18th International Conference on Renewable Energies and Power Quality (ICREPQ’20)

Granada (Spain), 2nd to 4th September 2020

Renewable Energy and Power Quality Journal (RE&PQJ)
ISSN 2172-038 X, Volume No.18, June 2020

Bio-inspired aerofoils for small wind turbines

R. Mulligan
Institute of Mechanical, Process and Energy Engineering, School of Engineering and Physical Sciences,
Heriot-Watt University
Riccarton, Edinburgh EH14 4AS (Scotland, UK)
Phone/Fax number:+0049 131 451 4374 / 3129, e-mail: rm96@hw.ac.uk

Abstract. This paper investigates the effectiveness of A. Biomimicry

modifying the design of wind turbine blades for small wind
turbines. Taking inspiration from nature, the two modifications Biomimicry is finding solutions to human design problems
explored were spanwise corrugations of the blades inspired by a by taking inspiration from nature’s strategies, developed
dragonfly’s wing structure, and flexible blades inspired by the from millions of years of evolution. Nature-inspired ideas
shape-morphing of birds and insect wings that adapt to the air already have prominent applications in the wind industry
conditions. where a company called Wind Power uses leading edge
Two types of corrugations were tested in a wind tunnel with the
modifications inspired by whale tubercles to increase
results that the corrugated nature appeared to have desired effects energy output of wind turbines blades by delaying stall [1].
on delaying stall by 4° and reduced the detriment of post-stall
behaviour. In addition, the corrugated skin displayed similar Fluid structure interaction models of plants with light
characteristics to the control aerofoil, with a small reduction in the weight and flexible characteristics allow them to tolerate
peak lift to drag ratio of 7.3%. In the second section, a rigid turbine high wind forces through energy dissipation to prevent
blade was compared a semi-flexible and fully flexible blade by structural failure [2]. Insects take advantage of wing
altering the Young’s modulus using the QBlade simulation tool. flexing during flight, with a numerical study on beetle
Although, spanwise deflection was high in the flexible blade, a wings [3] highlighting the significance of a flexible wing
reduction in the peak stresses was shown.
in changing the twist angle and camber so that less thrust
is required for the same amout on lift.
Key words. Aerodynamics, aeroelasticity, biomimetics,
boundary layer control, renewable energy B. Small wind turbine challenges

1. Introduction While birds have evolved to morph their wing shapes to

the flight and manoeuvre conditions, manufactured wings
While most research is focused on increasing the capacity require control to achieve wing morphing. The use of
of commercial wind farms, 1 in 10 people still lack access morphing wings for wind turbines has been explored in
to electricity. They live predominantly in rural areas, where principle [4], for horizontal-axis wind turbines [5] and
small off-grid wind turbine systems could be the best vertical-axis wind turbines [6].
solution in tackling energy poverty.
However, for small wind turbines, the added complexity
Utility-scaled turbines have a highly optimised shape and and costs of adding active control to wind turbines is
operational control, matching rotor rotation rate and blade excessive both, compared to their current manufacturing
pitch to the current wind conditions. In contrast, small wind costs and to their potential electricity yield. Viable options
turbines usually have a much simpler turbine blade shape to improve the performance small turbine rotors will have
and little control, especially in cases where the turbines are to rely on turbine blade design and passive control.
built in regions where local construction facilities might be Turbine blade design can consider the spanwise shape or
limited. In these conditions, many opportunities have the aerofoil shape, while passive control might be
arisen to optimise the energy production with new turbine achievable through passively morphing wings.
blade designs to improve aerodynamic performance under a
wide range of operating conditions with little active control. C. Flexible blades
One option is to find a suitable blade surface shape for a
rigid blade design, as explored in this paper in using a Wind turbines tend to be subject to deflections due to
corrugated blade surface. Another option is to explore aerodynamic loading during operations, and the use of
passive control with a flexible blade that has the potential of flexible blades as a form of passive pitch control has been
reducing the loading on the blade and/or the ability to previously explored. Wind tunnel studies have shown that
change pitch to work at optimal conditions for each wind an elasto-flexible lifting surface increases the chamber
speed. Both options can be found in nature. thus increases lifting capabilities and postpones stall onset

https://doi.org/10.24084/repqj18.488 753 RE&PQJ, Volume No.18, June 2020

[7], preventing the sudden drop in lift, and has been shown
to increase the energy production of the wind turbine by up
to 35% [8].

D. Aerofoil profile alterations

Research has looked at altering the entire shape of the

aerofoil surface to alter the aerodynamic properties. A study
by Manshadi [9] looked at the effect of a top jagged wall on
the top surface of an aerofoil to reduce the shock wave
interactions on the boundary layer to reduce overall wave
drag and weaken shockwaves during transonic flight. The
shock disintegration was most pronounced with triangular
jagged shapes and enables an opening to explore dragonfly-
inspired structures at high Reynolds numbers.

In studying the effects of skins which are corrugated on both Fig.1. Aerofoil designs
sides, Xia [10] concluded that a corrugated skin can delay
stall, where eddies form in the troughs which effectively The third aerofoil, Figure 1 (c), is based on research
smooth over the airflow like that of streamlined aerofoils. around corrugated skin on aerofoils to improve
Increasing the size of corrugations caused a reduction in lift performance. The design of these grooves is shallower and
and an increase in drag so a corrugated skin must be contains the groove structure all over the aerofoil surface.
optimally tuned. The shallower grooves should create smaller vortices and
cause less drag than the aggressive profile of the top
E. Aims and objectives corrugated aerofoil, while hoping for an improvement on
the lift/drag ratio. The leading edge of all three designs
This paper considers two bio-inspired methods for was identical to keep the design variables limited.
improving the performance of small wind turbines. First, the
effect of surface corrugation on aerofoil performance was The dragonfly inspired aerofoil in Fig. 1(b) will be here
tested in a wind tunnel and secondly, passive control referred to as the ‘top corrugated’ while the aerofoil in
through a flexible blade structure was explored using the Fig. 1(c) will be referred to as the ‘corrugated’ aerofoil.
QBlade [11] open source platform.
B. Wind Tunnel tests
2. Methodology
The three aerofoils described in §2.A were tested in a
A. Design of aerofoils recirculating wind tunnel with a square test section of
dimension 450 mm x 450 mm, with a design wind
The control aerofoil was a NACA 4416 aerofoil as shown velocity of up to 12 m/s, which resulted in tests covering a
in Figure 1(a) with its design details listed in table I. The range of Reynolds numbers from around 40 000 to 52 000.
span was the full length of the test chamber, effectively The wind speed was measured with an inclined manometer
giving an infinite aspect ratio. tube giving velocity readings with an accuracy of ±2.5%.

The second aerofoil, Figure 1(b), is inspired by the The aerofoil in the test chamber was held in place by a
corrugated dragonfly wing vein structure (Aeshna cyanea) force gauge rig where the front and rear lift forces, FL,f and
which has a deep groove corrugated structure, with the FL,r, and the drag force, FD, were measured. The reading
dimensions based on micro-CT images of a dragonfly wing accuracy of the load cells was better than ±0.1% but
cross section. Since the focus was on the top surface fluctuations in the readings resulted in a final uncertainty
boundary layer, the corrugations were placed on the top in the lift and drag forces comparable to the uncertainty in
only, and were slightly asymmetrical to mimic the natural the wind speed. From the load cells, the drag, lift, and
shape of a dragonfly’s wing. Based on previous studies, a moment coefficients, CD, CL and Cm respectively, were
change in shape of the top surface has demonstrated calculated as
improved lift and drag characteristics.
𝑭𝑫
𝑪𝑫 = (𝟏)
Table I. – Details of the Control Aerofoil: NACA 4416 𝟏
𝝆𝒃 𝒄 𝑼𝟐
𝟐
VARIABLE VALUE UNITS
Chord length, c 0.12 m
Span, b 0.45 m 𝑭𝑳,𝒇 + 𝑭𝑳,𝒓
𝑪𝑳 = (𝟐)
Maximum thickness/chord ratio, t/c 16 % 𝟏
𝝆𝒃 𝒄 𝑼𝟐
Position of maximum thickness 40 % 𝟐
Chamber 4 %
Zero lift angle -4 °

https://doi.org/10.24084/repqj18.488 754 RE&PQJ, Volume No.18, June 2020

 𝒅𝒇 𝑭𝑳,𝒇 − 𝒅𝒓 𝑭𝑳,𝒓 The wind turbine was modelled based on the Aerogenesis
𝑪𝒎 = (𝟑) 5 kW double-blade horizontal-axis turbine situated at the
𝟏
𝝆𝒃 𝒄 𝑼𝟐
𝟐 University of Newcastle, Australia with a rotor diameter
of 5 m.

where 𝝆 = 𝟏. 𝟐𝟐𝟓 𝐤𝐠 𝐦−𝟏 is the density of air, and b the The aeroelastic properties of this turbine have been studied
span and c the chord as given in Table I. using the NREL FAST simulation tool and validated with
experimental recordings. The turbine has yaw control
The lift and drag forces were measured for each aerofoil using a passive tail fin and has a cut-in wind speed of
over a range of angles of attack from –4° to 40°, set 3.5 m/s, a rated wind speed of 10.5 m/s which equates to a
manually with an uncertainty of –0.5° to + 1.5° rotor speed of 200 rpm at a tip speed ratio of 7. At the hub
height of 18 m, the measured average wind speed was
C. Blade Element and Structural modelling 7.5 m/s.

To explore the effectiveness of flexible blades, a numerical

approach had to be taken due to the national closure of
laboratory facilities during the COVID-19 pandemic. For
this, the QBlade simulation tool was used to modify an
existing reference rotor by changing the material properties
of the blades.

QBlade is a tool that implements Blade Element Momentum

(BEM) theory [12] to obtain the performance parameters
through solving the momentum balance of the wind turbine.
The blade was divided into 20 sections as shown in Fig. 3. Blade Design
Figure 3, and the lift and drag force are calculated locally at
each section. BEM is based on the assumptions of two- The blade in Figure 3 of radius 2.5m was modelled on the
dimensional effects only. Three-dimensional correction was NACA 4412 foil, using the in-built function to optimize
applied with the Prandtl tip loss factor, root deflection the chord length and pitch angle of each segment. In order
correction, and a Reynolds number drag correction. to study the effects of blade flexibility on the turbine
performance, a coupled aeroelastic simulation was
A turbulent wind field and wind shear is important to factor performed. It combined the aerofoil polar, structural
into simulations as small wind turbines are normally properties and a generated wind field as shown in Figure 2
situated near buildings and villages. The integration of the with a turbulence intensity of 18% and worked by
FAST module created by the National Renewable Energy transferring data of the pressure forces and node
Laboratory (NREL) [13] performs aeroelastic simulations. displacement between iterations in the FAST module. To
It can solve for unsteady flow and structural dynamics to obtain the results, the structural loads and turbine
consider the effect of wind fluctuations. performance were then post-processed using MATLAB.

It is used in the design and optimisation stage development To compare the results of blade flexibility three blades of
of horizontal and vertical axis turbines. The advantage is the varying Young’s moduli were modelled as represented in
low computing power required vastly speeds up research Table II.
and development of new concepts compared to the Table II. – Blade Material Properties
computer intensive CFD programmes. QBlade has been
validated with other solvers and experiments, proving a MATERIAL FLEXIBILTIY YOUNG’S
MODULUS (GPa)
high degree of accuracy and reliability.
Carbon Fibre Rigid 73

E-glass/Epoxy Semi-Flexible 40

Bamboo/Poplar Fully Flexible 10

E. Fatigue Damage Using Rainflow-Counting

The lifespan of the turbine is determined by the damage

caused by cyclic stress loading. The greatest stress occurs
at the blade root, and the equivalent stress [14] on this area
is calculating by combining the force and bending
moments imported from the FAST simulation:

𝑭𝒛 √𝑴𝒙 + 𝑴𝒚
Fig. 2. Flow diagram of the QBlade features and methods 𝝈𝒆𝒒 = (𝟒)
𝑨 𝑾𝒃
with screenshots from the application [11]

https://doi.org/10.24084/repqj18.488 755 RE&PQJ, Volume No.18, June 2020

Where Fz is the Axial force (N), Mx and My are the edgewise A. Effects of Corrugation on Performance
and flapwise moments (N.m) respectively, A is the circular
area of the blade root (m2) and Wb is the modulus of the root Figure 4 shows the lift and drag coefficients for the three
section of the blade (m3). test aerofoils against angle of attack. Comparing the lift,
the smooth aerofoil gives the highest lift reaching its
The continuous stress loading and unloading cycles can be maximum around 12°, however, suffers from rapid loss of
counted and measured and then sorted depending on its lift straight after separation due its smooth profile.
amplitude and mean into a 3D histogram using the rainflow
counting algorithm to determine the number of fatigue Both the corrugated and top corrugated shapes in Figure 4
cycles. With enough successive damage cycles, the blade show a significant improvement in delaying the stall by 4°
will break. According to Miners Rule [14], this occurs when and have a longer transition before complete boundary
the cumulative damage, D, reaches 1: layer separation. Both managed to avoid the rapid loss of
lift observed in the smooth profile. Dynamic stall can be
𝒏𝒊 problematic in wind turbine operation, where dynamic
𝑫=∑ (𝟓)
𝑵𝒊 instability reduces the fatigue life of the turbine. Previous
𝒊
studies have discussed how the vortices provide a suction
Where ni is the number of load cycles for a given stress force on the boundary layer to keep it attached to the top
amplitude/mean combination obtained from the rainflow surface of the aerofoil.
and Ni is the number of permissible load cycles to failure
for a given stress amplitude/mean combination [14] The corrugated aerofoil has a small reduction of 7.3% in
calculated as follows: the peak lift to drag ratio compared to the smooth aerofoil.
Aerodynamic performance greatly reduces for the top
𝒎 corrugated aerofoil, with a peak lift to drag ratio reduction
𝑿𝒕 + |𝑿𝒄 | − |𝟐𝜸𝑴𝒂 𝝈𝒎 − 𝑿𝒕 + |𝑿𝒄 || of 52.6% compared to the smooth aerofoil due to high drag
𝑵𝒊 = [ 𝜸 ] (𝟔) caused by an early separation of the boundary layer.
𝟐( 𝑴𝒃⁄𝑪 )𝝈𝒂
𝟏𝒃
The lift and drag results oppose the lift benefits shown by
Where several studies, with both corrugated aerofoils displaying
· Xt: Tensile Strength (MPa) a reduction in lift. In previous studies where corrugation
· Xc: Compressive Strength (MPa) increased lift, the whole aerofoil displayed a corrugated
· σm: Mean Stress (MPa) shape and it is possible that a sharp leading edge provided
· σa: Stress Amplitude (MPa) vortex lift to give the aerofoils an increased lift advantage.
Since the nose was kept rounded as a control for this study,
· 𝛄𝐌𝐚 , 𝛄𝐌𝐛 , 𝐂𝟏𝐛 : Safety Factors these vortex lift benefits were not displayed. The lift and
· Xc: Compressive Strength (MPa) drag results varied with the size of corrugation suggesting
· m: Slope of S-N curve that in future studies the corrugation shape can be
optimally tuned.
3. Results
This led to the suggestion that a morphing or flexible wing
surface could be used to reach more desirable lift and drag
The results are split into the two main work packages, first forces than a rigid aerofoil. The first stage of this research
covering the wind tunnel results comparing rigid aerofoil is outlined in the following section.
shapes, while the second part presents results on the
response of a flexible aerofoil and turbine rotor. B. Effects of Flexibility on Performance

The second set of work explored the response of a flexible

turbine blade on rotor performance in QBlade. The normal
and tangential forces recorded from the FAST simulation
were imported into the FEM module where the
displacement of the blade along the entire blade span in
the flapwise direction (figure 5a) and edgewise direction
(figure 5b) was plotted.

The flexible blade showed a significant increase in both

the spanwise and edgewise deflections. The flapwise
deflection should not be a problem unless the deformations
are large enough to hit the tower in which failure would
occur. The edgewise deflections indicate that the chamber
of the blade has changed. This could lead to slight
Fig. 4. Lift and drag coefficients versus angle of attack for improvements in blade performance.
corrugated, top corrugated and smooth aerofoils.

https://doi.org/10.24084/repqj18.488 756 RE&PQJ, Volume No.18, June 2020

 a)
E=Flexible
E=Semi-Flex
E=Rigid

Fig. 5. (a=left) Flapwise & (b=right): Edgewise

displacement along the blade radius under wind loading
conditions b)

The twist angle or torsion of the blade did not seem to differ
between the three blades, with a maximum percentage
difference of 0.3% between the rigid and fully flexible blade
occurring at the blade tip. This suggests that flexibility had
no effect on altering the angle of attack.

Next, in order to check whether an increased chamber

caused by the edgewise deflections improves the
performance, the power coefficient, Cp, was plotted against
the tips speed ratio, TSR, for an average turbulent wind c)
speed of 7.5 m/s in Figure 6. The performance curve for
each of the blades is almost identical despite the large
difference in deflections. The difference was around one
tenth of a percent. This result from the QBlade simulations
indicates that a flexible blade made from more sustainable
materials could produce comparable power to its fibreglass
counterparts.

Fig. 7 gives the rainflow for the (a) rigid, (b) semi-
flexible and (c) fully flexible blades, respectively.

The mean stress values in isolation cannot determine the

fatigue damage and the frequency and amplitude of the
stress waves must be accounted for.

The rigid blade in Figure 7a shows a high frequency

(signified by the high spike of 300+ cycles) of very small
amplitude waves. The small amplitude is likely not to
cause any significant damage but might be an issue if
occurs frequently. Figures 7b and 7c both show that
Fig. 6. Performance curve comparing the flexible, semi- increasing the flexibility of the blade causes a more
flexible and rigid blades. distributed rainflow where the stress waves varies quite
significantly in size over the simulated time period.
However, there needs to be an element of caution with these Although the magnitude of the stress waves may be larger
results as it could point to potential limitations with the than the rigid blade they occur much less frequently.
BEM theory. It was predicted that the performance of the
highly flexible blade would sway significantly to either a The damage calculated for each individual cycle was
high or low Cp. summed to produce the cumulative damage, D,
(Equation 5) to calculate the lifespan, shown in Figure 8.
C. Effects of Flexibility on Fatigue Life For the wind conditions of 7.5 m/s modelled in the initial
simulation, the rigid blade gives a far superior lifespan.
The rainflow counting algorithm in MATLAB analysed the However, the lines in Figure 8 intersect at around 10 m/s.
cyclic stress profile (computed using Equation 4) and was Due to the large gradient in the red line for the rigid blade,
used to assess the damage cycles of each stress the fatigue life quickly drops by three orders of magnitude
amplitude/mean stress combination. Comparing the mean and the blade would fail at wind speeds above 11 m/s.
stress values of each blade, the rigid blade gave the highest
mean stress value at 8 MPa, the semi-flexible blade showed
a slight reduction at 7.8 MPa, and the flexible blade gave a
significant reduction in the mean stress at 6.5 MPa.

https://doi.org/10.24084/repqj18.488 757 RE&PQJ, Volume No.18, June 2020

 This research has identified some promising underlying
aerodynamic processes which hint at how a turbine rotor
would passively adapt to the conditions. These two new
concepts were tested individually, and once proof of
concept is reviewed, higher computational effort is
required to introduce the two ideas into one system and
analyse using coupled fluid-structure interactions (FSI).
Then the flexibility and corrugations can be combined and
fully optimised for a small wind turbine system and
validated with physical test turbine.

The challenge remains to identify surface features,

material properties, and blade designs which respond
favourable to the challenging aerodynamic conditions for
Fig. 8. Life span of the flexible, semi-flexible and rigid small wind turbines.
blades for a wind speeds ranging from 7.5 m/s to 21 m/s.
Acknowledgements
The green squares and line represent the semi-flexible blade
which gives a slight reduction in the gradient, but it is the This work was supported by the Institute of Mechanical,
flexible blade, shown by the blue triangles and line, that Process and Energy Engineering at Heriot-Watt
strongly outperforms the other blades at high wind speeds. University in Edinburgh. Additionally, the author would
The results are clear that in areas which experience high like to thank Dr Wolf Früh for his supervision and support.
winds, a flexible blade could be beneficial in reducing the
blade loading and therefore increase the lifespan of the References
blade. The flexible blade managed to decrease the bending
moment and stresses on the blade root, just like the plants [1] F. Fish, P. Weber, M. Murray, L. Howle, “The tubercles on
were shown to bend to resist high wind forces. humpback whales' flippers: application of bio-inspired
Alternatively, in regions that experience low winds on a technology”, Integr Comp Biol (2011), Vol. 51, pp. 203-13.
regular basis, a more rigid blade would be a better choice. [2] E. d. Langre, “Effects of Wind on Plants”, Annual Review of
Fluid Mechanics (2008). Vol. 40, pp. 141-168.
This study correlated the life span of the blade directly from [3] T. Q. Le et al, “Improvement of the aerodynamic performance
by wing flexibility and elytra–hind wing interaction of a
the moments and forces on the blade root only, since the
beetle during forward flight”, Journal of The Royal Society
blade design was the focus point. Maintenance issues and Interface (2013). Vol. 10,
failures can come from a range of the mechanical and [4] X. Lachenal, S. Daynes and P.M. Weaver, “Review of
electrical components and a lifecycle analysis such as an morphing concepts and materials for wind turbine blade
FMEA requires an in-depth look at all components. applications”, Wind Energy (2013). Vol. 16, pp. 283 – 307.
[5] S.P. Evans, D.R Bradney and P.D. Clausen, “Aeroelastic
The fatigue life results proved to be highly sensitive to both measurements and simulations of a small wind turbine
the m parameter (S-N slope) and the compressive and tensile operating in the built environment”, Journal of Physics:
strength of the material in Equation 6. Material properties Conf. Ser. (2016), Vol. 753.
[6] D.W. McPhee and A. Bayene, “Fluid-structure interaction
tables vary between sources and whilst care was taken upon
analysis of a morphing vertical axis wind turbine”, Journal of
selection of these properties, it must be noted that the final Fluids and Structures (2016), Vol. 60, pp. 143 – 159.
lifespan result are sensitive to their choice. [7] J. Piquee et al, “Aerodynamic analysis of a generic wing
featuring an elasto-flexible lifting surface”, Adv. Aerodyn
4. Discussion and Conclusion (2019). Vol. 1, p. 20.
[8] V. Cognet, S. Courrech du Pont, I. Dobrev, F. Massouh and
While this research is preliminary, it highlights B. Thiria, “Bioinspired turbine blades offer new perspectives
for wind energy”, Proc. R. Soc. (2017). Vol. 473.
opportunities to improve the performance of turbines
[9] D. Manshadi, “Numerical evaluation of passive control of
through the exploitation of design options which are easy to shock wave/boundary layer interaction on NACA0012
manufacture, even in parts of the world where airfoil using jagged wall”, Acta Mechanica Sinica (2016).
manufacturing and maintenance facilities are limited. The Vol. 32, pp. 792-804.
corrugated aerofoils gave the advantage of delaying stall, [10] Y. Xia, O. Bilgen, M. Friswell “The effect of corrugated
with the corrugated skin proving a better option for skins on aerodynamic performance”, Journal of Intelligent
adaptation to wind turbine blades because it had similar lift Material Systems and Structures (2014). Vol. 25, pp. 786-
to drag characteristics to the smooth aerofoil. Flexible 794.
materials such as bamboo laminates are accessible even to [11] D. Marten, “QBlade Guidelines v 0.6”, TU , Berlin, 2012.
[12] M. O. L. Hansen, Aerodynamics of Wind Turbines: second
remote areas where local alternatives can be sourced.
edition, Earthscan, (2008).
Manufacturers face a problem in recycling fibreglass [13] J.M Jonkman, M.L. Buhl, “FAST users guide”, National
material at the end of life so flexible blades offers a Renewable Energy Laboratory, (2005).
sustainable solution that has been shown here to produce [14] Germanischer Lloyd, Guidelines for the certification of
comparable performance and increased design lifetime. wind turbines, Germanischer Lloyd, Hamburg (2010), pp.
197-204.

https://doi.org/10.24084/repqj18.488 758 RE&PQJ, Volume No.18, June 2020