Unfilled polyamide 6 is a superb engineering plastic and a poor filament. It contracts up to 2% as it cools, which on any part with a footprint means it peels itself off the plate. Most people who buy nylon and give up have run into exactly that, and no bed surface or chamber temperature entirely solves it.
Chopped carbon fibre solves it. The fibres physically restrain the polymer from contracting along the direction of the bead, and shrinkage drops from 0.8%–2.0% to 0.2%–0.6% — a quarter of the unfilled figure comparing the two minima, three-tenths comparing the two maxima. That is not an incremental improvement in a property; it is the difference between a material you can print and one you cannot, and it is why fibre-filled nylon rather than plain nylon is what most workshops actually own.
The heat number is the other half of the story
Heat deflection climbs to somewhere between 120 and 190 °C, against 60 to 100 °C for unfilled PA6. Two mechanisms are at work: the fibres carry load once the polymer softens, and they also nucleate crystallisation, giving a more crystalline matrix that holds its stiffness to a higher temperature.
Tensile follows, landing at 80 to 130 MPa — the highest range of anything on this site short of the aromatic polymers. This is one of the few materials where "stronger" is the honest word rather than marketing.
Ask this site's material selector for maximum strength and maximum heat resistance, with ease and cost weighted at one, on a machine that reaches 300 °C with an enclosure and a hardened nozzle, and PA6-CF comes second at a score of 0.8088 — behind only PPA, and ahead of the glass-filled version. ULTEM and PEEK are ruled out by the printer rather than by their properties: both need a hotter nozzle than the machine has. That is a fair picture of where this material sits. It is the top of what a very good workshop machine can actually run.
The fibre does nothing about the water
This is the trap for people who assume the fibre fixed everything. The matrix is still polyamide 6, still carries an amide group every sixth carbon, and still hydrogen-bonds with water as enthusiastically as it ever did. If anything the fibre–matrix interface gives moisture an extra place to sit.
- The polyamide drying schedule applies unchanged: 80 °C, 8 to 12 hours.
- Keep the spool in a heated dry box for the duration of the print, not only for the hours before it. At the temperatures involved, absorbed water hydrolyses the chain in the melt and the part is permanently weakened, not just cosmetically rough.
- A part that comes out with a rough, hazy surface and snaps along layers has almost always been printed damp.
Dimensional stability in service is likewise unchanged in kind, though the fibres reduce the magnitude. A PA6-CF part still moves with humidity, just less.
What the machine has to be
Nozzle 270 to 300 °C, so most printers are out on the hotend alone. Bed 70 to 100 °C, and a heated chamber rather than a closed box — a warm chamber both reduces the residual warping and improves layer bonding, which is where the fibre costs you.
A hardened steel or ruby nozzle is mandatory. Carbon fibre destroys brass in tens of hours, and it is worth remembering that everything the filament touches wears: drive gears, PTFE couplings and the extruder's filament path all have a shortened life.
Bed adhesion needs Garolite or a nylon-specific sheet. Plain PEI does not hold a polyamide reliably.
What PA6-CF is genuinely bad at
- Interlayer strength. The fibres lie in the plane of each layer, so Z gets no reinforcement while X and Y get a great deal. The anisotropy is larger here than in almost anything else, and part orientation is the most consequential decision you will make.
- Impact. It is stiffer and more brittle than unfilled PA6, which is genuinely tougher. If the part has to survive being dropped rather than loaded, the fibre works against you.
- Price. $60 to $120 per kilogram, plus a hardened nozzle, plus a dryer, plus a machine that can reach 300 °C.
- Surface finish and fine detail, both blunted by the fibre and by the larger nozzle it prefers.
- Being glued or painted, for the reasons set out on the unfilled polyamide page — nothing the fibre adds improves matters.
- Anything a cheaper material would do. This is a specification material; using it because it sounds impressive is expensive.
Where it is genuinely the right answer
End-use mechanical parts — brackets, mounts and structural components that must hold shape under both load and heat. Jigs and fixtures used near a soldering station or in a hot workshop. Drone and radio-control airframes, where the stiffness-to-weight ratio is what you are buying. Gears and pulleys that would run too hot for a copolyester. Tooling that has to be dimensionally stable while clamped.
If the part will be knocked rather than loaded, unfilled nylon is genuinely tougher — the PA6-CF against PA6 comparison works through that trade. And if the part must be electrically insulating, the glass-filled version is the one to look at, for reasons that have nothing to do with strength.