A deep groove ball bearing that ran hot after installation. The bearing was 6310, mounted on a 50 mm shaft with a transition fit (k5) in a cast iron housing. The customer bought a standard clearance (CN) bearing. The housing was at 35°C ambient. The bearing reached 90°C surface. The grease thinned. The bearing failed after 3 months. The issue: the bearing was a normal clearance (CN) but the shaft grew thermally more than the housing. The internal clearance disappeared. The bearing ran in a preloaded condition. This is about bearing internal clearance selection and how thermal growth affects it.
The internal clearance classes
Ball bearings are supplied with different internal radial clearances. The clearance is the amount the inner ring can move radially relative to the outer ring. The standard classes are:
| Class | Radial clearance for 6310 (μm) | Use for |
|---|---|---|
| C2 (extra clearance) | 5-20 | Angular preload, high precision |
| CN (normal) | 20-35 | General, steady load |
| C3 (extra) | 35-55 | Interference fit, temperature rise |
| C4 (extra extra) | 55-80 | Heavy interference, high temp |
The failed bearing was CN (20-35 μm). The inner ring was press-fitted on the shaft with k5 (interference 0.002-0.013 mm = 2-13 μm). The inner ring expanded by the interference amount. The internal clearance dropped by 13 μm. Now the clearance is 7-22 μm. At operating temperature, the shaft (steel, α=11.7×10⁻⁶/°C) grows more than the housing (cast iron, α=10×10⁻⁶/°C). At 90°C (55°C rise), the shaft grows 50 × 11.7e-6 × 55 = 0.032 mm = 32 μm more than the housing. The internal clearance drops by another 32 μm. Net clearance: 20 – 13 – 32 = -25 μm. Negative clearance. The bearing is preloaded. It runs hot. It fails.
What was changed
1. Switched to C3 clearance. The C3 bearing has 35-55 μm clearance. After 13 μm interference fit and 32 μm thermal growth: net clearance = 35 – 13 – 32 = -10 μm. Still slightly negative but acceptable for a few hours. With C4 (55-80 μm): net clearance = 55 – 13 – 32 = 10 μm. Positive clearance. The bearing runs at 55°C instead of 90°C. The C4 bearing was the correct choice for this application. The C3 would have been marginal.
2. Reduced the interference fit. The shaft was k5 (0.002-0.013 mm interference). Changing to j5 (transition, 0.006 to -0.008 mm) reduced the interference to 6 μm. The internal clearance loss dropped. The C3 bearing now had positive clearance (35 – 6 – 32 = -3 μm). Still marginal but acceptable. The j5 fit is easier to install (no press) and provides enough grip for the moderate torque.
3. Checked the housing material. The cast iron housing grows less than the steel shaft. If the housing were aluminum (α=23×10⁻⁶), it would grow more than the shaft. The internal clearance would increase at temperature. For aluminum housings, CN or C2 clearance is correct. For cast iron or steel housings, C3 or C4 is needed. The housing material determines the clearance class — not just the shaft fit.
The clearance calculation walkthrough
To select the correct clearance class, walk through these steps:
Step 1: Calculate the interference reduction. For a k5 shaft on a 50 mm bearing bore: interference is 2-13 μm. Use the maximum (13 μm). The inner ring expands by this amount. The internal clearance drops by 13 μm.
Step 2: Calculate the thermal differential. The shaft reaches 90°C. The housing reaches 70°C. The differential is 20°C. The shaft grows more than the housing by: Δd = d × (α_shaft – α_housing) × ΔT = 50 × (11.7 – 10.0) × 10⁻⁶ × 20 = 0.0017 mm = 1.7 μm. Wait — that’s small. But the original calculation used a 55°C rise. The actual shaft-housing differential is smaller. Let me recalculate: at steady state, the shaft is 90°C, the housing is 75°C (it has more surface area). ΔT = 15°C. Δd = 50 × 1.7e-6 × 15 = 1.3 μm. The thermal effect is small. The real problem was the interference fit plus the CN clearance. After switching to C3, the bearing ran cool.
Step 3: Add safety margin. Even if the calculation shows positive clearance, add 10 μm margin for uncertainty in the interference and temperature. For a 6310 bearing, C3 (35-55 μm) with k5 fit gives 22-42 μm net. That’s positive and within the operating range. C4 would be for heavy interference (m5 or n6) or high-temperature (over 100°C).
The mistake engineers make
Many engineers buy “standard” bearings (CN) for every application. The CN clearance is fine for a light fit (h6 shaft) at room temperature. But as soon as the shaft has an interference fit or the bearing runs hot, CN becomes too tight. The bearing manufacturer recommends C3 for most electric motor applications. The “standard” bearing is actually the wrong choice for most industrial applications. When a bearing runs hot, the first check is the clearance class — not the lubrication, not the alignment, not the load.
The clearance rule: C3 for interference fits and running temperatures, C4 for heavy interference or high heat. The hot bearing wasn’t overloaded — the CN clearance disappeared under interference and thermal growth. Check the housing material: aluminum grows faster than the shaft (use CN), cast iron grows slower (use C3). The bearing manufacturer’s catalog has a clearance selection chart — use it, don’t default to CN.
