ULTEM is a trade name for polyetherimide, and the 9085 grade is a blend of polyetherimide with polycarbonate. It is not on the shelf because it is the strongest polymer available or the most heat-resistant — PEEK beats it on both. It is on the shelf because of a certificate.
Flame, smoke and toxicity, which is the whole commercial story
Anything installed inside an aircraft cabin must pass flammability, smoke-density and toxic-gas-emission testing. Those requirements eliminate almost every thermoplastic outright, and 9085 was formulated specifically to clear them while remaining printable. The same reasoning applies to rail interiors under European standards. That is why a material this expensive has a supply chain at all: air-duct sections, bracketry, cabin trim and tooling for aircraft interiors are made from it in real production quantities.
If the flame and smoke certification is not part of your requirement, the case for 9085 largely disappears. Heat deflection of 150 to 175 °C is genuinely excellent. The tensile figure is not: 69 to 76 MPa printed flat, measured by ASTM D638 on Stratasys's own equipment, which puts it below several polyamides that cost a tenth as much.
Two things about that number are worth stating, because published figures for this material are quoted carelessly. The first is that around 100 MPa belongs to neat polyetherimide — ULTEM 1000 — and not to the 9085 blend, which trades some strength for the flame and smoke behaviour it exists to deliver. The second is orientation: printed on edge, so that the load crosses the layers rather than running along them, the same material tests nearer 42 MPa. NASA measured that collapse alongside 5 to 8% internal porosity in printed coupons, and put printed strength at a little under nine-tenths of the injection-moulded equivalent. Nothing about spending several hundred dollars a kilogram removes the need to orient the part properly.
The build-plate connection
Polyetherimide is the polymer of the beige-gold PEI sheets that sit on the top of most modern build plates. When people say they print onto "PEI", the material under the print is chemically a close relative of ULTEM.
This is more than a coincidence worth mentioning. It explains something practical: PEI sheets tolerate repeated heating to 110 °C without deforming, resist the solvents used to clean them, and do not react with the plastics printed onto them — all properties of polyetherimide as a class. It also explains why an ULTEM part is a poor candidate for adhesive bonding: the same chemical inertness that makes a build plate a build plate resists glue.
Amorphous, which makes it easier than its temperatures suggest
Unlike PEEK, 9085 does not crystallise. There is no ordered phase to form, so cooling rate does not decide what material you end up with, and there is no colour-change tell to read. Dimensions depend on thermal contraction alone, and that is far more predictable.
Shrinkage sits at 0.5% to 0.9%, comparable to polycarbonate, but with a much higher glass transition the stresses involved are severe. An actively heated chamber above 100 °C is the stated requirement, and without one large parts simply delaminate. Nozzle 350 to 390 °C, bed 140 to 160 °C.
Drying is a high-temperature operation: 120 °C for 8 to 24 hours, which no food dehydrator reaches. Polyetherimide absorbs enough moisture from ambient air to blister and lose strength, and there is no shortcut.
What a bracket costs
A bracket, 150 g, nine hours on a heated-chamber machine:
| Term | Basis | Cost |
|---|---|---|
| Filament | 150 g at $350/kg | $52.50 |
| Machine time | 9 h at $2.00/h | $18.00 |
| Failure allowance | 15% | $12.69 |
| Electricity | 8.1 kWh at $0.175/kWh, chamber drawing 900 W | $1.42 |
| Total | $84.61 |
The instructive comparison is with the material itself: the plastic is nearly two thirds of the total, which is unusual. On almost every other page on this site the filament is the smallest term of the four. Adjust it in the print cost calculator and note how little difference the electricity makes even at 900 W.
There is a further cost that a calculator cannot model. Several of the industrial platforms that run 9085 accept only authenticated material cartridges, so the price per kilogram is set by the machine vendor rather than by a competitive filament market, and buying elsewhere is not an option the printer will accept. Anyone budgeting a project around this material should establish that constraint before the spool price, because it usually dominates it.
What ULTEM 9085 is genuinely bad at
- Every printer this site covers. A 390 °C hotend, a 160 °C bed and a heated chamber together describe an industrial machine, not an enclosed desktop one.
- Cost per gram, at $250 to $500 per kilogram.
- Being bonded or painted, per the inertness above.
- Impact toughness, which is respectable rather than remarkable — it is a stiff, somewhat notch-sensitive material and it does not absorb blows the way a polyamide does.
- Prototyping. Iterating a design in a material this expensive on a machine this slow is a poor use of both.
- Being confused with ULTEM 1010. The 1010 grade is stronger, has a higher continuous-use temperature and carries biocompatibility and food-contact certifications that 9085 does not. They are not interchangeable and the number matters.
Where it is genuinely used
Aircraft cabin ducting, brackets and interior fittings, where the certification is the requirement. Rail interior components under the equivalent European fire standards. High-temperature tooling, jigs and soluble-core mandrels for composite layup. Electrical insulators and connector bodies, since polyetherimide holds its dielectric properties hot. Sterilisable medical trays and fixtures.
For a workshop rather than a certification-driven industry, the honest recommendation is elsewhere: if the need is heat, PPA reaches similar territory for a fifth of the price, and if the need is stiffness with heat, a fibre-filled polyamide such as PA6-CF prints on machines that already exist in workshops.