materials to know if such a sensor network would compromise the carbon fibre.
So many aspects of this craft feel like "quick and dirty" solutions to problems that ultimately are somewhere between inept and professional negligence. Yeah - sure you can buy a bunch of hardware off Amazon or Ali, but that doesn't make it fit for purpose ... and buying "time expired" carbon fibre from Boeing? Seriously? What the heck are you thinking?
2 comments:
There are many types of carbon fibre weave to make the cloth. The base material is much better in tension rather than compression, so better at containing pressure as a pipe/cylinder like the "Titan" than having external pressure trying to crush a hollow tube. Then, if you want to have a strong final product, you need to stick the whole thing in an autoclave to heat the whole assembly up to drive off gases between the weave and the impregnating epoxy resin. The alternative is a vacuum chamber, which isn't as good. Buying your own carbon-fibre weave cloth and treating it like a backyard fibreglass lay-up project is a recipe for failure.
There is a reason why the typical bathyscape pressure vessel is spherically-shaped to resist external pressure, not some cylinder with different materials glued on each end by diy amateur hour "engineers".
Also, most people would better understand what happened to Titan if the word "crushed" was used rather than "implosion" -- what's that when it's at home is most people's response. They have no idea what an implosion actually is.
Way back when I graduated as an engineer in '69, carbon fibre's big horny deal was Rolls Royce using it as material for the giant fan blades for their RB211 fanjet engine for the Lockheed 1011. The fan is what you see at the inlet of modern airliner engines. RR had their own special weave and glue, and the processed material was called Hyfil. Since fan blades are rotated, they operate in tension, not compression, so a good CF application. Then they tried the giant airgun bird test on the rotating blades, and they blew apart. Oops. Carbon fibre is not good at absorbing sudden shock,which breaks it. Formula 1 Race car front "wings" get sharded all the time, when cars hit each other in combat. Not bent like metal, but turned to shards.
I first used a carbon fibre chassis plate in my hobby of 1/8 scale RC race cars back around 1980.These weren't toys but 7 lb missiles with 2 horsepower engines and 60 mph top speeds, in about ooh, four seconds. You're looking at the 1988 Canadian All Wheel Drive champ here.The hobby devolved into miniature off-road buggies and died, while young people moved to stare at video screens rather than doing anything physical. Anyway, in the '70s, most of the RC race cars had aluminum chassis or fibre-glass like electronic circuit boards but four times thicker. I managed to drive my CF car into a building off the parking lot race course, Blew the CF chassis apart. Did the same with aluminum and it bent and could easily be straightened.
Anyway, I find Americans can be obtuse in their enquiries, and who really thinks the US Coast Guard knows its ass from a hole in the ground? Not me. I expect their report will be fuzzy wuzzy prevarication. Can't say CF is no good, why, the F-35 structure is made of it, and boy did they suffer delamination in wing load testing!
My point? Any backyard dumbbell can make a carbon fibre pipe in their shed. The chances they have a clue what they're doing is minimal. Cylinder versus sphere was mistake #1.
@Bill Malcolm - 2:48 AM:
Thank you for your input - you appropriately point out that Carbon Fibre has never been used in a compression context for good reasons.
I have limited experience with carbon fibre myself, but it’s well within my ability to comprehend - for example - the inherent problems with having mixed materials with different characteristics under pressure.
Post a Comment