A drone company approached us in week 7 of a ten-week contract with an MJF PA12 bracket that had just failed after 72 hours of salt spray testing. The datasheet said PA12 was chemical resistant. What the datasheet didn't say was: the part was printed in the Z-direction, the bolt holes were subjected to 180 N of preload directly across the layer lines, and the unsealed surface had absorbed 1.4% moisture before the load cycling even began. A USD 180 reprint in glass-filled PA12, printed in the X/Y direction, with an additional two hours of cyanoacrylate infiltration, survived 480 hours. Same geometry, same machine—a change in material decision, and the trial production shipped on time.
Material selection for 3D printing isn't just about flipping through datasheets. It's a decision that quietly determines "which processes you can use, how much the part will cost, how it will fail, and how much post-processing you'll have to undertake." Get it right early, and the entire project stays cheap; get it wrong early, and the cost reveals itself in salt spray tests.
Why Additive Material Selection is Different from Other Processes
In CNC machining or injection molding, you choose the material first, then find a factory. In additive manufacturing, material and process are inextricably linked: each printer has its own material list, and the polymer you want often directly determines which machine you need to rent. The "nylon" printed by SLS is not the same molecule as the nylon an injection molding expert refers to; for the same resin, different post-curing formulations can result in a 30% difference in modulus. This coupling is the source of most material selection errors.
The Four Material Families You Are Actually Choosing From
Every additive material you specify falls into one of four families, and each family is tied to a set of specific processes. Understanding the family first—rather than immediately chasing a specific grade—allows you to continue the conversation when a supplier tells you "PA12 is out of stock for six weeks, switch to PA11."
| Family | Common Processes | Strength vs. Traditional Materials | Best Use |
|---|---|---|---|
| UV Curable Resins | SLA, DLS, PolyJet | 60–80% of cast acrylic | |
| Nylon Powders | SLS, MJF | 90–110% of injection-molded PA | Functional enclosures, ducts, fasteners |
| Extruded Thermoplastics | FDM, FFF | 70–95% of injection-molded ABS/PC | End-use brackets, Ultem tooling |
| Metal Powders | LPBF, DMLS, Binder Jetting | 95–105% of forged parts after HIP | Aerospace, implants, heat exchangers |
| Composite Filaments | FDM with carbon fiber/glass fiber | 1.5–3 times the stiffness of base polymer | Structural jigs, drone arms |
Photopolymer Resins: When Surface Finish and Detail Are the Mission Itself
Photopolymer resins, cured layer by layer under UV light, can produce the smoothest surfaces and finest features possible with additive manufacturing—0.3 mm feature walls and Ra below 3 µm on SLA. The trade-off is susceptibility to UV light, creep under long-term loads, and a brittle fracture mode that worsens with each month on the shelf. An SLA prototype that looks like injection-molded clear plastic on day one will yellow by week eight and crack by year one. This type of material is suitable for investor photos, dental models, wind tunnel housings, and inspection samples—not for field-deployed hinges.
Within the same family, choose based on mechanical targets, not brand names. Tough/Durable resins have a tensile strength of approximately 45 MPa; high-temp variants can withstand 200 °C for short periods; elastic resins can print gaskets with 60A–80A hardness; dental and medical grades support IPL sterilization. Carbon DLS polyurethanes (EPX 82, RPU 70) truly cross into the end-use part domain—provided the part is never exposed to UV light, then the premium is worth it.
Nylon Powders: The Default Option for Functional Parts
If you're unsure what to print but the part must be functional, start with MJF PA12. It can survive a 1.2 m drop test, can have threaded brass inserts heat-staked directly into it, achieves a near-injection-molded surface after vapor smoothing, and costs about USD 110/kg to print. PA11 offers approximately 40% more elongation at break than PA12 and costs about 20% more—active hinges and snap-fits that require hundreds of cycles are worth the upgrade. Glass-filled PA12-GF increases the flexural modulus to about 3,100 MPa, significantly reducing creep, at the cost of a rougher surface and slightly reduced toughness.
The most common mistake with nylon is ignoring moisture absorption. Dry PA12 will absorb 0.9–1.4% water within two weeks at 23 °C, 50% RH, causing a 15% drop in modulus and a 0.3% increase in size. If your tolerance stack-up is calculated based on dry properties, and the part is then placed in a hot and humid workshop, the problem has already shipped. Either seal the part, specify "equilibrium" test conditions, or switch to lower-moisture-absorbing PA-CF.
FDM Thermoplastics: The Only Process Truly Printing Engineering Plastics
FDM is the only additive manufacturing route that truly extrudes "engineering thermoplastic pellets"—identical to the polymer families found in injection molding catalogs. ABS/ASA general-purpose materials cost USD 40/kg; PC is available in optical grades with a HDT of 115 °C; nylon, with or without fiber, covers the tough functional range; high-performance stars—PEEK, PEKK, Ultem 9085, Ultem 1010—cost between USD 450–900/kg but can pass aircraft interior certifications (FAR 25.853) and implant applications. The trap is anisotropy: an FDM PEEK specimen stretched along the layers has about 50% lower strength than one stretched perpendicular to the layers.
Metal Powders: When Parts Must Withstand Production Environments
Metal additive manufacturing is in a different cost bracket—a small bracket printed with LPBF costs USD 400–1,500, while MJF costs only USD 40. Material choice directly determines whether this cost is worthwhile. Choose based on service conditions, not on the cheapest factory quote.
| Alloy | Best For | Beware Of | Typical Post-Processing |
|---|---|---|---|
| 17-4 PH Stainless Steel | General prototypes, brackets | Prone to hydrogen embrittlement during printing | Solution annealing + H900 aging |
| AlSi10Mg | Lightweight structures, heat sinks | Fatigue life inferior to forged parts | T6 heat treatment |
| Ti-6Al-4V Grade 23 | Implants, aerospace brackets | Oxygen absorption, high cost | HIP + stress relief |
| Inconel 625 / 718 | High-temp gas paths, exhaust | Residual stress cracking | HIP + aging |
| 316L Stainless Steel | Marine, food, medical | Relatively soft as-printed | Solution annealing |
| CuCrZr / Pure Copper | Heat exchangers, coils | High reflectivity – requires green laser | Age hardening |
Those Properties the Datasheet Doesn't Emphasize
Every additive datasheet presents tensile strength measured on optimally oriented, freshly printed, well-conditioned specimens. Your part is not that specimen. Before finalizing, run every candidate material through the same five checks: Z-axis strength (expect 40–70% of X/Y), modulus after moisture equilibrium, condition after post-cure or HIP, UV half-life for outdoor use, and creep at service temperature. Five numbers, five bench tests, two weeks—and you'll intercept failure modes that would otherwise derail the entire project's tooling budget.
The True Cost of Each Material for Each Iteration
Material cost is only part of the iteration cost; machine time, post-processing, and scrap rates impact the total price as much as raw materials. The following ranges are based on a 120 × 80 × 40 mm enclosure, with quotes from digital manufacturing platforms in early 2026.
| Material | Process | Cost per unit (USD) | Lead Time | Surface Roughness (Ra) |
|---|---|---|---|---|
| Standard Resin | SLA | 35–60 | 2–3 Days | 2–4 µm |
| High Toughness Resin | SLA / DLS | 55–95 | 3–4 Days | 2–4 µm |
| PA12 | MJF | 45–85 | 3–5 Days | 8–12 µm |
| PA12-GF | SLS | 80–140 | 4–6 Days | 10–14 µm |
| PC / Ultem 9085 | FDM | 90–260 | 4–7 Days | 15–25 µm |
| 17-4 PH | LPBF | 420–780 | 7–12 Days | 8–15 µm + fine finish |
| Ti-6Al-4V | LPBF | 950–1,600 | 10–15 Days | 10–20 µm + HIP |
Three Material Decisions in the Last Twelve Months
A Surgical Retractor That Must Survive 200 Autoclave Cycles
A reusable surgical tray retractor for spinal surgery must undergo steam sterilization at 134 °C / 2.1 bar each time it enters and leaves the operating room. The team's first version chose MJF PA12 because the same machine had just passed validation for a single-use instrument six months prior. Two test specimens entered the autoclave, and both delaminated at the 40th cycle.
The critical design move: change the family, not the grade. PA12 does not belong in the repeated autoclave cycle, no matter which supplier's. The team switched to LPBF 316L with electropolishing. The cost per unit rose from USD 110 for PA12 to USD 640, but it passed over 200 cycles on the first try and avoided the FDA surface biocompatibility issues that PA12 would have raised. The 5.8-fold price increase was still the right decision because it closed the risk ledger.
Savings were realized downstream: certification testing was reduced from six weeks to fourteen days, and the clinical study timeline recovered the month lost on the PA12 path.
A Drone Arm That Needs Stiffness Without Weight
A 7 kg industrial quadcopter arm needs to maintain a rotational deflection within 0.4° under 12 Nm of torque, without adding weight. FDM's carbon fiber-filled PA (PA-CF) arms weigh 62 g each, with a flexural modulus of 2,400 MPa; compared to machined 6061 aluminum at 108 g and PA12-GF at 145 g. The PA-CF route cost USD 95 per arm, with a 30-hour print queue, delivering 48 sets within six days—no molds, no milling machine queue.
A Transparent Flow Channel That Cannot Use Resin
A diagnostic startup needed a 30 × 50 mm, optically clear flow channel with a two-year shelf life. SLA clear resin was the obvious choice—Ra 2 µm, 90% light transmission—until UV accelerated aging yellowed the part, reducing transmission to only 71% at 18 months. After switching to 3-axis CNC machined PMMA, transmission remained stable at 92% throughout accelerated aging, costing USD 140 per unit, more expensive than SLA's USD 55. In this battle, additive manufacturing lost—acknowledging this early allowed the product to launch on time.
"Do's" and "Don'ts" When Locking Down Materials
| Do | Avoid |
|---|---|
| Test both Z-axis and X/Y direction specimens simultaneously | Believing the headline tensile strength numbers on the datasheet |
| Allow parts to equilibrate to service humidity before testing | Testing parts immediately after removal from a drying oven |
| Specify post-curing formulas on the drawing | Leaving curing times to operator judgment |
| Request quotes for three materials simultaneously in the first round | Locking in a resin before tensile testing |
| Specify the alloy condition, not just the alloy | Writing "17-4 PH" without specifying H900 or H1025 |
| Include post-processing in iteration costs | Only comparing raw material price per kg |
Common Material Selection Mistakes
| Mistake | Why it Fails | How to Avoid |
|---|---|---|
| Assuming SLS PA12 = Injection-molded PA12 | Weaker interlayer bonding, different moisture absorption behavior | Test under actual use orientation and humidity |
| Using brand marketing names to specify resin grades | Specifications can subtly change annually | Specify by mechanical properties (MPa, HDT) |
| Using photopolymer resins for outdoor parts | 1,000 hours of UV exposure can drop modulus by 40% | Switch to PA or painted FDM |
| Choosing Ti-6Al-4V because it "sounds right" | Overkills on cost; 17-4 can handle most brackets | Match alloy to actual load requirements |
| Skipping HIP for fatigue-critical parts | Internal porosity becomes crack initiation points | Perform HIP on all rotating or cyclically loaded metal parts |
| Buying PEEK without validating the printer | PEEK requires a chamber temperature of 380 °C for crystallization | Validate the machine, not just the pellets |
Checklist Before Material Selection
Before locking in a material family, go through the following checklist. If you can't answer any question, the next step is empirical testing, not specification.
- Service temperature range (continuous and peak, including duration)
- Chemical and humidity environment the part will be in
- UV, ozone, radiation exposure over expected lifespan
- Load type: static, cyclic, impact, or continuous creep
- Primary failure mode of most concern, and the test to intercept it
- Regulatory path: biocompatibility, flame retardancy, aerospace, food contact
- Appearance requirements: paintable, aesthetic-critical, or internal part only
- Production ramp-up: 10 units/month or 10,000 units/month in Q4
Design Focus
Material selection is the earliest leverage point that determines whether an additive project remains inexpensive or becomes costly. First, choose the family, then the grade; specify by mechanical targets, not brand names; respect anisotropy and moisture absorption behavior buried in datasheet footnotes; and certify with bench tests in the actual service environment. By doing these four things well, the subsequent processes, post-processing, and quoting will become clearer on their own.
0 comments