From someone who works around advanced composites in aviation. Carbon fiber composites are strong in their narrow operation limits. Outside of those, stuff is like paper. Additionally, it requires inspections. If not done.... You have a mess just like this.

I have a family member who holds a PhD physics that has 40 years of experience in carbon fiber development and production. Having designed and built several carbon fiber production facilities all over the globe, I trust his expertise. He told me there was no way he would have ever even gone diving in the deep end of a swimming pool in that "contraption," as he called it, much less down to 12,000 feet. He knew right away based on the construction and lack of attention to detail and safety of the operation that they suffered a catastrophic implosion.

The primary problem with Titan was the temperament of it's creator. Too many corners cut for limited success.

I remember seeing the edited videos of the build process, and what stood out to me as a non-expert was the process of bonding the ends to the CF tube; the glue was hand-applied (and the surface apparently not cleaned beforehand), bare hands were touching the CF during the application process, it was not done in any sort of clean-room (lots of dirt and dust in the air), and there was exactly no squeeze-out when each sphere was slid onto the tube, meaning there were almost certainly voids.

The Titan is really compact now.

After seeing the accommodation, I can honestly say that there isn't enough money or any other temptation humans usually respond on in order to get me into such a ridiculous excuse of a submersible sardine can. Absolutely nuts.

Mike, I don't often comment on content I come across online, however I believe you have provided the clearest, most concise yet suitably detailed and realistic summary of what may have happened, with the available information and a solid understanding of materials and processes. It goes without saying that there could easily be a lot more maths and physics to quantify but then this would become a lesson in engineering rather than a clear relatable clip. The perfect balance, and your included animation is second to none. Well done sir!

I'm a layperson without any engineering knowledge, but I like what I'm hearing in this video. It sounds reasoned and I like that credit is given where due, caveats are clearly stated and safety and reliability is upheld beyond all else. I appreciate the expertise and knowledge and background work that has clearly gone into making this video - thank you.

Ive worked with Carbon fiber based competition race cars......it is a fantastic material given its strength and rigidity for the weight. However, every and any material must be tested exhaustively to make sure it is suitable for an application and not just "shell be right mate". Same goes for the attachment methods of the end domes......testing is needed.

The fact that the flanges at the end failed outwards is a clue to the failure mode of the pressure vessel. I don't agree that it would shatter uniformly like glass. The more likely initial failure is that one part of the cylinder became weaker than the rest and was forced towards the centre by the pressure - this compression on one side would (like if you were crushing a toilet roll) force other sections of the cylinder outwards for a fraction of time before complete failure occurs. That's what ripped the flange apart.

I think one of the main issues was the repeated stress on the Carbon fiber. It can only be subjected to these stresses for so many cycles. Also, you have to take into consideration of the four different materials. The carbon fiber, Titanium, the adhesive used and the view port.

Very well done. As an engineer, I appreciate someone who has the knowledge to present this topic. Thank you.

Working on the development of carbon fibre wide chord aero engine fan blades back in the late 60's, I have some knowledge of the processes and characteristics associated with CF. The development program for the fan blades was the result of the need for a light weight structure, capable of resisting the centrifugal forces on the blade at the tip, as well as absorbing the torquing stresses due to loading. Carbon fibre composite was the only material that met those criteria at the time. unfortunately, it failed the blade impact test, also known as the bird strike test, as mandated by the USA FAA, a nine pound bird fired into the engine running at full throttle, (it passed the UK CAA bird strike test using a seven pound bird). The carbon fibre delaminated, shredded! Years later, General Electric in the US adopted CF for the engines supplied to Boeing on the 777. They experienced the same problem, delamination until they came up with a design which involved three dimensional weaving, with the lay up having cross filaments incorporated in the weave specifically to prevent the delamination, apparently a sixteen hour layup cycle! If the Titan CF centre section had been constructed on the same lines, maybe this could have resulted in a structure capable of withstanding the enormous external pressure. BTW, I dispute carbon fibre composite has good compressional strength, a function of the matrix that binds it together, glue.

The big difference between composite and metals is metals can compress at a molecular level under pressure then expend back to normal when the pressure is removed. It can do this repeatedly and engineers can work out how many times it can safely do this. Composites on the other hand don't compress well because it's not a homogenous material. The carbon fibres and resin have different properties and on top of that, finished carbon fibre is not homogenous like Titanium or Steel. It will have differences throughout the material. Some parts may have a bit more resin, some less, some fibres might be closely aligned while others a little bit further apart. That doesn't make much difference in normal applications but 4KM down in the ocean under those sorts of pressure, small variations can have enormous consequences for the strength of the vessel as a whole. It was reported that the carbon fibers’ axial compressive strength is only 10%-60% of their tensile strength and their transverse compressive strength is 12%-20% of axial compressive strength. Titanium on the other hand, it's compressive strength is more than twice it's tensile strength. Stockton Rush was a snake oil salesman and he knew that's what he was. He knew he was taking a massive risk with his sub. The structure of his company illustrated this. The holding company was registered in the USA while the company that operated the sub and rented out the seats was based in the Bahamas, thereby avoiding US regulations.

Thanks for a comprehensive and accurate depiction of the flaws built into this vessel. When carbon fiber first came out, we (who worked in composits) were all taught that the 60/30 angle of layup would provide a vastly superior strength. When I first stated seeing similar reports on the tragedy, I figured it had to be a massive error in the laying up of the fiber. You illustrate this perfectly. I also thought his warning system had to be little more than wishful thinking. Lastly, I'm stunned he never utilized NDT like ultra sound. This was a programmed accident. Pride & confidence are not equal to good old-fashioned NDT & destructive testing. Thanks for such a good report. I subscribed. 👍🏻

I always appreciate your ability to take such complex concepts and communicate them visually and verbally in a way non engineers can grasp these very complex concepts in a more tangible and meaningful way. Thank you for your hard work. I always learn so many new concepts to learn more about.

Good analysis, another thing to point out here you didn’t mention. The carbon fiber tube(both models), were constructed in a warehouse, not a hermetic environment free from dust and other airborne particles that could’ve been introduced into the in between layers of the cabon fiber as it was being weaved. Also Ocean Gate removed a lot of their videos showing these construction processes(likely to hide evidence), and you can actually see air bubbles and imperfections in the layering process of the finished carbon fiber tube. Also a former engineer told Stockton to make the fiber tube at least 8 inches thick to ensure it was strong enough to withstand the ocean pressure at depth, Stockton went against this and it was only 5 inches thick. So I think imperfections in the tube hull combined with it not being thick enough as well as the flanges were weakening with each subsequent dive, and maybe water was beginning to ingress into the layers of the carbon fiber or finally penetrated the flanges and the glue and it was just enough to cause the catastrophic implosion. That’s my take on it anyways given what’s available.

Best explanation video i've seen about the titan sub. When they did put the first Titan sub model into the pressure chamber the cf started failing at the titanium dome at that glued joints. Thats why the cf hull was ribbeled appart in axial direction and also tearing appart the small titanium overlap parts. And i very much believe the exact same axial failiure of the cf hull at the glued titatium rings happened again, as it already happened with the test titan in that pressure chamber. There are photos on the internet about that ribbeled-off cf hull. Sorry for bad english. 👯

A couple of things, while this Carbon Fiber pressure vessel was being testing out in the Bahamas one day, it was struck by Lightning at least once. Pre-preg Carbon Fiber holds up VERY POORLY to Lightning Strikes as it usually causes serious Delamination's. Unless OceanGate incorporated some proper and well-placed bonding to the entire assembly, and within the CF material itself, there was likely Damages caused by that strike. Second, Anytime that you're bonding dissimilar materials with Epoxy, like Carbon Fiber to Titanium, you must be absolutely certain there are NO oils or other contaminants on either surface before bonding. Watch the released video of the OceanGate submersible assembly where the technician is wiping a Titanium mating ring w/ a rag (presumably with some acetone) and then handling that same area with his ungloved other hand ? ? ? Any good NDT Technician (non-destructive testing), or proper X-Ray composite viewer could have found any imperfections, delam's, distortions or defects within the materials used. It's Insane that they fired the manager who pushed for those NDT checks and they drove on so blindly without that assurance. Edited for spelling and Lightning Strike location.

I used to haul gravel on barges. When he said that the round port hole had 2700 tons pressure on it I imeaditly thought of a barge loaded with that weight. It would fill the deck area 15 feet high on a barge that is about 180 feet by 45 feet wide. That puts it in perspective.

I had not realized just how cramped it was on the inside, makes perfect sense. But somehow I was picturing something a bit less claustrophobic and horrible. Great video!

Had me think it was the scrubber that was the culprit for a while there.

Not getting on a sub with Mike, he didn’t learn the lesson about carbon fiber… The carbon fiber used was a fabric but what is holding that fabric rigid? It’s a resin, the only thing holding the layers of fabric together is that resin. Even under tension the resin will break down which is why the Boeing 787’s fuselage has a life cycle of 44,000 pressurized flights. Everyone who previously went on the sub discussed vary loud cracking sounds, which is a clear sign of delamination between the layers of fabric. If the sub had gone down and stayed down, it would probably still be intact. The issue is going down and coming back up, essentially the equivalent of a flight cycle. Each time it compresses and contracts, the cracks are getting bigger, and more frequent. Rush knew this but still neglected to do an x-ray because he knew it would be the end of Ocean gate, why not go out with a bag. Congratulations Rush you made it into history, I’m sure you’ll be in lots of engineering textbooks for decades to come.

Great piece of work. I worked in endurance testing high pressure equipment up to around 3000 Bar. This sub failed at around 400 Bar. The failures at these pressures progress through the structure as shock wave rather similar to an explosive event at around the speed of sound for that material. The stress is already in the carbon fiber and you get a cascade failure as the local support is removed by the failing material. I have no data for the speed of sound in carbon fiber in this configuration and pressure but it will around 3000 meters/sec per second much faster than the speed of sound in water at 1450 meters/sec. This means the tube may well failing faster than the water can keep up with . This would cause a huge mass of water accelerate conically inward ( the open end toward the direction of travel) in the direction of the end that did not fail first. As the in rushing water ( many tons) is at really high speed ( 150 mph +) when it collides in the center it creates an even faster flow towards the remaining end . When this hits the end you get the mother of all water hammer events blowing out the window. The violence of this event is difficult to comprehend. This explains why the window was not with the sub. The blown outward flange was on this end too so I think this could be included in this violent event. The whole thing would look a bit look like the copper in a shaped charge anti tank round going off. All this happened in less four thousandths of a second the victims could not have sensed any of this thank goodness.

Thank you for this video! It's the only video or article I've seen that actually attempted to show the Titan components, what was on board, what was in the tail cone, etc.

Finally a well researched video that doesn't zero in on the game controller (which was probably the most reliable part of the sub). The truth is that composite construction is commonly used in modern subermersibles. But usually fiberglass instead of CF and rated to shallow depths with a factor of safety of 10+ for both crush pressure and fatigue life, for the very reasons mentioned in this video--this construction is very sensitive to defects and it's hard to both validate the design AND the process without a large sample size to build confidence. And you certainly don't use the sound of fibers snapping to decide it's time to go back up.

Very interesting video. I like your take on this subject. 13:28 Yes titanium and carbon are bonded in aviation, but aviation is not a good analog for deep see usage. Carbon fibre composites compress much more than titanium. The haul would simply shrunk way way more that titanium faring under the deep see pressure. That isn't as big of a problem in aviation, where pressure differences are magnitudes smaller. This was a crucial potential failure point, but there was no proper examination or testing of this bond. Not to mention that they didn't even glue it in a proper clean and temperature controlled environment. Comparing Titan haul to an airplane wing is also not the best. Those wings are baked under pressure in autoclaves during the drying stage, which reduces the chance of any dry spots, bubbles and other imperfections. They are also thin enough that they can be properly examined (x-rayed) and tested. Nothing like that was done for Titan. I like your theory about axial fibbers, it's another example of how many corners did they cut and how botched the whole build was. I agree with your conclusions. Could you build a carbon fiber haul for submarine? Probably, but you need to go above and beyond with research, quality control and testing. Those guys didn't do you it, they were only interested in cutting costs.

Stockton Rush spent more effort on avoiding regulations than he did making sure his craft was safe for passengers. They say hubris is the fallacy of genius.

He did not want ultrasound testing on the hull.

The video of them attaching the titanium to the carbon fibre with epoxy smeared on is what gets me. No way to test those resin seals.

When this happened the first thought that came to mind was the difference in material compression between titanium and carbon composite. Primarily as how it would effect the glue joints and internal/external flanges where the 3 materials come together to form a bond/seal. The other thing is how thin and short the flanges were in comparison to the thickness of the material they were binding...

Great video. I hadn't looked beyond the headlines until I saw this video. I'm glad I found this first.

As an engineer and once scuba diver, I’ve always had issue with MSM assumptions about this loss. Starting with nonstop criticism of wireless game controllers. Any and all propulsion controllers need robust design and redundancy, no matter the specifics of their design. These controllers may have been perfect for the application. MSM automatically demonized this. As for carbon, its simply not just the material. Its was likely the lack of truly thorough testing and finite element analysis, destructive testing, ultrasound and independent third party testing and review. Not simply the “carbon”.

Your videos are always awesome. It was reported that the US Coast Guard has recovered the aft titanium dome and other parts of Titan, not previously recovered at the time of the incident.

We have repeated pressure cycling of a plastic cylinder glued to metal ends, and repeated temperature cycling of materials with dissimilar thermal expansion coefficients. Then we have compression forces at each end, pushing against the cyclinder ends. Then we have repeated pressure cycling of the plastic cylinder. Just take a look at the other submersibles that are capable of descending to the Titanic, they have thick metal pressure hulls. This was a Primark/Lidl submersible.

This is the first time I've watched one of your videos. I thought it was really interesting and well produced, avoiding all the speculation and just looking at the bare facts (and even questioning the mighty Real Engineering channel!). Subscribed 👍

Carbon fiber for those who don't know is essentially a rope. Think about the cables that hold some bridges, they are designed in a way that their tensile strength is much higher than their compression strength, the material is so strong that it could sustain strong compressive forces but only at a fraction of what they can support went tensed. Carbon fiber is by its design meant to be extremely impressive when tensed but it can only hold a fraction when compressed. Steel hulls are extremely strong under compression when using basic geometric shapes like a cylinder. To be an innovative application of carbon fiber one would have to understand how to weave it in a certain pattern and calculate a certain geometry for the submersible so that under uniform pressure the carbon fiber would have to extend in order to keep everything in place instead of contracting. This is admitting such a solution even exists. The other option would be to brute force the problem by creating such a thick and dense carbon fiber hull that even if the compression limit is only 10% of the extension limit it would still be enough to easily go back down to 5km and come back up. I am not even going into more technical details such as computer modeling. It is extremely easy to model the behavior of a cylindrical uniform Steel or Titanium hull (so much so that a noob could do it) whereas it is extremely complex to model a material composed of two or more due to the non linearity of the partial differential equations which results in bad behavior of the numerical solutions that your software will generate.

Appreciate the time and effort put into this presentation! The gradual, patient, and detailed layout of the discussion, with the intricate graphics and comprehensive editing, truly allows for better insight into the whole process behind the making of the Titan submersible.

The key to Carbon is it fails like glass and it suffers hidden stress failures from repeated deformation . We use it in sailing all the time and it is great stuff , light and strong . When it breaks it does not break a little bit .

It would be interesting to see what OceanGate's process was for generating the CF portion of the hull. Seeing videos of the layup process, I was surprised that there wasn't more attention paid to area cleanliness as well as temperature and atmosphere. A documented process review would prove interesting.

One basic thing about the whole sub business concept is people subjecting themselves to such an extreme risk just to take cell phone pics through a porthole. It isn't like climbing or scuba diving where you are intimately connected to the journey and process. Here you are just sitting and totally passive and then you glance out a window. There really is not much difference in the end than if they'd had just stayed on the surface and operated a robotic sub and taken pics that way. Maybe they could've done it while being bolted into a cold tube lol One thing people starting new businesses like this share is an almost blind self-confidence. You really need to be insanely cocky to not be cowed by all of the risk and not scared of failure. Unfortunately this self-confidence bled over into the engineering and safety aspects of the sub.

What has happened will set the carbon composite science back decades. Now any time someone brings up using them in any safety capacity this will het talked about and not tried. Proper testing is sooooo important.

I sometimes have a feeling I live in different reality than most people. There are people even here in this comment section confidently repeating that CF cannot be used for this application. Yet, it has been used 12 times successfully before it failed. The reality clearly proves that it is possible. Now it can be entirely possible also, that the limitations of the CF in terms of economics of it's use in this application are prohibitive, but technically, it clearly is capable of doing it. I have a sense that one of the reasons why Rush ignored the many warnings he got was exactly that the warnings were in a form of 'obviously you can't use CF' or 'this will never work' which made him discard them completely. The lesson here is that even in criticism you should be precise and accurate and not overstate your case. Your analysis here is good exampe of that approach.

As a guy who worked a lot around advanced composites on [military] aircraft, and also the ones used on sports car in my spare time, all I can say it that CFRP is light and strong, but it gives almost no warning when it has to give... Just look at F1 cars when they hit something, and you'll have the whole view of what happens when CFRP fails : it shatters like a little grenade ! Thousands of sharp little black carbon blades spinning in all directions. And to add to the injury, CFRP is not good at withstanding compressive loads. I mean way less than what it is able to do in terms of tensile strength. That's why many helium/nitrogen/oxygen/"gaseous whatever you want" high pressure tanks are made of composites in aviation and rockets and work very well. That's also why nobody ever decided to build a fiber-reinforced plastic submarine before. If composites were good for submarines, the military subs would have done it a long time ago. - Pressure trying to escape from inside = yes - Pressure trying to crush from outside = big no

This was a very reasonable overview of the craft, with great insights on possible weak points. The craft itself was working astounding 15 times, but as you pointed out - we (and them neither) had a clue, of what the safety margins were , and how much role material fatigue played in the catastrophic accident. That ist really unfortunate, because the idea itself still looks pretty promising - even from a technical standpoint! Thank you for carefully pointing out how a responsible path to such a submersible could look like.

This is the best KZfaq video on the Titan. No drama, just pure facts and very detailed as well. 10/10 I just want to add more on the “Real time monitoring system” Assuming the leaked transcript is real… The system actually seems to work as intended. It first sounded alarm at [09:28:16] and the last message from Titan wasn’t until [09:46:37] So that would’ve give them about 20 minutes or more to react… which should be more than enough to initiate emergency ascend procedure. but then 2 minutes after the alarm sounded, a message from Titan reads “no change with thrust the *rate of descent is increasing* At 35. Going to release the ballast now.” The only logical explanation I can think of is that the Titan had somehow, gained some weight… which strongly suggests that the water was leaking inside the hull between layers of the carbon fibre. So the process in which the hull failed does not seem to be immediate but took place over 20 minutes until it gave up and imploded. If not that, then Titan was most likely attacked by a giant squid which cling on to the hull (activating the alarm and increasing the rate of descent) and after 20 minutes of tentacle handling… finally crushed the hull into pieces. Again, this is based on if the leaked transcript is real. So far, I can’t proved that it is real, but I also can’t proved that it isn’t either. Note that the content of the leaked transcript all seems to check out with the available evidence. Also, both the authorities and Ocean Gate have not come out to denied the leaked transcript either… so most likely it is legit.

Ah ! finally we have someone who explains what happened. This is exactly what i explained to my friends as i work with carbon fiber since more than 20 years in racing car applications. The big problem come from the end caps, applying 5-6 times the pressure on the tube ends because of geometry, so after a while and many pressure variations the tube end got some wear coming from the resin aging through the process and the fiber becoming slighly wrinkled as you explained. So cracks start to form and one day you have a sudden major failure .

My claustrophobia is a blessing in disguise. This could never happen to me. I can't imagine wanting to look out that tiny window that you have to share with others.

This sub suffered from multiple systems failures in 80% of its dive attempts, and none of those had anything to do with the CF or game controller, I would love to see a follow up video on these failures

their are a few things in this vid that i have slight issues with, seeing the outter flange hanging after the thing has been salvaged from the bottom of the ocean, stuck on a ship and then unloaded at a port, and you see the flange hanging down and assume it got blown outwards? perhaps it got sheerd inwards while attached to the carbon then its weight and the salvage process makes it look now like its hanging outside, doesnt mean it got blown outwards. also something i still dont see anyone mention is the fact that the glue bonding the cylinder to the titanium flange had to not only take the stress of the dissimilar materials incomtatable compression properties each cycle, but the amount of hours they spent with the door hanging open in rough seas troubleshooting the long list of issues, the amount of missions that got scrubbed, each time with the door spending hours open with all that uneven force from that heavy ass door hanging from one side must have been putting serious load and fatuige on the bond that it just wouldnt deal with well like it would all the other loads its subjected to, i truly beleive its that hindge and hemisphere being swung open and kept open far too often that caused the most stress and fatuige on the bonding surface that it just does not look to me like it was designed with that load in mind, yes its got a well engeneered joint structure to put up with the stress of dives, where its all evenly compressed and both hemispheres are basically pressed inwards onto the bond while diving, but its got a sheering, twisting, peeling load put on it from the door

Thank you to the video creator for your insights. I'm an engineering contractor for some military in some part of the world. The problem here was too little testing on too radical of a design with too many lives at risk, and most importantly, naive patrons of the business who were too willing to risk their lives.This is more of an ENGINEERING MANAGEMENT problem than an engineering problem. There is a difference

you started by saying that the media made the carbon fiber hull as a bogie man and then you described how you think the carbon fiber failed. Wrinkles verse de laminating is really just semantics for the average person. The key is that the hull was untested and went straight to commercial use at extreme depth. Even aircraft with aluminum hulls at 1 atmosphere need to be regularly tested for cracks caused by regular pressurization.

Wow. Actually not a clickbait video like all the others. Thank you. I wonder, though, is carbon fiber really that strong in compression? I failed to find much about it (maybe because the opposite notion is trending so much). I am not an engineer, but I handled carbon fiber strips and composites, and based on that I would never imagine it being strong under compression. At most it would be good at keeping the epoxy in check and not deforming, but in reality it would be under tension (inward) and just keeping the geometry consistent enough to withstand the pressure. How much weight of the hull was CF and how much was the epoxy bonding? I guess epoxy made the bulk of the real strength, or am I completely wrong? In either case, any delamination (like the endcaps pushing on the end of the carbon fiber strands and pushing them apart) would spell disaster, once you compromise the intended geometry the whole strength is just gone, so hearing any kind of crackling should warrant immediate testing of the hull.

This contained some information I had not seen before. Thanks.

I think it's important to remember that airliners (representing a multi-TRILLION dollar business) have only just began implementing composites in the last 20 years (despite the immense weight reduction and thus efficiency it provides.) Composite airline parts are built with ample metal reinforcement and held to an unbelievably narrow tolerance. Airbus gave us a look into how they build the composite-based A350 and the complex, precision engineering required to build those parts is unbelievable. In fact, it is considerably *more* expensive to build with when compared to traditional aluminum airframes. If Boeing has only *just* figured out how to build with composites, it's best for silicon valley techies to leave it the hell alone.

I have worked with Titanium and Carbon Fabre for yachting purposes. My limited engineering knowledge is asking about differential thermal expansion and contraction... and Youngs Modulus is different for both materials... As you mentioned a rounded filament wound shape similar to SCUBA tanks would have been a better design... or a sphere... BUT then you have holes at the ends for access... and there would likely be a higher resin to fiber ratios due to the geometry in those areas --- plus more sharp radius curves === and possibly some cutting of the fibers for access ports. === and this could be a place of weakness --- it just doesn't take more than a pin hole to start cutting and rupturing materials at those pressures ... or worse yet a catastrophic failure... Look at the water jet cutters --- that could happen down under .... with no way to stop the intrusion.

This is probably the best presentation on what happened. Someone soon may be able to use carbon fiber installed to withstand pressure. It'll require better testing to insure safety over multiple dives.

9:27 While there might be some compressive forces on the wing, they’re minimal compared to the predominant tension forces depicted in that clip of wing flex. Thanks for the video.

Excellent overview. Thankyou for explaining to your audience the basic principles involved. I tend to agree with your assessment. I am a filament winder, and do deep sea and gaseous pressure vessels. We generally burst test to 2.5 times rated pressure and cycle test to 1.25 times rated pressure fill and empty up to 450000 cycles! This takes a long time and lots of power, but helps us sleep at night. Thanks for putting together this excellent overview with out the sensational hysteria. It must be very time consuming.

Interesting perspective! I've seen thunderf00t's take on the subject both are compatible as far as I understand. According to him the main culprit is the difference in compressibility between the joint (titanium ring) and the carbon fiber tube.

This reminds me of the darwinian advancement of early aircraft, when the designers were the test pilots. The good designers lived to spawn more designs...the bad ones not so much. 😮

I...can't help but feel anxiety just imagining sitting inside this thing while they bolt it shut. Fascinating video here, thanks for taking the time to do such a detailed breakdown. ... ugh... my word choice feels...off.

Great video, nice to see someone who actually understands the issue rather than just blaming composite material. One correction I think I have is the distribution of forces, you've stated as 3/4 in hoop and 1/4 axial. I thought the distribution was 2/3 in hoop and 1/3 in axial, although this is based on my knowledge of pressure vessels with internal pressure, i assume the same is true here just in a different direction. This is a based upon stress in hoop being pr/t and pr/2t in axial, i.e. giving a 2:1 distribution ratio.

Definitely found this video very insightful, thank you for creating and sharing your knowledge on this subject, will be looking forward to watching more of your stuff.

This was a well presented discussion regarding the Titan implosion with truly excellent graphics. But the one point of interest that was not mentioned is the elastic modulus of the two materials, carbon fibre hull and titanium end caps. They do not react the same under pressure, so over time there is going to be degradation.

Just looking at the interior photos gives me the creeps. I wouldn't have gotten locked in that thing on dry dock if they paid me $250,000.

Thank you for your balanced report that clarified the design issues with Oceangate. I now have a better understanding of the implosion and look forward to the final report.

The flange torn off the end cap says it all for me. The ends of the CF tube were only stopped from deforming under the compression of the endcaps by the flange which was not strong enough to resist the force of the ends of the tube pushing outward against it as they tried to deform under the compression. Once the flange went, there was nothing keeping the ends of the tube from deforming and squirting out around the sides of the endcap, at which point the pressure simply forced the endcap down the tube much like a mandril gets forced down a pipe to stretch it out. In a sense, the CF tube was being crushed on the ends by a hydraulic press. Forces Stockton likely never took into account in his designs.

Axial Compressive strength of carbon fibers is 10-50% of their tensile strength. And transverse compressive strength is 10-20% of axial compressive strength… even without defects

The hull was basically 5 inches of epoxy resin. I give the CF little value in such an extreme compression application. Middle of the hull would deflect and work more than the ends, since it had no internal bracing. Popping sounds were probably CF filaments sporadically failing. Obviously...any tiny defects would grow each dive cycle.

15:14 This is what I've been saying for so long; the excursion price was too LOW! It was so low that I now describe it as predatory. I say this because it put the Titanic excursion within the range of affordability to diehard fans (people who have wanted to see the Titanic up close their entire lives) who could now actually afford to re-mortgage their house to get on board the sub, and some of them did just that. I think these were the people that Stockton Rush was actually aiming at. Now even if there really is something to this whole "carbon fibre submersible" idea, it's become a poison chalice that no investors will touch for years to come. Thanks a lot Stockton Rush!

Ocean gate casually littering the ocean with every dive makes me heckin frustrated. Thanks for this great video

I ran across this on my feed and found this extremely interesting and well done, Great work!

The best information about the overall design of the submersible I found so far, thank you. One of the most important points: it has confirmed my suspicion the sub wasn't using variable buoyancy system, and its buoyancy has been solely controlled by solid balast. One of the unique (not necessarily bad) features of the contraption.

When i was a kid a bunch of us drew up plans to convert a 500 gallon propane tand to a sub and charge the other kids to take them down in it. The plans disappeared over the years. Now i know who took them.

Fantastic video! This is the most comprehensive and detailed description and model of how the submersible was constructed that I've seen so far. Also, this is the first time I've seen that video clip of the carbon fiber being laid axially.

I actually thought this was going to be one more of the videos I've seen about the topic with the same information, but I was really surprised to see how much more detailed and scientificly explained the whole incident was. I subscribed in the first minutes.

I appreciate the tone of this video and was delighted not once to hear the word “hubris”. We all know the Stockton Rush was impatient and too shortcuts. He ignored experts advice. But not all on board with him we’re innocent victims, as Nagolet knew all the shortcomings and had told in a text or email he would rather die instantly and catastrophically beneath the sea than a bored old man (or something to that effect) There was the usual media frenzy crazy speculation from incomplete information and scathing character assassination…when Rush was not all bad. He had started out in submersibles using his to take research crews and journalist. He had student educational programs. He actually was doing a service. But he became ambitious beyond his financial means and skipped the critical testing to failure and non/invasive periodic testing of the hull’s integrity with ultrasound as recommended-instead relying on Wile E. Coyote methodology acoustic sensors (seriously I imagine them arriving in an Acme box). Some experts have also mentioned a curing process (I am not an engineer) to strengthen the adhesive bond between the fibers. I kept pointing out your factual errors in that Rush made a second hull after the first one was making noises, and other things, but those minor error are corrected here. Rush went around certification, called passengers mission specialists, outright lied,etc but a complete villain? No. I even had to correct someone from calling him a psychopath. Really?

I always enjoy a technical video that uses animated graphics to explain/show construction and operation. It makes it so much easier to understand. The graphical simulations used in this video were excellent.

I really appreciate the thoughtful, rational way you explained this. I'm so sick of armchair experts who have no experience whatsoever in the design and construction of such a unique vessel. SR obviously made some literally fatal mistakes in how he treated risk and testing, but there's no reason that some company/institution couldn't spend the time and money to properly test and overspec such a craft successfully. He really was blind to the risks, or at least chose to ignore enormous issues with the whole design and build process, skimping on testing and lacking a willingness to seek outside expert advice and admit he needed such help. It would be amazing to see someone do it right and open up more of the ocean's depths and mysteries to people who ordinarily would have as much chance of experiencing such environments as the average person would to stepping foot on the moon.

So there are many comments here and some half truths and statements from people who sound like experts, but the stuff they say is just wrong. So being an Engineer that worked for the past 30 years in the aircraft industry and design and build high performance sailplanes from exclusively carbon fiber, I will add my 5 cents. 1. Carbon fiber is a fantastic material with some limitations. The compressive strength is lower than the tensile strength and for carbon roving my own measurement over a 20 year period in our University Materials Lab, I found the following. Compressive strength 670 MPa, tensile strength 1000MPa. 2. In compression the fiber absolutely gives the strength. The lower value is due to the fibers not being absolutely straight and therefore they buckle (Engineering term for structural instability of a column in compression) on a micro level. The resin only keep the fibers straight and contribute very little to the compressive strength as the resin strength is only about 70 MPa. 3. Somebody commented that the analogy of a wing is not correct as there is not really compressive loading in a wing. That is absolutely wrong. The structural element in a wing is basically a beam that resists the bending moments. In a beam the one side (bottom for aircraft) is always in tension and the other side in compression. We use carbon fiber primarily as the structural element in our beams and the upper flanges is stressed in compression. Because the compressive strength is well known, we can design save wing beams. As a matter of fact, the tensile strength is an useless property in this application and we are only interested in the compressive strength. 4. So to say CF cannot be used in compression is really an idiotic statement as it is untrue. The compressive strength is merely lower than the tensile, but still very high and usable. If you know the value accurately you can design safely with it. 5. The application of CF is a submersible is not necessarily wrong. 30 year ago nobody used CF in aircraft. Now it is well understood and used (787). At some stage somebody must push the limit and use this wonderful material in new applications like a submersible. The problem is that the problem must be well understood and the material be properly used. The compressive strength of CF makes it a good candidate for this application. The manufacturing method used was probably not completely suited for the loading of this design. My own analysis show a SF just above 1 for the limit loads. This is too close for comfort and a little more structure might have been the right thing to do. 6. I Agree with the comments that the fibre must protected against water ingress that will cause delamination. Regards Attie

10:49 No it doesn't. The composite is not isotropic, or even orthotropic, it's anisotropic, with maximum stiffness and strength aligned with the axial direction of the fibre and in tension only. This means the hoop arrangement of fibres carry axial forces only, i.e. internal pressure causes the fibres to be in hoop stress - this how this sort of composite is meant to be used. In compression the matrix material carries the force and fibre does nothing (you can't push a rope analogy). My PhD was in numerical modelling of crack propagation in long fibre metal matrix composites (MMC's) - titanium matrix/carbon fibre reinforcement, or aluminium matrix/carbon fibre reinforcement - which run in the hoop direction which is essentially what this composite was so I'm familiar with MMC failure modes. The upshot is, the designer of this vessel did not understand the stiffness properties of the composite he was using and paid for this with his and other's life. 11:20 Yes, exactly, and this is how it failed in hoop as well; remember the fibre is there to carry an axial load _in tension_ it cannot carry a compressive load; you can't push a rope.

The way they glued the titanium rings to the CF alone, absolutely ridiculous. They rolled the glue onto both parts, set the ring on top of the tube, done. No elaborate resin infusion technique or so to prevent voids in the glue, no vacuum bagging afterward, no autoclave curing. Nothing. They built pressure vessels like it was an Arts'n Crafts project for their Etsy shop or something. And then there's the combination of CF and titanium. I'd have to look up the exact numbers, but I'm pretty sure titanium has at least twice the the temperature expansion rate of carbon. They glued that hull together during summer judging from their clothes and sweaty foreheads, then repeatedly sent it down to the bottom of the ocean where it's 0°C. The internal sheer stresses in those halfassed glue joints must've been substantial.

Excellent video! Really appreciate the time you took to put this together. Did graduate level courses on composite materials in the process of getting my masters in ME. After this happened and the video was circulated of the team gluing the titanium end rings on in what looked like vacant warehouse it blew my mind that it survived 15 dives. Everything we created in my labs was done in a clean room type environment and that was something just for testing not to support and keep lives safe. They didn’t have on respirators, eye protection, even using clean rags and wearing gloves to keep hand oil off the bonding surfaces and out of the epoxy mix was apparently too difficult. While I agree with your analysis on how the sub failed and highly doubt it was the titanium to carbon fiber surface, this video just showed the overall lazy approach to working with a composite that even someone making a car hood or bumper would have had a basic level of understanding on and had avoided. With all of the failures that occurred during manufacturing and design I would argue that it proved carbon fiber as being a suitable material for submarine use.

By far the best explanation of what really happened. Thank you for taking the time and patience to put it all together.

I'm glad you stated that the transcript is most likely fake. For starters, it does not seem to be plausible to me that the dive would have continued with the abnormal descent rate the transcript implies.

I'm just a guy who is curious about science and I found this was fascinating, thank you for putting it together. As usual, the media just ran with whatever got the the most views, damn the science or facts.

Good point on the chamber cylinder axial loads. My take on this failure is in that carbon fibre is stronger than the epoxy to fibre bond strength. So the other loading on the hull material for a layered construction is radial. Without a radial tensile component in the build there was nothing to prevent the inner layers developing longitudinal ripples as the inner layers progressively delaminated under the compressive shrinking of the diameter. Mid layers are kept straight by surrounding material but the inner layers can wrinkle inwards to restore their original circumference. Also I would have opted for a shallow “V” tapered tube end socketing into the titanium end caps. The stripping of the titanium groove ring suggests that the hull ultimately on one side rather than Uni formally.

My theory: Tiny air bubbles at atmospheric pressure that were inadvertently included in the carbon fiber imploded one-by-one under extreme pressure. First the bubbles near the outer surface imploded, which weakened the outer surface, then deeper air bubbles imploded, until the carbon fiber still l@@ked like carbon fiber, but it was broken up around imploded air bubbles on a microscopic scale until it was unable to carry the load. Has anyone looked at THIS possibility?!

Oceangate Submersible Titan Implosion Analysis June 30, 2023 Jack Noble Composite barrel information from internet: 55.9 inch inside diameter; 95 inch length; 5 inch wall thickness; 0/90 CFRP laminate. With a CFRP modulus of 8 million psi in the 90 degree direction and because the stabilizing effect of the end rings is about 10 times the thickness away from the center, the failure mode will be buckling in the middle of the barrel. The stability formula in Roark & Young yields a failure pressure of 8900 psi. Dividing by the design pressure of 5700 psi give a factor of safety of 1.56. This gives a strain level of .0044 in the composite and a radial deflection of the barrel of .145 inches. At the ends of the barrel, although insufficient information on the rings is available to determine elastic properties, it appears safe to say that the radial deflection of the rings due to pressure will be much less than the barrel. This means that the rings will be holding the barrel out at the ends, flaring the barrel like a cooling tower and creating significant bending stresses in the barrel and where you have bending, you have interlaminar shear, the achilles heel of composite laminates. Metals typically have allowable shear stresses of over 50% of the allowable axial stresses while composite laminates have interlaminar shear stress allowables of 2% to 3% of the axial stress allowables. If only 1.5% of the pressure load on the barrel was transferred to the rings the shear stresses would be enough to cause delamination in the center of the thickness of the layup all the way around which would account for the noises heard on initial dives but would not be visible and would not cause failure because the critical stability section is at the middle of the barrel. On subsequent dives this delamination would progress toward the middle of the barrel creating smaller noises, also reported. This delamination is converting the 5 inch thick composite barrel into two concentric barrels 2.5 inches thick. The critical buckling pressure of two concentric barrels 2.5 inches thick is only 25% of the critical buckling pressure of one barrel 5 inches thick so the margin of safety would be gone long before the delamination reached the middle. Since there is only one design condition of overall pressure it may have been possible to tune the stiffness of the titanium ring to more closely match the deflection of the composite barrel and avoid this problem but the fact that the composite shear allowable was less than the applied pressure would almost certainly lead to the initiation of delamination somewhere on the barrel. The only protection from this delamination would be to install several hundred 5 inch long bolts through the shell and the potential for leaks would render that fix impractical. It appears that it is probably not possible to successfully use this material in this application.

What a fantastic depiction of the possible cause of this tragedy! Well done sir, your video was so well designed and gave me a big insight to the failure! Thank You Sir

This is by far the best 3D model of the Titan. Thank you for sharing!

Thanks for providing such a detailed and insight filled presentation. It’s an antidote to the flap, clickbait and sensational guesswork on some channels.

Very interesting video - makes sense (actually I have been thinking about the tube design; the thickness of the tube would mean that it would actually be under tension if it was being compressed (at least part of the tube) due to the 5 inch thickness. The longitudonal pressure (the 25%) was unopposed; This has parallels to the De Havilland Comet; hopefully similar testing is done to figure out what happened. A better design could have been a ‘thin’ (1 inch?) titanium tube, wrapped in 5 inches of carbon fibre, maybe with some external longitudonal braces also. But destructive testing should have been done.

Most comprehensive analysis of any video on this subject. Very well done.

Regarding your comment about the crew not wasting time typing messages to the surface ship during a crisis, that may be true for a normal person but this is Stockton Rush we’re talking about. He had a propensity of being flippant about danger and never seemed to stop talking so I think the messages could be credible.

Very well done. Best presentation I've seen overall. And not to take away from that in the least, am still not convinced the transcript is fake. The work in the main presentation is very solid.

Well done! I worked 33 years in the carbon fibre industry and I was wondering how the fibres were aligned in the cylinder. For a number of years I was part of a team that worked on very thick carbon laminate structures and worked out a way of curing them under pressure without wrinkling or deforming the laminate. This can be very difficult to do, but can be achieved. The combination of filament, winding, and hand layup would have been difficult to accomplish and very may have been the source of the problem.

This was a outstanding report - well researched, objective, and highly credible. Thank you!

I always thought it was where the end caps and Carbon Fiber met where it failed because of the different compression strengths. Also, I have read that the Carbon Fiber was bought for Boeing at a discount because it was past its shelf life if there is such a thing. Cheers thanks for Video.🍻