Functional prototypes, not display models.

There are two very different things people mean by "prototype". One is a shape you hold up in a meeting. The other is a part that has to go into the assembly, take the load, survive the temperature and tell you whether the design is right. We build the second kind.

The difference shows up in the decisions around the part rather than in the printer. A display model can be printed any which way. A functional prototype has a load path, and the layer orientation either supports it or quietly undermines it.

Why iteration speed matters more than unit price

On a prototype the interesting number is rarely the part cost. It is how many days pass between "this doesn't fit" and having the corrected part in your hand. With no tooling in the loop, a revision is a new file and a new build. That is the entire argument for FDM at this stage, and it is a strong one.

Because there is no tooling to protect, there is also no incentive to defend a bad design. If revision two needs a 2 mm change, make the 2 mm change.

What we look at before printing a prototype

• Where the load goes. Layer lines across a bending stress are the most common reason a printed prototype fails early and misleads you about the design.
• Which fits are real. Bearing seats, shaft clearances, connector cutouts and mating faces get planned; cosmetic surfaces do not need to.
• What it has to survive. Sunlight, solvents, a hot enclosure, repeated assembly cycles. Each pushes toward a different material.
• Whether the geometry is fighting the process. A wall that is 0.9 mm because CAD defaulted there, an unsupported 3 mm overhang, a hole that has to be round to work. These are cheap to fix in the model and expensive to discover on the bench.

When something in that list looks wrong, you hear about it during review, before the part is built, not after.

Typical prototype work

• Enclosures and housings for early hardware builds
• Brackets, mounts and structural interfaces
• Mechanism parts: linkages, carriers, guides, cam surfaces
• Fit-check and form-study parts ahead of a molded or machined design
• Replacement parts for equipment where the original is obsolete

From prototype to production

A prototype that works often just keeps going. Because the same process makes the next hundred, there is no requalification step between "it worked" and "build fifty". See low-volume production for how that scales, and what pricing does as quantity grows.

If the part eventually belongs in a mold, printed prototypes are still the cheapest way to be sure the geometry is right before you commit to tooling.