A right-angle helical-bevel reducer on a mixer failed after 13 months. The failure mode was scuffing on the bevel gear teeth. The gearbox was rated for 45 kNm output, running a 30 kW mixer at 40 RPM output. The oil was changed on schedule, at 2000-hour intervals. The oil analysis showed the right chemistry. The gear teeth looked like someone had sanded them.
Scuffing at 40 RPM is unusual. Scuffing normally shows up on high-speed gears, where the oil film can’t form fast enough. At 40 RPM, the pitch line velocity is about 2.5 m/s. That’s slow. But the oil was the problem, and it wasn’t a chemistry problem. It was a viscosity problem.
The viscosity number on the can isn’t what’s in the sump
The gearbox was filled with ISO VG 220 synthetic PAO oil. The nameplate on the reducer said “ISO VG 220.” The ambient temperature in the plant was 40°C in summer. The oil in the sump ran at 75°C under load. At 75°C, ISO VG 220 has a kinematic viscosity of about 45 cSt. That sounds fine. But the manufacturer’s rating for the gearbox assumed the oil would be at 90°C. Wait, no. The rating assumed the oil temperature would stabilize at about 70-80°C, and the viscosity at that temperature is the design point.
Here’s the detail that got missed. The gearbox nameplate said ISO VG 220, but it also had a footnote: “for ambient 20°C, use VG 320 for ambient 40°C.” The plant never read the footnote. The mixer ran at 40°C ambient, so the sump ran hotter, and the viscosity at operating temperature was below the design minimum. The oil film was too thin at the bevel gear mesh. Metal touched metal. Scuffing.
The oil analysis came back clean because the oil wasn’t contaminated. It was the right oil for the wrong ambient. A VG 320 would have held 60 cSt at 75°C instead of 45 cSt. That extra 33% film thickness was the difference between a 20-year gearbox and a 13-month failure.
How to pick viscosity for real operating conditions
The rating tables assume a nominal ambient and a nominal load cycle. Real installations have higher ambient, starts and stops, and load spikes. The rule of thumb that works: the operating viscosity at the sump temperature should be at least 40 cSt for helical and bevel gears, 70 cSt for worm gears.
Steps that actually work:
- Estimate the real sump temperature. If you can’t measure it, add 10°C to the nameplate assumption for every 10°C above 20°C ambient.
- Look up the oil’s viscosity-temperature curve (the VI number tells you the slope). A VI 150 PAO drops viscosity less with temperature than a VI 95 mineral oil.
- Check the viscosity at the estimated sump temperature, not at 40°C.
- If it’s below 40 cSt (70 cSt for worm), step up one ISO grade. One grade up costs nothing; a replacement gearbox costs a month of production.
The worm gear case that’s different
Worm gears need thicker oil than helical gears at the same temperature. The sliding contact at the worm mesh generates more heat and needs a thicker film to separate the surfaces. A worm reducer that runs at 60°C sump should have oil that’s at least 70 cSt at that temperature. That usually means VG 460 or VG 680, not VG 220. The common mistake is filling a worm reducer with the same oil as the helical reducers on the same line. It runs fine for two years. Then the bronze worm wheel wears out. The maintenance log says “bronze wheel lasted two years.” Nobody connects it to the oil.
We had a plant with six identical worm reducers, all running VG 220. The bronze wheels lasted 18-24 months. We switched to VG 680 synthetic. The first wheel is still in service after 5 years. Same load, same duty cycle. Only the oil changed.
What oil analysis actually tells you
Oil analysis is a maintenance tool, not a selection tool. It tells you the oil is clean and the additives are alive. It doesn’t tell you the viscosity is right for the application. The lab reports a 45 cSt at 40°C result and says “within spec.” The spec is wrong for the installation. You have to interpret the number against the actual sump temperature.
Two numbers from the oil report matter for gearboxes. The viscosity at 40°C and the viscosity at 100°C. Plot them on the VI curve and read the viscosity at your measured sump temperature. If you don’t have a sump thermometer, install one. It’s a $30 part and it turns oil selection from a guess into a calculation.
The change that fixed the mixer
The mixer gearbox was drained, flushed with the new oil, and refilled with ISO VG 320 synthetic PAO. The bevel gears were inspected. The scuffing was in the early stage — the teeth could be re-lapped rather than replaced. The gearbox ran at the same temperature. The oil at 75°C held 60 cSt. The scuffing stopped. That was four years ago. The gearbox is still running on the same set of gears.
The cost of the change: one oil change and a $100 sump thermometer. The cost of the original failure: a $12,000 gearbox and two weeks of downtime. The oil can was the cheapest component in the drivetrain, and it was the one that failed.
Pick gearbox oil by the viscosity at your real sump temperature, not by the nameplate. 40 cSt minimum for helical and bevel gears, 70 cSt for worm gears. Measure the sump temperature and do the VI calculation. The scuffed bevel gears weren’t a lubrication failure in the classic sense — the oil was fine and clean. It was just too thin at the temperature the gearbox actually ran.
