Analysis of Journal-Bearing Wear Damage Caused by Residual Metallic Particles in the Lubrication System of an Ultra-Supercritical Steam Turbine

Ultra-supercritical (USC) steam turbines are high-efficiency, large-capacity power-generation systems that generally operate under steam conditions exceeding approximately 600 °C and 25 MPa. Journal bearings are critical mechanical components that support rotor loads and maintain the stable position of the rotating shaft. In this study, an actual high-vibration trip event in the journal-bearing lubrication system of a 1050 MW USC steam turbine was investigated by integrating operational data, surface-damage characteristics, recovered foreign particles, material properties, and particle–film hydrodynamic screening. During the event, the T8 journal vibration displacement increased from approximately 62.5 to 225 μm, exceeding the 220 μm trip threshold, and the T8 bearing-metal temperature increased to approximately 118–120 °C near 500 rpm during coast-down. The maximum abrasive-wear depths measured at T5, T7, T8, and T9 were 0.79, 0.73, 0.71, and 0.55 mm, respectively. The expanded descriptive dataset comprised 30 representative recovered particles, with measured sizes ranging from approximately 0.08 to 2.5 mm and a descriptive mean of approximately 0.71 mm; the particles included Fe-based contaminants and Sn-bearing secondary wear debris. The bearing metal had a hardness of approximately 18 HV, whereas the recovered hard particles exhibited substantially higher hardness. Particle–film interactions at bearings T5, T7, and T8 were quantitatively evaluated as functions of rotational speed, lubricant viscosity, and particle size using particle response time, radius-based and local-shear-based nominal Stokes numbers and particle-to-film geometric relationships. Under identical rotational speed, lubricant viscosity, and particle-size conditions, the radius-based nominal Stokes number was identical among the three bearings, whereas the local-shear Stokes number varied with bearing geometry and maximum oil-film thickness. For the 200 μm reference particle at 3600 rpm, the T8 radius-based nominal Stokes number was 0.1359, whereas the local-shear Stokes number was 154.58, demonstrating the strong dependence of particle response on the selected fluid time scale. The observed directional wear, particle embedment and pull-out traces, hardness contrast, and conditional particle–film analysis are consistent with a particle-mediated three-body abrasive-wear mechanism. These findings indicate that particle behavior cannot be assessed from nominal particle size alone and that prevention of particle ingress and retention requires life-cycle contamination control during fabrication, installation, flushing, and commissioning.