Multi-scale modelling of polymeric deposition for sustainable additive manufacturing: towards customised implant fabrication

Droplet-based polymer additive manufacturing is increasingly considered viable for customised low-volume production, including patient-specific medical components. Industrial adoption is constrained by process variability and reliance on trial-and-error tuning, which increases waste, energy use and lead time while undermining repeatability. This thesis addresses these challenges for Arburg Plastic Freeforming, a material-jetting process combining injection-moulding plasticisation with discrete droplet deposition of thermoplastic granules. The work develops modelling-informed tools and evidence-based insights to improve reliability across scales and support feed-forward understanding. A progressive model–material strategy is adopted, using amorphous and semi-crystalline thermoplastics. At the macro-scale, crystallisation-driven distortion in semi-crystalline polymers is examined. A modified polyamide 6 formulation shift the crystallisation peak from 184.6 °C to 161.2 °C and reduces warpage from 5.72 mm to 0.06 mm (–98.9%). Mechanical characterisation reveals marked anisotropy and conditioning sensitivity: moisture conditioning lowers elastic modulus by up to 70 – 76% and changes ductility with raster orientation. At the meso-scale, an analytical framework derived from mass conservation and deposition kinematics links droplet geometry to part density as a practical quality proxy. The model is verified via mass-balance closure and evaluated against Archimedes-based density measurements, with parity analysis and error metrics showing mean absolute errors of approximately ~0.01 g/cm3. The analysis explicitly accounts for machine-induced shutter-frequency drift at high deposition speeds. At the micro-scale, a computational fluid dynamics analysis resolves droplet spreading, coalescence, and interfacial evolution in polycarbonate. Extrusion temperature governs morphology and wetting, with droplet volume increasing from ~0.0123 mm3 to ~0.0159 mm3 between 285 °C and 305 °C. A William-Landel-Ferry equivalent-time metric shows that bulk-like bonding strength is reached within ~10-5 – 10-4 s, with the most balanced response at intermediate temperature conditions. Overall, the thesis provides a physically interpretable basis for improving process reliability through distortion-aware material strategies, feed-forward density screening and mechanistic interpretation of bonding.