Spectral analysis of railway vehicle vertical vibration under normal operating conditions

ABSTRACT

This study focuses on railway vertical random vibration–based spectral analysis under normal operating conditions. Vibration acceleration measurements are acquired using sensors mounted on the car body, bogie frame, axle bearing, and primary suspension of an Athens Metro vehicle running without passengers on a straight track. Two distinct track segments and four speeds (50, 60, 70, 80 km/h) are considered. The analysis focuses on the 0–1000 Hz frequency range using data-based non-parametric and parametric spectral estimation methods. The results indicate good agreement of the methods, the prominence of the 5–170 Hz frequency range along with higher frequencies at around 300 Hz and in the range of 600–750 Hz on the car body and the bogie frame vibration, the relatively minor dependency of the dynamics on the track segment, but their significant dependence on train speed. Yet, resonant frequencies are affected to a lesser extent, with at least 20 of them appearing as prominent.

KEYWORDS:

• Power spectral density
• scalar and vector autoregressive modelling
• statistical analysis
• data-based methods
• railway vehicle dynamics
• vertical vibration signals

Acknowledgments

The authors are grateful to Attiko Metro S.A. and STASY S.A., Athens, Greece, for their support and assistance with the experiments. Particular thanks are due to Dr. G. Leoutsakos, Dr. A. Deloukas, Dr. E. Chronopoulos, Ms. K. Katsiana of Attiko Metro S.A.; and Mr. V. Tsakas, Mr. E. Paggios, Mr. C. Sarris of STASY S.A. Thanks also to Mr. K. Vamvoudakis–Stefanou, Ms. M.-C. Alexiou, and Mr. E. Loggovitis of the University of Patras for their assistance during the experiments.

Disclosure statement

No potential conflict of interest was reported by the authors.