A fully printable three-point bend fixture
Some specimens are far too delicate for commercial bend fixtures, which are built for metals and plastics and will damage a fragile sample before the test even begins. SunThru needed a three-point bend fixture matched to its own samples. I designed a fully printable 12-part jig in SolidWorks 2025, built entirely around gentle handling.
| Role | Design engineer (solo) | For | SunThru · early-stage hardware R&D |
|---|---|---|---|
| Tools | SolidWorks 2025 · VBA macro automation · FDM 3D printing | Timeline | 2026 · in progress |
| Result | 12-part fully printable fixture, in lab use | Status | First parts printed and in use |
The problem
Three-point bend testing is a standard way to measure flexural strength, but commercial fixtures assume a specimen that can take being clamped. When the sample cannot, the fixture itself becomes the thing that breaks it. SunThru needed a fixture sized to its own samples, gentle enough not to pre-damage them, and cheap enough to iterate: which meant 3D-printing it in-house.
Constraints
- Every part printable on the lab’s FDM printer, with no machined components
- 12 interlocking parts sharing one coordinate system, so dimension changes ripple predictably
- Standard M3 hardware throughout, with printable cross-hole channels
- Compliant contact geometry at the load nose and both supports, so the specimen sees a clean three-point load and nothing else
Parametric part design
A 12-part fixture with interlocked dimensions is a maintenance problem: change one dimension by hand and eleven other parts silently stop fitting. I kept every part on a shared coordinate system and automated the rebuilds with SolidWorks VBA, so a dimension change regenerates the affected parts consistently, with each part’s geometry checked against its expected volume before acceptance.
The parts themselves are designed for the printer: additive layer-friendly geometry throughout, and horizontal M3 cross-holes shaped as printable channels so nothing needs support material or post-drilling.
Where it stands
- 12 interlocking parts on a shared coordinate system, fully FDM-printable
- Standard M3 hardware throughout, with printable cross-hole channels
- Base plate and gauge plate printed and in use; remaining parts in staged delivery
- first test-series result or repeatability figure here
What I learned
The robust design decisions were the ones that thought like the manufacturing process: geometry built up in layers the way the printer builds it, holes shaped for printability instead of drills, and hardware chosen from one standard size. Designing for the process beat designing around it.