Model Form Error of Alternate Modeling Strategies: Shell Type Wind Turbine Blades

Abstract

In the U.S. there is a need for the Wind Energy sector to reduce maintenance costs and increase the continuous availability of wind turbines to compete with other forms of energy. In this respect, reliable modeling and simulation techniques play an indispensible role in allowing a better understanding of wind turbine blade vibrations. The purpose of this paper is to quantify the degrading effects of simplifications to the material cross section of a finite element (FE) model of a Suzlon S88-2.1 MW wind turbine blade. First, a code verification study with ANSYS Shell281 elements is performed to verify the element type selection. Next, a three-dimensional model of the blade is developed using NuMAD, pre-processing software and analyzed in ANSYS, FE modeling software. A solution verification study is performed to identify the mesh size that will yield converged solutions with manageable demands on computational resources. Five constitutive models are developed by incrementally simplifying the cross section of the blade. The most sophisticated model is idealized as the baseline to create synthetic experimental data, utilizing orthotropic materials and a composite cross section, in comparison to the least involved model which has an isotropic, smeared cross section. This manuscript explores the extent to which errors are increased as the model is simplified in order to reduce the number of input calibration parameters and computation time. Considering the baseline, the model form errors associated with the four models of lesser sophistication are estimated. The estimates of model form error conveniently quantify the effects of simplifications to the entire domain of a FE model wind turbine blade. Model form error will help to identify the tradeoffs between varying levels of sophistication in the physics, and will thus aid in future attempts to develop dependable FE models of shell type wind turbine blades.