PEEK Injection Molding Expansion and Shrinkage Analysis: A Complete Technical Guide

PEEK injection molding expansion and shrinkage analysis is essential for achieving tight dimensional control in high-performance engineering components made from polyetheretherketone. This comprehensive guide explains the crystallization behavior, thermal expansion characteristics, testing standards, and process parameters that govern PEEK dimensional stability during and after injection molding. Whether you are designing aerospace seals, automotive bushings, or medical implants, understanding PEEK shrinkage and expansion mechanisms will help you reduce scrap rates and meet stringent tolerance requirements.

Technical Overview

PEEK (polyetheretherketone) is a semi-crystalline, high-performance thermoplastic belonging to the polyaryletherketone (PAEK) family. With a melting point of approximately 343°C, a glass transition temperature of about 143°C, and continuous service temperatures up to 260°C, PEEK occupies a leading position in demanding engineering applications. However, these exceptional thermal properties come with significant dimensional behavior challenges during injection molding.

Because PEEK is a semi-crystalline polymer, it undergoes substantial volume changes during cooling and solidification. As melt transitions from the amorphous state to a crystalline structure, the material densifies, producing mold shrinkage that typically ranges from 1.2% to 1.8% for unfilled grades. This value is considerably higher than that of amorphous polymers such as PSU or PEI, which shrink only 0.5% to 0.7%. Filled grades behave quite differently: 30% glass-filled PEEK typically shrinks 0.3% to 0.5%, while 30% carbon fiber reinforced grades can achieve shrinkage as low as 0.1% to 0.3% with improved isotropy.

Expansion behavior is equally important. The coefficient of linear thermal expansion (CLTE) of unfilled PEEK below the glass transition temperature is approximately 47 µm/(m·°C), rising sharply to 108 µm/(m·°C) or more above Tg where the amorphous phase softens. Glass fiber reinforcement reduces CLTE to roughly 22 µm/(m·°C), and carbon fiber reinforcement further lowers it to approximately 15 µm/(m·°C). Accurate prediction and compensation of both shrinkage and thermal expansion are therefore critical for parts that must maintain tolerances across wide temperature ranges or that interface with metal components.

Key Factors Influencing PEEK Shrinkage and Expansion

Crystallinity and Mold Temperature

Mold temperature is the single most influential parameter for PEEK shrinkage. To develop full crystallinity (typically 30% to 35%), mold temperatures of 170°C to 200°C are required. When the mold runs too cold (below 150°C), an amorphous skin layer forms, immediate shrinkage decreases, but the part will exhibit uncontrolled post-molding shrinkage when exposed to temperatures above Tg in service. This delayed crystallization can cause dimensional drift, warpage, and part failure in high-temperature applications.

Melt Temperature and Injection Speed

PEEK melt temperatures normally range from 360°C to 400°C. Higher melt temperatures reduce viscosity and improve cavity filling but increase the total thermal differential during cooling, slightly increasing shrinkage. Excessive residence time at high melt temperatures also triggers thermal degradation, which alters molecular weight and can produce inconsistent crystallization behavior from shot to shot.

Holding Pressure and Time

The packing phase compensates for volumetric shrinkage during solidification. Insufficient holding pressure or premature gate freeze-off leads to excessive shrinkage, sink marks, and voids, particularly in thick-wall sections. Optimal holding pressure for PEEK is typically 50% to 80% of injection pressure, with holding time calibrated to part weight plateau verification.

Fiber Orientation and Anisotropy

Reinforcing fibers align with the flow direction during injection. This produces strongly anisotropic shrinkage: glass-filled PEEK may shrink 0.3% in the flow direction but 0.6% or more in the transverse direction. This differential shrinkage is a primary driver of warpage in flat plaques, gears, and structural components, and must be addressed through gate placement, flow balance, and fiber orientation simulation.

Post-Molding Thermal History

Stress relief and annealing treatments performed at 200°C to 250°C can intentionally complete crystallization and stabilize dimensions. Parts molded at lower mold temperatures benefit from controlled annealing, which adds approximately 0.5% additional shrinkage but eliminates unpredictable dimensional drift in service.

Moisture and Environmental Absorption

PEEK absorbs only about 0.5% water at saturation, so hygroscopic expansion is minimal compared to nylons. However, for ultra-precision applications with tolerances under 0.02 mm, even this small moisture-related dimensional change should be factored into the tolerance stack-up analysis.

Testing Samples

Standard Shrinkage Plaques and Bars

Molded flat plaques of 100 mm × 100 mm × 3 mm or standardized bars following ISO 294-1 Type D1/D2 geometries serve as the baseline specimens for mold shrinkage determination. Rectangular bar specimens of 127 mm × 12.7 mm × 3.2 mm are specified by ASTM D955. These geometries allow measurement in both flow and transverse directions to quantify anisotropy.

Molded Discs for Isotropic Evaluation

Center-gated discs of 100 mm to 150 mm diameter with radial flow patterns reveal the radial versus circumferential shrinkage differential, which is critical for predicting warpage in ring-shaped components such as seals, bearings, and valve seats.

Prototype and Production Components

Actual part geometries, including thick-wall bushings, thin-wall electronic housings, threaded medical implants, and machined stock shapes, are measured to validate simulation predictions and establish production process windows. First article inspection reports typically combine CMM dimensional data with process parameter records.

Thermomechanical Analysis Specimens

For CLTE measurement, specimens are cut from molded plaques in specified orientations, typically machined to 5 mm × 5 mm × 20 mm prisms for TMA testing per ISO 11359-2 or ASTM E831.

Testing Items

• In-plane mold shrinkage in flow direction and cross-flow direction, expressed as a percentage of mold cavity dimensions
• Post-molding shrinkage after 24-hour, 48-hour, and multi-day conditioning at 23°C/50% RH
• Annealing-induced shrinkage measured after heat treatment at 200°C to 250°C
• Coefficient of linear thermal expansion below and above glass transition temperature
• Volumetric shrinkage distribution through the part thickness
• Warpage and out-of-plane distortion, quantified by flatness or profile tolerance deviation
• Differential scanning calorimetry (DSC) crystallinity determination, typically 30% to 35% for well-molded PEEK
• Sink mark depth and internal void assessment in thick sections, using X-ray CT or ultrasonic inspection
• Density measurement per ISO 1183 to correlate with crystallinity and shrinkage
• Reheat shrinkage and dimensional stability during simulated service thermal cycling from -60°C to +250°C

Testing Methods

Mold Shrinkage Measurement

Mold shrinkage is determined per ASTM D955 or ISO 294-4. Specimens are molded under controlled conditions, and after a defined conditioning period of 16 to 24 hours at standard laboratory atmosphere, dimensions are measured with calibrated instruments such as optical comparators, vision measuring systems, or CMMs. Shrinkage is calculated as the difference between cavity dimension and molded part dimension divided by the cavity dimension, multiplied by 100. Measurements are taken 24 hours after molding and optionally after annealing to capture total shrinkage.

Thermomechanical Analysis for Thermal Expansion

TMA testing per ASTM E831 or ISO 11359-2 measures dimensional change as the specimen is heated at a controlled rate, typically 5°C/min, from room temperature to 300°C. The test yields CLTE values for the glassy region below Tg and the rubbery region above Tg. Anisotropic specimens cut parallel and perpendicular to flow reveal orientation-dependent expansion behavior in filled grades.

Differential Scanning Calorimetry for Crystallinity

DSC analysis per ISO 11357 measures the melting enthalpy of the molded sample, from which crystallinity is calculated using the theoretical enthalpy of 100% crystalline PEEK (130 J/g). This confirms whether mold temperature and cooling conditions achieved target crystallinity and predicts post-molding dimensional drift risk.

Thermal Cycling and Dimensional Stability Testing

Molded parts undergo programmed thermal cycling between extremes such as -40°C and +250°C, with dimensional measurements between cycles. This method captures hysteresis effects, permanent set from additional crystallization, and verifies that annealed parts remain dimensionally stable throughout the intended service envelope.

Injection Molding Simulation

CAE tools such as Moldflow, Moldex3D, or Sigmasoft are used with validated PEEK material databases containing pvT (pressure-volume-temperature) data, viscoelastic properties, crystallization kinetics models, and fiber orientation tensors. Simulation predicts local volumetric shrinkage, sink marks, and warpage before tool cutting, enabling compensation strategies such as steel-safe cavity sizing, gate relocation, and cooling channel optimization.

X-Ray Computed Tomography

Industrial CT scanning detects internal voids, density gradients, and full 3D geometric deviation of complex parts without destructive sectioning. This method is particularly valuable for validating wall thickness effects on shrinkage in medical and aerospace components with tight tolerance requirements.

Applications

Aerospace components including compressor seals, wear rings, fasteners, and clips rely on documented shrinkage data to achieve press-fit dimensions that remain stable across -55°C to +200°C service ranges. Carbon fiber reinforced PEEK brackets and structural clips demand anisotropy control to meet aircraft flatness and assembly tolerance requirements.

Automotive powertrain applications such as piston skirts, bearing cages, gear shift forks, seal rings, and transmission components exploit PEEK dimensional stability under continuous loads at elevated temperatures, where predictable CLTE ensures proper running clearances against metal counterparts over hundreds of thousands of operating cycles.

Medical device applications include sterilization trays, surgical instruments, dental abutments, spinal cages, and trauma implants. These parts must maintain dimensional accuracy after repeated autoclave cycles at 134°C, which sits below the glass transition but above typical molding mold temperatures in some cases, making full crystallization and annealing verification mandatory.

Semiconductor and electronics manufacturing equipment utilize PEEK vacuum chamber components, wafer carriers, test sockets, and insulators where thermal expansion mismatch with silicon, ceramics, and metals must be tightly controlled. Bearing-grade and unfilled PEEK bushings and wear pads in oil and gas downhole tools require shrinkage compensation to maintain interference fit integrity at 150°C to 200°C operating temperatures.

FAQ

What is the typical mold shrinkage rate of unfilled PEEK?

Unfilled natural PEEK exhibits mold shrinkage of approximately 1.2% to 1.8%, with the exact value depending on part geometry, wall thickness, mold temperature, and packing conditions. Shrinkage is generally slightly higher in the transverse direction than in the flow direction for unfilled material.

How does fiber reinforcement reduce PEEK shrinkage?

Glass and carbon fibers have near-zero thermal shrinkage and mechanically constrain the polymer matrix. A 30% glass-filled grade shrinks only 0.3% to 0.5%, while 30% carbon fiber reinforcement reduces shrinkage to 0.1% to 0.3% and greatly improves directional uniformity, minimizing warpage.

Why does PEEK shrink more than amorphous high-temperature plastics?

The crystalline phase packs polymer chains into an ordered, denser structure. During cooling, PEEK crystallizes by 30% to 35%, causing volumetric contraction far greater than amorphous engineering plastics like PEI or PSU, which solidify without a phase transition and shrink only 0.5% to 0.7%.

What happens if PEEK is molded with a mold temperature that is too low?

Mold temperatures below 150°C suppress crystallization, producing a partly amorphous part with lower immediate shrinkage but unstable dimensions. When the part later experiences temperatures above 143°C in service or during sterilization, additional crystallization occurs, causing unpredictable shrinkage, warpage, and potential fit failure.

Is annealing always necessary for PEEK parts?

Annealing at 200°C to 250°C is recommended when parts are molded at lower mold temperatures, when tolerances are tighter than 0.1%, or when service temperatures approach or exceed Tg. Parts properly molded at 180°C to 200°C with full crystallinity often require no annealing, though a stress-relief cycle still improves long-term stability.

How can shrinkage variation between shots be minimized?

Maintain stable melt temperature (360°C to 400°C), consistent mold temperature (170°C to 200°C for crystalline surfaces), sufficient holding pressure and time until gate freeze, uniform wall thickness design, and verified drying of resin at 150°C for 3 hours. Statistical process control on part weight provides the earliest indicator of packing consistency.

What tolerance capability can be expected from molded PEEK parts?

General molding practice supports tolerance grades around ±0.1% of nominal dimension for unfilled PEEK and ±0.05% or better for well-controlled filled grades on critical dimensions. Ultra-tight tolerances below 0.02 mm typically require mold shrinkage compensation plus post-molding machining or grinding of critical features.

How does PEEK thermal expansion compare with metals?

Unfilled PEEK expands about 47 µm/(m·°C) below Tg, which is roughly three times that of aluminum and six times that of steel. Carbon fiber filled PEEK at approximately 15 µm/(m·°C) approaches aluminum, making it the preferred grade for metal-interfacing components where thermal expansion matching prevents loosening or overstressing at temperature extremes.