Toward a 3D dynamic model of a faulty duplex ball bearing

• A new 3D ball-bearing model is presented and validated.
• A deformed outer race of a single axially loaded bearing is simulated.
• The model results are subjected to a sensitivity study.
• Duplex bearing with deformed outer race is simulated as well as tested.
• The simulation results are in a good agreement with the experimental results.

Bearings are vital components for safe and proper operation of machinery. Increasing efficiency of bearing diagnostics usually requires training of health and usage monitoring systems via expensive and time-consuming ground calibration tests. The main goal of this research, therefore, is to improve bearing dynamics modeling tools in order to reduce the time and budget needed to implement the health and usage monitoring approach.The proposed three-dimensional ball bearing dynamic model is based on the classic dynamic and kinematic equations. Interactions between the bodies are simulated using non-linear springs combined with dampers described by Hertz-type contact relation. The force friction is simulated using the hyperbolic-tangent function. The model allows simulation of a wide range of mechanical faults. It is validated by comparison to known bearing behavior and to experimental results. The model results are verified by demonstrating numerical convergence.The model results for the two cases of single and duplex angular ball bearings with axial deformation in the outer ring are presented. The qualitative investigation provides insight into bearing dynamics, the sensitivity study generalizes the qualitative findings for similar cases, and the comparison to the test results validates model reliability.The article demonstrates the variety of the cases that the 3D bearing model can simulate and the findings to which it may lead. The research allowed the identification of new patterns generated by single and duplex bearings with axially deformed outer race. It also enlightened the difference between single and duplex bearing manifestation. In the current research the dynamic model enabled better understanding of the physical behavior of the faulted bearings. Therefore, it is expected that the modeling approach has the potential to simplify and improve the development process of diagnostic algorithms.