WindTurbine

From DIT Experimental Gaming Group

Contents 1 Introduction 2 Methodology 3 Conclusions 4 Future Work

Introduction

Virtual Wind Turbine Model

The aim of this project is to try to measure, as accurately as possible, the amount of power a virtual wind turbine will produce within a virtual wind tunnel created using an open source physics engine called Open Dynamics Engine (ODE). The turbine design will be created using a 3D modelling tool called MilkShape3D. The results of the simulation will be compared to theoretical turbine results using mathematical formulas applied in industry to see how accurate the simulation is.

Methodology

A area of virtual wind was created. This wind was based on the models used in industrial level computational fluid dynamic program but scaled down so that it could be run on an ordinary PC. An approximate version of the Navier-Stokes equations used applied to the wind. A 3D wind turbine model was then created and was exposed to the virtual speed. The simulation was run at different wind speed and the amount of power the virtual turbine created was output to a spreadsheet file, where it could be used to form graphs that compared the results of the simulation with the results the turbine is expected to create in theory.

Conclusions

Results from Experiments

A series of tests were performed using different wind speeds. The power output of the each of the simulations when the turbine had reached a steady speed was plotted on a graph against the expected energy output derived using aerodynamic formulas. When the power output from the virtual turbine was scaled correctly, it was found that the power output of the turbine very closely resembled the expected power output. This proved that the simulation accurately predicted the correct power output of a turbine.

Future Work

1. The current simulation model will only accurately predict the results of horizontal axis wind turbines that depend on drag to cause the majority of their rotation. Lift should also be modelled in the simulation, as this would then cover all possible turbine models.
2. A user interface could be added into the simulation to make it easier to use and cause it to reach a larger audience.
3. Only one type of 3D model works for the simulation, .x files. Further file types could be implemented.
4. A genetic algorithm that takes a users turbine model and changes the pitch or length of the blade and tests if the new model produces more power could be implemented. This would mean that new turbine models could be discovered simply through computer brute-forcing.