I started 3D printing to make replacement parts. I've ended up using it to prototype mechanical components, enclosures, and tooling fixtures for software-driven hardware projects.
Here's what I wish someone had told me at the start.
Think in Tolerances, Not Dimensions
The most common beginner mistake is modeling parts to exact dimensions and expecting them to fit together perfectly. Filament-based printing doesn't work that way.
A 10mm hole modeled at 10mm will measure 9.6mm after printing. A 10mm peg modeled at 10mm won't fit in it.
Start every project by printing a calibration piece for your specific filament-printer combination. Establish your shrinkage factor. Then build it into your CAD workflow, not as an afterthought, but as a first-class constraint.
For clearance fits (things that need to slide), I typically add 0.2mm clearance per mating surface. For press fits, 0.05mm interference. For snap fits, you'll just have to iterate — there's no substitute for prints.
Material Selection Is an Engineering Decision
PLA is fine for prototyping. It's not fine for most functional applications:
- Not great above 60°C
- Brittle under cyclic loading
- Absorbs moisture and weakens over time
For anything that will see real use, consider:
- PETG: Better temperature resistance, more flexible than PLA, easier than ABS. My default for functional parts.
- ASA: Good UV resistance, reasonable mechanical properties. Use it for outdoor hardware.
- PA (Nylon): Excellent mechanical properties, great for anything that needs real toughness. Moisture control during printing is critical.
- PC: High temperature and impact resistance. Difficult to print well, but sometimes the right answer.
The material selection conversation should happen at the design stage, not after you've already modeled the part.
Anisotropy Is Real
Printed parts are not isotropic. The layer orientation has a massive effect on strength. A part might be rock solid under compression but fail immediately under tension applied perpendicular to the layers.
When designing functional parts, think about the primary load path and orient your print so the layers are perpendicular to the tensile stress direction. This often means printing parts in a less obvious orientation, sometimes with more supports.
For critical structural applications, add features to your design that mechanically interlock layers — ribs, gussets, interlocking geometry. Don't just rely on layer adhesion.
Wall Thickness and Infill Are Not Decorative
Slicer settings matter. A lot.
For functional parts:
- Perimeters/shells: More walls = more isotropic strength. I rarely go below 4 perimeters for structural parts.
- Infill pattern: Gyroid and honeycomb patterns distribute stress more evenly than rectilinear. Use them for anything that will see varied loading.
- Infill percentage: 40-60% for most structural parts. 100% infill is slower and doesn't always give you proportionally better properties.
These settings aren't just about the final product — they affect how you design the part. A design that relies on high infill density to work is a design that needs more thought.
The Iteration Process
Print → test → fail → understand why → redesign → repeat.
That's the process. Trying to avoid iteration by designing the perfect part on the first try is slower, not faster. Design for rapid iteration: small parts, quick print times, clear failure modes.
Document what you try. I keep a simple log for each project: print settings, material batch, what failed, what changed. It sounds tedious. Six weeks later you'll thank yourself.
The parts that have worked best in my projects are never the ones I designed once and printed once. They're the ones I printed eight times and understood deeply by the third failure.
That's the process.