Modeling of uncertainties for wind turbine blade design

Wind turbine blades are designed by a combination of tests and numerical calculations using finite element models of the blade. The blades are typically composite structures with laminates of glass-fiber and/or carbon-fibers glued together by a matrix material. This paper presents a framework for stochastic modelling of the load bearing capacity of wind turbine blades incorporating physical, model, measurement and statistical uncertainties at the different scales and also discusses the possibility to define numerical tests that can be included in the statistical basis. The stochastic modelling takes basis in the JCSS framework for modelling material properties, Bayesian statistical methods allowing prior / expert knowledge to be accounted for and the Maximum Likelihood Method. The stochastic framework is illustrated using simulated tests which represent examples relevant for wind turbine blades.