Instant gear contact can be simulated with contacting discs, which provides steady operating conditions and eliminates most of the dynamics and manufacturing tolerances involved in real gears, resulting in an accurately controlled contact condition. A high-pressure twin-disc test device was developed, where loading and rolling velocity can be varied continuously. It is equipped with disc bulk temperature, mean contact resistance and friction moment measurements. The test discs were grinded transversal to the disc rolling direction with proper crowning corresponding to the real gear flank properties. The test device was applied by studying the friction behaviour against the slide-to-roll ratio at different contact pressures, rolling velocities and surface roughness. The measurements were performed using mineral base oil in the range of operation conditions often used in industrial gears. In general, the measured friction coefficient behaviour correlates with earlier published results and is logical with measured bulk temperature and mean contact resistance. The limiting shear stress of the lubricant has an essential role in friction behaviour.
Lubrication conditions have a major effect on lifetime and friction in gear contacts. Monitoring of relative changes in protecting film thickness provides an opportunity to understand the operating conditions before an initial fault appears. In this study, the lubrication conditions were detected on-line using contact resistance and bulk temperature measurements, which were applied to a modified FZG gear test device. Measurements were made with polished gear surfaces at operating conditions that are typical of industrial gears. The presented mean contact resistance includes the data points in different positions along the line of action and thus also considers transient effects. The bulk temperature and trend curves of the mean contact resistance at different pitch line velocities, loads, and oil inlet temperatures are reported and discussed as measures for analysing the lubrication condition.
High-pressure lubricant properties are important when friction coefficients and power losses are evaluated in elastohydrodynamic (EHL) contacts. An approach was developed to determine the limiting shear stress and actual viscosity properties of lubricants using a numerical traction model based on elliptical EHL contact and traction curves, measured at a wide range of temperatures and pressures with a twin-disc test device. The tests were carried out in pure fluid film at high Hertzian pressures with finely polished surfaces. Each lubricant was tested at 135 test points, where traction coefficients and bulk temperatures were measured. The lubricant parameters in the traction model were adjusted so that the calculated results matched the experimental measurements at all test points. In general, there is a good correlation between the calculated results and the experimental measurements. The influence of pressure on limiting shear stress and actual viscosity is less significant for polyalphaolefin (PAO) oil than for mineral oil.
A wind turbine is equipped with lots of sensors whose measurements are recorded by the supervisory control and data acquisition (SCADA) system and stored every 10 minutes. The pitch subsystem of a wind turbine is of critical importance as it presents the highest failure rate. Thus, selecting the most essential features from the SCADA system is performed in order to detect faults efficiently. In this study, a feature space of 49 features is available, referring to the condition of a hydraulic pitch system. The dimensionality of this feature space (original input space) is reduced using a Deep Autoencoder in order to extract latent information. The architecture of the Autoencoder is investigated regarding its efficiency on fault detection task. This way, effect of new extracted features on the performance of the classifier is presented. A Support Vector Machine (SVM) classifier is trained using a set of healthy (fault free) and faulty data, representing different kind of pitch system failures. The data are acquired from a wind farm of five 2.3MW fixed-speed wind turbines. The performance metric used to evaluate their effect on data is F1-score. Results show that SVM using new extracted feature by Autoencoder outperforms SVM classifier using the original feature set, underlining the power of Autoencoders to unveil latent information.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.