5 Ways to Fix a Gear That Fails Its Safety Factor
By Borut Černe ·
The rating came back and the root or flank safety factor is below your target. Before you resize the whole gearbox, work through five levers in order of least disruption — and one check that comes before all of them.
First: make sure the gear is actually failing
A surprising share of "failing" gears are really failing inputs. Before touching the geometry, check:
- Torque and speed — is this the real operating point, or a worst case stacked on a worst case?
- Application factor — does it honestly describe your drive and driven machine, or was it copied from a previous project with rougher duty?
- Required life — demanded load cycles drive the permissible stress. A life target nobody asked for is free pessimism.
- Operating temperature (plastic gears) — polymer strength falls steeply with temperature, so an overly hot assumption sinks the rating before geometry gets a say.
If the inputs survive scrutiny and the margin is still short, work the levers.
Levers 1–3: geometry
1. Increase the module
The biggest hammer. A larger module means a thicker tooth, and root bending stress drops quickly with tooth size — this is the go-to lever when SF is the problem. The costs: at a fixed center distance, a larger module means fewer teeth, which worsens specific sliding and contact ratio and eventually risks undercut; with the tooth count kept, the gear simply grows. Either way the rating must be re-run, not extrapolated.
2. Widen the face width
More width spreads the same load over more tooth, helping both SF and SH without changing the mesh kinematics. It's the cheapest lever while axial space lasts. The limit is real, though: the wider the gear relative to its diameter, the harder it is to keep load evenly distributed across the width — shaft bending, misalignment and manufacturing deviations start to concentrate load at one end, and the load-distribution factor claws back the benefit.
3. Rebalance with profile shift
Profile shift is the precision tool. Shifting the pinion positive thickens its root — usually the weakest point of the pair — and at the same time lets you balance specific sliding between the two gears. It costs no size at all. The trade-offs: the shift budget at a fixed center distance is shared between the two gears, and too much positive shift thins the tooth tip and shrinks the contact ratio. Small moves, checked often.
Levers 4–5: material and manufacturing
4. Upgrade the material or heat treatment
The largest capacity jump that leaves the geometry untouched. Moving from a structural or through-hardened steel to a case-hardened or nitrided one raises the permissible stresses substantially — flank capacity most dramatically. The price: hardening adds process steps, distortion, and typically post-heat-treatment grinding. For plastic gears this lever reads "pick a stronger polymer or lower its temperature" — moving from a commodity thermoplastic to PEEK, or improving heat dissipation, shifts the whole strength curve.
5. Tighten the quality grade
The subtle one. Manufacturing accuracy (the ISO 1328 quality grade) feeds the dynamic and load-distribution factors in the rating: a more accurate gear runs with smaller internal dynamic loads, so the effective stress drops even though the nominal load is unchanged. It helps both SF and SH — but it buys margin with manufacturing cost, and only up to what your process can actually hold. Claiming a grade the shop can't deliver is a paper improvement.
Bonus lever: let the optimizer search
The five levers interact — module changes sliding, width changes load distribution, shift changes both — and hand-iterating the combinations is where afternoons go to die. Zenomiq's AI design mode turns the search around: mark tooth count, face widths, profile shifts or helix angle as optimization variables with bounds, set targets for the root safeties, sliding and contact ratio, and let the optimizer find a geometry that meets them. A volume weight keeps it from simply making everything bigger — which is, after all, the lever you were trying to avoid.
When to stop
Every lever spends something — size, mass, money or machining time. Once the rating passes with inputs you actually believe, additional margin is not "extra safety", it's cost without a requirement. Document the target you designed to and why, and stop there.
If you want the background on what those factors mean before pulling levers, read our explainer on SF vs SH. And when you're ready to try the levers live, the ISO 6336 gear design tool re-rates a pair in about the time it takes to change the input.
Frequently asked questions
Which lever helps flank safety (SH) the most?+
Anything that lowers contact pressure or raises contact strength: more face width, larger diameters (via module or center distance), and above all a harder flank — surface hardening is the single biggest step for contact capacity on steel gears.
Does increasing the module always help?+
It reliably thickens the tooth and helps the root. But at a fixed center distance a larger module means fewer teeth, which can worsen specific sliding, reduce the contact ratio, and eventually risk undercut on the pinion. Always re-run the full rating after the change.
Can profile shift fix both gears at once?+
Profile shift redistributes strength within the pair rather than adding new material. Shifting the pinion positive typically strengthens its root and balances specific sliding — but watch the mating gear and the tip thickness as you go.
What if none of the levers get me there?+
Step back before forcing it: re-examine the torque, life and application assumptions (they are wrong more often than the geometry), and then question the architecture — a larger center distance, a different ratio split, or two stages instead of one may solve cleanly what tooth-level tweaks cannot.
Do these levers apply to plastic gears too?+
The geometry levers do — module, width and profile shift work the same way. The material lever becomes polymer selection and temperature management, since a thermoplastic's strength depends strongly on operating temperature. Note that per VDI 2736, root bending governs for plastic gears.