Near race inclusions in bearing components and the resultant effect on fatigue initiation and component life

Document Type

Conference Proceeding

Publication Date



Steel cleanliness is of primary importance to the performance of tapered roller bearings in the rail industry, particularly under heavy axle load conditions. Non-metallic impurities present in these steels can manifest themselves as hard/brittle inclusions, which are detrimental to bearings in service due to the potential for developing sub-surface rolling contact fatigue (RCF). ASTM/ISO standards outline specifications for assessing the steel cleanliness based on bulk inclusion morphology, which limits inspection to a small area that may or may not be representative of the entire heat or section. Additionally, during manufacturing, there is the potential that these stray inclusions can end up being located at the near-race zone, which is the most susceptible to RCF failures due to the presence of high Hertzian contact fatigue stresses in this zone. Therefore, advanced methods for evaluating steel cleanliness are needed in today’s increasingly demanding railroad environment, and high (> 32.5-tonne) axle loads in many service environments. Previous published work by the authors had estimated steel cleanliness qualitatively from ultrasonic C-Scan images of entire bearing cups or cones and identified bearing quality based on bulk assessments of the overall inclusions present. An extension of this technique developed for this study is the identification and quantification of nearrace inclusions through an ultrasonic technique known as surface wave. For this study, bearing components were first subjected to ultrasonic surface wave testing to identify near surface defects, and then underwent simulated service life testing under heavy axle load conditions to assess the onset and propagation of RCF failure. Ultimately, the onset of RCF was then correlated back to the inclusions identified with ultrasonic surface wave scanning techniques. As a control, bearings produced from the same steel heat with similar inclusion content were used in order to maintain objectivity throughout the experiment, with samples from near-surface defect free components, and those with ultrasonic signatures indicative of near-surface defects being examined. The locations of fatigue failures were then superimposed with the ultrasonic scanning results to study the effectiveness of this examination and its predictive capabilities. The results clearly demonstrate the susceptibility of bearings to early RCF failure when near-race surface defects are present.

Publication Title

CORE 2016: Maintaining the Momentum