Epoxy Resin For Wind Turbine Blades Infusion Grades Process Windows
Epoxy Resin for Wind Turbine Blades: Infusion Grades & Process Windows
Vacuum-assisted resin transfer molding (VARTM) of wind turbine blade shells requires epoxy resin systems that combine near-Newtonian flow behaviour at injection temperature with latency sufficient to fill multi-metre preform stacks before gelation. Infusion grades are typically based on bisphenol A diglycidyl ether, bisphenol F diglycidyl ether, or their blends, modified with 5–20 wt% of a low molar mass reactive diluent such as 1,4-butanediol diglycidyl ether. The epoxy equivalent weight of the base resin is controlled between 170 g/eq and 185 g/eq according to ASTM D1652; batch-to-batch variation beyond ±3 g/eq requires stoichiometric adjustment of the hardener. Dynamic viscosity measured by rotational viscometry according to ISO 3219 at 25 °C generally falls between 200 mPa·s and 350 mPa·s; heating the mixed resin to 35 °C lowers the value to 80–180 mPa·s. A production infusion unit for a 70 m shell typically meters mixed resin at 12–25 kg/min through a 32-element static mixer and degasses the liquid at 10–30 mbar before injection. Residual gas loading in the mixed liquid is kept below 0.5 vol% to prevent void formation in the laminate, and resin inlet pressure is limited to 0.5 bar relative to atmospheric pressure to avoid core foam compression.
The candidate infusion resin must be evaluated against equipment-specific fill time. Darcy flow through a glass triaxial fabric stack with a saturated permeability of 2×10−10 m² and a laminate thickness of 60 mm produces fill times directly proportional to viscosity; a shift from 250 mPa·s to 350 mPa·s adds 35–45 % to resin arrival time at the far end of a 25 m preform. Mold surface temperature is controlled at 30–35 °C during injection to keep in-mold viscosity within the process window. Thermocouple data from shell molds show that resin temperature ahead of the flow front can be 4–8 °C lower than mold surface temperature in thick spar cap regions, reducing the local viscosity benefit and requiring compensation in fill simulations.
What Limits the Lower and Upper Viscosity Thresholds?
The lower viscosity limit is not defined by handling alone but by the crosslink density and fatigue resistance of the cured network. Reactive diluents reduce blend viscosity at the expense of glass transition temperature and modulus. Replacing 5 wt% of bisphenol A diglycidyl ether with 1,4-butanediol diglycidyl ether lowers mixed viscosity from 300 mPa·s to 180 mPa·s at 25 °C, but tensile modulus measured according to ISO 527-2 may fall from 3.1 GPa to 2.8 GPa, and glass transition temperature by 6–10 °C. The upper viscosity threshold is governed by preform permeability and foam pressure limits. Above 350 mPa·s, resin cannot fill high-density glass truss regions before gel, and the increased pressure differential can compress PVC foam core at 0.02–0.04 MPa, causing thickness deviation and resin-rich zones. For large blades, the practical upper mixed viscosity limit at the injection temperature is therefore 250–300 mPa·s, with the lower limit set at 150 mPa·s unless a higher reactive diluent content is accepted.
Pot life is the second gate parameter. Infusion grades for large shells are formulated for gel times of 240–600 min at 25 °C when measured on 100 g samples according to ASTM D2471. Bisphenol A epoxy with anhydride hardener can achieve 300–600 min at 25 °C but requires a cure cycle at 80–100 °C; modified aliphatic amine systems cure at 70 °C but gel in 120–240 min. On a 70 m shell, the operator must therefore manage resin batches of 200–400 kg and avoid feeding mixed resin older than 120–180 min into the static mixer because local accumulation in lines can initiate exotherm and create gel particles that block the inlet mesh.
| Parameter | Method | Infusion-grade range | Observation |
|---|---|---|---|
| Mixed resin viscosity at 25 °C | ISO 3219 | 200–350 mPa·s | Heated inlet recommended at upper bound |
| Mixed resin viscosity at 35 °C | ISO 3219 | 80–180 mPa·s | Wetting and fill pressure limited |
| Gel time at 25 °C | ASTM D2471 | 240–600 min | Batch size and feed age controlled |
| Peak exotherm, 100 g | ASTM D2471 | 140–190 °C | Not predictive of large laminate exotherm |
| Neat resin Tg after 7 h at 70 °C | ISO 11357-2 | 75–90 °C | Post-cure increases Tg |
| Neat resin tensile strength | ISO 527-2 | 60–85 MPa | Reactive diluent content dependent |
| Moisture absorption, 7 d in water at 23 °C | ISO 62 | 0.3–0.8 % | Amine adduct formation can raise uptake |
At a mold filling length of 25 m, the resin front movement is governed by the ratio of fabric permeability to resin viscosity and by the vacuum pressure gradient. In-plane permeability of unidirectional glass spar cap fabric under 55–58 % fibre volume fraction is approximately 1×10−11 m² to 8×10−11 m², while biaxial and triaxial shell fabrics show higher values of 1×10−10 m² to 5×10−10 m². A resin with 250 mPa·s viscosity and a 0.1 MPa pressure differential therefore fills a 2 m wide shell segment at rates of 0.2–0.6 m/min depending on stack architecture. Full scale injection is deliberately staged with multiple resin gates and flow media to keep fill time below 300 min; once the resin front passes 60 % of the preform length, priority valves are opened to prevent dry zones in root and trailing-edge regions. Flow simulation based on finite-volume or control-volume techniques is used to position gates and to adjust flow media coverage before the first production shot. The simulation results are validated by resin arrival sensors and in-mold pressure transducers at 10–15 m intervals.
Spar Cap Exotherm and the 90 °C Constraint
Exotherm measurements in spar cap laminates subject the infusion resin to a different thermal boundary than neat resin samples. The spar cap is often 50–70 mm thick with up to 60–65 % fibre volume fraction, and the local resin mass can raise adiabatic temperature rise above 100 °C if the cure cycle is not staged. A typical production cure cycle starts with 20–30 min at 50 °C to allow flow front stabilization, followed by 70 °C for 6–8 h. In-mold thermocouples record peak laminate temperatures of 85–95 °C in the thickest cap sections. Exceeding 90 °C for more than 30 min can cause surface print-through on the shell gel coat, risk core foam distortion, and generate residual thermal stress. For this reason, anhydride-based infusion grades with broad exotherm curves are preferred for spar caps, while faster modified amine systems are restricted to thinner shell skins and shear webs.
The degree of cure must reach at least 95 % before demolding. Differential scanning calorimetry according to ISO 11357-5 is used to measure residual enthalpy, and a value below 5 J/g is typically required for final release. The glass transition temperature determined by ISO 11357-2 after 7 h at 70 °C may be 75–85 °C; a subsequent post-cure of 4 h at 80 °C raises Tg by 5–10 °C and improves moisture resistance, but any post-cure above 90 °C on a foam-cored shell may induce core dimensional changes. Published data for the specific combination of anhydride-cured low-viscosity epoxy and high-density PVC foam in thick spar caps is limited; qualification therefore relies on representative panel tests and instrumented full-mold trials.
Root section infusion presents a different window than shell skins because the root preform can reach 100–150 mm in thickness with high unidirectional glass content. The practical injection viscosity at the root is often held below 200 mPa·s by increasing inlet temperature to 35–40 °C, and the fill time is extended to 4–8 h. Multiple injection gates are placed radially around the root, and the vacuum bag is staged to minimise dry fibre pockets between bolt-hole reinforcement rings. The resin feed system is divided into two or three degassed intermediate tanks to prevent mixed resin age from exceeding 120 min before injection. Failure to control feed age in the root lines can cause gelation inside the flow media or an increase in local viscosity that appears as a late-arriving dry band after mold pressure is released.
The bolt-hole root, however, cannot accept large quantities of reactive diluent because the root must carry bearing stress and fatigue load. Neat resin tensile strength measured by ISO 527-2 below 60 MPa or Tg below 75 °C is generally rejected for root sections. The root lamination also requires low moisture absorption and high interfacial adhesion to glass sizing; water absorption measured by ISO 62 is controlled to 0.5 % maximum for the cured resin. These requirements push root infusion formulations toward bisphenol A/F blends with 5–12 wt% reactive diluent and anhydride hardeners, which maintain long pot life but require the higher post-cure. Balsa and PVC foam core materials are pre-dried at 60 °C for 24 h if workshop relative humidity exceeds 60 %, because absorbed moisture increases the risk of void expansion and reduces resin adhesion at the core interface.
Qualification of an infusion-grade epoxy for a rotor blade normally requires a material and process approval package aligned with IEC 61400-5 and DNVGL-ST-0376. The resin supplier submits independent test reports for neat resin tensile properties according to ISO 527-2, flexural properties according to ISO 178, interlaminar shear strength according to ISO 14130, and unidirectional laminate fatigue data generated under tension–tension loading according to ISO 13003. The laminate test programme includes void content measurement according to ISO 7822, with acceptance below 2 % for structural regions and below 1 % for spar cap laminates. Glass transition testing according to ISO 11357-2 is performed on the exact cure cycle used in production; any change in hardener supplier or reactive diluent content greater than 2 wt% requires re-qualification because the stoichiometric ratio and network architecture shift. The approval dossier also documents the temperature–viscosity curve from 15 °C to 80 °C, the gel time curve, and the heat flow profile across the specified cure cycle.
| Process step | Parameter | Lower bound | Upper bound | Primary failure outside window |
|---|---|---|---|---|
| Resin pre-warm | Feed tank temperature | 25 °C | 35 °C | High viscosity or reduced pot life |
| Mold preheat | Tool surface temperature | 30 °C | 45 °C | Dry spots or exotherm-induced print-through |
| Vacuum application | Bag-side absolute pressure | 5 kPa | 20 kPa | Volatile degassing or insufficient compaction |
| Injection | Fill time | 180 min | 300 min | Dry spots or gel in feed lines |
| Initial cure | Mold temperature | 60 °C | 80 °C | Low conversion or laminate exotherm |
| Post-cure | Cycle | 70 °C for 4 h | 80 °C for 8 h | Low Tg or core foam distortion |
When Tooling Temperature Drops Below 18 °C
Tooling temperature below 18 °C affects the resin even before injection because viscosity rises non-linearly. For every 10 °C decrease, the viscosity of a bisphenol A/F infusion grade can increase by a factor of 2–2.5. At 15 °C, a resin rated at 250 mPa·s at 25 °C may exceed 500 mPa·s, making the fill time exceed the measured gel time in the feed tank. The mold temperature must therefore be pre-conditioned for a minimum of 8–12 h before layup if the workshop air temperature is below 18 °C. The resin drum is warmed in a heated enclosure to 25–30 °C and the static mixer jacket is set to 28 °C. These steps are mandatory because viscosity corrections after the resin has entered the preform are not possible; lowering inlet pressure to compensate for high viscosity only increases void content and creates resin-starved areas at the leading edge.
The same low-temperature condition also changes the curing reaction. Gel time at 18 °C can be extended by 40–70 % relative to 25 °C, but the initial cure may not reach full conversion if the mold cannot be ramped to 70 °C. Residual enthalpy after 8 h at 50 °C can remain above 10 J/g, requiring a longer post-cure. For this reason, low-temperature infusion is limited to parts where a delayed rise to cure temperature is acceptable and where the laminate thickness does not exceed 20 mm. Thicker sections are preheated or resin is warmed to 30–35 °C before the mold is closed. In winter production, operators monitor resin inlet and mold surface temperature with calibrated thermocouples and reject shots when the inlet resin temperature is below 25 °C or the mold surface is below 23 °C.
