Airframe material choice looks like a stiffness-per-gram spreadsheet exercise. In a startup it's really a manufacturing and iteration-speed decision: the best material is the one that lets you change the design next month without retooling. Here's how the options actually compare in UAV practice.
The contenders at a glance
| Material | Density (g/cm³) | Stiffness | Failure mode | Iteration speed | Best use |
|---|---|---|---|---|---|
| CFRP plate/tube | ~1.6 | Excellent | Sudden, brittle | Medium (CNC cut) | Primary structure, arms, spars |
| Aluminium 6061/7075 | 2.7 | Good | Bends, visible | Medium (CNC) | Motor mounts, hinges, heat paths |
| FDM polymers (PC, PA-CF) | 1.1–1.3 | Low–medium | Ductile-ish | Hours | Prototypes, brackets, ducts |
| MJF/SLS nylon (PA12) | 1.01 | Medium | Ductile | Days | Production brackets, enclosures |
| Glass fibre (GFRP) | ~1.9 | Medium | Progressive | Medium | Radomes, antenna covers, skins |
| EPO/EPP foam | 0.02–0.09 | Low (with spar) | Crushes, repairable | Fast (hot wire/mould) | Fixed-wing skins and wings |
| Wood (plywood/balsa) | 0.1–0.7 | Surprisingly good | Progressive | Fast | Prototypes, ribs, low-cost airframes |
Carbon fibre: the default, with three real gotchas
Flat CFRP plate and pultruded or roll-wrapped tube are the workhorses of multirotor structure — a plates-and-tubes architecture gets you a stiff, light frame using only 2D CNC cutting, which any local shop can do in days.
- It's conductive. Carbon shields RF — GNSS antennas mounted under a carbon top plate will underperform mysteriously (see RTK & GNSS guide). It also shorts anything that chafes against it; grommet every wire pass-through.
- It fails without warning. Aluminium bends and tells you a story at inspection; carbon looks fine until it snaps. Impact-damaged arms should be retired, not eyeballed. Your test program and maintenance schedule must assume hidden damage after any hard landing.
- 3D shapes are expensive. Moulded monocoque shells look great and cost tooling money you should spend later, at scale — not on rev B.
Aluminium: underrated in small UAVs
Aluminium earns its 2.7 g/cm³ where you need thermal conduction (motor mounts, ESC plates, companion-computer heatsinks — see onboard compute), threaded connections that survive repeated assembly, and crash-tolerant fittings. A carbon arm with 7075 motor-mount ends is a classic, sound combination.
3D printing: prototype like you mean it
Additive is the startup's unfair advantage — use it deliberately:
- FDM on your desk (PETG, PC, PA-CF): same-day iteration on brackets, antenna mounts, cable clips, sensor shrouds. Watch temperature: a black PETG part on a dark deck in summer sun will creep.
- MJF/SLS as a service (PA12, PA11): isotropic-ish, production-grade small parts without tooling. Many shipping UAVs carry MJF payload housings and duct work.
- Don't print primary structure on aircraft above a few kg MTOW. Layer adhesion under vibration and heat is a fatigue experiment you don't want to run with a customer payload attached — mind your mass budget instead of compensating with plastic.
Fixed-wing is a different game
Wings live on stiffness-per-span, not plate stiffness: the standard recipe is a carbon spar carrying bending loads inside foam or built-up skins carrying shape. EPO foam with a carbon tube spar dominates sub-10 kg wings because it's light, mouldable and absorbs belly-landing abuse. Composite-skinned wings (glass or carbon over foam cores) buy you surface finish and stiffness for a manufacturing-cost premium.
Anywhere an antenna must radiate through structure — GNSS, C2 links, video — use glass fibre, plastic or foam, never carbon. Design these "RF windows" in from day one; retrofitting them into a carbon shell is miserable.
A pragmatic material strategy for a startup
- Rev A–C: carbon plate + tube + FDM prints + off-the-shelf hardware. Zero tooling, week-long design cycles.
- First customer units: swap FDM for MJF nylon; add aluminium at heat and thread points; document every part for your compliance file.
- Scale (100+ units): only now consider moulded composites or injection tooling — with a design that flight hours have frozen.
Frequently asked questions
Is carbon fibre always the best drone frame material?
No. Carbon fibre wins on stiffness-to-weight, but it's conductive (RF and electrical short hazards), expensive to shape in 3D, and fails suddenly rather than bending. Many production drones mix carbon plates and tubes with printed or moulded polymer parts.
Can you 3D-print a production drone frame?
Yes, for small UAVs and for non-primary structure on larger ones. MJF nylon (PA12) and FDM in carbon-filled nylon are common for brackets, mounts and enclosures. Primary load paths on aircraft above a few kilograms are usually carbon tube or plate.
Does a carbon fibre frame block GPS and radio signals?
Carbon fibre is conductive and behaves like a partial RF shield. Antennas and GNSS receivers must be mounted outside or on top of carbon structures, never underneath or inside a carbon enclosure.