Bearing Vibration and Noise Testing to ISO 15242: Measuring Running Smoothness

Rolling bearings transmit load through rolling contact between rings and rolling elements, and any geometric imperfection, surface defect, or assembly error in these components produces vibration and audible noise during rotation. ISO 15242 defines a standardized method for quantifying this "running smoothness" by measuring vibration signals from a rotating bearing under controlled conditions. The test converts mechanical disturbance into velocity, acceleration, and peak signal levels, which can be compared against manufacturer or purchaser limits for acceptance. This article explains the underlying measurement principle, the signal parameters evaluated, specimen preparation and mounting, the procedural and instrumentation requirements, typical acceptance criteria, and the industrial applications in which bearing vibration and noise testing is applied alongside complementary examinations.

Principle & mechanism

Bearing vibration testing rests on the conversion of mechanical oscillation into an electrical signal by a transducer pressed against the rotating assembly. As the bearing turns under a defined radial or axial load, imperfections in the raceways, rolling elements, or cage excite periodic and random forces. These disturbances propagate through the outer ring to the transducer, which is typically a piezoelectric accelerometer or velocity sensor coupled through a mechanical loading arm. The sensor output is conditioned, filtered into defined frequency bands, and analyzed to yield band-limited vibration levels.
The mechanism links specific defect types to signal character. Waviness on raceways generates modulation at frequencies related to rotational speed and contact geometry. Surface roughness, contamination, and damaged rolling elements contribute broader-band energy and discrete peaks. Because the outer ring is held stationary under load in the standard configuration, its response is measured directly, giving a repeatable indication of the combined effect of component finish, geometry, and internal clearance. The method therefore assesses the integrated quality of manufacture and assembly rather than a single dimension.

Test Parameters: Velocity, Acceleration, and Peak Levels

ISO 15242 expresses bearing vibration through several complementary signal measures. The vibration velocity, usually quoted in micrometers per second as an RMS value, dominates low- and mid-frequency evaluation and correlates with perceived noise for many bearing types. Vibration acceleration, expressed in decibels or meters per second squared, weights higher-frequency content and responds to surface texture and localized defects. Peak values capture transient events such as debris indentation or rolling-element damage that RMS averaging may mask.
Each parameter is evaluated within specified frequency bands established by the measuring system. Spindle speed is fixed by the standard according to bearing size class, so that results remain comparable between laboratories. The applied radial or axial load, transducer type, and coupling force are likewise prescribed, because these variables shift the measured level. Test reports normally state the parameter type, band structure, and unit convention together with the measured values. Interpreting the set jointly is more informative than reading a single number: a bearing may pass a velocity limit while exceeding an acceleration limit, which points to a fine-surface or high-frequency cause rather than gross geometry error.

Sample Preparation and Mounting

Specimens must arrive clean and free of preservative that could damp or alter the vibration signal. The standard requires the bearing to be free of lubricant unless a defined lubricant quantity and type form part of the test condition, in which case the laboratory applies a stated grease or oil volume before measurement. Contaminants, residual acid from cleaning, or fingerprints on the raceways all raise measured levels and must be excluded through lint-free handling and controlled cleaning agents.
Mounting follows a defined sequence. The bearing inner ring is pressed onto a precision mandrel or spindle journal with a controlled interference fit, and the mandrel is clamped in the spindle with measured runout below a specified limit. The outer ring receives the prescribed load through the transducer pickup or a separate loading mechanism, oriented radially for radial bearings or axially for thrust bearings. The transducer is seated against the stationary outer ring at the designated location and force. Operators record the bearing designation, cleanliness steps, lubrication state, and mounting orientation, since any deviation from the prescribed fit or load invalidates comparison against catalog limits.

Measurement Procedure and Instrumentation

A typical measurement station comprises a vibration-isolated base, a variable- or fixed-speed spindle with low inherent vibration, a loading and transducer unit, signal conditioning filters, and analysis software computing RMS and peak values per band. The spindle background level must be verified periodically against a reference bearing or calibrated artifact so that the machine itself does not contribute to the result.
The procedure begins with a stabilization run that lets the lubricant distribute and the temperature settle. The measurement window is then triggered; during rotation at the standard speed, the system acquires the transducer signal, filters it into the defined frequency bands, and computes velocity and acceleration levels together with peak indicators. Multiple angular positions of the outer ring may be measured and averaged, because defect signals vary with load-zone orientation. The operator repeats the measurement to confirm stability and discards runs affected by external disturbance or electrical noise. Results are reported per band, per parameter, and per measurement point, along with spindle speed, applied load, lubrication condition, and ambient temperature, so the values can be traced to a defined state of the bearing.

Acceptance Criteria and Vibration Limits

Acceptance is a comparison exercise against limits that the parties agree in advance. Bearing manufacturers publish vibration classification levels per ISO 15242 for common series, grouping bearings into quietness grades according to band-limited velocity values. Purchasers in noise-sensitive sectors often specify a stricter grade or add acceleration and peak requirements on top of velocity limits. The test report therefore states which limit set was applied and whether each band and parameter passed.
Judging a result requires attention to context. A single band exceeding its limit while others pass suggests a specific frequency-related cause, such as raceway waviness at a particular order or a cage-related disturbance. Uniform elevation across all bands points to general surface finish, contamination, or inadequate lubricant distribution. Re-measurement after cleaning or re-lubrication can distinguish a specimen condition from a preparation artifact. When a bearing fails, the laboratory reports the failing parameter, band, and margin, allowing the manufacturer to trace the responsible process step. Acceptance decisions should always reference the exact spindle speed and band definitions under which the limits were established, since values from differing measurement conditions are not directly interchangeable.

Applications and Co-Testable Parameters

Vibration and noise testing to ISO 15242 supports several practical purposes across the supply chain. Manufacturers use it for batch release of quietness-graded bearings and for process control, tracking how grinding, honing, or washing changes shift band levels over time. Purchasers apply it at incoming inspection to verify that supplied lots meet the specified grade before assembly into noise-sensitive equipment. Failure analysts use band signatures to localize defects and separate manufacturing causes from transport damage or field contamination.
The test pairs naturally with complementary examinations on the same specimens. Geometric metrology of raceway waviness and roundness explains velocity-band findings at their source. Surface roughness measurement of raceways and rolling elements correlates with acceleration-band levels. Metallographic and hardness checks of bearing steel address material-related anomalies, and residual dirt analysis by gravimetric or microscopic methods identifies contamination contributions. Combining vibration results with these co-testable parameters lets the laboratory move from a pass or fail verdict toward a diagnosed cause, which is often the decisive output for quality disputes and process improvement programs.