Ball bearing breakaway torque must be rated as a maximum startup resistance, not treated as the constant running torque; published limits range from 1,400 to 11,000 mg/mm across specified miniature bearing sizes and fit classes [S1].
The rating applies to precision instruments, motors, positioning mechanisms, and other low-power drives where the available starting torque can be consumed by bearing friction, seal drag, preload, or cold lubricant behavior [S1][S3][S5].
Breakaway torque differs from running torque
Breakaway torque is the frictional moment that must be overcome to move a stationary bearing, while running torque is the torque needed to sustain rotation at a constant speed [S1][S5].
A drive can start a bearing successfully and still run below its expected efficiency if the running torque exceeds the actuator budget after startup. The specification should therefore separate the maximum stationary starting torque from the allowable running torque, rather than applying one value to both conditions [S1][S5].
For a general ball bearing, the calculated drive requirement must include the shaft, coupling, seals, lubricant, and any mechanism load reflected to the shaft. Precision-component designers are expected to compare total system torque input with the torque reflected by the driven components [S1].
Published limits show how size and fit change the rating
NHBB lists maximum starting torque under a defined axial test load and divides the limits into three internal radial-play classes: tight fit at 0.0001–0.0003 inch, normal fit at 0.0002–0.0005 inch, and loose fit at 0.0005–0.0008 inch [S1].
For SSRI-3, the maximum starting torque is 1,800 mg/mm at a 75 g thrust test load for tight fit, 1,500 mg/mm for normal fit, and 1,400 mg/mm for loose fit. For the larger SSRI-4, the maximum starting-torque limits are 1,800 mg/mm for tight fit, 1,500 mg/mm for normal fit, and 1,400 mg/mm for loose fit under a 75 g thrust test load [S1].
The contrast is material. The SSRI-2, SSRI-2 1/2, SSRI-3, and SSRI-4 use a 75 g test load and specify maximum starting torques of 1,800 mg/mm for tight fit, 1,500 mg/mm for normal fit, and 1,400 mg/mm for loose fit [S1]. These figures are product-specific limits, not universal ratings that can be transferred to an unrelated bearing.
Selection criteria for a usable startup rating
A valid rating needs at least three inputs: bearing size, internal fit or radial play, and preload or axial test load [S1]. NHBB's table lists bearing size, radial-play fit classes, test load, and maximum starting torque for its listed part numbers, while an unlisted bearing requires consultation with the factory [S1].
Fit is a direct selection variable because each NHBB size has a lower maximum for loose fit than for tight fit. In the listed SSRI-2 family, the maximum falls from 1,800 to 1,500 and 1,400 mg/mm as radial play moves from tight to normal and loose [S1].
Preload must also be tied to the operating configuration. Acceptance testing measures torque, preload, runout, stiffness, and other values to confirm whether they are within expected limits [S3].
That preload is suitable for controlling stiffness, but it may raise startup resistance. For a preloaded angular contact bearing, compare the selected preload and lubrication package against the actuator's breakaway capacity rather than relying on catalog load rating alone.
Temperature and lubricant can dominate cold startup
Starting torque was measured at low temperatures to characterize cold startup [S4]. Tests on three bearings covered temperatures from 0°C down to -90°C and -100°C, and the bearings were of the type intended for deployment in the REASON antennas for the Europa Clipper mission [S4].
Lubricant viscosity, seal design, rolling-element deformation, and raceway sliding all affect torque. Higher-viscosity lubricants increase resistance, and contact seals add drag torque that depends on seal design, material, and condition [S5].
Cold, shielded, and lightly preloaded options require different evidence
Open or lightly shielded bearings are not directly comparable with sealed or contact-sealed versions because the seal adds a separate drag component. The rating basis should identify whether the measured or calculated value includes seals, grease, cage, preload, and the load applied during startup [S1][S5].
Preloaded bearings can offer greater stiffness and repeatability, while a looser internal fit can lower the published starting-torque limit. The trade is not simply “preload is good” or “clearance is good”: it is a balance among startup torque, stiffness, service life, speed, and lubricant behavior [S1][S3][S5].
For cold service, the relevant comparison is torque capability versus the worst specified temperature. In the Europa Clipper deployment example, starting torque was measured for three bearing types from 0°C down to -90°C and -100°C [S4].
Acceptance testing closes the catalog-to-unit gap
Acceptance testing should measure the torque of the actual assembled bearing because catalog performance values are nominal averages with tolerances, and unit-to-unit variation is common [S3]. A torque trace can verify both starting and running torque against the design goal before shipment [S3].
Testing can also cover preload, runout, and stiffness. A U.S. Army tank fire-control example used two 3.0-inch inside-diameter angular contact bearings; the standard bearings met lubrication and preload requirements, but their starting torque exceeded the available rotation power after assembly [S3].
The remedy in that example was rework and redesign of the lubrication and preload requirements. A startup-torque acceptance limit would have exposed the mismatch before final system testing, making test-based angular contact bearing startup control a practical engineering response to that failure mode.
Procurement specification and acceptance plan
Specify the maximum breakaway torque at a stated temperature, axial test load, radial fit or internal clearance, lubrication condition, preload, and seal configuration. Retain separate limits for starting torque and running torque, because one value does not represent both operating phases [S1][S5].
Require a torque trace for critical or low-power mechanisms and define the measurement fixture, speed threshold for separating breakaway from running torque, sampling interval, environmental temperature, and pass or fail criterion. Kaydon identifies torque, preload, runout, and stiffness as measurable acceptance characteristics [S3].
For low-temperature service, add measured data at the minimum operating temperature and define lubricant type and conditioning. For an unlisted bearing size, obtain a manufacturer-specific limit rather than extrapolating the NHBB table from a neighboring size [S1][S4].
The next procurement checkpoint is a traceable specification containing one maximum startup torque, one running-torque criterion, the test load and temperature, and the fitted bearing part number. Any change to radial play, preload, lubricant, or seal should trigger revalidation because those variables directly change startup resistance [S1][S5].