The Material Question Behind Every Prototype
Most prototype material decisions get made by default: someone picks the resin the 3D printer is loaded with, or mirrors whatever the production part will eventually use. Both shortcuts cause problems, and they cause different ones.
The question that actually matters is what the prototype needs to prove. A form-and-fit check, a drop test, a thermal soak, and a customer demo are four different parts, and they do not all want the same material. Sorting that out first usually eliminates two-thirds of the candidate list.
What Each Material Is Actually Good At
PA12 (SLS nylon)
PA12 is the default for laser-sintered functional parts, and for good reason. Sintered nylon is close to isotropic — properties in Z are within roughly ten percent of X/Y in a well-tuned build, unlike FDM parts where layer adhesion dominates. That makes it the honest choice when a part will be loaded in an unpredictable direction.
It also tolerates snap fits. PA12 has enough elongation to survive a deflecting cantilever latch repeatedly, which is why it gets used for living hinges and clip features that would crack in a stiffer resin. Heat deflection typically sits in the 90 to 110 °C range at 1.8 MPa depending on the powder grade, so under-hood or near-motor applications are out.
Two things to plan around. Sintered surfaces are matte and slightly porous — as-built parts feel grainy and will wick moisture, so dyeing or sealing matters for anything that sees liquids. And PA12 absorbs water slowly from ambient air, which nudges dimensions over weeks. A part measured the day it ships may not measure identically a month later.
ABS
ABS remains the working horse of injection moulded parts because it is cheap, processes forgivingly, and takes secondary operations well. It bonds with solvent or cyanoacrylate, paints without priming drama, and accepts electroplating — which matters more than people expect for appearance models.
The downsides are shrinkage and weather. Moulded ABS shrinks around 0.5 to 0.7 percent, and that shrinkage is not perfectly uniform, so large flat parts warp. Add a rib-heavy design and you will fight sink marks and curl. UV resistance is poor; an ABS enclosure left outdoors chalks and yellows within a season.
For prototypes, ABS is the right answer when the production part will be ABS. Substituting it for PC or nylon in a prototype and assuming the fit result transfers is a mistake worth calling out explicitly.
Polycarbonate
PC earns its place on three properties: impact strength, transparency, and heat resistance. Notched Izod figures commonly run several hundred joules per metre — an order of magnitude above ABS — and heat deflection temperature sits around 130 to 138 °C at 1.8 MPa. If a prototype has to survive being dropped, or sit near something warm, PC is usually the only one of the three that works.
It is also the most demanding to process. PC must be dried before moulding, typically several hours near 120 °C, or it hydrolyses and loses strength. It is notch-sensitive in a way that surprises designers: the same part with a sharp internal corner will fail where a radiused one survives. And it scratches easily, so clear parts usually need a hard coat if they will be handled.
Tolerance: Where Prototypes Quietly Disagree With Production
Here is the part that causes the most rework. A sintered PA12 part typically lands within ±0.2 to ±0.3 mm, or roughly ±0.3 percent, whichever is looser — and that is after the process is dialled in. A machined prototype holds ±0.05 mm without difficulty, and tighter if the feature is simple and the shop is set up for it. A moulded part sits somewhere defined by the tool, not by the machine.
So if a snap-fit latch gap of 0.4 mm is what you are validating, a sintered part is not going to tell you the truth. Machine it, or go straight to a prototype tool. Conversely, if you are checking whether a housing is comfortable to hold, a sintered part delivers that answer in three days for a fraction of the cost.
Choosing by What You Are Testing
- Form, fit, ergonomics: SLS PA12 or a machined part if the interface is tight. Speed wins.
- Structural load in an unknown direction: PA12, or machined aluminium if stiffness matters.
- Impact or drop: PC, and be honest about the radius in the design.
- Optical clarity: PC or PMMA. Sintered nylon cannot do transparent.
- Heat above roughly 110 °C: PC among these three; above that, look at high-temperature resins or metal.
- Parts that will be moulded eventually: prototype in the production material, in the production process, if the schedule allows.
The Substitution Trap
Prototyping in a stand-in material is normal and often necessary. The trap is forgetting to correct for it. ABS and PC differ in shrinkage by enough to change a press-fit into a clearance fit. A nylon prototype that feels tough tells you nothing about how a glass-filled grade will behave, because the fibre orientation in a moulded part dominates its stiffness and warp.
If you must substitute, do it deliberately: note which properties you are approximating, add margin to the interfaces that depend on them, and re-validate in the real material before committing to tooling.
Finishing Changes the Answer Too
As-sintered PA12 is grey and matte. It takes dye well, can be bead blasted to a smoother feel, and can be sealed or painted for a presentable finish, but it will never look like a moulded part. ABS smooths beautifully with a vapour treatment and paints easily. PC can be polished to genuine clarity or vapour-polished, though this can relax fine detail.
If the prototype's job is to look like the product, budget for finishing. It is frequently a larger line item than the part itself.
Cost and Timeline, Honestly
Lead times vary with shop load, but the pattern is stable: sintered parts turn around in a handful of days with no tooling at all; machined prototypes take longer because every part occupies a machine; a prototype mould takes weeks but then produces parts at production-like cost and speed.
The crossover is the useful number to find. Below roughly twenty to fifty parts, tooling rarely pays. Above that, per-part cost on a soft tool drops dramatically — and you get the production material, which is usually the real reason to go there.
A Decision Rule That Works
Pick the cheapest process that answers the question you are actually asking, in a material whose relevant properties match the production intent. If you cannot name the property that matters, you have not finished defining the test.
That is the whole method. It avoids both failure modes — over-engineering a prototype into a small production run, and under-specifying one that gives a confident wrong answer.
Whether the answer is a machined part, a sintered one, or a short-run moulding, we run all three: CNC machining for tight-tolerance prototypes, 3D printing for fast iteration, and plastic injection moulding when you need production material and real per-part economics.
