ABSTRACT
With the growth of the share of wind energy in the energy market, the design and implementation of large scale wind turbines has become a common occurrence. However, since its inception the wind energy industry has experienced high gearbox failure rates [1]. In order to achieve their stated design life goals of 20 years, most systems require significant repairs or overhauls well before the intended life is reached [2-4]. Some firm conclusions about the nature of the failures have been made based on the work of Musial et al. [5]: (i) gearbox failures are not specific to a single gear manufacturer or turbine model, which are general; (ii) poor adherence to accepted gear industry practices, or otherwise poor workmanship, is not the primary source of failures; (iii) most gearbox failures do not begin as
gear failures or gear-tooth design deficiencies, and up to 10% of gearbox failures may be manufacturing anomalies and quality issues that are gear related; (iv) the majority of wind turbine gearbox failures appear to initiate in the bearings, and (v) it is believed that the gearbox failures observed in the earlier 500 kW to 1000 kW sizes 5 to 10 years ago may still occur in many of the larger 1 to 2 MW gearboxes being built today with the same architecture. With larger wind turbines, the cost of gearbox rebuilds, as well as the down time associated with these failures, has become a significant portion of the overall cost of wind energy [6]. Presently NREL is performing a long-term project to improve the accuracy of dynamic gearbox testing to assess gearbox and drivetrain options, problems, and solutions under simulated field conditions. In addition, in order to increase the longterm reliability of gearboxes and make their design more reasonable, there is increasing interest in utilizing time domain simulations and physical tests in the prediction of gearbox design loads, with the continual development of computer technology, simulation tools and measurement equipments. In several previous studies Klose et al. [7] performed an integrated analysis of wind turbine behavior and structural dynamics of a jacket support structure under combined wind and wave loads in the time domain. Seidel et al. [8] used the sequential coupling and the full coupling methods to simulate offshore loads on jacket wind turbines, and validated these methods using measurement data from the DOWNVInD project. Gao and Moan [9] performed long-term fatigue analysis of offshore fixed wind turbines using time domain simulations. Dong et al. [10] performed longterm fatigue analysis of multi-planar tubular joints for jacket-type offshore wind turbines using time domain simulations. Peeters et al. [11,12], Xing et al. [13] performed a detailed analysis of internal drive train dynamics in a wind turbine using multi-body simulations. However, there is, at present, limited literature concerning long-term time-domain based analysis of mechanical components, e.g., main shaft, gears and bearings, in the wind turbine drive train system under dynamic conditions. This is mainly due to the complexities involved in modeling and simulating the drivetrain with respect to the computation efforts and scale. Recently, Dong et al. [14] established and applied a long-term time domain based gear contact fatigue analysis of a wind turbine under dynamic conditions. In the present study, several practical problems of time domain based gear contact
fatigue analysis encountered in [14] are described and discussed. These are: (1) the rotation reversal problem of gears under low wind speed conditions, (2) the statistical uncertainty effect due to the time domain simulation and (3) simplified long term contact fatigue analysis of gear teeth under dynamic conditions. Several useful suggestions to address these issues are proposed.
