Research paperPreprintSubmitted to ISEC ’2713 pagesIEEE format
Machine-Tool Construction Practice Applied to a DIY Filament 3D Printer
Design, Failure Analysis, and Rebuild of “Boaz”
Abstract. Do-it-yourself filament 3D printers inherit their construction methods from the RepRap project, which optimized for self-replication from cheap parts rather than for accuracy. The paper tests whether five practices from machine-tool building transfer to this class of machine: exact constraint, interference fits on locating features, kinematic couplings, stiffness before squareness, and metrology performed during assembly.
The test article is Boaz, a CoreXY printer that was built, run in production, diagnosed, and then rebuilt. Both builds share the same design, kinematics and most components, and differ in how well the precision practices were executed, so the first build serves as a control. Eight performance criteria were fixed before construction. The first build failed at least four. The rebuild met seven of the eight, with runout at every pressed idler below 0.001 in, a bed mesh envelope of 0.2 mm over a 100 × 100 grid, and roughly 100 consecutive production plates with no first-layer intervention; the Y-axis resonance sweep capped acceleration at 9,500 mm/s² against a 20,000 mm/s² target.
Over a ten-day, 1,000-part PETG contract it held 0.2 mm functional tolerance on threaded parts at about 50 parts per day, against 60 for a $300 commercial machine alongside it. The throughput comparison was lost; the accuracy claim was not. The paper reports eleven construction rules with the observation that earned each.
- Additive manufacturing
- CoreXY
- Precision engineering
- Kinematic coupling
- Interference fit
- Metrology
- Input shaping
- Klipper
Preprint, under review. This version is dated 14 September 2026 and has not been peer reviewed. The paper has been submitted for review to the IEEE Integrated STEM Education Conference (ISEC ’27).
