PEEK Flange Bearing | Self-Lubricating Bushing

Lubrication is the main one: PEEK runs dry or water-lubricated, so the grease line, the schedule and the contamination risk all disappear, whereas a bronze bushing under boundary lubrication will scuff and eventually seize. PEEK is chemically inert across a far wider pH range and does not dezincify or pit in chloride service. It is electrically insulating, breaking stray-current paths. At 1.32 g/cm³ it is roughly 15% of bronze density. It will not gall or weld to a steel shaft.

On the other side of the ledger, bronze carries higher static load, conducts heat far better, and holds tighter clearance across temperature — PEEK’s 47 ppm/K expansion is the design constraint you must engineer around. On food-contact duty, traditional leaded bronzes (e.g. C93200) are restricted, though lead-free bronzes such as C89833 and RG5 are available and hold NSF/ANSI 61 & 372 certification — the PEEK advantage there is compliance combined with dry running, not compliance alone.

The flange earns its cost when there is genuine axial load or an axial datum requirement. It replaces a thrust washer plus retaining ring or machined shoulder, typically removing two parts and shortening assembly. It gives a repeatable axial reference at the bearing itself rather than at a downstream feature. It adds 20–40% heat-dissipation area, worth 5–15 °C of running temperature in dry duty. And the flange-to-seat interface forms a dust labyrinth.

Against that, material use rises 10–15%, an extra machining operation is required, and the flange root becomes a stress-concentration site that needs a proper radius. Where the load is purely radial and axial location already exists, a plain sleeve is the cheaper answer.

Flange OD is typically 1.4–1.8 × body OD — a Ø26 mm body pairs with a Ø36–47 mm flange. Thrust capability scales with the annular area π/4 × (D_flange² − D_bore²), so widening the flange buys capability far more efficiently than thickening it.

Flange thickness follows from a bending check on that annulus against the allowable stress (unfilled PEEK ~40 MPa, GF30 ~60 MPa at ambient, both derated for temperature), landing at 1.5–3 mm for light thrust, 3–5 mm medium, 5–8 mm heavy. Always specify a root radius R1.0–2.0 mm at the flange-to-body junction and a C0.3–0.5 mm outer chamfer. For thrust loads above roughly 60% of the derated allowable, or for any cyclic thrust, verify by FEA and keep peak root stress below 70% of allowable.

Life is set by the wear budget divided by the wear rate, and the wear rate is a function of PV, temperature and cleanliness. Indicative figures at the stated PV, ambient temperature and a clean HRC 50 shaft:

Flange axial wear typically runs at ⅕–⅓ of the radial rate; replace at 20% of flange thickness. Other end-of-life indicators: clearance beyond 2× the installed value, axial float > 0.5 mm, a step change in running temperature (> 20 K rise), onset of noise, or visible discolouration and embrittlement. Raise life most reliably by stepping up to the CF/PTFE blend, upgrading shaft hardness to HRC 55–60, filtering the environment, and lowering PV via a longer bearing (increase L, not just d).

• Unfilled: For impact and shock, frequent start-stop, best chemical breadth and the toughest flange — but the lowest PV and highest CTE.
• GF30: For elevated temperature under load. ~2× the stiffness, half the creep, CTE cut to 20–25 ppm/K; slightly higher friction and notch-sensitive.
• CF30: Where clearance must hold at high ΔT. CTE 15–20 ppm/K closely tracks steel, thermal conductivity ~4× higher (which directly raises the usable PV), wear resistance ~3× better; most brittle of the three, so avoid on impact-loaded flanges.
• CF/PTFE tribological blend: For all dry-running duty. Lowest friction (0.15–0.20 dry) and lowest wear rate by roughly an order of magnitude; mechanical strength is the trade-off.

In practice: dry running → blend; hot and loaded → GF30; tight-clearance hot → CF30; shock → unfilled.

Clean and deburr the housing bore and shaft; verify bore to H7 and the flange seat to ≤ 0.03 mm flatness. Press on the body face only — never through the flange — using a stepped pilot that centres in the bore, at ≤ 3 mm/s on a press, never by hammering. A light film of assembly lubricant on the OD reduces press force and scoring.

If interference exceeds 0.04 mm, warm the metal housing to 80–100 °C (do not exceed 120 °C) rather than chilling the bushing; cryogenic shrinking of PEEK is not recommended — at deep-cryogenic temperatures PEEK embrittles and the thin flange root can micro-crack. Confirm the flange is fully seated (gap < 0.05 mm, four-point variance ≤ 0.03 mm, contact ≥ 80 %). Then ream or bore to the finished size if that option was specified, and verify roundness ≤ 0.03 mm and concentricity ≤ 0.05 mm.

Bolt-fixed flanges: tighten diagonally in three stages to the specified torque (± 5%) with a thread-locking compound. Commission with a run-in at 30% speed and 50% load for 30–60 min, holding temperature rise below 50 K.

Yes — this is one of PEEK’s strongest positions. Water absorption is 0.10–0.15% at 24 h and 0.45% at saturation, with dimensional change under 0.3%, and PEEK resists hydrolysis in hot water and steam where most engineering thermoplastics degrade; after long-term immersion it retains over 95% of its properties.

Water acts as both lubricant and coolant, giving a PV limit of 0.5–0.8 MPa·m/s (reduce to ≤ 0.4 in silt-laden water). Hydrostatic pressure is not a limiting factor — PEEK is used routinely in deep-sea and downhole service, so ROV and subsea depth capability is set by the housing and clearance design, not by the polymer. Applications include water and wastewater pumps (≤ 80 °C), sludge scrapers and gates, chlorinated and ozonated systems, marine steering and winch linkages, and ROV joints. Specify GF30 for long-term seawater immersion, de-scale and de-foul on a 6-month interval, and for UPW service request the low-extractables grade with verified ionic (< 1 ppb) and particulate (< 0.1 ppb) release.

Four checks, in order:

1. PV: compute P = F_radial / (d × L) and V = π·d·n / 60,000, multiply, then compare against the grade’s PV limit after applying the temperature derating factor.
2. Thrust stress: F_axial / [π/4 × (D_flange² − D_bore²)] against the derated static limit.
3. Bore closure: press-fit closure (0.7–0.9 × interference) plus thermal closure (OD × Δα × ΔT) must leave the required running clearance at maximum temperature.
4. Retention: residual interference ≥ 0.01 mm at maximum temperature, accounting for the housing material’s own expansion.

If any check is marginal, the first lever is increasing L (lowers P and PV simultaneously), the second is grade selection, and the third is adding lubrication. Send the operating data and Allstar’s engineering team will run all four checks and return the calculation sheet.