By the time the final reports landed, the story of Titan had already become one of those rare disasters that people think they understand long before the engineering stops being guessed at and starts being documented.
The U.S. Coast Guard released its Marine Board of Investigation report on August 5, 2025, and the NTSB followed with its final report on October 2, 2025.
Together they stripped away the mythology and left behind something much colder and more useful: a chain of mechanical choices, missed warnings, and avoidable assumptions that ended in the pressure hull failure and implosion on June 18, 2023, about 370 nautical miles southeast of St. John’s, Newfoundland and Labrador.
If you want the simplest version of the answer, it is this: Titan did not fail because carbon fiber is automatically bad.
The reports are much more specific than that. The NTSB found that OceanGate’s engineering process was inadequate and that the company failed to establish the actual strength and durability of the Titan pressure vessel.
The result was a carbon fiber composite pressure vessel with multiple anomalies that did not meet strength and durability requirements, and the NTSB’s probable cause ties the hull failure to that inadequate process, plus flawed analysis of the pressure vessel monitoring data that allowed a damaged hull to keep operating.
That matters because the material choice by itself was never the whole story. OceanGate’s own stated design goal, as summarized by the NTSB, was a submersible rated to 6,000 meters with a 2.25 safety factor and a 10,000-cycle life limit.
In other words, the company was not trying to build a toy. It was trying to build a pressure vessel for repeated deep-ocean dives, and it set a target that, on paper, implied a substantial margin.
The trouble was that a goal on paper is not the same as a tested vessel in the real world.
The Titan never went through the kind of full validation and failure testing that would have established what the finished structure could actually withstand over time.
The pressure hull itself was a carbon fiber composite cylinder built up on a steel mandrel as five co-bonded layers, each nominally one inch thick.
Each layer consisted of 133 prepreg plies laid down in a repeated sequence, and the finished cylinder was then joined to titanium segments and domes.
This was not a simple monolithic tube. It was a layered structure with adhesive interfaces, and those interfaces turned out to matter enormously.
When investigators examined the wreckage, they found delaminations between layers 1 and 2 and between layers 3 and 4, and the separated interfaces showed signs that the layers had been rubbing against each other before the final dive.
The physical evidence showed that the cylinder had not been born perfect and did not remain perfect as it aged.
Investigators found porosity between plies, voids in the adhesive, wrinkles and waviness in the carbon fiber, and areas where wrinkles had been ground flush during manufacturing.
They also found that the average porosity content of the composite cylinder was about 2.7 percent.
That may sound like a small number until you remember that these are not decorative layers.
They are load-bearing layers in a structure meant to survive enormous external pressure. The NTSB concluded that those anomalies would have reduced compressive and shear strength, weakening the hull from the inside out.
The most important clue, though, was not hidden in the wreckage. It was hiding in Titan’s own data.
OceanGate had a real-time monitoring system on the vessel, which used eight acoustic emission sensors and eight hoop strain gages and eight longitudinal strain gages to watch the hull during dives.
But the system was already impaired. Three of the eight acoustic sensors did not register acoustic events during any of the 2022 dives.
That meant the monitoring was incomplete before the critical data problem even began. The decisive warning appears to have come on dive 80, on July 15, 2022.
OceanGate personnel on the Titan and on the surface heard a loud bang, and the RTM system recorded a burst of acoustic activity and a sudden jump in some strain outputs.
When the NTSB replotted the data the way an engineer should, strain against depth rather than strain against time, the pattern changed.
Through dive 80, the response had been linear. After dive 80, the low-depth response became nonlinear on subsequent dives, which is exactly the kind of fingerprint investigators associate with delamination.
Near the surface, separated layers can behave differently; under deep pressure they can squeeze back together and hide the problem.
The data was telling them the hull had changed. The problem was that OceanGate was not reading it in the way that best revealed the change.
The reports are even more direct on the point that dive 80 likely damaged the pressure vessel.
The NTSB concluded that Titan likely sustained one or more delaminations at the end of dive 80, caused by voids between the cylinder’s five co-bonded layers, and that those delaminations led to deterioration and weakening.
Titan then completed dives 81 and 82, which means the damaged hull kept going back down to Titanic depth after it had already shown the first detectable sign of internal failure.
The NTSB further says that after dive 82 the Titan sustained additional damage of unknown origin, and that this further deteriorated the vessel until the casualty dive, dive 88, when the local buckling failure led to the implosion.
That sequence is one of the most important parts of the story because it moves the failure from a single dramatic moment to a long, cumulative decline.
Titan successfully dove to the Titanic wreck site 13 times, but by the time of dive 88 the hull had been through enough cycles, enough damage, and enough uncorrected warning signs that its structural condition had worsened past the point of survival.
The final dive on June 18, 2023, reached a last known depth of 3,363 meters before the pressure hull failed.
The environment around the hull may have made things worse. The NTSB notes that after the 2022 expedition Titan was initially left uncovered in a parking lot in St. John’s, and later in 2023 it was moved indoors for the next expedition.
The board also points out that for the 2023 expedition the Titan and launch-and-recovery system were towed behind the support vessel Polar Prince rather than transported on deck as they had been on the Horizon Arctic in previous operations.
During that open-ocean tow, the Titan would have been exposed to about 2,900 miles of vibrations, impulse loads, and upset events, and the NTSB says additional damage after dive 82 may have come from unknown origins during that period.
The board does not claim it can prove exactly which environmental factor caused which final damage, but it is clear that the hull’s condition did not improve between storage, towing, and the next deep dive.
That is why Titan’s real failure story is so frustrating. It was not a case of one defective thing in isolation.
It was an engineering program that never fully closed the loop between design, testing, monitoring, storage, and operations.
OceanGate did not build and test a composite sub-scale model made with the same co-bonding process it later used in the full-scale hull, so it had no data on how that specific structure would deteriorate and fail.
The NTSB says OceanGate also never performed a cycle-life evaluation that would have established inspection intervals, the number of safe dives, or the warning signs that the hull was beginning to weaken.
Without that kind of baseline, the company had no reliable way to know what its monitoring data actually meant.
The vessel’s regulatory status mattered too. Titan was not registered under any flag state, was not classed by a classification society, and was not inspected by the U.S.
Coast Guard. The NTSB found that because OceanGate operated Titan with passengers for hire, it should have been inspected and should have complied with the applicable small passenger vessel rules.
Instead, it operated outside that framework. The Coast Guard and the NTSB both use the post-accident review to push for stronger, more consistent international rules for pressure vessels for human occupancy.
The Coast Guard’s report was released in August 2025, and the NTSB’s report recommends that the U.S.
Coast Guard propose that the International Maritime Organization make the existing PVHO guidance mandatory so that submersibles operating across borders are subject to consistent standards.
The final reports also make clear that OceanGate’s monitoring system was not the safety net it was supposed to be.
The RTM system stored acoustic emission and strain data during dives, but OceanGate’s interpretation of that data was flawed.
The NTSB says the company analyzed strain side by side with time rather than plotting it against depth, which obscured the problem.
When investigators used the depth-based approach, the nonlinear changes after dive 80 became obvious. The board’s language is plain: OceanGate’s flawed analysis of the monitoring system contributed to the continued operation of a damaged pressure vessel.
Even the search and rescue side of the story is treated as a serious engineering lesson in the reports.
The NTSB concluded that OceanGate did not notify search and rescue assets about its planned expedition, but despite that the U.S.
Coast Guard’s response coordination was effective and led to the timely discovery of the wreckage.
NOAA’s passive acoustic recorder, about 900 miles from the implosion site, later recorded the suspected acoustic signature of the event, which the Coast Guard made part of the investigation record.
That means the chain of evidence did not come from one sensor or one witness.
It came from a layered investigation that gradually locked the timeline into place. If there is a short version of the lesson, it is that Titan was not destroyed because carbon fiber is inherently evil.
It was destroyed because the actual finished hull never received the kind of engineering rigor needed to prove its strength, its fatigue behavior, and its failure mode before people were put inside it.
The NTSB says the company’s process was inadequate, that the hull contained multiple anomalies, and that the flawed interpretation of monitoring data allowed a damaged vessel to continue being used.
The final reports do not leave much room for mystery on that point. They are not describing a freak event so much as the end state of a system that had been deteriorating in visible and invisible ways for months.
There is a reason the reports spend so much time on design margins, inspection intervals, delamination, porosity, storage, towing, and data interpretation.
These are the places where engineering lives or dies. The hull’s material, the report says, was not the whole problem.
The whole problem was the process around it. That process missed the build defects, missed the significance of the acoustic event on dive 80, missed the meaning of the strain changes afterward, and missed the accumulated risk of continuing to dive a vessel that had already shown evidence of internal damage.
By the time Titan reached its 88th dive, the investigation concludes that the structure could not survive another descent to Titanic depth.
The final reports leave us with a difficult but useful answer. Titan was not a mystery that science could not touch.
It was an engineering failure that science, had it been properly applied, could have exposed earlier.
The tools existed. The warning signs existed. The problem was not lack of physics. It was lack of discipline.
And that is the lesson the Coast Guard and NTSB both push toward the future of Titanic-bound submersibles: if people are going to descend into the deep in pressure vessels, they need certified structures, proper testing, proper inspection, and data analysis that actually matches the way pressure behaves.
OceanGate’s version of innovation skipped too many of those steps, and the final reports say so plainly.
Disclaimer : This content may be created by AI for entertainment purposes. Any resemblance to real persons, events, or places is coincidental.
