Structural Optimization of a Horizontal Axis Wind Turbine Blade

Abstract

The main purpose of this project was the creation of a tool, based on the research performed by Scheldbergen [25] and Roscher [24], capable of performing Finite Element Analysis (FEA) on a Horizontal Axis Wind Turbine (HAWT) blade and minimize its weight by altering the various thicknesses locally, always satisfying a number of structural constraints. The first goal was the development of a geometry creation algorithm based on a Non Uniform Rational Basis Splines (NURBS) algorithm developed by Ferede [14], with a more complex spanwise/chordwise discretization technique in order to more accurately design a discretized HAWT blade. The second goal was the alteration of the material creation algorithm developed by Scheldbergen [25], so that different stacks at various blade locations can be assigned. Moreover, the aerodynamic loads generation algorithm of Scheldbergen [25] was modified to apply for different airfoils along the blade span. Also, Blade Element Momentum (BEM) theory was used for the calculation of the aerodynamic loads and the software XFOIL was used for the aerodynamic analysis of the airfoils (pressure distribution on the airfoil at corresponding angles of attack and Re numbers). An initial HAWT blade design was created, matching the SANDIA 5 MW wind turbine blade [23] geometry and materials. Furthermore, the apwise/edgewise stiffness, moments of inertia, mass and mode shapes were also validated. Additionally, the SANDIA blade was analyzed in DNV GL Bladed in order to validate the aerodynamic loads on the blade. Following, the tool was designed to process the geometry, the materials and the gravitational/ centrifugal/aerodynamic loads of the blade and translate them to an input for the FEA software MSC Nastran for Linear Static Analysis (SOL 101), Buckling Analysis (SOL 105) and Modal Analysis (SOL 103). The post processing algorithm of Scheldbergen [25] was modified to apply for a HAWT (instead of a VAWT), transforming the results of the FEA analysis into optimization constraints (ultimate load, buckling, tip deflection and fatigue). The final part of the algorithm involved the minimization of the blade's mass by altering the thickness at various locations, using the MATLAB function fmincon. Concluding, a case study of a smart rotor blade equipped with trailing edge ap (TRF) was examined. A _ 2 flap was used at 90-100% of the chord and 78-98% of the span. Two kinds of optimization were performed for the HAWT and smart rotor scenarios; one that included ultimate stress, tip deflection and fatigue constraints (UL-TD-FA) and one that also included buckling constraints (UL-BU-TD-FA). It was found that the buckling constraint was the limiting factor of the optimization and the fatigue constraint was the least dominant. The initial design of the smart rotor blade showed a 16.5% decrease of the maximum fatigue damage and a 2% decrease in the maximum stress. The optimized smart rotor blade mass was 0.5 % lower in the UL-TD-FA scenario and 2.5% in the UL-BU-TD-FA scenario compared to the HAWT blade. The tool created is capable of creating a flexible and fairly accurate representation of a HAWT and smart rotor blade. It also provides the ability to structurally analyze the blade under the most significant structural criteria and minimize the blade's weight under specific loading conditions. The tool constitutes a good starting point for more detailed structural analysis of smart rotors and more complex structural optimization of HAWT and smart rotor blades.