No abstract

The formation of white etching cracks in the 1 mm zone beneath the contact surface in steel rolling element bearings causes a premature wear failure mode called white structure flaking. The formation drivers of white etching cracks are contested, as are the initiation and propagation mechanisms of the cracks. Hydrogen diffusion into bearing steel sourced from the hydrocarbon lubricant or water contamination and transient operating conditions have been suggested as formation drivers. Extensive work has been conducted at Southampton to further understanding of white structure flaking and this paper summarises these evidences and the conclusions made. Serial sectioning has been used to map subsurface wear volumes of wind turbine gearbox bearings from service and large-scale test rigs, test specimens/bearings from laboratory under hydrogen charged conditions and non-hydrogen charged conditions. The process involves polishing of cross sections of test specimens/bearings at $3-5 mm material removal intervals typically over hundreds of slices, and this was used to map white etching cracks in their entirety for the first time. Serial sectioning has allowed a comprehensive investigation of the initiation and propagation mechanisms of white etching cracks and thresholds for their formation with respects to concentration of diffusible hydrogen, contact pressure and number of rolling cycles. From these studies it has been found that white etching cracks can form by subsurface crack initiation at inclusions under hydrogen charged and non-hydrogen charged conditions; hence it has been confirmed that this is one mechanism of WEC formation. Small/short sized sulfide inclusions, globular manganese sulfide þ oxide inclusions and small globular oxide inclusions between $1 mm and 20 mm in diameter/length predominated as crack initiators. In addition, detailed focused ion beam/transmission electron microscopic studies have been conducted to enhance the understanding of butterfly crack and white etching area formation mechanisms.

In this article, electrostatic charge sensing technology has been used to monitor adhesive wear in oil-lubricated contacts. Previous work in this area using FZG gear wear rig and pin-on-disc tribometers demonstrated that 'precursor' charge events may be detected prior to the onset of scuffing. Possible charging mechanisms associated with the precursor events were identified as tribocharging, surface charge variation, exo-emissions, and debris generation. This article details tests carried out to investigate the contribution of wear debris. Tests were carried out on a modified pin-on-disc rig using a sliding point contact and fitted with electrostatic sensors, one of which monitored the disc wear track and the other the disc surface just outside the wear track. Baseline tests used mild wear conditions with no seeded particles added to the entrained lubricant, whereas the high wear tests entrained seeded steel particles into the contact to promote wear. The wear debris produced dynamic charge features and the overall charging activities are directly related to the wear rate (i.e. charging levels increase with increasing wear rate). There appears to be a link between the net volume loss and the charge levels, relating charge directly to the increasing rate of debris production. Wear debris due to natural wear produced positive dynamic charge features, whereas debris from the seeded tests produced negative dynamic charge features. The polarity of the charge on debris is thought to depend on which charging and wear mechanism is predominant.

Online fault diagnostic technologies are fast emerging for detection of incipient faults on tribological components to avoid catastrophic failure. Vibration analysis has long been used to detect machine faults, but is sensitive to relatively severe conditions only. Electrostatic monitoring is a newly developed approach with the potential to detect precursor processes that indicate contact distress and wear. Recently, at the University of Southampton, both vibration and electrostatic sensors were implemented on a bearing testing rig to evaluate their effectiveness in detecting bearing faults. The results indicate that both types of sensor are sensitive to bearing deterioration shortly before complete failure. However, univariate plots of signals from both types of sensor only exhibit significant change when entering the severe wear stage. Therefore, multivariate techniques for detecting wear severity of components at different running stages need investigating. In this study, an unsupervised training method, called mixture-model-based clustering, that utilizes the expectation maximization (EM) algorithm is employed to develop further a wear detection technique. The choice and extraction of significant features from both vibration and electrostatic sensors are discussed as step one. The second step uses the clustering method to examine the behaviour of the extracted features during different running stages, and to quantify how good the sensors are at distinguishing wear severity. In the third step, a dynamic wear detection process is simulated. Clustering is applied to baseline data from a known healthy bearing and data from different wear stages to see if the data naturally group by wear condition. The result shows that the unsupervised clustering method is able not only to learn and detect wear conditions of the rolling element bearings with the developed statistical monitoring charts of occupation probability (OP) in the clusters and number of the trained clusters (NC), but also to obtain the advantage of detecting insignificant abnormalities that might be overlooked in the conventional plots.

No abstract