When bearings must perform with high reliably or within narrow limits, bearing acceptance testing may be needed
In critical bearing applications, you may need to know exactly how the bearings will perform before they go into service. This is especially true for applications such as machine tools, robotics, and electronic equipment where a bearing’s functional characteristics are the key to meeting system requirements. In such cases, bearing acceptance tests can provide the needed information.
Reasons to test
Engineering data or ratings in catalogs give a sufficiently accurate indication of bearing performance for most applications. But this data represents the estimated functional characteristics of bearings, based on calculations and nominal dimensions. By contrast, the actual performance varies from catalog values because of normal tolerance variations and unknown load and mounting influences. The degree of variation depends on the combined effect of bearing component tolerances after the bearing is assembled, and in some cases, mounted to the shaft and housing.
Moreover, bearing inspections verify that dimensions are correct, but they don’t provide any information about performance characteristics such as torque, stiffness, or accuracy. Acceptance testing eliminates uncertainties about performance by measuring key performance characteristics that are most important to the equipment designer. If any of these values don’t match the prescribed operating limits, it may be possible to rework components to bring the assembled bearing within the required limits. Or the test may indicate that the design can’t meet the application requirements, in which case, design options must be investigated.
Acceptance testing offers an added dimension to quality control, providing assurance that the bearings will operate as expected. This testing also reduces unexpected downtime to repair components that do not meet system requirements.
When to test
Generally, you should consider acceptance testing when:
• The end product must perform with high reliability, as in space-based applications.
• The end product is very costly.
• Failure poses a life-threatening situation.
• Repair is costly, difficult, or impossible, as with a space satellite.
• Comprehensive documentation must accompany the product. Such documentation can include measured dimensions, torque and stiffness values, and process certification.
• The equipment supplier must provide detailed accountability or an audit trail.
More specifically, you may need tests to verify performance in applications that impose narrow performance limits or conflicting requirements. For example, bearings are often preloaded to increase stiffness and reduce deflection, but preload increases the required starting and running torque. In low-power applications, this increased torque may exceed that available from the power source.
Other situations that require tests involve special types of bearings in unusual applications. For example, thin section bearings are often called upon to operate at low power with slowly rotating, highly loaded mechanisms — or with highspeed mechanisms that must start and stop quickly. In such cases, operating characteristics such as stiffness, accuracy, and torque are critical, but may represent conflicting requirements.
The following steps will help you determine if acceptance testing is needed:
1. Define the key system requirements such as bearing starting and running torques, stiffness, and accuracy, plus operating conditions such as temperature and available power. In some cases, the bearing’s functional characteristics can make or break the system’s performance.
A mounting drawing is helpful at this stage so that improper mounting methods, which may degrade performance, can be identified and corrected.
2. Ask the bearing manufacturer for an analysis that predicts bearing performance values under operating conditions similar to yours. If the predicted values approach the system limits, these attributes are candidates for acceptance testing.
Test requirements may range from a simple check of starting torque to a comprehensive set of measurements, plus QC documentation, including metallurgical certifications and heat treat records. The following elements are often included in a comprehensive acceptance testing plan:
• Traceability of parts by serial number.
• Runouts of assembled bearings. Measure radial and axial runouts with races either unrestrained (free-state) or mounted in shaft and housing fixtures.
• Face offset. This is the axial offset between outboard bearing faces.
• Internal fit — preload or clearance.
• Dimensional summary, which includes bore and outside diameters, component runout, and race widths.
• Load-deflection curves (axial and radial stiffness). Measure these parameters with preload applied and with races either unrestrained (free-state) or mounted in shaft and housing fixtures. The unrestrained condition is worse because the bearing lacks support and therefore may deflect more.
• Friction torque. Measure with preload applied and races either unrestrained or mounted in housing fixtures. Here, the mounted condition is worse because shaft and housing fit can increase the torque.
• Verification of lubricant type and quality.
You can request bearing acceptance testing in the form of a QC instruction, purchase order amendment, or a drawing requirement.
For complex acceptance tests, the bearing manufacturer and user may need to jointly establish a formal test plan, detailing the bearing functional characteristics that are important to the success of the application, plus the test procedures and acceptance criteria. This is especially true for very critical applications, such as satellites or missile guidance systems, where comprehensive testing is essential. In other cases, the manufacturer and user may agree on a suitable Acceptance Quality Level (AQL) from MILSTD- 105, which provides statistical test plans for consideration.
Where friction torque and stiffness must be measured, shaft and housing conditions can greatly alter the test results. Housing deviations may distort bearings, leading to erratic torque. Or a tight fit may reduce clearance and increase preload, leading to higher torque. For such tests, be sure to carefully define the operating conditions to be simulated.
In some cases, values obtained through unrestrained tests may provide adequate assurance if they can be correlated with prototype test values. Otherwise, these tests may require special fixtures to simulate the effect of the mounting arrangement. This provides greater confidence that the unit will function as expected.
Precision bearings for critical applications are typically assembled, lubricated, inspected, and tested in bearing manufacturer’s clean rooms. Cleanliness inspections involve counting microscopic contaminants from ball and race surfaces. If the count exceeds a predetermined limit, the bearing is recleaned and inspected again. Torque tests in particular are sensitive to contamination as well as shaft and housing fits. Therefore, manufacturers usually conduct such tests in a clean room. Simple checks, such as bearing runout, can be readily performed when the bearing manufacturer is notified early in the design process. But the most complicated procedures may add several days of inspection and documentation to the bearing assembly time.
The bearing manufacturer may be able to supply bearings in mounted subassemblies, which eliminates the uncertainty involved when the end user purchases separate components and then assembles them.
Michael Purchase is vice president-engineering, Kaydon Corp., Muskegon, Michigan.