Wear estimation for journal bearings with large static eccentricity under various wear evolution conditions using fast solution of bearing fluid film force

Although journal bearings (JBs) are a key component in rotating machinery to ensure stable dynamic performance, wear in JBs remains a critical issue that can lead to severe vibrations or failures. On the one hand, model-free methods (MFMs) offer rapid fault diagnosis but suffer from poor interpretability and weak extrapolation. On the other hand, hybrid physics-based methods (HPBMs) incorporate physical models to improve accuracy. However, they are limited by heavy computational demands. In this study, we propose a dimension-reduction approach coupled with a worn profile model derived from experimental measurement to achieve faster pressure distribution analysis. The results of a numerical validation are presented to demonstrate that the proposed method improves computational efficiency by a factor of approximately 30, and thus enables the application of computationally intensive methods such as genetic algorithms (GA) for online wear estimation. The proposed approach was combined with a GA to estimate wear in water-lubricated JBs using experimental data, and the results showed a close agreement between the estimated and measured values for worn clearance. Furthermore, analysis of the estimated profiles reveals distinct trends in the evolution of the effects of wear, which provides some valuable insights into the wear process of tested JBs. ● A dimension-reduction method is developed for wear estimation of journal bearings. ● A Fourier-series profile model provides smooth and accurate wear representation. ● The proposed method enhances efficiency by approximately 30 times over 2D models. ● Coupled with GA, the method enables online wear estimation from vibration data.