Proceedings of the Institution of Mechanical Engineers,Part J: Journal of Engineering Tribology

Elbows are widely used in high-pressure pipe manifolds but are prone to severe erosion by solid particles, threatening safety and reducing pipeline service life. The blind tee can enhance the erosion resistance. Based on liquid-solid erosion experimental results, numerical simulations were conducted to compare the erosion behavior of elbows and blind tees with different length-to-diameter ratios (1–2.5) under different outlet pressures (25–125 MPa), flow velocities (7.5–17.5 m/s), and particle mass flow rates (0.05–0.25 kg/s). The liquid-solid two-phase flow was modeled using the Euler-Lagrange approach with a modified E/CRC erosion model validated by experiments. Results show that increasing outlet pressure, velocity, and particle flow rate significantly enhances erosion, with velocity exhibiting an exponential effect and particle flow rate a linear effect. Compared to elbows, blind tees generate a buffer vortex that reduces the maximum erosion rate. The primary erosion area occurs at the tee intersection and the outlet pipe bottom. The maximum erosion rate decreases with increasing L/D, considering overall cost and space efficiency, L/D = 2 is the optimal choice. These findings provide valuable insights for the design and safe operation of high-pressure pipe manifolds.

Online detection of wear particles in marine machinery lubricants is critical for effective condition monitoring. However, current methods encounter significant difficulties in accurately identifying microscale wear debris owing to the high-frequency background noise and motion blur inherent in online ferrographic images. To address these challenges, this study proposes YOLO-RSMA, a novel detection framework based on YOLOv8 architecture. The model systematically integrates a Residual Efficient Multi-scale Attention (REMA) mechanism to retrieve global contextual features that are frequently attenuated in deep convolutional layers. Furthermore, the framework employs the Sophia optimizer to mitigate gradient instability within nonconvex loss landscapes, and incorporates the MPDIoU loss function to ensure precise geometric alignment for irregular particle morphologies. Validated on a custom dataset against industrial baselines and state-of-the-art methods, YOLO-RSMA demonstrated superior robustness under complex imaging conditions. The model achieves a mAP0.5 of 99.30%, representing a substantial improvement of 16.63% over the baseline, thereby offering an efficient and accurate solution for intelligent fault diagnosis.

This paper focuses on the study of the load-independent power losses generated by a complete planetary gearbox having a rotating ring gear and a fixed planet carrier. First, different configurations are studied to characterize the various losses such as those associated to bearings. The power losses generated by the complete gearbox is then measured thanks to a dedicated test bench. Those tests enable to determine a power loss distribution that is then validated using a thermal network analysis. Once the windage and the oil churning losses are estimated, their evolution with rotational speed and oil flow rate is investigated. The study reveals that the windage power losses, generated by the reduced scale planetary gearbox studied, do not increase with the cube of the rotational speed within the tested speed range. The oil churning losses evolve linearly with the oil flow rate and quasi linearly with speed. Finally, the evolution of the power loss distribution with respect to the rotational speed is presented.

Every year, significant amounts of non-biodegradable, toxic lubricants enter the environment. While legislative measures are being taken to reduce the environmental impact of this pollution, demand for environmentally acceptable lubricants (EAL) is increasing. These EALs differ in terms of their base oil and additives, thereby having a significant influence on the degradation behavior and degradation-related changes of tribological performance like oil film formation or boundary layer formation. However, the sensitivity of EALs to different degradation paths is not yet known. Therefore, the aim of this study is to determine the sensitivity of different lubricants to degradation based on their change in tribological performance. For this purpose, an ester- and polyglycol-based EAL as well as a mineral oil (polyalphaolefin) were degraded synthetically by oxidative and hydrolytic degradation. The degraded lubricants are visually, rheologically and chemically analyzed. In addition, the tribological performance in terms of oil film formation and tribological boundary layer formation was studied in a ball-on-disc tribometer. Afterwards the results are compared to their corresponding fresh lubricants. For the tested ester-based EAL no significant sensitivity to degradation could be shown, as no significant change in tribological performance occurred. Both polyalphaolefin and polyglycol proved to be sensitive to degradation leading to a significant reduction in the tribological boundary layer thickness. These results show, that mineral oil-based lubricants and EALs, as well as EALs themselves, differ in terms of their sensitivity to degradation. These effects must be considered in qualification process of lubricants.

Hydrogen is being introduced as a clean energy source for heavy-duty applications to reduce carbon emissions and environmental impact. However, its application in internal combustion engines poses significant challenges, particularly regarding lubrication performance and surface degradation under pressurised hydrogen conditions. Hydrogen diffusion at contact interfaces can alter tribofilm formation, potentially increasing wear and compromising component durability and efficiency—issues that remain insufficiently explored. This study examines the tribological performance of conventional lubricants in a hydrogen-rich environment using a custom-built tribometer housed within a 3-bar pressurised hydrogen vessel. Lubricants formulated separately with zinc dialkyldithiophosphate (ZDDP), molybdenum dithiocarbamate (MoDTC), and glycerol monooleate (GMO) in base oil were tested. Surface characterisation was performed using Raman spectroscopy and scanning electron microscopy coupled with energy-dispersive spectroscopy (SEM/EDS). The results reveal the presence of iron oxides and carbonaceous materials on sliding surfaces, which influence the formation and stability of ZDDP-, MoDTC-, and GMO-derived tribofilms. These interactions lead to variations in friction and increased wear, highlighting the need for lubricant formulations specifically designed for hydrogen-based systems.

Driven by the escalating requirements for efficiency and operational safety in power-generation units, data-driven forecasting of generator output has emerged as a critical research frontier. Generator performance is governed by a complex interplay between mechanical operating conditions and stochastic human factors. In this study, we performed online monitoring of turbine lubricating oil to assemble a comprehensive time-series dataset integrating oil wear signatures and unit power metrics. We propose a synergistic deep-learning architecture, the Transformer–LSTM, which leverages an LSTM-based auxiliary extractor to capture fine-grained local temporal patterns. Concurrently, positional encodings, timestamp embeddings, and value embeddings are employed to characterize global contextual features. By fusing these local and global representations, the model achieves a joint perception of short-term dynamics and long-range dependencies. Empirical results demonstrate that the proposed hybrid model significantly outperforms standalone LSTM and Transformer baselines across diverse evaluation metrics in terms of both predictive accuracy and robustness. This framework provides robust technical decision support for condition-based maintenance and optimized power-generation scheduling.

Lubricants containing nanoparticles have garnered significant attention due to their potential to improve the tribological performance of systems operating in demanding environments. In the current study, the enhanced stability of copper oxide (CuO) and molybdenum disulphide (MoS₂) nanoparticles (NPs) and their tribological performance, when used as lubricant additives, were experimentally tested to evaluate their synergistic effect on friction and wear performance. The oleic acid (OA) was used as a surface modifier agent to enhance dispersion stability. The influence of oleic acid on nanoparticle dispersion was evaluated using Dynamic Light Scattering (DLS) and Zeta Potential (ZP) techniques. The surface of nanoparticles was characterised by Fourier Transform Infrared Spectroscopy (FTIR) and Thermogravimetric Analysis (TGA). The tribological performance of the more stable nanolubricant, including friction reduction and wear resistance, was evaluated using a pin-on-reciprocating plate test rig. Surface characterisation of the wear tracks was performed using White Light Interferometry (WLI), Scanning Electron Microscopy with Energy Dispersive X-ray Spectroscopy (SEM/EDX), X-ray Photoelectron Spectroscopy (XPS) and Transmission Electron Microscopy (TEM) to understand how the improved stability affected the worn surfaces and tribofilm properties. NPs stability study confirmed that the best stability could be achieved with concentrations of 1% CuO/MoS₂ NPs and 1.5 wt. % OA. Tribology results indicated that CuO and MoS₂ NPs individually could reduce friction and wear, whereas their combination exhibited enhanced friction and wear reduction. This synergy in tribological performance has been attributed to the combined effects of nanoparticle size and enhanced dispersion stability, which enable effective nanoparticle entrainment between the contacting surfaces, leading to the formation of a tribofilm that provides anti-wear protection and contributes to friction reduction.

Surface topography is inherently multiscale. Macroscale topography is covered in microscale surface roughness, which has a substantial, quantifiable impact on overall performance. However, this difference in scales raises modelling challenges. Contemporary mixed lubrication research is dominated by two methods: the accurate but computationally expensive deterministic approach and the more efficient flow factors method that cannot predict local parameters (pressure, contact, etc.) accurately. The principal objective of this work is to develop a framework that will both predict local scale variables accurately and solve in a feasible amount of time.

Through the use of the Heterogeneous Multiscale Methods (HMM) the macroscale and microscale surface features can be modelled using separate domains, whilst maintaining the coupling between the scales. This homogenisation approach allows for both accurate predictions of local scale phenomena and computational efficiency. The HMM have been extensively validated for hydrodynamic/elastohydrodynamic lubrication. Within this work, the HMM are extended to model mixed lubrication and are validated against existing models and experimental data. To build further confidence, plots of simulated data are presented for various conditions on the Stribeck curve.

Both idealised and measured microscale topographies are applied and their impact quantified. The interaction between asperities in contact and lubricant flow has been fully resolved. This framework represents the first time that this microscale behaviour has informed the solution of a computationally efficient macroscale operating within the mixed lubrication regime. Average solution time is 8 h on a local PC with 16GB of RAM.

Condition monitoring can help to detect faults of rotating machinery early and thereby prevent failures. Rolling element bearings are one of the most important machine elements to be monitored. This study focusses on rolling element bearing fault detection and localization using high-frequency, structure-borne sound, so-called acoustic emissions (AE) sensors on a dedicated roller bearing test bench. One the one hand, the high-frequency signals (range 20–1000 kHz) are analyzed and on the other hand, a demodulation algorithm is employed to down-sample the signals to frequency range of common bearing frequencies (≤10 kHz) to allow a state-of-the-art fault localization using spectral analysis of these signals. The AE results are also compared to the commonly used spectral analysis of vibration signals using conventional, piezo-electric acceleration sensors (≤10 kHz). The results show that AE is on par with vibration signals for fault localization and outperforms vibration in detecting very small surface damages and starved-lubrication conditions.

A significant portion of the energy used to propel a vehicle is dissipated as rolling resistance (RR) in the tire-road interaction. Largely due to the hysteretic losses resulting from the cyclical deformations of the tire's viscoelastic material as the tread engages with asperities of different wavelengths, RR depends on the tire and pavement surface characteristics. However, factors related to vehicle and ambient conditions also play a role. In particular, specifics of battery electric vehicles (BEVs) like the increased tire load and different torque performance compared to equivalent-sized internal combustion engine vehicles are expected to increase the RR and induced energy consumption. This paper investigates the relationship between the pavement surface characteristics and the RR in a passenger BEV. To this end, experimental measurements of the RR were performed in conditions close to real driving. They involved an instrumented BEV equipped with two dynamometric wheels acting as tribometers, driven on a test track featuring several asphalt pavement sections of diverse texture and roughness levels, and differing in age and wearing course. Standardized descriptors of these surface properties were then correlated with the corresponding measured RR coefficients. Different segmentation strategies were applied to derive the values of the metrics. The results (1) validate correlations previously observed in drum- and trailer-based measurements and (2) reveal how local variations in pavement surface characteristics influence RR. These findings may support road pavement managers in the design of pavement management strategies that potentialize the environmental benefits offered by passenger BEVs.

This study investigates lubricant-induced power losses of tapered roller bearings. Rotational torque was measured with a KRL shear stability tester using five lubricants, including base oils and gear oils. Two conditions were examined: one without temperature control and one under controlled temperature. In the test without temperature control, the initial torque matched the predicted kinematic viscosity at the start. After a certain period, torque values converged across all oils. The relationship between torque and kinematic viscosity, calculated from oil temperature, shows that the torque depends primarily on kinematic viscosity, with no significant effects from other factors. In the test under controlled temperature, torque values correlated with the predicted kinematic viscosity and showed consistency with the calculated viscosity at each rotational speed. Thus, in this test as well, torque appears primarily dependent on kinematic viscosity. Comparison with torque values calculated using the SKF model of bearing friction suggests that the rotational torque observed in these tests primarily reflects changes in the viscosity-dependent rolling friction behavior. However, torque tended to be lower than the rolling frictional moment (M_rr) under conditions where the oil film was thin. As a result, the rotational test using the KRL shear stability tester may estimate energy loss due to the kinematic viscosity. However, it remains difficult to assess the influence of other factors, such as oil type and additives, with this test.