No. A labyrinth is a non-contact throttling device and cannot achieve zero leakage by definition. It controls leakage to a designed allowable value through clearance, tooth count and profile optimisation. Because PEEK tolerates rubbing, it can be built to a tighter clearance than a metallic seal of the same duty, which is the source of the leakage reduction. Where a true zero-leakage boundary is required, combine the labyrinth with a mechanical seal, dry-gas seal or buffer-gas system.

Two reasons, and neither is thermal expansion. First, PEEK’s tensile modulus (3.6–4.0 GPa) is roughly two orders of magnitude below steel, so contact loads are absorbed elastically instead of being transmitted into the rotor. Second, PEEK abrades sacrificially against a hardened shaft without galling, scoring or sparking, so a transient rub is a benign event rather than a machine-damaging one. This allows the cold assembly clearance to be reduced. Note that PEEK’s CTE is higher than steel’s, so the hot running clearance must be explicitly compensated — through CF30 grade selection or a metallic carrier.

PEEK-1000 (unfilled) offers the highest toughness and the broadest chemical resistance, and is preferred where impact and notch sensitivity matter. PEEK-GF30 raises stiffness and creep resistance and is electrically insulating, suiting loaded sections at elevated temperature. PEEK-CF30 provides the highest stiffness, the lowest friction, the best wear resistance, and — critically for labyrinths — a CTE of 15–20 ppm/K that closely tracks steel, giving the most stable running clearance across the temperature range. All three share the same 260 °C continuous ceiling.

260 °C continuous, with short-term excursions to 300 °C. This ceiling is set by thermo-oxidative degradation of the polymer backbone and is not increased by glass or carbon fibre reinforcement — reinforcement raises the heat-deflection temperature and load-bearing capability, not the chemical service limit. Above 300 °C, PEEK is not a candidate; specify metallic, abradable-metal or ceramic labyrinth seals.

Start from a diameter-proportional baseline (see the parameter table), then add the calculated differential thermal growth between the PEEK ring and the shaft over the full operating envelope, the manufacturing and installation tolerance stack, and the expected rotor vibration amplitude plus transient excursion. Finally, verify that the resulting leakage meets the specification via a CFD or bulk-flow calculation. Speed alone is not a sufficient input — an 0.10 mm clearance is routine at Ø50 mm and unachievable at Ø500 mm.

Ensure the bore and seal are clean and burr-free; use a pilot fixture to hold concentricity within 0.05 mm TIR; protect the tooth tips throughout handling; hold the axial location strictly to drawing; verify running clearance by feeler gauge or dial indicator before closing the machine; and run a no-load trial while monitoring vibration and acoustic signature.

At each shutdown, inspect the tooth tips for crazing, chipping and wear; measure the radial clearance and replace once it exceeds 150 % of the design value; trend the system leakage rate and investigate any step change; remove deposits from the throttling chambers, since fouling collapses the expansion volumes that produce the pressure drop; and log running hours against temperature and pressure to calibrate the replacement interval.

Under moderate duty — seal temperature ≤ 200 °C, surface speed ≤ 60 m/s, vibration within API limits — 20,000 to 30,000 operating hours (roughly 3–4 years of continuous service) is typical. Severe duty involving high temperature, high surface speed, frequent start-stop cycling or particulate-laden media may reduce this to 8,000–15,000 hours. Condition-based inspection is more reliable than a fixed calendar interval.