Dynamic Multibody Modeling of Spherical Roller Bearings with Localized Defects for Large-Scale Rotating Machinery

Area of Science:

• Mechanical Engineering
• Vibrational Analysis
• Condition Monitoring

Background:

• Early fault detection in rotating machinery is vital for industrial maintenance and cost reduction.
• Spherical bearings are critical components in many medium-to-large-scale industrial applications.
• Accurate simulation models are needed for effective predictive maintenance strategies.

Purpose of the Study:

• To develop and validate a multibody dynamic model of medium-sized spherical bearings.
• To assess the model's capability in identifying characteristic fault frequencies.
• To generate datasets for training diagnostic and prognostic algorithms.

Main Methods:

• Development of a multibody model in Simulink Simscape.
• Simulation of bearing dynamic behavior, including six degrees of freedom.
• Validation against experimental measurements from a heavy radial load test rig.

Main Results:

• The model accurately reproduced experimental signals, identifying characteristic fault frequencies.
• Simulated and experimental signals showed a good fit, confirming model validity.
• Localized defects on inner and outer races were effectively detected through vibration signatures.

Conclusions:

• The validated model is a promising tool for generating fault data for predictive maintenance.
• The model aids in improving diagnostic and prognostic algorithms for industrial machinery.
• It supports condition monitoring strategies despite minor amplitude discrepancies due to model simplifications.