Long Glass Fibre Compounding Windows and Coupling Agent Reactivity Guide

Long-glass-fibre-reinforced thermoplastics are differentiated from short-glass compounds by residual fibre length, which is controlled through compounding equipment configuration, melt temperature, specific mechanical energy input, and the reactivity of the applied coupling agent. Glass rovings with filament diameters of 10 µm to 24 µm, most commonly 17 µm, are fed downstream in twin-screw extrusion or impregnated in pultrusion melt-bath processes after the polymer has fully melted. The objective is to preserve fibre length above the critical fibre length for the polymer matrix; for polypropylene homopolymer with a tensile modulus of 1.5 GPa to 1.8 GPa and an interfacial shear strength of 15 MPa to 25 MPa when coupled, the critical length is typically 0.3 mm to 0.9 mm depending on filament diameter. Coupling agents used in long-glass compounding include amino-functional silanes, epoxy-functional silanes, methacryloxy-functional silanes, maleic anhydride-grafted polyolefins, and blocked isocyanate dispersions; each imposes a distinct thermal and kinetic window because the reactive group participates in condensation, ring-opening, or decomposition pathways during melt processing. Compliance with food-contact and appliance requirements may invoke FDA 21 CFR 177.1520 for polyolefins, Commission Regulation (EU) No 10/2011 for plastics intended for food contact, and REACH Regulation (EC) No 1907/2006 for coupling agent substances. Glass-fibre coupling agents are generally outside the substance restrictions of RoHS Directive 2011/65/EU unless the formulation contains a restricted flame retardant, stabilizer, or pigment. The following scenarios address specific polymer-coupling agent combinations, processing windows, and the failure modes observed on production-scale compounding equipment.

How Narrow Does the Processing Window Become in MAH-Grafted Polypropylene When Graft Loading Exceeds 1.0 wt%?

The most restrictive thermal window in long-glass-fibre polypropylene compounding occurs when the coupling agent itself is a chemically modified polyolefin with a high residual anhydride content and a peroxide-degraded melt flow rate. In a 40 mm co-rotating twin-screw extruder with L/D 52, barrel sections of 180 °C at the feed throat, 210 °C in the plastification zone, 225 °C in the mixing zone, and 235 °C at the die, the compounded melt temperature remains within 215 °C to 240 °C only when screw speed is held between 250 rpm and 350 rpm and the specific mechanical energy input does not exceed 0.25 kWh/kg. Glass roving is fed at barrel zone 7 of a 12-zone barrel after the PP/MAH-g-PP blend is fully molten; the fibre feed is followed by three distributive mixing elements rather than reverse kneading blocks to minimize fibre breakage. The grafting reaction between maleic anhydride and the aminosilane sizing is rapid at melt temperature, proceeding through ring-opening to amic acid and partial imide formation with water evolution. At MAH graft levels of 0.25 wt% to 0.8 wt%, the acid number of the compounded matrix typically remains between 5 mg KOH/g and 16 mg KOH/g when measured after polymer recovery; this range provides adequate interfacial bonding without excessive melt viscosity reduction. When MAH graft loading exceeds 1.0 wt%, the processing window narrows to approximately ±5 °C around a set melt temperature of 220 °C because acid-catalyzed β-scission accelerates above 230 °C, producing a rapid rise in melt flow rate, die drool, and strand melt fracture. At graft loadings above 1.5 wt%, tensile strength declines despite stoichiometric excess of anhydride because the matrix molecular weight has been compromised; the resulting compound exhibits higher MFR by ISO 1133-1:2022 at 230 °C and 2.16 kg, and lower notched Charpy impact by ISO 179-1/1eA. Injection moulding of the pelletized product is typically performed on presses with clamp force from 800 kN to 12,000 kN, with melt temperature limited to 240 °C and back pressure of 30 bar to 70 bar to avoid additional fibre attrition. The formulation gradient shown in Table 1 was generated on a 40 mm twin-screw extruder at 300 rpm with 30 wt% of 17 µm direct-drawn roving and illustrates the property cliff edge at excessive MAH content.

In a 40 mm co-rotating twin-screw extruder configured at L/D 44 with segmented barrel sections for atmospheric venting, the compounding of PA66 with amino-silane-sized 17 µm continuous roving is operated with a barrel profile of 250 °C in zone 2, 270 °C in zones 3–5, 280 °C in zones 6–8, and 275 °C at the die. The melt temperature must not exceed 285 °C for more than 20 s because aminopropyltriethoxysilane-modified surfaces undergo thermal scission at the amine-propyl linkage and PA66 undergoes oxidative yellowing and chain scission; the limiting data are visible as a colour shift from pale amber to dark brown and a drop in relative viscosity measured in 90 % formic acid according to ISO 307. Pre-drying of PA66 is mandatory to 0.05 % moisture or lower; at relative humidity above 60 %, drying time increases from 4 h at 80 °C to 6 h or more, and the presence of residual water shifts silanol condensation toward premature size-layer crosslinking before the fibre contacts the melt. The fibre is introduced through a side feed at zone 8, after which only low-shear distributive elements are used; screw speed is held at 250 rpm to 300 rpm for a 40 mm machine, producing a specific mechanical energy input of 0.22 kWh/kg to 0.28 kWh/kg. Under these conditions, the residual fibre length after pelletizing and injection moulding is 1.0 mm to 1.8 mm as determined by burn-off and optical image analysis according to ISO 22314. Addition of amine-based additives such as certain hindered amine stabilizer packages can compete with the silane surface for acid groups and should be avoided; the incompatibility appears as lower tensile strength in ISO 527-2 specimens and a greater proportion of glass pull-out on fracture surfaces. The processing window is bounded at the lower end by incomplete melting and high torque, and at the upper end by the 285 °C ceiling; within this corridor the aminosilane size remains reactive enough to transfer stress across the interface but does not covalently bond to the PA66 amide repeat unit in a single extrusion pass.

When PBT Melt Temperature Exceeds 265 °C, the Compounding Window Closes Rapidly Due to THF Evolution

Thermal excursions in PBT long-glass compounding are inherently self-accelerating because the polycondensation equilibrium releases tetrahydrofuran and additional carboxyl end-groups that catalyse further ester cleavage. The process window for PBT LGF with epoxy-silane-sized roving on a 40 mm twin-screw extruder of L/D 40 is typically defined by a barrel profile of 235 °C at the feed zone, 255 °C in the melt seal, 260 °C in the mixing zone, and 265 °C at the die. Melt temperature must remain below 268 °C because the rate of THF generation from butylene terephthalate repeat units becomes sufficient to cause strand porosity, die-hole wetting defects, and a measurable drop in intrinsic viscosity when the product is dissolved in 60:40 phenol/tetrachloroethane at 30 °C according to ISO 1628-5. Pre-drying of PBT at 120 °C for 4 h to 0.02 % moisture or lower is critical; if the feed hopper is not blanketed with dry air, hydrolysis of both the ester matrix and the epoxy silane can occur before the fibre is introduced. The epoxy silane coupling agent, typically 3-glycidoxypropyltrimethoxysilane applied at 0.4 wt% to 0.8 wt% on the roving, undergoes ring-opening with residual acid end-groups and surface silanols; excessive temperature converts this reaction into an uncontrolled crosslinking pathway that increases viscosity at the die and reduces interfacial wetting. On production lines, the failure mode at 270 °C and above is not fibre breakage but increasing die pressure due to pre-crosslinking of the size layer and a shift in melt viscosity from 120 Pa·s to more than 200 Pa·s at 1000 s⁻¹ shear rate as measured by capillary rheometry. The screw configuration uses two low-compression conveying zones after the roving feed and no intensive kneading blocks; specific mechanical energy is limited to 0.20 kWh/kg to 0.26 kWh/kg. The combination of epoxide-functional coupling agent and residual phenolic antioxidant must also be verified, because acidic degradation products from oxidized PBT can deactivate the epoxy group and reduce tensile strength in ISO 527-2 specimens. Published data for this specific epoxy-silane PBT configuration at extreme lower and upper melt temperatures is limited, and compounders typically run a narrow validation matrix before locking the extrusion profile.

At barrel set points of 230 °C to 250 °C, PA6 long-glass compounding with γ-aminopropyltriethoxysilane-sized roving is governed less by temperature than by moisture uptake and silanol condensation kinetics. The polyamide 6 matrix is dried to 0.06 % moisture, and the roving is preheated to 120 °C immediately before the side stuffer to reduce surface water without driving off condensation water needed for adhesion. The screw speed on a 45 mm twin-screw extruder with L/D 40 is set at 280 rpm to 320 rpm, and the die temperature is held at 245 °C; under these conditions the specific energy input is 0.18 kWh/kg to 0.24 kWh/kg and melt residence time after fibre addition is 10 s to 18 s. The APTES hydrolysis reaction at the glass surface follows a pH-dependent mechanism; in the weakly alkaline environment created by the amine group, the ethoxy groups hydrolyse to silanol, then condense with surface silanols to form siloxane bridges. The hydrolysis rate is fastest at pH 3.5 to 4.5 and at pH 8.5 to 9.5, while near-neutral pH 6.0 to 7.5 produces the slowest hydrolysis for many alkyltrialkoxysilanes; because the APTES amine group raises local pH, the reactivity is self-catalytic and can vary batch-to-batch if the roving moisture content is not controlled. The processing consequence is that drying the roving too severely below 0.05 % moisture at the surface delays hydrolysis and leaves the sizing layer brittle, whereas drying insufficiently produces free silanol condensation in the strand die and die-lip build-up. Table 2 summarizes the qualitative pH condition dependencies for APTES hydrolysis and condensation, based on supplier technical data and published sol-gel kinetic studies.

Inline LFT-D Process Pressure, Roving Pre-Heating, and Silane Hydrolysis Sensitivity to Dew Point

Long-glass direct compounding in an inline LFT-D line differs from pelletized LFT-G because the glass roving is fed continuously into a melt stream that is immediately transferred to a compression moulding unit without strand pelletization. The process pressure at the glass feed port in a 75 mm twin-screw extruder with L/D 36 is maintained at 5 bar to 15 bar; the melt pool must be sufficiently viscous to wet the roving but not so viscous that the fibre bundle is dragged into the upstream melt seal. Barrel temperatures for polypropylene are 200 °C to 230 °C, and the melt is discharged at 230 °C directly into a heated buffer or a hydraulic press tool. The roving is drawn from multi-end creels at tension levels of 100 cN to 300 cN; excessive tension causes fuzz breaking and uneven fibre distribution, while insufficient tension allows roving filaments to enter the melt as entangled bundles rather than dispersed filaments. The coupling agent on the roving, usually an amino-silane in PP LFT-D, is already partially condensed at the glass surface; additional hydrolysis occurs when the roving is exposed to ambient air with a dew point above 10 °C. Pre-heating the roving to 110 °C to 130 °C for 60 s removes surface water and improves wetting, but if the dew point is above 14 °C, condensation reoccurs before the fibre feed port unless the creel area is enclosed with dry air. The processing window for silane reactivity in LFT-D is therefore not only the melt temperature but also the atmospheric dew point and the time between roving pre-heater and melt entry. In production trials, a dew point increase from 8 °C to 16 °C was observed to reduce the average residual fibre length from 5.2 mm to 3.8 mm in compression-moulded plaques, with a corresponding loss of notched Charpy impact from 28 kJ/m² to 19 kJ/m² measured by ISO 179-1/1eA, because the pre-hydrolysed size layer creates higher inter-filament friction and changes the fibre breakage mode during screw mixing. The same line data also showed that reducing the roving pre-heat temperature to 90 °C increased strand wetting defects, while raising the dew point in the creel house produced batch-to-batch tensile strength variation exceeding 6 % in ISO 527-2 specimens.

Epoxy-functional silane-sized roving is specified for polycarbonate long-glass compounding because residual primary amines in APTES sizing accelerate carbonate cleavage at melt temperatures above 280 °C. The barrel profile on a 50 mm co-rotating twin-screw extruder with L/D 40 is set to 260 °C at the feed section, 285 °C in the melt zone, 290 °C in the mixing zone, and 285 °C at the die. A melt ceiling of 300 °C is applied because polycarbonate degradation becomes autocatalytic through phenolic end-group formation and carbon dioxide release, causing silver streaks and an immediate drop in melt strength. The epoxy-silane coupling agent does not participate in a ring-opening reaction with polycarbonate under these conditions; its primary function is to provide a hydrolytically stable siloxane interlayer on the glass surface and to reduce the catalytic effect of moisture at the interface. However, the use of amino-functional internal lubricants, certain antistatic agents, or amine-based colour concentrates is incompatible with polycarbonate long-glass compounding because the amine accelerates carbonate cleavage even when the silane itself is amine-free. The compound must be dried to 0.015 % moisture and processed with a screw speed of 180 rpm to 240 rpm; above this range, shear heating pushes the melt above 305 °C and strand quality degrades. Fibre loading is typically 20 wt% to 30 wt%; at 30 wt%, tensile strength ranges from 130 MPa to 145 MPa in ISO 527-2 specimens, but notched Izod impact by ISO 180/1A is sensitive to residual fibre length; if the average fibre length falls below 0.8 mm, impact strength drops below 10 kJ/m². Published data for this specific polycarbonate long-glass configuration with epoxy-silane-coupled roving is limited, and compounders must validate the melt residence time, moisture level, screw speed, and nozzle deposit tendency on their own machine configuration before transferring the process to a production line.