Why Bearing Selection Is Worth Getting Right
Bearing failures are among the most common and most preventable causes of machine downtime. A bearing that is correctly selected for its load, speed, and environment will run for its calculated design life with minimal maintenance. One that is undersized, misapplied, or running in the wrong environment will fail well before that — sometimes catastrophically. After years of seeing bearing failures in the field, the same root causes appear repeatedly: loads were underestimated, speed limits were ignored, or the selection was made by copying a previous design without checking whether the conditions had changed.
This guide covers the practical selection process: understanding the loads, applying the L10 life equation, and choosing between the main bearing types for your application.
Step 1: Define the Loads
A bearing sees two fundamental types of load: radial load (perpendicular to the shaft axis) and axial load (along the shaft axis). Most bearings experience a combination of both, and the selection process must account for both components.
Radial Load Sources
Radial loads come from the weight of components, belt and chain tension, gear tooth forces, and centrifugal effects. For a shaft with multiple radial loads at different positions, calculate the reactions at each bearing using static equilibrium (sum of forces and moments). Use the maximum radial load at each bearing position.
Axial Load Sources
Axial loads arise from helical gear thrust, worm gear thrust, inclined belt systems, weight of vertically oriented components, and intentional preload in precision systems. If a bearing must carry axial load, this must be explicitly accounted for in the bearing selection — deep groove ball bearings can handle moderate axial loads; angular contact and tapered roller bearings are designed for combined loading.
Dynamic Equivalent Load
For bearings carrying combined radial and axial loads, calculate the dynamic equivalent radial load P:
P = X × Fr + Y × Fa
where Fr is radial load, Fa is axial load, and X and Y are load factors from the bearing manufacturer’s tables (dependent on bearing type and the ratio Fa/Fr). For pure radial loading, P = Fr.
Step 2: Assess Speed — The n·dm Factor
Bearing speed capability is characterized by the n·dm factor (also called the speed parameter), where n is the rotational speed in rpm and dm is the mean bearing diameter in mm ((bore + OD)/2). This factor captures the surface velocity at the rolling contact.
Manufacturer catalogs list a limiting speed and a reference speed for each bearing designation. The reference speed assumes normal lubrication; the limiting speed is an absolute maximum. Operating above the limiting speed causes excessive heat generation, accelerated wear, and potential seizure. At high n·dm values, precision bearings with optimized cage designs and specific lubrication are required.
For grease-lubricated bearings, a conservative upper limit for standard deep groove ball bearings is roughly n·dm = 300,000 to 500,000 mm·rpm, depending on bearing series and grease type. Oil lubrication allows higher speeds.
Step 3: Calculate L10 Basic Rating Life
The industry-standard bearing life equation calculates the L10 life — the life in millions of revolutions that 90% of a group of identical bearings will reach or exceed under the given loading condition:
L10 = (C / P)p
where:
- C = basic dynamic load rating (from catalog, in kN)
- P = dynamic equivalent load (calculated above, in kN)
- p = 3 for ball bearings, 10/3 for roller bearings
To convert L10 from millions of revolutions to operating hours:
L10h = (L10 × 106) / (60 × n)
where n is speed in rpm. For a design target of 20,000 hours at 1,500 rpm, calculate the required L10h and work backwards to determine the required C/P ratio, then select a bearing from the catalog where C ≥ P × (L10 required)^(1/p).
Bearing Type Selection: Matching Type to Application
| Bearing Type | Radial Load | Axial Load | Speed | Typical Application |
|---|---|---|---|---|
| Deep groove ball bearing | Good | Moderate (both directions) | High | Electric motors, gearboxes, pumps |
| Angular contact ball bearing | Good | High (one direction) | High | Spindles, paired axial loading |
| Cylindrical roller bearing | Very high | None (N/NU types) | Moderate-high | Heavily loaded shafts, gearbox pinions |
| Tapered roller bearing | High | High (one direction) | Moderate | Wheel hubs, bevel gear shafts |
| Spherical roller bearing | Very high | Moderate | Moderate | Misalignment-prone shafts, conveyor rolls |
| Thrust ball bearing | None | Moderate | Moderate | Vertical shaft applications |
Seal and Shield Options
For most industrial applications, bearings with integral seals or shields simplify installation and maintenance. Shields (designation suffix Z or ZZ) are pressed steel covers that block large particles but are not contact seals — they do not retain grease as effectively. Seals (suffix RS or 2RS) use rubber-lipped contact seals that retain grease for life and exclude contaminants effectively, at the cost of slightly higher friction. For wet or dirty environments, contact-sealed bearings are almost always the better choice.
FAQ
Q: What happens if I use a bearing rated for a lower load than the actual application load?
A: An undersized bearing will have a calculated L10 life shorter than your design target. In practice, the bearing will fail early — the rolling elements, races, or cage will show accelerated fatigue damage. The failure mode is typically spalling (material flaking from the race surface), accompanied by increased vibration and noise. In high-load applications, the failure can be sudden rather than gradual.
Q: Should I always aim for the highest possible bearing life in my calculation?
A: Not necessarily. Over-sizing bearings wastes money and space, and very large bearings running at low load fractions can suffer from skidding (rolling elements sliding rather than rolling), which causes its own wear. Aim for a bearing life that matches your maintenance interval and machine design life, typically with a moderate safety margin. A factor of 1.2–1.5 above the required life is common practice.
Q: How do I account for shock loads in the L10 calculation?
A: Apply a service factor (also called application factor or load factor, designated Ka or fd) to multiply the calculated load P before applying the life equation. Typical service factors are 1.0 for smooth operation, 1.2–1.5 for moderate shock, and 1.5–3.0 for heavy shock loads. The specific values are given in bearing manufacturer handbooks for different application categories.
コメント