Filming and Milling Composites to Improve Productivity

• 1
  Key takeaways
• 2
  What filming and milling composites actually means in the cut
• 3
  Cutting parameters that keep the edge clean
• 4
  Why the machine and fixture decide the outcome
• 5
  When filming and milling composites is the wrong call
• 6
  Single-setup filming and milling vs. separate operations

Filming here is not a coating step. In the shop it means generating a clean, controlled face on the composite before the profile and pocket work begins, so the second operation can reference a surface that was cut in the same setup. On a mill-turn center the part stays clamped while the face is faced, the contour is milled, and the bore is turned. Nothing is released until the last tool leaves the cut.

The productivity gain is easiest to see on parts with a thin wall. A composite laminate at 1.5 mm to 2.5 mm thickness will move when you release the vise. If you film, mill, then flip the part to finish the back face, the second datum sits on a surface that has already sprung. The result is a taper that no inspection report can fix. Keeping the part in one fixture removes that whole error path.

There is also a fiber direction effect that pure metal work does not have. A carbon fiber laminate machines cleanly when the cutter travels against the fiber layup at a controlled angle. When you flip the part, the fiber direction relative to the tool path reverses. Edges that cut clean on side A can fray on side B. Single-setup processing keeps the relationship between tool path and fiber orientation consistent for the whole part.

None of this is new physics. It is the same argument that pushed mill-turn centers into metalwork: fewer setups, fewer datums, fewer chances to lose a part between operations. Composites just punish the mistake harder because the material does not forgive a bad edge.

• 1
  Face first, then profileFilm 0.2–0.5 mm off the face so the milled contour has a flat reference.
• 2
  Hold the part, not the fixtureVacuum or low-pressure clamping avoids crushing the laminate.
• 3
  Keep the fiber directionOne setup keeps tool path angle relative to layup constant.

Composites do not conduct heat away from the cut the way aluminum does. The resin at the tool tip softens around 180 °C to 250 °C for most epoxy systems, and once it smears, the fiber is no longer held in place. The tool starts pulling instead of shearing. So the controlling number is not the wear rate of the tool, it is the temperature at the contact point.

The practical answer is high surface speed with a controlled feed per tooth. For a 6 mm diamond-coated router in carbon fiber laminate, we run 250 m/min to 400 m/min surface speed and 0.05 mm to 0.10 mm feed per tooth. That combination keeps the chip load high enough to cut rather than rub, and the engagement time short enough that heat does not build in one spot.

Chip evacuation matters as much as the cut. Composite dust is abrasive and it will pack the flute if the air blast is weak. We run through-tool air at 6 bar to 8 bar, or flood coolant when the part geometry allows it. On a deep pocket, an air blast alone will not clear the corner, so the tool path has to open up the pocket in steps rather than spiral down into a corner that cannot vent.

Tool material choice is narrow. Uncoated carbide wears in minutes on carbon fiber. Diamond coating or PCD inserts hold an edge long enough to finish a run. For glass fiber the wear is less aggressive, but the dust is finer and the health controls are stricter. Either way, the tool change interval is set by edge quality, not by a fixed part count.

• 1
  Surface speed250–400 m/min for diamond-coated router in CFRP.
• 2
  Feed per tooth0.05–0.10 mm keeps the cut shearing instead of rubbing.
• 3
  Air blast6–8 bar through-tool to clear abrasive dust.

A composite part that is filmed and milled in one setup needs a machine that can reach the face and the contour without re-clamping. That is a five-axis or mill-turn problem. GreatLight runs 16 simultaneous 5-axis machining centers and 16 mill-turn centers, with travel up to 4,000 × 400 × 150 mm on the large frame. The rotary table is Ø400 mm, which sets the practical limit on parts that can be indexed without a second fixture.

The fixture is where most of the risk sits. Composite laminates are stiff in plane and weak through thickness. A vise with 4 kN of clamping force will crush a 2 mm laminate before the first cut. Vacuum chucks and low-pressure mechanical clamps spread the load. On a curved panel, a contoured support block under the part matters more than the clamp on top, because the part will deflect into any gap.

Thermal drift is a quiet error source on long cycles. A 40-minute composite cycle on a machine that has been running steel all morning will see spindle growth. We let the machine idle to thermal equilibrium before a tight-tolerance composite run, and we probe the datum again after the first part. On a ±0.005 mm feature, that probe cycle is not optional.

The payoff shows up in the inspection numbers. A single-setup part has one datum chain, so the stack-up is short. The documented qualification rate on our composite and metal work is 99.99%, and the reason is not better operators, it is fewer operations to get wrong.

• 1
  Reach without re-clamping5-axis or mill-turn keeps face and contour in one setup.
• 2
  Light clampingVacuum or low-pressure clamps protect thin laminates.
• 3
  Thermal soakLet the spindle reach equilibrium before tight-tolerance cuts.

If the part is a flat plate with open tolerances and no thin wall, a three-axis router with a vacuum table will beat a five-axis center on cost and cycle time. There is no datum problem to solve because there is no second face. Putting that job on a mill-turn center just adds setup time and machine hour rate.

Very thick laminates change the picture too. Once the wall goes past roughly 12 mm, the cutting forces rise and the heat stays in the cut longer. The single-setup advantage is still real, but the tool life and surface finish become the dominant cost, and the process needs to be judged on tool consumption per part rather than on setup savings.

Honeycomb and foam core panels are another boundary. The skin cuts fine, but the core crushes under clamp load and the dust behaves differently. Those parts usually need a dedicated core support and a different tool geometry. Running them through a standard composite setup will produce a part that passes a visual check and fails a tap test.

The honest rule is this: use the single-setup approach when the part has a second face that matters, a thin wall that springs, or a fiber direction that must stay consistent. For everything else, a simpler machine is usually cheaper.

• 1
  Flat plate, open toleranceThree-axis router with vacuum table is cheaper.
• 2
  Wall over 12 mmJudge on tool consumption, not setup savings.
• 3
  Honeycomb coreNeeds core support and different tool geometry.