Adequately designed and positioned surface textures are recognized as a promising way to increase load-carrying capacity and reduce frictional losses of journal bearings. The aim of this work is to analyze the local lubrication mechanisms of textures in journal bearings from microflow perspective, while considering the interactions between textures and the film formation in the whole bearing. For this purpose, hydrodynamic lubrication models of textured journal bearings are built. The results show that placing textures downstream of the high-pressure region leads to a reduced friction force, with a less severe loss of load-carrying capacity. The effects of textures on the load-carrying capacity include the positive micro-hydrodynamic pressure effect and the negative effect caused by the discontinuity of the high-pressure region. The micro-hydrodynamic pressure of textures can be generated on one hand by limiting pressure drop (cavitation) in the divergent gap and on the other hand by the inertia effect. For the friction, the vortex inside textures affects the friction force by influencing the maximum shear stress at the minimum oil film. In turn the vortex is influenced by the bearing lubrication film. The research provides the fundamental reference and theoretical basis for the design and optimization of textured journal bearings.
Surface textures in journal bearings offer significant potential for reducing friction and enhancing energy efficiency. However, the complexity of texture configurations necessitates an accurate and efficient performance prediction model to properly design textured journal bearings. To address this issue, this study develops a machine learning (ML)-based surrogate modelto predict friction in textured journal bearings. First, Computational Fluid Dynamics (CFD) models employing a dynamic mesh algorithm are developed to generate accurate data sets. Furthermore, three ML methods are trained and compared to select the most suitable prediction method: artificial neural network (ANN), support vector regression (SVR), and Gaussian process regression (GPR). Among these ML methods, ANN shows the best prediction performance. Given the high computational cost of CFD simulations, the prediction accuracy of the ANN-based surrogate model is further enhanced without the need for additional data sets. This enhancement is achieved through an architecture design based on cross-validation and further optimization utilizing the genetic algorithm. Eventually, the average prediction accuracy is improved to 98.81% from 95.89%, with the maximum error reduced to 3.25% from 13.17%. These findingsdemonstrate the potential of ML in the performance prediction in textured journal bearings and provide a promising approach for broader applications in developing highly efficient and accurate ML-based surrogate models, particularly in cases with limited available training data sets.