Principle-Verification experimental platform for wind turbine gearbox under equivalent wind speed excitation

Wind turbine gearboxes are key components of wind power drivetrain systems; their vibrations are strongly affected by nonstationary input excitations arising from actual wind-field conditions. In practical operation, factors such as wind shear, tower shadow, and wind-speed distribution can induce periodic fluctuations in aerodynamic torque and rotational speed, which are transmitted along the drivetrain, producing amplitude modulation, frequency modulation, and complex sideband structures in gearbox vibration signals. Although existing gearbox test platforms are mainly based on constant-speed driving or simplified loading conditions, it is difficult to reproduce the dynamic input characteristics of wind turbines under complex wind-speed disturbances. Conventional experimental platforms, which mostly focus on single-fault simulation or single-signal acquisition, cannot fully meet the requirements of fault mechanism analysis, feature extraction verification, and experimental teaching. To address these issues, this study develops a principle-verification experimental platform for wind turbine gearboxes under equivalent wind-speed excitation, with the aim of reproducing wind-induced low-frequency disturbances in the laboratory and investigating the influence of equivalent wind-speed excitation on gearbox vibration response and fault feature evolution.

The platform consists of a drive unit, a transmission unit, a power generation unit, and a control and test unit. The drive unit uses a servo motor and reducer structure and programmable speed commands (e.g., constant speed, step speed, ramp speed, random disturbance, and periodic disturbance) controlled through a programmable logic controller (PLC). The transmission unit includes a planetary gearbox and a parallel-shaft gearbox that facilitate comparative analysis of different gearbox structures. Based on wind shear and tower shadow effects, an equivalent wind-speed model is established, which can be transformed into a programmable input disturbance signal to reproduce the low-frequency fluctuations caused by the spatiotemporal distribution of wind speed at the drivetrain input side. Considering a single broken tooth fault of the planetary gearbox sun gear as a typical local fault, four operating conditions are designed: the normal state under constant wind speed, the normal state under equivalent wind speed, the broken-tooth fault under constant wind speed, and the broken-tooth fault under equivalent wind speed. Vibration signals are collected under the same sampling conditions, and time-domain analysis, frequency-domain analysis, and Hilbert envelope demodulation are used to identify rotational frequencies, gear meshing frequencies, fault characteristic frequencies, and modulation sidebands.

The experimental results reveal that, under constant wind-speed conditions, the gearbox vibration amplitude is relatively stable and the waveform fluctuation is small, whereas under equivalent wind-speed excitation, the vibration amplitude increases and the waveform fluctuation becomes more obvious, indicating that the low-frequency disturbance can be effectively transmitted through the drivetrain. In the frequency and envelope spectra, additional modulation components related to equivalent wind speed are observed, especially sidebands at three times the rotor rotational frequency around the gear meshing frequency. For the sun gear broken-tooth fault, periodic impact components appear in the time-domain signal, and the fault characteristic frequency and its harmonics become prominent after envelope demodulation. Under the combined action of equivalent wind-speed excitation and the local tooth fault, the vibration spectrum presents a more complex sideband structure because of the superposition of wind-induced and fault-induced modulations.

Thus, the experimental platform can effectively reproduce equivalent wind-speed excitation, simulate typical gearbox faults, and collect vibration signals for fault feature analysis. The results also verify that equivalent wind-speed excitation can induce obvious low-frequency modulation characteristics in gearbox vibration and that the platform can identify broken-tooth fault features under both constant and disturbed input conditions. This platform provides experimental support for wind turbine gearbox fault mechanism analysis, feature extraction method verification, complex working condition simulation, and related experimental teaching.