Modeling of PEEK crystallization kinetics under transient thermal conditions

This thesis develops a unified kinetic model that captures poly-ether-ether-ketone (PEEK) crystallization over a broad temperature 𝑇 window (near glass transition temperature 𝑇g ~155°C < 𝑇 < near melting temperature 𝑇m ~310°C); and across cooling rates spanning from quasi-isothermal to ~10³ °C/min, which are inaccessible by conventional differential scanning calorimetry (DSC). Existing crystallization models are typically calibrated on narrow isothermal ranges and slow constant-rate experiments using conventional DSC. As a result, they break down under the fast, transient cooling profiles characteristic of additive manufacturing (AM) and often require recalibration for different conditions. Consequently, no experimentally validated, mechanistic framework exists that can take an arbitrary thermal history 𝑇(𝑡) representative of AM and reliably predict crystallinity. This work addresses that gap by utilizing experimental datasets collected via both (DSC) and fast scanning calorimetry (FSC), enabling measurement of isothermal and non-isothermal crystallization kinetics under AM-relevant conditions. A parallel dual Avrami model was applied to the isothermal and non-isothermal data for calibration, capturing both the faster primary and the slower secondary crystallization contributions through temperature-dependent weight factors that reproduce the observed dome-shaped trend across 𝑇g < 𝑇 < 𝑇m. The final model is extended with two novel cooling-rate-dependent scalars, a rate-aware induction-time law, and a temperature-dependent “maximum isothermally achievable enthalpy function” so that it outputs absolute enthalpy. Validation against a measured AM-like transient cooling program shows that the unified model reproduces the crystallization onset time and enthalpy evolution with R²≈0.95, providing a practical route to crystallinity prediction and process-path design for PEEK AM.