Carbon fiber prepreg on a polycarbonate (PC) matrix is quietly becoming a favorite in places where a part has to be light, strong, and very hard to set on fire. Think aircraft interiors, automotive pieces, and electrical enclosures. The trouble is you cannot grab a bag of flame-retardant PC and call it a prepreg resin. The stuff has to be formulated, compounded, and pelletized so it stays processable and still wets carbon fiber the way it should during impregnation.
Below is how the formulation, the pelletizing line, and the equipment actually fit together, plus the failure modes that eat batch quality when nobody is watching.
Why flame-retardant PC earns its place in prepreg
PC gives you impact strength and heat resistance, which is the whole reason it shows up in structural parts. The catch is neat PC drips when it burns. Those molten drops carry flame straight down to whatever is underneath. In any regulated application that is a dealbreaker.
So the real job is a formulation that reaches UL-94 V-0 without losing the melt flow you need to impregnate fiber. That is four demands at once: flame resistance, flow, fiber wetting, and stability under heat. Tip any one of them and you either get a resin that refuses to flow or a part that fails the burn test.
The formulation, component by component
A resin that actually works for carbon fiber prepreg pulls together six groups:
- PC resin is the matrix and fills the rest of the formula, usually 65 to 85 wt%
- A phosphorus flame retardant, almost always BDP or RDP, at 8 to 15 wt%
- An anti-drip agent, normally PTFE, at 0.3 to 1.0 wt%
- An antioxidant pair, Irganox 1010 with Irgafos 168, totaling 0.2 to 0.6 wt%
- A lubricant at 0.2 to 0.5 wt%
- An optional impact modifier such as a silicone-acrylate copolymer at 2 to 6 wt%
The PC grade gets picked wrong more often than it should. Too low a melt flow index and the material fights you through pelletizing and impregnation. Too high and the finished part gives up its strength. Start in the middle and adjust from there.
BDP and RDP are phosphorus-based and they sit well with PC, which is why they are the default. They get you to UL-94 V-0, and most of the time 8 to 12 wt% is enough. You push toward 15 wt% only for special requirements. Liquid BDP in particular wants to be warmed to 50 to 100°C before it goes in, or it clumps instead of dispersing.
PTFE looks like a throwaway addition at 0.3 to 1.0 wt%, but it does the heavy lifting on the drip problem. A little of it keeps the melt from running and lifts the rating more than you would expect.
The antioxidants are the quiet ones. Irganox 1010 as primary, Irgafos 168 as secondary, each around 0.1 to 0.3 wt%, stop the PC from oxidizing and breaking down at processing temperature. Cut them short and the polymer chain comes apart inside the extruder.
Lubricant keeps flow and release smooth and takes the friction down between melt and steel. The optional silicone-acrylate modifier is there for toughness in the final composite.
Getting the batch mixed
With liquid phosphorus retardants like BDP, warm the retardant to 50 to 80°C first. Then run it with the PC powder and the rest of the additives in a high-speed mixer for 3 to 10 minutes. What you are after is the PC soaking up the liquid so it does not sweat back out or clump in the extruder later. Do this and the burn rating holds steady. Skip it and it drifts, batch to batch, for reasons nobody can pin down on a bad day.
Pelletizing, step by step
The route is melt compounding on a twin-screw extruder, then pelletizing. Three stages.
Drying
Dry the PC and powder at 110 to 120°C for 2 to 4 hours, tuned to the grade and retardant you are using. Moisture needs to drop below 0.02%. PC drinks water, and undried material hydrolyzes at temperature and the properties fall apart. There is no way around this step.
Premixing
Weigh everything to the ratios above with PC as the balance, and blend in the high-speed mixer. If the impact modifier is in, add it here and pull the PC number down so the total lands on 100 wt%.
Extrusion and pelletizing
Push the premix through the main hopper of the twin-screw. It melts, mixes, leaves through the die, cools in a water bath, and a strand cutter chops it into pellets.
On a six-zone machine the temperatures usually run:
- Zone 1, the feed: 200 to 230°C
- Zone 2: 250 to 260°C
- Zone 3: 260 to 270°C
- Zones 4 through 6: 270 to 280°C
- Die head: 280°C
- Screw speed: 250 to 350 rpm
Pellets come off around 3 to 4 mm. A vibrating screen knocks out the long, short, and misshapen ones so the size stays even.
The two controls that decide pellet quality
Vacuum vents belong at the melting and metering sections. They pull out the volatiles, leftover moisture and low-molecular junk the retardant may have carried in. Vent poorly and those volatiles become pores in the prepreg down the line.
Then there is residence time. PC degrades if it sits hot too long, and the way you control that is screw configuration and screw speed, not the temperature you set on the panel.
What the line needs
You want a hot-air circulating oven for drying, a high-speed mixer for preblend, a loss-in-weight feeder for accurate dosing, and a co-rotating intermeshing twin-screw as the heart of it. Co-rotating screws shear hard, which is exactly what dispersing additives into thick PC demands. After the die, a water bath, a pelletizer, and a screen.
If the resin is headed for a melt-impregnation line, die-face or underwater pelletizing pays off. Those give round, even pellets that feed without fuss.
The five ways this goes wrong
The flame retardant will not stay put. Premix liquid phosphorus retardant with PC powder at the wrong time or temperature and it blooms to the surface, and the rating wanders. Warm it, add it in stages, or move to a masterbatch where the retardant is already spread through a carrier.
Drying that you cannot trust. PC is sensitive enough that a half-done dry drops viscosity and degrades the polymer, and the prepreg comes out weak. Put a dew-point meter online and hold the drying air at -40°C or better.
Heat kills it. PC has a narrow window. Keep melt temperature under 320°C, do not let material linger in the hot zone, tune the screw, and watch the actual melt temperature instead of the setpoint.
Ugly pellets. Uneven颗粒 throw off the feed in impregnation and the fiber content in the prepreg goes inconsistent. Underwater or die-face cutting fixes the roundness, and the screen handles grading.
Too stiff to flow. PC’s viscosity eases with a low-molecular-weight resin like SAN, which also helps it wet the fiber. Just watch what that does to flame resistance and strength before you lean on it.
What it really comes down to
The chemistry here is not the hard part. The losses come from drying that was a little loose, dispersion that was a little uneven, and residence time that ran a little long. Tighten those three and the rest takes care of itself.