Archived

This content is available here for research, reference, and/or recordkeeping.

Abstract

This work develops a GPU-accelerated, bead-resolved digital twin for FFF that reconstructs super-elliptic filament geometries from G-code and simulates transient heat transfer via voxel-level activation synchronised to deposition order. Applied to a parametric study of two metal–polymer composites (Cu–PLA and IN718–PLA) across five super-ellipse exponents (n = 2.0, 2.7, 2.8, 2.9, 3.0), the framework reveals a sharp thermal regime transition: a 43% increase in interlayer contact width (n = 2.0 → 2.7) and 53% (n = 2.0 → 2.9) produce negligible thermal change (|Cohen’s d| < 0.02 for IN718-PLA and < 0.2 for Cu-PLA), while n = 3.0 triggers +150–178% increases in bonding windows, −55 to −60% reductions in cooling rate, and a 27–45-fold increase in thermal reactivation. The transition amplifies across temperature thresholds (2× at 170 °C, 15.8× at the glass transition), reverses the deposition-order thermal gradient (r = −0.83 to +0.64 for IN718-PLA and −0.55 to +0.81 for Cu-PLA), and causes surrogate model failure (cross-regime R² = −4.6). These findings establish bead morphology as a sharp morphology-driven threshold requiring regime-aware, material-specific digital twin. At the transition, weld-strength increases while cooling-rate decreases, indicating improved bonding uniformity and reduced residual-stress heterogeneity.

Document Type

Article

Publication Date

1-1-2026

Notes/Citation Information

Publisher Copyright: © 2026 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.

Digital Object Identifier (DOI)

10.1080/17452759.2026.2675817

Archival?

Archival

Share

COinS