Dynamic nonstationary representations of a railway vehicle considering wheel polygonalization under variable speed conditions

Wheel polygonalization is one of the most common failures of the wheelset, which can directly affect the safety and comfort of railway vehicle operation. In the actual condition, traction/braking torque continuously acts on the polygonal wheel fault can result in the modulation of its vibration representation and then influence the effect of the feature exaction of the polygonal wheel faults. Considering this phenomenon, this study aims to clarify the forming mechanism of fault signals under variable speed conditions to improve the accuracy of feature extraction and ensure the safe operation of vehicles. This paper derived the Lagrangian equations with the dissipation function of a railway vehicle model with polygonal wheel fault under traction/braking conditions. After that, the modulation representation under constant speed, linear variable speeds, and nonlinear variable speeds is investigated, respectively. Finally, the vibration evolution of polygonal faults under different working conditions is visualized using statistical features. Compared with the constant speed, the curve in the time domain has more high-frequency fluctuations under variable conditions, and the phase modulation causes the half-wave asymmetry of the waveform. In the time-frequency domain, the continuous input of torque increases the harmonics frequency and side frequency response of fault response and excites the resonance frequency of the vehicle system. The dynamic evolution of fault statistical characteristics is positively correlated with speed and fault severity, and the fluctuation is sharper under braking conditions. The results can provide the theoretical support for feature extraction, interpretable features, and intelligent diagnosis of polygon faults.

Automated summary

Wheel polygonalization is a common failure in wheelsets that can impact the safety and comfort of railway vehicles. This study aims to clarify the formation mechanism of fault signals under variable speed conditions for improved feature extraction accuracy. The results show that the dynamic evolution of fault statistical characteristics is positively correlated with speed and fault severity, with sharper fluctuations under braking conditions.