Pultrusion is the only continuous manufacturing process capable of producing constant cross-section composite profiles at industrial throughput. Unlike batch processes — SMC compression, prepreg lay-up, or filament winding — a pultrusion line converts fiber, resin, and heat into finished product in a single uninterrupted operation. But getting the parameters right is where most production engineers lose yield: pull speed too fast, wet-out is incomplete. Die temperatures too high, resin cures prematurely in the inlet. This guide breaks down the three core process parameters — pull speed, die zone temperatures, and resin bath configuration — with practical optimization ranges backed by industrial data.
Pull Speed: The Throughput vs. Quality Trade-Off
Pull speed dictates line throughput and directly controls the degree of cure before the profile exits the die. Typical pultrusion lines operate between 150 and 1,500 mm/min, depending on part thickness and resin chemistry. The limiting factor is through-thickness thermal lag: a 6 mm thick profile at 500 mm/min stays in a 900 mm die for 108 seconds, which must be sufficient to transfer heat from the die wall to the centerline, initiate crosslinking, and achieve a glass transition temperature above demolding requirements.
Faster pull speeds reduce cycle time but risk undercure at the part centerline, leading to blistering, poor interlaminar shear strength, and surface defects. Slower speeds guarantee full cure but reduce output and risk resin advancement in the injection chamber. The practical optimization uses a cure-degree model: thermocouple probes embedded at the die exit should read crosslink density >85% for polyester and vinyl ester systems, and >92% for epoxy systems. For a standard unsaturated polyester / E-glass system, these are the industrial benchmarks:
| Profile Thickness | Recommended Pull Speed | Die Residence Time | Exit Temperature |
|---|---|---|---|
| 2-3 mm | 800-1,200 mm/min | 45-75 sec | 140-155°C |
| 4-6 mm | 400-700 mm/min | 80-135 sec | 145-165°C |
| 8-10 mm | 200-400 mm/min | 135-270 sec | 150-175°C |
| 12+ mm | 100-250 mm/min | 220+ sec | 155-180°C |
An often overlooked factor is the exotherm contribution: resin systems with higher reactivity — MEKP initiator at 1.5-2.0 phr versus BPO at 1.0 phr — can sustain higher pull speeds because the exothermic peak accelerates centerline cure. Monitoring the die exit temperature gradient (surface vs centerline) with a dual-probe setup provides real-time feedback for pull speed adjustment without destructive testing.
Die Temperature Profiling: Three-Zone Strategy
The pultrusion die is typically heated in three distinct zones — preheat, gelation, and cure — each with a specific thermal profile. A common mistake is treating the die as a uniform heat source, which leads to resin curing in the tapered inlet (choking the die) or exiting undercured at the product end.
The preheat zone (first 15-20% of die length) raises the resin-fiber matrix from ambient to 80-110°C, lowering viscosity before gelation begins. This is when fiber wet-out completes — and when die taper geometry matters most. The gelation zone (next 30-40% of die length) maintains 110-140°C to initiate crosslinking while the resin is still mobile enough to consolidate. The cure zone (final 40-50%) ramps to 140-180°C — or higher for high-Tg epoxies — to complete polymerization and build interlaminar strength.
The following zone temperature table represents an optimized profile for an unsaturated polyester / continuous glass roving system producing a 6 mm flat bar at 500 mm/min in a 900 mm die:
| Die Zone | Length (mm) | Setpoint (°C) | Resin Temperature (°C) | Function |
|---|---|---|---|---|
| Preheat (Zone 1) | 150 | 120 | 80-100 | Viscosity reduction, final wet-out |
| Gelation (Zone 2) | 300 | 145 | 110-135 | Crosslinking initiation |
| Cure (Zone 3) | 450 | 170 | 145-165 | Complete polymerization |
A critical but underappreciated detail is the die entrance taper angle. Taper angles of 2-4° on each side generate sufficient consolidation pressure to eliminate voids without creating excessive back-pressure that stalls roving feed. Angles below 1.5° risk dry spots from insufficient compaction. Angles above 5° generate frictional heating and potential fiber breakage at the inlet. This geometry-thermomechanical coupling is specific to each profile cross-section and fiber volume fraction — a 60% fiber volume fraction UD rod requires a sharper taper than a 45% FVF rectangular profile because of the higher packing density.
Resin Bath and Wet-Out: Injection Box vs. Open Bath
Two resin delivery systems dominate modern pultrusion: the traditional open bath (dip tank) and the more recent injection box (closed die). The choice between them affects wet-out quality, emissions compliance, and processing versatility.
Open baths are simple, low-maintenance systems where fiber rovings pass through a resin-filled tank before entering the die. They work well with low-viscosity polyester and vinyl ester resins (200-500 cP) and are forgiving of fiber tension inconsistency. However, open baths expose resin to atmospheric moisture and styrene evaporation — a regulatory concern in many jurisdictions — and require frequent viscosity monitoring as monomers evaporate.
Injection boxes inject resin under pressure (2-10 bar) directly into the die inlet through a manifold, wetting the fibers milliseconds before gelation begins. This closed system eliminates VOC emissions, maintains consistent resin viscosity by preventing monomer loss, and enables faster pull speeds because the resin spends less time exposed to air and heat before entering the die. The trade-off is higher capital cost and sensitivity to fiber distribution uniformity — an uneven roving pack leads to resin-rich and resin-starved channels that injection pressure alone cannot correct.
Practical Optimization: Data-Driven Parameter Tuning
The most reliable way to dial in pultrusion parameters for a new profile is the step-change method: run the line at three pull speeds (300, 500, and 700 mm/min), measure product properties at each, and interpolate to the optimal setpoint. The key quality metrics are short-beam shear strength (ASTM D2344), Barcol hardness, and visual surface quality. A well-optimized pultruded profile should achieve 85-95% of the resin manufacturer’s published cured properties, with the 5-15% gap attributable to fiber alignment imperfections and microvoids inherent to the continuous process.
Barcol hardness alone is a surprisingly effective in-line quality check. A reading of 45-55 on a Barcol 934-1 impressor for a cured polyester part indicates adequate crosslink density at the surface. Readings below 40 suggest undercure at the surface — often from insufficient die zone 3 temperature or excessive pull speed. Readings above 60 can indicate over-cure with associated embrittlement, particularly in thin sections where the exotherm overshoots the setpoint.
Frequently Asked Questions
What is the fastest pull speed achievable in pultrusion?
For thin profiles (2-3 mm) using highly reactive urethane acrylate resins and radio-frequency preheating, pull speeds of 2,500-3,000 mm/min are commercially achievable. For standard polyester/E-glass systems at 6 mm, practical limits are 600-800 mm/min without RF assist. The record in published literature is 5,000 mm/min for a 2 mm polyester rod using microwave-assisted pultrusion.
Why does my pultruded profile have surface cracks?
Surface cracks typically result from one of three causes: (1) die zone 1 temperature too high, causing resin skin cure before centerline — the cured skin is dragged against the die wall at a different rate than the uncured core; (2) exit temperature too high causing thermal shock when the profile hits ambient air; (3) insufficient internal mold release agent, typically 0.5-1.0 phr for most systems. Increase mold release and verify die zone temperatures before adjusting pull speed.
Can I pultrude carbon fiber with the same parameters as glass fiber?
Not directly. Carbon fiber’s higher thermal conductivity (10-15 W/m·K vs 1-1.3 W/m·K for E-glass) means faster heat transfer to the centerline — which sounds beneficial but can cause premature gelation in the die inlet if zone 1 is too hot. Reduce zone 1 setpoint by 10-15°C when switching from glass to carbon. Carbon fiber’s lower electrical resistivity also complicates RF and microwave-assisted pultrusion.
How do I prevent roving breakage during pultrusion?
Roving breakage is usually a tension problem, not a fiber quality problem. Ensure all rovings from the creel to the guides maintain consistent tension within 5-10% of the mean. A tension differential exceeding 20% between adjacent rovings causes the tighter roving to carry disproportionate load and snap. Ceramic guide eyes at all redirection points reduce friction compared to steel, and tension compensation devices at each creel position are worth the investment for high-value profiles.
For a deeper analysis of how pultruded profiles compare with aluminum extrusions in structural applications, see our carbon fiber vs aluminum tubes structural comparison.
For authoritative process data and industry standards on pultrusion quality control, refer to the CompositesWorld Pultrusion Resource Center.
Standards & References
The following industry standards are referenced in this article:
