Fireproof Plastic Materials for 3D Printing

• 1
  Key takeaways
• 2
  How a flame-retardant polymer actually stops a flame
• 3
  What UL 94 and FAR 25.853 each measure
• 4
  Where printed flame-retardant parts hold up, and where they do not
• 5
  When to machine the part instead of printing it
• 6
  Six steps to qualify a flame-retardant material for your part
• 7
  Printed flame-retardant parts against machined flame-retardant stock

A flame needs three things at once: fuel, oxygen and heat. Flame retardants attack the fuel and the heat, not the oxygen. The most common additive family for 3D printing is halogen-free phosphorus or nitrogen chemistry, usually blended at 15 to 30 percent by weight. When the polymer reaches roughly 300 to 350 °C, the additive decomposes first and forms a char layer on the surface. That char is a thermal barrier. It slows the rate at which fresh polymer is pyrolysed into flammable gas.

The second mechanism is endothermic. Some additives release water or carbon dioxide as they break down, which cools the reaction zone. Others work in the gas phase by scavenging the free radicals that carry the flame. You will see the difference on a burn test: gas-phase systems self-extinguish faster but produce more smoke, while condensed-phase char formers leave a heavier residue and tend to drip less.

Why does this matter to a printed part? Fused deposition modeling builds a surface from overlapping beads. Between those beads are micro-voids and a thin skin of partially degraded polymer. That skin has a slightly different composition than the bulk, because the nozzle has already taken it above the melting point once. The flame reaches this altered surface first. A resin printed on a laser platform has a different issue: the photoinitiator residue and the unreacted monomer on the surface can ignite earlier than the cured bulk.

So the polymer formulation tells you half the story. The other half is how the part was built. This is why a V-0 pellet does not automatically give you a V-0 printed component.

• 1
  Char layerForms at 300–350 °C and blocks heat transfer to the bulk.
• 2
  Endothermic releaseWater or CO2 cools the reaction zone.
• 3
  Radical scavengingInterrupts the gas-phase chain reaction; more smoke.
• 4
  Surface skinReprocessed polymer at the bead boundary ignites first.

UL 94 is a small-scale screening test. A specimen is clamped vertically and a flame is applied for 10 seconds, then removed. The afterflame time is recorded. The flame is applied a second time. For V-0, the total afterflame for both applications must stay under 10 seconds per specimen, no specimen may burn for more than 30 seconds after either application, and nothing may drip that ignites a cotton pad below. V-1 relaxes the time to 30 seconds per specimen. V-2 allows the flaming drips.

Two details trip people up. First, UL 94 reports a thickness. A material rated V-0 at 1.5 mm is not rated at 0.8 mm unless the certificate says so. Second, the rating belongs to a specific color and a specific formulation. Changing the pigment or moving to a different supplier's grade invalidates it until you retest.

FAR 25.853 is the FAA rule for aircraft interiors, and it is not the same question. It references 14 CFR Part 25 Appendix F, which runs a 60-second vertical burn for interior panels and a 12-second test for smaller parts. The 60-second test also caps burn length at 152 mm and limits drip flame time to 3 seconds. For larger cabin surfaces, the rulebook adds heat release rate and smoke density tests under a radiant panel. A part can pass UL 94 V-0 and still fail the smoke density limit.

For electronics you will also meet IEC 60695-11-10, which is technically equivalent to the UL 94 vertical method, and glow-wire tests such as IEC 60695-2-11 for parts near current-carrying connections. Glow wire is often the more relevant test for a housing that sits next to a terminal block.

• 1
  UL 94 V-0Afterflame under 10 s total, no flaming drips.
• 2
  UL 94 V-2Flaming drips allowed; weaker grade.
• 3
  FAR 25.85360 s vertical burn, burn length under 152 mm.
• 4
  Smoke and heat releaseSeparate limits for larger cabin surfaces.

A printed part does well when the wall is thick enough that the bead boundaries are buried under solid material. In practice that means 2 mm and up for FDM with a 0.4 mm nozzle, and 1.5 mm and up for SLS. Enclosures for a control board, ducting that carries warm air, brackets inside a cabinet, and covers that only need to resist a brief ignition source all fall in this range. These parts are usually non-structural and can tolerate 1 to 3 percent porosity.

The trouble starts with thin walls. At 1 mm, an FDM wall is two or three beads wide, so a large share of the cross-section is boundary. The char layer that forms is discontinuous, and the flame finds a path along the layer lines. We have seen parts fail a V-0 coupon at 1.2 mm that passed at 2.5 mm from the same spool. If your design has a 1 mm rib, either thicken it or plan to machine it.

The second weak point is any feature that traps powder or resin. A hollow channel in an SLS part holds un-sintered powder. That powder has a huge surface area and ignites far more readily than the fused wall. Blind holes and internal lattice should be avoided unless you can prove the powder was removed. For SLA, the same applies to trapped liquid resin in internal cavities.

The third issue is heat deflection, which is often forgotten. Flame-retardant additives usually lower the heat deflection temperature by 10 to 25 °C compared with the unmodified polymer. A part that passes the burn test can still sag at 80 °C in service. Check the HDT on the datasheet alongside the flammability grade.

• 1
  Good fitEnclosures, ducts, covers, brackets with walls over 2 mm.
• 2
  Poor fitRibs under 1.2 mm, internal lattices, trapped powder or resin.
• 3
  Watch HDTAdditives can drop heat deflection by 10–25 °C.

Machining removes the boundary problem entirely. A flame-retardant PEEK or PC plate is consolidated under heat and pressure, so there are no bead voids and no uncured surface. The flammability grade applies to the bulk material at the thickness you specify. That makes the certification argument much simpler when a customer asks for the test report.

At GreatLight we machine flame-retardant plastics on the same 127 CNC machines used for metal work, including 16 simultaneous 5-axis centers. For a part like a switch housing or a sensor bracket, we hold ±0.005 mm on critical features and reach Ra 0.8–1.6 μm on sealing faces. Wall thickness can go down to 0.5 mm on a supported rib, which is not achievable with a printed bead.

The trade-off is shape freedom. A machined part needs tool access, so deep internal channels and organic lattice geometry are out. If the design is a box with ribs, a machined route is usually faster and cheaper in small lots, and it produces a part you can inspect with a CMM. If the design is a curved duct with internal flow guides, printing wins.

One practical note on mixed builds. It is common to print the complex shell and machine the small flame-critical insert that sits near the ignition source. That gives you the geometry you need and a documented material grade exactly where the standard applies.

• 1
  Choose machiningThin ribs, sealing faces, tight tolerances, small lots.
• 2
  Choose printingInternal channels, organic shapes, low load paths.
• 3
  Mixed buildsPrint the shell, machine the flame-critical insert.