This map lines up the five main families against efficiency, noise, load capacity, backlash, lubrication, and cost so a process engineer can short-list the right geometry in one pass — for a deeper primer see the industrial gear encyclopedia entry and the cost-stack detail in worm gear TCO analysis.
Spur Gears: Cheap, Efficient, Loud
Cut or shaved spur gears hold AGMA Q6–Q8 accuracy (≈0.05–0.10 mm total composite error) and carry surface contact stresses up to roughly 1.0–1.4 GPa for case-hardened steel at 10⁷ cycles.
The trade-off is straight-cut tooth engagement: impact loading on every mesh produces a 6–15 dB(A) noise penalty versus helical equivalents at the same pitch line velocity, and the lack of tooth overlap means a single tooth carries the full transmitted load at any instant. For low-speed (below ~8 m/s), low-noise-tolerant, budget-bound applications, spur gears remain the rational pick; for anything above that line, the noise and vibration penalty starts to dominate [S1].
Helical Gears: Smoother, Quieter, with Axial Thrust
Typical helix angles of 15°–25° balance smoothness against the axial thrust component, which the supporting bearings must absorb. Double-helical (Herringbone) variants cancel thrust entirely and are common in heavy-mill drives above 5 MW.
For high-speed (above 10 m/s), high-load, or noise-regulated service, helical is the default [S1]. Sliding contact along the helix also demands EP-grade or synthetic PAO lubricants; a 32-grade mineral oil that works for spur at 70 °C sump can scuff a helical at the same load.
Bevel Gears: Right-Angle Drives at a Cost
Bevel gears redirect motion between intersecting shafts (typically 90°), with straight-bevel, spiral-bevel, and hypoid variants each tuned to a different speed-noise band. Spiral-bevel units match helical for smooth running and carry AGMA Q9–Q12 precision (≈0.025–0.05 mm) for automotive and aerospace final drives.
Pitch-line sliding makes lubrication the limiting factor: EP additives, ISO VG 68–320 synthetics, and sump temperatures held below 90–100 °C are typical.
Worm Gears: High Ratio, Self-Locking, Heat-Limited
A single worm stage delivers reduction ratios of 5:1 to 100:1 in one mesh, with optional self-locking when the lead angle is below the steel-on-bronze friction angle. That makes worm gear pairs the standard pick for hoist, lift, and conveyor hold-back duty, where back-driving would be a safety hazard. [S1]
The trade-off is severe: mesh efficiency commonly lands at 30–90%, with each mesh generating 30–60% of input power as heat that the housing must dump. A 7.5 kW worm reducer at 40:1 can reject 3–4 kW of heat continuously; above that envelope, a forced-oil or fan-cooled housing is mandatory, and synthetic PAG or ester lubricants are typically specified to push sump temperatures down. The full cost stack — energy loss, lubricant change intervals, and seal wear — is laid out in the worm gear TCO breakdown. Bronze worm wheels wear progressively under shock load, so worm units are the wrong choice for cyclic high-inertia drives even though the ratio envelope is tempting [S1].
Planetary Gears: High Torque Density, Backlash-Sensitive
Planetary (epicyclic) stages split the load across three or more planets, raising torque density by a factor of roughly 3 versus a comparable shaft-mounted gear of the same envelope. Standard stages deliver 97% efficiency each, and a 3-stage planetary can hit ratios of 1,000:1+ in a compact package, which is why servomotor and wind-turbine pitch and yaw drives lean on this family.
Planetary geometry is sensitive to manufacturing tolerance: AGMA Q6+ quality and precision-ground planets are typical for backlash below 5 arc-min, and zero-backlash harmonic-drive variants are the only option when positioning repeatability is sub-arc-minute. The downside is complexity — 6+ bearings per stage, internal lubrication channels, and a housing that must keep the planet carrier concentric. Maintenance windows on industrial planetary units commonly run 5,000–10,000 hours for oil changes; skipping them can double planet-bearing wear rate [S1].
Selection Criteria: Matching Family to Duty
Match the family to four variables — input speed, ratio, duty cycle, and noise limit — before material or lubrication is even discussed. Below ~8 m/s pitch-line velocity and under 70 dB(A) noise targets, helical or spiral-bevel is the safe default. Above that line, helix angles and precision class need a written spec, and the housing thermal rating must be verified against continuous-duty losses.
For drives over 5 kW, an efficiency delta of 3–5 percentage points between two otherwise-equal stages translates to thousands of kWh per year of waste heat and measurable kWh-bill impact; this is the single most overlooked number on most gear-selector worksheets. Where safety demands non-back-driving, worm or dual-stage helical-with-brake are the only rational picks, and a forced-cooling or oil-pump heat exchanger is non-optional once continuous input power exceeds ~3 kW [S1]. A side-by-side comparison of the five families is summarised in the table that follows for quick reference.
Material, Lubrication, and Failure-Mode Constraints
Material choice follows the operating envelope. Case-carburised 16MnCr5 / 20MnCr5 through-hardening steels dominate below 1.0 GPa tooth-bending stress, while nitriding 31CrMoV9 and through-hardening 42CrMo4 cover the higher-temperature, higher-load range up to roughly 200 °C sump. For food-grade or low-noise service, POM (polyacetal) or PA66 plastic gears on metal shafts run quiet but cap out at 0.2–0.3 GPa tooth contact stress and below 80 °C continuous. [S1]
Lubrication is the second failure driver: EP gear oils (ISO VG 68–680) with anti-scuff additives are baseline for steel pairs, while synthetic PAO or PAG fluids extend drain intervals to 10,000–20,000 hours at higher bulk temperatures. Common failure modes — macropitting, micropitting, scuffing, tooth-bending fatigue, and wear — are all governed by the combined surface stress, slide-to-roll ratio, and oil temperature; under-rated lubrication can cut calculated L10 life by 50% or more before a single tooth shows visible damage [S1]. Sealed-for-life units are a clean answer for low-duty service but lock out the oil-condition-monitoring that a high-availability line should be running.
Trackable Signals for Future Spec Revisions
Three near-term signals are worth tracking: ISO 6336 tooth-strength calculation revisions, AGMA quality grade updates, and supplier-published synthetic-lubricant approvals for worm and hypoid service. The industrial gear encyclopedia page is updated as standards and vendor spec sheets evolve, and the linked worm-gear TCO reference covers the energy and lubricant cost stack in detail. [S1]
Component reference pages worth checking: industrial adhesive, and industrial borescope.