Start with a simple piece of physics: when an aircraft pushes through air, the air piles up ahead of it and gets hot, and the heating grows roughly with the square of the speed. The material requirement is a staircase, and each temperature step up demands a change in the material composition of the vehicle.
At Mach 2, the nose of the Concorde reached about 127°C. Hot enough to stretch the airframe by a couple of hundred millimetres, but cool enough for aluminium. At Mach 3, the SR-71 ran hundreds of degrees hotter and had to be built from titanium, with corrugated panels designed to expand and joints that famously wept fuel on the runway until the aircraft warmed up and sealed itself. And somewhere past Mach 4 or 5, the temperature staircase becomes too hot for metal altogether. The air itself becomes a blowtorch; leading edges see well over 1,000°C. At this point, a hypersonic vehicle is not an aircraft that can handle high temperatures; it is a thermal protection system with an engine attached.
A supersonic revival is now underway and yet, it remains comfortably within the lower bounds of the supersonic speed range. Boom Supersonic, a startup that has broken the industry’s mould in the best way, flew its XB-1 demonstrator past Mach 1 last year – the first independently built civil jet ever to do so – and did it repeatedly without an audible boom reaching the ground.
But what Boom’s hardest problem turned out to be was not the shape of the aeroplane – it was that no established engine maker would build them an engine. They are now building their own factory in order to build their own. Even at Mach 1.7, where aluminium and carbon fibre are still acceptable materials, the gating challenge turned out to be industrial, not conceptual.
The Seduction of Design
Hypersonics has a public image – a rendering, actually: a sleek black wedge, computational fluid dynamics in colour, a waverider surfing its own shockwave. The imagery gives the strong impression that this is a design problem; that somewhere there is a shape yet to be discovered, and ‘winning’ is merely a matter of computing that shape.
Computational fluid dynamics (CFD) image of the NASA X-43A at Mach 7.
In reality, we are not short of shapes, not since the sixties at least. The X-15 flew at Mach 6.7 in 1967, well before the moon landing. The Space Shuttle re-entered at Mach 25, more than a hundred times across thirty years. The aerodynamics textbooks were written before modern computing. Designs for hypersonic aircraft have existed in glorious abundance for seventy years.
If design is not the bottleneck, what is?
Death in the Valley
I was struck by the notable lack of one company at this year’s Farnborough Airshow, a British company I really admired. Reaction Engines was, by wide agreement, home to one of the most brilliant propulsion concepts of the age: the SABRE engine, whose precooler could take air arriving at over 1,000°C and chill it in a fraction of a second, through a heat exchanger with thousands of kilometres of fine tubing. It was not a model; it was built and tested, and it worked. It represented nearly four decades of world-class engineering.
All the same, the company entered administration in late 2024. The best design in the field did not fail for want of cleverness. It failed in the long, expensive valley between a validated component and a manufactured, certified, priced product. The valley where capital runs out faster than milestones arrive. Again, the design was not the bottleneck – the “Valley of Death”, covered extensively in these pages, had struck.
The Limitations of Reality
Astute readers will recognise that valley. The materials that survive the extreme temperatures of hypersonic speeds (carbon-carbon, ceramic matrix composites, refractory ceramics) have existed for a while. What does not is the marriage of material, manufacturing route, and believable cost-per-part.
The process routes are measured in furnace-weeks; scrap a part at final inspection and you have lost an entire season, not just a component. The parts are coated systems, so they are certified as systems and inspected as systems. Every vehicle needs leading edges, panels, control surfaces, and nozzles in different materials, all joined together perfectly (and the joints between dissimilar materials in a thousand-degree gradient are a whole other conversation). None of this can be simulated on a computer – the only way to find out is at the furnace door.
The Best of All Worlds
In my previous essays, I argued that the number gating advanced composites is “cost per part at rate”: what the thousandth part costs, once the line is running. The trap for hypersonics: hypersonic programmes have no thousandth part. Fleets are tiny, vehicles are sometimes expendable, and production runs are measured in dozens, not thousands.
The field that needs manufacturing learning the most is the field that does not support enough volume to drive costs down. Every part is a first article; every first article carries the whole overhead on its back. Costs stay astronomical, so fleets stay tiny, so costs stay astronomical, and so on and so forth.
Foundation, Foundation, Foundation
Now overlay the part that captures government attention: sovereignty. The world has only recently learned that “designing chips” and “being able to make chips” are different national capabilities. Manufacturing is a strategic asset: painful to lose and near-impossible to rebuild quickly. High-temperature materials are running the same experiment now.
The fibres at the heart of the best ceramic composites come from a handful of plants worldwide; export rules mean the relevant materials and machines cannot simply be bought; and the tacit knowledge lives in people’s heads and production lines, not in books or manuals. A nation can hold every design in the field and still be unable to build the vehicle. Manufacturing capability is not a support act to the technology. It is the technology.
How to Open the Bottleneck
So the way through is not a better shaped wedge. It is better manufacturing: processing routes borrowed from the polymer industry’s cheap, fast, automatable playbook; material systems that cure in hours rather than infiltrate over weeks; multilayer structures that reserve the exotic, slow, precious ceramics for the stagnation line and build the acreage from something a factory can execute; structures moulded into one another rather than fastened, so the joints stop being the pain point. My readers will know I believe inorganic matrix composites are a key that fits this lock. But the argument stands by itself: whoever makes high-temperature structures cheap to manufacture at small volumes will easily find dominance.
The shapes are waiting, filed neatly for decades, in aerodynamics textbooks. The race is not to design the vehicle of the future. It is to become the kind of country, and the kind of industry, that can afford to build it.
Dr Richard Grainger CEng FRAeS is a co-founder and the CTO of HTMS Composites, a UK company based in Malmesbury developing next-generation ceramic matrix composites for aerospace, defence, motorsport, and energy storage. He holds a PhD from the University of Sheffield in self-healing, self-sensing carbon fibre composites. Before founding HTMS, he worked at the Advanced Manufacturing Research Centre, Rolls-Royce, and the National Composites Centre. He co-founded HTMS to address the lack of a high-quality UK or European ceramic composite supply.
Thank you for reading. We always welcome fresh perspectives and new contributions. If you are interested in writing for Fox and Lion or have a piece that you would like to publish with us, please feel free submit a pitch via the Submissions Portal. We warmly welcome active and former servicemembers, and members of the defence tech community. Additionally, if you are hiring, contact us to get your open role featured on our Defence Tech Careers jobs board.
