Why OceanGate Titan Imploded - A Carbon Fiber Composites Perspective
This video is a bit of a break from the norm on this channel as I usually enjoy talking about nature, skiing etc but as someone who has worked with composite fibres like carbon fibre for over 10 years I was absolutely shocked at the engineering and assembly practices of the Oceangate Titan submersible. In this video I wanted to discuss a few of the things I have observed in the construction which may have lead to the catastrophic failure of the pressure hull and subsequent tragic loss of life.
This is of course all just my speculative opinion and is not fact. Much of my composites knowledge is self taught and empirical and I would not claim to be an expert in any way. I also do not do not wish to offend anyone who may have been affected by this event. Thoughts go out to the victims and families involved.
00:00 - Intro
00:37 - What failures may have caused the Implosion
01:05 - Design flaws in the carbon fibre pressure hull design
02:26 - Mechanisms of carbon fibre failure
03:01 - Audible cracks in the carbon fibre pressure hull
03:58 - What I think the pressure hull should have looked like
04:14 - Strength to buoyancy ratio?
04:40 - Glue bond between the carbon hull and Titanium interface cap
05:33 - Why you don't degrease components without gloves on
06:20 - Where is the clean room for the hull assembly?
06:49 - Was the glue mixed and applied correctly?
09:26 - No glue squeeze out from join when assembled?
09:40 - The bonding equipment I would have used
10:03 - How to bond dissimilar materials like Titanium and Carbon fibre
10:47 - What can we learn from the imploded Titan wreckage
12:16 - Composites could work but need way more testing
4:38 Hi, I’m a naval architect! “Strength-to-buoyancy ratio” is complete and utter nonsense. The buoyant force is only a function of the volume of displacement of the hull (provided that the hull isn’t flooded). A fully-immersed cylinder with a 2m outer diameter will have exactly the same buoyant whether it is made of carbon fiber or steel or titanium. If your requirement is a certain inner diameter, the extra thickness will get you some more buoyant force, but you could get the same extra buoyant force by simply making the metal hull larger to achieve the same outer diameter, with the added benefit of slightly more available volume inside the submersible. TLDR: Archimedes is right and Stockton Rush is wrong.
Fantastic debunking! Thank you sir!
Thank you. I thought I was being dumb there when he came out with that one.
@@JamieKunkaGlad to help! It’s relatively rare that my expertise is relative to something in the news, but when it is, I’m more than happy to help elevate the public discourse.
@@michaelimbesi2314 that is true , the buoyancy is proportional to the volume regardless of what materials the hull is made of. but i think what rush was referring to is that he wanted to achieve a certain buoyancy with the lightest materials possible to reduce the weight of the sub.
He was obviously referring to how using a lightweight material like carbon fiber reduces the overall weight of the sub vs using titanium. It shouldn't be so hard to read between the lines. You're just twisting his words to make them sound as bad as possible really.
I have been an aviation maintenance technician for 32 years, specializing in aircraft structural repair. I spent 10 years in The Air Force, 5 working on the B-2 Bomber and had a lot of composite repair training. I am not an engineer, just a repair technician. Watching the video of the Titan’s construction made me cringe. To me, they did nothing right and everything wrong. Watching them “clean” the Ti ring and bond it to the carbon made me gasp out loud. We had a saying, “Cleanliness is next to bondliness”.
Since you’re here: So does anything about the Titan’s construction make sense for the aerospace industry? Because I’ve wondered if Stockton was taking what he apparently knew as an aerospace engineer and just assumed that a vehicle for underwater works the same way as planes.
I too was appalled at whàt I saw of the layup! I've worked on aircraft,, boats, roçkets, atomic submarines since 1957! Worked with autoclaves And HYDROCLAVES. Also oñ filement winding machines. Yes, they were very naive iñ every respect, from the mixing of the àdesive to handling of materials! Did they use a vacuum chamber to remove the air from the adhesive? What was used to thicken the resin? Many questions!
Other than a superficial resemblance to aerospace structures, no. As stated in the video, filament winding in only one direction is a naive design. I've never done any safety critical bonding just hobby stuff, and you always do the correct prep for whatever it is you're working on. Step 1 no matter what process is wear gloves to keep your skin oils off the bonding interface. Step 2 is usually mechanical abrasion or chemical etching. Step 3 is keeping the temperature and humidity and sometimes pressure at the right level for the duration of the cure.
Every step of this is wrong and that's only the tip of the iceberg. Absolutely stunned that it didn't implode on its maiden voyage.
Hi Chris, Retired USAF AFRL/ACO 2016 here. As a SME I applaud you for composite repair skills. I started the Mobile Training Team at the request of HI. In 2008 and personally trained over a hundred airmen in the field. In my opinion there were both design flaws and grossly inadequate manufacturing processes. This puts an unneeded smear on composite structures. The joint was the problem! Adhesives are great in shear not so good in tension terrible in peel strength. Some of the factors stressing the joint: dissimilar material stiffness, dissimilar material expansion due to temperature changes, hemispheres flatten out under pressure causing the bonding surface to dish out and put the joint in tension on the outer circumference. I was the principle manufacturing engineer on filament wound cases for the space shuttle boosters Air Force program to do a polar orbit. Canceled because of the Challenger disaster. In testing using water pressure the reverse loading condition of the submersible the joint was a big hurdle. At near 1000 psi the half dome hemisphere inflated like a balloon and the composite tube barely moved radially. This caused the Clevis joint on the dome to open up and expose the interior end of the pin to the pressure. The 2000 pounds of force on the end of the pin was like pulling the trigger on a gun. Pins were spit out like a machine gun as the joint failed in microseconds. The AFT joint failed 6000 psi water pressure intruded compressing the air to less than a golf ball size and blowing out the forward porthole and the after shockwave blew the composite tube apart. Only epoxy in that direction no fibers along the axis.
6:17 Rush handing him the glue on a tray, while he's on the topmost step of a random stepladder is really a metaphor for the whole operation.
Even myself, a layperson with 0 years working with composite fibres and 2-part epoxy-style adhesives, was staring with disbelief as they slid the two pieces together, by hand, IN AN OPEN WAREHOUSE AND IN STREET CLOTHES. *I've seen woodworkers do more prep work before applying layers of topcoat to a piece of furniture* - wearing full-body suits in a sealed environment, and so on. And no one's life is on the line if the finish being applied to the furniture you're working on has a bad spot or three. A bad spot or three on a deep-sea submersible, on the other hand, means instant death. Quite literally. Just unbelievable. You can't pretend you know what you're doing when building something like this. LIVES ARE ON THE LINE, YOU DAMN WELL BETTER BE SEXTUPLE-CHECKING EVERY TINY DETAIL AND ASKING EVERY EXPERT YOU CAN GET YOUR HANDS ON TO OFFER THEIR BEST TIPS. Firing everyone who told him "no" and hiring nothing but kids fresh out of college (who could be browbeaten or intimidated into never telling him "no") probably didn't help much, either. Why hire industry experts? Those guys are *EXPENSIVE* and will tell you "no" to your face. College grads, on the other hand, tend to work for a fraction of the salary and tell you what you want to hear, 'cause you're the CEO! The hubris on display throughout this whole sad saga borders on insanity, like Stockton truly believed that the Universe would bend to his whims, that all his "innovative engineering" was infallible, and that he could defy the laws of physics through sheer grit and determination (while saving a bunch of money in the process).
You hit the nail on the head there, hubris and ego above lives and proper engineering practices. A very sad saga indeed.
@@JamieKunka I am also a layperson but with an extensive background in a range of technical disciplines . GLUES RESINS AND EPOXIES HAVE CAUSED MANY DEATHS if not mixed properly ...BOSTON MA .. BIG DITCH TUNNEL PANEL DROP is a typical example ..it killed the occupants of a minivan when it dropped because the EPOXY used to anchor the suspension bolts ran like molasses in JULY. Besides the obvious being addressed by the expert analyst ...RUSH was too full of own self generating BS , he believed in his own BS as most legends of one`s own mind do enough so as not to listen to anyone, but his own misguided mind.
The engineer and those workers have blood on their hands now. Did any of them ride down in the Titan death trap at anytime I wonder? Who was the inspector working on this submersible, he or they too have blood on their hands. So did SR but he died in the implosion but I'm not sad that he went down with his flimsy submersible. PH should have known better and sounded the alarm, but would SR have listened to him?
@@MissX905 Stockton seemed to take anyone questioning his "innovation" as a personal attack, so I seriously doubt it.
What's even worse is the shop that made this "vessel" and glued it the way they did. Granted they probably did it based on the spec provided but I work in that industry and when we see shit that is an obvious no no we say something. The shop that made the shit is liable in a way as much as ocean gate in this instance. But that's my opinion
Ironically, a Pringles can is actually better engineered than the titan sub was.
Because a Pringle’s can was engineered.
And like a Pringles can... once the Titan popped; it just couldn't be stopped...
Titan was just a giant Tylenol.
Rush's workshop is little better than mine (I am a musical instrument maker working at home). I take greater care in gluing and bonding than these guys do. Amazing. I will give him some credit: he went down with the ship.
so Rush got ripped off from the sounds of it
Rush may not even deserve much credit for "going down with...". OceanGate's former finance director said that Stockton Rush asked her to be the head pilot for the Titan after the former pilot, David Lochridge, previously said he was fired for noting concerns about the sub's safety. She quit because of that.... or maybe she had limited experience with Game Controllers? Rush seemed eager to keep the money coming in and appearances up. Would it be too much to speculate if the finance director knew other valuable information that was best left buried at the bottom of an Ocean?
Yes, I heard had they used titanium and steel for the body instead of carbon composite even with other problem, they could have averted the implosion.
@@kamilebrahimoff3589 could have but didn't
if you think about it, it was kinda like watching a turdlog do the disco circle before disappearing down the black morass
No BS. No sensationalism. Solid content. Thank you for sharing your experience on these materials.
One positive from this incident is the case study it gives to all future engineering students. Universities should come up with courses covering all the engineering concepts behind this failure.
I hope this incident gets dissected and every little lesson we can get out of it is learned. I hope that this tragedy is not wasted.
Yep, Oceangate University. -Don't do this 101. -Not like that 202. -No frigging way 301.
"Didn't see any glue squeeze out" I design loudspeakers and I would have fired the vendor for this - for a loudspeaker. Absolutely insane.
You would think that thing would be assembled in an airtight chamber and then put in an autoclave
That’s a fundamental tenet for applying adhesives. I’ve done everything from speaker reconing to carpentry. Insane.
@@thecloneguyz ..by workers wearing PSA, with a prezise construction frame to hold all parts in place by µm, measured and leveled with lasers. Direct control of the joint with Xray scanners after leaving the autoclave and and and...
I don’t think it’s an actual glue. It’s more like a filler putty as the pressure holds the body together no glue is needed that’s the way I see it especially as the fiber shrinks with compression. If it were a set bond type glue it definitely would crack and not adhere. I think of it more as a playdough type material that never cures . But what do I know it’s just a mindset. They had external collars that wrapped around to secure and seal at shallow depths as the depth increases the pressure holds everything secure. Until the hill starts cracking over and over of your expired super cheap deal carbon fiber. It was actually expired so he got it at a discount
“Don’t just stand there , get some glue!! Anyone?
Absolutely insane the carbon was only applied this way, anyone who ever 3D printed a tube knows that it breaks in the layers, this is no different, actually impressive it survived as many dives as it did.
I pictured the water pressure like an ax trying to chop a piece of wood, and how much easier it is to chop with the grain than against it. Also, since the fibers were all in the same orientation, a few layers buckling could cascade through the entire carbon fiber hull. I think the strength of the carbon fiber hull would easily have protected against implosion at least several times better if it had each layer going in a different orientation than the layer before it. Aside from that, I didn't like the fact the carbon fiber hull can't be bolted into. I would never trust diving under water in something bonded together by nothing but glue. When I first saw them gluing it together, my first thought was a massive "Why?!" It was kind of like witnessing an idiot in a horror film about to get themselves killed.
To be fair, Rush clearly states the fiber orientation is 0/90 (1:33 mark), meaning equal amount of fibers in the axial direction and the hoop direction. Still I agree that some +/- 45 degrees also would have made the hull stronger. Rush justified not having +/-45 by saying that there are no torsional moment, which is technically true. Problem is, this is only true as long as the hull is completely intact. As soon as there is a little damage to the hull. The +/- 45 degree fibers would have helped redistributing the loads around the defect. I am a structural engineer with carbon fiber experience.
@@davidosterberg how would they run a axial layer with a setup like that? Wouldn't the turned ends of the mandrel prevent them? I guess they removed the ends or there would have be axial layers at the top only???
@@AORD72 That is an excellent question, that I don't know the answer to.
@@AORD72look up ElectroImpact AFP. That's how the built that hull.
It should have been double prooffed. Meaning that if they were going to be operating at 6,000 feet, it should have been built for and tested at 12,000 ft.
That would require a crazy level of over-engineering.
correct it should’ve been common sense to have a Margin ,, if you’re towing a 7000 pound trailer, get a truck with a 9000 pound towing capacity, not a 7000 pound towing capacity.. I mean, this is nearly common sense in the engineering world, it’s astonishing. He was pushing it at the limit.
@@andychow5509100% more than expected is incredibly reasonable when human lives are concerned.
You're soo off I'm guessing your completely uneducated in the topic and just throwing numbers out that sound good. Most submersibles do 120% their operating depth, your talking about 200%. Titan didn't even have the extra 20% either, and since the carbon fiber tore on the very first dive it definitely wasn't even rated for Titanic's depths let alone had a safety margin.
@@sertank735 that's what an uneducated person would say, since industry standards are 120% and there's never been a problem till now. Titan didn't have that safety margin, it wasn't built right to handle the depths it was going to, the carbon fiber tore on the first dive, and yet you think its gonna somehow do double its operating depth? Geez, your so ignorant about submersibles idk why your even acting like you know what should be done and what is a good level of safety. 120% is actually more than enough, but your sub has to not only be built right but be able to handle 100% of it's operating depth without it falling apart every dive. Carbon fiber isn't a good submersible material period, nor should we take this accident and try to change industry standards when this sub had nothing to do with industry standards, and actually went so far off industry standards it imploded. The problem started at the design phase not at depth, the sub was just so bad it couldn't even operate 20-40% it's operating depth without the carbon fibers tearing. Every other submersible does 120% and has never had a single problem operating at 100%, because they didn't use carbon fiber and their design was actually good and safe. Just get educated if you wanna act like you know what your talking about, because clearly you don't.
I've been wondering about the titanium/carbon fibre interface as well. The bonding procedure taking place in an open industrial workshop and not in a cleanroom. The lack of proper safety equipment; scaffolds etc. No one suited up for clean operations, no vacuum chamber to pull air from the glue joint. The mixing of the glue in an open environment. It all seems quite amateur and hasty.
When I saw the clips of the assembly, I was thinking "Surely they had a vacuum chamber somewhere. Why didn't they mention it?". I don't know that much about epoxy, but I know that anyone doing serious work with it has to find a way to get rid of the inevitable air bubbles as part of the cure schedule. That work looked like me patching the hull of my sailboat.
Or my attempts at fiberglass objects!
True, the biggest problem of all was the carbon composite, which could not handle the water pressure at that depth, so it imploded.
There was like 40 megapascals of force pushing that titanium cap against the carbon fiber. That adhesive was just gap filling once that thing started down. The issue isn't going to be some dust, it is going to be whether the interface at the edges can mechanically tolerate the cycles.
@@kamilebrahimoff3589the thing is - it did handle that depth, but it could not repeatedly handle that depth. The egomaniacal fool at the wheel (Stockton) confused the two as well.
Doomed from the get go ... no Clean Room, no Auto-Clave and no AS9102 certification to keep it all on the straight and narrow. But hey, that's just what my 40 years in Aerospace Composite Engineering tells me. Just imagine if Rush had invested in a "Thread laying machine" (TOWS) like we use on our F-35 Program ... much stronger ... and Rush wouldn't have been able to buy that expired (and obviously "re-tested" once) CF Tape from Boeing ! And THAT folks is why we call it "HALT" ... highly accelerated lifetime testing ... "bend it until it breaks" so you have a baseline. IMHO, the Hull imploded and had NOTHING to do with the Titanium flanges/dome; the acrylic view port and domes blew off like Champagne corks (Water cannot be compressed and it had to go somewhere) ... studs smooth as a shiny new whistle ~ zero threads left nor CF tear away on the mounting flanges (there are ZERO “Adhesives” on the professional/FMS market that don’t call for application of a “Primer” prior to the actual Adhesive). His "Design Engineers" sucked ... the hull was in "hoop" yet zero internal stringer supports nor sleeve to help keep it equalized; hull thickness would have needed to be approximately 17.6 inches thick without said support structure. All the "popping noise" on previous dives was the CF delaminating, transitioning from "Elastic" to "Plastic" state ... the fibers ripped out of the resin matrix bed and never able to re-seat; Rush would have known this had he bothered to perform NDI (ultrasonic or X-Ray) after each dive. Pity his hubris got in the way. Good presentation Jamie.
Agree. Rush fired his design engineer after he raised safety issues. Rush was a narcissicist. When I listen to Rush speak he is an obvious noob.
Yes, you’d think ultrasonic testing would have revealed the defects before the vessel was submerged. And also the ruptures that developed under pressure.
@@maxenielsen yes, it would have identified the delamination areas ... "X-Ray" for more accurate locations ... but he was such a cheapskate he skipped it.
Excellent video. However, the biggest problem is that carbon fiber has a much lower modulus of elasticity than Ti. The tube wanted to shrink by 1/2" at depth, but the Ti wouldn't let it, resulting in an ENORMOUS stress concentration on the inside of the tube at the junction of the Ti, resulting in buckling.
...or the adhesive bond failed prematurely.
im a former professional glass fibre pipe system builder (power plants etc, 7-8years) and I absolutely agree with the issues you point out. TBH it looked that they had very little clue about what they were doing and was just winging it, hoping for the best. Also I think the different thermal expansion coefficient of the materials used could play a role.
No glue squeezing out of the end caps when they laid it up definitely threw a red flag for me too the first time i saw that video. Very good analysis vid 👍
Same with me, I have no idea about this technology, but seeing they place the end cap on top of the carbon fibre and no glue coming out made me nervous
I try for some glue squeeze-out with woodworking projects, but I’m also fond of using screws In addition to glue. *lots* of screws.
@@dennisyoung4631 - you should have built that submarine of course
@@dennisyoung4631 I'm work in a cabinet shop and another thing I notice was that they didn't even have to tap it in place it just slid right down on it
@@Blasko86 got a question for you, then. When I am putting pieces of 1/2” nominal birch plywood in slots/grooves/Rabbets/?, I commonly try for a “drop-in” fit, one where there’s (hopefully!) no needing to bang things with a mallet. (I’m working alone, still relatively new to woodworking, and am older, chronically ill, *and* disabled. Most of the shelving units, bookcases, etc. are more *disability accommodations* than conventional furniture.) I use hide glue, because it has a long open time, is relatively easy to clean up, and doesn’t leave visible stains under the finish. I then put in #6 countersunk screws. Am I full of for doing it that way? Unlike metalworking, I haven’t done it for a living, nor did I take college-level machining classes. I did my share of reading, and I have asked people about it who have more experience than I do.
The mating of the titanium to the Carbon Fibre is something I've had thoughts on. Years back I used to install strain gauges in the aviation industry - mostly aluminium alloys and titanium, but also on carbon fibre. The process for metal and CF involved using different solvents to degrease the parts, and keying the surface was also a little different. The process I was taught required titanium to be warmed with infrared lamps to drive out surface and subsurface contamination like oil, during which solvents were used until the surface was clean, and various fine grade oxide papers were used to sand the surface in figure eight motion to key the surface to bonding the gauge to...which then had a light acid applied and neutralised...the process with CF was similar, but less intensive, without heat lamps. The main difference was the type of glue applied for bonding. We used a special proprietary bonding agent manufactured by strain gauge manufacturer. The CF used a different bonding agent based on a cyanoacrylate type glue. My understanding back then was that the same glue/bonding agent could not be used for titanium and CF, so was surprised to see a glue being used to mate both surfaces. Also surprised the surfaces didn't seem to be keyed - especially the titanium - ready to receive the glue! Another concern was how shallow the join was!
I've always called the deliberate roughening of a surface to be bonded giving it "tooth" - similar to how the term is used to describe how paper will hold graphite from a pencil when drawing. Absent that increased surface area, and scrupulous cleanliness, you have almost zero chance of forming any effective bond to a metal surface - and, the adhesive which you choose to apply must be to some degree corrosive to the metal in order to achieve a chemical bond. Further, you join needs to ave significant resistance to slippage, much like mating screw threads have a force resistance projecting perpendicular to the direction of tear-out. And, these slip-joints had absolutely nothing in the way of resisting elements to hold against tear-out forces. Just a bad job, physically and chemically, even before you look at the actually assembly "procedure"...
I dont know anything about these materials how do u bond carbon fibre to titanium (if not with glue)
Great to hear the interesting details from your real world bonding experience. This just goes to show the length and precision we must go to do make such bonds.
Meh why not just get 4 guys in t-shirts to slap some bondo on it, squeeze it together, and call it a day?
@@inthefade :D
The force on the endcap bond is closer to 28000 psi. ( the full surface area of the end cap applied to the surface of the carbon fiber tube) Critical epoxy mixtures are mixed by weight, not volume. Air was certainly trapped in the ring when the joint was made, not enough pressure to squeeze out the air bubbles. Also, an autoclave bond is required for high strength composite layups. I am surprised this sub lasted more than one dive.
Very interesting to hear your calculations on the pressure. I’m uncertain if any pressure was applied to the interface cap when it was bonded or do you think they were just relying on the weight of the machined part pressing down on the hill?
@@JamieKunka From looking at the sloppy ness of the bond, I do not think they applied extra weight at assembly
@@JamieKunka The surface area of the Carbon fiber tube is approx 1300 in sq. 5 psi of bonding pressure needs 6,500 lbs of weight.
@@glennnovember3266 Using rough numbers for the dimensions I got similar results: -139MPa (-20ksi) for longitudinal stress. What is interesting is how they must have thought they had a margin of safety with the 0/90 layup since the resin alone was probably rated at -250MPa. I am also curious of the hygrothermal stresses of curing a 5" thick wall. This guy was definitely way off of charted territory on this design..
Rush built this craft with inadequate funding. A bunch of guys building something in their garage. Sad.
This is by far the best video I've seen on the topic. All of the points you make are right on, and there are a couple you didn't mention. I've 10 years experience related to the construction, testing and deployment of aluminum cylindrical pressure vessels rated for full ocean depth, and 6+ years winding thick carbon fiber cylinders. Regarding implosion testing, the weakest side of the aluminum cylinder begins to buckle, putting the interior surface of the aluminum under tension, and any resulting fracture begins there. Carbon fiber cylinders are strong under tension, that is, high pressure from the inside surface, but there is no lengthwise compression strength with carbon fiber. More to the point, when under extreme internal pressure, the carbon fiber cylinder actually grows physically in the radial direction. If under external pressure as with the Titan, the strength in the radial direction is unknown to me, but if the composite is pressed inward by the uniformly-exerted pressure, it is likely to shrink slightly. If this rate of dimensional change is even slightly different than the titanium end rings, this would produce shear forces at the glue joint during a dive. Each dive could stress and weaken the glue joint more and more. The only way I know to effectively bond titanium to carbon fiber is by a difficult process that involves more than just glue, and was not mentioned. Lastly, for carbon fiber with a high number of layers, as shown, many things, especially the fact that the epoxy shrinks while curing, often results in voids and tangential cracks internal to the cylinder which cannot be seen, but are revealed when a test section is cut from the cylinder and inspected. My supervisor has told me that metals expand with temperature, while carbon fiber shrinks with temperature. I don't know if that's true, but if it is, it would also be a problem. Thank you for your excellent video.
I’ve taken a submarine design course (and used Paramarine for a project) and never heard of “strength-to-buoyancy ratio”. There is a factor called “weight limited designs” which the structure material plays into; it’s why HY80 has been superseded by HY100 and HY130 in newer classes of US subs: these are stronger materials and can thus be thinner and therefore less weight to the hull. But buoyancy has a lot of other factors not just the hull material. On the titanium cleaning clip, good catch! But based on where the rigging is laid out, it’s unclear if this would be the bonding surface or just the external surface. Either way, the controls in the entire process are COMPLETELY negligent at BEST.
▪ _"A former OceanGate finance director said Stockton Rush asked her to be the head pilot for the Titan."_ ▪ _"She told The New Yorker_ (Magazine) _that she quit when the OceanGate CEO urged her to replace the former pilot."_ ▪ _"The pilot, David Lochridge, previously said he was fired for noting concerns about the sub's safety."_
"Fine I'll do it myself".
I just found your video, excellent job, my first alarm bell was the way the CF was wound one layer on top of the last with no overlap or weave pattern, you don’t need to be an engineer to see all the flaws of this design
For generations to come, videos like these will train future engineers how NOT to design a submersible. Good job Sir, in training us laymen on complex engineering subjects.
MOST if not all (in that small community) already KNOW and knew this submersible’s design was a tragic accident (time bomb) waiting to happened
I hadn't thought about that! This wreck is the 'How NOT to build a submersible' like the Forestall (or 'Forestfire' as my naval friend says), is the standard video on how NOT to fight an aircraft carrier fire.
Unfortunately these findings are nothing new, similar things have repeated in the history over and over again. What we learned from this: RTFM with materials&glues etc and let certification people do their work and go through everything or work with them to establish best practices to ensure safety.
@@timop6340 🤡 statement
@@nadagabri5783yes, I'm wondering if someone could of sent a lawyer or a letter to Everett Washington explaining the dangers and get the occupational license revoked so they couldn't operate. This could of caused unwanted press coverage for oceangate and exposed them before a tragedy could occur..
It’s stunning this deathtrap made it more than one dive.
I hear Rush patented his “real time hull health monitor” system. He also said in previous dives the cracking was loud and persistent. Since we know that cracking sound was the sound of a hull progressively weakening, his monitoring system was obviously another useless idea.
the loud cracking sounds was cleverly fixed by asking passengers to listen to music
@@DelphiAmnestieddamn I didn't see that way. That's insidious of him
No, not useless. Just useless in the context that it was employed in.
The more time that passes the more I hate that guy. This is all his fault.
I’ve have no expertise in these things but it seems like cracking sounds equals damage. Damage that can’t be properly repaired or even evaluated. To continue to use something you know is damaged for a mission that dangerous is insane.
If there’s one silver lining we get out of this, it’s that hopefully many composites professors and industry SME will use this as a warning to future engineering students. Similar to how Tacoma Narrows Bridge is well-known and taught in the engineering community.
Actually there is already a lot of research and examples, in fact it's on every label of composite adhesive products
@@scifi_shop There's research and there's stories like this one that are good examples to convey that research at each step of the way.
Don't hold your breath. It isn't like engineers didn't already know ALL of this before the sub imploded. Some of them even tried to warn them. Still, they did it anyway and killed 5 people.
@@stargazer7644 We all do because we have experience. I'm talking about professors teaching new engineering students. I see how my OG comment didn't make that clear. Edited. It's a big event people know of that they can use as a starting point to a set of lectures.
I don't think this is really comparable with Tacoma Narrows. Tacoma Narrows was, as I understand it, a bridge that failed because we didn't know you need to consider XYZ when building a bridge (specifically, the possibility of aerodynamic flutter causing resonance). Its failure taught us lessons that could be incorporated in future suspension bridge design. Titan's failure seems to have been caused by ignoring things that we already understand well. There's no lesson to be learned, because we already knew not to do all that stuff. Unless your point is that now there's a practical demonstration of what happens if you ignore these things.
I'm skeptical about the emphasis on the glue, although it is certainly indicative of an overall sloppy attitude to quality control (and health and safety). The external pressure will tend to push the end caps harder onto the carbon fibre tube the deeper the dive, closing any voids without putting any sheer load on the glue. Since the pressure increases progressively during the dive any water leaks through the glue interface should have been apparent before they became dangerous. The carbon fibre tube may well have compressed more than the titanium ring, but the latter seems to have an internal channel the inside edge of which would have broken off if this were the root cause of the failure. My money is still on a simple and entirely predictable fatigue failure of the carbon fibre tube.
John, That’s the way I see it too. Extreme compressive cycles on a material designed for strength in tensile loads.
The hoop stress means this carbon tube is gonna shrink in diameter. If the titanium shrinks at a different rate, massive shear stress. Temperature changes can result in insane shear stress. I did a simulation on a pool window of poor design: 300,000 lbs due to thermal stress alone
@@ZebraLemur However, as I said in the OP, it is not a simple butt joint. The carbon tube is glued into a channel in the titanium ring. Under the strain you describe, the inner lip of this channel would be stressed and would have to break off for there to be significant movement of the glue joint. The titanium parts seem to be completely undamaged in the images we have seen. Conceivable that localised damage was deliberately or accidentally hidden as the parts were lifted ashore, but it will still be known to the investigators and should resolve the matter one way or the other. The fact that the carbon tube has not been recovered, or even significant intact pieces of it, also argues for this as the original failure.
Yeah but you're not thinking about the fact that the carbon-fibre compresses at a different rate than the titanium and what would that do to the mating surfaces?
@@thecloneguyz mating surface - nothing good. Just saw another video on this sub where a professor from Europe runs modeling software on this vessel under pressure. That model of the failure looked like the Pringles can being squeezed, everything was over in just a handful of milliseconds. The highest stress was at the Ti-CF joint. After the joint failed and the cylinder shattered the two titanium end caps were driven inward at high speed and ‘clapped together’ in the center where the cylinder used to be. After seeing this I have now come around to seeing the Ti-CF joint as the biggest problem. No matter what the adhesive or procedure used, this joint is the location of the highest concentration of stresses. Since there were several successful dives to this depth before this incident, material fatigue cycles must have played a role. Perhaps the glue joint finally failed, or the carbon fiber itself became weaker and deflected more on this dive, or both, or something else (he also identified the acrylic window as a weak point). The titanium hemispheres that were recovered looked brand new from the brief unloading video.
Recent graduate in Naval Architecture & Marine engineering here - take everything I have to say with a grain of salt because again, just graduated and by far not an expert! But my 2 cents on the "strength to buoyancy" term you mentioned. My first reaction is this is the first time I've ever heard this term utilized, I also have not spent much time with submersibles as they are definitely a niche in the industry. However my next reaction is that term seems a tad bit silly since the buoyant force must be equivalent to the weight of the object (Archimedes Principle), thus in reality strength to buoyancy is practically the same thing as saying strength to weight. Strength to weight is also not really a metric used all too often in Navarc, when designing commercial ships (think tankers, container ships, dry bulk etc) weight is certainly a concern from a cost and strength perspective, but it's often not the case that in commercial design you compare the strength to weight ratios of various materials, it's just not as much of a concern since the body of water will just displace more volume to account for a heavier ship - and often these commercial ships are volume limited. In other words, commercial ships often reach maximum capacity of their cargo holds long before they're in any dangerous territory of displacing too much water. Weight becomes much more a concern when talking about high performance yachts such as the Americas Cup sailing vessels. But the usage of composites in these craft is vastly different that in a submersible... The question that still troubles me is that someone must have realized that Carbon Fiber does not have the same strength in compression that it does in tension? I wonder if the strength to buoyancy ratio he was using accounted for the different compressive strength of carbon fiber, since pressure vessels under seas have to deal with compressive loading (greater ambient pressure) while airplanes often deal with tensile loading (lower ambient pressure).
Thanks for the comment and information and great to hear from a Navarch graduate!
"Squeezing a can of Pringles" Exactly the analogy I've been using to explain to friends that the Titanium rings overlap with the CFRP was way too short.
Yes - but a longer overlap poorly bonded would probably still failed. But agree - so short!! And the 90/0 fibre winding - insane!!
I don't agree completely with that analogy, because it implies the initial failure had to be radial. However, don't forget that in the sub, there is also an extreme amount of pressure on the end caps pushing in, thus reducing the need for much ring-tube overlap, and I suspect the initial failure was the tube itself in axial direction, not radial. And the reason for that, is the 1:3 scale model Oceangate tested to failure. You can see a picture of it at 3:46 The tube is still round, it's the end that is pushed in, delaminated/shattered and has separated from the metal end rings and caps (not in the picture). So a better analogy could possible be that you take a Pringles can and stand or jump on the lid so the tube buckles/shatters, and not the other way around. And if that was the cause, it would probably not have made any difference if the ring-tube overlap was ten times longer.
@@Kowalski301 do you know how they tested the 1/3 model? Cycle testing or just loaded up until failure?
@@AORD72 We have to assume they did some cycling before full pressure to collect more data, especially since they made a big deal about the so called "real time monitoring" system that would detect abnormal movements and acoustics in the hull, that would change from one dive to the next dive, and that the scale model had fitted some of these sensors, but I can't say that for certain. Stockton only talk about the scale model test very briefly and doesn't specify it any more other than they took it to failure and the shockwave blew out all the pressure sensors in the chamber. They did the tests at the University of Washington so there should have been at least some smart people in the area...
@@acebacker1 Love all you instant engineers.
I think the problem was the difference in stiffness between the carbon fiber and the titanium end caps. Under the pressure¸ the CF cylinder would have compressed in the radial direction much more than the titanium ring. This would lead to radial shear and bending stresses in the glue joint.
Exactly what I was thinking.
There were just so many pathways to failure, about half a dozen that seemed inevitable. Various things to go wrong in the bond, same in the layup. Rush really had a psychological blind spot in his intolerance for criticism.
As a materials engineer, the difference in modulus of elasticity between the titanium and composite seems like a likely cause of failure here. Computer modeling was likely used to calculate the stresses involved, but there can be many reasons for adhesive bonds to not meet the properties listed by the manufacturer. Consequently, extensive testing of the bond should have been done before risking human life in the vessel.
I was thinking this too. Seems like you would want the more elastic material on the outside so pressure increases the bond rather than pulling it apart when it's inside the titanium.
This is a definite possibility, and is what my first thought was on hearing that the sub had titanium end caps and a carbon fiber pressure hull.
So fascinating. It’s very uncomfortable watching how DYI this build was! 😮
Not a fascinating as all the INSTANT Marine engineers that appeared.
@@blackhd92 right? incredible stuff
I am a naval architect and marine engineer. I spent over 33 years working for major oil companies. All of my work was in design, engineering, maintenance, environmental issues, and spill response. I spent a considerable amount of time working on fatigue cracking particularly in critical areas where cracking could result in catastrophic failures. I have no experience with carbon fiber. I did at one point own a fiberglass sailboat. When I first heard the reports of the missing submersible and heard that it was a carbon fiber cylinder with titanium end caps, I told my wife that there was no hope of rescue. When the videos of the hull construction started to be shown, I told her that the design was flawed and was destined to fail. I agree with your comments on using a single radial wrap was a critical error. There should have been a weave pattern to strengthen the cylinder to handle the axial compression from the end caps. Even my sailboat's hull was laid up with crossing wraps of fiberglass. The Titan was subjected to multiple dives where the carbon fiber tube was subject to fatigue from the compression and relaxing of the axial load of diving and resurfacing. The arrangement of the rings and the bells leaves much to be desired. Securing the rings to the carbon fiber with glue alone, with no mechanical backup raises questions. The use of 18 bolts to secure the bells seems to be odd since a pipe flange at 6000 psi requires 44 bolts. I heard in several reports that the hatch end bell only used 17 bolts because of difficulty in tightening the 18th bolt. The lack of Classification Societies' input was a fatal flaw. Let's assume that review and approval have a zero-time and zero dollars impact. Skipping Classification meant there was no third-party review of the design, there was no fatigue analysis, there were no failure modes and effects analysis, there were no inspections and certifications of components during construction, and there were no in-service inspections. I told my wife if I had a disposable income sufficient to waste $250,000 to visit the Titanic to look out a porthole, please excuse the expression, "I wouldn't be caught dead on the Titan."
The man was an amateur, how did he get in charge of all this... the vessel at the absolute minimum should have been tested to 8k psi double the normal if I were in charge.. and I'm amateur..
When you say "pipe flange at 6000 psi" I assume you mean 6000 psi on the internal surface of the pipe? That's quite a different load on the bolts to 6000 psi of external pressure like the sub would experience.
It’s a miracle it survived the dives it did
Kudos for the slide discussing the necessity for surface preparation of the titanium for adhesive bonding - that's better detail than I've seen in any of the other discussion to date.
Just shocked that this was glued together. I'm surprised it sustained one dive without popping.
Richard Stockton Rush had poor eyesight - he was also blind to Titan’s fatal safety limitations.
It feels like Stockton was just guessing. Had some interesting ideas, implemented them haphazardly, cutting corners, didn't test them in any way, ignored warnings, and just YOLOd it. It wouldn't have been so bad if he was just risking his own life, but he also convinced 4 others and took them down with him.
People are allowed to think before going with him. He wasn’t forcing anyone. Like getting in a fast car with someone who says it will all be fine but there’s always a risk. And even someone without knowledge in engineering could see this was made to fail
@@PT-mj3bktrue. Problem is he probably didn’t properly convey the actual risks. When I went skydiving, I also had to sign a bunch of waivers and consent forms. The company still had a responsibility to provide a safe experience - if the instructors were inexperienced, planes and equipment poorly maintained, corners were cut, etc., I would have had no idea. I simply knew their track record was good (no deaths in over a decade) and trusted the company. Oceangate depended on these paying customers and Rush made all sorts of misleading claims about the safety of the vehicle and went out of his way to convince clients (e.g flying to meet with them) - it’s impossible as a customer to accurately assess the risk if the company is dishonest.
@@sue8412 Yes, and to add to that, Stockton claimed they collaborated with NASA, Boeing, and U Washington, but that was a lie. Stockton was also warned by his own employee as well as external experts about his fault design, but he ignored them.
@@PT-mj3bk A better fast car parallel would be if the driver failed to tell you the car was improperly maintained and that it had had some issues on previous rides that were brushed aside. Also, the car is not actually road safe, but the driver failed to mention it. Also, in case of a crash, the airbags may or may not deploy, but the driver also forgot to mention this and other details that are necessary to make a fully informed decision.
You can not tell a Narcissist anything, they will do the opposite.
There is more cleanliness and care taken in boat fiberglass work and auto body painting than what is shown in Titan construction.
Right? Painting a car: every surface is prepped and meticulously wiped down to get rid of every spec of dust. Paint is sprayed on in a paint booth that has an air filtration system by techs in full body suits. And then there's ocean gate mixing up batches of glue by hand and applying it the same way in what looks like a dirty industrial facility.
You'd think a deep sea sub with novel hull material ought to be legally obliged to go through the necessary tests and certifications.
If not in international waters, they would be. That's how they got around the certification.
I’m okay with people taking their own dumb risks for exploration or sport. We don’t make wing suit athletes certify their wings. But OceanGate better have been very very thorough and honest with every design detail in their marketing etc before they had those people sign wavers. If you’re up front and honest about the risks, I think people should have the right to do stupid things like this to an extent. I feel bad for the kid though. What an irresponsible father.
If it carried passengers it must be ":classed". hence, they were crew, supernumery.
Exactly, an agency should have done a final inspection to clear it for the dive.
It didn’t have to because the submersible only operated in international waters and was carried through territorial waters as cargo on the deck of a ship. By operating only on the high seas and not being registered in any country, it was literally outside the law. Rush deliberately chose to operate it only in international waters because doing so meant that he was literally outside the jurisdiction of any country.
A quick search and we find that the 'circumferential' stresses are double that of axial stresses in a 'thin walled cylinder'. But I've also found that so-called 'transverse' strength of carbon fiber is a lot less than when applying along the length of the fiber. So even if there are no 'torsional' stresses, certainly there are axial loads, and being applied perpendicular to the fibers. That seems to be a BIG problem (IMHO).
Even if there weren't any air bubbles entrained in the adhesive during mixing and application, there would be microscopic gas bubbles formed by the leaving groups coming off of the monomers as they polymerized to form the join. Normal procedures to remove these bubbles include curing at elevated temperature under vacuum in order to draw the gases out of the adhesive.
Bonding area is pretty important and a tapered sleeve on outside with 3-4 x the overlap would have possibly preserved the bond. However he mentions not seeing any bonding agent squeeze out....When we bond and pot things with thickened epoxy a real problem is when you slip one thing over another or inside another, you scrape the glue off the sides and push it into the end which also traps air. The ends aren't even touching 100% . You go back later for a repair and cut through to inspect and that is what we see.
Most deep sea vehicles are made as a sphere whereas the Titan was a cylinder but I have not read much about the difference of how a cylinder would act under extreme pressure compared to a sphere, the shape of the Titan may be part of the problem as well as the materials and construction methods.
CRP is still a sort of mystery material today. I've been in aviation manufacturing and we don't know exactly what it's hysterysis/longevity really is. And that's after 40 years.
Note: The guy wiping the ring down with a shop rag appears to be removing stray dust from the opposite side of the parts from the actual bonding groove. This is also a necessary step, as stray dust can migrate as the part gets craned up and over the CF cylender... hoever, he's not wearing a "bunny suit" and thus can leave skin cells and hairs behind after wiping down. Not a good situation on the non-bonding surface, but definitely a poor practice indeed on any surface where you want to achieve a bond.
Totally agree with you ...
I bet he’s feeling pretty shitty. Yeah, wipe with cloth with one hand and then rub over it with his other bare hand. Yikes!
Excellent point about fibre orientation. Single direction creates perfect path for cracks progression. I guess this is the key design flaw. Picture in 2:45 is showing perfectly what probably happened.... 👍
This was a really thorough, informed and thoughtful analysis of the various problems with the design and manufacturing of the submersible. I'm not an Engineer, rather I'm a physicist, and some of these issues seem disappointingly trivial, even to someone with my lack of relevant knowledge and experience. Not correctly texturing the surface of the carbon composite and/or the titanium dome before bonding the two... From Year 7 D&T I can remember being told to rough up surfaces a little before gluing things together, give the glue something to grab on to.
I still haven't hear anyone explain how carbon fiber would be desirable to hold OUT pressure. It seems perfect for holding IN pressure as in an air tank but it seems (to me) that the crush resistance could only be a tiny fraction of its expansion resistance.
When I worked with a HYDROCLAVE our carbon fiber seemed to shrink and turn into a softer like graphite material! Unlike auto claves it could obtain pressures around 3,000 p. S. I . . Carbon takes on many different forms! OceonGate probably never considered using a HYDROClAVE because they are very expensive. They seemed to operate more like boat shop!
I've been in composites for 30 years I've made helicopter parts blades nhra drag cars I agree the 4 in wound lay up was a mistake I also seen no mechanical adhesion sand roughing up of titanium nothing but I think the lay up was the big mistake it's weak that way especially under compression
Intuitively, carbon fibre going only one direction seems like a terrible terrible idea. I realise the primary strength will be going around, but that leaves no room for error if any part is weakened a bit.
Without fibres thing the other way, the point where the titanium ends and the fibres have to stand alone would be under enormous stress.
The primary strength will be going around IN TENSION. Carbon fibers are like wet noodles, they have no strength in compression. The only strength here is in the compressive strength of the resin.
@@stargazer7644 wrong, they're not like a "wet noodle", the compressive strength can be up to 50% of their tensile strength. Also, the epoxy holding them together has great compressive strength.
@@xpusostomos Carbon fibers are THREAD. You can hang your car from a single 1/4 inch carbon fiber rope in tension with well over a 2 times safety margin. So you're telling me you can balance that same car upside down on the end of that same 1/4 inch rope in compression? Have you ever actually handled carbon fiber fabric? Carbon fiber alone has almost no compression strength. In a composite the resin is just about the ONLY strength carbon fiber has in compression. Using carbon fiber to make submarine hulls is like using concrete to make tubes for pressure tanks.
@@stargazer7644 There is much you don't understand. Firstly, the sub wasn't made of carbon fibre cloth, but strands of fibre, all going the one direction. Secondly, to make a tube lose structural integrity by pressing on it the inside of the tube would be under tension while the outside would be under compression. Thirdly, the compressive strength is measured when the fibres are in alignment with the force applied. In a rope under tension, they are kept in alignment by the tension. Under compression you've got a problem keeping those fibres perpendicular to the force. But the sub is not a rope trying to hold up a car. Fouthly, carbon fibre has about 4x the compressive strength of steel. So let's say you wanted to support a car with some material by compression. What's a good material? Most people would say steel, that's what every car lift is made of. That's what every set of car ramps is made of. And yet carbon fibre is 4x better. Fifthly, don't listen to me, use Google and find out for yourself.
You've pointed out some of the questions I had about the carbon fiber section. The two part epoxy applied with a mixing nozzle, reorienting the fiber direction in subsequent layers, and proper surface prep should have been no-brainers. I have yet to hear anything about how the carbon fiber section was cured. Aircraft industry standards for curing composites is to bag the part (on the mandrel) and apply a vacuum inside the bag to compress the layers before and during oven curing. Overall, it seems either shortcuts were knowingly taken, or someone didn't know what they were doing. But what do I know...I'm just a 68 year old white guy that worked in manufacturing laminated aircraft components (including heat forged carbon fiber airframe connectors).
Saw this I think on Reddit. One reason SpaceX switched from carbon fiber to stainless steel pressure vessels (fuel tanks) was the difficulty of detecting structural flaws. I'm a lay person but my father is an architect and builder of skyscrapers who also went to Princeton only years before Stockton Rush. The federal regulations and codes in the construction industry, especially high rises, are insane but are very necessary. I'm sure the "crackling sounds" they heard was that carbon fiber breaking up! What else? Or those seals coming loose or that damn Plexiglas porthole that was inches too thin. Stockton was always talking about people in the deep sea exploration community using a "safety argument to disrupt innovation." The problem is you can't argue with time tested engineering and physics. If you're going to use carbon fiber, it must be in a woven pattern, not a single weave. That isn't strong enough. And that video of those guys applying the glue? Why weren't they using an autoclave? Wearing gloves? It's insane. That thing was built by students who had no idea what they were doing, lead by an arrogant greedy man who skimped on materials and had no regard for anyone else's safety pr lives, not even his own. It's all very sad.
I have 28 years experience in the middle east as technical manager in composites, and agree with you that the bond between cf and titanium was a question mark. For me, i will never introduce different materials for this application and compensate with a huge value for manufacturing constant not less than 10, and a value of 3 for cyclic load constant..Also, i agree with you, he should have placed layers of helix winding. Filament wound should be one-piece such as this kind of application.
When you hear the guy talking about and wondering, and almost bragging about just 'How far out the box' he is operating, it makes you feel sick that he could do this AND take others with him AND charge them a quarter million dollars at the same time. He will be remembered as one of the worlds biggest Wankers
I think that this is the best analysis of the OceanGate sub. And the only one that I've seen that's placed any attention on the bonding and process and the fact that the titanium ring was recovered with not a bit of fiber on it.
Thanks, this video was far more useful in understanding possible failure modes than others I've watched. Well done!
It also strikes me the carbon strands at the ends of the tube will be irregular where the machine laying it on the tube former changes direction (eg at 2:16). There would have to be some bunching of the carbon fibres on the inside of the bend formed as the head changes direction because of the width of the tape. There would also be some inconsistency as to where these bends lie on top of each other, just where there are the loads transferring the axial compressive forces from the titanium ends onto the tube. I guess the idea was that the resin glue would spread the load, but it does look to be an area where repeated loadings would lead to settlement and cracking.
In a cylinder, pressure (internal or external), will be two times in the hoop direction (circle around the cylinder) but one in the axis direction. For rockets and internal pressure you either add end to end loops or use a 27 degree “magic wrap angle” which provides needed end to end strength without giving too much hoop stress. I heard that there were layers of multiaxial carbon fabric added by hand in between the zero degree uniaxial wrap. But haven’t seen documentation.
Very interesting about that magic angle. I do wonder if a design more like an internal pressure vessel, would have done much better over time. Will be interesting to find out if any other axis were used…
Thanks for your professional input on the matter. My cousin, who professionally fixes yachts, said a similar thing, just with fewer details and with a lot of swearing. I use home made carbon/kevlar composites only for fixing RC toys. Before I touched the stuff I watched about 40 minutes of tutorials on Easy Composites website and without the slightest exaggeration I'm way more qualified to fabricate composite parts than any of those Ocean Gate engineers.
Sad but mostly true !
This is the best video I've seen so far because you are looking at specifics, not just criticizing Rush for a lack of concern about safety. My guess is they just didn't have the budget for a proper job, whatever that might be.
The carbon fiber and titanium have different compression strength and behavior. While they seem snug during the construction, at depth there would be a small but significant difference in the diameter of the fiber cylinder and titanium ring which would put a lot of force on that little layer of epoxy. I’m surprised it lasted as many dives as it did.
Excellent informative video without sensationalism. Thanks
Very interesting: It's actually worthwhile to know in detail what you're doing. Impressive knowledge, Sir!
Hi Jamie, some excellent views, thank you. It would be obvious to me that because the hull had no end to end fibres, that when the centre of the cylinder deformed due to the pressure it would mean that the INSIDE wall of the hull becomes under extreme tension from end to end, this would then lead to a radial crack at the centre which then propogates through the hull to the outer wall.
Precisely! This is absolutely what I think happened as well.
The whole world is on a morbid curiosity kick with this sub.🙂
I’ve heard a few people say this. But I don’t think it’s correct. In Rush we have a man who hand-waved every rule and expert warning and ploughed ahead with a bad idea, seemingly implemented with a corner-cutting attitude. I see a lot of people grappling to understand this attitude, the level of hubris, the number of warnings etc. I see a lot of people incredibly saddened by the needless waste of live. And I see a lot of people dumbfounded that in todays age of safety and compliance and regulation this guy was able to build a test vehicle, avoid certification, and avoid any regulation. The above seems to be the overwhelming sentiment I’ve come across. But there has also been plenty of talk about how fast the implosion would have happened and the immense pressures involved and how scary it would be ti get locked into that thing.
Also what about the microscopic bubbles that are cured into the matrix as the epoxy cured? a pressure vessel that sees the extreme pressure that that hull did would compress and uncompress with each dive breaking strands of fiber with each cycle. He heard this crackling and yet he did nothing to address it.
It's now 9 months later and I still find myself digging for more quality Titan content. I have no relation to the sea, diving or engineering and have zero understanding of why I'm STILL fascinated with every angle of this event. Really good content btw! A+
Thanks and hope it was interesting!
Thank you for sharing your experience and knowledge, this was a really informative video even I could understand, with the bonus that unlike some other videos about this, you make it clear each step that you're talking about possibilities, and we've got to wait for the final report before we actually know what happened. Going off your comments, that I'm guessing are best practices ones in your field learned over many years of "OK, that isn't gonna work", it's almost like they kind of had some idea of carbon composites, but they didn't actually understand it from experience. Kind of like someone who plays flight sims and then thinks they can fly a real airplane.
I too have experience in composite materials although it is sadly out of date and barely relevant here being with glass fibre reinforced polyesters. Epoxies were just coming in when I left the industry. Nevertheless I have maintained an interest in the engineering of such structures. I have followed the analyses of this incident with great interest and sought out calmly presented knowledgable sources like yours. Unfortunately I think the shock jocks and sensationalists will win the day achieving so much air time that REAL research into this topic will be buried in the BS. I agree there is a future for composite materials in this area but will we ever see it because lay people have been frightened too badly to permit their use?
It worked 11 times to Titanic on 30 other shallower missions. They were close. Didn't realize their margins were so small. 10% more plies and that thing might last for years and years. I'm still suspecting that bonded joint could be the root cause. We need to see the acoustic data from their hull monitoring system for all the previous dives. I'm not convinced that system would give them enough Time to ascend and get to safe depths. When composites finally let go, it's usually very quick.
I'm not sure 10% would make a great difference. The failure mode of rigid CF in a relatively malleable matrix, in compression, is surely quite different from that of homogenous materials subjected to cyclical tensile strain. I wouldn't expect the logarithmic gains you normally get from most materials, in most circumstances.
This is so important! Thank you for sharing 🌻
Very interesting hypothesis, I also believe the interface joints failed causing catastrophic compression. The two dissimilar materials definitely pose a very high risk of separation at the pressures experienced at that depth. I've seen pressure tanks for SCUBA made of carbon fiber, which are lighter than their metal counterparts, but are limited in their number of pressure cycles. It will indeed be interesting to see how future test and experimentation allow carbon fiber to be used in future under sea exploration. Very well done video on your part.
Nice to hear from someone with experience in carbon fiber composites. (CFCs) - titanium construction. A couple of questions come to my (4 year mech eng degree) mind. (1) water will seep into any tiny pore or crack due to capillary action. What is known about water at 5500-6000 psi forcing its way into tiny cracks of the resin matrix & promoting delamination or such? Or would the compressive stress tend to close the cracks? Maybe no one knows the answer since most deep sub manufacturers want nothing to do with composites in their pressure hulls. (2) I noticed what looked like a step in the circular edge of the CFC cylinder, that is, a reduction in diameter from the main wall down to where the Titanium end caps/rings are glued on. To my mech eng mind this screams out “possible stress concentration” which is double deadly under cyclic load. Are CFC subject to stress concentration due to changes in section (or notches) to the same extent as metals? The edge may be designed with a proper radius (or not) I can’t tell watching on a tablet.
Look up a report from last year, "review of composite hulls in deep water"... they even were kind enough to include pictures.
Physics does not care what material it is concerning stress concentrators , section changes, material changes, mismatched stiffness , mismatched thermal coefficients, and inadequate manufacturing processes all add up to knock down your theoretical safety margins. I believe the joint failed it is the hardest part to engineer, second most likely hoop layers were micro buckled by subsequent hoop layers as they alternated between axial pre impregnated mushy high temp epoxy layers and room temperature gelling epoxy but not fully cured as layers were added. Autoclave final cure did nothing to consolidate layers just cured the epoxies.
Are you aware of them hand laying layers of 90 over their 0 degree pre preg filament wind?
@@JamieKunka yes but the axis is 0 degrees hoops are 90 degrees. This is done routinely when winding interface skirts for rocket motor cases. It is a problem on thick walled buildups keeping the laminate consolidated as you build up. Squeegeeing epoxy over the voids between tows is not a substitute for compaction.
@@JamieKunka Real information on the final manufacturing is sketchy, as different videos are shown as time passes. Either way if it was all prepreg that would be ridiculous to build up 5” of uncurled material.. I once had a carbon fiber flywheel project for U of Texas we had to build up and cure in stages to keep the fiber straight. We could not do it in one shot. Inter leaving prepreg and wet winding that thick is also problematic. A real tragedy and it will take a year for the NTSB to issue a final report.
I'm an engineer Jamie. EXCELLENT, use of your expert analysis. Thank you.
Very good analysis of what and why likely happened the Titan implosion. Thanks. IMHO it's even intuitive that the carbon fiber layers should be put ALWAYS in several (many, even some layers at random?) different angles to get all its properties, strength and to diminish the advance of delamination and its speed of propagation. The RTM that Stockton included into his carbon fiber submersible was a good idea (indeed as per the transcript, it took about 19 minutes since the moment when the RTM alerted something was wrong with the carbon fiber hull to the moment of the implosion). So, if just he would have laminated the layers at many different angles, likely he had time enough to save the situation. Also, it's pretty clear that the delamination was already quite advanced since the beginning of that last dive because it descended faster than predicted (and yet faster as the submersible was diving deeper) and it then did ascend very, very slowly despite all their efforts, which was "unclear why rate is small" (I guess due to the delamination the carbon fiber hull was being compressed, deformed and so bending slightly and more towards the mid section of the tube, firstly silently, then loud enough for the RTM to note it and, finally, with terrifying bangs and cracking sounds...). As the hull starts to give in to the pressure and bends inwards, it displaces less water, hence it loses buoyancy. The carbon fiber composite is very hard and light material... but fragile when its limit is exceeded. So for submersibles, even being well constructed with carbon fiber, I think it's better to rely on other materials which are able to deform without cracking, like high quality stainless steel. And also the shape it's very important. Needless to say that the sphere is the best shape for this task. But, the shape of a rugby ball it's better than a cylinder shape (because the arc shaped structures are much more robust than the straight lines) as it's so widely used in construction of dam retaining walls and in architecture.
Very interesting, covered a lot of basic simple steps that I haven't seen mentioned. Sometimes it's the simple things that get ya.
Looks like EA9394, or similar adhesive. It's gray in color, would not need thickening agents, and smells eerily like peanut butter. Solid content. 👌
Very good ID on the adhesive!
Dose the adhesive you refer to contain asbestos?? Something about it sounds familiar!!?
@@robertchilders8698 No.
as a lay person the meer idea of 'gluing the sub' seems insane to me.
I liked your analysis it makes things clearer thank you
This my friend is a very good video very well done I think you are spot on and I don’t know why no one else has realized this or talked about it!
At last,someone who knows what they are talking about, doing a video on this. Great job and clearly explained. The passengers were doomed from the start it seems
He speaks, dresses well, and with confidence. A recipe for disaster Because the above overwhelm the non-credible and risky things he is talking about Like a lot of Start Up CEOs
Thank you. A very informative video. The design and construction of the submersible seems to have been an excercise in incorporating as many catastrophic failure modes as possible. I'm sure regulators around the world are sharpening their pencils.
A very measured, considered and informative review, Jamie. Thank you.
If it is correct the submarine was descending too fast and barely ascending later, could it be the carbon fibre cylinder had been soaking up water? Perhaps, if some of the fibres was actually damaged, voids had been created and water made the submarine heavier? So when the crew was weighed they should rather had been checking the submarine?
Very good point! High pressure water ingressing jnto the voids would not be good and as you say disrupt the buoyancy.
If the online transcript that's out there is true SR was hearing sounds in the aft part of the submersible.
@@MissX905 - the audible monitor was all red, perhaps it was water
The carbon fiber absorbing water wouldn't make it significantly less buoyant. The craft had thousands of pounds of ballast to give it buoyancy.
But, but, but the Rhino Liner!!!
The critical error was starting with two dissimilar materials. Just insane to begin with. Those two ends of the sub probably blasted off with extreme force.
Wow this is like, hilariously wrong. *nicknorthcutt's prior mechanical engineering experience consisted of Sunday morning Looney Tunes*
I cannot believe there was no reinforcement through the entire length of the hull to join the end caps together. It would be like having concrete without reinforcement. Utterly insane. Its almost like this was a highschool project.
Thanks for your expert observations! It's becoming increasing clear OceanGate didn't have any business building a submersible.
And design wise, i would think the titanium ring should be designed to fit within the carbon fiber tube, not outside.
By at least an inch. The pressure would press the tube into the seal, actually reinforcing the connection.
That would be a horrible way to do it. The force on the end cap would then be trying to slide the rings deeper into the carbon fiber tube. The only thing resisting that would be whatever you used to secure it to the carbon fiber surface. With it capping the ends, the force tends to press everything tighter together.
@@stargazer7644 not totally inside, but like we use for prescription bottles for my aunt who has limited mobility, where it is L shaped rather than the I beam shape seen here. Yes, if the endcap is totally inside the tube you have made a compression cylinder which spends all its time planning to crush the passengers.
These guys put less care into bonding those pieces than I would putting a screen protector on my phone. You know these guys phones have air bubbles between the screen and screen protector.
That's OK just make sure to wipe with a dirty greasy cloth before attempting to apply it
Y'know, to make sure it bonds really well 😂
The footage of the wreckage overlayed with Stockton Rush talking about his sub is some dark humour right from the depths of hell and I am here for it.
Jamie, this is so far the most factual analysis done about this accident. Most of the videos posted are mere opinions. You point out aspects that are not far from the cause of the accident.
What gets me is that Stockton knew that the carbon fiber tube would shrink by 3/4" (19 MM) at the depth of 12,500 feet. I know enough that carbon fiber is not known for it's flexibility properties. Stockton himself said it was concerning that the carbon fiber tube would make a cracking sound at depth.
I'm not an engineer, but it seems to me that with outside pressure, the sub was relying more on the strength of the epoxy that was binding the carbon fiber together rather than the fiber itself.
The sad truth is the whole sub community was aware of what was going on and only tried warning Stockton. The sub design would never pass certification, it wasn't tested correctly prior to taking passengers. Crush hull depth should have been tested at 16,000 to ensure 12,500 was safe to dive..typically, you add 20%...Stockton only went over less than 700 feet.
Good video! For consideration, forgoing the predominant error regarding use of Carbon Fibre tow axially wound as a total misapplication of engineering principles, the geometry of the mating surfaces between the titanium ring and hull was completely misunderstood by Stockton-the-Princeton-engineer as well.
Yes!!! Great video and I think you a spot on regarding the potential failure areas of Titan.
There are hundreds of decades old videos out there of nerds pointing out how carbon fiber is terrible for handling pressure.
I'm assuming you mean compressive pressures? Carbon fibre makes fabulous lightweight pressure vessels when under tension.
@@cameronwood9385 Composite SCBA air cylinders are extremely popular with first responders due to their light weight. The applicable retesting regulations do reveal the limited life of these cylinders, even under compression loads: carbon fiber & fiberglass: retest every 3 years, max life of 15 years. Steel and aluminum: retest every 5 years, max life is unlimited as long as the cylinder passes its pressure and inspection tests. I assume that someone is concerned with composite cylinder degradation over time.
The air bubbles helped keep them warm at depth.
I think the reason why we see no carbon on the end caps is due to the fact that there is a large change in stiffness at this interface, which is basically the definition of a stress concentration. Whatever caused the failure would have had cracks propagate quickly to this joint and then fly 360 degrees around in the blink of an eye. Basically, the sub could have been perfectly adequate in design and hit by a torpedo and I would expect a similar outcome.
Nice summary. Everything you say makes sense, and makes me wonder why Stockton Rush the -III- Turd had such lax manufacturing procedures.