Choosing the right gear type is one of the first decisions in any transmission design, and it quietly sets the efficiency, noise level, physical envelope and running cost of the whole gearbox for years. Four factors drive the choice. The first is the transmitted load and how it is applied, whether it is steady torque, repeated shock, reversing duty or a mix of all three. The second is speed: low-speed reducers quietly tolerate geometries that a high-speed drive would never accept, because impact and sliding grow with pitch line velocity. The third is the shaft arrangement, parallel, crossed or right-angle, which often eliminates two of the four candidate geometries on the spot before any numbers are crunched. The fourth is the balance between noise, backlash, efficiency and unit cost, because the technically perfect gear is rarely the cheapest to buy, install, run and maintain over a ten-year service life. This guide walks through the four most common gear types, compares them side by side in plain prose, and ends with the parameters you should put on the table when you ask a supplier for a quote.
Spur gears are the default starting point for almost every design team because they are the simplest, the cheapest to cut and the easiest to specify, inspect and replace. Their teeth run parallel to the shaft axis, so contact along the face width is purely rolling, with no axial thrust to speak of. That makes them predictable, forgiving on alignment and friendly to standard deep-groove bearing arrangements. The mechanical limit is that engagement starts and ends abruptly along a line across the tooth, producing impact at the meshing frequency and a characteristic whine that becomes noticeable as soon as pitch line velocity climbs past a few metres per second. Spur gears shine on low-to-moderate speed, high-load, parallel-shaft drives where noise is not the primary concern: machine tool headstocks, conveyor drives, small industrial reducers, printing presses, packaging machinery and starter drives for internal combustion engines. They are also the geometry of choice for backlash-sensitive positioning axes, because tooth-to-tooth play is easier to control on a straight-tooth form than on any curved or helical variant. Above a few metres per second of pitch line speed, or when the drive must sit close to people or to a precision instrument, spur gears start to lose ground to helical designs, and the noise budget is usually the reason.
Helical gears cut the teeth at a helix angle on the cylinder, so contact begins at one edge of the tooth and sweeps across the face width gradually instead of slamming in all at once. The result is a higher contact ratio, smoother load transfer, lower noise at the same pitch line velocity and the ability to carry more torque on the same pitch diameter. For noise-sensitive or high-speed machinery, automotive manual and dual-clutch transmissions, electric vehicle reduction drives, industrial compressors, blowers and high-speed spindles, helical gears are almost always the answer, and in many cases they are specified even when a spur gear would technically carry the load. The trade-off is axial thrust: the helix angle pushes the pinion and gear apart along the shaft, so bearings must take a combined radial and axial load, and efficiency drops slightly because of the sliding friction along the tooth flank. Double helical or herringbone gears cancel the thrust by pairing opposite hands on one shaft, but they cost more to manufacture, demand a precise gap between the two tooth rows and are harder to deburr and inspect. When noise and speed dominate the spec, the small efficiency penalty and the bearing upgrade are almost always worth it, and the long service life pays back the premium.
Worm drives pair a screw-like worm with a bronze worm wheel and are the only common geometry that delivers a high reduction ratio in a single stage, often twenty to one up to sixty to one, and in custom designs above a hundred to one, between non-intersecting shafts at ninety degrees. The single-stage ratio collapses what would otherwise need three or four spur or helical stages into one compact, quiet, sealed envelope, which is why worm reducers are so common in compact machinery. The same sliding contact that makes the ratio possible also creates two traits that buyers care about deeply: self-locking, in which the wheel cannot back-drive the worm and the load is held in place without an external brake, and low efficiency, typically in the forty to seventy percent range, which means more heat to dissipate, more careful lubrication and a higher electricity bill over years of continuous duty. Worm sets are the right call when a compact right-angle reducer, quiet operation and a hold-the-load behaviour matter more than peak efficiency, conveyor drives, lifting equipment, valve actuators, food packaging machinery, small machine tool heads and stage equipment are classic examples. They are the wrong call when the gearbox runs continuously for hours at a time, because the heat generation and bronze wheel wear become a real operating penalty, and the efficiency cost shows up on the energy bill.
Bevel gears sit on shafts that intersect, almost always at ninety degrees, and transmit motion around a corner where parallel gears simply cannot reach. Straight bevel gears are the closest equivalent of spur gears on a cone: easy to make, easy to set up, quiet at low speed and noisy above it. Spiral bevel gears curve the teeth in an arc for gradual engagement, in the same way helical gears improve on spur gears, and they run smoother, carry more load and tolerate higher pitch line speed, at the price of more complex geometry and more expensive cutting. Zerol bevel gears sit in between, with curved teeth but a zero spiral angle, and share the thrust behaviour of straight bevels. The practical pain point across all bevel types is axial thrust, which changes direction with handedness and with the direction of rotation, so bearing selection, preload and mounting distance must be specified together with the gear blank. Bevel gears are the only sensible choice when two shafts must meet at an angle: automotive differentials, printing and textile machinery, machine tool angle heads and rotary tables, right-angle output stages of gearboxes, and hand tools such as angle grinders and right-angle drills. If the two shafts are parallel, bevel gears are never the right answer, no matter how compact the housing looks on paper.
Put side by side, the four geometries line up quite predictably on the engineering dimensions buyers actually compare. On single-stage reduction ratio, worm gears win outright, then spiral bevel, then helical, then spur, which is why worm drives are the default for compact high-ratio boxes. On per-stage efficiency, spur and helical lead at roughly ninety-six to ninety-eight percent, straight bevel is close behind, and worm drives trail because of the inherent sliding friction between the steel worm and the bronze wheel. On noise at a given pitch line speed, helical and spiral bevel are the quietest, worm sets are surprisingly quiet because the sliding contact absorbs tooth impact, and straight bevel and spur are the noisiest as soon as speed climbs. On axial thrust, spur gears need none, which simplifies bearings; helical and spiral bevel produce a predictable thrust that the bearings must handle; and worm drives put a heavy separating force on the shafts, which pushes bearing and housing sizing upward. On purchase cost, spur is cheapest, then straight bevel, then helical, then worm, because the worm wheel usually runs a bronze-on-steel pairing that adds material and machining cost, and the worm itself is ground to a precision thread. On backlash control, spur and straight bevel are easiest to tune by shimming centre distance, while worm and spiral bevel demand more careful setup because the contact pattern moves with mounting position. You can read the same comparison in more detail in the dedicated notes on spur versus helical behaviour and on worm versus bevel applications, then cross-check the efficiency and maintenance expectations in the worm drive deep dive, and confirm the module and pitch chosen will carry the torque you calculated.
A practical selection flow collapses all of this into a short sequence of questions. First ask whether the input and output shafts intersect. If they meet at an angle, almost always bevel gears, and pick spiral bevel if speed or noise is a concern, straight bevel if cost and low speed dominate. If the shafts are parallel and the required ratio is high above ten to one in a single stage, think worm even though the worm geometry places the shafts at ninety degrees by design, because that is the one case where worm competes on ratio rather than on shaft angle. If the shafts are parallel and speed is low to moderate and noise is acceptable, spur gears are the cheapest, simplest and most reliable answer. If speed is high, noise is a concern, or load is heavy on a compact envelope, move to helical gears, and choose double helical when axial thrust on long slender shafts is unacceptable. If the application must hold a load without an external brake, or needs a single-stage ratio above twenty to one in a small box, worm is the default despite lower efficiency. At the end of the flow, two or three candidates usually remain, and the choice is settled by the noise budget, the thermal capacity of the housing and the price target.
When the geometry is narrowed to one or two candidates, the inquiry to the supplier should carry a short, specific list of numbers so the recommendation is credible rather than generic. Give the input speed and the required output speed, which sets the ratio and confirms whether a single stage is realistic. Give the torque on both shafts, and whether the load is uniform, moderate shock or heavy shock, because that changes the service factor and the face width the supplier should quote. State the centre distance or the available envelope, because that constrains the pitch diameter and often rules out one of the candidates. State the required accuracy grade, whether DIN 3961, ISO 1328 or AGMA class is more familiar to your customer, especially if the drive is for positioning or for high-speed duty where pitch error shows up as vibration. State the preferred material and heat treatment, steel grade through-hardened, case-carburised, nitrided, or even plastic, or let the supplier propose it against your load and duty cycle. Finally, state the annual or one-off quantity, because the same gear geometry can be hobbed, shaped, ground, broached or even 3D printed depending on batch size, and unit cost swings hard between those routes. With these six numbers on the table, the supplier can confirm or push back on the gear type you started with, rather than sending a price list that pretends to fit a transmission nobody has actually sized yet.