Imagine a conference room with three engineers inside it. One is a chemist who passes her days wondering how molecules meld themselves together. One is a physicist with an ability to describe exactly how a wave of stress propagates through a solid. The third is a mechanical engineer with twenty years’ experience asking one question: when you bolt this part onto an aircraft and fly the thing through a thunderstorm, will it survive?
They are arguing.
Not because anyone is wrong. Because they all are right in a sense, somehow. And that, in one picture, is composite engineering.
Composites are not really materials. They are arrangements. You have fibers, usually carbon or glass or aramid, and you put them in a matrix which is most of the time a polymer resin. The fibers carry the load. The matrix keeps everything in place and also transfers stress between fibers. Simple in theory.
This basically means that every single decision around that arrangement is related to a different area of science. Pick a new resin and suddenly you have presented your chemist with a riddle about the nuclear physics of cure kinetics and crosslink density. Rotate the fiber by 15 degrees, and the physicist has to recalculate every stress calculation because the material is no longer behaving as it was yesterday. If you add a single ply to the layup, the mechanical engineer has to redistribute the weight, change how it will be attached and figure if that new thickness would fit into an assembly tolerance someone wrote on a drawing in 2007.
Composite engineering cannot be really done by one person. You can try. People do try. The end result is typically one that looks great on paper, yet fails.
In nearly every composites project, there comes a point where the chemist and the mechanical engineer disagree about cure temperature. Why the chemist wants it to go higher is that, as a result of doing so, the resin will crosslink more completely and be tougher at end of life. The mechanical engineer wants to keep it lower because the tooling will not survive much heat or else it can warp, and the mundane residual stresses that build up while cooling will at some point quietly kill the part three months from now.
Both are correct. Individually, neither can prevail in the argument. They have to pick a number out of thin air, somewhere in the middle and that only works because the physicist is sitting at the corner running a finite element model showing precisely how the part will deform as it cools. The physicist did not initiate the dispute. The physicist will end it.
Three disciplines. One number. This is how composites actually get designed.
And the neat thing is this arguing makes parts better than any one expert could perfect on their own. The Boeing 787 Dreamliner is approximately 50% composite by weight, representing on average around 70,000lbs. That number is not achieved by someone intelligent sitting in a room by himself. It was found by thousands of engineers from different specializations haranguing for over a decade on questions of resin chemistry, fiber placement, fastener design, lightning strike protection and a hundred other issues that nobody on the team could have resolved without input from someone with a fresh perspective.
This is one of the things that you do not learn in school. The biggest portion of the value an experienced composites engineer brings is translation.
The chemist says the cured resin has a glass transition temperature of 185 degrees Celsius. That resonates with the mechanical engineer, who says “great, you can fly hot.” The chemist is referring instead to 185 being the dry glass transition temperature, and once the resin bathes in humid air on the runway it falls by 30 degrees, sometimes more. So, it would not actually be rated for 185. If you're honest about the environment, it's rated for possibly 150.
Nobody lied. Nobody made a mistake. The chemist went on to provide a technically correct answer to a question the mechanical engineer had no idea how to ask. That gap is where parts fail. Bridging that gap is what really good composite engineering is about.
The same story applies between mechanical engineers versus physicists. It is to determine the strength of a laminate by the mechanical engineer. Its return is a tensor, a multidimensional mathematical object which embodies how the material would behave in every direction at every instant. The mechanical engineer was looking for a single number. The physicist returned twenty-one of them. Someone, normally a third party who's worked both sides of the fence, has to come in and indicate which of those numbers count for this segment on this aircraft.
The most fun part of bringing three disciplines together in the same room is that nothing can be predicted about what comes out.
Take toughened epoxy resins as an example. Decades ago, the workhorse aerospace epoxy was strong but brittle. We expected that dropping a wrench on a composite panel could lead to barely visible impact damage, the type of microcracking that does not appear at the surface yet nevertheless dramatically reduces compression strength. This was the kind of problem that would not be solved behind the bars of a single discipline.
The chemists persisted in varying the resin's molecular structure to toughen it. They made it harder by also making it easier in some other way. The mechanical engineers continued to attempt to design through the issue by simply incorporating more material, which added weight that ultimately undermined the entire purpose of composites. Physicists continued modeling and shaking their heads at the crack propagation.
Then someone (or three or four people, depending on whom you ask) had the brilliant idea to include small rubber particles mixed into the resin. These rubber pieces resemble small shock absorbers that stop cracks before they can develop. It started with a polymer chemistry idea, but it didn't click until a mechanical engineer explained that the failure was due to crack growth because of impact and not pure tensile failure.
It opened up an era of robust composites that can take a beating. It is now standard practice. That does not just come from one field. This arose out of the conflict between domains.
Building the part is harder than designing it.
Each assumption in the design phase hits bricks on the shop floor. The fiber orientations the physicist optimized assume that plies are oriented within plus or minus 2 degrees of the design angle. Even with expert technicians, hand layups can wander to 5 degrees or more in corners and bends. The chemist's ideal cure cycle presumes that the resin temperature is uniform throughout the entire part. In an actual autoclave, for a component with thick and thin sections, the temperature can be 15 to 20 degrees Celsius different among locations. In the mechanical engineer's model of stress, a value of less than 2 percent void content is assumed.
Thus, they do not just debate the design. They argue during manufacturing too. The chemist examines the part for quality. The physicist distresses at every ply drop and fastener hole. The mechanical engineer looks at a part that needs to be manufactured repeatedly by hundreds of technicians on dozens of shifts for the next two decades.
The job is to get all three to agree on what acceptable looks like.
The next generation of composite technologies will make the collisions of disciplines even louder.
Take thermoplastic composites. Thermoplastics fibers are thermoplastic matrices which can be melted and reformed rather than curing once and remaining there forever like traditional thermoset resins. This paves the way for quicker methods to fabricate parts using techniques such as induction welding or out-of-autoclave processing, and it simplifies repair. But it also means the chemistry is totally different, that the processing windows are tighter and that the long-term behavior under load has yet to be established. Every discipline has to unlearn what it thought it knew.
Then we have nano-engineering composites; introducing particles smaller than a virus into the matrix to modify electrical conductivity, thermal performance or mechanical properties. But for now, chemists must converse with materials physicists who study quantum effects. Physicists need to talk to nanotech experts. In the corner of the room sit the mechanical engineers, who must determine how to certify a part with critical features that are too small to be seen with a microscope.
Add in the context of automated fiber placement, digital twins and inline inspection technologies, and you can fully grasp why the smartest composite engineers spend most of their days translating between specialists rather than designing parts. That's getting more and more complicated materials. Well, if that happens your teams are getting more diverse. The arguments are getting louder.
And the parts? The parts are getting better.
Still, when outsiders see a composites team for the first time, they often find themselves shaking their heads in confusion. They expected something orderly. What they got was a chemist drawing the molecular structures on a whiteboard, while a mechanical engineer brought up a CAD model on his tablet and a physicist scrolled through finite element results on his laptop, all three talking over one another for ten minutes before inexplicably converging to an answer none of them would have come in with.
That is not dysfunction. That is the method.
Composite design cannot be performed from the corner of a building. The materials are too complicated, the failure modes too nuanced, and the consequences of a misstep too great. You want the chemist who knows what happens on a molecular level. The physicist who knows how this translates to bulk behavior. You need the mechanical engineer who knows how that bulk behavior couples with everything: the rest of the aircraft, the rest of the mission, hell, the rest of world. And you need all three, willing to debate in good faith until something productive emerges.
This line of argument makes for an uncomfortable job. It is slow. It does not photograph well. You cannot fit it easily in a pitch deck.
That is also how the finest composite parts in history were built. And it will be how the next ones are too.
The best composite engineering does not take place when everyone agrees. It occurs when three disciplines do not shut up until the answer is right.
Mentis Sciences dwells in this messy middle. Advanced composites, aerospace materials; defense engineering. Once chemists and physicists and mechanical engineers stop trying to win arguments but instead try to solve the real problem that matters.
