Guangde Zhou, Menghao Zhan, Dan Huang, Xiaolong Lyu and Kanghao Yan
By seamlessly integrating physical laws, physics-informed neural networks (PINNs) have flexibly solved a wide variety of partial differential equations (PDEs). However, encoding…
Abstract
Purpose
By seamlessly integrating physical laws, physics-informed neural networks (PINNs) have flexibly solved a wide variety of partial differential equations (PDEs). However, encoding PDEs and constraints as soft penalties in the loss function can cause gradient imbalances, leading to training and accuracy issues. This study aims to introduce the augmented Lagrangian method (ALM) and transfer learning to address these challenges and enhance the effectiveness of PINNs for hydrodynamic lubrication analysis.
Design/methodology/approach
The loss function was reformatted by ALM, adaptively adjusting the loss weights during training. Transfer learning was used to accelerate the convergence of PINNs under similar conditions. Additionally, the iterative process for load balancing was reframed as an inverse problem by extending film thickness as a trainable variable.
Findings
ALM-PINNs significantly reduced the maximum absolute boundary error by almost 80%. Transfer learning accelerated PINNs for solving the Reynolds equation, reducing training epochs by an order of magnitude. The iterative process for load balancing was effectively eliminated by extending the thickness as a trainable parameter, achieving a maximum percentage error of 2.31%. These outcomes demonstrated strong agreement with FDM results, analytical solutions and experimental data.
Originality/value
This study proposes a PINN-based approach for hydrodynamic lubrication analysis that significantly improves boundary accuracy and the training process. Additionally, it effectively replaces the load balancing procedure. This methodology demonstrates considerable potential for broader applications across various boundary value problems and iterative processes.
Peer review
The peer review history for this article is available at: https://publons.com/publon/10.1108/ILT-07-2024-0277/
Details
Keywords
Kanghao Yan and Dan Huang
In high-pressure pumps, due to the interaction of asperities on the upper and lower surfaces, the piston–cylinder interface suffers severe lubrication and sealing problems during…
Abstract
Purpose
In high-pressure pumps, due to the interaction of asperities on the upper and lower surfaces, the piston–cylinder interface suffers severe lubrication and sealing problems during mixed lubrication. This study aims to establish a mixed thermo-elastohydrodynamic (EHD) model for the lubrication gap to determine how working conditions affect the lubricating characteristics and sealing performance of the interface.
Design/methodology/approach
A mixed thermo-EHD lubrication model is established to investigate the lubricating characteristics and sealing performance of the interface between the piston and cylinder. The model considers piston tilting, thermal effect, surface roughness and bushing deformation. The interface lubricating characteristics and sealing performance under different working conditions are calculated by the proposed numerical model.
Findings
A higher inlet pressure contributes to an increase in the minimum film thickness. Increased shaft speed can significantly reduce the minimum film thickness, resulting in severe wear. Compared to roughness, the impact of the thermal effect on the interface sealing performance is more significant.
Originality/value
The proposed lubrication model in this study offers a theoretical framework to evaluate the lubricating characteristics and sealing performance at the lubrication gap. Furthermore, the results provide references for properly selecting piston-cylinder surface processing parameters.
Peer review
The peer review history for this article is available at: https://publons.com/publon/10.1108/ILT-03-2023-0072/
Details
Keywords
Kanghao Yan and Dan Huang
Multitilting-pad journal bearings (MTPJBs) used in large-scale hydraulic turbines often suffer from complex operating conditions, which greatly influence the overall performance…
Abstract
Purpose
Multitilting-pad journal bearings (MTPJBs) used in large-scale hydraulic turbines often suffer from complex operating conditions, which greatly influence the overall performance of the rotating machine. The purpose of this study is to establish a thermal-elastic-hydrodynamic lubrication model for MTPJBs that can predict the static and dynamic characteristics of high-speed and heavy-load MTPJBs under different operating conditions.
Design/methodology/approach
A thermo-elasto-hydrodynamic lubrication model considering the turbulence effect is proposed for high-speed and heavy-load TPJBs, which is solved using the coupled finite difference method and finite element method. The model considered the turbulence effect, thermal energy diffusion, viscosity–temperature–pressure relationship and elastic deformation of the pads. The influences of the operating conditions on static and dynamic characteristics of tilting pad journal bearings were analyzed in depth.
Findings
The operating conditions have a strong effect on the static properties of the bearings. The dynamic characteristics of the TPJB were the most influenced by the shaft speed. The effects of the load direction on the dynamic properties of the TPJB were much stronger than those of the static characteristics.
Originality/value
This study used analytical methods and models to provide theoretical guidance for evaluating lubricating characteristics, assembling conditions and overall health.