President Vane-Stomper Said No
How one heel in one meeting foreclosed thirty years of commercial fan blade architecture at Pratt & Whitney.
President Vane-Stomper Said No
How one heel in one meeting foreclosed thirty years of commercial fan blade architecture at Pratt & Whitney.
By Herbert Roberts, P.E. | Inventor's Mind | Tuesday Engineering Column
1. The Victory Lap
Last March GE Aerospace published a quiet little victory lap. Thirty years of polymer composite fan blades. Three hundred million flight hours. The GE9X engine — the most advanced widebody powerplant ever certified — about to enter service on the Boeing 777X. Four engine programs across three decades, each one walking the same composite fan blade architecture further down the runway: GE90, GEnx, LEAP, and now GE9X.
It is real engineering. It deserves the lap. The composite fan blade is one of the most consequential material innovations in the history of commercial jet engines, and GE got there by doing the unglamorous thing — picking a material in 1991, partnering with Snecma to fund and build the manufacturing capability in San Marcos, Texas, and then walking the yield curve from under thirty percent to over ninety-seven percent over the following thirty years. That is not a marketing campaign. That is an institutional bet, sustained across four CEOs and four engine programs, paid down one blade at a time.
I read the GE press release with admiration. I also read it as a forensic engineer reads an NTSB report. Because every one of those thirty years, Pratt & Whitney, where I once worked and looked across the technology gap at GE and recognized how far they were changing the aviation baseline, was on the other side of that bet. And the reason we were on the other side comes down to one man, in one meeting, and one heel.
2. The Hopscotch
Carbon fiber polymer matrix composite — PMC — is one of those materials that does not care about industry boundaries. The chemistry is the chemistry. The fiber is the fiber. The resin is the resin. What matters is whether anyone is willing to do the work of learning what the material can do.
The hopscotch went like this. Aerospace structures in the 1960s — radomes, fairings, the second-stage motor cases on solid rocket boosters. Then a long pause while the manufacturing infrastructure caught up. In the 70's carbon fiber made its way into some structural elements on tje L1011 and the Harrier jet, not primary structural elements but milestones focused on lowering weight and adding stiffness when needed in daily operating aircraft.
Then in 1981, John Barnard at McLaren rolled out the MP4/1 — the first carbon fiber monocoque chassis in Formula 1. By the end of that month every team in the paddock had hired composite engineers and ordered prepreg by the yard. Within a season the entire grid was running carbon or making plans to introducing PMCs in their cars. F1 made the technology decision to advance into PMCs in weeks, because F1 has no funding intermediary between the engineering team and the result on Sunday afternoon.
Lotus had ideas in the same era — the twin-chassis Lotus 88 with carbon-Kevlar — but Lotus was structurally underfunded to think at the system level. McLaren had the system focus and the institutional commitment to treat composites as a program, not a part. Same lesson, same pattern, every time this material lands in a new industry: the winner is not the one with the smartest individual. The winner is the one with the institutional commitment to walk the curve.
Ten years after McLaren, GE Aerospace introduced the GE90 with the first composite fan blade ever flown on a commercial jet engine. Twenty-two blades, 128-inch fan, world-record thrust. Carbon fiber had hopscotched from rockets to Formula 1 to the front end of a Boeing 777. The material had arrived in commercial aviation. The question was who would walk its curve.
3. F1 in Weeks, GE in Decades — Same Right Answer
The interesting thing about the hopscotch is what it tells you about timescale. Formula 1 made the composite decision in weeks. GE made it across decades. Two industries with nothing in common except a willingness to take the material seriously, and they both arrived at the same answer in their own time. The variable was never the calendar. The variable was institutional commitment.
That is the diagnostic question for any organization staring at a new material, a new architecture, or a new manufacturing capability. The question is not how fast the industry moves. The question is whether the organization is structurally capable of recognizing what the material wants to be and committing to walk its curve. F1 said yes in weeks. GE said yes across thirty years. Both of those answers are correct.
Pratt & Whitney said no. And the no came from a single chair, in a single division, occupied by a single man with a single heel.
4. President Vane-Stomper Said No
The President of Pratt & Whitney's Military Engine Division — and I will not put his name on this page, because the finding is not about a man, it is about a chair and what the man in the chair did with it — held a famously low opinion of polymer matrix composites. He called them rags and glue. He said it in front of his engineering staff. He said it in front of his program managers. He said it in front of the people who were trying to build the future of his business.
And then, in a meeting whose specifics I will not narrate but whose outcome was witnessed and remembered by everyone present, he stomped on a composite vane set. A real one. On the floor. With his foot. To make a point.
That stomp was not a tantrum. A tantrum would have been forgivable. Engineers stomp on things. Engineers throw things. Engineers say things they do not mean. The vane-stomp was something else. It was a managerial communication. The President of the Military Engine Division — the only chair at Pratt with the Air Force funding required to mature a rotating PMC capability — was telling his organization, in the most legible possible body language, that polymer matrix composites would not be permitted to mature into a primary structural part on his watch. Not on a fan blade. Not on a stator. Not on anything that mattered.
The downstream consequence was immediate and structural, and it closed both ends of the pipeline at once. The Military division — the only division at Pratt with the Air Force funding to mature a rotating PMC capability — was now off the board. The Commercial division had its own PMC work, but it was scoped from the first design review to never threaten a rotating-part decision. The PW4000 commercial engine carried a thirty-inch non-rotating PMC vane — a real composite part on a real commercial engine — but it was designed in every respect like a metal vane. Conservative geometry. Conservative attachment. Conservative load path. Mild weight savings. Its design philosophy could not have been converted to a rotating fan blade if you had handed the team another decade and another budget. Commercial PMC at Pratt was confined to static structures designed by metal-vane rules. Two ounces here. Three ounces there. Acoustic panels. Inlet rings. The kind of parts where a composite failure would not propagate into the engine core and where the engineering investment could be capped at a level that would not threaten anyone's worldview. We had the material. We had the engineers. We had the test rigs. What we did not have, in either division, was permission to point any of it at a rotating part that would actually matter.
5. The Bridge That Got Stomped
Here is the part of the story that converts a bad call about a material into a strategic catastrophe about a company. To understand it, you have to understand how Pratt & Whitney's commercial technology pipeline actually worked at the time. (I have not worked there for years, so none of what I observed then may not be true today, and I wish them well, 8 now and forever.)
Commercial engine development is too capital-intensive to fund from airline revenue. The airlines do not pay for blank-sheet R&D. Outside investors do not write checks for fan blade chemistry. Even GE, the eventual winner of the commercial PMC race, had to bring Snecma in as a 50/50 joint venture partner — CFAN, founded in December 1991, manufacturing facility opened in San Marcos, Texas in February 1993 — to fund and execute the manufacturing scale-up. That is how big the bet is. The company that made the right call on PMC still could not fund the factory alone. They needed an international partner with sovereign-grade patience and a thirty-year yield curve that started under thirty percent and is still being walked today, north of ninety-seven.
So how do American jet engine companies fund advanced technology development? They use the military pipeline. The Air Force-funded ATEGG and JTDE programs pay for advanced engine component R&D in the military engine division. The technology matures on government money, foced on developing what government request to maintain a high technology arsenal of aircraft and other forms of defense and offensive capabilities. To fulfill the government’s technology request companies win contracts to produce the technology and follow a full development cycle, real hardware, real qualification, paid for by the customer of record. Then the commercial division inherits the matured capability knowledge and adds to it to scale it up using commercial investment funds that would never have been sufficient to develop the technology from scratch. Military leads. Commercial follows. That is the playbook. That is how major commercial engine technology at Pratt & Whitney reaches the airlines for the last fifty years.
Which means the play that was on the table in the mid-1990s was obvious to anyone who understood the funding architecture. Pitch a PMC fan blade as a proposed upgrade for the F119 — Pratt's internal designation PW5000 — the engine that powers the F-22 Raptor. Use PMC to produce a lighter and stiffer fan blade for the thirty-two-inch fan stage (win - win). Deliver a real military requirement for advancing the engine performance. Meet a real Air Force interest. Using real ATEGG and JTDE money on the table to pay for the work. Use the government's check to learn three things at fighter scale: the PMC blade design itself, the blade-to-disk attachment design (the hardest mechanical interface problem in any composite fan blade), and the manufacturing process control — the layup, the autoclave, the non-destructive inspection, the field repair. Mature all three on military money. Then, when the commercial division was ready to scale, take a qualified, flying, government-paid thirty-two-inch capability and walk it up to commercial fan diameters with commercial investment funds. Not a blank-sheet bet. A scale-up of a known capability.
That play required exactly one signature. The President of the Military Engine Division. The man with the heel.
He said no. The bridge was closed. And because the bridge was the only structurally available path between Air Force PMC dollars and a commercial PMC capability at Pratt & Whitney, closing it foreclosed the entire architecture. Not the blade. The architecture. He took someone else's future away — the commercial division's future — and the commercial division had no vote, no veto, and no alternative funding mechanism large enough to develop the capability from scratch. They were structurally dependent on a peer who refused to let the work happen.
6. Four Compounding Advantages, Foreclosed
It is tempting to file the vane-stomp as a bad call about a weight-saving feature. That undersells the damage by an order of magnitude. The composite fan blade is not a weight-saving feature. It is an architectural enabler that opens four compounding advantages, and the man with the heel said no to all four.
First: direct-drive at high tip speed. The composite blade can spin at modern fan tip speeds without the centrifugal load problems that limit hollow titanium. The GE9X fan spins at roughly twice the rotational speed of Pratt's Geared Turbofan fan, with no reduction gearbox between the fan and the low-pressure turbine. That direct-drive architecture is not available to a hollow titanium fan blade at modern bypass ratios. The material chooses the architecture.
Second: arbitrary three-dimensional aerodynamic geometry. The composite layup can be shaped to whatever surface the computational fluid dynamics says is optimal. Swept tips. Forward-leaning leading edges. Compound twist. Whatever the flow field wants, the layup delivers. Hollow titanium fights you at every step — the blade has to survive forging, machining, bonded internal cavities, and heat treatment, and every one of those steps is a geometry-constraining operation. With hollow titanium, you end up designing the blade the manufacturing process will let you make, not the blade the aerodynamicist wants. With PMC, the aerodynamicist gets to draw.
Third: lower blade count. The GE9X dropped from twenty-two blades on the GE90 to sixteen on the GE9X — not because they wanted lighter. Because each composite blade was doing more aerodynamic work, and the design closed at a lower count. Lower blade count means higher per-blade loading, better propulsive efficiency, fewer parts to manufacture, fewer parts to inspect, fewer parts to fail. That is a compound benefit the GTF will never access with hollow titanium.
Fourth: a thirty-year S-curve runway. GE90, GEnx, LEAP, GE9X. Four certified programs across three decades, each one walking the same material capability further — lower blade count, more aggressive aerodynamics, larger fan diameter, higher bypass ratio. Pratt's GTF, by contrast, is a point solution. Legitimate engineering — the gearbox is a real accomplishment and the engineering team that made it survive commercial duty cycles deserves their bows — but the architecture has no growth path. You cannot spin the fan faster. You cannot lower the blade weight. You cannot redraw the aerodynamic surfaces. The engine you certify today is the engine you fly in twenty years, give or take a performance improvement package.
Four compounding advantages. Each one a thirty-year S-curve in its own right. All four foreclosed by one heel.
7. Thirty and Counting Years Behind
Here is how the symmetry resolves. GE subtracted blade weight and got direct-drive, high-speed, large-diameter fans on a four-program runway with a thirty-year manufacturing learning curve walked alongside a French partner in San Marcos, Texas. Pratt & Whitney could not subtract blade weight, so they added the mass of a planetary gearbox between the fan and the low-pressure turbine to reach a similar bypass ratio by a different architectural path. Both engineering teams chased the same propulsive physics. Both got there. One walks an S-curve into the next generation. The other carries hardware for the life of every engine to compensate for the material capability that was crushed under a heel in the mid-1990s.
And here is the part that has no shortcut. The CFAN yield curve — under thirty percent in 1993, over ninety-seven percent today — is a thirty-year manufacturing learning curve that does not reset for newcomers. You cannot buy your way onto it. You cannot license your way onto it. You either started in 1991 with a willing international partner and the institutional commitment to walk the curve, or you did not start. Pratt did not start. The bridge was closed.
The forensic finding is short. The material was permitted to do what materials do — hopscotch from one industry to the next, on its own physics, on its own timeline. The funding architecture was available. The play was obvious. The signature required was held by a single chair. The man in the chair said no, and to make sure the no was understood, he put his foot through a vane set in front of his engineering staff.
Thirty years later, GE Aerospace runs its victory lap on three hundred million flight hours and a fifth-generation composite fan program. Pratt & Whitney runs the GTF — legitimate engineering with no growth path — and watches a learning curve they were never allowed to start.
Thirty and counting years behind.
— Herbert Roberts, P.E.
32 years in aviation R&D | 8+ years forensic engineering consulting