Acceleration signal with DTCWPT and novel optimize SNR index for diagnosis of misaligned cardan shaft in high-speed train. (June 2020)

• Record Type: Journal Article
• Authors: Hu, Yongxu; Zhang, Bing; Tan, Andy Chit
• Is Part Of: Mechanical systems and signal processing. Volume 140(2020)
• Publication Date: 2020-06
• Subjects: Misalignment -- Cardan shaft -- Acceleration signal -- Complex wavelet -- SNR index -- High-speed train; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2020.106723
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: Systematic studies of kinetics and dynamics of misaligned cardan shaft in rotation. Confirming of predominant acceleration due rotation of misaligned cardan shaft. Novel algorithm for identification of cardan shaft misalignment in train. New experimental designing to simulate actual cardan shaft misalignment. Abstract: The cardan shaft transmits rotating motion from motor to gearbox where the axes of rotation of motor and gearbox are not in the same plane. The cardan shaft can be misaligned due to offset at the cross-shaft. The offset causes severe unbalance force on the shaft which eventually leads to catastrophic failure of the whole train. Therefore, it is essential to detect symptom of cardan shaft failure in advance. General diagnostic methods for detecting cardan shaft misalignment are through analysis of vibration signals from sensors located on the rotating system. Unfortunately, the signals are usually corrupted by other mechanical interferences and noise which make vibration analysis unreliable. Hence, it is necessary to eliminate mechanical interferences in the signal and result in improving the detection of shaft misalignment. This study aims to diagnose cardan shaft misalignment by using acceleration signal and dual tree complex wavelet packet transformation (DTCWPT) method to extract the relevant features. The acceleration response of misaligned cardan shaft is analyzed using the multibody dynamics model of Chinese high-speed rail (CRH) built in software. Both DTCWPT and Hilbert transform (HT) methods are applied to decompose the acceleration signal and an envelope of the signal is created. A novel index system is proposed to optimize the signal-to-noise ratio (SNR) index to improve the detection of shaft misalignment. The amplitude of Fourier transform (AFT) coefficient of acceleration signal is then used to assess the degree of shaft misalignment. The results from simulation study and experimental tasks confirmed the capability of the proposed technique in detecting cardan shaft misalignment. This paper presents a new approach to accurately and quantitatively diagnose cardan shaft misalignment.