I have watched literally hundreds of videos since the Titan disappeared and this is, by far, the best and clearest explanation for a total layman to understand. Thank you! 😊
Thanks. My goal is to try and keep the engineering principles simple for the average person to get an idea. The super technical math and material science behind it all is not necessary to get an idea of what it doing on.
That demo beginning at 16:55 minute mark of this video with the aluminum sheet ripping off with stretch pressure after poking that tiny hole in the middle was eye opening. Excellent analysis of the design flaws of the Titan submersible.
Great Video. One thing that really struck me is the environment where they were haphazardly applying the bonding agent. It was in an open warehouse setting, no one is wearing bunny suits or even hair coverings. I saw another video where one of the workers was touching the surfaces with bare hands as he was applying the sealant. Ideally this would be done by a robotic machine in a clean room.
Naval and aerospace construction has changed... Titanium and aluminum plates can be automatically welded on both sides of the material simultaneously and with laser and inert gas that prevents material deterioration... Robotics is of little use if you don't know how to do things... Publishing stickers with robotics is like applying the same thing to a riveter or a broom to sweep the garden... The aerospace industry and shipbuilding have changed... and the company's engineers were unaware of this topic. .. The new technology is legal in America and Europe...
I was thinking the same thing. While I don't have experience with this bonding agent, I am aware that impurities in bonds can cause problems because you have slightly weaker spots compared to others. I've seen this with various props I've made in the past. So doing this on a submersible that costs millions of dollars, was "pushing the limits", and where lives were at stake, you'd think they'd want to do everything they can to ensure everything in the process is basically in a clean room wherever possible.
awesome job! im obsessed with titan5 story and didnt fully understand why it is poor design to mix carbon fiber with titanium ends until your drawing and explanation. many thanks!
There was a video a few weeks ago by someone who found a text on submarine pressure vessel design. This a classic submarine hull profile. Apparently it stated that if the cylindrical pressure vessel varies from true circular by 0.5% you lose 30% of your depth capability, so it was essential to not allow it to distort under pressure.
Very well explained an excellent job thank you. When it comes down to it, the Titan DSV was a perfect example of Mr Rush's ego. Hubris or Hybris, whichever one you want to call it was definitely a major factor in play for this whole project. When your own ego and arrogance allows you to basically not believe what other experts (IOW the people who Mr Rush called the old guys) in the field are telling you and basically laughing them off because you know better was the recipe for this disaster. The Titan was a ticking time bomb there's no question about that. Sadly, when that time bomb finally went off it took five souls with it. RIP to those who perish and condolences to their families.
This is a very good discussion of the stresses on an object ! I have only seen low resolution pictures of the tube being manufactured, so I may be wrong, but I assumed the Titanium ring had a U-channel machined in it so the smooth CF tube would fit into this channel. That might help support the ends of the tube to reduce the chance of collapse right at the end of the tube. That still leaves a large area of the tube where any irregularity could lead to catastrophic failure. The demo with the aluminum sheet is very illuminating! It shows how even a small discontinuity can propagate so quickly.
Key things a lot of people forget. The sub has been successful in reaching the Titanic multiple times so there may have been no defects at construction. However, the Titan had been struck by lightning, in the Bahamas. Lightning can do damage on a molecular level, including to metal and even rock. It rearranges molecules. The molecules around the point of the lightning hit were scavenged and pulled towards the point of the strike. On some rock types a direct strike leaves a little Olympus Mons shaped raised bump, by scavenging material from the surrounding area in the instant moment of the strike, rearranging the molecules of rock. Not melting it in the shape. Then there is the Fulgurite, commonly called "petrified lightning". At the point of a lightning striking the ground, rock, organic matter, soil, sediments are fused on a molecular level into a glass like stone called a Fulgurite. I collect them. The next day during a dive is when OceanGate staff reported hearing distressing loud crackling sounds in the hull, that Stockton Rush totally dismissed.
@@k53847 They didn't repair or replace the carbon fiber hull and were diving the next day, after minor repairs. The first cracking sounds in the aft of the submersible were reported by the crew, during a the dive after they lightning strike, when they were still in the Bahamas. They didn't even build a secondary prototype CF hull to put through appropriate pressure tests. Did you see the leaked transcript yet? The Titan descended dangerously too fast, then he reported cracking sounds in the aft, and his real time hull monitoring (RTM system) alerts were all read. They dropped ballast, and even had to jettison the frame, but were still too heavy and ascending wayy too slow. Stockton also lost main power so had to route to a backup battery. Check it out, if you have not yet!
Excellent job!!! I was amazed at the lack of concern for the popping noises they were hearing from day one in the sub. What the hell did they think was making those noises? Just plain stupid in my view 🤷♂️
Excellent explanation, thanks. I was thinking along these same lines. Repeated compression and relaxing of the huge pressures along the length of the cylinder with each dive had to be causing some flexing back and forth. Not hard for me to imagine that causing stress to the "glued" joints, culminating in popping loose an end bell. I can't believe anyone would go down to those huge pressures at that depth in a vessel being held together with a thin layer of epoxy. Mind boggling. I think even a few hundred feet would be a big gamble. Rush was just going cheap to keep his costs down to maximize profit. Even though he was an aeronautical engineer, I would think any engineer would have thought better of gluing metal to carbon fiber and thinking it could withstand that level of forces, and repeatedly.
I have been intrigued by this story. Had me looking through an old physics book and it made me realize I need to buy an old mechanics of materials book as well! (previous engineering undergrad, but not working in engineering)
@@theeverydayengineer1 You provided a very nice explanation. Looking forward to hearing an eventual investigation conclusion but honestly I don't expect anything more than about what you said here. Maybe porthole. Isn't it wild though, that Mr. Rush thought that this was acceptable, to take that risk with no additional testing? That's part of why this is such an interesting (and macabre) story.
@@antontonableis it really that interesting? It’s just another tale is a rich guy getting away with things that a normal person wouldn’t be allowed to get away with, like not paying taxes. I was far more interested in why 500 people who died in the sea in Greece on the same day and why they didn’t get anywhere near as much press coverage, although the answer to both is obvious. They were poor.
@@StoutProper I'm interested due to the mechanics and engineering involved. A capsized vessel (albeit very terrible) is something that has been going on for thousands of years.
@@antontonable yeah, spend billions of taxpayers dollars on a rescue attempt they knew was pointless for 5 rich people but let 500 poor people die without any such efforts because it’s “boring”. Funny that you weren’t interested in the Titan before people died, yet the engineering was all still there. You’re but really interested in the engineering, you’re just interested in the deaths of 5 rich people, and couldn’t give a fig about 500 poor people dying. People like you with a lack of moral fibre and without an ounce of human decency disgust me
Great explanation. This tragedy was a classic case of "can't see the forest for the trees". Always listen to critiques they are not in front of the tree!
It is important to see the forest but also to identify the tree... because naval and aerospace construction has changed... and still people keep talking about Space X that does not have the technology and you cannot go to Mars in a steel tank stainless...IT'S A FRAUD...The submarine should have been built with lasers with the new technology admitted in 2020...THE SUBMARINE SHOULD BE MANUFACTURED AS A ONE-PIECE ONE-PIECE MONOCOQUE HIGHLY RESISTANT TO SEA PRESSURE...
13:50 You are making a mistake here: The contact is not made on the carbon fiber but almost only on the adhesion of the epoxy to the carbon fibers and the epoxy itself. It's almost only epoxy here. Because of the winding method of the carbon fiber which is transversal. Delamination can therefore occur in the same direction as the shear stress you have presented. It's actually even worse than expected.
I had watched a video on what might be the text communication with the Titan and the surface support ship. There where read flags when the text says the Titan decent was too fast, and once they realized a problem such as a bus failed they could not come back up. They also mentioned that their monitors had gone from green to red as the composite was delaminating and crackling.
that alleged "last text comm from Titan" has never been judged authentic, or validated by anyone in the Oceangate crew or anyone who had contact with them right after the event. In fact, it actually mimics a Titan text comm from 2019, when they had trouble dropping ballast at the bottom, but did finally make it successfully back. Also it first appeared anonymously on Tik Tok, not ever a reputable source.
I think there's a major clue to what happened in the fact that the crew reported that they were ascending at a very slow rate, even after jettisoning ballast and the landing frame. Possibly, there was a lot of water infiltration into the weave layers of the composite hull segment, apparently from the aft dome area, and this water was rapidly changing the buoyancy of the submersible. Water getting into the composite layers would also then quickly weaken the structural integrity, leading to catastrophic implosion when it failed. Just my thoughts.
Agree. Along with that, I suspect that the structure was flexing inwards, thus reducing it's volume and thus, displacement. It's displacement being the only thing that gave it upward thrust.
Thank you for your simple and clear explanation of the possible (most likely) cause of the Titan failure. I have watched the news coverage of the debris recovery and it appears the Ti rings appear to be very clean, there seems to be no broken pieces of carbon fiber still glued to the ring? Its almost like the Ti rings popped off cleanly, along with the front hatch and porthole (unless the savage crew unbolted them during recovery process) Hopefully, the NSTB will release an interim report and followed up by detailed analysis of the Titan loss.
Splendid break down and examples! Helps us non engineers get it. The more I watch, the more I can’t help but wonder about dives 1 and 2. I do wonder will we get a final assessment that presents in a manner much like in this video, the degradation but spread across 3 dives.. This submersible successfully went down to these depths 2 other times. I almost forget that point when I hear and start to understand the views and commentary. Maybe there are a crazy amount of interviews going on with those of the 1st and 2nd dives.. When I see that train car implode in that popular video playing everywhere now, it’s hard to believe this vessel had a life before that dive of 6/18.
It had something like 13 successful dives to the Titanic. The only other CF deep sub I know was did one dive (into challenger Deep) and then they condemned it because they had no real idea as to the fatigue life. This was all part of the original plan for that sub, not because of any problems they detected.
@@k53847 There is the "big disconnect". Was it two previous complete dives to Titanic's bottom depth or thirteen complete dives? I can't get a complete list of dive dates and bottom depths attained over the entire lifetime of Titan.
You are absolutely correct in your analogy, also they could have tried to mitigate the voids in the adhesive joint with the use of an autoclave or at least a vacuum bag, I saw no evidence of such measures during the construction of the hull, basically he knew nothing of strengths and weaknesses of CF. He could not of payed me enough to get in that thing.
The union with the titanium is not as you say. The fiber structure go complete into the molded titanium. It means that there is not a different thickness in the fiber border. However, it seem to be the same problem. When the fiber cilinder compress in the middle, the fiber in each union with the titanium will deform and it only depend on the epoxi.
Hello mate. Thanks for your video, I am actually a Civil Engineer and I am looking at it from a different way, instead of taking a cross-section which is a circle, if you take a longitudinal section of 1 inch thickness through the carbon fiber cylinder and the two ending titanium rings, you end up with sort of a beam of 2.6 meters in length and 1 inch in width and a depth of 5 inches (the thickness of the carbon fiber), the beam is now supported from both ends by the two titanium rings, a very simple structure with a uniform load from the top of 5600 psi, I believe the concentrated load in the middle will be massive and the carbon fiber depth needs to be far more than 5 inches, I also believe if Stockton put a third titanium ring in the middle between the two rings, but from inside only as a support to prevent the cylinder buckling in the middle point which is the point of max buckling, it would have made a massive difference in safety. Am I right? And sorry for the long comment. Cheers
I totally agree. I think he’s missing the point here talking about the pressure on the end caps, & he seems to forget that these force arrows on the materials are acting in all directions. Every test I’ve seen (including a rudimentary scale model in a pressure vessel) has shown that it’s the deformation of the tube (inwards at the middle) that caused this, pretty much blowing the end caps off.
Yes I was also thinking a titanium ring in the center could have helped, I was also thinking that a few cross members from the end caps holding that ring in place could also help. Then you would have a sort of titanium frame that is wrapped in carbon fibre.
That loud popping they heard was that epoxy separating and or the carbon fiber separating from the load on the spheres pushing end to end. This thing should have been x-rayed or sonographed BEFORE and then after the first dive when they heard the loud popping, it would have shown what was failing by comparing a known good exam (BEFORE ANY DIVE).
Probably want to take a comment, act like it's their own & create 10 threads themselves. I don't doubt that whatsoever. People love to take thoughts & ideas from others & act like they came up with it.
I have to admit, when you started drawing a circle to show how pressure applies to it I went "yeah yeah I know all this just get to the good stuff" but as soon as you used the 'keystone' analogy I was HOOKED! I wish I had teachers like you! 😊 I'm nobody and even MY first thoughts were that carbon fibre was designed to add tensile strength to incompressible epoxy so surely all he achieved was to give it the ability to delaminate. All of his education was aerospace, with at most 0-1 bar more pressure inside than, and he admits he fell in love with carbon fibre so I guess he built a thicker 787 fuselage somehow forgetting that the pressure is on the outside not the inside and it's 400x greater... nobody is that stupid I know, but this IS a guy who only found out a freshly replaced or rebuilt thruster was operating in reverse when he got to 4000m. I'm only certified to dive 55m but I already know that if my drysuit inflation fails I'll be cling-wrapped and literally incapable of ANY movement long before 40m which is 1/100th of the depth this thing was facing. We also stop and check each other's gear (no bubbles or potential snags etc) at 6m - apparently he checks his thrusters work when he hits 3800m 🤬 Death by narcissism 😥
1. The keystone beneficial shape is lost in the cylinder longitudinally, making it susceptible to buckling or shear inward, requiring much thicker wall than a sphere to avoid this problem. 2. A joint between dissimilar materials requires a thickened cross section, a flange - NOT a machined down narrower cross section as shown. 3. Given that carbon fiber is the strongest of the materials present, the glue is necessarily weaker, requiring a cross section several times larger than the cross sect of carbon composite; again a flange is required. 4. Carbon fibers are embedded in an epoxy resin, and therefore will fail at the resin, regardless of how strong is the carbon. 5. Using all titanium would have eliminated several problems: dissimilar deflection, delamination, weak compressive strength of fibers. 6. An ellipsoid shape rather than cylinder would have gotten better performance out of the same material and thickness. 7. The technology of bottling compressed gasses contains 5400 psi, 10,000 psi, and higher - so experimenting with human lives on new 'bottle' designs is completely unethical. They could have done multiple cycle testing in the lab on a miniature. -
Great explanation! The diagram and props, using laymans terms, assist in understanding the engineering, material issues, etc leading to the Titan disaster. Thanks!
I am not so sure I buy into that transcript. I see a few fishy things. It is very possible they may have heard cracking of the composite before it failed...if it was that and not the window.
Hi man. Minor point of clarification - It appears that the construction of the tube to the adapter pieces is that the titanium adapter is shaped around the full size of the carbon fiber tube - that is , there's no notching in the carbon tube; its the titanium ring cut to fit the tube. Sub Brief's 'The Titan Tragedy' shows a decent vid as they attach one of the rings to the tube. Also, there's a vid floating around where Rush claims - although Ive not seen it - where the carbon fiber wrapping did do crossing as well as straight. Still, there's also pics of items being attached to the inner hull liner - which according to thier on YT vid from 6 years ago was 1.5" stainless steel -- with shop screws poking in. Yikes. Of interest to this, is how there was, at least that theyve shown, zero fiber brought up from the bottom. The viewport is also, apparently, MIA. Personally, id love to see high res pics of the titanium end caps and adapter rings - the fact the port is gone, a look at the attachment points from viewport to hull will be telling, as will whatever witness marks remain on the end caps and adapter rings. End of the day, Using carbon fiber for a hull in this manner has been, well, crushed, permanently. Had Rush used a Titanium hull all the way through, odds are they'd of been fine - at least until the subpar viewport punched in.
From video I saw there is at least a notch on the outside of the carbon fiber hull as they are doing the glue up. I can see it with my own eyes. My hand drawing may not be to scale though. Even if there was no notch the shear stress at the joint would still apply since they didn’t taper the titanium step over down gradually.
@@theeverydayengineer1 Huh. Maybe we saw different vids then? The one i saw, quoted on SubBrief, the line i see appears to be glue. Really though, it doesnt terribly matter if it was or not. From the second i first saw that sub, there was no way in hell i wouldve set foot in it - Hell i dont even like getting on damned Ski lift gondolas! Anyway, new sub, will keep an eye out for ya!
I'm an architect. Your basic analysis is spot-on. Your drawing / section of the flange to tube connection doesn't look quite right though. The Ti ring has what appears to be a 90 degreee 2" overlapping flange at both the interior and exterior carbon fiber tube interface, with a flat direct contact surface between them. That aside, the likely failure mode was indeed loss of adhesion between the Ti end-cap rings and the CF tube when the CF tube deflected at it's center due to the enormous pressure it was (repeatedly) subjected to. The inherent defelction problem was compounded by what appears to be an idiotic failure to weave the fibers in a way that exploited their enormous tensile strength. Had the first 1/3 or so of the INTERIOR weave included a lot of courses PARALELL to the logitudinal (long) axis of the tube, they might have resisted the tensile stresses caused by deflection in it's center. Combined with a resin matrix reinforced with....uh....titanium or steel fibers / nanoparticulates or something with more compressive strength and ductility than whatever the hell they glued it all together with they might have been able to reduce the deflection of the tube. The other problem appears to be insuffcient overlap of the Ti shoe-box style flange with the CF tube. There's an old saying in mechnical / civil engineering: "If it don't look right, it probably isn't" Even a lay person can see that the proportions of that ring and it's flanges "don't look right". I'd expect to see at least a 1:1 ratio between the 5" direct mating surface and the overlap between the Ti ring flanges and the CF tube. In other words those flanges should have been at least twice as deep. At least. More would be better, and a two 4' sleeves glued with a slow thin slow-cure reinforced expoxy and mated at the center with another ring to fully enclose the CF tube would have provided MUCH higher strength, containment, reduncandancy and resistance to flexural failure of the adheisive bond between the Ti and CF. But then Numb Nuts Rush the innovative billionaire would have a boring old Ti submraine rienforced with a Carbon Fiber tube, instead of a pure CF submersible with a couple of Ti endcaps slapped on with some JB Weld. Then there's the whole issue of repeately load cycling the assembly without any way to asesses fatigue performance short of destructive testing. This is exactly how Oceangate SHOULD have done thier R&D to establish a safe lifecycle for the sub: 1) Build 5 identical full scale subs 2) Take them to the Mariana Trench 3) Send 4 of them down (unmanned!) to twice the sevice depth you plan to use them at for paying customers. 4) Repeat this twice as many times as you plan to keep your final operational submersibles in service. Asssuming none implode, cut them up and asesses material performance. Redesign if anything looks questionable. 5) Send the 5th sub down until it implodes. Study the wreckage to identify ultimate failure mode. After all of that, rate your sub for commercial sevice, again, with service depth and lifecycle 1/2 of what the design is proven to through direct testing. Yup - it's gonna cost a fortune, and the resulting operational submraine will cost twice as much as "necessary" from a purely engineering standpoint. It also won't implode.
My cross section may not be to scale, but it was intended to have 2 flanges that fit over the composite tube (with some machining of a step in the composite based on videos I saw. I keep thinking how they could have made that seam better, and a low viscosity resin sucked in under vacuum could have been a possible better choice. Also the flanges on the titanium end cap should have been longer and had a taper to them to gradually change the force loading at the transition between the titanium and cap and the composite tube. The bond joint is just one of a multiple possible failures. They may never know what really happened.
I'm less familiar with their design but I saw some images and saw the raised floor, as well as the wall in the back. From the supposed transcript of the event, it sounded like they heard cracking from the rear. It also sounded like they were increasing depth at a fast rate as well as having buoyancy issues after dropping weights. I'm wondering if the aft bulkhead would obscure a small pinhole/crack leak in the adhesive as you mentioned and let water in where it couldn't be seen, displacing their air and dropping buoyancy. Combine reduction of buoyancy and existing failure in progress and you get an event.
Very doubtful. I don’t think a pinhole type leak would happen, and then at a future point in time a catastrophic implosion. To me that transcript is a hoax. Too many flaws in it like someone is trying to make a juicy story. The submersible would not take on water like a ship to have buoyancy issues. I think the back wall was just a panel to cover up electronics and maybe some other support systems that may had been back there.
One of the better analyses of the problems with the adhesive joint where the titanium ring flanges were (sort of) adhered to the composite cylinder. We could go into a several hours long discussion about surface preparation - or the lack thereof - in order to give the adhesive more "bight" to the titanium. However, almost nobody has gone into any thought about the joint's lack of resistance to tear-out forces as designed. Much as we use anchor bolts when attaching equipment to a reinforced concrete foundation, it would be smarter to run bolting rods tied into a reinforcing cage within the composite, such that the titanium ring flange could be positively held in place by the clamping force. Such a design however would in my opinion not be sufficient to save the vessel, as the use of composite material brings with it the flaws of delamination and debonding with service cycles. And, CF composites in particular tend to break nearly instantaneously, not giving you enough warning to back away from the impending failure.
Great video. I'm familiar with how vacuum chambers are used to remove air pockets from epoxy even in hobby projects, should that have been done here? (ignoring that they would still have a huge problem with the CF). Also, Rush wasn't onboard every time. There was at least one time when it went down on an expedition with another staff member piloting it. That was the time they had installed one of the two horizontal thrusters on backwards and so the sub was spinning in circles on the sea floor. They had to remap the game controller while down there to fix it.
A comment below suggests he must have been aware of this. I’m not sure- if you suffer from hubris you kind of neglect reality. He overestimated his abilities. And he had achieved some remarkable success, so nobody could talk reason into him. But clearly the worst part : he took 4 victims with him.
You are perfect to analyse the failure of Titan. Being structural engineer I wonder why the Titan was not structurally analysed by competent authority before and tested throuhly three diamensionaly ? Behaviour of composite material depends on strength/weakness of each other with strength of bonding material which is basics of composit material and it is widely used any many areas of engineering or manufacturing item.
Thank you so much for explaining this in simple terms. My dad’s an engineer and let me tell you when he explains stuff to me is like he’s speaking a different language 😅
Honestly, when I saw the clip of them mixing the ‘glue’ and the application I was floored. Glueing dissimilar materials is so hard. What were they thinking.
Liked, subscibed. A great explanation. I'd like to add that the weakening that's described due to the delamination, could allow compression of the tube. That would possibly be within the modulus of the inner metal tube and be unnoticed, but the reduced volume would result in lower displacement and thus buoyancy, (upward thrust).
Having watched almost all reputable tech channels (yours among them) in this matter I've come to the conclusion that they started diving the Titan with a 50-50% chance of survival from the get go , diminishing this chance after every dive to as low as 0% on the fateful day .... In my 27+ years of experience with composite materials in the military aviation field I've never expected to see such disregard not only to safety regulations and certifications , but a blatant disrespect to science itself ....
Thanks. I used to go tutoring to engineering students, and I frequently do training for my employer. I try and use analogies to things we experience in everyday life. Basic engineer principles are not hard if you can visual them.
Defies logic to have titanium endcaps GLUED with a mechanical bond at best to a dissimilar material. Plus the glue, titanium, carbon composite, and acrylic window all react differently to pressure and temperature. There needs to be an in depth investigation or book written about Stockton Rush, billionaire killer.
Yes, the mix of materials was not a good idea. The acrylic windows have been the go to for submersibles since fay one, but he flirted with disaster on such a large diameter window.
Very good analysis. Especially for the non-engineer. Easy to say this now, but after forty plus years of building carbon composites, my initial reaction to seeing the pressure vessel, was complete disbelief. I just had to search my memory for the number of times I'd engineered a compressive structure. And all i could come up with were three. And those required extensive omni-directional laminates. I'm still at a loss as to how this was even considered. And even if attempted, how both vacuum and spectral inspection was not in the mix. Thanks for the video. Keep up the good work.
Thanks. I tried to hit the main points for the average person and not make a crazy long video trying to break down every little technical anomaly. I do not have the expertise to say much about use of composites in compression. I am actually curious about studies using carbon fiber in such a manner.
@@theeverydayengineer1 You did fine. Always effective to use simple practical examples to get more complex concepts across. The reality is, composites, regardless of whether they're carbon, kevlar, etc., are rarely used in compression unless they are part of, or complement, a more complex construction of other materials. That's why, for instance, composite wing structures, even on light aircraft, will incorporate substructures to distribute compressive loads, while allowing tension loads to be somewhat independent through a limited vector. From my perspective, the real issue with the Titan's use of this material and process, is that we go into composites knowing that the only way to validate integrity, is by use of pre-pregs under vacuum. And then, non-destructive testing. With metals, we can control the matrix. With composites, that matrix is only known or trusted, on the material level. Not the construction level. Thanks again. Keep it up.
Very reasonable idea about water seeping into the glue joint and then destroying layers of carbon fiber. This explains the newly appeared text messages exchange with the submarine, where the crew heard "sounds" from the aft and was alerted for some minutes before contact had been lost, i.e. the destruction was not instantaneous.
The carbon fiber/epoxy hull of the Titan features alternating layers of UD prepreg in the axial direction, and wet-wound carbon fiber filaments in the hoop direction. So its not just in one direction as many self proclaimed internet expert are saying. Also the titanium end caps are held in place by the pressure itself when under water. The glue are just there to stop it from falling off when not under pressure. Its not an area of concern.
Thank you for the information on the hull. I was certain it had to have axial fibers to hand bending stresses on the surface and during loading and unloading. Interesting that they mixed pre-preg and wet-winding. Yes, the titanium ends would be kept in place by the water pressure along under water. The adhesive would only serve a purpose on the surface.
It most definitely is an area of concern, because the groove is complete responsible for transferring the end cap loads to the composite shell. The epoxy has a strong influence on this load introduction.
@@marconawijn405 You can use anal lube instead of epoxy glue and as long as it stay in place until you start diving, it will be perfectly fine as the immense pressure press the titanium ring against the hull of the sub.
World's best and first correct analysis available anywhere. Let me add a little quantitative analysis (which is so easy btw that it can be done in five minutes). Compressive stress in the carbon fiber polymer (CFRP) wall of the Titan submersible was about 200 newtons per millimeter squared at Titanic depth. This corresponds precisely to the stress level where CFRP typically fails under compression. Only under tensile load (which is not applicable underwater), CFRP can withstand about 10 times more. What's more, OceanGate used unidirectional (circumferential) CFRP, as you have also stated. This means that in the longitudinal direction, the hull was only about half as strong, at around 100 newtons per millimeter squared. It's a miracle the ship even made it to Titanic a few times without imploding. It was like doing pole vault, but with a rope as the pole.
With all fibers in the hoop direction - the strength in the axial direction is about 5% as compared to the hoop direction. And - in compression the strength of CFRP is about 60% that of tensile strength - not 10x less.
@@DavidG2P Tensile strength 900-2500 MPa - OK, although 900 MPa would be for plain-weave fabric rather than tape. The 900 MPa for tape would be an extremely low number (read: shitty tape which belongs in a scrap bin). Compressive strength 120-420 MPa - incorrect. For a tape with 2500 MPa tensile strength, the strength in compression should be 1300-1500 MPa. If you are getting 12-420 MPa - this means that the test is incorrect (like very long non-stabilized test coupons which fail essentially in buckling). 60% of tensile is a rule of thumb value. These are the results I have been getting on a number of carbon fiber materials in the last 30+ years. Fibers like (somewhat old but still widely used) T300 fibers reliably produce values of about 2100 MPa in tension and 1300 MPa in compression, on coupons with 35% resin content, tested to ASTM standards. Compressive strength transverse: 60-150 MPa is OK, as this is consistent with strength of resin in compression - although 150 MPa would call for a rather fancy resin. We are talking here about the failure stress - so, if you can - please indicate a source of "Compressive strength longitudinal 120 - 420 N/mm²"
@@DavidG2P If so - then these numbers are either incorrect, or refer to something else, or it simply is a typo. I have spent the last 40 years designing, analyzing and testing composite structures (aircraft).
So I assume that once water ingress occurred, potentially in the region of the glue layer, the pressure of the water at depth would actually work to push the titanium rings away from the composite hull axially and vice versa causing it to decouple. This seems more likely than the sheer at the interface.
Nicely done! Very good use of diagrams and props. Your example poking a small hole in a sheet of aluminum foil was brilliant! One minor correction, I believe you are using seabed level crush depth pressures in your calculations. The sub appears to have imploded at less than half the total depth of the anticipated dive.
Since the loss of text and location comms occurred 1.5- 1.75 hours into a dive that typically was 2 hours to seabed, they were probably much farther down than less than halfway, given that implosion would likely occur within minutes of their primitive hull breach warning system. Anyway, investigators will probably be able to make some estimate of their depth, based on the distribution of the wreckage, and the contents of the REAL text communications and locator documentation of the last 2 hours of titan. (The alleged "last text comm" originated from an anonymous Tik Tok poster, which has never been a reputable source for truth.) Very reasonable to use pressures at seabed as an example, when outcomes would be similar even at lesser depths.
Just realised that the fracture line in the titanium to carbon Junction was only held by resin and no fibre. The fibre was only wound in one direction around the circumference, hence parallel to the stress line. Your crack lines of delimitation should be vertical, not accross that junction horizontally.
I still think there was horizontal layers mixed in to support all the bending stress of holding the two titanium end caps on. I can’t prove it, nor disprove it, however. Also the cracks would form in the resin between the layers of fiber, hence the horizontal fracturing.
@@theeverydayengineer1I found an article in a 2017 trade magazine dedicated to composites where the company that designed and manufactured the hull of the Titan (then called Cyclops 2) talks about it. I'm not sure if I can post urls in comments so I'll just quote the relevant passage: "The biggest challenge, Spencer reports, was developing a manufacturable design that “would produce a consistent part with no wrinkles, voids or delaminations.” And without use of an autoclave. Spencer opted for a layup strategy that combines alternating placement of prepreg carbon fiber/epoxy unidirectional fabrics in the axial direction, with wet winding of carbon fiber/epoxy in the hoop direction, for a total of 480 plies."
You forgot to discuss the matters of the epoxi matrix, the lack of pressure treating in an autoclave, and the different and opposing contraction and expansion temperature coefficients of the carbon fiber cylinder and of the titanium caps. The latter contract when temperature drops, while the carbon fiber matrix expands. Otherwise this is an excellent presentation. Thanks!
Thanks. Yes, there is soooo much to potentially talk about, but I tried to keep it as short as possible and get to likely the most important issue with the design for the average person to understand.
I’ve done 3D printing where the layers were perpendicular to the direction of force, and the layers separated cleanly. If the submersible was rolled in the loop fashion, and then a section was under a sheer force like you described at 13:52, could the machined mating point on the carbon fiber ring snap as the layers are perpendicular to the direction of force?
It would be more likely, but it also looks like they laid down a pretty wide fiber that would have been woven providing support in the perpendicular direction, but it would be not nearly the same as actually putting in fibers in helical or horizontal direction.
I was kind of waiting for someone to explain that carbon fiber composite, when fully cooked, became very good at... tension, not compression. Compression was just trying to spread out fibers and separate them from each other, not using any of their molecular strengths. When you only count on the surrounding cooked resin bond to do the job of resisting to those multi-million pounds of pressure, this is basically what happens.
Would the fact that it was salt water possibly seeping in make it worse? Salt is corrosive, plus when the water evaporates, salt crystals form which could physically force layers apart.
Add in temperature, coefficient of expansion, and every mechanical joint. Coming from a race car background I’m aware of testing metallic components for stress fractures such as Magnaflux inspection, however I am unaware of such a process for carbon fiber, I assume there is SOME process but wonder if the part size make the process restrictive for a dive hull.
There is ultrasonic testing to test for delaminations and voids. There are very big machines that can do this. But that might only be practical when the tube was initially built, and would not be practical after the vessel was assembled.
Because the net force squeezing in is equal to the area of the circle since some of the force is being directed to try and pinch the hemisphere in the perpendicular direction.
I wonder whether the cold water could effect the glue and make it brittle and unflexible. It going from the surface temperature to 12k feet down must have some effect, even if they descend slowly.
The water temperature is not all that much below freezing at that depth. That isn't all that cold to likely create a brittle fracture environment. But it is something that could be a possibility depending on the exact material used.
Question: I get the keystone kind of trying to keep the pressure equal, on the cylinder shape, instead of you getting the keystone shape 360 degrees, you only technically have it in one direction and less so as you kind of turn it, until once you get to 90 degrees it has lost the keystone-like protection, would this also drastically weaken its ability to resist being crushed, or would the few degrees of keystone protection (for a lack of better choice of words) offer sufficient resistance to compression?
You are correct. The "keystone" in a sphere necks down on all four sides from the outside to the inside, while in the cylinder it only necks down on 2 sides. The two sides are sufficient in a cylinder. Making the wall a little bit thicker can help regain any compressive strength that may be lost.
Have you considered, (disregarding the issues of the end caps etc.) with a hoop wrapped Carbon Fibre with no end to end fibre reinforcement that when the pressure in the centre of the cylinder pushes the cylinder in at the middle, the INNER cylinder wall will be under extreme tension as it is being STRETCHED by the deflection of the five inch thick wall section. This tension on the inner wall would then lead to the resin between the fibres cracking and propogating around the cylinder to the outside wall in a perfect radial crack in the centre of the cylinder. The hull is effectively ONLY held together from these tensional forces by the epoxy resin itself, so a CAD model of these forces should use resin only without reinforcement to see how the resin cracks under tension.
Yes. I tried to keep the video as short as possible and go with my gut feel on the windings. Some commenters said indeed there was horizontal and/or helical layers. I figure there would have to be. I plan to do another video on basics of composites.
Very informative, I feel you are spot on with your analysis, with the leaked messages from ship to Titan, just before the messages ended Stocton Rush said he could here crakkling sound in the aft, I bet it was delimiting in the rear section between the Titanium and glue and Carbon fiber.
I have a question: wouldn't the radial compressive force on the cylindrical faces of the hull have a sort of stiffening effect on the composite fiber plies that would counteract delamination due to the compressive forces of the two end domes?
@@theeverydayengineer1 I also don't really understand why the joint was made with such abrupt changes in section, groove in the ring and tooth in the cylinder. It should be general knowledge that such a thing will create stress concentrations and so on... They really seem to have taken a tourist's approach to the project.
@@theeverydayengineer1 I was wondering: if the taper on the titanium ring was on the inner side, the pressure on the end caps would make them work as a wedge, opening the end of the cylinder. Combining this with the radial pressure acting directly on the carbon fiber cylinder, would this amount to clamping radially the cylinder, avoiding the crushing of the fibers? Combined with the taper it may have improved the life of the structure.
With the way that the tube was laminated, the stress at the titanium/tube glue joint perpendicular to the tube means that the strength of the tube in compression is just the shear strength of the resin. This is also true at the half length of the tube where compression deformation of the tube would be putting tube outside length wise compression stresses and inside tension stresses on the weakest material, the resin.
You have provided the most reasonable cause and effect analysis of what could have led to the implosion. My observations watching how they manually MIXED the glue instead having it mixed in a controlled environment. The application of the adhesive was also manually applied and was not uniform. My question is what were the engineers on staff not doing the due process to determine that the Titan was not not structurally sound. I believe they also share responsibility as engineers in whatever judicial actions are given to the company! Where were the safety and quality folks? Probably helping mixing the peanut butter that HUBRIS Stockton was so proud of. I am not an engineer and had the same observations as you did. Thank you! I have subscribed to your channel!
Best news is that 1) a licensing body for these vehicles may be created from this event and 2) their deaths happened faster than the human nervous system can transmit information.
I watched a number of videos where "engineering analysis" of possible causes of Titan failure have been presented. I am an engineer with 40 years of experience designing and analyzing composite structures - and composite is about the last material I would pick for this application. Having said that - it is obvious to me that vast majority of these presentations are made by people who have no idea or understanding of composites. Yes - the choice of material was incorrect. Yes - moisture ingress into the matrix, manufacturing flaws, porosity, bonded joint failure, difference in thermal expansion, difference in stiffness (titanium caps vs composite cylinder), cyclic fatigue, impact damage, microcracks, application of cylindrical shape etc. - all these COULD be contributing factors. What is astonishing - is how little attention is paid to the LAYUP - that is, orientation and placing of the fibers. Pressure at 4000m is about p=40 MPa. With 4.7ft diameter and 5" wall thickness - the ratio of R/t is about 5.6. Hoop stress is about Sr=p*R/t = 220 MPa Axial stress is about Sa=p*R/2t = 110 MPa Carbon tape in tension fails at about 2100 MPa, and about 1200 MPa in compression (dry, RT). But in TRANSVERSE direction (perpendicular to the fibers) it can only take about 80 to 180 MPa in compression - as this is limited by the strength of the matrix alone. Compare axial stress and the strength of epoxy matrix... If the cylinder was manufactured by winding ALL fibers in the hoop direction (which I still can't believe - because this would be total idiocy) - then the matrix would be stressed almost to its ultimate stress in repeated dives. Most of the comments and analyses concentrate on "wrong material" (yes, for this application), number of "Empire State buildings pressing on Titan" (BS), inability of carbon fibers to take compression (BS) etc. - while totally missing the basics of composite design and analysis. If there really were no fibers in the axial direction - then this was the most probable reason for failure.
Question: I've never worked with carbon fiber composite, but I know from experience with insulated aerial lift booms, that glass/resin composite is highly hygroscopic if fibers are exposed. Is the same true for carbon fiber composite?
@@theeverydayengineer1 I was musing more along the lines that since water seems to accelerate delamination in FRP at normal atmospheric pressure, I can only imagine that any surface fissure, subjected to water at 300+ atmospheres, would also speed delamination.
Well explained. As an Engineer you know that we have to wait for the full investigation report, other systems such as the porthole were also contentious, so we don't know the exact probable cause but I am with you in the theory that this is the most likely scenario. (I am a BEng CEng Automotive).
It looked like the window was blown out when it imploded. I can't imagine they would remove it just to create a hoist point, every investigator would freak out.
@@k53847 yes, but it could well have failed due to the immense forces which would have been generated at the point of implosion, it will be interesting to find out.
Some comments from an experienced composites engineer. 1. Like you I was very concerned by the appearance of only hoop winding in the shell, though I have to wonder how the vessel survived even it's first dive if this is the case. 2. You describe what is termed a 0-90 layup. There are also helical winding techniques where the fibers run at an angle. 3. Carbon fibers are very stiff. They are not limp. Those who work with them often get stabbed by an individual fiber. They really hurt. This is in contrast to Kevlar fibers which are quite strong but very flexible. Properly designed and bonded carbon fiber structures can withstand compressive loads quite well. One common example is the top of helicopter blades which are under compressive forces. 4. The joints between dissimilar materials terrify me. Thermal expansion differences. Differing Young's moduluses when stressed. The sharp 90 degree corner on the titanium flange ring is a tremendous stress riser almost certain to cause problems. I have had to design mounting of steel rings on rtp structures. The best approach is generally to provide some sort of cushion between the materials. 5. In the photos, the area of the glue line attaching the flange appears to be much too small. Socket joints on fiberglass piping less than a foot in diameter are generally deeper. 6. Components for sizable pressure vessels are generally not cast. Casting issues can cause many problems with grain structure. Shells are rolled from plate. Heads are spun from plate. Smaller items may be machined from rolled stock or forged. 7. While you are correct about the advantages of spherical shell, a cylindrical shell can also be used. A cylinder will be heavier for the internal volume, but the navies of the world use huge cylindrical shells for their submarines. These large cylinders are reported to be designed for as deep as 3000 feet which is just as much a design challenge for a large cylinder and 20000 feet for a small cylinder.
I almost mentioned helical winds but I wanted to keep it simple for the average person. I have worked at times with tubes with helical and hoop winds. It looks like I cut out a small section during editing that stated composites under compression are not necessarily a bad thing, just that additional considerations to factor in. The titanium end cap should have tapered the flange to almost nothing over several inches to transition the stress load. I tried to keep the video short enough to keep attention span. I was shocked I was still over 20 minutes.
"One common example is the top of helicopter blades which are under compressive forces." Let's not forget about centrifugal forces...A helicopter rotor forms a cone with tension on the blades - this is what keeps it in the air. There is in fact very little compression if any. But on the wing of an aircraft - yes. Generally you are right elsewhere :)
Sounds like basic design was fundamentally flawed. There seems to be many mods that could make it stronger. For example, 1. Considering tube thickness. 2. Using strong backs which buttess the tube. 3. Using a taper shape for the matiing surfaces of the end tube. Just like the mandrel in a milling machine where the harder you press, the tighter the fit when trying to machine steel at 3000 rp m and you don't want it to come loose
Can we assume Titan was taking on water in unseen areas because of the rapid descent and the difficulty in climbing to the surface once mission was cancelled? If yes, then this would indicate a failure much earlier than any of the monitoring systems detected?
I think that is completely bogus. Where would it take on water? It would have to take on a significant amount, that would have had to be internal to the pressure vessel, and they almost surely would have heard it coming in. That transcript has too many fishy things in it in my mind. Someone tried too hard to embellish the story in my opinion.
As an "aeronautical engineer", Stockton Rush was out of his element. The bad choice of the cylinder to extend passenger capacity was a commercial decision. Business determination can not over ride the laws of physics. If the cylinder war deformed by the water pressure the glued seams could separate. Stockton Rush's most notable quality was that he was a was persuasive salesman, but he was not a scientist. .
If i’m not mistaking - what they attempted, was a purely layered all-conposite tube, between the titanium half-dome ends. from an engineering standpoint, i feel that if they had: 1 put segments of titanium rings also inside the laminate, and 2. titanium rods - connecting the two end-half spheres, lenghtwise along the lateral tube, then they would’ve solved most issues of compression, while maintaining a weight advantage of having most of the cylinder composed of composite. aside from that, i don’t see the argument for weight savings being relevant in a submarine- on a plane, yes, but in a sub, where tou wish to decrease buyancy - not really
You could do something like that with the rings and the rods, but that would make the internal cabin smaller, which isn't good for making a sub for tourism. Yes, trying to have weight savings in mind for a submersible is a little overkill.
Great explanation. Thank you. It looks like Stockton Rush was trying to create new physics ( well - it just doesn't work like that) it was stupidity but not innovation
Also the plastic binder's specific gravity increases at a far greater rate than carbon fibre under increasing pressure causing laminations due to the binders plastically reduced volume seting up rapid fatigue failure under repeated loading/trips. NOTE: Boeing engineeers should have condemned the carbon fibre for destruction once it passed it's "use-by" date?
I have watched literally hundreds of videos since the Titan disappeared and this is, by far, the best and clearest explanation for a total layman to understand. Thank you! 😊
Thanks. My goal is to try and keep the engineering principles simple for the average person to get an idea. The super technical math and material science behind it all is not necessary to get an idea of what it doing on.
That demo beginning at 16:55 minute mark of this video with the aluminum sheet ripping off with stretch pressure after poking that tiny hole in the middle was eye opening. Excellent analysis of the design flaws of the Titan submersible.
He made that very simple to understand and to a simple man like me everything he said was logical.
That is the true but nothing that the true
Agreed !
Great Video. One thing that really struck me is the environment where they were haphazardly applying the bonding agent. It was in an open warehouse setting, no one is wearing bunny suits or even hair coverings. I saw another video where one of the workers was touching the surfaces with bare hands as he was applying the sealant. Ideally this would be done by a robotic machine in a clean room.
Naval and aerospace construction has changed... Titanium and aluminum plates can be automatically welded on both sides of the material simultaneously and with laser and inert gas that prevents material deterioration... Robotics is of little use if you don't know how to do things... Publishing stickers with robotics is like applying the same thing to a riveter or a broom to sweep the garden... The aerospace industry and shipbuilding have changed... and the company's engineers were unaware of this topic. .. The new technology is legal in America and Europe...
I was thinking the same thing.
While I don't have experience with this bonding agent, I am aware that impurities in bonds can cause problems because you have slightly weaker spots compared to others. I've seen this with various props I've made in the past. So doing this on a submersible that costs millions of dollars, was "pushing the limits", and where lives were at stake, you'd think they'd want to do everything they can to ensure everything in the process is basically in a clean room wherever possible.
There was a video of a dude cleaning the titanium with a dirty rag and then touching it with his other bare hand right after.
awesome job! im obsessed with titan5 story and didnt fully understand why it is poor design to mix carbon fiber with titanium ends until your drawing and explanation. many thanks!
I am glad I could help you.
It's not poor design, it's depends on method applied and application. That's what's poor in this instance.
Thanks for this explanation, for those of us not engineers but who are analytical! It is the best video I have seen and understood!
Well at least the Titan tragedy has one good result, IT PROVED THAT BUILDING A SUB WITH CARBON FIBRE CAPSULE IS A VERY BAD IDEA!
Very good explanation, thank you!
Unfortunately this submersible was, as we say in my country "on a one spit of life".
There was a video a few weeks ago by someone who found a text on submarine pressure vessel design. This a classic submarine hull profile. Apparently it stated that if the cylindrical pressure vessel varies from true circular by 0.5% you lose 30% of your depth capability, so it was essential to not allow it to distort under pressure.
Very well explained an excellent job thank you.
When it comes down to it, the Titan DSV was a perfect example of Mr Rush's ego. Hubris or Hybris, whichever one you want to call it was definitely a major factor in play for this whole project.
When your own ego and arrogance allows you to basically not believe what other experts (IOW the people who Mr Rush called the old guys) in the field are telling you and basically laughing them off because you know better was the recipe for this disaster.
The Titan was a ticking time bomb there's no question about that. Sadly, when that time bomb finally went off it took five souls with it. RIP to those who perish and condolences to their families.
I would agree that ego got in the way in trying to revolutionize a niche market.
This is a very good discussion of the stresses on an object !
I have only seen low resolution pictures of the tube being manufactured, so I may be wrong, but I assumed the Titanium ring
had a U-channel machined in it so the smooth CF tube would fit into this channel. That might help support the ends of the tube to reduce the chance of collapse right at the end of the tube.
That still leaves a large area of the tube where any irregularity could lead to catastrophic failure.
The demo with the aluminum sheet is very illuminating! It shows how even a small discontinuity can propagate so quickly.
As a professional engineer with 30 years experience you pretty much summed it up in easily understandable terms. Nicely done!
Btw. I’m a 50+ year old engineer. No way in hell would have signed off on that
Thanks.
By far the best analysis! Thank you for breaking it down in such simple terms!
thanks
Finally! This is the best titan and whiteboard video of all time!
Thanks. I tried to keep it simple without getting too in depth.
Fellow student engineer - Amazing explanation and really well articulated video!
Key things a lot of people forget. The sub has been successful in reaching the Titanic multiple times so there may have been no defects at construction. However, the Titan had been struck by lightning, in the Bahamas. Lightning can do damage on a molecular level, including to metal and even rock. It rearranges molecules. The molecules around the point of the lightning hit were scavenged and pulled towards the point of the strike. On some rock types a direct strike leaves a little Olympus Mons shaped raised bump, by scavenging material from the surrounding area in the instant moment of the strike, rearranging the molecules of rock. Not melting it in the shape.
Then there is the Fulgurite, commonly called "petrified lightning". At the point of a lightning striking the ground, rock, organic matter, soil, sediments are fused on a molecular level into a glass like stone called a Fulgurite. I collect them. The next day during a dive is when OceanGate staff reported hearing distressing loud crackling sounds in the hull, that Stockton Rush totally dismissed.
They supposedly either repaired or replaced the CF hull after that, I've heard both. Not sure how you'd repair it...
@@k53847 They didn't repair or replace the carbon fiber hull and were diving the next day, after minor repairs. The first cracking sounds in the aft of the submersible were reported by the crew, during a the dive after they lightning strike, when they were still in the Bahamas. They didn't even build a secondary prototype CF hull to put through appropriate pressure tests.
Did you see the leaked transcript yet? The Titan descended dangerously too fast, then he reported cracking sounds in the aft, and his real time hull monitoring (RTM system) alerts were all read. They dropped ballast, and even had to jettison the frame, but were still too heavy and ascending wayy too slow. Stockton also lost main power so had to route to a backup battery. Check it out, if you have not yet!
@@lunamaria1048 I saw a video on that transcript. Well, at least they saved money by having the UW interns design the sub electrical system.
This is an excellent explanation! This video definitely deserves more views!
Excellent job!!! I was amazed at the lack of concern for the popping noises they were hearing from day one in the sub. What the hell did they think was making those noises? Just plain stupid in my view 🤷♂️
Excellent explanation, thanks. I was thinking along these same lines. Repeated compression and relaxing of the huge pressures along the length of the cylinder with each dive had to be causing some flexing back and forth. Not hard for me to imagine that causing stress to the "glued" joints, culminating in popping loose an end bell. I can't believe anyone would go down to those huge pressures at that depth in a vessel being held together with a thin layer of epoxy. Mind boggling. I think even a few hundred feet would be a big gamble. Rush was just going cheap to keep his costs down to maximize profit. Even though he was an aeronautical engineer, I would think any engineer would have thought better of gluing metal to carbon fiber and thinking it could withstand that level of forces, and repeatedly.
I have been intrigued by this story. Had me looking through an old physics book and it made me realize I need to buy an old mechanics of materials book as well! (previous engineering undergrad, but not working in engineering)
I am rusty myself on some things. You just don’t use everything they teach you.
@@theeverydayengineer1 You provided a very nice explanation. Looking forward to hearing an eventual investigation conclusion but honestly I don't expect anything more than about what you said here. Maybe porthole. Isn't it wild though, that Mr. Rush thought that this was acceptable, to take that risk with no additional testing? That's part of why this is such an interesting (and macabre) story.
@@antontonableis it really that interesting? It’s just another tale is a rich guy getting away with things that a normal person wouldn’t be allowed to get away with, like not paying taxes. I was far more interested in why 500 people who died in the sea in Greece on the same day and why they didn’t get anywhere near as much press coverage, although the answer to both is obvious. They were poor.
@@StoutProper I'm interested due to the mechanics and engineering involved. A capsized vessel (albeit very terrible) is something that has been going on for thousands of years.
@@antontonable yeah, spend billions of taxpayers dollars on a rescue attempt they knew was pointless for 5 rich people but let 500 poor people die without any such efforts because it’s “boring”.
Funny that you weren’t interested in the Titan before people died, yet the engineering was all still there. You’re but really interested in the engineering, you’re just interested in the deaths of 5 rich people, and couldn’t give a fig about 500 poor people dying. People like you with a lack of moral fibre and without an ounce of human decency disgust me
Great explanation. This tragedy was a classic case of "can't see the forest for the trees". Always listen to critiques they are not in front of the tree!
It is important to see the forest but also to identify the tree... because naval and aerospace construction has changed... and still people keep talking about Space X that does not have the technology and you cannot go to Mars in a steel tank stainless...IT'S A FRAUD...The submarine should have been built with lasers with the new technology admitted in 2020...THE SUBMARINE SHOULD BE MANUFACTURED AS A ONE-PIECE ONE-PIECE MONOCOQUE HIGHLY RESISTANT TO SEA PRESSURE...
13:50
You are making a mistake here:
The contact is not made on the carbon fiber but almost only on the adhesion of the epoxy to the carbon fibers and the epoxy itself. It's almost only epoxy here.
Because of the winding method of the carbon fiber which is transversal. Delamination can therefore occur in the same direction as the shear stress you have presented. It's actually even worse than expected.
I had watched a video on what might be the text communication with the Titan and the surface support ship. There where read flags when the text says the Titan decent was too fast, and once they realized a problem such as a bus failed they could not come back up. They also mentioned that their monitors had gone from green to red as the composite was delaminating and crackling.
that alleged "last text comm from Titan" has never been judged authentic, or validated by anyone in the Oceangate crew or anyone who had contact with them right after the event. In fact, it actually mimics a Titan text comm from 2019, when they had trouble dropping ballast at the bottom, but did finally make it successfully back. Also it first appeared anonymously on Tik Tok, not ever a reputable source.
I think there's a major clue to what happened in the fact that the crew reported that they were ascending at a very slow rate, even after jettisoning ballast and the landing frame. Possibly, there was a lot of water infiltration into the weave layers of the composite hull segment, apparently from the aft dome area, and this water was rapidly changing the buoyancy of the submersible. Water getting into the composite layers would also then quickly weaken the structural integrity, leading to catastrophic implosion when it failed. Just my thoughts.
Agree. Along with that, I suspect that the structure was flexing inwards, thus reducing it's volume and thus, displacement. It's displacement being the only thing that gave it upward thrust.
Thank you for your simple and clear explanation of the possible (most likely) cause of the Titan failure. I have watched the news coverage of the debris recovery and it appears the Ti rings appear to be very clean, there seems to be no broken pieces of carbon fiber still glued to the ring? Its almost like the Ti rings popped off cleanly, along with the front hatch and porthole (unless the savage crew unbolted them during recovery process) Hopefully, the NSTB will release an interim report and followed up by detailed analysis of the Titan loss.
Splendid break down and examples! Helps us non engineers get it.
The more I watch, the more I can’t help but wonder about dives 1 and 2.
I do wonder will we get a final assessment that presents in a manner much like in this video, the degradation but spread across 3 dives..
This submersible successfully went down to these depths 2 other times. I almost forget that point when I hear and start to understand the views and commentary.
Maybe there are a crazy amount of interviews going on with those of the 1st and 2nd dives..
When I see that train car implode in that popular video playing everywhere now, it’s hard to believe this vessel had a life before that dive of 6/18.
It had something like 13 successful dives to the Titanic. The only other CF deep sub I know was did one dive (into challenger Deep) and then they condemned it because they had no real idea as to the fatigue life. This was all part of the original plan for that sub, not because of any problems they detected.
@@k53847 There is the "big disconnect". Was it two previous complete dives to Titanic's bottom depth or thirteen complete dives? I can't get a complete list of dive dates and bottom depths attained over the entire lifetime of Titan.
You are absolutely correct in your analogy, also they could have tried to mitigate the voids in the adhesive joint with the use of an autoclave or at least a vacuum bag, I saw no evidence of such measures during the construction of the hull, basically he knew nothing of strengths and weaknesses of CF. He could not of payed me enough to get in that thing.
The union with the titanium is not as you say. The fiber structure go complete into the molded titanium. It means that there is not a different thickness in the fiber border. However, it seem to be the same problem. When the fiber cilinder compress in the middle, the fiber in each union with the titanium will deform and it only depend on the epoxi.
Hello mate. Thanks for your video, I am actually a Civil Engineer and I am looking at it from a different way, instead of taking a cross-section which is a circle, if you take a longitudinal section of 1 inch thickness through the carbon fiber cylinder and the two ending titanium rings, you end up with sort of a beam of 2.6 meters in length and 1 inch in width and a depth of 5 inches (the thickness of the carbon fiber), the beam is now supported from both ends by the two titanium rings, a very simple structure with a uniform load from the top of 5600 psi, I believe the concentrated load in the middle will be massive and the carbon fiber depth needs to be far more than 5 inches, I also believe if Stockton put a third titanium ring in the middle between the two rings, but from inside only as a support to prevent the cylinder buckling in the middle point which is the point of max buckling, it would have made a massive difference in safety. Am I right? And sorry for the long comment. Cheers
I totally agree. I think he’s missing the point here talking about the pressure on the end caps, & he seems to forget that these force arrows on the materials are acting in all directions.
Every test I’ve seen (including a rudimentary scale model in a pressure vessel) has shown that it’s the deformation of the tube (inwards at the middle) that caused this, pretty much blowing the end caps off.
Yes I was also thinking a titanium ring in the center could have helped, I was also thinking that a few cross members from the end caps holding that ring in place could also help. Then you would have a sort of titanium frame that is wrapped in carbon fibre.
@@repawnd1 Indeed, it is like creating a titanium cage covered by a thick layer of carbon fiber
But we shouldn't also forget that the front acrylic window was only certified for 1300 meters, this could be also a very probable point of failure
@@AhmedBahgat True, then no matter how strong the rest was it was all pointless :/
I'm picturing Stockton Rush with an index finger in each ear, while shout-singing "LA-LA-LA-DEE-DAA-DEE-LA-LAAAAAAAA!"
Must've been a common sight.
Thanks for the refresher back to my statics classes thirty years ago. I still understand it but am glad I don't have to solve those problems again!
This is, by far, the best explanation I’ve seen of why this sub was death trap.
There seem to have been multiple reasons it was a death trap, not just one flaw
That loud popping they heard was that epoxy separating and or the carbon fiber separating from the load on the spheres pushing end to end. This thing should have been x-rayed or sonographed BEFORE and then after the first dive when they heard the loud popping, it would have shown what was failing by comparing a known good exam (BEFORE ANY DIVE).
They refused to examine the carbon fiber hull at all, they claimed it's impossible and saved money by not buying x-rays or sonographs.
I think that ideally you'd CT it before it does the first dive and then after each dive. "Not having your pressure hull implode ... priceless".
@@k53847 Yes sir, you need a known good exam before putting it in the water, then check later after each dive and compare to the known good exam.
Yeah, but if you X-ray the pressure vessel, you might have to cancel a dive that you accepted $1M for. They might make you give the money back!🙄
Very concise. I saw this on Reddit and commented on it there. Reddit has since removed the original post, for whatever reason.
Probably want to take a comment, act like it's their own & create 10 threads themselves. I don't doubt that whatsoever. People love to take thoughts & ideas from others & act like they came up with it.
Best explanation I've seen on this tragic occurrence. Thank you.
I have to admit, when you started drawing a circle to show how pressure applies to it I went "yeah yeah I know all this just get to the good stuff" but as soon as you used the 'keystone' analogy I was HOOKED! I wish I had teachers like you! 😊 I'm nobody and even MY first thoughts were that carbon fibre was designed to add tensile strength to incompressible epoxy so surely all he achieved was to give it the ability to delaminate.
All of his education was aerospace, with at most 0-1 bar more pressure inside than, and he admits he fell in love with carbon fibre so I guess he built a thicker 787 fuselage somehow forgetting that the pressure is on the outside not the inside and it's 400x greater... nobody is that stupid I know, but this IS a guy who only found out a freshly replaced or rebuilt thruster was operating in reverse when he got to 4000m.
I'm only certified to dive 55m but I already know that if my drysuit inflation fails I'll be cling-wrapped and literally incapable of ANY movement long before 40m which is 1/100th of the depth this thing was facing. We also stop and check each other's gear (no bubbles or potential snags etc) at 6m - apparently he checks his thrusters work when he hits 3800m 🤬
Death by narcissism 😥
1. The keystone beneficial shape is lost in the cylinder longitudinally, making it susceptible to buckling or shear inward, requiring much thicker wall than a sphere to avoid this problem. 2. A joint between dissimilar materials requires a thickened cross section, a flange - NOT a machined down narrower cross section as shown. 3. Given that carbon fiber is the strongest of the materials present, the glue is necessarily weaker, requiring a cross section several times larger than the cross sect of carbon composite; again a flange is required. 4. Carbon fibers are embedded in an epoxy resin, and therefore will fail at the resin, regardless of how strong is the carbon. 5. Using all titanium would have eliminated several problems: dissimilar deflection, delamination, weak compressive strength of fibers. 6. An ellipsoid shape rather than cylinder would have gotten better performance out of the same material and thickness. 7. The technology of bottling compressed gasses contains 5400 psi, 10,000 psi, and higher - so experimenting with human lives on new 'bottle' designs is completely unethical. They could have done multiple cycle testing in the lab on a miniature. -
Great explanation! The diagram and props, using laymans terms, assist in understanding the engineering, material issues, etc leading to the Titan disaster. Thanks!
The unverified transcript shows the crew heard a crackling sound at the aft bulkhead. This could well have been the sound of de-lamination.
I am not so sure I buy into that transcript. I see a few fishy things. It is very possible they may have heard cracking of the composite before it failed...if it was that and not the window.
Hi man. Minor point of clarification - It appears that the construction of the tube to the adapter pieces is that the titanium adapter is shaped around the full size of the carbon fiber tube - that is , there's no notching in the carbon tube; its the titanium ring cut to fit the tube. Sub Brief's 'The Titan Tragedy' shows a decent vid as they attach one of the rings to the tube.
Also, there's a vid floating around where Rush claims - although Ive not seen it - where the carbon fiber wrapping did do crossing as well as straight. Still, there's also pics of items being attached to the inner hull liner - which according to thier on YT vid from 6 years ago was 1.5" stainless steel -- with shop screws poking in. Yikes. Of interest to this, is how there was, at least that theyve shown, zero fiber brought up from the bottom. The viewport is also, apparently, MIA. Personally, id love to see high res pics of the titanium end caps and adapter rings - the fact the port is gone, a look at the attachment points from viewport to hull will be telling, as will whatever witness marks remain on the end caps and adapter rings.
End of the day, Using carbon fiber for a hull in this manner has been, well, crushed, permanently. Had Rush used a Titanium hull all the way through, odds are they'd of been fine - at least until the subpar viewport punched in.
From video I saw there is at least a notch on the outside of the carbon fiber hull as they are doing the glue up. I can see it with my own eyes. My hand drawing may not be to scale though. Even if there was no notch the shear stress at the joint would still apply since they didn’t taper the titanium step over down gradually.
@@theeverydayengineer1 Huh. Maybe we saw different vids then? The one i saw, quoted on SubBrief, the line i see appears to be glue.
Really though, it doesnt terribly matter if it was or not. From the second i first saw that sub, there was no way in hell i wouldve set foot in it - Hell i dont even like getting on damned Ski lift gondolas!
Anyway, new sub, will keep an eye out for ya!
Well-presented and produced, the best on YT.👍
It is the first video I watch that make totally sense, thank you.....
I'm an architect. Your basic analysis is spot-on. Your drawing / section of the flange to tube connection doesn't look quite right though. The Ti ring has what appears to be a 90 degreee 2" overlapping flange at both the interior and exterior carbon fiber tube interface, with a flat direct contact surface between them.
That aside, the likely failure mode was indeed loss of adhesion between the Ti end-cap rings and the CF tube when the CF tube deflected at it's center due to the enormous pressure it was (repeatedly) subjected to. The inherent defelction problem was compounded by what appears to be an idiotic failure to weave the fibers in a way that exploited their enormous tensile strength. Had the first 1/3 or so of the INTERIOR weave included a lot of courses PARALELL to the logitudinal (long) axis of the tube, they might have resisted the tensile stresses caused by deflection in it's center. Combined with a resin matrix reinforced with....uh....titanium or steel fibers / nanoparticulates or something with more compressive strength and ductility than whatever the hell they glued it all together with they might have been able to reduce the deflection of the tube.
The other problem appears to be insuffcient overlap of the Ti shoe-box style flange with the CF tube. There's an old saying in mechnical / civil engineering:
"If it don't look right, it probably isn't"
Even a lay person can see that the proportions of that ring and it's flanges "don't look right". I'd expect to see at least a 1:1 ratio between the 5" direct mating surface and the overlap between the Ti ring flanges and the CF tube. In other words those flanges should have been at least twice as deep.
At least. More would be better, and a two 4' sleeves glued with a slow thin slow-cure reinforced expoxy and mated at the center with another ring to fully enclose the CF tube would have provided MUCH higher strength, containment, reduncandancy and resistance to flexural failure of the adheisive bond between the Ti and CF. But then Numb Nuts Rush the innovative billionaire would have a boring old Ti submraine rienforced with a Carbon Fiber tube, instead of a pure CF submersible with a couple of Ti endcaps slapped on with some JB Weld.
Then there's the whole issue of repeately load cycling the assembly without any way to asesses fatigue performance short of destructive testing. This is exactly how Oceangate SHOULD have done thier R&D to establish a safe lifecycle for the sub:
1) Build 5 identical full scale subs
2) Take them to the Mariana Trench
3) Send 4 of them down (unmanned!) to twice the sevice depth you plan to use them at for paying customers.
4) Repeat this twice as many times as you plan to keep your final operational submersibles in service. Asssuming none implode, cut them up and asesses material performance. Redesign if anything looks questionable.
5) Send the 5th sub down until it implodes. Study the wreckage to identify ultimate failure mode.
After all of that, rate your sub for commercial sevice, again, with service depth and lifecycle 1/2 of what the design is proven to through direct testing. Yup - it's gonna cost a fortune, and the resulting operational submraine will cost twice as much as "necessary" from a purely engineering standpoint.
It also won't implode.
My cross section may not be to scale, but it was intended to have 2 flanges that fit over the composite tube (with some machining of a step in the composite based on videos I saw. I keep thinking how they could have made that seam better, and a low viscosity resin sucked in under vacuum could have been a possible better choice. Also the flanges on the titanium end cap should have been longer and had a taper to them to gradually change the force loading at the transition between the titanium and cap and the composite tube.
The bond joint is just one of a multiple possible failures. They may never know what really happened.
Yep, clearest explanation so far.
I'm less familiar with their design but I saw some images and saw the raised floor, as well as the wall in the back. From the supposed transcript of the event, it sounded like they heard cracking from the rear. It also sounded like they were increasing depth at a fast rate as well as having buoyancy issues after dropping weights. I'm wondering if the aft bulkhead would obscure a small pinhole/crack leak in the adhesive as you mentioned and let water in where it couldn't be seen, displacing their air and dropping buoyancy. Combine reduction of buoyancy and existing failure in progress and you get an event.
Very doubtful. I don’t think a pinhole type leak would happen, and then at a future point in time a catastrophic implosion. To me that transcript is a hoax. Too many flaws in it like someone is trying to make a juicy story. The submersible would not take on water like a ship to have buoyancy issues. I think the back wall was just a panel to cover up electronics and maybe some other support systems that may had been back there.
One of the better analyses of the problems with the adhesive joint where the titanium ring flanges were (sort of) adhered to the composite cylinder.
We could go into a several hours long discussion about surface preparation - or the lack thereof - in order to give the adhesive more "bight" to the titanium.
However, almost nobody has gone into any thought about the joint's lack of resistance to tear-out forces as designed.
Much as we use anchor bolts when attaching equipment to a reinforced concrete foundation, it would be smarter to run bolting rods tied into a reinforcing cage within the composite, such that the titanium ring flange could be positively held in place by the clamping force.
Such a design however would in my opinion not be sufficient to save the vessel, as the use of composite material brings with it the flaws of delamination and debonding with service cycles. And, CF composites in particular tend to break nearly instantaneously, not giving you enough warning to back away from the impending failure.
Very well explained, unlike others.
Great video. I'm familiar with how vacuum chambers are used to remove air pockets from epoxy even in hobby projects, should that have been done here? (ignoring that they would still have a huge problem with the CF). Also, Rush wasn't onboard every time. There was at least one time when it went down on an expedition with another staff member piloting it. That was the time they had installed one of the two horizontal thrusters on backwards and so the sub was spinning in circles on the sea floor. They had to remap the game controller while down there to fix it.
A comment below suggests he must have been aware of this. I’m not sure- if you suffer from hubris you kind of neglect reality. He overestimated his abilities. And he had achieved some remarkable success, so nobody could talk reason into him. But clearly the worst part : he took 4 victims with him.
This is an excellent video. Thanks for explaining it so clearly.
You are perfect to analyse the failure of Titan.
Being structural engineer I wonder why the Titan was not structurally analysed by competent authority before and tested throuhly three diamensionaly ?
Behaviour of composite material depends on strength/weakness of each other with strength of bonding material which is basics of composit material and it is widely used any many areas of engineering or manufacturing item.
Thank you so much for explaining this in simple terms. My dad’s an engineer and let me tell you when he explains stuff to me is like he’s speaking a different language 😅
Do you think them diving at a faster rate had a role in the preasure not equally being distributed?
Honestly, when I saw the clip of them mixing the ‘glue’ and the application I was floored. Glueing dissimilar materials is so hard. What were they thinking.
Thank for such interesting and profound, yet easy to understand explanation!
Liked, subscibed. A great explanation. I'd like to add that the weakening that's described due to the delamination, could allow compression of the tube. That would possibly be within the modulus of the inner metal tube and be unnoticed, but the reduced volume would result in lower displacement and thus buoyancy, (upward thrust).
Having watched almost all reputable tech channels (yours among them) in this matter I've come to the conclusion that they started diving the Titan with a 50-50% chance of survival from the get go , diminishing this chance after every dive to as low as 0% on the fateful day .... In my 27+ years of experience with composite materials in the military aviation field I've never expected to see such disregard not only to safety regulations and certifications , but a blatant disrespect to science itself ....
Best explanation I've seen! You would be a great teacher, if you aren't one.
Thanks. I used to go tutoring to engineering students, and I frequently do training for my employer. I try and use analogies to things we experience in everyday life. Basic engineer principles are not hard if you can visual them.
Defies logic to have titanium endcaps GLUED with a mechanical bond at best to a dissimilar material. Plus the glue, titanium, carbon composite, and acrylic window all react differently to pressure and temperature. There needs to be an in depth investigation or book written about Stockton Rush, billionaire killer.
Yes, the mix of materials was not a good idea. The acrylic windows have been the go to for submersibles since fay one, but he flirted with disaster on such a large diameter window.
Is it possible the glue could have contributed to the failure if the carbon fiber warped too much?
Very good analysis. Especially for the non-engineer.
Easy to say this now, but after forty plus years of building carbon composites, my initial reaction to seeing the pressure vessel, was complete disbelief. I just had to search my memory for the number of times I'd engineered a compressive structure. And all i could come up with were three. And those required extensive omni-directional laminates. I'm still at a loss as to how this was even considered. And even if attempted, how both vacuum and spectral inspection was not in the mix.
Thanks for the video. Keep up the good work.
Thanks. I tried to hit the main points for the average person and not make a crazy long video trying to break down every little technical anomaly. I do not have the expertise to say much about use of composites in compression. I am actually curious about studies using carbon fiber in such a manner.
@@theeverydayengineer1 You did fine. Always effective to use simple practical examples to get more complex concepts across.
The reality is, composites, regardless of whether they're carbon, kevlar, etc., are rarely used in compression unless they are part of, or complement, a more complex construction of other materials. That's why, for instance, composite wing structures, even on light aircraft, will incorporate substructures to distribute compressive loads, while allowing tension loads to be somewhat independent through a limited vector.
From my perspective, the real issue with the Titan's use of this material and process, is that we go into composites knowing that the only way to validate integrity, is by use of pre-pregs under vacuum. And then, non-destructive testing. With metals, we can control the matrix. With composites, that matrix is only known or trusted, on the material level. Not the construction level.
Thanks again. Keep it up.
Very reasonable idea about water seeping into the glue joint and then destroying layers of carbon fiber. This explains the newly appeared text messages exchange with the submarine, where the crew heard "sounds" from the aft and was alerted for some minutes before contact had been lost, i.e. the destruction was not instantaneous.
The text messages have not been authenticated yet.
The carbon fiber/epoxy hull of the Titan features alternating layers of UD prepreg in the axial direction, and wet-wound carbon fiber filaments in the hoop direction.
So its not just in one direction as many self proclaimed internet expert are saying.
Also the titanium end caps are held in place by the pressure itself when under water. The glue are just there to stop it from falling off when not under pressure.
Its not an area of concern.
Thank you for the information on the hull. I was certain it had to have axial fibers to hand bending stresses on the surface and during loading and unloading. Interesting that they mixed pre-preg and wet-winding.
Yes, the titanium ends would be kept in place by the water pressure along under water. The adhesive would only serve a purpose on the surface.
It most definitely is an area of concern, because the groove is complete responsible for transferring the end cap loads to the composite shell. The epoxy has a strong influence on this load introduction.
@@marconawijn405 You can use anal lube instead of epoxy glue and as long as it stay in place until you start diving, it will be perfectly fine as the immense pressure press the titanium ring against the hull of the sub.
World's best and first correct analysis available anywhere.
Let me add a little quantitative analysis (which is so easy btw that it can be done in five minutes). Compressive stress in the carbon fiber polymer (CFRP) wall of the Titan submersible was about 200 newtons per millimeter squared at Titanic depth. This corresponds precisely to the stress level where CFRP typically fails under compression. Only under tensile load (which is not applicable underwater), CFRP can withstand about 10 times more. What's more, OceanGate used unidirectional (circumferential) CFRP, as you have also stated. This means that in the longitudinal direction, the hull was only about half as strong, at around 100 newtons per millimeter squared. It's a miracle the ship even made it to Titanic a few times without imploding. It was like doing pole vault, but with a rope as the pole.
With all fibers in the hoop direction - the strength in the axial direction is about 5% as compared to the hoop direction. And - in compression the strength of CFRP is about 60% that of tensile strength - not 10x less.
@@les8489
Tensile strength 900 - 2,500 N/mm²
Compressive strength longitudinal 120 - 420 N/mm²
Compressive strength transverse 60 - 150 N/mm²
Roughly factor 10 and 2, respectively
@@DavidG2P Tensile strength 900-2500 MPa - OK, although 900 MPa would be for plain-weave fabric rather than tape. The 900 MPa for tape would be an extremely low number (read: shitty tape which belongs in a scrap bin).
Compressive strength 120-420 MPa - incorrect. For a tape with 2500 MPa tensile strength, the strength in compression should be 1300-1500 MPa. If you are getting 12-420 MPa - this means that the test is incorrect (like very long non-stabilized test coupons which fail essentially in buckling). 60% of tensile is a rule of thumb value. These are the results I have been getting on a number of carbon fiber materials in the last 30+ years. Fibers like (somewhat old but still widely used) T300 fibers reliably produce values of about 2100 MPa in tension and 1300 MPa in compression, on coupons with 35% resin content, tested to ASTM standards.
Compressive strength transverse: 60-150 MPa is OK, as this is consistent with strength of resin in compression - although 150 MPa would call for a rather fancy resin.
We are talking here about the failure stress - so, if you can - please indicate a source of "Compressive strength longitudinal 120 - 420 N/mm²"
@@les8489 the numbers come from CFRP companies named PRK and Roechling.
@@DavidG2P If so - then these numbers are either incorrect, or refer to something else, or it simply is a typo. I have spent the last 40 years designing, analyzing and testing composite structures (aircraft).
Very good analysis. Thanks
So I assume that once water ingress occurred, potentially in the region of the glue layer, the pressure of the water at depth would actually work to push the titanium rings away from the composite hull axially and vice versa causing it to decouple. This seems more likely than the sheer at the interface.
Nicely done! Very good use of diagrams and props. Your example poking a small hole in a sheet of aluminum foil was brilliant! One minor correction, I believe you are using seabed level crush depth pressures in your calculations. The sub appears to have imploded at less than half the total depth of the anticipated dive.
Since the loss of text and location comms occurred 1.5- 1.75 hours into a dive that typically was 2 hours to seabed, they were probably much farther down than less than halfway, given that implosion would likely occur within minutes of their primitive hull breach warning system. Anyway, investigators will probably be able to make some estimate of their depth, based on the distribution of the wreckage, and the contents of the REAL text communications and locator documentation of the last 2 hours of titan. (The alleged "last text comm" originated from an anonymous Tik Tok poster, which has never been a reputable source for truth.) Very reasonable to use pressures at seabed as an example, when outcomes would be similar even at lesser depths.
Just realised that the fracture line in the titanium to carbon Junction was only held by resin and no fibre. The fibre was only wound in one direction around the circumference, hence parallel to the stress line. Your crack lines of delimitation should be vertical, not accross that junction horizontally.
I still think there was horizontal layers mixed in to support all the bending stress of holding the two titanium end caps on. I can’t prove it, nor disprove it, however. Also the cracks would form in the resin between the layers of fiber, hence the horizontal fracturing.
@@theeverydayengineer1I found an article in a 2017 trade magazine dedicated to composites where the company that designed and manufactured the hull of the Titan (then called Cyclops 2) talks about it.
I'm not sure if I can post urls in comments so I'll just quote the relevant passage:
"The biggest challenge, Spencer reports, was developing a manufacturable design that “would produce a consistent part with no wrinkles, voids or delaminations.” And without use of an autoclave. Spencer opted for a layup strategy that combines alternating placement of prepreg carbon fiber/epoxy unidirectional fabrics in the axial direction, with wet winding of carbon fiber/epoxy in the hoop direction, for a total of 480 plies."
You forgot to discuss the matters of the epoxi matrix, the lack of
pressure treating in an autoclave, and the different and opposing
contraction and expansion temperature coefficients of the carbon
fiber cylinder and of the titanium caps. The latter contract when
temperature drops, while the carbon fiber matrix expands.
Otherwise this is an excellent presentation. Thanks!
Thanks. Yes, there is soooo much to potentially talk about, but I tried to keep it as short as possible and get to likely the most important issue with the design for the average person to understand.
Very easy to understand video!
This is an excellent video. Thank you.
I’ve done 3D printing where the layers were perpendicular to the direction of force, and the layers separated cleanly. If the submersible was rolled in the loop fashion, and then a section was under a sheer force like you described at 13:52, could the machined mating point on the carbon fiber ring snap as the layers are perpendicular to the direction of force?
It would be more likely, but it also looks like they laid down a pretty wide fiber that would have been woven providing support in the perpendicular direction, but it would be not nearly the same as actually putting in fibers in helical or horizontal direction.
I was kind of waiting for someone to explain that carbon fiber composite, when fully cooked, became very good at... tension, not compression. Compression was just trying to spread out fibers and separate them from each other, not using any of their molecular strengths. When you only count on the surrounding cooked resin bond to do the job of resisting to those multi-million pounds of pressure, this is basically what happens.
Would the fact that it was salt water possibly seeping in make it worse? Salt is corrosive, plus when the water evaporates, salt crystals form which could physically force layers apart.
Add in temperature, coefficient of expansion, and every mechanical joint.
Coming from a race car background I’m aware of testing metallic components for stress fractures such as Magnaflux inspection, however I am unaware of such a process for carbon fiber, I assume there is SOME process but wonder if the part size make the process restrictive for a dive hull.
There is ultrasonic testing to test for delaminations and voids. There are very big machines that can do this. But that might only be practical when the tube was initially built, and would not be practical after the vessel was assembled.
Why did'nt you use the surface of the hemisphere to calculate the force "squeezing" on the ends of the CF tubular body?
Because the net force squeezing in is equal to the area of the circle since some of the force is being directed to try and pinch the hemisphere in the perpendicular direction.
@@theeverydayengineer1 I will leave this without comment...just don't get involved in design of any pressure vessels.
I wonder whether the cold water could effect the glue and make it brittle and unflexible. It going from the surface temperature to 12k feet down must have some effect, even if they descend slowly.
The water temperature is not all that much below freezing at that depth. That isn't all that cold to likely create a brittle fracture environment. But it is something that could be a possibility depending on the exact material used.
Excellent explanation with no nonsense
Question: I get the keystone kind of trying to keep the pressure equal, on the cylinder shape, instead of you getting the keystone shape 360 degrees, you only technically have it in one direction and less so as you kind of turn it, until once you get to 90 degrees it has lost the keystone-like protection, would this also drastically weaken its ability to resist being crushed, or would the few degrees of keystone protection (for a lack of better choice of words) offer sufficient resistance to compression?
You are correct. The "keystone" in a sphere necks down on all four sides from the outside to the inside, while in the cylinder it only necks down on 2 sides. The two sides are sufficient in a cylinder. Making the wall a little bit thicker can help regain any compressive strength that may be lost.
@@theeverydayengineer1 thanks
What about force's aground the OD of the Carbon Fiber tube which cause it to bend inward and snaping the glue joint.
11:30 this is exactly what i think, too. Especially when i see the viscosity of that glue.
Have you considered, (disregarding the issues of the end caps etc.) with a hoop wrapped Carbon Fibre with no end to end fibre reinforcement that when the pressure in the centre of the cylinder pushes the cylinder in at the middle, the INNER cylinder wall will be under extreme tension as it is being STRETCHED by the deflection of the five inch thick wall section. This tension on the inner wall would then lead to the resin between the fibres cracking and propogating around the cylinder to the outside wall in a perfect radial crack in the centre of the cylinder. The hull is effectively ONLY held together from these tensional forces by the epoxy resin itself, so a CAD model of these forces should use resin only without reinforcement to see how the resin cracks under tension.
Yes. I tried to keep the video as short as possible and go with my gut feel on the windings. Some commenters said indeed there was horizontal and/or helical layers. I figure there would have to be. I plan to do another video on basics of composites.
Very good explanation, best so far! Thanks
Glad it was helpful!
Very informative, I feel you are spot on with your analysis, with the leaked messages from ship to Titan, just before the messages ended Stocton Rush said he could here crakkling sound in the aft, I bet it was delimiting in the rear section between the Titanium and glue and Carbon fiber.
Those messages have not been confirmed to be authentic. As one commenter said, why would Oceangate first leak the messages to a TH-camr in Pakistan?
I have a question: wouldn't the radial compressive force on the cylindrical faces of the hull have a sort of stiffening effect on the composite fiber plies that would counteract delamination due to the compressive forces of the two end domes?
To some degree I would agree with that. It is a complex thing to think about.
@@theeverydayengineer1 I also don't really understand why the joint was made with such abrupt changes in section, groove in the ring and tooth in the cylinder. It should be general knowledge that such a thing will create stress concentrations and so on... They really seem to have taken a tourist's approach to the project.
@@ottaviocolombo5805 Yes...how they did not have a long smooth reducing taper on the titanium ring that slides over the composite is wild.
@@theeverydayengineer1 I was wondering: if the taper on the titanium ring was on the inner side, the pressure on the end caps would make them work as a wedge, opening the end of the cylinder. Combining this with the radial pressure acting directly on the carbon fiber cylinder, would this amount to clamping radially the cylinder, avoiding the crushing of the fibers? Combined with the taper it may have improved the life of the structure.
With the way that the tube was laminated, the stress at the titanium/tube glue joint perpendicular to the tube means that the strength of the tube in compression is just the shear strength of the resin. This is also true at the half length of the tube where compression deformation of the tube would be putting tube outside length wise compression stresses and inside tension stresses on the weakest material, the resin.
You have provided the most reasonable cause and effect analysis of what could have led to the implosion. My observations watching how they manually MIXED the glue instead having it mixed in a controlled environment. The application of the adhesive was also manually applied and was not uniform. My question is what were the engineers on staff not doing the due process to determine that the Titan was not not structurally sound. I believe they also share responsibility as engineers in whatever judicial actions are given to the company! Where were the safety and quality folks? Probably helping mixing the peanut butter that HUBRIS Stockton was so proud of. I am not an engineer and had the same observations as you did. Thank you! I have subscribed to your channel!
Best news is that 1) a licensing body for these vehicles may be created from this event and 2) their deaths happened faster than the human nervous system can transmit information.
I watched a number of videos where "engineering analysis" of possible causes of Titan failure have been presented.
I am an engineer with 40 years of experience designing and analyzing composite structures - and composite is about the last material I would pick for this application.
Having said that - it is obvious to me that vast majority of these presentations are made by people who have no idea or understanding of composites.
Yes - the choice of material was incorrect. Yes - moisture ingress into the matrix, manufacturing flaws, porosity, bonded joint failure, difference in thermal expansion, difference in stiffness (titanium caps vs composite cylinder), cyclic fatigue, impact damage, microcracks, application of cylindrical shape etc. - all these COULD be contributing factors.
What is astonishing - is how little attention is paid to the LAYUP - that is, orientation and placing of the fibers.
Pressure at 4000m is about p=40 MPa.
With 4.7ft diameter and 5" wall thickness - the ratio of R/t is about 5.6.
Hoop stress is about Sr=p*R/t = 220 MPa
Axial stress is about Sa=p*R/2t = 110 MPa
Carbon tape in tension fails at about 2100 MPa, and about 1200 MPa in compression (dry, RT).
But in TRANSVERSE direction (perpendicular to the fibers) it can only take about 80 to 180 MPa in compression - as this is limited by the strength of the matrix alone.
Compare axial stress and the strength of epoxy matrix...
If the cylinder was manufactured by winding ALL fibers in the hoop direction (which I still can't believe - because this would be total idiocy) - then the matrix would be stressed almost to its ultimate stress in repeated dives.
Most of the comments and analyses concentrate on "wrong material" (yes, for this application), number of "Empire State buildings pressing on Titan" (BS), inability of carbon fibers to take compression (BS) etc. - while totally missing the basics of composite design and analysis.
If there really were no fibers in the axial direction - then this was the most probable reason for failure.
Question: I've never worked with carbon fiber composite, but I know from experience with insulated aerial lift booms, that glass/resin composite is highly hygroscopic if fibers are exposed. Is the same true for carbon fiber composite?
The fibers themselves have hydroscopic properties, but that won’t do much to prevent a tiny crack from being split open by hydraulic pressure.
@@theeverydayengineer1 I was musing more along the lines that since water seems to accelerate delamination in FRP at normal atmospheric pressure, I can only imagine that any surface fissure, subjected to water at 300+ atmospheres, would also speed delamination.
@@kendalgoodson ah. Yes. Any surface fissure with that pressure can’t be good.
Well explained. As an Engineer you know that we have to wait for the full investigation report, other systems such as the porthole were also contentious, so we don't know the exact probable cause but I am with you in the theory that this is the most likely scenario. (I am a BEng CEng Automotive).
It looked like the window was blown out when it imploded. I can't imagine they would remove it just to create a hoist point, every investigator would freak out.
@@k53847 yes, but it could well have failed due to the immense forces which would have been generated at the point of implosion, it will be interesting to find out.
Some comments from an experienced composites engineer.
1. Like you I was very concerned by the appearance of only hoop winding in the shell, though I have to wonder how the vessel survived even it's first dive if this is the case.
2. You describe what is termed a 0-90 layup. There are also helical winding techniques where the fibers run at an angle.
3. Carbon fibers are very stiff. They are not limp. Those who work with them often get stabbed by an individual fiber. They really hurt. This is in contrast to Kevlar fibers which are quite strong but very flexible. Properly designed and bonded carbon fiber structures can withstand compressive loads quite well. One common example is the top of helicopter blades which are under compressive forces.
4. The joints between dissimilar materials terrify me. Thermal expansion differences. Differing Young's moduluses when stressed. The sharp 90 degree corner on the titanium flange ring is a tremendous stress riser almost certain to cause problems. I have had to design mounting of steel rings on rtp structures. The best approach is generally to provide some sort of cushion between the materials.
5. In the photos, the area of the glue line attaching the flange appears to be much too small. Socket joints on fiberglass piping less than a foot in diameter are generally deeper.
6. Components for sizable pressure vessels are generally not cast. Casting issues can cause many problems with grain structure. Shells are rolled from plate. Heads are spun from plate. Smaller items may be machined from rolled stock or forged.
7. While you are correct about the advantages of spherical shell, a cylindrical shell can also be used. A cylinder will be heavier for the internal volume, but the navies of the world use huge cylindrical shells for their submarines. These large cylinders are reported to be designed for as deep as 3000 feet which is just as much a design challenge for a large cylinder and 20000 feet for a small cylinder.
I almost mentioned helical winds but I wanted to keep it simple for the average person. I have worked at times with tubes with helical and hoop winds. It looks like I cut out a small section during editing that stated composites under compression are not necessarily a bad thing, just that additional considerations to factor in. The titanium end cap should have tapered the flange to almost nothing over several inches to transition the stress load. I tried to keep the video short enough to keep attention span. I was shocked I was still over 20 minutes.
"One common example is the top of helicopter blades which are under compressive forces." Let's not forget about centrifugal forces...A helicopter rotor forms a cone with tension on the blades - this is what keeps it in the air. There is in fact very little compression if any. But on the wing of an aircraft - yes.
Generally you are right elsewhere :)
Sounds like basic design was fundamentally flawed. There seems to be many mods that could make it stronger. For example, 1. Considering tube thickness. 2. Using strong backs which buttess the tube. 3. Using a taper shape for the matiing surfaces of the end tube. Just like the mandrel in a milling machine where the harder you press, the tighter the fit when trying to machine steel at 3000 rp
m and you don't want it to come loose
Super informative video, well done.
Can we assume Titan was taking on water in unseen areas because of the rapid descent and the difficulty in climbing to the surface once mission was cancelled? If yes, then this would indicate a failure much earlier than any of the monitoring systems detected?
I think that is completely bogus. Where would it take on water? It would have to take on a significant amount, that would have had to be internal to the pressure vessel, and they almost surely would have heard it coming in. That transcript has too many fishy things in it in my mind. Someone tried too hard to embellish the story in my opinion.
As an "aeronautical engineer", Stockton Rush was out of his element. The bad choice of the cylinder to extend passenger capacity was a commercial decision. Business determination can not over ride the laws of physics. If the cylinder war deformed by the water pressure the glued seams could separate. Stockton Rush's most notable quality was that he was a was persuasive salesman, but he was not a scientist. .
If i’m not mistaking - what they attempted, was a purely layered all-conposite tube, between the titanium half-dome ends. from an engineering standpoint, i feel that if they had: 1 put segments of titanium rings also inside the laminate, and 2. titanium rods - connecting the two end-half spheres, lenghtwise along the lateral tube, then they would’ve solved most issues of compression, while maintaining a weight advantage of having most of the cylinder composed of composite. aside from that, i don’t see the argument for weight savings being relevant in a submarine- on a plane, yes, but in a sub, where tou wish to decrease buyancy - not really
You could do something like that with the rings and the rods, but that would make the internal cabin smaller, which isn't good for making a sub for tourism. Yes, trying to have weight savings in mind for a submersible is a little overkill.
Spot on, great analysis.
Thanks.
Wow! It is a brilliant explanation, basically the sub was an amateur design
Thanks.
Youve heard about the 17 of the 18 bolts being torqued?
Top bolt was too far away.
Great explanation. Thank you. It looks like Stockton Rush was trying to create new physics ( well - it just doesn't work like that) it was stupidity but not innovation
Also the plastic binder's specific gravity increases at a far greater rate than carbon fibre under increasing pressure causing laminations due to the binders plastically reduced volume seting up rapid fatigue failure under repeated loading/trips.
NOTE: Boeing engineeers should have condemned the carbon fibre for destruction once it passed it's "use-by" date?