Experimental investigation on a Jeffcott rotor with combined coupling misalignment using time-frequency analysis

Abstract Rotating machinery, such as turbo-jet engines, operate at a high rotational speed and passes through critical zones. The dynamic response of high-speed machines is critical for long-term stability and functioning. In this work, a fast and effective method for detecting coupling misalignment utilising time-frequency analysis (TFA) based on both the adaptive noise added complete ensemble empirical mode decomposition and wavelet-based denoising is presented. This novel and innovative method detect the coupling misalignment feature via the amplitude modulation aspect in the envelope analysis of the fault-containing intrinsic mode function. The Hilbert spectrum analysis provides spontaneous frequency and spectral energy in the time-frequency domain. The experiments were performed for various rotor accelerations and combined parallel and angular coupling misalignments using a laboratory test rig. The suggested approach gives excellent denoising efficiency and can improve misalignment identification accuracy. Additionally, it may be highly helpful for machinery that starts and stops often.