Dust Effect on the Performance of Wind Turbine Airfoils

doi:10.4236/jemaa.2009.12016

Paper Menu >>

Journal Menu >>

J. Electromagnetic Analysis & Applications, 2009, 1: 102-107

doi:10.4236/jemaa.2009.12016 Published Online June 2009 (www.SciRP.org/journal/jemaa)

Dust Effect on the Performance of Wind Turbine

Airfoils

Nianxin Ren1,*, Jinping Ou1,2

1School of Civil Engineering, Harbin Institute of Technology, Harbin, 150090, China; 2School of Civil and Hydraulic Engineering,

Dalian University of Technology, Dalian, 116024, China.

Email: *[email protected]

Received January 9th, 2009; revised February 23rd, 2009; accepted March 3rd, 2009.

ABSTRACT

The full two-dimensional Navier-Stokes algorithm and the SST k-



turbulence model were used to investigate incom-

pressible viscous flow past the wind turbine two-dimensional airfoil under clean and roughness surface conditions. The

NACA 63-430 airfoil is chosen to be the subject, which is widely used in wind turbine airfoil and generally located at

mid-span of the blade with thickness to chord length ratio of about 0.3. The numerical simulation of the airfoil under

clean surface condition has been done. As a result, the numerical results had a good consistency with the experimental

data. The wind turbine blade surface dust accumulation according to the operational periods in natural environment

has been taken into consideration. Then, the lift coefficients and the drag coefficients of NACA 63-430 airfoil have been

computed under different roughness heights, different roughness areas and different roughness locations. The role that

roughness plays in promoting premature transition to turbulence and flow separation has been verified by the numeri-

cal results. The trends of the lift coefficients and the drag coefficients with the roughness height and roughness area

increasing have been obtained. What’s more, the critical values of roughness height, roughness area, and roughness

location have been proposed. Furthermore, the performance of the airfoil under different operational periods has been

simulated, and an advice for the period of cleaning wind turbine blades is proposed. As a result, the numerical simula-

tion method has been verified to be economically available for investigation of the dust effect on wind turbine airfoils.

Keywords: NACA 63-430 Airfois, Lift Coefficient, Drag Coefficient, Roughness Height, Roughness Area, Roughness

Location

1. Introduction

As is known to all, the world energy crisis is more and

more serious in our modern society. Wind energy, which

is the most mature technology among so many kinds of

clean and renewable energy, is developing with an

amazing speed now. The crucial to rotor design is the

subject of airfoils. One of the most critical problem for

wind turbine rotors is degradation of the performance,

and the unpredictability of stall due to dust accumulation

on blade surface area. The objective of this work is to

provide a better understanding for the effect of blade

surface roughness on the performance of the wind turbine

thick airfoil. It is very useful to enable wind turbine

designers to predict loads and energy losses during wind

turbines operating under dust conditions, it is necessary

to qualitatively and quantitatively know the change in the

aerodynamic properties due to the dust accumulation on the

surface of the blade. So, the analysis of airfoil surface

roughness has practical application interest in addition to

academic interest. Now, the importance of the dust effect on

the performance of the wind turbine airfoil is well realized.

Generally, roughness has a large effect on the flow

dynamic processes. Therefore, the stall-regulation phe-

nomena in wind turbines are affected by a high degree

due to the increase of the blade surface roughness. De-

spite some previous experimental and numerical work

where surface roughness is involved, the information that

has been obtained on this subject still remains far from

complete [1-3], and the processes of boundary-layer

separation and stall phenomena, which occur on the wind

turbines blade in the presence of surface roughness, are

not fully understood. To obtain more detailed information

about the effect of the surface roughness on the lift coef-

ficients and the drag coefficients of mid-span thick air-

foils, the full two-dimensional Navier-Stokes algorithm

and the SST k- turbulence model are used to investigate

incompressible viscous flow past the wind turbine two-

dimensional airfoil with surface roughness. The NACA

63-430 airfoil is chosen to be the subject, which is widely

used in wind turbine airfoil and generally located at

mid-span of the blade with thickness to chord length ratio

Dust Effect on the Performance of Wind Turbine Airfoils 103

of about 0.3. The lift coefficients and the drag coeffi-

cients of NACA 63-430 airfoil have been computed un-

der different roughness heights, roughness areas, and

different roughness locations. The role that roughness

plays in promoting premature transition to turbulence and

flow separation has been verified by the numerical results.

Furthermore, the degenerate trend of performance for

NACA 63-430 airfoil under different operation periods

has been simulated.

2. Numerical Modeling

2.1 Governing Equations

The flow past the airfoil was modeled by the full Na-

vier-Stokes equation for two-dimensional, viscous, im-

pressible flow. The continuous equation and Momentum

equation based on Reynolds averaged N-S equations are

as follows:

u

 (1)

ij ''

jijj

(uu)

(u)u

p(u

tx xxx









 

  (2)

where i, j=1, 2; =1.255kg/m3; =1.789410-5 kg/(ms).

2.2 SST k- Turbulence Model

The shear-stress transport (SST) k- model was devel-

oped by Menter [4] to effectively blend the robust and

accurate formulation of the k-



model in the near-wall

region with the free-stream independence of the k-

model in the far field. The SST k- model is described as

follows:

ijj

() k

() ()

txx x

ikkk

kkuG YS



 



 





(3)

ijj

() () ()

txx x

uGYDS



 

 



 



 





(4)

where k represents the turbulence kinetic energy; 

represents the specific dissipation rate; Gk represents the

generation of turbulence kinetic energy due to mean ve-

locity gradients; G represents the generation of ; k

and  represent the effective diffusivity of k and , re-

spectively; Yk and Y represent the dissipation of k and 

due to turbulence; D represents the cross-diffusion term,

calculated as described below; Sk and S are user-defined

source terms.

The SST k- model is similar to the standard k-

model, but includes the following refinements:

·The standard k- model and the transformed k-

model are both multiplied by a blending function and

both models are added together. The blending func-

tion is designed to be one in the near-wall region,

which activates the standard k- model, and zero

away from the surface, which activates the trans-

formed k- model.

·The SST model incorporates a damped cross-diffusion

derivative term in the  equation.

·The definition of the turbulent viscosity is modified to

account for the transport of the turbulent shear stress.

·The modeling constants are different.

These features make the SST k- model more accurate

and reliable for a wider class of flows (e.g., adverse

pressure gradient flows, airfoils, transonic shock waves)

than the standard k- model. Other modifications include

the addition of a cross-diffusion term in the  equation

and a blending function to ensure that the model equa-

tions behave appropriately in both the near-wall and

far-field zones.

2.3 Modeling

The whole computational zone consists of a semicircle

with the radius of 10m and a rectangle with the length of

25m (Figure 1). The length of numerical airfoil which

locates near the center of the semicircle is 1m. The height

of the grid near the airfoil surface is 210-5m (Figure 2).

Reynolds number is 1.6106 [5].

3. Numerical Results

3.1 Numerical Simulation under Clean Surface

Condition

First of all, to ensure that the numerical model is avail-

able for the free-stream flow past the airfoil, the numeri-

cal simulation under clean airfoil surface condition was

made to compare with the wind tunnel experimental data

[5]. The simulative condition was according to the wind

Figure 1. The whole computational zone

104 Dust Effect on the Performance of Wind Turbine Airfoils

Figure 2. Local amplified grids for the airfoil

tunnel set-ups, for example, Reynolds number 1.6106;

free stream turbulence level of 1%. The lift coefficients

and the drag coefficients of the NACA 63-430 airfoil

were computed under the angle of attack between 0 and

25 degree, based on the full two-dimensional Navier-

Stokes algorithm and the SST k- turbulence model. The

comparison between numerical results and experimental

data shows in Figures 3-4.

In the above two figures, it was obvious that the nu-

merical results had a good consistency with the experi-

mental data, despite small discrepancy in comparison of

drag coefficient curves. Therefore, the numerical model

was confirmed to be available for the free-stream flow

past the NACA 63-430 airfoil. Then, in the following of

the paper, the numerical model would be used to investi-

gate the airfoil surface roughness effect.

3.2 Numerical Simulation under Different Sur-

face Roughness Heights

It is supposed that all the surface of the airfoil is rough-

ness area and the angle of attack is 10.6 degree. That’s

because at this angle of attack the numerical results have

a perfect consistency with the experimental data. Then,

numerical simulations of the NACA 63-430 airfoil under

different surface roughness heights has been done and the

results are shown in Figure 5.

Figure 3. Comparison of lift coefficient curves

NACA 63-430 Airfoil

0.00

0 5

Angl e)

Dra

Figure 4. Comparison of drag coefficient curves

Figure 5. Numerical results of Cl and Cd under different

roughness heights

In Figure 5, it is evident that the lift coefficient curve

decreased very rapidly when the roughness height is less

than 0.3mm. However, the lift coefficient curve decreases

very slowly when the roughness height is more than

0.3mm. The same trend of the change also happened in

the drag coefficient curve, which increases very rapidly

when the roughness height is less than 0.3mm, but in-

creases very slowly when the roughness height is more

than 0.3mm. That’s because the surface roughness plays

a role in promoting premature transition to turbulence

and flow separation. When the roughness height is less

than 0.3mm, the effect of roughness is very obvious. But,

when the roughness height is more than 0.3mm, the

whole airfoil has already fully become turbulent bound-

ary and at the same time the flow separation is very serious,

so the effect of roughness on the performance of the airfoil

seems no longer evident. Therefore, the roughness height of

0.3mm could be viewed as a roughness critical height.

Besides lift coefficients information and drag coeffi-

cients information of the airfoil, the numerical results

contain more information of the flow past the airfoil un-

NACA 63-430 Airfoil

C ue

l and Cd val

.00.5 1.0

ughness height(mm)

Lift coefficient

Drag coefficient

NACA 63-430 Airfoil

Lift coefficient

Drag coefficient

Roughness height (mm)

0.0 0.5 1.0

C1 and Cd value

1.2

1.0

0.8

0.6

0.4

0.2

0.0

NACA 63-430 Airfoil

0.0

0.2

0.4

0.6

0.8

1.0

Angl )

L t

1.2

1.4

510 1520 25

e of attack(degree

ift coefficien

Numerical results

Experimental data

0 5 10 15 20 25

Angle of attack(degree)

umerical results

Experimental data

NACA 63-430 Airfoil

Lift coefficient

1.4

1.2

0.8

0.6

0.4

0.2

0.0

0.05

0.10

0.15

0.20

0.25

51015 202

e of attack(degre

g coefficient

Numerical results

Experimental data

NACA 63-430 Airfoil

umerical results

Experimental data

0 5 10 15 20 25

Angle of attack (degree)

Drag coefficient

0.25

0.20

0.15

0.10

0.05

0.00

Dust Effect on the Performance of Wind Turbine Airfoils 105

der different roughness conditions, which are also very

useful to clarify the physical mechanism of the surface

roughness effect. For example, the turbulent intensity

distribution and pressure coefficient distribution of the

airfoil surface are shown in Figure 6 and Figure 7, re-

spectively.

In Figure 6, it is clear that the turbulent intensity of the

airfoil under clean condition is very small, however, with

the roughness height increasing from 0 to 0.3mm, turbu-

lent intensity increases very obviously, especially for the

leading edge of the airfoil. In Figure 7, the area enclosed

by the pressure coefficient curve of the airfoil obviously

decreases with the roughness height increasing from 0 to

0.3mm, especially for the location at the front 50% of the

chord length.

Furthermore, the two pictures could be better to dem-

onstrate and clarify that surface roughness plays a role in

promoting premature transition to turbulence and flow

separation. By the way, it also could be found that the

increase of turbulent intensity and the change of the

pressure coefficient are no longer evident when the

roughness height is more than 0.3mm.

Figure 6. Turbulent intensity curves under different rough-

ness heights

Figure 7. Pressure coefficient curves under different rough-

ness heights

3.3 Numerical Simulation under Different

Roughness Areas

To clarify the effect of different roughness areas on per-

formance of the airfoil, simulations under the angle of

attack 10.6 degree and roughness height 0.3mm were

done. The roughness area is denoted by the ratio of the

chord length covered with roughness, which is calculated

from the leading edge. The simulation results are shown

in Figure 8. It is obvious that the lift coefficient curve

decreases rapidly when the roughness area is less than 0.5.

However, the lift coefficient curve decreases very slowly

when the roughness height is more than 0.5. It can be

concluded that the front 50% of the chord length more

easily promotes premature transition to turbulence and

flow separation. In the flowing section, further study and

explanation will be given.

3.4 Numerical Simulation under Different

Roughness Locations

Furthermore, to see the effect of different roughness lo-

cations on performance of the airfoil more clearly, simu-

lations under the angle of attack 10.6 degree and rough-

ness height 0.3mm have been done. The whole NACA

63-430 airfoil is averagely divided into 10 parts, and each

part is with a length of 0.1m. The simulation results are

shown in Figure 9. The decrease percentage of the lift

coefficient is defined by:

NACA 6

% 100%

clean roughness

decrease

clean

Cl Cl





(5)

From the Figure 9, it could be seen that the influence

of surface roughness located at the front 50% of the

chord length is more obvious than that located at the back

50% of the chord length. On one hand, it is because the

roughness located at the front 50% of the chord length

more easily promotes premature transition to turbulence

and flow separation. On the other hand, it is also because

the airfoil at font 50% of the chord length is generally

thicker than that at the back 50% of the chord length.

Therefore, the roughness location at 25% of the chord

length had the most obvious effect, where is the location

of the largest thickness.

It is worth to be noticed that, when the roughness loca-

tion is at the trailing edge 10% of the chord length, the

lift coefficient under surface roughness condition is about

3% larger than that under clean surface condition. So, it

could be concluded that properly increasing the surface

roughness of trailing edge could be benefit to promote the

airfoil’s lift coefficient to some degree, which has a good

agreement with experimental data [6]. The conclusion

would be useful for practical projects.

3-430 Airfoilclean

0.0

0.1

Turbulent intensity

0.1

0.2

00.2 0.40.6 0.81

x/c

0.1mm roughness

0.3mm roughness

1mm rou

hness

0 0.2 0.4 0.6 0.8 1

x/c

NACA 63-430 Airfoil

Turbulent intensity

0.2

0.15

0.1

0.05

clean

0.1mm roughness

0.3mm roughness

1mm roughness

NACA 63-430 Airfoil

-2.5

0 0.20.40.60.8 1

-2

Pressure coefficient

1.5

-1

0.5

1.5

x/c

clean

0.1mm roughness

0.3mm roughness

1mm rou

hness

NACA 63-430 Airfoil

1.5

Pressure coefficient

0.5

-0.5

-1

-1.5 clean

0.1mm roughness

-2 0.3mm roughness

1mm roughness

-2.5

x/c

106 Dust Effect on the Performance of Wind Turbine Airfoils

Figure 8. Lift coefficient under different roughness areas

Figure 9. Lift coefficient under different roughness locations

3.5 Numerical Simulation Under Different

Operational Periods

In this section, considering the effect of the blade surface

dust accumulation, the performance of the wind turbine

NACA 63-430 airfoil under different operational periods

has been studied. The relationship between surface

roughness and operational period is determined by field

investigation, which was done by Khalfallah [2]. Then,

the roughness heights and the roughness areas under

different operational periods can be calculated by the

following linearization function proposed by Khalfallah,

which is not taken the rain washing effect into consid-

eration. As a result, the specific four operation periods

are list in Table 1.

0.08 0.02

DT (6)

/25

DT (7)

where, Dd is the dust size (diameter in mm); T is length of

operational period in months; Da is the dust area in per-

centage of the chord length.

Subsequently, the former available numerical model

has been used again to investigate the performance of the

NACA 63-430 airfoil under the four operation periods

and the numerical results are shown in Figure 10 and

Figure 11.

It can be seen that, the dust effect on the performance

of wind turbine airfoil is obvious when the operation pe-

riod is 5months. The decrease of the lift coefficient and

the increase of the drag coefficient are remarkable to a

large degree, especially for the angle of attack around 13

degrees. That’s because the roughness height of the

5months operational period is 042mm, which is larger

than the critical value 0.3mm. As the operation period

increasing from 5months to 12.5months, the roughness

area is increasing from 0.2 to 0.5, which just reaches the

critical roughness area value (0.5). Therefore, the lift co-

efficient further decreases to a large degree at the angle

of attack less than 13 degrees. But, when the operational

period is longer than 12.5 months, the change of the lift

coefficient is rather small. That’s because that, when the

operational period is longer than 12.5 months, the rough-

ness height and roughness area both larger than the criti-

cal values proposed in the previous section. As a result,

the effect of the surface roughness on promoting prema-

ture transition to turbulence and flow separation tends to

be neglectable.

NACA 63-430

Table 1. Roughness heights and roughness areas according

to different wind turbine operational periods

T(months) Dd (mm) Da

2.5 0.22 0.1

5 0.42 0.2

12.5 1.02 0.5

20 1.62 0.8

Figure 10. The comparison of lift coefficient curves during

different operation periods

-1

x/c

10%

20%

30%

5%15%25% 35% 45%55% 65% 75%85% 95%

Cl decrease

percentage

NACA 3-430 6

5% 15% 25% 35% 45% 55% 65% 75% 85% 95%

x/c

30%

20%

10%

-10%

C1 decrease

percentage

Dust Effect on the Performance of Wind Turbine Airfoils 107

·The performance of NACA 63-430 airfoil is more sen-

sitive to roughness location at the front 50% of the

chord length than roughness location at the back 50%

of the chord length. In particular, proper roughness

height at the trailing edge can be benefit to promote the

lift coefficient to some degree. This conclusion is very

useful for the wind turbine blade designer to further

improve the performance of the blade.

·Considering the dust accumulation effect, the per-

formance degenerate trend of NACA 63-430 airfoil

under different operation periods has been studied. As

a result, the period of 3 months without any rain is

proposed as the proper period of cleaning the blade

surface.

As a result, the numerical simulation method has been

verified to be economically available for the investigation

of the dust effect on the performance of the wind turbine

airfoil.

Figure 11. The comparison of drag coefficient curves during

different operation periods

In practical project, it is of great economic interest to

keep the good performance of the wind turbine blade, so

it is very meaningful to periodically clean the dust of the

blade surface when there is no rain washing for a long

time. Considering the performance degenerate trend of

the wind turbine airfoil according to the operational pe-

riod, the period of 3 months is proposed for the proper

period of cleaning the blade surface.

5. Acknowledgments

The support of the National Science Foundation of China

under project No. 50538020 and the National Science and

Technology Planning under project No. 2006BAJ03B00

is gratefully acknowledged.

4. Conclusions REFERENCES

In this paper, the full two-dimensional Navier–Stokes

algorithm and the SST k- turbulence model were used

to investigate incompressible viscous flow past the wind

turbine NACA 63-430 airfoil under clean and roughness

surface conditions. The key findings can be summarized

as follows:

[1] R. Van and W. A. Timmer, “Roughness sensitivity

considerations for thick rotor blade airfoils,” in 41st

Aerospace Sciences Meeting, Reno, USA, pp. 472-480,

2003.

[2] G. K. Mohammed and M. K. Aboelyazied, “Effect of dust

on the performance of wind turbines,” Desalination, Vol.

209, No. 1-3, pp. 209-220, April 30, 2007.

·The full two-dimensional Navier-Stokes algorithm and

the SST k- turbulence model have been verified to be

available for predicting the performance of the wind

turbine airfoil under clean and roughness surface condi-

tions. The numerical results under clean surface condi-

tion have a good consistency with the experimental data,

despite small discrepancy in comparison of drag coeffi-

cient curves.

[3] W. H. Wade and P. R. Alric, “Numerical prediction of

unsteady vortex shedding for large leading-edge

roughness,” Computers & Fluids, Vol. 33, No. 3, pp. 405-

434, March 2004.

[4] F. R. Menter, “Zonal two-equation model k–w models for

aerodynamic flows,” in 24th Fluid Dynamics Conference,

Orlando, Florida, 1993.

[5] P. Fuglsang, I. Antoniou, and S. D. Kristian, “Wind tunnel

tests of the FFA-W3-241, FFA-W3-301 and NACA

63-430 airfoils,” http://www.risoe.dk/rispubl/VEA/veapdf/

ris-r-1041.pdf.

·The lift coefficient and the drag coefficient of NACA

63-430 airfoil is influenced more obviously by the

roughness height less than 0.3mm than by the

roughness height more than 0.3mm. In other words,

the performance of airfoils is more sensitive to small

roughness height. As a result, the roughness height

of 0.3mm is proposed to be roughness critical

height.

[6] N

. S. Bao, F. E. Huo, and Z. Q. Ye, “Aerodynamic

performance with roughness on wind turbine airfoil

surface (in Chinese),” Acat Energiae Solars Sinca, Vol. 4,

No. 8, pp. 458-462, 2005.