ABSTRACT
Computational Fluid Dynamics (CFD) software was used to compare the performance of a handmadewind turbine blade with that of a conventional factory made model. The geometry wassimplified to 2D aerofoils and the surrounding flow field was analyzed at a Reynolds’s number of80,000. It was found that the lift/drag characteristics of the two aerofoils across a range of angles ofattack were virtually identical, meaning that the torque force exerted on the wind turbine bladeswould also be identical and therefore as would the power outputs of the two turbines.Small scale wind turbines can be used to provide powerto remote areas of the developing world that are faraway from any existing electrical grid system. Theelectricity they supply can be used to provide light in themornings and evenings which can allow children tostudy and further their opportunities in later life oradults to continue working and provide that little bit ofextra income for their families that could allow them towork their way out of poverty. Unfortunately, at a costof thousands of pounds, factory built small scale windturbines are expensive, even for reasonably well offcitizens of the developed world.However, it is possible to build smallscale wind turbines by hand, using basic workshop toolsand techniques .This analysis will give the clear idea about how the hand made wind turbines are efficient as compared to factory made turbines.
A BASIC HOME MADE TURBINE
Fig.1
BASIC AERODYNAMIC THEORY OF HORIZONTAL AXIS WIND
TURBINES (HAWTS)
Lift-based Horizontal Axis WindTurbines (HAWTs) have todaybecome the standardmechanism for harnessing thepower in the wind andconverting it into electricity.They evolved from the graingrinding drag-based windmillsof yesteryear. As aeronautics took off during the last centurywith the commercialisation ofaeroplane technology, moreand more became known aboutwing technology. The threebladedwind turbines we see allaround us today evolved fromaircraft wing theory and areessentially three wings boltedonto a generator with acommon central axis. They uselift from the oncoming wind torotate themselves about thiscentral axis, which pointshorizontally into the wind,hence the name HAWT. VerticalAxis Wind Turbines (VAWTs) are also used, but the higherefficiencies of HAWTs have made them the dominant technology in today’s society.
Fig.2
THE AEROFOIL
An aerofoil (or airfoil in the USA) is a 2D shape capable of producing a reactive lift force when inmotion relative to the surrounding air. Most commonly known as the cross-sectional profile of anaircraft’s wing.
Fig.3
LIFT AND DRAG, THRUST AND TORQUE
Fig.4
Figure shows that how the lift and drag forces are defined on an aerofoil as forces perpendicular andparallel to the airflow. As the air travels over the top of the aerofoil, it accelerates and consequentlypressure decreases in this area. Lower pressure on the top of the aerofoil than the bottom creates asuction force called lift. Lift is the force that keeps aeroplanes in the sky. Drag on the other hand,acts in the same direction to the airflow and is generally considered a nuisance, as for example in anaeroplane, extra fuel must be used to overcome the drag forces. Drag forces arise mainly from friction between the viscous fluid and the surface of the aerofoil (skin friction drag) and thedifference in pressure between the leading and trailing edges of the aerofoil (form drag).
Fig.5
When comparing different aerofoils, it is often more useful to look at lift and drag coefficients ratherthan the total lift and drag forces on acting on the aerofoil. Lift and drag coefficients are nondimensionalnumbers used to quantify the amount of lift or drag on a given aerofoil under a givenset of flow conditions, e.g. Reynold’s number or AoA.
The Tip Speed Ratio (TSR) is a commonly used parameter in wind turbine design and relates thespeed at which the blade tip is travelling (ωr) to the velocity of the oncoming wind:
When calculating the power output of a wind turbine, it is often more relevant to resolve theresultant aerodynamic force into torque and thrust rather than lift and drag. Torque is the useful
force that causes the blades to rotate, whilst thrust is the redundant component that merely pushesagainst the tower. The power produced by a wind turbine can be calculated very simply as follows:
BOUNDARY LAYERS
The boundary layer is a thin layer of fluid near the surface of an aerodynamic body which is affectedby the friction between the surface of the body and the viscous fluid flowing around it. This frictionresults in the force of skin friction drag. The properties of the boundary layer depend on theReynold’s number of the overall flow and on the local conditions. The properties of the boundarylayer around an aerofoil can be grouped together into three categories:
Laminar flow – at low Reynold’s numbers layers of fluid slide smoothly over each other in a thinboundary layer resulting in low skin friction drag. The flow over an aerofoil will usually start offlaminar before transitioning into turbulent as it travels away from the leading edge. The position ofthe transition point is dependent on Reynold’s number and the localised flow conditions.
Turbulent flow – at higher Reynold’s numbers, or after the transition point, the flow near the surfacebecomes unstable and the boundary layer thickens, leading to higher skin friction drag.
Separated flow – a high adverse pressure gradient occurring towards the trailing edge of the aerofoilcan cause the flow to separate. Flow near the surface reverses direction and a large turbulent wakeis created from swirling eddy currents that dramatically increases skin friction drag.
REYNOLD’S NUMBER
Reynold’s Number is a ratio of the inertial forces to the viscous forces acting within a fluid. It can beused to characterise flow types: for example laminar flow occurs at low Reynolds numbers, whereviscous forces are dominant, and is characterised by smooth, constant fluid motion, while turbulentflow occurs at high Reynolds numbers and is dominated by inertial forces, which tend to producerandom eddies, vortices and other flow instabilities.
INTRODUCTION TO COMPUTATIONAL FLUID DYNAMICS (CFD)
Computational Fluid Dynamics (CFD) is a powerful tool used to model the real life behaviour offluids. It allows the optimisation of design parameters without the need for the costly testing ofmultiple prototypes. What is more, it is also a powerful graphical tool for visualising flow patternsthat can give insight into flow physics that otherwise would be very difficult and costly to discoverexperimentally, if possible at all. Governing equations exist to model fluid behaviour, but it is notalways possible to apply them to many of the complex flow patterns we see in the real world directlyas there would be too many unknown variables. However, CFD involves creating a computationalmesh to divide up real world continuous fluids into more manageable discrete sections. Thegoverning equations for fluid flow can then be applied to each section individually, but as theproperties of each section are inevitably linked to its neighbouring sections, all the sections can besolved simultaneously until a full solution for the entire flow field can be found. This methodobviously requires a huge amount of computational power, nevertheless with the advancement ofmodern computing, solutions that would take months to compute by hand can now be found inseconds using nothing more than an ordinary desktop or laptop computer.
THE MODELLING PROCESS
The modelling process consists of first taking the real world fluid geometry and replicating this in thevirtual environment. From here, a mesh can be created to divide the fluid up into discrete sections.Boundary conditions must then be entered into the model to designate parameters such as the type of Conventional aerofoil geometries, with their characteristics and applications.of fluids to be modelled or the details of any solid edges or flow inlets/outlets. The simulation is thenready to be run and when a converged solution is found, it must be carefully analysed to establishwhether the mesh is appropriately modelling the flow conditions. Generally, some form of meshrefinement will be necessary to put in further detail around the areas of interest.
CFD FOR WIND TURBINE ANALYSIS
CFD allows virtual experimentation with and consequently optimisation of the design parameterssuch as airfoil shape or angle of attack across a wide range of operating conditions. It is veryattractive to industry as it saves both time and effort during the design process when comparedalongside traditional experimental methods. However, the degree of confidence in the results isdependent on many factors and as a result; data should be compared with and validated againstexperimental findings wherever possible.
MODELLING SOFTWARE
The mesh generation programme ICEM CFDwas used in conjunction with the solver
FLUENT VERSION 6.3.26to perform the CFD analysis for this project.
MODELLING STRATEGY
A 2D model of an aerofoil section from the blade was created, with the aim ofcalculating lift and drag data for the aerofoil at varying AoAs. This data can give an estimation of therelative performance of the modelled wind turbines. The aerofoil at the blade tip was chosen as thisis the part of the blade that generates the most lift and therefore its performance is most critical tothe overall power output of the turbine. Varying the AoA simulates the varying wind conditions thatthe turbine is likely to experience in service.
2D MODELLING DOMAIN
In order to accurately simulate free-stream conditions, a far-field boundary was used at a distance ofat least 12 chord lengths from the aerofoil surface to create the control volume for the analysis.To model varying AoAs, the free-stream airflow was rotated, whilst the aerofoil remained horizontal.A wind velocity of 4.5m/s with a TSR of 5.5 was used, giving an inlet velocity of 25.155m/s. As thegeometry was relatively simple, a structured quadrilateral mesh was used to maximise the accuracy of the model FIGURE 6 IS FACTORY MADE AEROFOIL.
FIG.6
FIG.7 HOME MADE AEROFOIL
TURBULENCE MODEL
A number of different turbulence models were suitable candidates for modeling the flow over a 2Daerofoil. The k-ε RNG and k-ω SST models are both popular choices, however during preliminarymodelling it was found that the S-A (Spalart-Allmaras) model most accurately predicted the lift anddrag characteristics of the NACA0012 aerofoil. The S-A model was designed specifically for low-Reynold’s number aerospace calculations and has been shown to give good results when simulatingboundary layers subjected to adverse pressure gradients.. Dueto the turbulent nature of the physical flow conditions that are being modelled close to and beyondthe stall angle, complex time-varying simulations would be required to correctly simulate thisbehaviour. As a result, the data obtained from this simple steady-state model in this region cannotbe considered reliable. With regards to the drag coefficient, it demonstrates that theS-A model gives a far better match to the experimental data. As a result, it was decided to use the SAturbulence model for the main analysis.
RESULTS
Using the validated model, simulations were run of the factory made wind turbine blade tip aerofoilalongsideof hand made wind turbine blade tip aerofoil. The angle of attack was varied between 0°and 15° at a Reynold’s number of 80,000. This would seem to suggest that wind turbines using either of these profiles wouldproduce similar amounts of power.
DRAG CO-EFFICIENT COMPARISON:
FIG.8
LIFT CO-EFFICIENT COMPARISON:
FIG.9
VIOLET FACTORY MADE WIND TURBINE BLADE TIP AEROFOIL
ORANGE HOME MADE WIND TURBINE BLADE TIP AEROFOIL
EVALUATION
Although the homemade aerofoil may seem to have an identical performance to the factory made aerofoil, the model used was very simple and a number of factors that will influence the performance of a wind turbine blade in real life were not included, for example:
(1)3D effects – Real wind turbine blades are both twisted and tapered and encounterincreasing relative wind speeds towards the tip due to the rotation of the blade. Vortices arealso created at the ends of the blades due to the difference in pressure between the upperand lower surfaces of the aerofoil. In order to correctly model the geometry of a windturbine blade, a 3D model would need to be built.
(2)Transition point – The point along the surface of the aerofoil at which the flow transitionsfrom laminar to turbulent is of critical importance to determining the drag on the aerofoil. To date, no computational techniques are capable of predicting the location of this point and experimental measurements must be taken to determine its location. As a result, the entire boundary layer was modelled as turbulent and the drag is likely to be an overprediction.
(3)Idealised geometry – The wind turbine blade tips were modelled as ideal aerofoilgeometries, however this would not be the case in the real world. In particular, a hand-made wind turbine blade is likely to be full of defects arising from poor manufacturing technique, especially at the tip where the size of the aerofoil relative to the size of the tools is smallest.
These defects, although they may seem small, could have a critical impact on the aerofoil’s performance as they could trip the flow from laminar to turbulent and consequently increase drag. The surface roughness of the wind turbine blades was also not modelled, which could also similarly affect the drag.
(4)Accuracy of CFD – As shown by Figure 15, the model still has a significant degree ofinaccuracy, especially near and above the stall angle. Further modelling is required to increase the accuracy of the model, in particular unsteady simulations to more accurately determine performance around stall angle.
(5)Experimental validation – Although the model was validated using the homemade aerofoil,the flow physics may be slightly different around the two test aerofoils and as a result experimental data for these would give more confidence in the results.
CONCLUSION
FIG. 8 & 9 clearly indicates that the performance of the two wind turbine blade tip aerofoils isvirtually identical. This result is highly unexpected as the geometries of the two aerofoils are very different. This would seem to suggest that wind turbines using either of these profiles would produce similar amounts of power. The homemade aerofoil is a far simpler shape to manufacture, as the lower surface is effectively a flat surface and therefore it is a far more appropriate design for low cost hand manufacturing as virtually no performance is sacrificed.
Homemade wind turbine has been shown to have comparable performance to that of a
Factory made wind turbine. In the simplified model of the aerofoils at the blade tips, both exhibited virtually identical lift and drag characteristics, implying that the torque force exerted on the blades and consequently the power produced by the turbine would also be identical. However, the simple model neglects many important factors such as 3D effects, geometric variations deriving from manufacturing defects and the location of the transition point. As a result, further modelling and/or experimental work is required to give more confidence in the results of this study.