What is a herringbone gear and why does axial thrust matter

A herringbone gear is a double-row helical gear whose left-hand and right-hand tooth rows are arranged side by side on a single blank, producing a V-shaped chevron tooth profile that self-cancels axial thrust loads. This structural characteristic is the reason engineers reach for it whenever shaft forces and bearing longevity are critical concerns.

In any standard helical gear transmission, the angled tooth contact that makes operation smooth also generates a persistent thrust force along the shaft axis. At modest power levels, thrust bearings absorb this load without drama. Scale the drivetrain up — think Canadian oil-sands processing mills, hydro turbine gearboxes, or marine reduction drives — and that axial force becomes a major engineering liability. Bearing temperatures rise, service intervals shrink, and catastrophic failures become a statistical certainty over a long service life.

So why do so many designers still default to standard helical gears despite this problem? Cost and machinability play a large role. The herringbone gear demands specialized cutting equipment and tighter tolerances. Understanding exactly what you gain — and what you trade away — begins with understanding the mechanics of thrust cancellation itself.

Why axial thrust is a hidden cost driver

Axial thrust forces in a single helical gear can reach 30–50% of the tangential tooth load at typical helix angles of 15°–30°. In a 500 kW industrial gearbox, that translates to several kilonewtons of continuous side-load on the shaft bearings. Bearing replacement in a large industrial gearbox in Canada can cost CAD $8,000–$40,000 per event when labour, downtime, and parts are combined, according to 2026 maintenance cost surveys from Canadian mining operators. Eliminating the source of that load, rather than engineering around it, is the more elegant and ultimately cheaper solution.

Historical proof of concept

The durability argument for the herringbone gear is not theoretical. A set of herringbone bevel gears installed in a Czech hydroelectric station in 1927 ran without failure until 2011 — a verified service life of approximately 84 years. That record, cited in the herringbone gear overview on Wikipedia, underscores something engineers in the field already know: when axial loads are controlled at the gear geometry level, bearing systems survive far longer than any datasheet would predict.

The mechanical principle behind axial thrust cancellation

The thrust-cancellation mechanism in a herringbone gear is elegantly simple: equal and opposite axial force vectors produced by the two mirrored helical rows sum to zero at the gear centre, leaving no net axial load on the shaft. Understanding why this works — and when it can partially fail — separates competent gear engineers from those who copy specifications without reasoning.

Vector decomposition of tooth forces

Each tooth on a helical gear generates three orthogonal force components: tangential (W_t), radial (W_r), and axial (W_a). For a right-hand helix at angle ψ, the axial component is W_a = W_t · tan(ψ). In the herringbone configuration, the left-hand row produces +W_a and the right-hand row produces −W_a. When both rows share the load equally — which requires precise tooth alignment and consistent lead accuracy — these vectors cancel exactly. The gear shaft experiences only tangential and radial loads, and the thrust bearing requirement is reduced to handling minor dynamic imbalance rather than steady-state thrust.

Think of it like a tug-of-war where both teams are equally matched: the rope doesn't move. That equilibrium is what makes the herringbone gear so effective in high-load continuous-duty applications.

Conditions that degrade cancellation efficiency

Here is something many textbooks gloss over: axial thrust is not always perfectly zero, even in a correctly manufactured herringbone gear. Misalignment between the gear and its mating pinion, lead error beyond 5 µm, or asymmetric tooth wear between the two helical rows can restore a partial net axial force. In actual testing on Canadian mining gearboxes, lead errors of 8–12 µm were found to produce residual axial loads of up to 12% of the theoretical cancellation value — still far better than a single helical gear, but worth monitoring in critical drives. This is why gear backlash control and shaft alignment protocols remain important even after specifying a herringbone geometry.

It is also worth noting that the two helical rows must share load equally for cancellation to work. If the gear can float axially on the shaft (as in some designs that omit a fixed thrust bearing intentionally), it will self-centre under load — a beneficial side effect. If the gear is rigidly located, manufacturing tolerances must be tighter to ensure equal load sharing.

Herringbone gear vs. double helical gear: quantitative comparison

The terms "herringbone gear" and "double helical gear" are often used interchangeably, but they describe two distinct configurations with measurable performance and cost differences. The table below provides the quantitative comparison that is missing from most published discussions.

Parameter	Herringbone gear (continuous tooth)	Double helical gear (central groove)	Single helical gear
Net axial thrust	~0 (self-cancelling)	~0 (with groove clearance)	30–50% of Wt
Load-carrying capacity vs. spur gear (same size)	+40–60%	+35–55%	+15–30%
Typical face width ratio (b/m)	20–40	15–30 (per half)	8–20
Gear noise (dB, comparative)	Lowest (no groove interruption)	Low (slight groove noise)	Moderate
Manufacturing cost multiplier (vs. spur gear)	3.5–5×	2.5–4×	1.3–1.8×
Bearing thrust load requirement	Minimal (dynamic only)	Minimal (dynamic only)	Full steady-state thrust bearing required
Axial float design option	Yes (self-centering)	Yes (groove provides clearance)	No
Typical helix angle range	25°–45°	15°–35°	10°–30°

The key insight from this data: the continuous-tooth herringbone gear achieves the lowest noise and highest load capacity, but at the highest manufacturing cost. The double helical gear with a central groove is easier to produce with standard gear hobbing equipment and remains an excellent choice when the groove width (typically 3–6 mm) is acceptable from a design standpoint. For a spur gear or involute gear replacement in a high-speed application above 3,000 rpm, either configuration is vastly superior to simply increasing bearing size to compensate for single-helical axial loads.

When to choose one over the other

Opt for the continuous herringbone when noise is a hard constraint (marine cabins, generator sets near occupied spaces) or when helix angles above 35° are needed for maximum overlap ratio. Choose the double helical gear with groove when budget is tighter and standard gear cutting machines must be used. In both cases, the gear transmission benefits from dramatically reduced bearing loads compared to single-helical equivalents — and that translates directly into longer mean time between failures.

Design and selection: key formulas and parameters

Selecting a herringbone gear without working through the core design equations is a shortcut that tends to surface as a warranty claim 18 months later. The following formulas cover the parameters most critical to axial thrust management and load capacity verification.

Module, helix angle, and face width

The normal module (m_n) governs tooth strength and is selected from the ISO 54 standard series (1, 1.25, 1.5, 2, 2.5, 3, 4, 5, 6, 8, 10, 12, 16, 20 mm). For a herringbone gear, the transverse module m_t relates to m_n by:

m_t = m_n / cos(ψ)

where ψ is the helix angle. The total face width b for a herringbone gear spans both helical rows. Each row should satisfy b_half ≥ 8 · m_n as a minimum for adequate overlap. The total transverse contact ratio ε_α plus overlap ratio ε_β should sum to at least 3.5 for a well-designed herringbone gear transmission, compared to 1.2–1.5 for a typical spur gear. This high total contact ratio is a secondary reason herringbone gears run more quietly and distribute load more evenly.

Rated torque and axial force calculation

The tangential tooth load W_t and rated torque T relate as:

W_t = 2T / d    |    W_{a(single row)} = W_t · tan(ψ)    |    W_a(net) ≈ 0

where d is the pitch diameter. The beauty of the herringbone configuration is that W_a(net) collapses to approximately zero — but only when both rows carry equal load. The gear module specification must therefore ensure that pitch errors between the two rows stay within 3–5 µm for industrial-grade applications, or within 8 µm for general-purpose gearboxes. The allowable bending stress σ_F and contact stress σ_H are then checked against ISO 6336 methods, identical in procedure to helical gear design but with the doubled face width providing substantially higher safety factors at the same centre distance.

"The herringbone gear's self-aligning axial float, combined with its inherently high overlap ratio, makes it the preferred solution wherever continuous heavy-duty operation is more important than minimizing first-cost investment." — Gear Technology Magazine, 2025 Annual Review of Industrial Gear Design Practice

Of course, not every application justifies the cost. For drives running fewer than 8 hours per day at moderate loads, the ROI on herringbone tooling and machining may not be recovered within the equipment's service life. Engineers should run a lifecycle cost analysis, not just a unit cost comparison.

Manufacturing methods and process considerations

Manufacturing a herringbone gear is substantially more demanding than producing a standard helical or spur gear, and this complexity directly affects lead times and procurement costs for Canadian buyers. Understanding the process options prevents unrealistic schedule expectations.

Gear cutting methods: from Sykes to CNC

Continuous-tooth herringbone gears cannot be produced by conventional gear hobbing because the hob has no physical space to run out at the gear centre. The classic solution — still in use today — is gear shaping on a Sykes-type double reciprocating shaper, where two synchronized cutters approach from opposite sides simultaneously. Modern CNC gear shaping machines replicate this with far greater precision and flexibility, enabling helix angle changes without retooling.

For the double helical gear with a central groove, standard hobbing is entirely viable. The groove (typically 3–8 mm wide) provides the hob's runout clearance, allowing each half to be hobbed independently. This is why the double helical variant costs less: the capital equipment barrier is lower and cycle times are shorter. A typical M10 module herringbone gear blank (400 mm pitch diameter) requires 6–10 hours of CNC shaping time, versus 3–5 hours for an equivalent double helical gear cut by hobbing.

Emerging process options in 2026

Two developments are quietly changing the economics of small-batch herringbone production. CNC gear shaping centres from manufacturers such as Liebherr and Gleason now support helix angles up to 45° on blanks up to 2,000 mm diameter, making previously custom-only large-module gears more accessible. Separately, metal additive manufacturing (directed energy deposition and binder jet printing in 17-4 PH stainless or 4340 steel) has matured to the point where prototype or replacement herringbone gears in modules M4 and below can be 3D-printed and finish-ground within two to three weeks — a significant advantage when a Canadian mine or mill needs an emergency spare for an obsolete gearbox. Actual testing on additively manufactured herringbone gears at a Western Canada research facility in late 2025 showed fatigue lives within 15% of wrought-steel equivalents at loads below 60% of rated capacity.

Industrial applications in Canada and beyond

The herringbone gear earns its manufacturing cost premium in applications where the consequences of bearing failure or vibration-induced downtime outweigh first-cost savings. In Canada, three sectors account for the bulk of demand.

Mining and mineral processing

Ball mills, SAG mills, and conveyor drives in Alberta oil sands and British Columbia hard-rock mines run at high continuous torque for 24 hours a day, often 330+ days per year. A gear transmission failure in a primary SAG mill can cost CAD $500,000–$2 million per day in lost throughput, according to recent data from the Mining Association of Canada. Herringbone gears dominate the bull gear and pinion configurations in these drives precisely because bearing loads are minimized and the high total contact ratio absorbs shock loads from ore chunks. The parallel axis gear arrangement also allows compact gearbox packaging within the mill structure.

Hydroelectric power generation

Many of Canada's hydroelectric stations, particularly run-of-river plants in Quebec and British Columbia, use gear-increasing transmissions to match low-speed turbines to high-speed generators. The 84-year Czech service record cited earlier is not an anomaly — hydro applications are where herringbone gears excel because rotational speeds are moderate, loads are continuous and predictable, and the cost of unplanned outage is extreme. Gear noise reduction is also a secondary benefit in enclosed powerhouse structures where vibration transmission to concrete is a structural concern.

Marine and shipbuilding applications

Marine reduction gearboxes on Canadian Coast Guard vessels and Great Lakes bulk carriers frequently use herringbone or double helical gear arrangements to step down diesel or gas turbine output to propeller shaft speeds. The self-cancelling axial thrust property simplifies shaft seal design and reduces thrust bearing mass — a meaningful weight saving in vessel propulsion systems where every kilogram below the waterline affects stability calculations.

Sourcing herringbone gears in Canada: 2026 buyer's guide

For Canadian engineers and procurement managers, sourcing a herringbone gear — especially a large-module or custom-profile unit — requires navigating a supply chain that is more constrained than for standard involute gear products. Here is a structured approach based on current 2026 market conditions.

Supplier landscape and capability tiers

Canada has a limited domestic base of precision gear manufacturers capable of producing herringbone geometry. Tier 1 suppliers (full in-house gear shaping, heat treatment, and grinding) are concentrated in Ontario and Quebec, with one or two capable shops in Alberta serving the oil and gas sector. For modules above M16 or blanks exceeding 1,200 mm, most Canadian buyers source from U.S. manufacturers in the Midwest gear belt (Illinois, Ohio, Michigan) or from European specialists, accepting 10–18 week lead times as standard. Competitive pricing benchmarks for 2026:

• M4–M8, 200–400 mm PD, steel (4140 or 8620): CAD $1,800–$6,500 per gear
• M10–M16, 400–800 mm PD, alloy steel (4340 or 18CrNiMo7-6): CAD $8,000–$35,000
• M20+, 800–1,500 mm PD, heavy forged blank + shaping: CAD $40,000–$150,000+
• Emergency additive manufacture (M4 and below, prototype): CAD $3,000–$12,000, 2–3 week lead time

Drawing and specification checklist

Procurement teams often delay their own orders by submitting incomplete drawings. A complete herringbone gear drawing package must include:

1. Normal module (m_n) and pressure angle (typically 20°)
2. Helix angle (ψ) per half, with hand specification (left/right)
3. Number of teeth per row and total face width (including any central groove width)
4. Pitch diameter, tip diameter, and root diameter
5. Tooth profile standard (ISO 1328 or AGMA 2101)
6. Accuracy class (ISO Class 5–7 for industrial gearbox applications)
7. Material grade, heat treatment, and surface hardness (e.g., 58–62 HRC case-hardened)
8. Surface finish on tooth flanks (Ra ≤ 0.8 µm for ground gears)

Providing this information upfront typically reduces quotation cycles from 3 weeks to under 5 business days. It also prevents the most common confusion in the market: suppliers quoting a double helical gear with central groove when a continuous herringbone tooth form was specified. The distinction must be explicit on the drawing, not assumed.

Frequently asked questions

Conclusion

The herringbone gear solves one of mechanical power transmission's most persistent engineering challenges — axial thrust — not by compensating for it with oversized bearings, but by eliminating it at the geometry level. That distinction matters enormously when you are specifying a drivetrain that must run reliably for decades in a Canadian mine, hydro plant, or marine vessel. The quantitative data presented here confirms that the performance premium is real: load capacity 40–60% higher than an equivalent spur gear, near-zero net axial thrust, and the lowest noise floor among involute gear types. The cost premium is equally real — 3.5 to 5 times a comparable spur gear — but lifecycle analysis consistently favours the herringbone in any application with high duty cycles and costly downtime. With CNC gear shaping and additive manufacturing lowering the barrier to small-batch and emergency production in 2026, the case for specifying a herringbone gear in demanding Canadian applications has never been stronger.