Publication of design trade data for a bearing housing material often begins with the tribological case, because a housing that functions as the bearing surface removes an entire bushing layer. Unfilled polyoxymethylene copolymer running against a ground steel shaft exhibits a dynamic coefficient of friction of 0.25–0.35 under unlubricated conditions when measured by ASTM D1894 or ISO 8295; unlubricated steel-on-steel contact can exceed 0.60 and is generally unacceptable without oil, grease, or a separate insert. This friction difference allows a molded POM housing to serve as both structural cage and bearing surface, eliminating the bushing inventory, press-fit operation, and secondary lubrication channel. On a production-scale injection molding line with 1,200 kN clamp force and a single core pin forming the shaft bore, supplier technical bulletins indicate that bore ovality can be held to 0.05–0.10 mm when the mold surface temperature is kept at 80–100 °C and the residual moisture of the granulate is below 0.10% by weight. Melt-temperature recommendations are 190–230 °C for copolymer grades and 200–215 °C for homopolymer grades; excursions above 240 °C can produce formaldehyde by thermal scission, creating bubbles, porosity, and an operator exposure hazard. The same single-shot processing route is not available for a machined steel or bronze housing, which requires multiple cutting stations and often a pressed-in bearing liner.

What Load and Speed Conditions Favour a Polymer Housing?

Bearing capacity in POM is not a simple stress limit. Manufacturer technical bulletins quote limiting pressure–velocity values of 0.08–0.15 MPa·m/s for unfilled copolymer against a hardened steel counterface with surface roughness Ra 0.2–0.4 µm, when evaluated using ASTM G99 pin-on-disc geometry. In practice, the housing must keep the projected bearing pressure below 3–5 MPa at surface velocities under 0.05 m/s if a high-cycle life beyond 10⁵ cycles is required; published data for this specific configuration is limited, and component-level testing is recommended. At shaft speeds above 0.5 m/s, frictional heating in the POM surface layer approaches the material’s continuous-use temperature range, and a lubricated metal or metal-hybrid housing is normally selected. The metal-bearing failure modes are different: surface fatigue, lubricant breakdown, and corrosive pitting, not the thermal–mechanical softening that limits POM. ISO 7148-1 and ASTM D3702 provide standardized methods for wear testing of bearing materials under specific geometries. The trade table therefore partitions the operating envelope by PV, ambient temperature, and lubrication regime rather than declaring a universal material winner.

Thermal expansion differences are often decisive in a housing designed to retain a rolling-element bearing outer ring. Unfilled POM exhibits a coefficient of linear thermal expansion of 100–120 × 10⁻⁶ K⁻¹ by ISO 11359-2, compared with 23 × 10⁻⁶ K⁻¹ for a cast aluminum housing, 12 × 10⁻⁶ K⁻¹ for structural steel, and 27 × 10⁻⁶ K⁻¹ for a die-cast zinc alloy. For a 40 mm bore diameter and a temperature rise from 23 °C to 60 °C, the POM bore increases by approximately 0.16–0.20 mm, whereas the equivalent steel bore changes by less than 0.03 mm. This movement must be absorbed by larger initial clearance or by designing the assembly so that the shaft and housing form a matched thermal pair. In precision spindle carriers and gearbox bearing seats, the mismatch usually excludes unfilled POM unless a steel outer sleeve or a low-expansion filler is introduced. POM also absorbs 0.2–0.4% water at saturation when immersed at 23 °C according to ISO 62; the dimensional contribution is small relative to polyamide, but in a continuously humid enclosure the moisture-induced change can add about 0.2% linear movement and must be summed with the thermal term when calculating the installed bearing clearance.

Mating a rolling-element bearing outer ring with a POM housing requires a fit strategy that differs from metal. In steel, a light press fit of 0.01–0.02 mm interference generates a high radial pressure because steel modulus is high; in POM the same interference develops a lower radial pressure, but the polymer relaxes. Supplier design guides recommend increasing the POM interference to 0.05–0.10 mm on a 25 mm bore only when the wall is thick enough to keep the hoop stress below the material’s creep-sensitive range, typically 25–35 MPa at 60 °C. If the wall is too thin, the outer ring can rotate inside the housing after temperature cycling. This failure mode—bearing ring rotation inside a plastic seat—is common in appliance gearcases when the POM housing is molded with a uniform wall of 4 mm or less and subjected to 40–60 °C internal air. The fix is either a windowed metal insert, a keying feature, or a change to a higher-modulus glass-filled POM grade. Thus the design trade table must include the local hoop stress and fit geometry, not only the material’s tensile properties.

When the housing cross-section must remain stiff at elevated temperature, the trade table shifts toward metal

Flexural modulus for unfilled POM is 2,500–3,000 MPa by ISO 178, while published tensile-modulus values for die-cast zinc alloys are roughly 85,000–96,000 MPa and structural steels range from 190,000–210,000 MPa. Matching the radial stiffness of a zinc housing may therefore require a POM wall thickness three to five times greater, which consumes packaging space and lengthens heat flow paths. Sustained radial load produces an additional creep loss. Under 10 MPa stress at 60 °C, POM copolymer can exceed 1% tensile creep strain after 1,000 h in long-term tests conducted according to ISO 899-1, while metals show negligible creep at equivalent stress. In bolted housing flanges, clamp load decreases as the polymer creeps. Torque-to-yield audit measurements on production assemblies have shown clamp load reductions of 30–50% after 500 h at 50 °C when metallic compression limiters are absent; the housing does not fracture, but bearing seating pressure decreases and the outer ring can begin to rotate. This failure mode is documented in supplier technical bulletins. Metal inserts and limiting washers are therefore specified in POM housings when preload loss must remain below 10% over the service interval.

Systematic comparison of design inputs for a light-to-moderate load bearing housing is summarized in the table below.

Chemical compatibility is the next filter in the trade table. Supplier resistance charts show POM copolymer withstands aliphatic hydrocarbons, automotive fuels, and many mild aqueous solutions at 23 °C. Strong acids at pH below 4, oxidizing agents, and concentrated formic acid attack the acetal group; field failures in bearing housings used in hot chlorinated alkaline washdown are documented as surface pitting after 200–500 h of immersion in screening tests conducted according to ISO 175. Zinc housings may corrode under similar alkaline cleaning unless chromate conversion coated, while steel requires oil film or plating. POM’s low saturated moisture uptake of 0.2–0.4% prevents the gross dimensional shifts typical of polyamide, but exposure to hot water above 60 °C can reduce tensile strength over time due to hydrolysis. Published data for this specific hot-water bearing housing configuration is limited; field validation is required before a stainless steel housing is replaced.

Does POM Maintain Dimensional Control After Injection Molding?

Unfilled POM copolymer exhibits volumetric shrinkage of 1.8–2.5%, and the linear shrinkage can vary by 0.1–0.3% depending on wall thickness, gate geometry, and mold temperature. ISO 294-4 defines the method for measuring injection molding shrinkage. A bearing bore formed by a core pin can become oval by 0.03–0.08 mm if gate freeze-off and packing are inconsistent. In production, a 1,200 kN clamp force machine using a 45 mm screw and a 25 mm diameter runner gate holds a bearing bore tolerance of ±0.05 mm only when hold pressure exceeds 80 MPa and cooling time is set to maintain a mold surface at 90 °C. Underpacking below 80 MPa hold pressure produces sink opposite the gate and excess clearance in the bearing seat. Tooling engineers therefore specify an initial POM bore tolerance two or three times wider than the equivalent machined metal seat, or add a post-machining step. A machined steel housing can routinely hold ISO 286-1 IT6 or IT7 bore tolerances, which remains the decisive reason metal is retained in high-precision bearing applications.

In high-speed bearing housings, the heat-generation term q = μFv must be compared with the thermal conductivity of the housing material. Unfilled POM has a thermal conductivity of 0.30–0.37 W/m·K according to ISO 22007-2 or supplier data, roughly one-hundredth that of aluminum at 180 W/m·K and one-fiftieth that of steel at 45 W/m·K. A polymer housing slows heat rejection from the bearing ring, raising the local contact temperature. For a 20 mm shaft running at 1,000 rpm with a dry friction coefficient of 0.2, the surface heat flux may exceed 0.5 W/mm² under moderate load; this value must be validated by component testing because the actual contact area is not equal to the projected area. At this heat flux, an unlubricated POM surface can soften or melt in seconds, whereas an aluminum housing conducts heat quickly enough to protect the bearing. This thermal conduction gap is an underrecognized entry in the design trade table and explains why POM housings are generally limited to intermittent motion or low duty cycles, while continuous high-speed shafts remain in metal housings with oil or grease lubrication.

Low duty cycle, high bearing preload in a 25 mm hinge housing

In a low-cycle industrial hinge bearing housing with 25 mm bore and 2–3 kN radial load, POM can be used if the load is distributed over enough projected area to keep pressure below 5 MPa. The housing is molded with a 6 mm minimum wall and a steel compression limiter around each bolt; the limiters carry the clamp load so the polymer is not required to maintain bolt preload. With a shaft surface finish of Ra 0.4 µm, the assembly runs dry for intermittent motion at speeds under 0.05 m/s. In field testing, glass-filled or internally lubricated POM grades have been used to raise stiffness and PV, but published data for this specific configuration is limited. The material choice is validated by measuring bore growth after 1,000 cycles and measuring clamp load at 24 h; if bore growth exceeds 0.15 mm or clamp load loss exceeds 15%, the design must shift to a metal housing with a pressed bushing. This decision boundary—not the material label—determines the final selection on the design trade table.