Design of a small wind turbine with CFD

Computational Simulation Techniques WS 2015

Final Exam Project

Submitted by: Iqbal Meskinzada

Matriculation #: 2267075
Lecturer: Prof. Dr.-Ing. Rainer Stank
Due Date: 31st of May 2016

Table of Contents
1. Introduction 3
1.1. Design conditions and Airfoils.............3
1.2. Calculation of Angles of Attack and Twist and the Lift Coefficients .............6
2. Building up Geometry ....................6
2.1. Assembling the turbine.............8
3. Meshing...............9
4. Pre-processing.............11
5. Main processing..14
6. Post-processing ..15
7. Theoretical power calculation . ..18
8. Results and Discussions .............18
9. References...19

Hamburg University of Applied Sciences (HAW) |CWi Project

1. Introduction
Computational fluid dynamics, usually abbreviated as CFD, is a branch of fluid mechanics that uses
numerical analysis and algorithms to solve and analyze problems that involve fluid flows. In the
branch of Renewable Energies, a lot of procedures are based on transport processes like fluid or heat
transport or on mixing processes. Therefore, fluid dynamics and especially computational fluid
dynamics are playing an important role. The advantage of computational methods is that they provide
a practical solution of the exact governing equations for almost all engineering problems.
Computational methods can be characterized as an art of replacing the governing nonlinear partial
differential equations by numbers and advancing these numbers in space and or time to obtain a final
numerical description of the complete problem of interest. The end product of computational methods
is indeed a collection of numbers which is in contrast to a closed-form analytical solution. Generally,
the objective of most engineering analysis is a quantitative description of a problem [1].

1.1. Design conditions and Airfoils

This report is a detailed explaination of the designing of a small wind turbine using numerical methods
with ANSYS software package. The primary purpose of this task is to replace the blades on the
existing wind turbine,TESPE. Blades on this turbine were very simple in terms of design, there is no
twisting along the blades, constant angle of attack from the hub to the tip and a uniform cross
section. These properties reduce the performance of the turbine. In this report, the aim is to the
improve the performance of the turbine by changing the blades shape, determine the optimum angle of
attack and twist angle. The rotor blades of the turbine should consist of the two airfoils NREL S822
and S823.

Figure (1): Profile of the airfoil NREL S822, smooth and streamlined, selected for the tip

Since blade tickness decreases from root to the tip [2] and the airfoil NREL S822 is streamlined and
narrow, it is suitable for the tip position and NREL S823, shown in figure (3), for the hub.
Hamburg University of Applied Sciences (HAW) |CWi Project

Figure (2): Polar diagram of airfoil NREL S822, with optimum angle of attack of 4.2
and corresponding lift coefficient of 0.675.

Figure (3): Profile of the airfoil NREL S823. Thicker, selected for the hub

Hamburg University of Applied Sciences (HAW) |CWi Project

Figure (4): Polar diagram of airfoil NREL S823. with optimum angle of attack
of 13.5 and corresponding lift coefficient of 12.8.

Chord length varies over the radius and has to be

determined at the given radii. The chord length between
the given airfoil section at the give radii varies linearly.
For radii smaller than r=0.15R there is no airfoil.
The ratio of the chord length at the hub to the chord
length at the tip is 2.5. The twist of the airfoil sections is
done with respect to a normal axis at 0.25 chord length as
shown in figure (5). The blade radius is 3.8m and the
number of blades are 4.
F igure (5): Positions for twisting

Given tip speed ratio and rated wind velocity are

of attack is defined as.

Blade angel of attack,

Chord length distribution,

and

respectively.The angle

. (1)

. (2)
)

Hamburg University of Applied Sciences (HAW) |CWi Project

1.2. Calculating the Angles of Attack and Twist and the Lift Coefficients
The first step is to determine the optimum value for aerodynamic parameters such as angles of attack,
chords length and twist angles at different rotor diameters. This can be done using the given polar
diagrams of the airfoils and design conditions. The aim is to determine the maximum left coefficients
over various angles of attack such that the ratio of the hub chord length to that of the tip remains less
than 2.5. The lift force increases with the increase in anagle of attack until the blades stall where the
the lift coefficient decreases. Therefore, there is a range for optimum angle of attack. Most often the
best angle of attack lies within a range of (4-16 degrees) for such small size wind turbine. This is a
trial-&-error method to find the best angles from the given polar diagrams. The determined angles are
presented in table (1). Since the blade cross-section will not be uniform, different chord lengths should
be calculated at the given radii usnig the given formula. Assumed lift coefficients and the
corresponding chord length and twist angle are calculated in table (1).
Airfoil
S822
S822
S823
S823

R
3.8
3.8
3.8
3.8

r(R)
3.8
2.85
0.95
0.57

Lambda D
4.4
4.4
4.4
4.4

Z
4
4
4
4

AoA []
4.2
4.2
13.5
13.5

0.674
0.674
1.28
1.28

chord length c(r) AoA Twist

0.4
4.42
0.53
7.22
0.73
17.72
1
31.79

chord lenght ratio

2.498

Table (1): Calculated values for chord lengths and twist angles

2. Building Up the Geometry

To create the geometry of the blades, first of all the data points pertaining to S823 should be imported
in the ICEM and then connected all together to make the first airfoil S823, it consist of five curves.
Then this airfoil needs to be offset by a distance of 950cm from the origin and 20 degrees twisted. The
twist of the airfoil sections is done with respect to a normal axis at 0.25 chord based. Similarly, the
data points for the airfoil S823 are imported and merged with the exiting data points. Airfoil S822 will
be positioned at 2.85m from the center point and offset to the tip. Then it will be scaled and twisted
accordingly. All twist angles and offsets are based on the calculation of table (1).
1

Importing geometric points for both

Scalling and offseting of the airfoils.

Twisting

Figure (6): Building the blade: 6.1: imported data points. 6.2: scaled and offset airfoils, 6.3: twisted airfoils.

Hamburg University of Applied Sciences (HAW) |CWi Project

The next step is to create the surfaces and this can be done from the Create/Modify Surface option.
The following figure shows the generated sufaces over the blade.

d
Figure (7): Blade surface generation

To complete the blade geometry, the above created blade should be replicated around the center point
by 90 degrees to create a four-bladed turbine as indicated in figure (8).

Figure (8): Four-bladed rotor, surface generation

Hamburg University of Applied Sciences (HAW) |CWi Project

2.1. Assembling The Turbine

Next task is to insert this rotor geometry to the old given tespe file ICEM input file Tespe which is a
complete geometry of small-sized wind turbine. However before doing this, the blade geometry in this
file must be removed. In addition to that, there are some other adjustments as well that need to be
done. Gondel needs to be rotated (45 degrees) and inlined with blades direction. The inner and outer
interfaces should be resized such that they accommodate the large blades and leave some space
between the tip and outer interface. Initially there would be an intersection between the lit-Bret and
outer-interface of the blades, so some modifications are needed. After all the required corrections are
made, the geometry is ready for meshing. The completed geometry is presented in the following
figure.

Figure (9): Four-bladed turbine, adjusted into the tespe file

Hamburg University of Applied Sciences (HAW) |CWi Project

3. Meshing
Meshing is one of the most important influences on CFD simulation accuracy since it depends on the
quality of the meshing. Upon completion of the geometry, the next step is meshing. Tetrahedron mesh
type is selected for this particular problem. To get a better control over the meshing process, it is
divided into two parts. The first part is meshing of the environment, the surrounding box including the
interfaces of the rotor. The second part is the blades and the interfaces corresponding to the hub and
rod. The student version of the ANSYS has limitations on the number of meshes (512,000), therefore
some trial and error work is involved until acceptable number of meshes achieved. In the first step,
the surrounding is meshed. In the Global Mesh Setup, a scale factor of 1 and maximum element of
1600 were selected. The maximum mesh sizes are entered in Part Mesh Setup and a total of 120,000
meshes were obtained for the first part. The following figure shows the meshed surrounding.

Figure (10): Wind turbine surrounding area meshed.

Similarly meshing is carried out for the second part which consisted of the rotor as well as the
interfaces covering it. A total of about 263,244 meshes were obtained for the second part. Making a
total of about 383,244 meshes. The following show shows the meshing of the blades and associated
interfaces.

Hamburg University of Applied Sciences (HAW) |CWi Project

Figure (11): Resulted meshed blades, tetrahydrons

Figure (12): Resulted meshed for the Nabel, Holm and Blades

After the meshing is complete and accurate, the solver setup should be set for ANSY CFX, for
writing the solver input for the pre-prosessing phase.

Hamburg University of Applied Sciences (HAW) |CWi Project

10

4. Pre-possessing
After meshing process is completely done and checked. The next step is pre-processing. In general, in
pre-processing stage, the following major steps are involved.
.

Import or Create geometry

Define element type and real constants
Define material properties
Applying Loads
Applying constraints, boundary conditions (fixed edge, simply supported)

However, in our case we have the existing old tespe wind turbine file, in which all these steps have
already been performed and we just need to apply the same properties to our new file. Therefore, a
new case should be introduced into the CFX-pre and the meshed file should be imported. The general
idea in this phase is to create domains for the environment the surrounding box and the wind
turbine. Then inlet, outlet and circumference interfaces should be introduced. After these setting have
made, all the properties of the pre-file should be checked and compared against the original given
pre-file (tespe). Below figures are some of the snaphshots for the applied properties

Figure (13): Velocity inlet and constant pressure outlet on the surrounding area

Hamburg University of Applied Sciences (HAW) |CWi Project

11

Figure (14): Velocity and pressure on the rotor interface

Hamburg University of Applied Sciences (HAW) |CWi Project

12

Below figure indicates the properties that were incorporated into the new wind turbine pre-file which
is the same as the old tesp file.

Figure (15): Applied properties to the new pre-file, barrowed from the old tespe file

After these consideresations are given, we run the simulation process.

Hamburg University of Applied Sciences (HAW) |CWi Project

13

5.

Main Processing
After providing the input for all necessary parameters in Pre-Processing phase, simulation in Main
processing using CFX-Solver Manager shall be initiated. Graphs in the following step are attached
below.

Figure (16): Solution curves for wind turbine torque

Figure (17): Solution curves for momentum and mass

Hamburg University of Applied Sciences (HAW) |CWi Project

14

6.

Post-processing

Upon successful completion of the main-processing, the results are generated for the
post-processing. At this stage, the output power from the rotor needs to be calculated
by using the following formula.
.. (3)
Where is the angular velocity, n rotational speed and M, the torque. The rotational
speed can be calculated using the following formula.

.. (4)

= 121.63 rpm

The angular velocity can be calculated by using below formula:

. (5)

= 12.73 rad/s
The total output power of the wind can be calculated through formulla (3), where M is torque which
was determined from the post-prossesing phase ( M = 750 N*m). For an angular veclocity of 12.73
rad/s and a torque of this much, the power generated by the wind turbine would be:

= 9547.5 W
Figures 18 and 19 shows velocity and pressure contours respectively and figure 20 shows streamlines
(100) coming from the blades.

Hamburg University of Applied Sciences (HAW) |CWi Project

15

Figure (18): Velocity contour

Figure (19): Pressure contour

Hamburg University of Applied Sciences (HAW) |CWi Project

16

Figure (20): Streamlines from the turbine blades indicating the nature of air flow lines leaving the blades

Hamburg University of Applied Sciences (HAW) |CWi Project

17

7. Theorethical Power Calculation

The theoretical maximum power efficiency of any design of wind turbine is 0.59 or in other words, no
more than 59% of the energy carried by the wind can be extracted by a wind turbine. This is called the
power coefficient and is defined as: Also, wind turbines cannot operate at this maximum limit. The
Cp value is unique to each turbine type and is a function of wind speed that the turbine is operating in.
Once we incorporate various engineering requirements of a wind turbine - strength and durability in
particular the real world limit is well below the Betz Limit with values of 0.35-0.45 common even in
the best designed wind turbines [3].
Wind turbines work by converting the kinetic energy in the wind first into rotational kinetic energy in
the turbine and then electrical energy. The energy available for conversion mainly depends on the
wind speed and the sweap area of the turbine.
. (6)

Where is air density, v wind speed, A sweap area of the rotor and Cp the power coefficient. With a
blade radius of R= 3.8m, the sweap area of the rotor is:

) (

Here for our wind tubine, the efficieny of the turbine is assumed as 30%. With the density of air as
1.225 kg/
and rated wind speed of 11m/s, we can then use equation (6) to calculate the power
converted from the wind into rotational energy in the turbine as follow:

) (

) (

= 11089,252 W

8. Results and Discussions

Power calculation from both the numerical and theoretical method seems valid. Comparing the above
theoretical output power with the we obtained from the numerical simulation, the differece is around
(11089,252 W - 9547.5 W = 1542W) which is acceptable. Certainly, throughout the this simulation
and the numerical solution ,there might have been some errors. For instance in the meshing, low
number of meshes could have impact on the final result.
Hamburg University of Applied Sciences (HAW) |CWi Project

18

9. References
[1]. Computational Simulation Techniques. Wind turbine design with CFD lecture notes, Chapter 1

Introduction Pages 2-6, Prof. Dr.-Ing. Rainer Stank, Hamburg University of Applied Sciences
HAW Hamburg.
[2] Wind Turbine Lecture notes , Chapter 4 Aerodynamics of Wind Turbines Page 6, by Prof.
Dr. Timon Kampschulte,Hamburg University of Applied Sciences HAW Hamburg, Faculty of Life
Sciences.
[3] Wind Turbine Power Calculations, RWE npower renewables Mechanical and Electrical
Engineering Power Industry, The Royal Academy of Engineering.
[4] Aerodynamics of Wind Turbines 2nd Edition, Martin O. L. Hansen, 2008.
[5] - AERODYNAMIC STUDY OF A SMALL-SCALE WIND TURBINE. E. Barbiera, P. Bucelloa, S.
Dhersa, P. Mosquera Michaelsenb, B. Pritzb, N. Bottinic and M. Micheloudc. aCentro de Mecnica
Computacional, Instituto Tecnolgico de Buenos Aires, Av. E. Madero 399, Buenos Aires, Argentina,
pbucello@itba.edu.

Hamburg University of Applied Sciences (HAW) |CWi Project

19