Gearbox and Planetary Reducer Selection for ASEAN Industrial Drives

Thermal Power, Not Torque, Is What Limits a Gearbox in a 40 °C Plant

If your shop floor sits at 35–40 °C, the rating that decides whether an industrial gearbox survives is thermal power, not the torque or kW figure on the front of the datasheet. Bonfiglioli's 3/H heavy-duty catalogue prints both, side by side, at 20 °C and 40 °C ambient. On size 3/H 11L2 at 30.3:1 the mechanical rating is 202 kW — but basic thermal capacity is 80 kW at 20 °C and only **36 kW at 40 °C**: 45 % of the temperate figure, and roughly 18 % of the rating you probably selected on. For precision servo planetary reducers the equivalent trap is backlash: an arc-minute figure is measured at 1–2 % of rated torque, not a property under working load. Ask for thermal power at your ambient, torsional stiffness, and the torque the backlash was measured at, and most quotes stop being comparable the way the sales sheet implies.

What the catalogue actually says

The definition is plain: thermal power is "the maximum power that can be mechanically transmitted by the gear unit, in continuous operation, without the internal temperature reaching a value that could damage the gear unit components." The selection rule is one line — basic thermal capacity must be at least the input power consumption — and it is the check a tropical installation fails first.

Two verified rows show the shape of it. On 3/H 15L3 at 179.2:1 and 1500 rpm input, basic thermal capacity falls from 99 kW at 20 °C to 57 kW at 40 °C — a 42 % loss; on the 3/H 11L2 row above, 55 %. The conditions behind them matter as much: continuous operation, horizontal mounting, under 1000 m altitude, ISO VG 220 oil, and "installation in large open spaces (air velocity > 1.4 m/s)." A reducer bolted inside a guarded machine base, under a mezzanine or against a wall never sees 1.4 m/s. Neither assumption describes a typical ASEAN production hall.

The Cures Lose Capacity in the Same Heat That Created the Problem

The usual fix is a shaft-mounted cooling fan. The same catalogue quantifies it: on that 3/H 15L3 row, fan-assisted capacity is 168 kW at 20 °C and 108 kW at 40 °C — still a 36 % loss. Note the manufacturer's caveat: "the forced ventilation efficiency is greatly reduced for intermittent operation and drive speeds below n1 = 900 min-1." Slow, intermittent drives — mixers, screw conveyors, sludge rakes — run hot and gain least from a fan.

If the fan is not enough, the next step is an oil cooling unit, and the derating repeats. The catalogue's air/oil units exchange 5.6 / 7.5 / 16 / 21.6 kW at 20 °C ambient, but 3.4 / 4.5 / 9.6 / 13 kW at 40 °C: about 40 % less capacity at the ambient that made you buy the cooler. Its own worked example: a 3/H 21L3 at 40 °C carrying 275 kW has 130 kW basic and 220 kW with the fan — the fan alone still fails. Size the cooling chain at your ambient, or buy the accessory twice.

"3 Arc Minutes" Is a Measurement Condition, Not a Property

Precision planetary reducers are sold on an arc-minute number, and buyers pick the smallest of three quotes. The number is real but conditional. As gearbox manufacturers describe the test, "the load used to perform the test is not arbitrary. The typical value is 1-2 % of nominal torque capability" — deliberately low, so torsional windup is minimised and lost motion alone is measured. At 40 % of rated torque the shaft, gears and carrier deflect, and the lost motion your machine actually sees is backlash plus windup. So compare two numbers: backlash in arc minutes **and** torsional stiffness in Nm per arc minute — then ask at what torque it was measured.

Separate gear quality from backlash as well. Listings in this class commonly cite ISO 4 grade gear grinding; in the ISO accuracy system a lower grade number means tighter tolerance, and the grade describes tooth geometry — runout, pitch, pressure angle and helix deviation. That influences backlash but does not state it. Where a supplier does commit to a figure, as the [ZY AB Series](/en/products/jiangsu-zhiyuan-precision-technology-zy-ab-series-high-precision-planetary-reducer) listing does with backlash below 5 arc minutes, that is the right disclosure — the follow-up is simply "at what torque, and what stiffness?" In this class generally, "ultra-low backlash" without a number is a marketing phrase, not a specification.

Ratio Is an Inertia Decision Before It Is a Torque Decision

On servo axes the ratio is usually chosen to meet output torque, and the machine then refuses to tune. A gearbox divides reflected load inertia by the square of the ratio — reflected inertia is load inertia ÷ ratio², plus the gearbox's own. As one drive supplier puts it, "a 10:1 gearbox has a 100:1 impact on the inertia ratio." A 1:1 load-to-motor ratio is easiest to control; beyond roughly 10:1, tuning becomes materially harder.

So the order is: compute reflected inertia, choose the ratio that puts the inertia ratio where your drive can tune it, then verify output torque and speed — not the reverse. A ratio chosen on torque alone can be correct on paper and impossible to commission. The inline [ZY PLX](/en/products/jiangsu-zhiyuan-precision-technology-zy-plx-series-high-precision-planetary-reducer) and [ZY ZPX](/en/products/jiangsu-zhiyuan-precision-technology-zy-zpx-series-high-precision-planetary-reducer) and right-angle [ZY AFR](/en/products/jiangsu-zhiyuan-precision-technology-zy-afr-series-high-precision-planetary-reducer) cover the same ratio range in different envelopes — geometry is a separate decision, and both come after inertia.

Sealed or Breathing: Two Different Tropical Failure Modes

Precision planetary reducers in this class are commonly sealed and grease-filled, and several list IP65 protection — the [ZY AE](/en/products/jiangsu-zhiyuan-precision-technology-zy-ae-series-high-precision-planetary-reducer), [ZY AF](/en/products/jiangsu-zhiyuan-precision-technology-zy-af-series-high-precision-planetary-reducer) and [ZY PH](/en/products/jiangsu-zhiyuan-precision-technology-zy-ph-series-high-precision-planetary-reducer) listings all do. IP65 is worth having in washdown areas, but it rates dust and water arriving from outside: it says nothing about humid air sealed in at assembly, or about condensation. For a sealed unit the buying question is seal and grease life in hot, humid duty, not the IP digits.

Oil-sump industrial gearboxes have the opposite problem, because they breathe. As the unit heats it expels air; as it cools overnight it draws ambient air back in, and in a tropical plant that air is close to saturated. Lubrication-industry sources identify the breather as the main moisture ingress path; the cure is a desiccant breather, not a plain vent cap. Two commissioning details follow: many gearboxes ship with a sealed transport plug that must be swapped for the working breather before start-up — put it on the acceptance checklist — and the catalogue warns that outdoor units need protection "to prevent direct exposure to atmospheric agents and solar radiation." Finally, ISO VG grades are defined by viscosity at 40 °C, and the 3/H thermal figures assume ISO VG 220: ask for the lubricant recommendation and first drain interval **at your ambient band and expected sump temperature**.

What to Put in the Purchase Order

• Thermal power at **40 °C ambient**, stated separately from the mechanical rating, plus the assumed air velocity and mounting position
• The service factor the quote assumes, and for how many hours per day and starts per hour — industry practice maps AGMA Class I / II / III to roughly 1.0 / 1.41 / 2.0
• For servo reducers: backlash in arc minutes, the torque it was measured at, and torsional stiffness in Nm/arc-min
• Reflected inertia and the resulting inertia ratio at the quoted gear ratio
• Desiccant breather specified, and the transport-plug swap on the commissioning checklist
• Lubricant grade and first-change interval for 35–40 °C ambient
• Seal type, with spare seals and breather elements quoted alongside the unit — see our guide to [spare parts and after-sales for Chinese machinery](/en/news/spare-parts-after-sales-chinese-machinery-asean)

We do not publish prices for these reducers and will not invent one. What makes quotes comparable is insisting that every supplier answers the same seven items in writing.

Conclusion

Buy on thermal power at your real ambient, choose ratio from inertia then confirm on torque, and treat backlash figures as measurements with conditions. Browse the [gearbox and reducer category](/en/products?category=gc-gearbox-reducer), read our [industrial robot ROI guide](/en/news/industrial-robot-roi-asean-factories) and [CNC control comparison](/en/news/cnc-control-siemens-fanuc-mitsubishi-asean), or send your load, duty cycle and ambient temperature through [contact](/en/contact) — we will ask the supplier the seven questions for you.