Session: 14-06-01: Applied mechanics and materials in micro- and nanosystems

Paper Number: 166961

Design of Aerodynamically Efficient Offshore Wind Turbine Blades 

Offshore wind turbines are now a vital element of sustainable power generation due to the growing need for renewable energy. For deep-water applications, the NREL 15 MW wind turbine is designed to outperform available offshore wind turbines based on power generation and structural reliability. In this study, a thorough Fluid-Structure Interaction (FSI) simulation of the NREL 15 MW wind turbine blade is presented to assess its structural and aerodynamic performance in extreme environmental settings, further enhanced through Machine Learning (ML)-based optimization.

The simulation implements a one-way fluid-structure interaction framework of Finite Element Analysis (FEA) and Computational Fluid Dynamics (CFD) to capture the complex interplay between structural deformations and aerodynamic forces. The CFD domain is modeled in ANSYS Fluent using the Reynolds-Averaged Navier-Stokes (RANS) approach with the k-ω SST turbulence model to resolve the complex flow behavior around the blade, and the structural response is calculated in ANSYS Mechanical, where the blade is modeled as a composite structure with realistic material properties. In one-way FSI modeling, the pressure information generated in CFD Fluent is mapped node-to-node to transfer pressure forces into the structural module to investigate the stress and deformation under load.

To improve computational efficiency and design optimization, a Machine Learning-based surrogate model has been introduced. Training data is generated from high-fidelity simulation and experimental results from NREL database, and advanced ML algorithms, including Artificial Neural Networks (ANN) and Gaussian Process Regression (GPR), are employed to predict aerodynamic forces, stress distributions, and blade deflections. These ML models significantly reduce computational cost by allowing rapid evaluation of multiple design iterations without the need for repeated high-fidelity simulations.

The simulation accounts for operational wind speeds and extreme load conditions, including hurricane-level gusts, to assess the blade's performance and reliability. The results reveal significant variations in lift and drag forces across the blade span due to deformation-induced changes in the angle of attack. Structural analysis highlights critical stress concentrations near the root region, emphasizing the need for reinforcement in high-load areas.

The study indicates aerodynamic efficiency loss due to structural deformation occurs above the rated wind speed (above 20 m/s), which is crucial for optimizing blade design. The ML-based optimization framework helps mitigate this issue by predicting and adjusting critical design parameters to minimize performance degradation. Additionally, ML-driven fatigue analysis is conducted to estimate the lifespan of the blade under cyclic loading conditions, providing a more accurate assessment of long-term structural integrity.

In this study, the integration of FSI and ML in the design phase enables the prediction of real-world performance, reducing uncertainties in extreme weather conditions. This study underscores the importance of combining FSI simulations with ML-driven optimization in offshore wind turbine design, offering a robust methodology for assessing structural integrity and aerodynamic performance. The results contribute to the development of more reliable and efficient wind energy solutions, supporting the global transition to sustainable energy sources.

Presenting Author: Gazi Raihan University of New Orleans

Presenting Author Biography: Gazi Raihan is a graduate student who is doing his Ph.D. in Mechanical Engineering at the University of New Orleans. Currently, he is working under Dr. Uttam Chakravarty, Associate Professor, Department of Mechanical Engineering at the University of New Orleans. His current research focuses on offshore wind turbine blade design.

Authors:

Gazi Raihan University of New Orleans
Uttam Chakravarty University of New Orleans