Related Articles
Wreckage of the Titan’s innovative carbon fiber hull was found separated into three distinct layers, US National Transportation Safety Board engineer Donald Kramer has told a Coast Guard hearing into the fatal implosion of the OceanGate submersible in 2023.
Although Kramer would not offer an opinion on what caused the hull to delaminate into separate layers, he testified to multiple problems with the hull, beginning with its manufacture in 2020.
Using samples of carbon fiber saved from its construction, as well as dozens of pieces recovered from the seabed, the NTSB gave the most complete picture to date of the experimental nature of the Titan’s hull.
After the Titan’s first hull was found to have a crack and delamination following deep dives in 2019, OceanGate switched manufacturers to replace it.
The new manufacturer, Electroimpact, used a multistage process to wind and cure the five-inch-thick hull in five separate layers. Each layer would be baked at high temperature and pressure before being ground flat, having an adhesive sheet added, and another layer built on top. The idea of this multistep process was to reduce wrinkles in the final hull that the company believed had caused test models to fail short of their design depths.
However, Kramer testified that the NTSB found several anomalies in the fresh hull samples. There was waviness in four of the five layers, and wrinkles that got progressively worse from layer to layer. The NTSB also found that some layers had porosity—gaps in the resin material—four times larger than specified in the design. It also recorded voids between the five layers.
On Monday, Roy Thomas, a materials expert from the American Bureau of Shipping, told the hearing: “Defects such as voids, blisters on surface, and porosity can weaken carbon fiber, and under extreme hydrostatic pressure can accelerate the failure of a hull.”
OceanGate did not make any additional test models using the new multistage process.
The NTSB was able to recover many pieces of the carbon fiber hull from the seafloor, one still attached to one of the submersible’s titanium end domes. In a report issued simultaneously with Kramer’s testimony, the NTSB noted that there were few, if any, full-thickness hull pieces. All of the visible pieces had delaminated into three shells: the innermost of the five layers, a shell made of the second and third layers, and another with the fourth and fifth layers. Like an onion being peeled, the hull had largely separated at the adhesive joining the layers.
Most PopularGearThe Top New Features Coming to Apple’s iOS 18 and iPadOS 18By Julian ChokkattuCultureConfessions of a Hinge Power UserBy Jason ParhamSecurityWhat You Need to Know About Grok AI and Your PrivacyBy Kate O'FlahertyGearHow Do You Solve a Problem Like Polestar?By Carlton Reid
The scene was set for Kramer on Monday, with testimony from Phil Brooks, OceanGate’s director of engineering until March 2023. Brooks was asked about a loud bang that had been heard when the Titan surfaced from a successful dive to the Titanic in July 2022.
About 10 minutes after the submersible surfaced, its crew and passengers heard a bang “as loud as an explosion,” according to a witness last week. When the Titan returned to its support ship, Brooks downloaded data from acoustic sensors and strain gauges mounted on its carbon fiber hull.
Five of the eight acoustic sensors captured the noise. The other three showed nothing and were likely nonfunctional, according to Kramer. On Monday, NTSB investigators noted that the acoustic sensors had not recorded useful data for several preceding dives.
The strain data, measuring stretching in the hull, also showed a shift coinciding with the bang. Brooks had seen a similar but larger change in strain just before an earlier scale model of the Titan’s hull failed dramatically in testing. “A sudden shift is something to be concerned about,” he testified. “But it was fairly minor, on two out of the 16 strain gauges.”
Brooks also admitted, however, that three of Titan’s strain gauges had never worked since it was rebuilt and that he was “probably not” qualified to decide on the significance of the change in the strain data. Brooks trained as an electronics engineer and computer scientist, and he deferred on mechanical issues to CEO Stockton Rush, one of five people who died in the Titan.
Brooks said that Rush believed that the bang originated in an external metal frame that held the hull and that Rush gave the go-ahead for future dives. “Without my being a qualified mechanical engineer, his explanation sounded plausible,” said Brooks.
Most PopularGearThe Top New Features Coming to Apple’s iOS 18 and iPadOS 18By Julian ChokkattuCultureConfessions of a Hinge Power UserBy Jason ParhamSecurityWhat You Need to Know About Grok AI and Your PrivacyBy Kate O'FlahertyGearHow Do You Solve a Problem Like Polestar?By Carlton Reid
One former OceanGate engineer told WIRED they thought the bang could have been layers in the carbon fiber parting or the hull separating from the titanium end rings: “The degree of the separation can’t be known, but it could have been the beginning of the end.”
The Titanic dives continued without Brooks or Rush doing any further investigation of the noise or consulting with outside experts.
“We didn’t see any further shifts in strain data. No loud bangs,” said Brooks. “Nothing with the acoustic monitoring that seemed out of the ordinary.”
In fact, Kramer’s own analysis of strain data showed that the changes persisted for at least the next three dives. The NTSB did not have sensor data from the Titan’s 2023 missions.
At the end of the 2022 dives, Brooks lobbied to bring the Titan back to OceanGate’s headquarters for testing. “We really wanted to bring the sub back to Everett, Washington, and look for cracks,” he testified. “It was frustrating because it was left at St. John’s [in Newfoundland] at the dock. We were told it was cost-prohibitive. They were low on money.”
A maintenance log for the Titan shows no repairs or adjustments were made to the submersible between late July 2022 and late March 2023.
The hearings continue this week, although it will likely be many months before the Coast Guard issues its final report into the accident.
