🌿 CFRP · GLASS · ARAMID · STACKUPS

Deep Hole Drilling of Composites

Carbon fiber and glass laminates fail differently from metals. They don’t cut — they delaminate, fuzz, and burn. The hole exit is where composite drilling goes wrong: without backup support and a sharp diamond tool, up to 60% of aerospace parts can be rejected for delamination. This guide covers tooling, parameters, and backup strategy for CFRP, glass, aramid, and metal–composite stacks.

0.05–0.06mm/rev feedOptimum for thermoset CFRP
30–45m/min VcResearch optimum, CFRP
<4 / >4 mmDiamond tool ruleCVD coat · PCD tip
35.9%Less delaminationFoam backup plate

Why Composites Are Different

💡 The exit is the enemy: Drilling-induced delamination has been shown to reduce the ultimate tensile strength of open-hole CFRP laminates by ~5.8% and fatigue life by ~26.5% (research on hole quality vs mechanical behavior, Li et al., Composite Structures). In aerospace assembly, delamination and uncut fibers cause up to 60% of drilled-part rejections. Everything below is engineered around that single statistic.
Difference vs MetalWhat HappensConsequence
Heterogeneous structureHard abrasive fibers + soft resin matrix alternate in the cutRapid flank wear, edge dulling, fiber pull-out
AnisotropyEach ply has its own fiber orientation crossing the cutting edgeFuzzing, uncut fibers, variable cutting force
Low thermal conductivityFriction heat stays in the layer ahead of the drill, worst at exitMatrix melting, burns, glass fibers char
Weak interlaminar bondThrust force exceeds the critical interlaminar strengthDelamination, microcracks between plies
No plastic chipFibers shear, matrix crumbles into dustDust hazard, abrasive chip slurry in coolant
⚠️ Never transfer metal cutting logic: In metals, higher speed raises temperature and tool wear; in composites, thrust force is the master variable. Feed rate is directly proportional to both thrust force and delamination — ANOVA in glass-fiber studies attributes up to 73.4% of thrust-force variance to feed rate alone.

Material Behavior: CFRP, Glass & Aramid

MaterialFiber HardnessDominant DamageCooling / DustNotes
Carbon (CFRP)Very hard, abrasiveDelamination, uncut fibers, edge chippingDry or MQL; carbon dust conductiveWorn tools produce the greatest damage
Glass (GFRP)Abrasive, fibrousFraying, fuzzing, splintering, matrix meltDry + dust collectionGlass drill temps roughly double basalt; diamond grit best
Aramid (Kevlar®)Tough, resists shearingFiber pull-out, fuzzing, uncut fiber endsLow RPM, sharp edgeTough fibers must be sheared, not torn
Thermoplastic matrix (PEEK/PEKK)Matrix softens with heatMelting, burrs, dimensional driftCryogenic / MQLOptimum Vc ~130 m/min found for PEKK/CFRP
✅ Tool-grade rule of thumb: HSS is cheapest but short-lived (~500 holes between sharpenings in fiberglass). Tungsten carbide lasts far longer and re-sharpens. Diamond-edged tools run roughly 3× faster than carbide and give the cleanest, fuzz-free holes — at the highest cost.
💡 Relief is everything: Studies on glass and carbon fibers report best results with drill relief angles (α) exceeding 10°. Flatter relief = more rubbing on the already-weak resin layer.

The Delamination Problem

Delamination is the interlaminar fracture where the drill pushes the last uncut plies apart instead of cutting them. It comes in two forms:

TypeLocationCauseSeverity
Peel-up delaminationHole entrance (top plies)Helix angle pulls plies upward away from the toolLess critical
Push-out delaminationHole exit (bottom plies)Thrust force exceeds interlaminar strength of the last plyMost critical — undermines structural health

What stops it

⚠️ Rigid backup is not always better: Aluminum support gives the lowest initial delamination factor (1.095) but causes built-up edge, higher tool wear, and increased vibration — and actually increases delamination by 17.59% by the 200th hole. Soft, compressible, damping materials (foam, cork, rubber) outperform rigid plates over production runs.

Tooling: Diamond, Geometry & Point Angle

<4 mm
diameter
CVD diamond coating (Lach Diamond rule)
>4 mm
diameter
PCD-tipped drill
~12 μm
diamond coat
Kennametal recommended thickness
≤10 μm
edge radius
Sharp edge before coating
118–140°
point angle
118° common CFRP optimum
10:1
tool-life gain
Diamond-coated vs uncoated carbide
Tool TypeBest ForLimitations
CVD diamond-coated carbideComplex geometries, small drills <4 mm (Seco Feedmax C1/C2, Sandvik R854, OSG)Coating dulls the edge; generally not resharpened
PCD-tipped (brazed on carbide)Larger drills >4 mm; higher speeds; longest life (Seco CX1/CX2, CoreHog, ISCAR)Flat wafers limit edge geometry; braze joint near the hot edge
Uncoated solid carbideLaminate stacks where no single tool satisfies every materialShortest life; quality preserved at cost of tool life (AMAMCO)
Diamond-grit edge drillsFiberglass; ~3× faster than carbideMandatory dust collection; cannot run in bushings
✅ Geometry guidance: Composite drills want a high helix angle, severe clearance, a high-rake gash for easy entry, and an edge kept sharp — ≤10 μm radius before coating (Kennametal tests). Double-angle points redirect axial force to reduce uncut fibers; flat 180° points are specified for stacked metal–composite holes to break metal chips.

Parameters by Material

There is no universal CFRP parameter set. Optima depend on fiber architecture (woven vs unidirectional), matrix (thermoset vs thermoplastic), tool geometry, and diameter. These are defensible starting points from published studies.

MaterialVc (m/min)Feed (mm/rev)DrillSource / Notes
Woven CFRP (0/90°)50–900.2–0.4Ø8 mm carbideIsik et al., Processes 2024, SA/GA-optimized
Thermoset CFRP15–450.05–0.2Ø3–5 mm twistBolat et al., Micromachines 2023; optimum 45 m/min, 0.05 mm/rev
CFRP, delamination-min~300.06118° point carbideKarabük University Taguchi study
PEKK/CFRP (thermoplastic)~130LowPCD / coated carbideOptimum Vc for min delamination, 2025 study
Glass fiber (GFRP)Low RPM 5,000–15,000 spindleModerate steady pressureDiamond-grit or carbide 55° pointGill Corporation machining guidance
CFRP/Ti stackSplit per layerLow, variable at interfaceDouble-margin one-shot PCDMetal layer dominates wear & heat
💡 Watch the speed curve: Delamination follows a nonlinear (parabolic) relationship with cutting speed — it falls then rises, so an intermediate optimum exists. Do not assume “faster is better” because thrust drops with speed.
✅ Dry is normal: Most CFRP drilling runs dry. When coolant is used, MQL and chilled air help; liquid nitrogen (cryogenic) gives the cleanest holes but needs infrastructure. Never flood water-based coolant over hot carbon dust.

Chip & Dust Control

Composites don’t make metal chips — they make abrasive dust and broken fiber fragments. In deep holes this dust packs the flutes, burns the matrix, and gauges the bore. Dust extraction is mandatory, and carbon dust is conductive and a fire/electrical risk.

1
Extract at source

Vacuum at the drill point or machine enclosure. Carbon dust is conductive — keep it out of electrical cabinets and bearings.

2
Peck for deep holes

Lift the drill periodically to clear dust and chips. This prevents galling, binding, and heat buildup (Gill Corporation guidance).

3
Match chipbreaker to stacks

In metal–composite stacks, a flat-point / chipbreaker geometry breaks the long metal chip before it whips and damages the composite wall.

4
Filter coolant aggressively

Glass and carbon dust is highly abrasive; fine filtration extends tool and seal life in any flooded application.

⚠️ Fire and electrical hazard: Carbon fiber dust is electrically conductive and can settle on live electronics. Use ATEX-rated / explosion-protected vacuum equipment and ground the work area. Fiberglass dust is an irritant — use respiratory protection per the SDS.

Deep & Long Holes in Composites

ChallengeConsequenceCountermeasure
Chip evacuation over lengthDust packs flutes, friction burns the matrixPeck cycles, internal coolant, diamond-edge tools
Entry drift / off-square startDeviation grows with depth — bell-mouth entrySpot-face or pilot hole; drill templates with hardened bushings
Exit unsupported over long spanExit deflection releases stored energy → delaminationBackup plate or support under the final plies
Wear over the hole lengthForce climbs as edge dullsMonitor torque/thrust; regrind on schedule
Multiple-diameter boresStep transitions catch uncut fibersStep drills with controlled point transitions
💡 Pre-drill then enlarge: For close-tolerance holes (±0.005” or less) drill a pilot first, then finish to size. Holes over 10 mm should be pre-drilled at 5–6 mm and enlarged progressively. A 1 mm pilot hole reduced thrust by up to 55% in high-speed drilling trials.
⚠️ Straightness in stacks: When drilling stacked laminates clamped together, center-to-center tolerance usually suffers on the last part in the stack. Drill against a template and clamp firmly to control the cumulative error.

Hole Quality & Inspection

DefectDetectionAcceptance Typical
Delamination factor (Fd)Ultrasonic C-scan, edge microscopyDepends on ply count; exit damage most critical
Uncut fibers at exitVisual / optical inspection of last ply circumferenceNone for structural holes
Surface roughness (Ra)Profilometer; waterjet reference ~6–7 μm RaDrilled CFRP typically 1–4 μm Ra
Diameter / toleranceAir gauge, CMMH8 typical in aerospace stack drilling (SAE)
Sub-surface damageX-ray CT, high-resolution ultrasonic3-D damage-zone extraction used for structural assessment
✅ Damage quantification: Delamination is normally scored with a delamination factor (ratio of max damage diameter to hole diameter) or by damage-area methods. Research using central composite design predicts delamination from parameters with average error as low as 1.2% — a good reason to lock parameters once validated.

Hybrid Stackups: Metal + Composite

Riveted and bolted aircraft joints are drilled as stacks — CFRP/Al, CFRP/Ti, or three-layer CFRP/Al/Ti — in one shot so holes stay aligned. The metal layer dominates tool wear and heat, and the composite layer dominates quality risk. This is where the process earns its reputation.

Layer TransitionProblemCountermeasure
CFRP → AlMetal burr whips and scratches the composite wallFlat-point / chipbreaker one-shot tool; PCD edge
CFRP → TiTitanium chips are hot, long, and hard to break; heat burns the CFRPLow speed, split parameters, chilled air or MQL
Ti → CFRP interfacePerformance mutation at the interface; oversize holesDouble-margin tool design; variable feed at interface
Wear accumulationStandard twist drills wear rapidly in single-shot workPCD or coated carbide; diamond/ta-C coatings

✅ Why single-shot

  • Perfect hole-to-hole alignment across plies
  • Fewer setups and assembly hours
  • Consistent H8 quality demonstrated (SAE)
  • Widely used: CFRP/Al is the most common aviation stack

❌ The cost

  • Rapid, severe tool wear at metal layers
  • Heat & burn marks on CFRP in Ti stacks
  • No single tool suits every material perfectly
  • Tool life sacrificed to protect hole quality (AMAMCO)
💡 Literature anchor: Jebaratnam & Hassan (Int. J. Advanced Manufacturing Technology, 2024) review process enhancement for single-shot CFRP/Al drilling; Hassan & Abdullah (2022) optimize customized twist drills for CFRP/Al. Both converge on PCD/coated carbide, optimized feed, and interface-aware parameters.

Alternatives: Waterjet & Laser

When delamination cannot be tolerated, or the stack includes materials that ruin a drill, non-mechanical methods step in. They each trade one quality for another.

💧 Abrasive Waterjet (AWJ)

  • Smoothest cut edges, ~6–7 μm Ra
  • No thermal damage, no HAZ
  • No tool wear — great for stacks

❌ Waterjet limits

  • Severe delamination & chipping at hole exit
  • Kerf taper 4–6°, burr, waviness
  • High power cost; needs drying

🔥 Laser

  • Fast, zero cutting force
  • Micro holes possible (e.g. 50 μm in 1.25 mm CFRP)
  • UV laser gives narrow kerf, small HAZ

❌ Laser limits

  • Heat-affected zone: matrix recession, fiber swelling up to ~50%
  • Pulse-laser taper up to 37.5°, HAZ to 650 μm
  • Can reduce workpiece stiffness by up to 30%
🌿
Plain CFRP, short hole→ Twist drill + backup
🛡️
Fatigue-critical assembly→ Mechanical + low feed
🔗
Metal–composite stack→ One-shot PCD / waterjet
🔥
Thick CFRP, no tool wear→ Laser-mechanical hybrid
✅ Best of both: Laser-mechanical compound drilling — laser removes material, twist drill finishes — cut thrust to 53.8% of the delamination threshold, reduced main-edge wear 60.87%, and shrank delamination 55% versus direct drilling (published thick-CFRP study).

Troubleshooting Composite Holes

SymptomLikely CauseFix
Exit delaminationFeed too high at break-through; no backupReduce feed at exit; add cork/foam backup; drill pilot
Fuzzing / uncut fibersDull edge, low relief angle, wrong point angleRegrind to sharp edge ≤10 μm; relief >10°; 118° point
Burns / matrix melt at exitHeat concentration in the last pliesLower RPM, increase peck, MQL/chilled air
Oversize hole / driftOff-square entry, worn tool, thin stack flexSpot-face entry, drill template with bushings, clamp stack
Rapid tool failureWrong tool for stack; CVD coat on >4 mm drillSwitch to PCD; match geometry to the metal layer
Burr on metal ply in stackMetal chip whipping in the composite boreFlat-point chipbreaker, lower feed, one-shot PCD

Key Safety Points

⚠️ Carbon dust is conductive: Carbon fiber dust can short electronics and create electrical hazards. Use ATEX-rated extraction, ground the setup, and keep dust away from switchgear and bearings.
⚠️ Fiberglass and resin dust: Fine glass and matrix dust is a respiratory irritant and can cause skin irritation. Extract at source and use appropriate respiratory and skin protection per the material SDS.
⚠️ Diamond tooling: PCD and CVD-coated tools are brittle and expensive. Never let them run in bushings (diamond-grit tools), avoid interrupted cuts, and inspect for edge chipping before every cycle — a chipped edge scrapes rather than cuts.

Keep Reading