Ball bearing vibration data for detecting and quantifying spall faults.

Area of Science:

• Mechanical Engineering
• Reliability Engineering
• Data Science

Background:

• Ball bearings are critical components in electromechanical systems, particularly in energy and aerospace sectors.
• Bearing failures can significantly reduce system service life, necessitating early fault detection and quantification for safety-critical applications.
• Vibration signals from bearing fatigue faults (spalls) exhibit complex excitation mechanisms influenced by fault size and operating conditions.

Purpose of the Study:

• To provide a comprehensive dataset of vibration measurements for ball bearings with artificially seeded fatigue faults.
• To characterize vibration excitation mechanisms associated with different fault sizes and operating conditions.
• To facilitate the validation of condition monitoring techniques for industrial and aerospace applications.

Main Methods:

• Artificial seeding of realistic fatigue faults (spalls) on bearing races using a precise machining process.
• Acquisition of vibration data under various controlled operating conditions and fault sizes.
• Detailed documentation of datasets, including fault characteristics and operating parameters.

Main Results:

• The dataset captures diverse vibration signatures corresponding to different fault sizes and operating conditions.
• Identified common vibration excitation mechanisms inherent to bearing spalls.
• The data provides a basis for analyzing the relationship between fault characteristics and vibration signal behavior.

Conclusions:

• The provided vibration datasets are valuable for researchers and engineers in condition monitoring.
• Understanding vibration signal characteristics under various fault conditions is crucial for reliable electromechanical systems.
• This data facilitates the development and validation of advanced fault detection and quantification methods for bearings.