Introduction

Additive manufacturing of biomedical scaffolds demands implant-grade dimensional fidelity under clean-room constraints. In this thesis, I optimized a magnetic planar-drive material-extrusion printer (Beckhoff XPlanar) to fabricate polycaprolactone (PCL) lattice structures for implant applications. The practical challenge was to define a robust process window that preserved geometric accuracy and bridge integrity while remaining reproducible in operation.

I investigated how motion planning, deposition (layer) height, raster width, mover levitation height, and thermal stability influenced fidelity. Guided by a structured Design of Experiments (full-factorial followed by Box–Behnken), I quantified main effects and interactions using dimensional metrology and Digital Image Correlation, and I identified layer height as the dominant driver of geometric error.

Model predictions (peak accuracy ≈ 98.289%) were validated through confirmation prints (observed 97.17–97.71%). Bridge collapse at the mathematical optimum motivated a shift from a single “best” point to a practical operating window, supported by standard operating procedures for motion and thermal control. The outcome was a reproducible parameter set and documentation that enabled consistent, implant-grade prints in a clean-room environment.