Processing Method and Characteristic Analysis for Sliding Friction Test of Rubber Blocks on Wet Surface

Accurately acquiring the dynamic friction characteristics between rubber and wet road surfaces is critical for improving vehicle braking safety and optimizing tire design. However, sliding rubber's friction coefficient test results are easily disturbed by noise and exhibit non‐stationary and nonlinear features, which restrict traditional analytical methods like Fourier transform. Though Empirical Mode Decomposition (EMD) is applicable to such signals, it suffers from mode aliasing. To solve these problems, this study adopts the improved adaptive noise‐based ICEEMDAN to process rubber friction signals via signal analysis and reconstruction. First, a double sliding window method is used to identify steady‐state friction intervals adaptively and eliminate transient data in acceleration/deceleration stages. Then, the signal is decomposed into Intrinsic Mode Functions (IMFs), from which effective components are extracted for reconstruction using a framework combining hypothesis testing and variance contribution rate thresholds. Results show that this method enhances experimental repeatability and reduces friction coefficient volatility. The friction coefficient decreases with rising speed (200–2000 mm/s) and load (200, 400 N). Different rubbers' frictional properties are mainly dominated by their dynamic mechanical properties, especially loss modulus, laying a foundation for further understanding rubber‐road friction and optimizing rubber formulations.