PEEK (polyetheretherketone) exhibits pronounced viscoelastic expansion behavior when it enters the viscous flow state above its melting point of approximately 343°C, directly affecting die swell, dimensional accuracy, and part quality in extrusion and molding processes. This technical article systematically analyzes the expansion mechanism of molten PEEK, covering its molecular origins, measurable characterization parameters, standard test methods, and practical industrial applications.
Technical Overview
PEEK is a semi-crystalline, aromatic high-performance thermoplastic characterized by a glass transition temperature of approximately 143°C, a melting point of approximately 343°C, and a recommended melt processing window of roughly 360–400°C. Within this window, PEEK exists in the viscous flow state: the crystalline phase has fully melted, molecular chains are mobile and entangled, and the material can undergo irreversible flow while simultaneously retaining a measurable elastic memory. This dual viscous-elastic character is the fundamental reason why molten PEEK does not simply flow passively through a die or mold, but instead expands, swells, and recoils once flow constraints are removed.
Three distinct expansion phenomena are observed when PEEK is in the viscous flow state. The first and industrially most significant is elastic die swell (the Barus effect), in which an extrudate emerging from a die expands to a diameter larger than the die orifice. The second is thermally driven volumetric expansion, described by the pressure-volume-temperature (PVT) behavior of the melt, in which specific volume increases with temperature and decreases with applied hydrostatic pressure. The third is degradation-driven expansion, which occurs when excessive melt temperatures or long residence times cause chain scission or crosslinking reactions that release volatile decomposition products, generating bubbles and foaming within the melt. A thorough mechanism analysis must distinguish among these three, because they respond to completely different processing variables.
Molecular and Viscoelastic Origins of Melt Expansion
The dominant mechanism of die swell is the storage and release of recoverable elastic energy in the entangled polymer melt. As PEEK is forced through the converging entry region and the shear field of a capillary or die channel, the long-chain molecules are stretched and preferentially oriented along the flow direction. According to entanglement network and tube-model (Doi-Edwards) descriptions of polymer dynamics, this deformation stores elastic free energy and is quantified macroscopically by the first normal stress difference N1, which grows with shear rate. At the die exit, the wall constraint and driving pressure disappear, the oriented chains recoil and relax toward their equilibrium random-coil conformation, and the axial contraction produced by this recoil manifests as lateral expansion of the extrudate.
The magnitude of elastic expansion is governed by the competition between the relaxation time of the melt and the residence time in the die, commonly expressed through the Weissenberg number. For PEEK at approximately 380°C, melt relaxation times are typically on the order of milliseconds to a few hundredths of a second, so at high shear rates the chains cannot fully relax within the die and substantial swell remains at the exit. Reported extrudate swell ratios for unfilled PEEK generally fall in the approximate range of 1.1 to 1.7, increasing with shear rate, decreasing with higher melt temperature, and decreasing with longer die land length (higher L/D ratio), because a longer residence time in the die permits greater relaxation. Rigid fillers such as glass or carbon fibers strongly suppress swell, often reducing the ratio to approximately 1.0–1.2, since the fibers neither store elastic energy nor transmit orientation recoil.
Thermally Driven Volumetric Expansion
Independent of elasticity, the molten PEEK phase itself expands volumetrically with temperature. PVT measurements indicate that the melt thermal expansion coefficient of PEEK is on the order of 6–8 × 10⁻⁴ K⁻¹, meaning that the specific volume increases measurably between 350°C and 420°C, while applied pressures on the order of several hundred bar compress the melt significantly. Upon cooling from the viscous flow state, the amorphous melt first contracts thermally and then undergoes an additional, abrupt specific-volume reduction at the crystallization temperature as the ordered crystalline phase forms. This crystallization shrinkage partially counteracts the recorded die swell in slowly cooled extrudates and must be separated from the purely elastic contribution when analyzing mechanism data.
Expansion Induced by Thermal Degradation and Volatiles
PEEK is thermally robust, with thermogravimetric 5% mass-loss temperatures typically above approximately 535°C in nitrogen, but oxidative and long-residence degradation begins to matter well below that in processing practice. Prolonged exposure above roughly 420–430°C, or extended residence at processing temperatures, can initiate chain scission and crosslinking reactions that evolve small volatile species such as carbon monoxide, carbon dioxide, and phenolic fragments. These gases nucleate bubbles within the viscous melt, producing a permanent, non-recoverable expansion of the extrudate that is mechanistically distinct from elastic die swell. Residual moisture, although PEEK absorbs only about 0.1–0.5% water at saturation, can similarly generate surface defects and microvoids if resin is not dried prior to processing.
Key Parameters Governing Expansion Magnitude
- Shear rate at the die wall, which controls stored elastic energy and N1
- Melt temperature, which shortens relaxation times and accelerates in-die relaxation
- Die geometry, including L/D ratio, entrance angle, and land length
- Molecular weight and molecular weight distribution of the PEEK grade
- Filler type and loading, such as glass fiber, carbon fiber, PTFE, or graphite
- Residence time in the hot flow channel and thermal history of the melt
- Draw-down ratio applied by haul-off equipment downstream of the die
- Hydrostatic pressure and cooling rate, which control PVT expansion and crystallization contraction
Testing Samples
Unfilled PEEK Resin Grades
Virgin unfilled PEEK pellets in standard, medium, and high viscosity grades constitute the baseline sample set. Because melt elasticity scales with molecular weight, comparing grades of different viscosity classes isolates the molecular weight dependence of the expansion mechanism. All pellets are dried, typically for 3 hours at 150°C, to reduce moisture below approximately 0.02% so that hygroscopic effects do not contaminate the elasticity data.
Filled and Modified PEEK Compounds
Compounds containing 30% glass fiber, 30% carbon fiber, and bearing-grade formulations with PTFE, graphite, and carbon fiber are tested to quantify how rigid and particulate fillers suppress die swell and alter the viscoelastic mechanism. These samples reveal the transition from entanglement-driven chain recoil to filler-orientation and matrix-filler interaction effects.
Post-Processing Melt Specimens
Extruded strands collected at defined shear rates, 3D printing filament spools, and injection molded plaques are analyzed after solidification. Density mapping along the extrudate axis, bubble inspection, and shrinkage measurement on molded plaques allow the elastic, thermal, and crystallization contributions recorded in-line to be reconciled with the final solid-state geometry.
Testing Items
A complete characterization program for the viscous flow state expansion mechanism of PEEK includes the following measurable items, each of which maps to one or more of the mechanism components described above.
- Melt mass-flow rate (MFR) and melt volume-flow rate (MVR) for lot-level QC of flow and elasticity shifts
- Apparent and true shear viscosity as functions of shear rate from 10² to 10⁴ s⁻¹
- Die swell (extrudate swell) ratio determined at multiple die L/D ratios and shear rates
- First normal stress difference N1 as a function of shear rate
- Storage modulus G′, loss modulus G″, complex viscosity η*, and relaxation time spectrum
- PVT specific volume as a function of temperature and pressure across the melt range
- Melt density and melt thermal expansion coefficient
- Thermogravimetric degradation onset and mass-loss temperatures in nitrogen and air
- Glass transition, melting, crystallization, and crystallinity by DSC
- Extrudate density profile and void/bubble content in solidified strands
Testing Methods
Capillary Rheometry with On-Line Die Swell Measurement
High-pressure capillary rheometry in accordance with ASTM D3835 or ISO 11443 is the primary method for mechanism analysis. A piston-driven rheometer maintained at 380–400°C forces molten PEEK through capillary dies of selected L/D ratios, for example 10/1, 20/1, and 40/1, with defined entrance angles. Bagley orifice corrections remove the entrance pressure contribution, and Rabinowitsch corrections convert apparent to true shear rate. A laser diameter gauge mounted in a heated isothermal chamber directly downstream of the die measures the free extrudate diameter before cooling and crystallization distort it, yielding the true elastic swell ratio. Plotting swell ratio versus shear rate for each L/D separates the entrance-contribution from the in-die relaxation contribution.
Oscillatory Rotational Rheometry
Parallel-plate oscillatory rheometry per ASTM D4440 or ISO 6721, typically with 25 mm plates at a 1 mm gap under nitrogen at 360–400°C, provides the linear viscoelastic fingerprint of the melt. Frequency sweeps from approximately 0.1 to 100 rad/s yield G′ and G″, and the Cox-Merz relationship allows complex viscosity to be correlated with capillary shear viscosity. The G′–G″ crossover and the low-frequency plateau behavior define characteristic relaxation times that control how much elastic energy survives the die transit and is released as expansion at the exit.
Melt Flow Rate Testing
MFR/MVR testing per ISO 1133 or ASTM D1238, commonly conducted for PEEK at 400°C with a 2.16 kg load and at higher custom loads, serves as a rapid screening method. Systematic shifts in flow rate between resin lots indicate molecular weight drift that will translate directly into changes in die swell and expansion behavior on the production line.
PVT Dilatometry
Piston-dilatometer PVT measurement, for example following ISO 17744, records specific volume along isobaric heating and cooling paths at pressures from 200 to 1000 bar. These data quantify the pure thermal expansion of the viscous melt and the crystallization-induced specific volume drop on cooling, allowing the crystallization shrinkage component to be deconvoluted from the total dimensional change observed at the die exit.
Thermal and Compositional Analysis
DSC per ISO 11357 establishes Tg, Tm, crystallization temperature, and crystallinity, confirming that samples enter a fully molten state during rheological testing. TGA per ISO 11358 at 10°C/min in nitrogen and air defines the degradation ceiling that bounds the safe processing window and separates benign elastic expansion from degradation-driven bubble formation. Density gradient column testing per ASTM D1505 and Karl Fischer moisture analysis complete the specimen characterization.
Applications
Understanding and quantifying the expansion mechanism of PEEK in the viscous flow state is not an academic exercise; it directly determines dimensional control, surface quality, and process stability across every major melt-processing route for this polymer.
- Profile, rod, and tube extrusion, where die orifices must be oversized and shaped to compensate for the measured swell ratio in order to hit final dimensions
- Wire, cable, and magnet wire insulation coating, where draw-down must be balanced against die swell to maintain stable insulation wall thickness at high line speeds
- 3D printing filament production, where the tight diameter tolerance of 1.75 mm filament requires precise swell control, particularly for carbon-fiber-filled PEEK grades
- Injection molding, where packing pressure, holding time, and mold shrinkage predictions depend on the PVT and elastic behavior of the melt, and where excessive injection speed can cause jetting and flash
- Film and sheet extrusion, where die lip design and chill-roll gaps must account for exit expansion and subsequent crystallization contraction
- Melt spinning of PEEK fibers and monofilaments for filtration, braiding, and conveyor applications
- Medical implant, aerospace, oil and gas, and semiconductor components, where PEEK parts must meet strict dimensional and void-free quality requirements
In each of these processes, the practical engineering responses derived from mechanism data include lengthening the die land, raising melt temperature within the safe window, reducing shear rate, applying controlled draw-down, adding rigid fillers, and enforcing strict residence time limits to avoid the degradation-driven expansion regime.
FAQ
What exactly causes PEEK to expand in the viscous flow state?
The dominant cause is viscoelastic die swell. Shear and converging entry flow in the die stretch and orient the entangled molecular chains, storing elastic energy quantified by the first normal stress difference. When the extrudate exits the die, the constraints vanish, the chains recoil, and the melt expands laterally. Secondary contributions include pure thermal volumetric expansion of the melt and, under abusive conditions, bubble formation from thermal degradation volatiles.
How large is die swell for unfilled PEEK?
Typical literature values place the extrudate swell ratio of unfilled PEEK at roughly 1.1 to 1.7, depending on shear rate, melt temperature, and die L/D ratio. Swell increases with shear rate and decreases with higher temperature and longer die land. Fiber-filled grades typically show much lower ratios, often close to 1.0 to 1.2.
Does raising melt temperature reduce expansion?
Yes, within the safe processing window. Higher temperature shortens the melt relaxation time, so more elastic energy relaxes inside the die before the exit, reducing elastic swell. However, temperatures above approximately 420–430°C, or long residence times, risk degradation-driven bubble formation, which is a different and permanent expansion mechanism.
How do glass or carbon fibers affect the expansion behavior?
Rigid fibers strongly suppress die swell because they do not store recoverable elastic energy and they restrict matrix chain recoil. A 30% carbon fiber PEEK compound can show a swell ratio close to unity, which is one reason filled grades are preferred for high-precision extrusion and filament production, although fiber orientation introduces its own anisotropic dimension effects.
How is elastic die swell distinguished from degradation-driven expansion in testing?
Elastic swell is fully established within a short distance of the die exit, is reproducible at a given shear rate and temperature, and disappears when shear rate is lowered or L/D is increased. Degradation-driven expansion appears as internal bubbles and voids, worsens with residence time and temperature, correlates with TGA mass loss, and is irreversible. Combining on-line laser swell measurement with extrudate density and microscopy cleanly separates the two.
Which standards govern the testing of PEEK melt expansion behavior?
Capillary rheometry follows ASTM D3835 or ISO 11443, oscillatory rheometry follows ASTM D4440 or ISO 6721, melt flow rate follows ASTM D1238 or ISO 1133, PVT dilatometry follows ISO 17744, DSC follows ISO 11357, TGA follows ISO 11358, and density determination follows ASTM D1505.
How can die swell be minimized in production?
Effective countermeasures include increasing die land length, raising melt temperature within the safe window, lowering wall shear rate through die redesign, applying controlled draw-down between the die and haul-off, using fiber-filled grades where dimensional precision outweighs isotropy, enforcing strict drying and residence time control, and iterating die orifice geometry using measured swell ratios from capillary rheometry.
