These guidelines are intended to be a design aid for wind turbine designers. The complete list of guidelines is provided below. • DG01 Loads Analysis • DG02 Strength Analysis • DG03 Yaw and Pitch Rolling Bearing Life Design Guideline • DG04 Gearbox Specification • DG05 Control and Protection System Specification and Design All of these documents-with the exception of the present document (DG03)-are in draft form and are available only by request sent to NREL, Sandy.Butterfield@NREL.Gov. Modern wind turbines use large turntable bearings at the root of each blade to enable pitch angle changes and thus aerodynamic performance and load control. Yaw bearings are used for angular realignment of the nacelle into the predominant wind direction. These applications require long periods in nearly stationary positions with large stochastic loads. Due to this demanding load environment and the fact that bearings exist in the critical load path, their design becomes critical to the safety and reliability of most turbine designs. This document attempts to introduce modern bearing-design practice and its relation to the unique requirements of wind turbine applications. The fundamental theory presented here is available in other textbooks referenced in this document, and included in the References section found at the end of this report.

An interdisciplinary systems analysis is presented for high-speed gas turbine engine mainshaft roller bearings which will enable the designer to meet the demands for ever higher rotative speeds and operating temperatures. The latest elastohydrodynamic experimental traction data are included. Analytical results cite a need for better definition of the rolling friction portion of the total traction. A fluid mechanics model for the detailed analysis of fluid drags is developed based upon a turbulent vortex-dominated flow and includes the effect of lubricant flow through the bearing. A complete thermal analysis including dynamic and thermal effects upon bearing dimensions and resulting clearances is also included. Heat transfer coefficients are given in detail. Shaft power loss and cage slip predictions as a function of load, speed, and lubricant supply correlate well with available experimental data.

The results of life tests of 388 caged, needle-roller bearings under combinations of load, speed, and amplitude of oscillation are analyzed with Weibull statistics. The rolling bearing theory of Lundberg and Palmgren is extended to cover oscillating roller bearings and is favorably compared with the life tests. A formula for basic dynamic capacity in oscillation is developed for amplitudes of oscillation which are less than the critical amplitude. Capacity formulas are also presented for both ball and roller bearings for any amplitude of oscillation.

This paper presents the analysis of the stress field in a hollow sphere in the vicinity of the contact area. The sphere is subjected to a normal load applied through a flat plate. The elastic contact shape and extent are developed for a load of 1000 lb applied to a 1-in-dia hollow ball with a 0.08-in-thick wall. Hollow ball shell bending stresses have a significant effect upon the subsurface stress field. Fatigue life estimates for the hollow ball vary significantly depending upon the selection of decisive stress amplitude. Comparison of the maximum value and location of the reversing orthogonal subsurface shear stress with solid ball data according to the Lundberg-Palmgren dynamic life theory predicts a 91.6 percent life reduction for the hollow ball contact. The use of the unidirectional subsurface shear stress results in a prediction of hollow ball contact life over 30 times the solid ball contact life.

A Dynamic capacity for oscillating rolling bearings was published in 1968 and correlated with available laboratory fatigue life data. That development of the Dynamic Capacity extended the classic fatigue life theory of Lundberg and Palmgren (1947 and 1952) to oscillating rolling bearings. The calculation of the Dynamic Capacity is simplified as a modification of present ABMA and ISO load rating and life standards for continuously rotating rolling bearings. The simplified formulas agree with the Harris, 1991, text book formulation for oscillation amplitudes (greater than the critical amplitude) which cause an overlapping of stressed contact areas by adjacent rolling elements. Oscillation amplitudes less than the critical amplitude result in separate, discrete contact areas on each raceway. Use of the Harris equations will lead to overestimation of the fatigue for oscillation amplitudes which are less than the critical amplitude.