Most specifiers treat single helical versus double helical as a geometry pick — cheap and compact on one side, heavy and quiet on the other. The choice is not binary.
Once a properly sized thrust bearing is on the table, single helical with bearing support becomes a third option that API 613 and API 617 explicitly allow. The math that forces the decision — F_a / F_t ratio, torque density per face width, allowable pitch-line velocity, and cost per kW — rarely gets run before the spec lands on herringbone. Below the heavy-mining and rolling-mill threshold, that math often points the other way.
What Is a Single Helical Gear?
A single helical gear is a cylindrical gear with teeth cut at a single helix angle β, typically 15°–20° for industrial drives. Single helical dominates standard industrial reduction from 0.12 kW packaging drives up to 200 kW conveyor and mixer duty.
The geometry creates a problem the specifier must price: axial thrust. Every newton of tangential tooth force produces an axial component equal to F_t · tan(β). At β = 20°, that is 36 percent of the tangential load pushing along the shaft — thrust the bearings have to carry.
Per AGMA 2001-D04 rating practice, this is not optional padding; it is the force that sets bearing selection, housing rigidity, and thrust-washer specification. The decision between a 15° and a 20° helix angle is a real tradeoff between load smoothing and bearing cost. Specifiers who want the physics in detail can work through what causes axial thrust in helical gears before running their own F_a calculation.
What Is a Double Helical Gear? (And How Is It Different From Herringbone?)
A double helical gear has two helical tooth sets of opposite hand cut on the same wheel, so the two axial-thrust vectors cancel internally. This is how you get the torque-smoothing benefit of a 20°–45° helix angle without exporting thrust to the bearings. The advantages are documented — higher contact ratio, better load sharing, near-zero net axial force — and the disadvantages are real too: the parts are harder to manufacture, harder to inspect, and cost 1.5 to 2.5 times a single-helical equivalent at the same face width.
ISO 1328-2 lists double helical and herringbone as separate gear classification categories despite functional equivalence. DIN 3990 explicitly covers double helical for load-capacity rating.
Double Helical vs Herringbone: The Manufacturing Distinction
Double helical gears and herringbone gears are not the same part, though the terms are often used interchangeably. A true herringbone has a continuous V with no central gap. A double helical has a relief groove in the middle so the cutting tool has somewhere to exit — that groove is the signature of the manufacturing route.
True herringbone is hobbing-impossible because the cutter collides with the opposite helix, which is why the industry relied on the Sykes gear shaper (1910s, cylindrical guides with round cutters) until the 1980s, and on the Wuest central-groove patent for double helical. Both geometries deliver thrust cancellation; the difference is the tooling envelope.
Why the Cost Premium Holds
The premium is not paying for the V-shape. It is paying for a manufacturing constraint. Double helical halves cannot be assembled in the field from two separate single-helical gears — the phasing tolerance is tight enough that the gears must be machined as integrated units.
That constraint, plus the AGMA quality grade the geometry demands in heavy service (typically Q10–Q12 under ANSI/AGMA 2000-A88, equivalent to A2–A4 under the newer ISO 1328-1 numbering), fixes the cost gap. Specifiers who see “AGMA Class 10” in a spec should confirm which standard era is being cited; the same numeric grade means opposite things under the two systems.
A double helical gearbox quote reflects this constraint at the assembly level, not just the gear pair.
Choosing Single vs Double Helical: Axial Thrust, Torque Density, Speed, Cost
Spec selection turns on axial thrust, torque density per face width, allowable pitch-line velocity, and cost per kW. Each runs against the application’s duty class to give a defensible answer.
Axial Thrust Ratio
The torque calculation must include the axial component. F_a = F_t · tan(β) is the standard force relationship. The ratio grows fast with helix angle: 27 percent at β = 15°, 36 percent at β = 20°, 58 percent at β = 30°.
Worked example — a single helical gear pair with z = 15/45 teeth, normal module 2.5 mm, β = 20°, and 100 Nm of input torque. F_t = 5,000 N, F_r = 1,971 N, F_a = 1,820 N. That 1,820 N is real bearing load, not a rounding error.
The common shortcut — “if there’s thrust, specify double helical” — skips the bearing-sizing step. API 613 (special-purpose gear units) and API 617 (centrifugal compressors) both require thrust bearings on single-helical pinion shafts and only recommend them on double-helical.
The standards treat single-helical-plus-thrust-bearing as a sanctioned specification, not a workaround. For the moderate-power range, that path is usually cheaper than paying the double-helical premium — and there are other options to eliminate axial thrust in helical gears that deserve a price check first.
Torque Density per Face Width
At equal face width, double helical carries more torque because load is shared across two helices and contact ratio is higher. The density gain depends on helix angle and quality grade, but the direction is consistent. Double helical earns its size advantage above roughly 1 MW per stage and in heavy continuous duty where housing footprint is constrained.
Below that band, a wider single-helical face width is cheaper than splitting the tooth into two helices.
Allowable Speed and Efficiency
Per-stage mesh efficiency for both single and double helical lands in the 94–98 percent band. The geometries are mesh-equivalent at the teeth. The efficiency delta sits in the bearings, not in the tooth flanks.
In API 613-class machines, the high-speed pinion’s thrust bearing typically debits 0.5–1.5 percent of nameplate. Kingsbury tilting-pad data shows that doubling the oil flowrate raises power loss 25 percent at 4,000 rpm and 38 percent at 13,000 rpm. Discharge geometry can swing the same loss by another 60 percent.
Double helical eliminates this thrust source entirely through its self-cancelling axial geometry. The lifecycle-cost gap therefore widens at 8,000-plus hours per year of duty, well beyond what the per-stage gear number alone suggests.
Pitch-line velocity ceilings diverge only at the top end — above about 50 m/s, double helical’s higher contact ratio becomes structurally necessary at fixed helix angle. In the 15–20° helix band most industrial drives live in, both geometries hit the same speed targets.
Cost per kW Including the Thrust-Bearing Alternative
The true cost comparison is three-way, not two-way: single helical with a light bearing, single helical with a heavy thrust bearing sized per API 613/617, and double helical. The single-helical premium for a heavy thrust bearing is a fraction of the double-helical part-cost premium in the moderate-power range.
Above the heavy-continuous threshold — rolling mills, heavy-mining crushers — the thrust-bearing path stops being economical; the bearing scales badly with power and double helical removes the thrust rather than carrying it.
| Factor | Single Helical | Double Helical |
|---|---|---|
| Axial thrust (β = 20°) | 36% of tangential | ~0% net (internally cancelled) |
| Torque density at equal face width | Baseline | ~1.3–1.6× baseline |
| Stage efficiency | 94–98% | 94–98% |
| Practical speed ceiling | ~50 m/s at β = 15–20° | >50 m/s at β = 20–45° |
| Relative cost per kW | 1.0× | 1.5–2.5× |
| Typical duty class | Light-to-medium, up to ~1 MW/stage | Heavy-continuous, above ~1 MW/stage |
Single vs Double Helical by Industrial Application
The boundary is set by service factor and continuous-duty hours, not by power alone. Textile, packaging, food processing, and standard mining-conveyor drives are single-helical territory — service factor 1.0–1.5, intermittent or single-shift duty, moderate shock.
This duty is addressed by the standard helical gearbox catalog (R/MD inline, F/MP parallel-shaft, K/MJ helical-bevel up to 50,000 Nm) paired with appropriately sized thrust bearings.
Rolling mills, heavy-mining crushers, cement-kiln drives, and 24/7 conveyor trains with high shock cross the line: service factor 2.0 or above, continuous duty, power above the 1 MW-per-stage band. That is where double helical earns its cost — and where heavy-duty gearbox ranges (MTH helical, B/MTB helical-bevel) become the natural spec.
A useful benchmark: the USS Gary Works hot-strip mill (United States Steel, Gary, Indiana) runs 3,000–12,000 hp per stand. After roughly twenty years the original gears had failed — surface distress, cracking, a catastrophic broken pinion shaft.
The rework partnership with Drive Systems Technology Inc. respec’d replacement gears to AGMA Quality Class 10–12 with carburized pinions, modified profiles, 0.020–0.040 inch backlash, shot-peened roots, and CBN hard finishing. Six years into post-rework service the new gears showed virtually no wear.
Heavy continuous-duty rolling-mill service is the territory where the 1.5–2.5× premium pays back. Below that band, a correctly sized single-helical drive with the right bearing is the defensible spec.
Single vs Double Helical: The Specifier’s Decision Rule
Pick single helical with a properly sized thrust bearing when the drive sits in the moderate-power range, service factor is below 2.0, duty is intermittent to single-shift, and the application is not on the heavy-mining or rolling-mill list. Pick double helical when two or more of those conditions flip — power above roughly 1 MW per stage, service factor at or above 2.0, 24/7 continuous duty, or a shock-load profile the bearing alone cannot absorb economically.
Run the F_a · tan(β) number first; price the thrust bearing against the part-cost premium second; let gearbox service factor and continuous-duty hours settle the edge cases. The spec decision stops being gut-feel the moment the math lands on the page.
