ECN publication
Ontwikkeling en validatie van een stochastisch windbelastingsmodel. Deel 1. Windmodellering
Published by: Publication date:
ECN 1-12-1996
ECN report number: Document type:
ECN-C--96-096 ECN publication
Number of pages: Full text:
66 Download PDF  

The considerations that led to the choice of a stochastic turbulencemodel for normal wind conditions are described in terms of spectral density functions and spatial coherence functions for the three components of the wind speed vector. This turbulence model should be used in design load calculations on horizontal axis wind turbines. The model should comply with the IEC 1400-1 standard and should produce design life time fatigue load spectra which are at least as severe as the corresponding load spectra based on the Dutch 'Handbook for Wind Turbine Design Version 3'. For the longitudinal component the IEC standard proposes three combinations of spectral density function and coherency function, which are not interchangeable. Based on a qualitative evaluation of the spectra of rotationally sampled wind speed it is concluded that the Von Karman spectral model is not adequate for fatigue life time calculations, leaving only the exponential coherency function in combination with either the Kaimal spectrum or the Von Karman spectrum as an acceptable choice. Based on validation calculations, reported in part 3 of this report, it is finally concluded that the Von Karman spectrum in conjunction with the exponential coherency function should be used. The IEC standard only specifies spectral density functions for the transverse and vertical component. The missing spatial coherency functions for these in-plane components are taken from other sources, most notably ESDU. In addition, the complete set of spectral functions of the Von Karman isotropic turbulence model is derived. This enabled the assessment of the importance of cross couplings between the three components and showed how the direction of the separation influences the coherency functions. It is concluded that only the cross coupling between the u- and w-component is of importance. The direction of the separation does have a significant influence on the coherence. Because a uniform coherence facilitates efficient simulation some measures are taken to eliminate the directional dependence of the coherence. To ensure good quality wind simulation guidelines are given to determine values of important parameters, e.g. the length of the wind field simulation, the time step in the wind speed time series and the number of points in the rotor plane at which the wind speed time series are generated. In the second volume the aeroelastic models of three wind turbines are constructed and calibrated. In the third volume these models are used to validate the stochastic turbulence model proposed in this volume. 12 figs., 1 tab., 2 appendices, 24 refs.

Back to List