Open AccessThe effect of mean stress on damage predictions for spectral loading of fibreglass composite coupons
Author(s) -
Sutherland Herbert J.,
Mandell John F.
Publication year - 2004
Publication title -
wind energy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.743
H-Index - 92
eISSN - 1099-1824
pISSN - 1095-4244
DOI - 10.1002/we.125
Subject(s) - exponential function , diagram , bilinear interpolation , residual , structural engineering , stress (linguistics) , turbine blade , composite number , residual strength , mathematics , turbine , engineering , statistics , algorithm , mathematical analysis , mechanical engineering , linguistics , philosophy
Abstract In many analyses of wind turbine blades the effects of mean stress on the determination of damage in composite blades are either ignored completely or characterized inadequately. Mandell et al. have recently presented an updated Goodman diagram for a fibreglass material that is typical of the materials used in wind turbine blades. Their formulation uses the MSU/DOE fatigue database to develop a Goodman diagram with detailed information at 13 R‐values. Using these data, linear, bilinear and full Goodman diagrams are constructed using mean and ‘95/95’ fits to the data. The various Goodman diagrams are used to predict the failure stress for coupons tested using the WISPERX spectrum. Three models are used in the analyses. The first is the linear Miner's rule commonly used by the wind industry to predict failure (service lifetimes). The second is a non‐linear variation of Miner's rule which computes a non‐linear Miner's residual strength based upon an exponential degradation parameter. The third is a generalized non‐linear residual strength model that also relies on an exponential degradation parameter. The results illustrate that Miner's rule does not predict failure very well. When the mean full Goodman diagram is used, the non‐linear models predict failures near the mean of the experimental data, and when the 95/95 Goodman diagram is used, they predict the lower bound of the measured data very well. Published in 2004 by John Wiley & Sons, Ltd.