Gears are the workhorses of mechanical power transmission. If you are a mechanical engineer building a gearbox, a conveyor drive, or a robot joint, you will eventually face a simple question: how do I pick a gear pair that will not fail in service? This tutorial walks through spur gear design step by step, using realistic formulas and one complete worked example, the kind of calculation you would actually do in a design office rather than in a textbook.
😀 Why start with spur gears? Because they are the simplest, most forgiving tooth form, and every concept you learn here carries over to helical and bevel gears. Master the spur gear, and the rest of the gear family becomes a variation on the same theme.
1. How a Gear Pair Actually Transfers Power
When two gears mesh, the driving tooth pushes the driven tooth along the line of action. The contact point moves smoothly if the tooth profile is an involute, which is why virtually all industrial gears use involute tooth shapes. The involute guarantees constant velocity ratio, which means zero vibration from the tooth geometry itself. Two quantities define the gear size: module (m) in millimeter or diametral pitch in inch-based systems, and number of teeth (z). The pitch diameters are simply m times z.
Power enters as torque times rotational speed. A speed reduction multiplies torque by roughly the gear ratio, minus friction and other losses. That is the physical heart of every gearbox on a production floor.
2. Choosing Gear Types: A Quick Field Guide
Not every application wants a spur gear. Use this mental checklist:
- Spur gears: parallel shafts, high efficiency, noisy at high speed, cheapest to make.
- Helical gears: parallel shafts, quieter and stronger because teeth engage gradually, but they generate axial thrust.
- Bevel gears: intersecting shafts, typically 90 degrees, used in differentials and right-angle drives.
- Worm gears: perpendicular, non-intersecting shafts, huge reduction in one stage but lower efficiency.
- Planetary gears: compact high-ratio reducers used in robots and winches.
For this tutorial we focus on spur gears, where the classic equations make the physics transparent.
3. The Core Formulas You Will Use Every Day
Three formulas show up in almost every gear calculation. First, the gear ratio i equals the ratio of driven to driving teeth, i = z2/z1. It is also the ratio of input speed to output speed. Second, the center distance a = m(z1+z2)/2. Getting this right is essential for the housing and shaft layout. Third, pitch line velocity v = pi * d * n / 60000 in m/s when the pitch diameter d is in mm and speed n is in rpm. These three formulas unlock every sizing question that follows.
Always compute the gear ratio twice, once from teeth and once from speed, to catch data-entry errors early. In my own design reviews, this simple double check has caught more mistakes than any finite element analysis ever has.
4. Checking Bending Stress: The Lewis Equation
Tooth bending is the most common calculated failure mode. The Lewis equation models the tooth as a cantilever beam and gives the bending stress sigma_b = Ft / (m * b * Y), where Ft is the tangential load on the tooth in newtons, m is the module in mm, b is the face width in mm, and Y is the Lewis form factor, a dimensionless number that depends on the tooth count. Notice that stress falls when you increase module or face width, so a bigger, wider gear can always survive more load.
In practice you apply a velocity factor Kv and a service factor Ka to account for dynamic loads and operating severity. The allowable bending stress comes from the gear material and its surface treatment. A hardened and ground gear can run at much higher bending stress than a soft through-hardened gear, which is why automotive gearboxes use case-hardened low-alloy steels.
5. Checking Surface Durability: Contact Stress
Pitting – small craters on the tooth flank – is the second dominant failure mode. It is governed by the Hertzian contact stress at the pitch line, sigma_c = C * sqrt(Ft * (i+1) / (b * d1 * i * E)), where C collects constants, E is the combined elastic modulus, and d1 is the pinion pitch diameter. Surface strength depends heavily on hardness: the Brinell hardness of the gear steel basically sets how much contact stress the flank can tolerate. This is why case hardening and nitriding matter so much.
Many beginners size gears for bending only and then wonder why their gearbox fails after a few months under load. The honest answer is usually surface pitting. Always check both bending and contact stress, then design defences against the weaker one.
6. Worked Example: A 20:1 Two-Stage Reducer
Suppose you must transmit 3 kW from a 1440 rpm motor down to 72 rpm output for a conveyor. The total ratio is 20. Instead of one huge stage, use two stages of about 4.47:1 each. Round to 4.5 and 4.44, giving overall about 20. Choose module m = 2.5 mm for the first stage, pinion 18 teeth and gear 80 teeth, face width 30 mm. The pitch line velocity is about 4.7 m/s, well inside the range for a quality machined gear.
Running the Lewis check with a tangential load near 1200 newtons and a form factor near 0.3 gives a bending stress comfortably below the 180 MPa allowable for a 40Cr steel in the soft condition. The contact check also passes with several hundred megapascals, provided the flanks are hardened to around 55 HRC for the second stage where the torque is highest. The numbers work, and the gearbox fits in a housing about 300 mm long. That is the difference between a theory problem and a deliverable design: every equation maps to a physical dimension you can machine.
7. Material Selection and Heat Treatment
Pick the gear steel by duty. Low-cost applications use 45 or 40Cr medium carbon steel, normalized or quenched and tempered. Heavily loaded gears use 20CrMnTi case-hardened to 58 to 62 HRC, then ground, which is the standard recipe for automotive and industrial reducers. Plastic gears made from POM or nylon suit light, quiet, lubrication-free mechanisms such as small printers and toys. When vibration or water is present, stainless and special alloys enter the picture, but they always cost more.
Heat treatment order matters: rough machine, heat treat, then finish grind. Grinding after hardening restores the accuracy that distortion removes, and it leaves a compressive surface layer that resists crack initiation.
8. Lubrication, Backlash, and Accuracy Grades
No gear design is complete without thinking about oil. Splash lubrication works up to about 12 m/s pitch line speed; above that you need forced oil jet delivery directly at the meshing zone. The oil grade comes from the pitch line speed and load: heavier loads and slower speeds use higher viscosity oils such as ISO VG 220, while high-speed lightly loaded trains run VG 68 or even lower. Add EP (extreme pressure) additives when contact stress is high, because the additives form a sacrificial film that prevents microwelding.
Backlash is the intentional clearance between teeth. Some backlash is essential to allow thermal expansion and to keep the film of lubricant in the mesh. Too little backlash causes jamming and noise; too much causes impact and rattle, especially in reversing drives. A CNC feed axis typically asks for very small backlash or none at all, which is why precision gearboxes advertise their backlash in arc minutes.
Machining accuracy is graded from ISO 1328 classes. A grade 7 gear is a reasonable general-purpose industrial gear. Grade 6 suits machine tools and automotive main gears. Grade 5 and better are precision and instrument gears. Every drop in grade number approximately doubles the machining cost, so buy only as much accuracy as the application truly needs.
9. Common Failure Modes and How to Prevent Them
- 🛠️ Tooth breakage: usually fatigue bending; add rams, increase module, or use a stronger steel.
- 🕳️ Pitting: surface fatigue; harden the flank, improve the oil film, or reduce contact stress.
- ⚙️ Scoring and scuffing: local welding from high temperature; change oil to an EP grade and improve cooling.
- 🔥 Wear with debris: contamination; filter the oil and protect the gearbox from dust.
- 💥 Tooth chipping: usually impact or overload; review the service factor and duty cycle.
Every failure mode has a signature sound, temperature, and wear pattern. Record them. A gearbox that talks to you, through vibration and oil analysis, fails predictably instead of catastrophically.
10. From Paper Design to Production
Hand the geometry to the shop as a clean drawing with the backlash, accuracy grade, heat treatment, and surface finish explicitly stated. If you can, generate the tooth profile from the actual hobbing or grinding tool data, so the design and the machine agree. Many teams now export gear geometry directly to CNC and use CMM inspection at the end, closing the loop between calculation and measurement.
Conclusion
Spur gear design boils down to a short checklist: compute the ratio, size the module and face width with the bending check, verify the contact stress, choose material and heat treatment, then handle lubrication, backlash, accuracy, and failure prevention. The worked example shows that a realistic 20:1 reducer can be sized on one page of calculations with no fancy software, as long as you respect the physics. Whether you are building a conveyor gearbox or a robot actuator, these are the calculations that separate a working machine from a recurring maintenance problem. Keep this tutorial beside your desk, run both stress checks every time, and your gear designs will quietly outperform the ones that were guessed.
11. Quick Sizing Rules to Remember
When you are in a hurry, keep these thumb rules in mind. A face width of 6 to 10 times the module is a safe starting point for steel spur gears. A minimum of 17 teeth on the pinion avoids undercutting at the standard 20 degree pressure angle. If your pinion has too few teeth, cut the teeth with positive profile shift rather than accepting the weak, undercut geometry. And when input speed is high, prefer helical teeth from the start, because they cut vibration dramatically and cost surprisingly little extra in a production run.
Finally, document your calculation assumptions in the design file. Future engineers, including you in six months, will thank you for writing down the service factor, the material grade, and the lubrication plan. A gearbox is a product, but the design rationale is the real engineering asset.
Key takeaway: gear design is a sequence of two strength checks plus one reality check. The bending check protects the tooth, the contact check protects the flank, and the reality check is simply asking whether the gearbox fits, ships, and survives the duty cycle you promised the customer.
Start your next gearbox with these pages open, crunch the numbers twice, and let the physics decide the module instead of guessing from a previous project.
