The best engineering plastic for an unlubricated gear is not necessarily the strongest resin or the grade with the lowest published friction coefficient. It is the material that keeps tooth wear, deformation, and temperature within acceptable limits throughout the actual duty cycle. For moderate loads and controlled temperatures, POM is a useful starting candidate; PA can offer toughness but requires careful moisture assessment; PEEK becomes relevant when thermal or chemical conditions rule out lower-cost alternatives.

None of these choices is valid independently of the mating gear, tooth geometry, and operating environment. A dry-running gear pair is a tribological system: changing the counterface, surface finish, or material formulation can change wear behavior enough to reverse a preliminary material ranking.

Define what “without lubrication” means in service

Dry operation should mean that the gear pair meets its requirements without externally supplied oil or grease. A plastic containing internal lubricating additives may still satisfy that requirement, but those additives must be acceptable for the application. Restrictions on contamination, outgassing, food contact, or electrical performance can eliminate an otherwise promising formulation.

Also distinguish between a gear that runs continuously and one that moves briefly, then holds position. Continuous rotation creates repeated tooth loading and sustained frictional heating. An indexing mechanism may generate less average heat but impose high acceleration loads. A stationary loaded tooth faces creep rather than sliding wear. These applications can require different materials even when their nominal torque is identical.

The operating specification should identify:

• Continuous, peak, starting, and stall torque, including how long each load persists.
• Rotational speed, duty cycle, direction reversals, and required operating life.
• Ambient temperature, enclosure ventilation, humidity, and exposure to cleaning agents or process chemicals.
• Mating material, surface condition, gear geometry, mounting stiffness, and available backlash.
• Acceptable wear, transmission error, noise, debris, and loss of positioning accuracy.

“Maintenance-free” is not a measurable acceptance criterion. A gear can continue transmitting torque while wear has already made it unsuitable for accurate positioning or quiet operation.

Separate structural capacity from dry-running capacity

A gear tooth must resist root bending fatigue, contact damage, and excessive deformation. Dry operation adds a coupled wear-and-temperature problem. Friction generates heat at the mesh; increasing temperature reduces polymer stiffness and can accelerate creep and wear. Changes in tooth shape then redistribute loading, potentially concentrating it on fewer teeth.

Start with tooth forces and geometry rather than resin tensile strength. For a given torque, a smaller pitch diameter produces a higher tangential force. Module, face width, root geometry, contact ratio, and alignment affect how that force is carried. Room-temperature tensile strength does not provide a permissible gear-root fatigue stress, especially under elevated temperature or repeated reversing loads.

Polymer-specific gear rating methods, such as the relevant guidance in the VDI 2736 series, provide a more appropriate framework than transferring steel-gear allowable stresses to plastics. Their usefulness still depends on having suitable material data and respecting the method’s scope. A calculation using generic resin properties should be treated as screening, not proof of service life.

Published pressure–velocity limits deserve similar caution. Values obtained from bearings or simple sliding specimens cannot be applied directly to gear teeth. Tooth contact combines rolling and sliding, with sliding speed changing across the contact path. Gear geometry, thermal dissipation, and intermittent contact differ from the conditions used to generate many supplier tribology figures.

The relevant thermal question is not simply whether the housing stays below the resin’s maximum service temperature. Tooth bulk temperature and local contact heating must remain compatible with fatigue, stiffness, and wear requirements. Heat-deflection temperature is a laboratory comparison property, not an allowable continuous operating temperature for a loaded gear.

Use material families to narrow the shortlist

POM is a reasonable baseline, not an automatic approval. Its relatively low moisture sensitivity simplifies dimensional control compared with many polyamides. It is worth evaluating for compact mechanisms where stable backlash and dry sliding behavior matter. However, favorable friction does not compensate for insufficient root strength or creep resistance at the actual tooth temperature. Homopolymer, copolymer, and modified grades require separate assessment.

PA selection must include its conditioning state. Moisture absorption can increase dimensions while reducing stiffness and changing toughness. A gear evaluated immediately after molding may behave differently after equilibrating in service. That affects mesh clearance, tooth deflection, and positioning accuracy. Testing only dry-as-molded parts can therefore give a misleading result for humid operation.

PEEK is justified by a demanding operating envelope, not by its reputation alone. It becomes a serious candidate when elevated temperature or chemical exposure makes POM or PA unsuitable. An unfilled PEEK grade should not be assumed to outperform a wear-modified lower-temperature polymer in every dry mesh. The thermal advantage must translate into a verified system-level benefit.

A useful shortlist therefore names exact grades rather than just polymer families. “PA gear” or “PEEK gear” leaves unanswered the questions that determine performance: molecular formulation, reinforcement, lubricating additives, processing condition, and available fatigue or wear evidence.

Fillers can solve one failure mode while creating another

Glass or carbon fibers can increase stiffness, reduce creep, and change thermal expansion. Those benefits may help preserve tooth geometry under load. They do not automatically improve dry-running life. Fibers exposed at the working surface can increase counterface abrasion, and their orientation creates direction-dependent properties.

This matters particularly in injection-molded gears. Flow paths, gate location, and weld lines influence fiber orientation and local strength. A tensile specimen molded under different conditions may not represent the tooth root. Reinforced materials can also shrink unevenly, complicating tooth accuracy and concentricity.

PTFE, graphite, and other tribological additives may reduce friction or wear by modifying the interface and supporting transfer-film formation. Their effectiveness depends on contact conditions and the mating surface. A formulation optimized for sliding against steel may not perform similarly against another polymer.

The decision should follow the limiting failure mode. If tooth deformation controls performance, reinforcement may be valuable. If counterface damage or debris is unacceptable, an unreinforced or differently modified grade may be preferable. If frictional heating limits operation, a tribological formulation deserves evaluation—but its structural properties still need checking.

Choose the mating surface and clearances with the resin

A plastic–metal pair and a plastic–plastic pair present different thermal and wear conditions. A metal gear can provide a heat-conduction path, although its usefulness depends on material, shaft connection, and the surrounding assembly. Two polymer gears generally provide a less effective heat-removal path, making sustained dry operation more sensitive to duty cycle and enclosure conditions.

For a metal counterface, control roughness, burrs, machining marks, and any surface treatment. A rough flank can abrade the polymer, but an extremely smooth finish is not automatically optimal for every formulation. Transfer-film behavior and wear must be established for the actual surface. Hardness alone does not describe counterface suitability.

For plastic–plastic contact, assess the specific material pairing rather than assuming that two individually wear-resistant plastics will work well together. Adhesion, debris behavior, and heat accumulation can differ from results obtained against steel.

Backlash must accommodate manufacturing variation, thermal expansion, moisture-related growth, shaft movement, and housing expansion. Insufficient clearance can turn dimensional change into binding and rapid heating. Excessive clearance can increase impact during reversals and undermine positional accuracy. Apply the tolerance assessment to the assembled drivetrain, not only to the gear drawing.

Increasing face width can lower nominal loading, but the benefit depends on alignment and load distribution. A wider gear that contacts mainly at one edge may gain little useful capacity. Similarly, tooth-profile modifications intended to compensate for elastic deflection should be evaluated with the chosen polymer’s temperature-dependent stiffness.

Validate the production gear, not just the material sample

Supplier friction and wear data are valuable for eliminating unsuitable candidates. They become substantially more useful when the supplier identifies the test configuration, load, speed, temperature, counterface, roughness, and conditioning state. Without that context, a numerical comparison between grades may compare different experiments rather than different materials.

Bench tests should use representative geometry, production processes, and mating surfaces. Machined prototypes can support early screening, but they do not reproduce all effects of injection molding, including shrinkage, weld lines, residual stress, and fiber orientation. Conversely, prototype molding that differs materially from production tooling may also fail to predict final performance.

A meaningful validation program combines representative duty cycles with the conditions most likely to expose the governing failure mechanism:

• Continuous operation at the most demanding sustained load and thermal condition.
• Starts, stops, reversals, and specified short-duration overloads.
• Loaded dwell periods where tooth creep could affect accuracy or restart behavior.
• Relevant humidity conditioning and chemical exposure.
• Assembly tolerance extremes that challenge clearance and alignment.

Monitor temperature, backlash growth, tooth-profile change, drive torque, noise, and debris rather than recording only whether the gear breaks. Separate initial running-in from subsequent wear: an early surface adjustment may stabilize, whereas accelerating wear can indicate a developing thermal or geometric problem.

Accelerated testing requires particular care. Increasing speed or load may change the failure mechanism from gradual wear to thermal softening, making a simple life extrapolation invalid. Acceleration is useful only when there is evidence that the test preserves the mechanism relevant to service.

Make the selection against an explicit failure limit

A defensible choice links the exact material grade and production process to a defined operating envelope and measurable end-of-life criteria. The selected gear should retain adequate fatigue capacity, acceptable dimensional accuracy, and stable wear behavior—not merely survive a short dry-running demonstration.

If every candidate overheats, binds, or loses tooth accuracy, the constraint may be the drivetrain rather than the resin. A larger pitch diameter, revised module, improved alignment, better heat removal, or reduced duty cycle may deliver more benefit than moving to a more expensive polymer. If those changes are unavailable and testing still fails, lubrication-free operation is not established.

The final specification should preserve the details that made validation successful: resin grade, additive package, conditioning requirements, production method, mating surface, clearances, and load limits. Replacing any of these with a nominally equivalent option can change dry-running behavior. The reliable choice is the validated gear pair within its proven operating envelope, not the polymer name in isolation.