| Designing for fatigue has always been a formidable task for the wind turbine engineer. An integrated approach to wind turbine fatigue design is outlined which interfaces existing computer codes that generate simulated wind turbulence (SNLWIND-3D), model wind turbine dynamics (YawDyn and ADAMS), and estimate fatigue life (LIFE2). This approach is employed to investigate potential sources of fatigue damage to wind turbines. Three wind turbine models are used including a downwind, three-bladed, rigid hub machine, an upwind, three-bladed, rigid hub machine, and a downwind, two-bladed, teetering hub machine. For each model, stress cycles at the blade root are calculated from flap moment cycles. Conversion factors for this task were determined based on assumptions regarding fatigue load levels the blades are designed to tolerate. The load to stress conversion and interface of dynamic analysis and fatigue life programs were expedited by the program Dyn2LIFE developed as part of this project.;Results confirm the necessity for accurate models of both wind turbulence and wind turbine to yield meaningful fatigue analysis results. The importance of a complete fatigue analysis including fatigue damage calculation is demonstrated as a means toward understanding the significance of fatigue design decisions. This project reveals several specific sources of fatigue loads.;Although turbulence is a well-known, yet uncontrollable, source of fatigue cycles, it is shown that resultant fatigue damage may actually be due to controllable machine response. For the three-bladed machines, yaw is found to play a significant role in fatigue damage rates. High yaw rates of the free yaw machine are shown to yield gyroscopic load cycles far larger than most other normal operation load cycles. For the fixed yaw machine, yaw error is shown to be a significant source of fatigue damage. Differences in the response of rigid and teetering hub machines make the latter type of turbine more susceptible to rare, extreme load events, primarily because normal operating loads are reduced by design. Hence, these extreme load events prove to better serve as design drivers for teetered turbines, whereas normal operation loads are found to be the greater concern for rigid hub machines. |
