Transcription
The Coast Guard final report was released on August 5th, 2025. It was 335 pages long and had a great deal of information that helps explain why Stockton was so confident in both holes that he built. In this video, I'm going to tell you the story about his complete confidence in the second hall and how and why Stockton came to this conclusion. The strength analysis he was presented was excellent. The MBI final report states it this way. In summary, the safety factors reported for the final Titan hole design were well above the required thresholds for aerospace applications, ensuring a high margin of safety under the anticipated operating conditions.
In a previously released part one video, I told the story about why Stockton had confidence in the build of his first hole, the one that cracked in the Bahamas. Stockton had even more confidence in the second hole, the one that imploded. He was told by carbon fiber engineers that this hole was ready for Titanic depths. When I review the basis for the engineers conclusions and Stockton's decisionmaking, I think you will be shocked by the process. Stockton really thought the second hole was indestructible >> and he told us that it was basically indestructible. >>
The Coast Guard public hearings left unanswered questions about who helped Ocean Gate engineer the second hall. Phil Brooks was the only Ocean Gate engineer employed at the time the second hole was being built that testified publicly, but he was a software engineer. His testimony was very important and mentioned some interesting things about the time period of building the second hole, but his story alone never filled in all the blanks. He never even mentioned the main company Stockton had hired to do the computer analysis of the second hole. Kier Aerospace. I will tell you this, Kier was the Boeing of the second hole. I suspected Kier's involvement because in the MBI documents concerning strain gauge placement, there was a single image that had the Kier name on it, but that was it. It was the only clue. We had to wait for the 335page MBI final report to learn about Kier's work. And then the day after the final report was released, the Coast Guard released additional information. 1430 pages of transcripts. These transcripts included some private Coast Guard interviews. A significant interview was with Dan Scoville and information provided here allowed us to completely figure out the second hole story. Let's jump right into understanding the story of Kier and the second hole. Watch until the end to methodically see the poor engineering decisions laid out. And at the end of the video, Jeff Ostro will comment on my Titan book.
So, no one from Kier testified publicly at the hearings and the 1430 pages of transcripts do not include an interview with a Kier representative. It's unclear if the Coast Guard spoke to Kier or if Kier just responded in writing to specific Coast Guard questions. I am sure that Kier was all lawyered up because they were intimately involved with the second hall and they had no maritime experience. Dan Scoville was made the director of engineering after Tony Nissen was fired. So Dan was the second director of engineering and involved with building the second hole for Stockton. In his private testimony, Dan states that he found out about Kier in a trade magazine because they advertised themselves as experts in carbon fiber finite element analysis. Later he would also state that NASA had worked with Kier and had confidence in them. Dan and Stockton had to hire Kier to do the FEA calculations for the hole because there was no one at Ocean Gate who would have been able to run their Hypersizer software if they bought the software product outright from Kier. Here's how Dan described it in his testimony transcript. So we did a finite element analysis on that. I didn't I don't have the skill set to do that. Neither does anyone at the company, at least not at this point. So, we hired a third-party company to do that. So, we called a company called Hypersizer. He meant Kier. Their software is called Hypersizer. I found them in a trade publication. They specialized in FEA for carbon fiber. So, they sell this product. I called them up and said, 'Hey, there's like nobody here is going to run your software. This isn't our gig. You know, we're not going to be able to buy your software and run it because we're not going to be able to get in and get the result out because we don't know what we're doing.' Even though Kier was an aerospace company, Dan states they responded like this. Would you guys please contract with us to run your own software for us and analyze our hole? They thought it was a cool idea. They liked a different thing to work on because, you know, they work aerospace.
Here are the two requests for proposals and purchase orders from Ocean Gate for Kier to do the FEA with their hyperscizer software. This is for the second hole referred to here as Cyclops 3. Request for proposal for engineering services to be performed for Ocean Gate. On December 31st, 2019, OceanGate issued a purchase order for $33,000 to Kier Aerospace. On May 29th, 2020, Oceanate issued a purchase order to Kier Aerospace for $37,500. I want to interject an interesting aside here. During the same time period, in the early 2020 time period, for some reason, Oceangate was simultaneously reaching out to Boeing again, but Boeing ignored them. Listen to Mark Naggley. >> Uh, I think the last time Boeing was in discussion with Ocean Gate was uh March of 2020 where we declined uh to respond to a request for proposal. I have a feeling Boeing did not want to deal with Ocean Gate and Stockton any longer.
Now it's time to go through the MBI final report and review the collaboration between Kier and Oceanate. Again, we don't know exactly how Kier provided the information to the Coast Guard, but this is the content in the MBI final report. In terms of engineering this missionritical structure, the carbon fiber hole, the information is shocking. As I go through these entries, try to understand how Kier, a professional engineering company, felt they were in a position to draft an engineering conclusion about the predicted strength of the hole and present it to Oceanate. And from Ocean Gate's perspective, how could they possibly trust the analysis? Because with an analysis such as this, you must have good data input to get reliable data output. You will see what I mean soon. The scope of their work was limited to static stress analysis did not extend to fatigue, damage propagation, crack growth or dynamic loading which were outside the services Kier provided. According to Kier, the CAD model used for the analysis was created prior to the contract and provided by OceanGate based on previous work completed by Boeing. So all the work was static, nothing dynamic. And OceanGate passed to them some of the work done by Boeing, which must have been information from the 2013 feasibility study. The emphasis on static analysis is repeated. The first contract included an FEA with both 2D and 3D models. This included evaluating the deformation of the vehicle under static loads, analyzing the interfaces between carbon fiber and titanium, and determining the theoretical performance envelope of the pressure hole.
Kier did do something smart. They knew they had to have metal domes to test the scale model holes because during the Spencer scale model test, the carbon fiber domes always failed. For the 1/3 scale tests, Kier designed a steel ring fixture to mimic the deformation profile of a titanium end dome. After the first one-third scale model test, Kier analyzed the failure load and measured wrinkles attempting to adjust the finite element model stiffness to match the observed failure load. But concentrate on this next part because here starts the failure to fully evaluate the test holes. However, even after adjustments, the FEM was unable to predict the failure load in the second test due to the magnitude of wrinkles in the carbon fiber layers. Kier recommended that a highfidelity damage progression model would be needed, a service their company did not offer. So the wrinkles were so bad on the first scale model test hole, they could not adjust the computer model for the second test hole predictions. They recommended a better modeling method and it seems that Oceangate did not pursue it. So it basically means that the test models were not giving useful data to be extrapolated to the full-scale whole computer modeling. Furthermore, the test models were fabricated with a single layer and cure technique. So they did not match the fabrication technique of the full-sized hull. >> We built two of them. The issue on those two we there were not they were not multicure. They were just single cure to and I can't remember the exact depth but it was probably around 2800 m and both test models failed at about 2800 m because they would fail at the wrinkle. This was a weak area. So basically the one-third scale test data was worthless. Dan Scoville did not think they would get to the Titanic if they could not fix the wrinkle problem. So basically the bottom line was are we going to do multicure or not? And so my mindset was we're never going to the Titanic because we know this thing is going to get a wrinkle. We know that if it gets a wrinkle it's going to fail based on all the other testing we've done. All the third scale models we have made, we know that the hole will fail at, you know, low depth, not making depth, right?
Kier did help Ocean Gate with preparing for the deep ocean test facility regarding strain gauge placement for the testing. And this is why there was this one collier image in the early MBI documents. It had to do with strain analysis of the titanium dome. And this was the plan for strain gauge placement. During DAFF testing, there were two on the dome. But even when working on the titanium dome analysis and Kier trying to do a proper job for ocean gate, ocean gate showed their incompetence. For the full-scale test, Kier contributed by advising on strain gauge placement in the dome. Collier subsequently reviewed strain gauge data collected during the test observing good correlation in trends with the analysis model. However, they could not compare absolute strain values because Oceangate did not provide the necessary gauge factor for the strain gauges. I can't fully explain this strain gauge controversy based on this paragraph, but various makes and models of strain gauges made of different materials have what is called a gauge factor and you need to know exactly the model you are using to fully evaluate the data it is giving you. There was documentation for the model of strain gauge they were using. So this information should have been available. But apparently Ocean Gate could not provide Kier with exactly what they wanted. So the conclusion was this. However, Kier was able to inform Ocean Gate that the relative magnitude between gauges was comparable to the model. So this is another example of poor attention to detail. They just accepted strain gauge assumptions for the domes. That was less data than what would have provided a full picture of strain.
This is an interesting entry. Since the contract was hourly based, Kier only performed tasks that were explicitly requested by Oceanate. This ensured that the scope of work remained focused and aligned with the client's needs and objectives throughout the duration of the project. The contract was hourly based. I believe this paragraph is code for the fact that Oceangate had a strict budget and did not want to pay for anything extra.
Kier responded to a question from the Coast Guard about cycle life for the hole. In response to a question asking Kier if they had modeled fatigue for the new Titan hole in order to determine how many deep sea dives would be possible before failure. Kier provided the following written response. Not by Kier. Our company does not have expertise in this area. When we asked Oceangate about this topic, we were told that their acoustic monitoring system would be used to detect damage growth in real time and that this capability had been demonstrated on the previous iteration of the carbon fiber hall. This is another miss on a crucial point. Ocean Gate had no prediction on how many dives the hole would tolerate.
Regarding the manufacturing, again, Oceangate handcuffed Kier because they could not provide good data. Kier was unable to model manufacturing process defects or knockdown factors for the full-scale Titan due to a lack of essential data from OceanGate. Knockdown factors are strength reduction factors that need to be considered in the modeling. Boeing emphasized this in the feasibility study from 2013. You can't proceed with modeling if you can't factor these in. Mark Negley had a classic line about this considering the effects of defects. We did a feasibility study basically a a a concept or a trade study output and didn't consider uh effects of defects or anything you know that would have been in a a detailed configuration and analysis.
Early in the project, Kier recommended testing with thick coupons 1 to 1.5 in to derive accurate material strength allowables. But Ocean Gate later stated that such testing was not feasible due to the high cost of the required test facilities. Just as Tony Nissen, Boeing, and NASA had told Oceangate, the proper way to understand the strength of the final hole is to fabricate test coupons which represent exactly the fabrication process to know the final strength of the fabrication compared to the predicted strength. This paragraph used the term allowables regarding the carbon fiber system. Not having a materials expert on staff at Ocean Gate was a huge mistake. Dan Scoville, who was the director of engineering at the time and was responsible for communicating and coordinating the fabrication of the hull by Kier Electro Impact and Janicki. But he was an electrical engineer, not a materials engineer. This is what he told the MBI. They had to create sheets called the allowable sheets. The hyperscizer guy asked us was what are the allowables. I'm scratching my head like what's that mean? And so then he explained it to me again. It's a little bit outside my area. I don't think it's the data sheet for just the carbon fiber. I don't think it's the data sheet for the resin. I think it's like the combination. highly technical stuff. If you're not in the industry or you're maybe a material science guy, you might be struggling as an electrical engineer. I was struggling with that stuff, right? So, anyways, we were able to get the allowable sheet from somewhere because Boeing or somebody had used that material. Boeing spends hundreds of thousands of dollars to get this allowable. It goes through testing to figure this out. They don't necessarily want to hand it to all their competitors, right? But somehow we had the allowable sheet for this toé material we used. That's how we tried to get the correct numbers, the numbers we're looking for. We found it in the allowable sheet. And there's a class A allowables and a class B allowables. You know, I don't know when you get to use which one, but depending on what aerospace thing you're doing.
This admission by Dan is interesting. Stockton put Dan in a bad position. He did not have enough experience to provide constructive input on the carbon fiber processes in this hall. It's clear that Oceangate needed a materials expert on staff. Sure, they got allowables numbers from Boeing for the toé product laminate, but who evaluated that in the context of the five layer co-bonded hall? We don't know. This is another example of a variable in the design process, which adds to the question about the final strength predictions.
These next paragraphs discuss how problematic this was since manufacturing defects are typically accounted for by modifying material strength and Kier was not contracted to derive material properties for the final Titan Hall. The company did not offer this service. OceanGate did not provide the necessary material strength data, nor was relevant publicly available data accessible for 5inch thick laminate structures. When the issue of manufacturing defects was raised, Oceangate indicated that it would be addressed during fullcale testing of the hull. During one-third scale testing, Kier was asked to account for wrinkles by modifying the holes structural stiffness in those areas. However, this approach was found to be unviable during further testing and was not applied to the full scale Titan. The one-third scale test also did not provide sufficient data to estimate manufacturing process knockdown factors for the full scale hull. Kier needed two critical pieces of information to estimate manufacturing knockdown factors for the Titan. The strength reduction as a function of defect magnitude and the actual magnitude of the defects on the full-sized hole. So did Kier have the first piece of critical information? The one-third scale holes which had large defects such as wrinkles were unrepresentative of the full-scale design due to the differences in size and manufacturing processes and therefore could not provide useful conclusions about how defects would affect a full-sized hole. Did Kier have the second piece of critical information? Additionally, Ocean Gate did not provide data on the magnitude of defects for the full-sized hole, and measurements would have required non-destructive inspection or dissecting the hole, both of which were deemed unfeasible by Ocean Gate.
So, after all these problems, with proper data input and modeling and the whole strength predictions, Ocean Gate just blew off the problems. They assumed the hole was strong enough and they could rely on real time monitoring. Oceangate decided that the strength of the hole would be sufficiently validated through testing and real-time monitoring with acoustic sensors. Oceangate was confident that the RTM approach would be viable for assessing structural health on the final Titan hole well ahead of a catastrophic failure. But really the worst part of this story is this. After all these inconsistencies with the modeling process, amazingly Kier proceeded to give Ocean Gate an assessment of the safety of the hole. Let's quickly review the factors contributing to potentially poor computer modeling and whole strength predictions with the Ocean Gate and Kier collaboration. Only static stress analysis was done. The scale models gave worthless data. There was strain gauge specification confusion. Whole cycle life or number of dives was never predicted. Knockdown factors were unclear. No coupon testing was done to validate fabricated material strength. No nondestructive inspection was done to evaluate defects in the fabrication. No engineer of record for the co-bonding process. Accepting real time monitoring data during actual dives for validation versus valid engineering before building the hole. No engineer of record for grinding down wrinkles, which we will discuss next. Based on this problem list, wouldn't you think that Kier would have backed off and told Stockton, "We have such poor data that our modeling is not reliable and we can't rate the hole." But no, they gave the hole a fantastic rating. This is how they did it. This is what they reported to Stockton.
There are a few ways engineers work with margin of safety and factor of safety. But for this discussion, what they reported to Stockton to consider was a margin of safety and a factor of safety greater than one. The safety factor as was determined by Kier of the Titan design was reported to Oceanate in terms of margin of safety and factor of safety. A positive MS or a factor of safety greater than one indicates the structure is not expected to fail under the applied load. The analysis used composite material properties specified by Ocean Gate for the T800/3900 system with classical aerospace methods employed to check stress and strain in the laminate for the carbon fiber holes material strength. The FS at 4500 meters of seawater was reported with an additional 1.25 factor applied in the axial direction. The FS was 8.2 and the MS was 7.2 in the general area and the FS was 4.6 and the MS was 3.6 near a stress concentration at the titanium fitting. In the hoop direction, the FS was 6.6 6 and the MS was 5.6 in the acreage and the FS was 6.0 and the MS was 5.0 near the titanium fitting. These values far exceed the typical FS requirements for commercial aviation which is an FS greater than 1.5 and space launch which is an FS greater than 1.4. For buckling of the carbon hall the FS was 5.6. To put this in context, space launch customers such as NASA typically require an FS of at least 2.15 for buckling in cylindrical structures. So those numbers are more than great. They are amazing. But based on what we know about the process of arriving at these numbers, how can they be trusted? So it was Kier who gave Stockton great confidence in the second hall. In summary, the safety factors reported for the final Titan hole design were well above the required thresholds for aerospace applications, ensuring a high margin of safety under the anticipated operating conditions. This is why Stockton thought the second hole was indestructible. Quote, a high margin of safety under the anticipated operating conditions. Meaning this structure used as a hole could withstand 380 atmospheres of pressure. But Stockton did not want to pay Kier to dig deeper and analyze more. As was typical for Kier projects, Kier offered to conduct a deeper investigation and comparison between the test results and the analysis to help pinpoint all possible differences. Ocean Gate turned down Kier's offer and informed Kier that they were going to proceed with construction of the remainder of the final Titan hole.
Dan's private testimony adds great information to what Phil Brooks said in his testimony about working with Electro Impact. when it came time to do the fullscale hall, um Stockton did not want to do the multicure and I believe that EI um said that they would not build it they would not build it without doing a multicure that we you know we had to do it that way or take it take it someplace else. Electro Impact wanted to prevent wrinkles and strongly suggested a co-bonding technique with 1 in thick layers. So at no charge to Ocean Gate, Electro Impact built two test structures on the full diameter mandrel. For one, they did about a twoft length of full diameter hole, but just two layers individually cured to test the peel ply process between layers. We did, I believe, a full diameter, full thickness called a slice. So these were not full length, but one of them was like 2 feet plus two layers. That was to test the shield ply layer that goes between the two layers that get cured separately. Electro Impact was so insistent that we needed to do multiple cures. They paid for that themselves. They were so insistent. You're not going to get a good hole. You've got to do these multiple cures. Stockton kept saying no. Electro Impact then did a second test. They did about a two-foot length of a full diameter hole with four or five layers to prove the co-bonding concept. And then we did a full thickness, but again only probably 2 feet long, four or five inches thick, I believe test part. Dan was impressed with the results because with these short segment tests, they were seeing the best results ever in terms of no wrinkles. But of course, it's always a challenge dealing with Stockton. And Dan tells an interesting story about how he proceeded with the full-scale hall and the five layer co-bonding technique. I thought there was no chance we'd make it to the Titanic, but since Stockton never stopped me, I continued forward. I got the first layer cured. I don't remember what the discussion was, but he didn't like totally freak out at me, you know? It was kind of like, okay, well, whatever, and maybe we'll do more. Then the second layer, he didn't tell me no. Got it cured. I think this went through the third layer. Then he said, we'll just cure the final two inches all at once. I was like, why? You're so close. You've spent 80% of the money. Just do it. We were able to cure every inch. and we had a hole that we thought was, you know, that was the first point I was like, okay, this might happen, right?
However, Dan conveniently leaves out one of the most controversial things about the second hole that he was involved in, about sanding down wrinkles and high spots on each of the four layers of the co-bonded hole. So, the layers must not have been perfect. Phil Brooks gave a lot of details about this during his testimony. >> So when we started doing the multicure, what was done was they would do the first layer and then cure it and then we had a group of people with sanders and built a like a tent and they had to dress in uh very protective gear because carbon fibers is dangerous to breathe. and they would sand down any any imperfections, any sort of blisters or anything like that and smooth it out as best they could. I know I know that uh Dan Dan Scoville um basically had come up with this method to do this and it was done at every multi-cure layer. >> And how would they remove these imperfections? >> Uh just by sanding. >> By sanding down on the carbon fiber, >> sanding sanding down the carbon fiber. Yes. I I'm not sure what type of it was a obviously a power sander that they Yeah, they would go over the whole thing and sand it down. >> Sanding down the carbon fiber layer with dozens of sanding spots per layer and possibly hundreds over all four layers is not routine. >> Would it be common practice to uh sand or grind those off to make the um surface level for the next layer that will be put on? >> No, that is that is not common practice. So you're removing material, you're making a discontinuity. >> Or would you recommend sanding down wrinkles to smooth out the layer for the next layer? >> You know, we we would have looked at that. We would have characterized it. Um and then we would have done an analysis of the reduced performance.
Look at this picture of the outside of the finished hole. They did not sand this layer because it is the last layer, but every one of these bumps is an example of what they would have sanded away on layers one through four. Don Kramer from the NTSB examined the cut off ends, which amounts to about 8 in of hull and found one grind spot on the first layer, about eight on the second layer, about 14 on the third layer, and 24 on the fourth layer. Kramer saw grind spots where up to 12 plies were removed. Since each 1 in layer is 133 plies. If 12 plies are removed, that is removing about 9% of the thickness of that layer at that spot. Who exactly from Kier Electro impact or Janicki said this idea was acceptable? There is no information from Kier that accounted for this disruption to the laminate layers in their modeling. Since Ocean Gate did not do destructive testing, they had no idea what the paracity of the laminate was. Kramer tested the cut off ends and found an average paracity of 2.6% which is 5.7 times greater than the toé specification. who signed off on the final prediction of strength of the hole without understanding how much pocity was present in the one-inch layers. In pieces of the wreck recovered from the ocean floor, Kramer found that the adhesive film between the layers was not 100% adhered. And even worse, there was evidence that it was breaking down. Who signed off on a final prediction of the strength of the hole without understanding the strength of the film adhesive bond?
So Stockton thought the hole was indestructible because the engineers on the project were working down to the level of Ocean Gate and not forcing Ocean Gate to work up to the level of solid engineering. The hall actually performed well up to the end of dive 80 when the big bang was heard and the strain gauges showed an abrupt change in strain. If Stockton had used the worrisome data from his real time monitoring system that he was so proud of and stopped diving after dive 80, we would be having a completely different discussion and Stockton would have been remembered as the innovator he wanted to be remembered by. If you have enjoyed this content about the decisions made by Oceangate, the mindset of Stockton and the detailed whole engineering explanations, then you should consider getting my book. I will let YouTuber Jeff Ostro, another Titan disaster expert, tell you about the book. Links to the book and my other Titan YouTube videos are in the description. Please consider subscribing, and thank you for your time.
Oh, and by the way, I highly suggest you pick yourself up a copy of this nice paperback book here. This is called Titan to the Titanic and Back 13 and a half times by Gordon Telipan. And what I really like about this book, it has everything you could ever possibly want to know about the Titan from the history of the company all the way to the end to the last day of the implosion. And um this is about 300 pages. You can also get the Kindle version and download that. And you can also get a PDF version which is really cool because you can search. So, this right here will save you from having to sift through hundreds and hundreds of hours of videos all over the web because all of the answers and all of the key points that you need to know about this entire engineering disaster is right here in the book, including all of the mistakes made during the manufacturing of the Ocean Gate Titan submersible carbon fiber hull. And what I find especially useful is these key talking points that witnesses made during the Marine Board investigation there. And it it's got the timestamps from the videos, too. So, you can go directly to the US Coast Guard's live streams from the Marine Board investigation, and you could directly find those timestamps in there and listen to the conversations that were had.