Fractal and multifractal characterization of wear surface dynamics for tribopairs in bi-directional seawater pump

Traditional roughness parameters are inadequate for characterizing the multi-scale wear surface features for tribopairs in bidirectional seawater pump. This study employs fractal theory, specifically using the box-counting method to calculate the fractal dimension and multifractal spectrum, to investigate the dynamic evolution of carbon fiber polyether ether ketone (CF-PEEK) / 17-4PH stainless steel tribopair under bidirectional sliding conditions. In line with the actual operating conditions of the bidirectional seawater pump, ball-on-disk wear tests were conducted at a load of 20 N, a sliding speed of 0.4 m/s, and commutation frequency (C = 0 ∼ 80). The sensitivity of the fractal dimension was found to be 5.6 times higher than that of traditional surface roughness (R_a) at C = 80, indicating its potential as a high-resolution tool for dynamic monitoring. The multifractal spectrum analysis reveals that lower commutation frequencies (C ≤ 20) result in increased surface fluctuations, whereas higher commutation frequencies stabilize the lubrication layer. Fractal theory effectively captures the dynamic wear surface characteristics, while the multifractal spectrum provides a foundation for monitoring lubrication states and optimizing maintenance strategies for the tribopairs in bi-directional seawater pump.