A Stanford-spinout startup is developing a continuous composite manufacturing process it says can produce structural parts larger than the machine making them, using a self-propagating chemical reaction to eliminate the need for ovens, autoclaves or conventional molds.
The chemistry originated at the California university, where researchers were working on a chemically recyclable resin for composite applications. Wind blades were the initial target, given the apparent overlap between recyclability requirements and structural performance. That thesis did not survive contact with the sector’s economics. What emerged was that the resin’s exothermic behaviour could be exploited in ways conventional composite manufacturers spend considerable effort avoiding. Standard practice in composite processing aims for isothermal conditions, keeping tool, laminate and everything else at uniform temperature to prevent localised overheating.
I spoke toPerseus Materials’ CEO and co-founder, Dan Lee, about how the Knoxville, TN-based company inverts that logic.
Founded in 2022, and now approaching double-digit headcount, their process sits somewhere betweenpultrusionand Continuous Fibre Composite 3D Printing (CF-3DP). Pultrusion offers excellent fibre volume fractions and processing consistency, but produces only constant cross-section parts, requiring full retooling for any geometry change. CF-3DP trades geometric freedom for speed. Perseus is attempting to occupy the space between them by shrinking a conventional one-metre pultrusion die down to roughly one centimetre and making it actuatable, so the cross-section can change continuously as the part is pulled through.
The cure mechanism draws on a phenomenon known as self-propagating high-temperature synthesis. The polymerisation reaction releases enough heat to initiate curing in adjacent material, which in turn initiates its neighbours, propagating through the laminate without requiring sustained contact with a heated tool. “Instead of the heat coming from the tool, it comes from my neighbour,” Lee said. Because the reaction is self-propagating, the part does not need to remain in contact with the die once initiated, allowing thick laminates to be pulled through at around 30 centimetres per minute.
Pressure, normally supplied by an autoclave, is applied mechanically by actuators built into the adaptive die, reaching 90 to 120 PSI in open air. There is a trade-off: the harder the actuators squeeze while the part is being pulled, the more they deflect laterally, degrading dimensional tolerance. The current system can hold shape to within one or two millimetres, depending on how much clamping force is applied, a constraint Lee acknowledges is still being characterised across different geometries.
Because the process is continuous rather than batch-based, part length is no longer constrained by machine size. The company’s first paid pilot is in wind blade components, where structural members run between 60 and 130 metres. Conventional tooling must exceed the length of the part being made. Perseus claims its machine length remains constant regardless of how long the part is. Lee frames the scalability argument in stark terms: “How do we get to 10,000x production rates at half the cost? Doubling production by doubling capex, to me, is fake scalability.”
Perseus Materials Backing and Business Model
Lockheed Martin is among Perseus’s backers, drawn, Lee says, by the ability to produce airfoil-geometry parts at low cost with moulding-grade quality. Large featureless panels, wing skins, spars, and stiffening elements are the primary targets. The process trades off the fine-feature capability of conventional 3D printing for high throughput on structurally continuous, large-format work.
The company is not selling machines. Lee is blunt about why: the usability threshold required to hand a system to a third-party operator is, in his assessment, far beyond where the technology sits today. Perseus is instead operating as a contract manufacturer, accepting component orders and handling some assembly and finishing. Its current pitch is the economics of a standardised production run combined with enough geometric flexibility to avoid the retooling costs that make short composite production runs prohibitively expensive.
Source: 3D Printing Industry
