The 4 processes you actually need to know
There are dozens of 3D printing technologies. For hardware prototyping, you'll use one of these four. Here's how to choose.
| Process | Best for | Tolerance | Cost (cm³) | Lead time |
|---|---|---|---|---|
| FDM (Fused Deposition) | Internal brackets, large housings, fast iteration | ±0.5mm | $0.50–2 | Same day |
| SLA (Resin) | Snap-fits, fine detail, surface finish | ±0.1mm | $3–8 | 1–2 days |
| SLS (Nylon powder) | Functional parts, living hinges, complex geometry | ±0.3mm | $4–10 | 3–4 days |
| MJF (HP Multi Jet Fusion) | Production-like parts in small batches | ±0.2mm | $5–12 | 3–5 days |
Decision tree
- Does it need to look like injection molding? → SLA or MJF
- Does it have snap-fits or living hinges? → SLS (Nylon PA12 is flexible enough for both)
- Is it bigger than 200mm? → FDM (others get expensive fast)
- Is this your 5th+ iteration? → SLA or SLS only — FDM tolerances have eaten your cumulative budget
- Is it a one-off jigsaw bracket for an assembly test? → FDM, even ugly
Tolerance budgets that actually work
Marketing materials say SLA is "±0.05mm." That's a lie. Realistic tolerances for well-printed, well-designed parts:
- SLA: ±0.15mm for features <30mm, ±0.25mm beyond
- SLS: ±0.3mm constant (Nylon shrinks)
- FDM: ±0.4mm (depends heavily on orientation and printer)
- MJF: ±0.2mm (most consistent of the four)
For snap-fits, we use the conservative budget: design the cantilever thickness +0.3mm tolerance, then sand down if needed. A snap-fit that needs 0.8mm cantilever and you print at 0.8mm will sometimes fit, sometimes not. Print at 1.1mm and sand to 0.85mm for guaranteed fit.
Material cheat sheet
Skip the catalog. Here are the four materials you'll use 95% of the time:
- PLA (FDM): Cheap, fast, ugly. Internal only. Don't quote this to investors.
- PETG (FDM): Functional, food-safe-ish, better temperature resistance. Default for mechanical prototypes.
- Tough Resin (SLA): Looks like ABS, feels like ABS, fails like ABS if you drop it. Great for fit-and-finish tests.
- PA12 Nylon (SLS): The workhorse. Living hinges, snap-fits, functional gears. Use this.
Finishing that matches injection molding
Once you've printed, the surface is usually rough or has visible layer lines. To match the look of an injection-molded part:
- SLA: Sand with 400 → 800 → 1200 grit, then spray filler primer, then sand again, then paint. 4-6 hours per part.
- SLS: Vapor smoothing with solvent (we use a proprietary process, but you can get 80% there with a hand-applied finish coat).
- MJF: Already looks "production-ish." Just dye the color and you're done.
- FDM: Don't try. Investors can tell.
Our rule: if the part goes in front of a customer, we SLA or MJF. If it stays internal, FDM is fine.
3D printing for injection-molding tooling
The underrated use case: 3D-printed injection molds for low-volume runs (50-500 units). We use this for our customer projects all the time.
- Print the mold in SLA tough resin or SLS metal-filled
- Machine it to spec
- Inject 50-300 parts before the mold wears out
- Total cost: 1/10th of an aluminum tool
This is how we deliver first batches in 7-10 days without $30k tooling.
Common mistakes
- Printing at 100% infill. 30% gyroid is stronger than 100% rectilinear for most loads.
- Not orienting for strength. Layer lines are weak. Put stress perpendicular to layers, not along them.
- Forgetting draft angles. Anything going into a mold needs 1-2° draft. SLA prints snap apart without it.
- Ignoring post-cure. SLA parts need UV cure or they're weak for 24 hours.
- Trusting online quotes. Lead times are off by 2x in our experience. Plan for it.
When to hand off to us
If your prototype needs to survive:
- Drop tests (1.5m onto concrete)
- IP rating (splash, dust)
- UV exposure (a week of sun)
- Repeated mating cycles (>100 insertions)
...then 3D printing alone isn't enough. We typically bridge with SLA + spray-coating for ~50-100 units, then move to injection molding for the real production run.
Send us your CAD at [email protected] for a free process recommendation.