The Five Techniques That Cover Almost Every Prototype

Five processes handle the overwhelming majority of rapid prototyping work: FDM printing for cheap, fast form checks; SLA printing for fine detail and smooth surfaces; SLS printing for tough, functional nylon parts without support structures; CNC machining for real engineering materials and tight tolerances; and urethane casting for short runs of production-like parts. No single technique wins everywhere — mature projects mix them, matching each prototype's question ("does it fit?", "does it survive?", "will it mold?") to the process that answers it cheapest and fastest. Here is how they compare in practice.

FDM: The Workhorse for Speed and Cost

Fused deposition modeling extrudes melted thermoplastic layer by layer. It is the least expensive and most accessible process — simple parts cost single to low tens of dollars from a service, often with next-day turnaround.

  • Best for: early form and fit checks, jigs and fixtures, large cheap parts, brackets and enclosures where surface finish is secondary.
  • Limits: visible layer lines, weaker along layer boundaries, modest fine-detail resolution.
  • Materials: PLA, PETG, ABS, ASA, plus engineering blends with carbon fill.

SLA: Detail and Surface Finish

Stereolithography cures liquid resin with a laser or light source, producing smooth, highly detailed parts that can approach injection-molded appearance after light finishing.

  • Best for: appearance models, small intricate features, snap-fit detail checks, master patterns for casting.
  • Limits: standard resins are more brittle than molded thermoplastics and can degrade with UV exposure; less suited to hard functional abuse.
  • Cost: moderately above FDM for comparable size.

SLS: Functional Parts Without Supports

Selective laser sintering fuses nylon powder, with unfused powder supporting the part — enabling complex geometries, living hinges, and interlocking assemblies impossible to print otherwise. Parts are genuinely tough and close to production-plastic behavior.

  • Best for: functional testing, enclosures with snaps and threads, ducting, parts heading for field trials.
  • Limits: grainy matte surface, limited material palette (mostly nylons and TPU), pricier than FDM.

All three printing processes — and how to buy them well — are covered deeper in our 3D printing guide and our review of 3D printing services for prototypes.

CNC Machining: Real Materials, Real Tolerances

CNC cuts parts from solid blocks of the actual production material — aluminum, steel, POM, polycarbonate, PEEK. When the question is strength, thermal behavior, or precision fits, machined parts answer it in the true material rather than a printed proxy.

  • Best for: load-bearing parts, tight-tolerance interfaces, metal components, validating designs in production-intent material.
  • Limits: costs rise steeply with geometric complexity; deep pockets, internal features, and organic shapes may need multiple setups or become impractical.
  • Cost: typically low hundreds of dollars and up per part, driven by machine time and setups — full numbers in our guide to CNC machining for prototypes.

Urethane Casting: The Bridge to Production

A silicone mold, cast around an SLA master, yields parts in polyurethane resins formulated to mimic production plastics and rubbers — including overmolded and colored parts. One mold typically produces on the order of a couple dozen parts.

  • Best for: pilot batches of tens of units, investor and trade-show units, field-test fleets — production look and feel without tooling investment.
  • Limits: per-part cost far above molding; mold life limits quantity; lead time of days to a couple of weeks.

Choosing: Match the Technique to the Question

  1. "Does the shape work in my hand?" — FDM, upgraded to SLA when appearance matters.
  2. "Do the mechanisms and snaps survive use?" — SLS, or CNC in final material for critical parts.
  3. "Does it meet tolerance and strength specs?" — CNC.
  4. "Can I put twenty production-like units in front of users?" — urethane casting.
  5. "Are we ready for thousands?" — that is production territory; see our manufacturing technologies guide for the handoff to tooling.

Budget note: iteration count matters more than per-part price. Three cheap FDM loops that catch design errors early beat one expensive "perfect" prototype that reveals them late — the cost dynamics are laid out in how much a prototype costs, and the overall build-test-learn strategy in our prototype development guide and prototyping pillar.

Prototype with Engineers Who Pick the Right Process

Projects House runs all five techniques daily — choosing per part, per iteration, so clients pay for answers rather than for impressive plastic. Tell us what you are building through the contact form and we'll recommend the fastest, most economical prototyping route for your product.