![]() |
Charlotte Stonestreet
Managing Editor |
| Home> | POWER TRANSMISSION | >Gearboxes | >Specified on paper, failed in service: what continuous duty really demands |
Specified on paper, failed in service: what continuous duty really demands
19 October 2026
Most continuous-duty gearbox failures trace back not to the calculation, but to what the datasheet leaves out. Thomas Parker sets out the four factors that decide service life, and the questions OEM design teams should answer prior to selection

WHEN I investigate a gearbox failure in continuous-duty machinery, the cause is rarely a calculation error. The torque was right. As was the ratio. The unit was sized correctly against the datasheet. But it failed. Why?
What decides service life sits outside the datasheet: how often the machine starts and stops, what the driven equipment does to the shafts, how much heat the gearbox can shed, and what service factor really means. Together, they explain why a gearbox that looked comfortable on paper becomes a vulnerability very quickly.
Duty cycle and start-stop frequency
A torque rating assumes a pattern of use. Eight hours of steady running is a very different life from 24 hours of short cycles. Every start, reversal or stall drives peak torque through the teeth and bearings, often well above running torque.
A conveyor indexing 20 times an hour continuously sees more than 170,000 starts a year. That is fatigue loading no nominal torque figure captures. I always ask for the real operating pattern: hours per day, starts per hour and the inertia the motor must accelerate.
What the driven equipment does to the shafts
Catalogue ratings usually assume a pure torque load. Belt and chain drives pull sideways on the output shaft, overhung pulleys add leverage, and helical gear or fan loads push along the shaft axis.
Bearings and seals carry these radial and axial loads, not the gears. That is why so many failed units I see have healthy teeth but worn-out bearings. Mount a pulley further out than the catalogue assumes and bearing life drops drastically. Check overhung load ratings against the actual layout, including whether the load reverses.
Thermal capacity is not mechanical capacity
Mechanical capacity tells you whether the gears and shafts survive the load. Thermal capacity tells you whether the unit can shed the heat generated in the mesh, bearings and seals during continuous operation. Exceed it and the oil thins, seals harden and wear accelerates, even though every stress calculation passes.
Ambient temperature, airflow and mounting position all move that limit, and cold matters as much as heat. Below zero, oil thickens, starting torque climbs and seals lose flexibility. A unit specified for a warm factory floor may not cope with a freezer.
Service factor is a duty correction, not a safety margin
Service factor is still widely read as headroom: choose 1.5 and you have 50% to spare. That is not its purpose. It adjusts a rating established under standard conditions to reflect the real duty: running hours, shock level and start frequency.
A machine with heavy shock loads and frequent starts may need every bit of a high service factor simply to reach normal life. Treat it as spare capacity and the duty quietly uses it up, while the design team believes in a margin that was never really there.
How they combine: a refrigerated transport example
These factors compound. Frequent starts raise peak torque and heat, cold raises starting torque further, and overhung loads press on bearings already working against stiff lubricant. Each might sit inside its limit alone; together they do not.
We’ve seen this before with an OEM of refrigerated transport equipment. Its gearboxes were correctly sized for nominal torque, yet failures appeared in service. With the OEM's design team, we looked at what the units had to do: start and run reliably below zero. We specified for that duty rather than the nominal torque figure, and gearbox failures went to zero.
The original calculation was not wrong. It answered a different question from the one the application was asking. That’s where application knowledge matters.
Questions to answer before selection:
- An OEM design team should answer these before selection, not after commissioning:
- How many hours a day will it run, and how many starts, stops and reversals per hour?
- What inertia does the motor accelerate, and what peak and shock loads follow?
- Where do radial and axial loads act on each shaft, how large are they, and do they reverse?
- What are the lowest and highest ambient temperatures, and how much airflow will the unit get?
- Does the thermal rating cover continuous running in that environment?
- Which service factor applies to this duty, and why?
If any answer is a guess, that is where the risk sits. Answering them at the design stage may cost a few hours. Answering them after commissioning costs downtime, warranty claims, redesigns, and potentially reputational damage. Which outcome would you prefer?
Thomas Parker is sales manager at Brown Geartech
- Successful field tests for astronaut smart glove
- Smart cities expected to be commonplace within 10 years
- UK manufacturing output & new orders fall at fastest rates since 2012
- £365m funding announced for new UK aerospace R&D projects
- What rising raw material prices mean for electronics
- ERA Foundation celebrates 25 years
- Robotics & AI Industry Showcase
- Common Control Platform for automotive manufacturers
- Deus Robotics secures $3m
- Improved closed loop control
- GEARBOX IS CHEAPER TO REPLACE THAN REPAIR
- PLANETARY/SPUR GEARBOXES
- PLANETARY GEARS WELL SUITED FOR RETROFIT
- AGEING WORM GEARS REPLACED BY PLANETARY GEARS IN SEWAGE PLANT
- REHFUSS BRINGS QUALITY AND VERSATILITY TO GEARBOX DESIGN AND SELECTION
- FLEXIBLE COUPLINGS
- GETTING THE BEST OUT OF GEARBOXES
- A Guide to the Selection
- Greater Corrosion Resistance
- Fits existing parameters


















