01Executive summary
Three milestones that removed the link between geometric complexity and cost, and then took twenty-eight and thirty-two years respectively to reach certified service.
Chuck Hull built parts by curing photopolymer layer by layer in 1983. A large airliner entered service in 2011 with a carbon fibre fuselage and wing, forty-eight years after usable carbon fibre was first produced. A printed metal component entered certified service in a commercial jet engine in 2015, consolidating around twenty separate parts into one. The recurring theme is that the manufacturing demonstration is early and the qualification is long.
02Additive manufacture: what it actually changes
Conventional manufacture is subtractive or formative. Machining removes material, so internal features unreachable by a tool cannot be made. Casting and moulding require the part to leave the tool, so undercuts and enclosed voids are impossible or require cores. In both, cost rises with geometric complexity because complexity means more operations, more tooling or more setups.
Building layer by layer breaks that link. An internal cooling channel that follows a curved surface, a lattice that varies in density through a part, or twenty components fused into one costs essentially the same to print as a simple block of the same volume. Complexity becomes free; volume and build time become the cost drivers.
Conformal internal features
Cooling channels that follow the surface they cool, rather than straight-drilled approximations. This alone justifies the process in tooling and in hot-section components.
Part consolidation
Assemblies become single components. Every joint removed is a leak path, a fastener, an inspection and a failure mode removed with it.
Anisotropy and residual stress
Properties differ between build direction and in-plane, and rapid local solidification leaves residual stress that can distort a part or crack it during build. Orientation is a design decision.
Support, surface and post-processing
Overhangs need supports that must be removed, sometimes from places nothing can reach. As-built surfaces are rough. Stress relief, hot isostatic pressing and machining of critical faces are usually required.
The process was demonstrated in 1983. Certified service in a critical aerospace application came in 2015. The gap is not machine capability; it is the evidence needed to argue that a part will behave as designed. Every powder lot, every machine, every build orientation and every parameter set potentially changes the material. Establishing allowable design properties requires enormous coupon testing, and establishing that production remains within the qualified envelope requires in-process monitoring and rigorous control. As with composites in the earlier series: the process is the material, so the material qualification is a manufacturing qualification.
03Composite primary structure: the long path to a certified airframe
Carbon fibre reached usable stiffness in 1963. It was used in secondary structure — fairings, control surfaces, interiors — for decades before a large airliner flew with a composite fuselage and wing in 2011. The intervening period was spent building the analytical methods, manufacturing processes and certification evidence needed to trust it with primary load paths.
| Aspect | Aluminium airframe | Composite airframe |
|---|---|---|
| Fatigue | Governing design case; crack growth and inspection intervals dominate | Far less fatigue-critical, permitting higher cabin pressure differential and larger windows |
| Corrosion | Requires protection schemes, drainage and inspection | Fibres do not corrode, but galvanic attack of adjacent aluminium fittings must be prevented |
| Damage detection | Cracks are visible or found by established methods | Impact damage may leave no visible mark while removing compressive strength internally |
| Manufacture | Sheet forming, fastening, riveted joints | Fibre placement and curing in autoclave; a fuselage barrel made as one piece |
| Repair | Well-established riveted patch methods, performed widely | Bonded repairs requiring controlled conditions, cure and specialised inspection |
| Lightning | Aluminium conducts strike current readily | Conductive mesh must be built in to provide a path and protect the structure |
The lightning row is a good illustration of second-order consequences. Replacing a metal skin removes a property nobody specified because it was inherent — electrical conductivity — and it must then be deliberately reintroduced. This is exactly the pattern seen with grid inertia in the previous series. Whenever a material or technology is substituted, the audit question is what was being provided incidentally by the thing being replaced.
04Designing for a process that removes constraints
Engineers trained on subtractive manufacture carry constraints that are no longer binding, and the most common failure in adopting additive manufacture is printing an existing design. That captures none of the benefit and usually costs more than machining it.
- Re-examine the requirementStart from what the part must do, not from its current geometry. The existing shape encodes the old process’s limitations.
- ConsolidateLook for assemblies where joints exist only because the parts could not be made together. Each removed joint removes failure modes.
- Put material on the load pathTopology optimisation and lattice structures place material where stress requires it, which is rarely a shape a machinist could produce.
- Design for the buildOrientation, support requirement, thermal distortion and removable support access must be resolved during design, not after.
- Plan the qualificationDecide early what testing establishes properties, what monitoring shows the process stayed in envelope, and what inspection finds internal defects.
For simple geometry in volume, conventional processes remain far cheaper and faster, and this is not close. Additive earns its place where complexity is high, volume is low, lead time matters, part consolidation removes significant assembly, or the geometry is impossible otherwise. Selecting it as a default rather than against those criteria produces expensive parts that a mill would have made better.
05Takeaways for current practice
- Do not print an existing design. Its geometry encodes constraints the new process does not have.
- Count the joints you can delete. Part consolidation removes leak paths, fasteners, inspections and failure modes at once.
- The process is the material. Qualification of powder, machine, orientation and parameters is the real programme cost.
- Audit what the incumbent material provided incidentally. Conductivity, damping, thermal path and corrosion behaviour are frequently unspecified because they were inherent.
- Expect decades between demonstration and certified service. The interval is filled with evidence generation, and planning for it is the difference between a programme and a prototype.