OceanGate Titan Disaster Occasions Misinformation About Carbon Fiber Composites

SELF-DECLARED "EXPERTS" ON SOCIAL MEDIA AND IN THE NATIONAL PRESS ARE CRITICIZING THE USE OF CARBON FIBER COMPOSITE IN THE SUBMERSIBLE

Jun 24, 2023

With a bevy of talking heads chiming in about the cause of failure in the Titan’s structure, there is a danger that misinformation about carbon fiber will unfairly degrade its perceived place in modern boatbuilding.

Look, I am certainly not a deep-diving submersible engineer or any kind of expert on ocean research HOVs. I am, however, a 30-plus year veteran of yacht and small-ship building. I am, as well, pretty familiar with carbon fiber composite construction in vessel shells, masts, and other boat parts. And what I do know is that the past few days have seen more drivel spoken and written about carbon fiber composites than during the previous ten years in total.

In particular, I’ve seen it asserted that carbon fiber should not have been used in the Oceangate Titan because:

“Carbon fiber is not as strong as steel or titanium.”
“Carbon fiber might be passable for pressure vessels where the load is on the inside (as in a pressurized dive tank), but insufficient for pressure vessels where the load is on the outside of the structure (e.g., for a submersible).
“Carbon fiber composite material tends to delaminate under pressure.”
“Carbon fiber composites are brittle and exhibit a tendency to fail suddenly and catastrophically, unlike ductile metals which first distort visibly.'‘

Kryptonite be damned, Carbon Fiber Man is stronger than the Man of Steel

Comparing the “strengths” of various materials is not butt-head simple; it requires a number of qualifiers. First and foremost, it’s important to understand that the strength of a material is expressed in unit strength per unit of cross-sectional area — for example, in lbs per sq inch (PSI).

Second, the strength of a material can be the same or different when it is loaded in tension (i.e., when the load works to stretch it) versus when it is loaded in compression (i.e., when the load works to squeeze it).

Third, some materials, such as steel and titanium, are “homogeneous”, that is, they are of uniform composition throughout and cannot be separated into different materials. As well, their mechanical properties are the same in all directions. However, some non-homogeneous materials — specifically plastic composites — are made up of long-strands of reinforcing fibers set into a polymerized (plastic) matrix. These materials exhibit differing strengths and other mechanical properties, depending on whether the primary loading is aligned with the axes of the majority of their reinforcing fibers or applied at some angle oblique to those axes. In general, carbon fiber composites fall into this latter group and, therefore, require very careful and astute engineering if one is to take full advantage of their mechanical properties.

Last but not least, because some materials are much denser than others with which they may be compared (i.e., they weigh significantly more per cubic unit, say, per cubic foot) a reasonable comparison of strength, whether in tension or compression, requires relating load carrying ability to unit weight, say, lbs load per lb of structural weight.

The example I like to use for this is a comparison between steel and aluminum. Steel is a lot stronger per square inch of cross-sectional area than aluminum, but it is also a lot heavier per cubic foot — 490 lbs per cu foot for steel vs 169 lbs for aluminum. Thus, aluminum weighs about one third what steel does per unit volume. On average, steel is about twice as strong as aluminum per unit of cross-sectional area. Consequently, the rule of thumb is that, for a given load, you can build an aluminum structure at about two-thirds the weight of a steel one, all other factors held constant.

Okay, you say, but how does carbon fiber fit in with this?

Carbon fiber composite carries about five times the load as steel per unit weight, when that load is properly engineered to align with the majority of the reinforcing fibers. (And provided, of course, that the polymer matrix is an appropriate epoxy plastic.) So, a carbon fiber compression hull skin in a deep diving submersible could be five times stronger for the same weight as a steel one. Moreover, it could also be five times thicker — which is, in itself, beneficial because structural rigidity is significantly enhanced by thickness.

It’s being asserted in the current public discussion of the Titan disaster that CF composite may perform well in tension but not so well in compression. The point of the assertion seems to be that the hull of deep diving submersible is loaded in compression and so is a misapplication for carbon fiber composite. But although the assertion is true — in a sense — it’s also misleading. CF composite is somewhat less strong in compression than CF composite in tension, but it is still stronger than steel per unit weight whether in tension or compression.

Some argue that CF composite in compression will tend to delaminate (have its layers come apart from one another) because the compressive loading is taken primarily in the epoxy polymer matrix. However, I personally doubt that the CF reinforcing goes entirely unloaded when the composite is in compression, and I for one would want to see systematic testing to prove that claim before giving that theory any credence.

Well, what about titanium and the fly in the ointment?

Titanium has about the same strength as steel per unit of cross-sectional area but is 45% lighter per unit of cross-sectional area. Thus, like aluminum, titanium has a higher strength to weight ratio than steel. Titanium does have other valuable properties, such as exceptionally high resistance to corrosion, which needs to be counted in a corrosive environment such as seawater. Titanium is also very, very expensive.

The fly in the ointment — the Achilles heel, the slip twixt the cup and the lip, whatever — in all of this is the seriously big problem of attaching metal components to CF plastic composite parts. Put bluntly, it’s a royal pain in the butt.

I know this from first-hand experience as one time I had to install a couple of CF Cardan shafts on a set of water-jet drives which were being driven by a pair of MTU 16V-4000 engines (4,500 BHP each). The main part of the Cardan shaft was a carbon fiber tube (CF for maximum stiffness, minimum weight and rotational momentum/vibration) onto each end of which a stainless steel, constant-velocity universal joint had to be installed. There was so much torque involved we could not rely on adhesive bonding alone, so the CV joints had to be mounted on a short length of stainless steel tube that was a sliding fit onto the outside of the CF tube. And we drilled to close tolerances and bolted right through the two concentric tubes. The procedure was costly and time consuming, both to engineer and to execute properly.

Carbon fiber composite is not a ductile material. This means it can also be “notch sensitive” — that is, it does not stretch at the leading end of a developing fissure in response to stress (deformation under load), to relieve stress at that leading point of a developing fracture. With the result that, when CF laminate finally gives way, it does so in a sudden and dramatic way. Tjerefore, you have to be exceedingly careful to avoid creating stress concentration points when drilling and cutting CF laminate for the installation of metal fittings. And the metal fittings themselves have to be very carefully engineered and designed.

Of course, none of this sheds any conclusive light on what the cause of the Titan’s catastrophic structural failure was. Some pundits have ventured that an inadequately depth-rated port light (fixed window) gave way. But again, I have yet to see any convincing evidence of that. The point is there are plenty of other possibilities having to do, for example, with the attachment of the Titanium pressure hull ends to the carbon fiber main hull tube — something I believe would have been exceedingly difficult to do properly with anything approaching a low probability of failure.

It’s likely, we’ll never know with any certainty what the precise proximate cause of the disaster was. But I, for one, do not for a second believe the initiating cause was a failure in the carbon fiber laminate. Rather, I think it significantly more likely that the CF pressure hull skin was breached elsewhere at a site of an installation of a metal fitting or fittings, and that this led, in a chain reaction, to a catastrophic implosion. And I sincerely hope that carbon fiber composite technology will not be unfairly besmirched, as it was following the 1979 Fastnet Race disaster, when several CF rudder stocks failed badly in the storm primarily because yacht designers and boatbuilders did not, at that time, yet fully understand the ins and outs of carbon fiber laminates.

As to the captain and crew of the Oceangate Titan, whatever mistakes may have been made, they paid the price in full. May they rest in peace.

— Phil Friedman

And if you have another minute or two, please share this post with some of your friends and connections, or leave a comment — or both.

Discussion about this post

Ready for more?