This study develops a kinetic model that captures poly-ether-ether-ketone (PEEK) crystallization over a temperature window from glass transition () to melting () temperature, and across cooling rates from 5 to ~103 °C/min. The framework is a parallel dual-Nakamura formulation whose isokinetic parameters {} are obtained from a bi-level non-linear regression of isothermal crystallization tests conducted using a flash-differential scanning calorimeter (FSC). The weight partitions the faster primary and slower secondary crystallization and is represented by a physics-based analytical function that captures its dome-shaped temperature dependence. A maximum isothermally achievable enthalpy function is introduced so that the model predicts enthalpy natively under arbitrary thermal profiles. To extend this isothermal backbone to non-isothermal conditions, two explicit cooling-rate-dependent scalars are introduced, and , which shift and limit attainable crystallinity at high cooling rates respectively. Finally, a rate-dependent induction time relation is added to adjust the onset of crystallization. Calibrating these rate functions against non-isothermal experiments, while keeping the isokinetic parameters fixed, yields a single isothermal–non-isothermal model that predicts under arbitrary profiles. Model performance is validated using an interrupted FSC experiment with a multi-segment cooling program that mimics a local transient thermal history of PEEK during additive manufacturing. The sample is cooled through successive constant-rate segments with intermittent quench–remelt cycles to probe the accumulated crystallinity along the path. Without additional fitting, the model predicts the measured enthalpy evolution with R2 ≈ 0.95. The framework thus provides a practical route for predicting polymer crystallinity under processing-relevant thermal histories.
This work presents a microstructure-informed pathway for assigning a material symmetry class to distinguish between tensor components and scalar effective thermal conductivity (ETC) values derived from directional measurements. The framework combines directional thermal measurements with three-dimensional statistical quantification of microstructural features (fibers and voids) to assess whether symmetry assumptions required for tensorial interpretation are justified. Three distinct microstructures of short carbon fiber-reinforced polyetherimide composite were analyzed, with the microstructure statistics altered by the melt extrusion additive manufacturing process parameters. The directional temperature-rise history in the material samples was measured using the Transient Plane Source sensor. The statistics obtained from 3D images of microstructural features were used to assess the material’s anisotropy class to justify the applicability of the transverse isotropic regression method for ETC. One microstructure exhibited characteristics consistent with a statistical transverse isotropy idealization, enabling inference of the ETC tensor; the others did not, and their directional ETC values are treated as test-specific parameters obtained from isotropic model fits. The results also demonstrate that microstructure parameters may strongly influence directional thermal transport. More broadly, this work highlights the need for microstructure-informed justification when interpreting directional measurements as tensor components rather than configuration-dependent scalars, underscoring a critical unresolved gap in the experimental characterization of general anisotropic ETC tensors.