With the increase of length and axle load of heavy haul trains, the wheel wear of freight wagons gets more attention. Wheel eccentricity is usually caused by wheel wear and the error of reprofiling quality. The periodic impact vibration induced by the wheel eccentricity, which influences the safety and riding comfort of railway vehicles. The vibration of freight wagons is complex because of random excitation of the track irregular and the operation speed of freight wagons is usually under the varying speed condition. Those factors result in the difficulty of the detection of the wheel eccentricity. Focusing on this problem, a dynamics model of freight wagon is established and then simulated under a varying speed condition. The acceleration of the axle-box is extracted. The adaptive chirp mode decomposition method is adopted to detect wheel eccentricity because of its outstanding merits in processing the non-stationary signal. The results indicate that this method is capable of decomposing and separating the harmonic components induced by the wheel eccentricity. The instantaneous frequency estimated by adaptive chirp mode decomposition is also in good agreement with the actual fault frequency under the varying speed condition.

Wheel eccentricity; Condition monitoring; Freight wagons; Varying speed; Adaptive mode decomposition.