Numerical Procedure for Gear Dynamics Simulation of Wind Energy Drivetrain Using Tabular Contact Search

Abstract eng:
Accurate prediction of dynamic load variations in a wind turbine gearbox is crucial for its design and performance evaluation [1]. In particular, due to the high gear ratio and the large torque and speed variations of the input shaft connected to rotor blades, gears are subjected to severe contact and it is reported that 20-year design life of the drivetrain is often hard to achieve [2]. In this investigation, a numerical procedure for the analysis of multibody wind energy drivetrain model, which considers the effect of micro-geometry of a gear pair on the dynamic transmission error, is developed. In order to parameterize the gear tooth profile, two surface parameters are introduced as shown in Fig. 1 and the location of contact point on the arbitrary surface can then be defined using the surface parameters. The gear tooth profile function can be provided by either analytical or tabular form with the three- layer smoothing spline function for accurately obtaining its higher order derivatives. Despite the accurate surface representation used in the model, use of the online contact detection that involves multi-point contact leads to extensive computational time when user-defined or measured gear tooth profiles are considered for the failure mode analysis. For this reason, in this investigation, the contact geometry analysis for a single gear pair is performed as a function of relative rotation of each gear wheel prior to the dynamic simulation and then the contact point are stored in a look-up contact table. The look-up contact tables are then interpolated in the dynamic simulation as a function of relative rotation of a gear pair for determining the location of contact point at every time step. This leads to a computationally fast contact search algorithm with considering the precise contact geometry of gear pairs. The several numerical examples are presented in order to demonstrate the use of the numerical procedure developed in this investigation.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket