The first pressure vessel was designed and fabricated in collaboration with the University of Washington Applied Physics Lab, and Spencer Composites and constructed entirely from carbon fiber, including the end caps.
The pressure vessel was approximately 1/3 of the size of the full-sized vessel. The walls of the scale model were made of multiple layers of carbon fiber material and resin. The vessel was assembled from three main components: a center cylinder, and two dome end caps. Fabrication of the end caps required 28 unique winding patterns of the carbon fiber material to form a spherical shape. After the carbon fiber cured, the sphere was cut in half to form the two dome end caps.
The vessel was lowered vertically into the pressure test chamber at the University of Washington and suspended by way of a bridle. Ballast was hung from the bottom of the vessel to prevent it from floating and contacting the pressure chamber end cap once the chamber was filled with water.
The test was conducted in a series of steps where the pressure was increased to a pre-set level and held at that pressure for a period of time (dwell time) to monitor results. Nineteen strain gauges were installed inside the pressure vessel and monitored throughout the test. The initial test plan included 9 steps with a dwell time of 5 to 10 minutes at each step.
At a pressure of 4,285 psig (the pressure at approximately 10,000ft depth) one of the hemispherical end caps failed just outboard of one of the large stainless-steel inserts. The end caps were thought to be the most likely area of failure as the use of carbon fiber for hemispheres such as these really pushed the boundaries of carbon fiber design and fabrication.
Following the failure, the team immediately removed the model from the test chamber and, within hours, the two end caps were flown to Spencer Composites -- the manufacturer of the carbon fiber vessel, to modify the design and rebuild carbon fiber endcaps for further testing.
The OceanGate team, along with engineers from Spencer Composites and the University of Washington Applied Physics Lab conducted a detailed analysis of the model and the test data. Despite the setback on the domes, the test provided valuable proof that filament wound carbon fiber hemispheres are viable to depths of 10,000 feet and demonstrated that the benefits of carbon fiber are real.
The second pressure vessel was also designed and fabricated in collaboration with the University of Washington Applied Physics Lab Collaboratory, and Spencer Composites. It was comprised of three main components: a center cylinder, and two 6" thick aluminum end plates. Unlike the first pressure test where we tested a scale model with carbon fiber hemispheres on each end of the cylinder, replaced the carbon fiber endcaps with 6" thick aluminum end plates in order to independently test the cylinder.
The pressure vessel was successfully tested to 6000 psig, the equivalent of 4200 meters.
The third vessel used the newly constructed carbon fiber domes from Spenser Composites for the endcaps and the same hull that was used in Test 1 & 2 the team conducted a third test to validate that the carbon fiber model is capable of withstanding an external pressure of 9,500 psig -- corresponding to operating in the ocean at a depth of about 6,685 meters (21,935 feet).
As with the first two tests, the model was lowered vertically into the pressure test chamber and suspended by way of a harness. Lead weights are hung from the bottom of the model to prevent it from floating and contacting the pressure chamber lid when water is added to the chamber. The chamber was partially filled with water and the pressure test began.
The test was conducted in a series of steps where air was pumped into the chamber to increase the internal pressure to pre-set levels, and then held at each level for a period of time (dwell time) to monitor results. During the test, the pressure inside the scale model remained at 1 atmosphere (14.7 psig) -- the air pressure we experience at sea level. During the test, the team assessed the effects of pressure on the model by monitoring strain data to detect any deformation of the carbon fiber.
Results showed that while the hull was unaltered, the end caps comprised of 2 carbon fiber hemispheres were not capable of withstanding the immense pressure. As a result, the team decided to proceed with titanium for the forward and aft domes on the full-size pressure vessel.
The acoustic monitoring system we developed was able to detect all events uncovered by the Boeing system and many more small events, thus giving us high confidence in our new system.
The primary objective of fourth and final scale model test was to validate that our proprietary health monitoring system and its ability accurately predict failure of the carbon fiber pressure vessel when subjected to pressure long before a catastrophic event.
To achieve the objective, we intentionally increased the pressure inside the test chamber until the carbon fiber pressure hull failed. It was only by breaking the model that we could confirm that the acoustic monitoring can reliably predict failure. Because our scale model had previously experienced two rapid decompressions during two of the first three tests, we expected to see structural failure during this test at a lower pressure than we otherwise would have.
As with the first three tests, air was pumped into the chamber to increase the internal pressure to pre-set levels, and then held at each level for a period of time (dwell time) to monitor results. During the test, the pressure inside the model remained at 1 atmosphere (14.7 psig) -- the air pressure we experience at sea level.
We tested a number of threshold levels to eliminate data noise and only show data to display meaningful information illustrating the hull status. We saw only low signals until we transitioned beyond 5,000 psig. The noise levels we saw are displayed on a relative scale and merely represent the digital value of an analog input. The maximum level (aka the saturation level) was 32,000 units. Prior to leaving 5,000 psig most events were between 2,000 units and 4,000 units (~6%-12% of saturation). Once we left 5,000 psig, we saw 12,000+ level acoustic events and then some at full saturation (32,000). The data is consistent with the lower level values representing resin adjusting and the higher numbers indicated snapping of the carbon fibers themselves.
Once we saw the large events, we slowed down our test stair-step pressurization process. The hull was very quiet after a long dwell (hold time) at 6,000 psig so we continued to 6,500 psig – equivalent to 4200 meters deep (15,000 feet). After about 2 minutes at 6,500 psig, the number of events had not declined, as we had previously seen during dwells at lower pressures but seemed to increase - and then we heard the loudest bang imaginable as the hull imploded. The energy released equated to ~ 1.4 sticks of dynamite in roughly 0.5 seconds.
When we removed the cylinder and examined the failure point it was clear it occurred at the “sticky” joint between the cylinder and aluminum slab and then permeated down the cylinder.
The time from the first group of large acoustic events to failure was over 35 minutes and a change of 1,500 psi (the equivalent of approximately 3,500 feet of depth). This significant advance warning is exactly what we had planned for and gives us confidence that we can rely on this real time system for monitoring hull integrity throughout every dive.
Real-time health monitoring system makes it possible for the team to take a stair-step approach to test diving Titan. Over the course of the next several weeks, the OceanGate operations team will conduct a series manned dives, starting in the Port of Everett Marina and slowly working towards the deepest areas of the Puget Sound. Upon completion of the shallow dive testing, operations will mobilize to the Bahamas for deep sea validation where the team will continue an incremental dive approach to 4000 meters. Our CEO, Stockton Rush, will be the first to pilot Titan to 4000 meters.
Conduct continuous and extensive site surveys to inspect underwater infrastructure, wrecks or sensitive environmental habitats without resurfacing.
Collect research data in real-time with first-hand views, onboard collaboration and the flexibility to modify your mission profile while still on site.
The ultimate stage to film the treasures of the deep in a highly adaptive vessel designed to illuminate the depths, capture vibrant images and document natural habitat.
A unique underwater testing platform to conduct a wide array of experiments to test equipment or expedite sensitive research in deep ocean environments.