Following the loss of the OceanGate Expeditions Titan tourist submersible, the company’s use of carbon fiber composite material to construct the sub’s pressure hull has come into question, as the industry standard has been to make them from titanium.
But a Rhode Island company that specializes in building carbon fiber pressure vessels for submersibles has compiled an impressive record of safety and reliability that says the material can be used in this application, as long as the vessel is designed and tested properly.
Composite Energy Technologies (CET) provides carbon fiber pressure vessels to commercial and government customers such as the Office of Naval Research that have never failed in their dives to much deeper sites than Titanic, said president Chase Hogoboom in an interview with Design News.
Deep-sea veteran and University of Rhode Island professor of oceanography Bob Ballard has been making trips to the bottom of the ocean since taking the U.S. Navy’s Alvin submersible down in 1960. He is best known for having located the site of the sunken RMS Titanic.
“We’ve made thousands and thousands of dives with different vehicles, whether they’re French, whether they’re Russian, whether they’re American, and we’ve never had, ever, in the history of these extreme deep-diving programs, ever lost a vehicle,” Ballard noted in an interview broadcast by ABC News. “So this is a first.”
Every one of those vessels that have compiled the record of 100 percent safety was made of titanium. “Naturally you go to, ‘How does this vehicle differ from the vehicles we’ve been using for many years?’” said Ballard. “It did have a very experimental hull and obviously that hull imploded. This has never happened in our deep submergence world, beginning in 1960, he said.
While Ballard is contemplating an indictment of carbon fiber in the fatal implosion of the Titan’s pressure hull, filmmaker and deep-sea explorer James Cameron seems ready to convict the material. “I was dead-set against any use of composites for an external pressure vessel,” he said in the same ABC News broadcast with Ballard. “They’re great for making vessels with internal pressure. You can make scuba tanks out of composites all day long. That’s kind of what they’re great at,” he said. “But for external pressure, they’re a terrible material.”
But CET has already proven that carbon fiber composites can be used safely in this application. “I’ve heard Cameron’s comments,” said Hogoboom. “That’s just not the case,” he asserted.
Carbon fiber is better in tension than in compression, Hogoboom concedes, but with appropriate design and manufacturing, it is an effective material for both, he said. “It is a more efficient material in tension, but it works very well in compression when used correctly.”
Of particular concern is cycle fatigue, which could be why Titan failed only after making repeated dives and then failing at a depth it has survived previously.
“We all knew that the danger was delamination and progressive failure over time with microscopic water ingress and ... what they call cycling fatigue," Cameron said in an interview with Reuters. "And we knew if the sub passed its pressure test it wasn't gonna fail on its first dive ... but it's going to fail over time.”
Safe Track Record
But with appropriate construction, that’s not a problem, CET asserts. To prove it, Hogoboom points to the record of his company’s products in deep-sea applications. “We’ve built vessels that we’ve cycled 200 times (to deep-sea pressures) and then brought to implosion and those fail at the same depth as new ones.”
The key is diligence in designing and testing the composite structures, Hogoboom explained. “We have a very high confidence in the strength of what’s been built,” he said. “We use engineering models, but we test to failure to validate what’s been modeled. That’s a crucial step that OceanGate has skipped," according to Hogoboom. “They never brought an exact clone to failure.”
Some important aspects of the manufacturing process include the layout and fiber orientation of the carbon fiber cloth used. But as Design News learned during our experience laying up carbon fiber at Lamborghini, manufacturing variances can undermine the intended results. “Those variances can determine the ultimate strength of the item,” he said.
Composite Energy Technologies
Performing the work consistently and then testing the resulting parts ensures dependability, according to Hogoboom. CET works with the Woods Hole Oceanographic Institution in Massachusetts to test most of its products, though it tests its very large Unmanned Underwater Vehicles (UUVs) that are of a similar size to the Titan sub at the U.S. Navy facility in Annapolis, Md. “We’ve built carbon UUVs that are nearly 30 feet long with a large diameter that we’ve tested many times in lab facilities with all the equipment running. There’re ways to do this in a controlled environment.”
The company also has real-world results to back up those tests. “There are also examples of carbon UUVs with over 6,000 operational hours at over 6,000 meters,” said Hogoboom. “That’s in the real-world environment after many cycles.” As a result, CET delivers carbon fiber submersibles that reliably work at deep depths. “I feel very comfortable with what we do but it has taken a lot of testing to build up that confidence,” he said.
OceanGate, meanwhile, said it was conducting rigorous safety testing, but the company seems not to have tested to failure. Cameron outlined his own company’s design and testing program as a contrast, telling ABC News “We spend so much time and energy on that and we use all the computerized tools available today, finite element analysis. We worked on our sphere for our deep-sea vehicle that went to the Challenger Deep for over three years, just in the computer, before we even made the thing. Then of course we pressure tested it over and over and over. So this is a mature art and many people in the community were very concerned about [Titan].
Guillermo Sohnlein, co-founder of OceanGate Expeditions but no longer with the company, told Times Radio, “I know from first-hand experience that we were extremely committed to safety and safety and risk mitigation was a key part of the company culture.”
OceanGate has a 2019 video on its YouTube channel in which CEO Stockton Rush, who died aboard Titan in the accident, describes the company’s safety measures. “All of our test program has been about incremental testing we started over two years ago with (submersible) Cyclops One, which allowed us to test our launch and recovery system, the launch platform, and a number of really important operational and electronic processes -- motor controls things like that out here. This is really focused on one thing and that's the pressure vessel and making sure that that component, which is clearly the most critical component of the sub, is safe and capable of handling depths down to 4,000 meters repeatedly with people on board."
However, OceanGate doesn’t seem to have taken the step of having Titan certified as capable of doing that. “There were a number of people who signed a letter to OceanGate recommending that they not take passengers down unless they get certification from one of the major certification agencies, like the American Bureau of Shipping or the Norsk group or Lloyd’s, one of the respected certification agencies, and that idea was not heeded,” said Cameron.
“The certification protocols that other deep submergence vehicles except this one, that carry passengers, especially paying passengers, all over the world in tropical waters, deep coral reefs, other wreck sites, and so on, the safety record is the gold standard, absolutely,” he added.
Rush also declined to crowdsource know-how from the deep-sea community, which is described as being tight-knit and communicative. “When you’re working in these specialized scientific environments, it is a supportive community that is interested in sharing and learning from each other,” explained Hogoboom. Rush passed on the opportunity to tap into the others’ knowledge and experience, he said.
“I had some email exchanges with that guy Rush a handful of years ago,” he recalled. “I thought there would be some interest in some of the data we had collected. He seemed interested but not enough to come by and visit.”
There was another missed opportunity more recently, according to Hogoboom. “He had the OceanGate sub at the University of Rhode Island (which is near CET’s headquarters) a couple of years ago. In advance of that visit, I shared with him some of the work we were doing in that space. It was a bit of a surprise that he wasn’t interested.”
For example, in addition to testing some of its pressure vessels to failure, CET performs thermal imaging and ultrasonic scans on every part to verify that it has been manufactured properly. Then the company provides the details of those scans to its customers so that they can perform the same checks at regular intervals in the vehicle’s life to test for evidence of the cycling fatigue that worries Cameron. “It can be tested against the baseline when it left the factory,” Hogoboom said. So far, they haven’t found any changes over time.
In contrast to this system, OceanGate employed a real-time monitoring system that Rush patented to watch for signs of trouble. Evidentially this system did provide an alert, because according to Ballard, Titan had dropped its ballast weights and was ascending, apparently in response to the alarm, when it imploded. The U.S. Navy subsequently confirmed to the U.S. Naval Institute that its underwater listening network recorded the sound of an implosion at that time. “I heard it had this health monitoring system,” said Hogoboom. “I’m not sure of the effectiveness of that if you are operating on such thin margins.”
Rush’s patent promises that his system is capable of “validating the ‘health’ of a material or structure in real-time, determining alarm conditions, predicting failure conditions, and the like.” In practice, it seems to have not provided sufficient warning time. “This OceanGate sub had sensors on the inside to give them a warning when it was starting to crack,” said Cameron. “I think if that’s your idea of safety then you’re doing it wrong.”
In addition to the possibility of a failure of the carbon fiber hull, Titan’s implosion could have occurred due to a failure at the joint between that hull and the titanium cap that bolted the passengers inside.
“You’ve got two different materials that expand and contract at different rates,” pointed out Hogoboom. “That would be concerning to me, especially since an O-ring seal is part of that assembly.”
Crewed or UUV
A crucial difference with CET’s pressure vessels is that they are used only for unmanned underwater vehicles, not ones that carry people. As with the launch of expensive satellites, the cost of failure of such vehicles can be extreme, so even though no lives are at stake, the company and its customers take the utmost care to ensure safety and so far that has resulted in a record of success, Hogoboom said.
“We do take the matter very seriously,” he said. “These missions are very expensive. There is the equipment, people paid to deploy it, and also the potential to lose the information gathered. To have something fail that is part of a very expensive effort, it is certainly a lot of responsibilty.”
Despite his bullishness on carbon fiber and CET’s rigorous test regime, the company isn’t ready to put people inside its pressure vessels, he said. “They are for underwater housings for equipment. We’ve never put a person inside. We have a ways to go before I would feel comfortable doing that.”