Computational thermal analysis of large-format additive manufacturing for CF/PAEK with integrated localized heating

Large-format additive manufacturing (LFAM) of high-performance polymers like carbon fibre-reinforced polyaryletherketone (CF/PAEK) faces significant thermal management challenges, where uncontrolled cooling rates and thermal gradients can compromise interlayer bonding and are key drivers for residual stress development. While localized heating has emerged as a promising strategy to modulate thermal histories, predictive models capable of capturing its coupled effects with deposition dynamics in LFAM remain underdeveloped. This study presents a high-fidelity finite element framework to simulate the transient thermal behaviour in LFAM with integrated localized heating. Using the Abaqus AM module, the model incorporates a moving double-ellipsoid heat source to represent pre-deposition heating, sequential element activation for material deposition, and dynamic cooling boundaries. The framework is employed to systematically investigate the influence of critical process parameters-including localized heating power, nozzle-to-heater distance, layer thickness, and printing speed–on the thermal profile at a representative interfacial location. Results demonstrate that localized heating effectively elevates the thermal baseline, reduces cooling rates, and extends the dwell time above the glass transition temperature, thereby promoting conditions favourable for interlayer diffusion. The analysis reveals a strong, non-linear coupling between heating power and printing speed in setting the pre-deposition interface temperature. Furthermore, an optimal balance between layer thickness and heater penetration depth is identified to maximize thermal build-up while avoiding geometric instability. This computational work elucidates the thermal mechanisms governing LFAM with auxiliary heating and provides a validated foundation for optimizing thermal management strategies. The developed framework paves the way for implementing digital twins and physics-informed surrogate models to accelerate the development of robust, high-quality LFAM processes for advanced thermoplastic composites.