Glossary term

Bearing Preload

Intentional load or interference applied to a bearing to control internal clearance, stiffness, position, vibration and running accuracy.

Branch: Mechanical Engineering
Glossary type: quantity
Content: Glossary term
Updated: Jun 25, 2026
Revision: v1.0.0 · reviewed

Definition

quantity

Bearing preload is an intentional internal load or displacement applied to a bearing to remove clearance, control stiffness or set position before external service loads are added.

Bearing preload can be created by springs, spacers, locknuts, shims, matched bearing sets, interference fits, thermal assembly or adjusted axial position. Correct preload can improve stiffness, running accuracy, gear mesh control, spindle precision and vibration behavior. Excess preload can collapse clearance, increase heat, reduce lubricant film, shorten life, damage races and make a machine fail after thermal soak.

Bearing preload is an intentional load or displacement applied inside a bearing arrangement before external service loads are added. The purpose is to control internal clearance, shaft position, stiffness, running accuracy, gear mesh, spindle precision or vibration behavior.

Preload is useful only when it is controlled. Too little preload can leave looseness, poor positioning, skidding, cage instability or gear mesh variation. Too much preload can create heat, lubricant-film loss, race stress, early fatigue, false brinelling during standby, motor overload or bearing seizure after thermal soak.

Engineering Role

Bearings are not only catalog load-carrying parts. In a machine they set the location and dynamic boundary condition of shafts, rotors, gears, screws, spindles and wheels. Preload changes:

internal clearance and shaft position;
bearing stiffness and damping;
natural frequency and critical-speed margin;
friction heat and lubricant film thickness;
sensitivity to thermal growth and fit variation;
bearing life under combined preload and service load.

The right preload depends on bearing type, speed, temperature, lubrication, fit, duty cycle, accuracy requirement, installation method and failure consequence.

Preload and Deflection

A simple stiffness screen writes preload as:

F_p=k_b\delta_p

where $F_p$ is preload force, $k_b$ is bearing or bearing-pair stiffness in the loaded direction, and $\delta_p$ is imposed elastic displacement.

Example: a bearing pair has an approximate axial stiffness:

k_b=45\ \text{kN/mm}

The assembly spacer imposes:

\delta_p=0.025\ \text{mm}

The preload screen is:

F_p=45(0.025)=1.125\ \text{kN}

That value must be compared with the bearing supplier setting range and the thermal condition. A preload that is acceptable cold may become excessive when the shaft, inner ring or housing temperature changes.

Operating Clearance

For a simplified clearance budget:

C_{op}=C_0-\Delta C_{fit}-\Delta C_T-\Delta C_p

where $C_0$ is initial internal clearance, $\Delta C_{fit}$ is clearance loss from fits, $\Delta C_T$ is thermal clearance loss and $\Delta C_p$ is the clearance removed by preload adjustment.

If:

C_0=40\ \mu\text{m}

\Delta C_{fit}=12\ \mu\text{m}

\Delta C_T=18\ \mu\text{m}

\Delta C_p=15\ \mu\text{m}

then:

C_{op}=40-12-18-15=-5\ \mu\text{m}

The negative value means the simplified screen has entered preload. That may be intended for some angular-contact or tapered-roller arrangements, but it is not automatically safe. The heat, load, speed and bearing setting method must agree with that condition.

Thermal Effect

When the inner ring or shaft runs hotter than the housing, radial clearance can reduce. A rough differential-expansion screen is:

\Delta C_T\approx D(\alpha_i\Delta T_i-\alpha_o\Delta T_o)

where $D$ is representative bearing diameter, $\alpha_i$ and $\alpha_o$ are expansion coefficients for inner and outer paths, and $\Delta T_i$ and $\Delta T_o$ are their temperature changes.

For similar steel paths with the inner path $25\ \text{K}$ hotter than the outer path:

D=60\ \text{mm}

\alpha=12\times10^{-6}\ /\text{K}

the clearance loss screen is:

\Delta C_T=60(12\times10^{-6})(25)=0.018\ \text{mm}=18\ \mu\text{m}

This is enough to change a lightly cleared bearing into a preloaded one. That is why hot checks and temperature trends matter during release.

Friction Heat Screen

A simple preload-related friction torque screen is:

\displaystyle M_f\approx fF_p\frac{d_m}{2}

where $f$ is an empirical friction factor, $F_p$ is preload and $d_m$ is mean bearing diameter. Friction power is:

P_f=M_f\Omega

If:

f=0.002

F_p=1.5\ \text{kN}

d_m=60\ \text{mm}=0.060\ \text{m}

then:

M_f=0.002(1500)(0.030)=0.090\ \text{N m}

At:

n=3600\ \text{rpm}

\displaystyle \Omega=\frac{2\pi n}{60}=377\ \text{rad/s}

so:

P_f=0.090(377)=34\ \text{W}

If preload rises to $5.0\ \text{kN}$ after thermal growth or incorrect adjustment, the same screen gives about $113\ \text{W}$ . That extra heat may reduce viscosity, shrink clearance further and create a self-reinforcing problem.

Bearing Life Screen

Preload is part of the load seen by the bearing. A simplified equivalent-load screen is:

P_{eff}=P_s+F_p

where $P_s$ is service equivalent load. If the original load is:

P_s=3.0\ \text{kN}

and preload adds:

F_p=1.0\ \text{kN}

then a simplified ball-bearing life ratio is:

\displaystyle \frac{L_2}{L_1}=\left(\frac{3.0}{4.0}\right)^3=0.422

The screen says life could fall to about $42\%$ of the no-preload load case. Real rating calculations are more detailed, but the direction is clear: preload must be intentional and documented.

Diagnostic Evidence

Excess preload may show rising bearing temperature, increasing current draw, high-frequency vibration, poor coast-down, grease purge, discoloration, race distress or a temperature rise after thermal soak. Under-preload or excessive clearance may show looseness, cage noise, unstable position, poor gear mesh, variable runout, vibration that changes with load or poor repeatability.

The strongest diagnosis combines installation records, bearing type, fit calculation, spacer or nut setting, torque record, temperature trend, vibration trend, lubricant condition and inspection evidence. A spectrum alone is not enough.

Validation and Release

A defensible preload review states bearing type, arrangement, clearance class, fit, spacer stack, adjustment method, preload target, supplier limit, lubrication, speed, temperature, thermal growth assumption, measurement method, uncertainty and release criterion.

Release should be withheld when the setting method is undocumented, temperature continues to climb, preload changes after thermal soak, bearing current or torque is high, vibration changes after adjustment, fit calculations consume most clearance or there is no evidence that the installed bearing matches the specified clearance class.

Common Mistakes

Do not treat preload as a way to hide poor alignment or loose fits. Preload can mask movement while increasing heat and stress.

Do not use room-temperature clearance as the operating condition. Fits, shaft temperature, housing temperature and lubricant temperature can change clearance enough to reverse the decision.

Do not assume more preload means more precision. Precision machines need controlled preload, not maximum preload.

Limits

The equations above are screening tools. Real bearing preload depends on contact angle, rolling element geometry, race conformity, lubricant film, centrifugal effects, fits, housing stiffness, thermal gradients, mounting sequence and supplier-specific rating methods.

The practical goal is to prove that the installed bearing has the intended clearance or preload in the condition that matters, with acceptable heat, vibration, life margin and repeatable release evidence.

REF

Disciplines