A gear reducer on a conveyor. It was rated for 500 N·m output torque. The conveyor needed 200 N·m (40% load). After two hours of operation, the reducer housing was too hot to touch (85°C). The customer thought it was defective. It wasn’t — it was operating at 40% torque load but in an enclosure with no ventilation. The thermal rating was the limit, not the torque rating. This is about reducer thermal sizing.
Two ratings, not one
Every gear reducer has two ratings:
- Mechanical rating: the maximum torque the gears can transmit without tooth failure. For our 500 N·m reducer, that’s 500 N·m. We’re at 200 — fine.
- Thermal rating: the maximum continuous torque the housing can dissipate without overheating. For a 500 N·m reducer in free air, that’s about 300 N·m. In an enclosure, it drops to 150 N·m. We’re at 200 — over the thermal limit.
The reducer overheats because the gear mesh generates heat (about 3-5% of the transmitted power is lost as heat). The housing must radiate that heat to the air. If the housing is in a sealed enclosure, the heat can’t escape. The oil temperature rises. At 85°C, the oil breaks down. At 100°C, the seals fail. The reducer dies.
The heat generation
The heat generated by the gear mesh is:
P_heat = P_input × (1 – η)
Where P_input is the input power and η is the gear efficiency (0.95 for a helical gear reducer, 0.85 for a worm gear). For a 2 kW input at 95% efficiency: P_heat = 2000 × 0.05 = 100 W. That’s the heat the housing must dissipate. A standard reducer housing has a thermal resistance of about 0.5°C/W. For 100 W: ΔT = 50°C. If ambient is 30°C, the housing reaches 80°C. That’s the limit.
In an enclosure, the thermal resistance doubles (0.9°C/W). ΔT = 90°C. Housing reaches 120°C. That’s over the limit. The reducer overheats.
What I did
1. Ventilated the enclosure. I added a fan on the reducer’s enclosure. The forced air convection drops the thermal resistance from 0.9 to 0.4°C/W. ΔT = 40°C. Housing reaches 70°C. That’s fine.
2. Added a cooling fan on the reducer. Some reducers have an optional fan kit. The fan blows air over the housing fins. Thermal resistance drops to 0.25°C/W. ΔT = 25°C. Housing reaches 55°C. Even better. The fan kit costs $80. The reducer replacement (or oil change every month) cost $500.
3. Upsized the reducer. If ventilation isn’t possible, I go to a larger reducer. A 700 N·m reducer has a bigger housing (more surface area). The thermal rating at 200 N·m is now well within range. The larger reducer runs cooler because it’s oversized. This is the brute-force fix, but it works.
The duty cycle
Thermal rating is for continuous duty. If the reducer runs intermittently (10 minutes on, 10 minutes off), it has time to cool. The thermal limit is higher. I check the duty cycle. For our conveyor, it runs 24/7 continuous. No cooling time. The thermal rating is the bottleneck. For a machine that cycles (10 seconds on, 20 seconds off), the thermal limit doubles. The reducer can handle more torque because it cools between cycles.
The oil
At high temperatures, the oil degrades. I specify synthetic gear oil (ISO VG 220) instead of mineral oil. The synthetic oil runs cooler (lower friction) and tolerates higher temperatures. For continuous-duty reducers, I also schedule annual oil changes. Mineral oil at 80°C lasts 6 months; synthetic lasts 2 years. The oil change is cheap insurance.
The rating I check: thermal, not just mechanical. A reducer running at 40% torque can still overheat if it’s in a sealed box. The heat from gear mesh must escape. Ventilate, fan, or upsized. The reducer that burned at 40% load wasn’t over-torqued — it was overheated. The mechanical rating says “it can push it”; the thermal rating says “it can survive pushing it.”
