Submitted in fulfillment of the requirements for the degree of Master of Science in Mechanical & Aerospace Engineering

I hereby declare that this manuscript, entitled “A VLES and Two-Way Dynamic Fluid-Structure Interaction Study of Wind Turbines”, is the result of my own work except for quotations and citations, which have been duly acknowledged. The first method aimed to create a 2-way dynamic FSI of wind turbines with the application of the advanced VLES turbulence model.

Introduction

Background
Aims and objectives
Scope and Constraints
Novelty and Contributions
Problem Statement
Hypothesis

This approach can lead to a detailed study of wind turbine vibration and noise and multidisciplinary design optimization (MDO). Implementation of preCICE to combine DAFoam and CaculiX with a shell element structural model that can be used for future transient multidisciplinary design optimization (MDO) and for vibration and noise reduction (previously [7] foam extend 4.0 was used which can only use solid elements and cannot apply to MDO).

Literature Review

Overview: Fluid-Structure Interaction approaches

A range of model building techniques and coupling strategies are currently available for investigating the FSI behavior of blades. ANSYS and UG modeling software were used to complete the modeling of the blade object.

Turbulence Models

RANS Model
LES Model
VLES Model

The largest eddies contain most of the turbulent energy and are responsible for the dominant structures in the flow. LES simulations are computationally demanding due to the need to explicitly resolve large structures.

Methodology

ALE method

PISO algorithm

Software overview

Coupling in preCICE

It requires Linux (Ubuntu) operating system, but can also be used on Windows by creating a subsystem for Linux. DAFoam will be used for implementing VLES in it with code modification and dynamic link library building. PreCICE is an open source software used for coupling used for multiphysics analysis such as fluid-structure interaction simulations [30].

The preCICE developers design adapters for fluid solvers to couple them with structural solvers for FSI. By applying a new methodology to the entire field of study, it will be possible to overcome the problems of wind turbine analysis as currently performed and gain a deeper understanding of the physics of the actual flow. A Linux operating system is required to use the required software.

It is done using WSL (Windows Subsystem for Linux), which allows installing the Ubuntu 20.04.4 terminal environment. As can be seen when executed, it waits for the fluid solver to also start to perform FSI. The same can be seen in Figure 3.3, where the fluid solver is waiting for the solid solver to start.

Results and Discussion

FSI Test Cases

Elastic 3D Tube
Perpendicular flap

The obtained results of this FSI example showed that the connection of OpenFOAM and CalculiX was successful and working correctly. While the fluid participant reads the displacement data, the solid participant reads the forces at the interface. As can be seen, the fluid area has moved except for the area where the flap was installed.

This case also shows that the link in preCICE works correctly and can be applied to further work.

Figure 4.3. Flow traveling through the tube.

NREL Phase VI wind turbine

Fluid model
Solid model

The blade is fixed in the middle to avoid translation and allow rotation, for which the rolling frame approach is applied. For solid solvent mesh generation it was necessary to use ANSYS software as well as Salome. The mesh was then transferred to the Salome software for defining the necessary nodes and transferring the mesh to CalculiX.

Nevertheless, the mesh was changed several times due to the error that occurred while running the solver. By applying the force to the tip of the blade, an attempt was made to create the rotation and the. One node was created to fix the leaves from translation in x, y and z axes and allow rotations.

Another node is created for applying the force, which is shown in Figure 4.15.

FSI Trials

FSI run

Initial trials
Misalignment of the geometries
Geometry with a spar
Geometry with three spars
Software Limitation

When the deformation is too high, OpenFoam exits with the floating point exception error. It was found that the geometry in the structural solver was significantly larger than that in the fluid solver. Then, to make the meshes exactly the same and not have a misalignment problem, it was decided to transfer the OpenFoam mesh to the Calculix mesh.

It was achieved by using the FoamToFluent tool in OpenFoam and by saving the Ansys Fluent mesh in UNV format for CalculiX. Prepomax is a software tool that can be used with the finite element analysis (FEA) software CalculiX. The next step was to reduce the deformations by using the blade with the spars inside.

After several attempts and spending a lot of time, it was found that the combination of preCICE software with OpenFOAM and CalculiX for two-way dynamic fluid-structure interaction (FSI) simulations involving rotational motion encountered challenges and limitations. Although a powerful tool for static structural simulations, it may have limitations when dealing with dynamic or rotational motion in the context of FSI simulations. Rotational motion introduces additional complexities such as the need to account for centrifugal forces, mesh deformation, and precise coupling to the fluid domain.

Low-Fidelity FSI

Overview

They are commonly used in inviscid flow simulations, where the fluid viscosity is assumed to be negligible. Potential flow theory assumes that the flow is irrotational and incompressible, allowing further simplification of the governing equations [35]. Solving the potential flow equations can provide valuable insights into the overall flow behavior, but it neglects important aspects such as boundary layers and flow separation that are crucial in high-fidelity simulations.

Note that these low-accuracy flow field equations provide a basic representation of fluid flow and may not capture all the complexities present in real-world scenarios.

QBlade

Aero-elastic coupling
Lifting Line Free Vortex Wake Method
The Chrono Library

Initially, a convergent circulation distribution search is performed for the bound vortex of the vane. In this iteration, the unsteady aerodynamic model along with various corrections is used to calculate the lift and drag characteristics of the airfoils. The airfoil lift, drag, and moment parameters are used to interpolate blade plate forces and moments to the structural dynamics discretization from the aerodynamic discretization.

One of the aerodynamic analysis methods implemented in QBlade is the Lifting Line Free Vortex Wake (LLFVW) method. The LLFVW method in QBlade takes into account the influence of the bound vortices on the flow field and calculates the induced velocities caused by the vortices emitted by each lifting line element. Based on the effective speed and the local geometry of the blade section, QBlade calculates the lift and drag coefficients for each element.

A vortex model of the flow field and the invisible potential flow concept serve as the basis for the LLFVW technique [30]. A combination of the free stream velocity 𝑉∞, the induced velocity 𝑉𝑖𝑛𝑑, and the blade motion 𝑉𝑚𝑜𝑡 yields the relative velocity 𝑉𝑟𝑒𝑙. In addition, the Chrono DLL is linked to the QBlade source code, allowing use of the Chrono module and its associated functions within the QBlade simulation.

Figure 6.1. The flowchart of aero-elastic coupling [36]

Results in QBlade

NREL 5 MW wind turbine
NREL Phase VI wind turbine

Power factor is a measure of how efficiently a wind turbine converts the kinetic energy of the wind into electrical energy. It is defined as the ratio between the actual power obtained by the wind turbine and the maximum possible power that can be obtained from the wind [37]. The maximum possible power that can be extracted from the wind is given by the Betz limit, which is approximately 59.3% (or 0.593) of the total kinetic energy in the wind.

The tip speed ratio is a dimensionless parameter that compares the rotational speed of the wind turbine rotor to the wind speed at the tip of the rotor. It is defined as the ratio between the tangential speed of the tip of the rotor blade and the wind speed [37]. Where: ω is the rotational speed of the wind turbine rotor, R is the radius of the rotor, and V is the wind speed.

The twist angle is found to be higher closer to the center of the blades by almost 14 degrees, which is illustrated in Figure 6.7. The geometry of the NREL Phase VI wind turbine is shown in the following figure. The trends in the results are similar, despite the parts having the highest twist angle.

CFD Simulation

Mesh

In Figure 6.13, the results for the yaw angle from simulation and experiment are plotted together for comparison.

Simulation results

The simulation can also be attempted with higher wind speed values to analyze the turbulence eddies. This structure can later be compared with the results for the implementation of the VLES model in future works. The figure below shows the pressure coefficients obtained by CFD simulation of the URANS model, which are then compared with experimental data at different ranges from the work of Hand et al.

It is clear that the pressure coefficients from the simulation and those from the experiment agree well with each other.

Figure 7.4 represents the vorticity contours around the blades. The simulation can also be tried with higher windspeed values in order to analyze the turbulence eddies

Conclusions & Future works

Fluid-structure interaction analysis of NREL phase VI wind turbine: aerodynamic force evaluation and structural analysis using FSI analysis. High-fidelity 2-way FSI simulation of a wind turbine using fully structured multiblock meshes in OpenFoam for accurate aeroelastic analysis. Simulation of fluid structure interaction of a wind turbine rotor in complex flows, validated by field experiments.

Fluid-structure interaction modeling of horizontal axis wind turbine blades based on CFD and FEA. Effects of wind and rotational speed on fluid-structure interaction vibration for offshore wind turbine blade. A comparative numerical study of four turbulence models for horizontal axis wind turbine flow prediction.

On the Wake Characteristics of the NREL Phase VI Wind Turbine Under Turbulent Inflow Conditions. Implementation, optimization and validation of a nonlinear lift line-free vortex wake module within the wind turbine simulation code qblade. CFD Investigation of the Aerodynamic Properties of a Small Size Wind Turbine from NREL PHASE VI Operating with a Stall Regulated.