Influence of Turbulent Excitation Force Spectrum Models on Fretting Wear Prediction of Foreign Objects in Nuclear Steam Generator

Foreign objects trapped between the tube bundles of a nuclear steam generator can undergo fretting wear with the tube walls due to excitation by turbulent flow effects in the complex flow field, posing a threat to equipment safety. This study aims to investigate the impact of turbulent excitation force spectrum models on the conservatism of fretting wear predictions. Initially, experimental measurements were conducted to capture the fluctuating time-history signals of turbulent lift and drag forces acting on foreign objects trapped in a scaled steam generator tube bundle region. Based on the experimental data, envelope spectrum models of the turbulent excitation forces were developed using various methods, including segmented linear, segmented logarithmic, polynomial, and maximum enveloping. By introducing quantitative evaluation metrics-Mean Excess Error (MEE), Relative Excess Error (REE), Smoothness Index (SI), and Power-law Consistency Error (PCE), a systematic comparison was performed to assess the fitting accuracy and conservatism of different envelope models across the full frequency range. The results indicate that the choice of envelope spectrum model significantly affects the inversion results of the excitation forces. Simple segmented linear or low-order polynomial envelope models show poor fitting at low frequencies, leading to an overestimation of the turbulent force input. In contrast, traditional empirical models may fail to fully envelope the experimental data in the mid-to-high frequency range. The comparative analysis reveals that for the drag force spectrum, an eight-segment logarithmic envelope model demonstrates the best overall performance across all metrics, with its high-frequency decay slope showing good agreement with classical turbulence theory. For the more complex lift force spectrum, an eighth-order polynomial envelope model best captures its multi-peak structure, yielding the highest fitting accuracy. This study provides a quantitative basis for selecting force spectrum models that balance conservatism and accuracy in engineering assessments.It offers important references for accurately evaluating the fretting wear risks associated with foreign objects trapped in steam generators.