PEEK Machining: PEEK vs Acetal, Nylon and PTFE for Machined Parts

The parts measured perfectly at the bench on Friday. On Monday the bores were tight, the flat plate had bowed, and the assembly would not close. Nothing was machined wrong. The plastic simply absorbed moisture from the air and expanded, and the drawing had specified a tolerance that only makes sense for metal.

Machining engineering plastics is less about cutting forces than about what the material does after the cutter leaves. Moisture uptake, thermal expansion, creep under clamp pressure and residual stress from the extrusion process all move dimensions, and each polymer moves differently.

Here is how PEEK, acetal, nylon, PTFE and UHMW-PE compare, and how to specify a plastic part that still fits after a week in service.

The four movements that ruin plastic parts

Every tolerance problem in a machined plastic part comes from one of four sources.

Moisture absorption. Polymer chains accommodate water molecules between them, which pushes them apart. Nylon is the worst offender; PTFE and UHMW-PE are effectively immune.

Thermal expansion. Plastics expand five to ten times more than aluminum per degree. A part that is inspected at 30 °C in a warm shop and used at 10 °C in a cold room changes size measurably.

Creep. A clamped plastic part continues to deform under sustained load. Vise pressure at 9 am is not the same geometry at 5 pm.

Machining stress. Cutting induces surface stress, and when material is removed from one face of an extruded sheet, the balance tips and the part bows.

Moisture absorption, quantified

The nylon row explains most field failures. A 100 mm nylon part that goes from dry to saturated grows by roughly 1.5 to 2 percent in the thick direction, which is 1.5 to 2 mm. No tolerance can absorb that. Nylon works when the part is designed around it and conditioned before final machining, not when a metal drawing is reissued in nylon.

PEEK: expensive, and worth it when the temperature is real

PEEK holds its properties to 250 °C continuously, resists most chemicals including steam and fuels, and has the best strength-to-stiffness combination of any unfilled engineering plastic. It also costs ten to twenty times as much as acetal, so let temperature, chemistry or a regulatory requirement drive the choice.

Unfilled PEEK machines well with sharp carbide but is notch sensitive and chips at exit edges; back up through holes with a sacrificial piece.

Glass-filled PEEK is abrasive: it wears carbide quickly, so production runs justify diamond-coated or PCD tooling.

Carbon-filled PEEK is stiffer and dimensionally more stable — around 20 GPa modulus and much lower expansion — which makes it the choice for precision structural parts in aerospace and semiconductor equipment.

Acetal: the default for mechanical parts

POM-C machines more like brass than like plastic. Chips break cleanly, finishes come off the tool, and tolerances of ±0.02 mm are achievable on a stable machine. It is cheap, stable, and handles most dry mechanical applications up to 90 °C.

Its weaknesses are creep under sustained load and poor resistance to strong acids and chlorinated water. It also chips at sharp internal corners, so radius everything.

For a bracket, spacer, gear blank or fixture pad that sees shop temperature and moderate loads, acetal is almost always right.

Nylon, PTFE and UHMW-PE: three specialists

Nylon is tough. It absorbs impact without cracking, resists abrasion, and works well for wear pads, rollers and gears that see shock loads. Its dimensional instability makes it a poor choice for precision fits, and its strength drops when it absorbs moisture — a dry 80 MPa tensile can fall to 50 MPa conditioned.

PTFE has the lowest friction of any solid, a continuous service temperature of 260 °C, and near-zero moisture absorption. It also cold flows under load, wears poorly and cannot be glued easily. Use it for seals, gaskets and chemical handling, not structure.

UHMW-PE is the abrasion champion: chain guides, wear strips, chute liners. It creeps badly, machines with a fuzzy burr, and should not run above about 80 °C.

Clamping, tooling and stress relief

Plastic deflects roughly twenty times more than aluminum under the same load, so vise pressure that would be gentle on a metal part can bow a PEEK or acetal part by 0.1 mm and leave the error in the finished surface.

Use soft jaws machined to the part profile, support the work directly under the cutting zone, and clamp only as hard as the cut requires. For thin plates, vacuum fixtures or double-sided tape on a flat plate often beat a vise.

Cutting tools should be sharp, polished, high-rake carbide with two or three flutes and generous chip clearance. Dull tools rub, generate heat, and leave a melted smear rather than a cut surface. Air blast is usually better than flood coolant for plastics because it avoids absorption and staining, and it clears chips without soaking the part.

For tight tolerances, rough machine, anneal, then finish. PEEK benefits from a stress-relief cycle around 200 °C for a few hours before the finish pass. Acetal and nylon should be conditioned to the expected service humidity before final machining, and the drawing should say so.

Temperature matters at inspection too. Measuring a plastic part in a shop at 28 °C and expecting it to hold at 5 °C is optimistic; a 10 °C change on a 100 mm acetal part is 0.11 mm.

Selecting a material

If the application is borderline, send us the temperature, the chemical exposure and the tolerance stack. When we quote a plastic part at JSHS, we usually recommend a material and a machining sequence together, because the sequence matters as much as the polymer. You can request a quote with the drawing and service conditions attached.

Frequently asked questions

Why does my PEEK part change size after machining?

Two reasons: residual stress in the extruded stock releasing when material is removed, and moisture uptake. Annealing the blank before finishing, and roughing then finishing in separate operations, removes most of the first. The second is small for PEEK, around 0.1 to 0.2 percent.

Can PEEK replace metal in a structural part?

Sometimes, if the load is modest and the temperature would soften aluminum. At 3.6 GPa against 69 GPa, PEEK deflects about nineteen times more than aluminum unless the section grows.

What tolerance can be held on machined acetal?

±0.05 mm is routine, and ±0.02 mm is achievable on a rigid setup with temperature control. Below that, moisture and thermal movement dominate and the tolerance becomes a measurement exercise rather than a manufacturing one.

Is glass-filled PEEK harder to machine?

Yes. The glass fibres are abrasive and wear carbide quickly, so tool life drops. Diamond-coated or PCD tooling is normal for production, and edges need support.