Journal of The Institution of Engineers (India): Series C

Investigation on the Magnus Effect on a symmetric NACA 0021 Airfoil: A Numerical

and Experimental Study
--Manuscript Draft--

Manuscript Number: IEIC-D-20-00500

Full Title: Investigation on the Magnus Effect on a symmetric NACA 0021 Airfoil: A Numerical
and Experimental Study

Article Type: Original Contribution

Section/Category: Aerospace Engineering

Keywords: Airfoil, symmetric, spinning cylinder

Corresponding Author: Md. Abdus Salam, M.Sc

Military Institute of Science and Technology
Dhaka, Dhaka BANGLADESH

Corresponding Author Secondary

Information:

Corresponding Author's Institution: Military Institute of Science and Technology

Corresponding Author's Secondary

Institution:

First Author: Md. Abdus Salam, M.Sc

First Author Secondary Information:

Order of Authors: Md. Abdus Salam, M.Sc

M.A Taher Ali, PhD

Md. Quamrul Islam, PhD

Order of Authors Secondary Information:

Funding Information:

Abstract: This paper addresses the outcome of implementing a spinning cylinder at the leading-
edge of symmetric NACA 0021 airfoil at velocity ratios (Uc/U) inferior to previously
conducted studies with higher velocity ratios (≥ 1.0). In an attempt to study the efficacy
for applying Magnus effect both numerical and experimental approaches are carried
out. The analysis in this study was conducted in different angle of attacks ranging from
0° to 20°, in different spinning velocities of the cylinder (0, 3000, 3500, 4000, 4500 and
5000 rpm) and in 20 m/s free stream velocity (U). There is a significant correlation
between the spinning of the cylinder and the increase in the coefficient of lift (CL) and
escalation in stall angle (αe) when the cylinder is placed at the leading edge and
rotated with a low speed (Uc). This study focused on the effects of low velocity ratios of
the cylinder as there were very less contributions of the previous studies on this
aspect. This concept can be beneficial in short take-off and landing airfoils, wind
turbine, thick airfoil for RC aircrafts etc.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Title Page

Investigation on the Magnus Effect on a symmetric NACA

0021 Airfoil: A Numerical and Experimental Study

Md. Abdus Salam1,2, Dr. M. A. Taher Ali3, Dr. Md. Quamrul Islam4
1,2
PhD Student and Research Scholar, Department of Mechanical Engineering, Head, Department of
Aeronautical Engineering, Military Institute of Science and Technology, Bangladesh, Email:
[email protected]
3
Professor, Department of Aeronautical Engineering, Military Institute of Science and Technology, Bangladesh.
4
Professor, Department of Mechanical Engineering, Military Institute of Science and Technology, Bangladesh.

Abstract:

This paper addresses the outcome of implementing a spinning cylinder at the leading-edge of symmetric
NACA 0021 airfoil at velocity ratios (Uc/U) inferior to previously conducted studies with higher
velocity ratios (≥ 1.0). In an attempt to study the efficacy for applying Magnus effect both numerical
and experimental approaches are carried out. The analysis in this study was conducted in different angle
of attacks ranging from 0° to 20°, in different spinning velocities of the cylinder (0, 3000, 3500, 4000,
4500 and 5000 rpm) and in 20 m/s free stream velocity (U). There is a significant correlation between
the spinning of the cylinder and the increase in the coefficient of lift (CL) and escalation in stall angle
(αe) when the cylinder is placed at the leading edge and rotated with a low speed (Uc). This study
focused on the effects of low velocity ratios of the cylinder as there were very less contributions of the
previous studies on this aspect. This concept can be beneficial in short take-off and landing airfoils,
wind turbine, thick airfoil for RC aircrafts etc.

Keywords: Airfoil, symmetric, spinning cylinder

1
To whom any correspondence should be addressed.
Manuscript (excluding authors' names and affiliations)

Investigation on the Magnus Effect on a symmetric NACA

0021 Airfoil: A Numerical and Experimental Study

Abstract generalizability and practicality of this method. This

study aims to broaden the existing findings by
This paper addresses the outcome of implementing investigating Magnus effect on a thick symmetric
a spinning cylinder at the leading-edge of symmetric NACA 0021 airfoil at Uc/U ≤ 0.2 and Re 2 x 105.
NACA 0021 airfoil at velocity ratios (Uc/U) inferior
to previously conducted studies with higher velocity
ratios (≥ 1.0). In an attempt to study the efficacy for
applying Magnus effect both numerical and There is a huge volume of published studies
experimental approaches are carried out. The comprising both experimental and numerical tactics
analysis in this study was conducted in different which focuses on upgrading the performance of
angle of attacks ranging from 0° to 20°, in different aircraft or airfoils. Blowing and suction in boundary
spinning velocities of the cylinder (0, 3000, 3500, layers, optimization of the jet length and slot
4000, 4500 and 5000 rpm) and in 20 m/s free stream geometry, control of the boundary layer using a
velocity (U). There is a significant correlation moving surface (MSBC) etc. are some of the
between the spinning of the cylinder and the prominent methods that have augmented the airfoil
increase in the coefficient of lift (CL) and escalation performance [1-6]. The elementary perception of
in stall angle (αe) when the cylinder is placed at the MSBC via a spinning cylinder to intensify the
leading edge and rotated with a low speed (Uc). This aerodynamic performance can be described by
study focused on the effects of low velocity ratios of Magnus effect; a concept which is dated back to
the cylinder as there were very less contributions of 1852 pioneered by scientist Gustav Magnus.
the previous studies on this aspect. This concept can Extensive series of studies have been conducted to
be beneficial in short take-off and landing airfoils, incorporate this concept on account of the growth in
wind turbine, thick airfoil for RC aircrafts etc. lift generation on aircraft. Because of the adverse
pressure gradient, the free stream velocity and
velocity gradient reduces. Consequently, the wall
shear stress also drops to zero as adverse pressure
1. Introduction gradient rises to a certain extent. This compels the
free stream air to flow in opposite direction and thus
At present, aircrafts are constantly being used in creates flow separation. Addition of momentum
replacement of other modes of transportation energy to the flow due to the spinning cylinder
primarily because of the speed, efficiency and provides with an increase in velocity gradient which
convenience both for passenger and freighter in turns delays flow separation.
aircraft for military as well as civilian purposes.
Therefore, increasing the efficiency of the aircraft is
becoming progressively challenging. Modifying the
aircraft structure can increase the aerodynamic A number of investigations on this concept was
performance by increasing CL, range, αe and conducted with symmetric airfoils with a leading-
endurance which decreases the operational and edge spinning cylinder. By far the most detailed and
maintenance cost of the aircraft. Integrating a extensively investigated approaches throughout
spinning cylinder at the airfoil leading edge injects numerous numerical and experimental methods by
extra energy to the free stream airflow and thus Modi et al. Nevertheless, their researches have not
limits the possibility of flow separation to a certain been able to account for lower Uc/U and effect of CD
extent. The detected growth in CL is attributed to the [7-10]. A qualitative numerical investigation by
increase in Uc/U but with the downside of Ahmed et al. was carried out on the role of MSBC
increasing coefficient of drag (CD). in 2014 on symmetric NACA 0024 airfoil and it was
found that CL had increased by 36% and αe got
delayed by 122% using modified airfoil [11]. To
better understand this method, Huda et al. carried
Previous studies on this concept were mainly out numerical analysis on NACA 0010 and reported
focused on Uc/U > 1.0, yet the consequences of 145% increase in CL in the modified airfoil although
lower Uc/U were neglected or insufficient. Also, their study was partial to only Uc/U = 0 and 2. [12].
many studies seemed to overlook the effect of lift In a numerical analysis which was set out to estimate
induced drag and drag due to the increasing appropriate gap between the airfoil body and
vibration of the high-speed spinning cylinder. This spinning cylinder, Abdulla et al. concluded that a 3
resulted in an incomplete representation and lack of mm gap would be necessary for best performance
for symmetric NACA airfoil [13]. In a recent The gap between the cylinder and the airfoil body
experimental study, Salam et al documented 24% was 3 mm and the cylinder diameter was chosen to
increase in CL and 20% increase in αe for NACA be 15 mm to provide the 3 mm gap easily. The
23018 airfoil for Uc/U = 0.2 [14] and explored the spinning speed of cylinder was selected as 0, 3000,
effects of vibrational drag of the cylinder. 3500, 4000, 4500, and 5000 rpm which corresponds
to Uc/U = 0, 0.12, 0.14, 0.16, 0.18 and 0.2 to
facilitate the accuracy of the chosen velocity ratios.
At a distance of 37.5 mm from the airfoil leading
The paper explores the possible outcomes of edge, a steel rod was implanted into the airfoil to
applying concept of MSBC for a thick NACA 0021 facilitate the provision of changing angle of attack.
airfoil and highlights the usefulness of applying this A portion of that steel rod (10 mm) was hooked onto
method for various aerodynamic applications. It was the 3-component balance which was attached with
the intention of this study to fill a gap in the ongoing the end support plate. The 3-component balance was
literature by shedding lights on the effects of lower used to determine lift and drag. A coupler was
Uc/U on aerodynamic performance of a thick introduced to hold the steel rod and the cylinder as it
symmetric airfoil. This paper is composed of five was required to change the angle of attack of the
sections. The first section contextualizes the study airfoil as well as change the location of the spinning
by providing a short summary of the literature. The cylinder. Thus, it provided similar orientation for the
next section provides the methodology for airfoil and cylinder at all cases specially when
numerical and experimental analysis. The next changing angle of attack. A motor was required to
section draws key findings of the research. An propel the cylinder and the motor was needed to be
uncertainty analysis for the experiment was carried installed outside of the test section. So, both walls of
out in next section. In the fifth section, an analysis the test sections included a groove. A cap was
of the results is presented. utilized to hold one end of the cylinder connected to
bearing and rear sidewall disk. The rear sidewall
disk guided the cylinder shaft to the 3-component

Figure 1. Experimental and Numerical setup (a) coupler with motor; (b) airfoil; (c) cylinder; (d) sidewall
disk; (e) wires to connect electronic speed controller and battery; (f) inlet of the numerical domain; (g)
sidewall of the numerical domain; (h) outlet of the numerical domain

balance. The coupler was used to attach the other

end of the cylinder with the motor shaft. The
2. Methodology connecting arm was used to hold the motor as well
as to provide similar orientation while changing
2.1 Experimental Methodology angles of attack. The connecting arm was attached
with the airfoil through the front sidewall disk to
Prior to undertaking the experiment, an open loop facilitate a sturdy support.
suction type subsonic wind tunnel of model AF-100
with a 300 mm x 600 mm x 300 mm working section
was selected to conduct the experiment to keep the
Uc/U in a precise range for certain rotational speeds 2.2 Numerical Methodology
of the cylinder. The maximum free stream velocity
offered by the wind tunnel was 35 m/s. The aspect
ratio of the wing was 2 with an area of 0.045 m2,
chord length of 0.15 m and span of 0.3 m. A groove
was formed at the leading edge to provide the space
for spinning cylinder as shown in Figure 1.
Figure 1 represents the numerical setup for the study validation was achieved for nearly 500000 nodes.
where the inlet velocity was provided as 20 m/s, the To avoid the rise in numerical cost with slight
sidewalls are kept as stationary wall and the outlet accuracy in approximation, the number of nodes was
was considered as pressure outlet. A relatively large not increased further. Similarly, to avoid the risk of
numerical domain was utilized so that the near-flow less accuracy, the number of nodes was not
field was not affected by the outer boundary. In the decreased further. The numerical data was verified
numerical analysis, the 2nd order upwind method with the data of Holst et al. with a negligible
was applied to discretize the governing equations. maximum deviation of 2.3% at higher angles of
SIMPLE scheme was used to achieve pressure- attack [15].
velocity coupling. The convergence criteria was set
to O(6) in case of each and every dependent
residuals. To discretize the numerical domain an
unstructured grid was applied. In the present study, 3. Results
the aerofoil is considered to be thick aerofoil (21%
of the chord length). Therefore, near the leading CL and CD of modified NACA 0021 at Re 2 x 105
edge the laminar inlet condition would become and six different Uc/U are presented.

(a) (b)
Figure 2. CL vs Angle of attack (AOA) for unmodified and modified NACA 0021 airfoils for (a) numerical
approach at Re 2 x 105, (b) experimental appraoch at Re 2 x 105,
turbulent. However, since the experiment was A relation between CL and AOA for modified and
carried out in the wind tunnel, the flow was expected unmodified NACA 0021 airfoil for numerical and
to be fully turbulent as it entered the wind tunnel test experimental approaches has been shown in Figure-
section. In addition, the leading-edge spinning
2. For each case, the spinning increased CL and αe
cylinder acts as a trip and the transition point may
start from the leading edge itself. Therefore, the inlet for all Uc/U except Uc/U = 0. At Uc/U = 0, the CL
turbulent intensity specification of 5% can be decreased relative to the unmodified airfoil because
justified. In the present study, y+ values over the of the absence of addition energy involvement and
aerofoil range between 5 and 6 and increase near the the whole configuration restricted the scope for
cylinder to range between 12 and 16. For the masa smooth flow. The most significant finding is that the
and momentum conservation, the Reynolds maximum increase in CLmax was 35% and maximum
averaged form of the governing equation is given as:
delay in αe was 50% relative to the unmodified
𝜕𝑢̅𝑖 airfoil. As the spin of the cylinder increased, more
=0 amount of external energy was injected into the free
𝜕𝑥𝑖
stream airflow and it caused increase in velocity
𝜕𝑢𝑖 𝜕𝑢̅𝑖 𝜕𝜌̅ 1 𝜕 2 𝑢̅𝑖 𝜕𝜏𝑖𝑗 𝑅 gradient and consequently delay in flow separation
+ 𝑢̅𝑗 = − + +
𝜕𝜏 𝜕𝑥𝑗 𝜕𝑥𝑖 𝑅𝑒 𝜕𝑥𝑗 𝜕𝑥𝑗 𝜕𝜏 and this phenomenon is also reflected in figure 2.
The likely interference of the vibration of the
In the analysis, the domain was divided among about spinning cylinder cannot be ruled out. In the
500000 cells, consisting the mean cell size between experimental analysis, there were some irregular
2 × 10−4 m and 5 × 10−2 m. The node validation
pattern after αe compared to numerical analysis. This
study was carried out and it showed that the grid
interesting finding can be attributed to the irregular extreme amount of uncertainty in case of measuring
pattern of vibration of the cylinder due to the air density was ±0.02 kgm3 which accounted for
rotation. overall ±1.69% of the uncertainty in case of air
density. The extreme amount of uncertainty in case
of measuring chord length and wing span by a digital

(a) (b)
Figure 3. CD vs Angle of attack (AOA) for unmodified and modified NACA 0021 airfoils for (a) numerical
approach at Re 2 x 105, (b) experimental appraoch at Re 2 x 105

height meter was ±0.001 m which accounted for

overall ±0.67% of the uncertainty in case of chord
The observed correlation between CD and AOA is length and ±0.33%. in case of wing span. The
plotted in figure 3 showing mixed behaviour with extreme amount of uncertainty in case of measuring
respect to spin speed of the cylinder unlike the air velocity was ±0.1 m/s which accounted for
overall ±0.5% of the uncertainty in case of air
predicted pattern of behaviour in figure 2. As the
velocity. As the velocity term is squared in the
flow separation area over the surface of the wing equation for CL and CD, so the overall uncertainty
increased due to the inclusion of the cylinder, CD became ±1.01% in case of air velocity. The typical
also increased to a certain extent. In case of amount of uncertainty in case of measuring spin
numerical analysis, CD remained almost constant speed of the cylinder was ±12.6 rpm which
ranging from 0.2 to 0.21 whereas in case of accounted for overall ±4.2% of the uncertainty in
experimental analysis CD varied in a wide manner case of spin velocity. The characteristic amount of
uncertainty in case of measuring mass during
ranging from 0.13 to 0.21. This wide range of CD can
calibration was ±0.01 kg which accounted for
also be attributed to the irregular pattern of vibration overall ±0.2% of the uncertainty in case of
of the cylinder due to the rotation. The oscillation of calibration. Therefore, in the experimental
the cylinder created irregular gaps between the approach, the total amount of uncertainty was (1.012
cylinder and airfoil body and thus it obstructed the + 1.692 + 0.332 + 0.672 + 4.22 + 0.22)1/2 = ±4.71%.
airflow inside the gap to a certain extent in some
cases. A possible explanation for the regular patterns
in numerical analysis can be the absence of the
5. Conclusions
vibration of the cylinder unlike the experimental The study was set out to investigate the impact of
analysis. Magnus effect applied to a thick symmetric airfoil in
terms of lift and drag. It was the aim of the study to
objectively examine this impact in both numerical
and experimental approaches and found that flow
4. Experimental Uncertainty Analysis separation can be delayed to a certain extent if this
concept is applied. Contrary to the past literatures
It was possible that the findings of this study might which attributed this concept only for higher Uc/U,
represent some degree of errors and to determine this study provided a means of filling the research
that, an uncertainty analysis was carried out. The gap on this field. One of the most noteworthy
findings of this study was the realization of cylinder control as applied to two-dimensional airfoils.
oscillation and the negative impact of it. The insights J. Air. 28, 104-112 (1991).
gained from this investigation may assist the future
10. V. Modi, Moving surface boundary-layer
works where the vibration of the cylinder can be
control: a review. J. Fluid. Struct. 11, 627-663
minimized in an optimized manner. The current
(1997).
research also sheds a light on the issue of practicality
of including a leading-edge spinning cylinder for 11. S. Ahmed, A. Nazari, E. Wahba, Numerical
thick airfoil in aerospace applications. This concept analysis of separation control over an airfoil
can be implemented in many cases such as short section. Int. Rev. of Aero. Eng. 7, 61-68 (2014).
takeoff and landing (STOL) airplanes where high CL
at low AOA is necessary. Also, this research can be 12. N. Huda, T. Ahmed, S. Ahmed, A. Salam, R.
Afsar, M. Faisal, T. Ali, Study of NACA 0010
useful in some applications of wind turbines and
symmetric airfoil with leading edge spinning
primary and secondary control surfaces of airplane.
cylinder in a subsonic wind tunnel. Int. Conf.
Moreover, additional configurations of the position
Mech. Eng. (2015).
of spinning cylinder on different applications can be
analyzed to provide contributions to the current 13. A. Najdat, H. Mustafa, Effect of gap between
literature. airfoil and embedded spinning cylinder on the
airfoil aerodynamic performance Res. Dev.
Mat. Sci. 3, 1-10 (2018).

References: 14. M. A. Salam, M. T. Ali, M. Q. Islam,

Experimental Analysis of Aerodynamic
1. K. Yousefi and R. Saleh, Three-dimensional Performance on asymmetric NACA 23018
suction flow control and suction jet length Aerofoil incorporating a Leading-Edge
optimization of NACA 0012 wing. Meccanica Rotating Cylinder. In IOP Conference Series:
50, 1481-1494 (2015). Materials Science and Engineering, 912,
042027 (2020).
2. K. Yousefi, R. Saleh, P. Zahedi, Numerical
study of blowing and suction slot geometry 15. D. Holst, F. Balduzzi, A. Bianchini, B. Church,
optimization on NACA 0012 airfoil. J. Mech. F. Wegner, G. Pechlivanoglou, L. Ferrari, G.
Sci. Technol. 28, 1297–1310 (2014). Ferrara, C. Nayeri, C. Paschereit, Static and
dynamic analysis of a NACA 0021 airfoil
3. D. You and P. Moin, Active control of flow section at low Reynolds numbers based on
separation over an airfoil using synthetic jets experiments and numerical fluid dynamics. J.
J. Fluid. Struct. 24, 1349-1357 (2008). of Eng. for Gas Turbines and Power-
transaction,, 141, (2019)
4. H. Zhu, W. Hao, C. Li, Q. Ding, B. Wu,
Application of flow control strategy of
blowing, synthetic and plasma jet actuators in
vertical axis wind turbines. Aeros. Sci.
Technol. 88, 468-480 (2019).
5. K. Yousefi, R. Saleh, The effects of trailing
edge blowing on aerodynamic characteristics
of the naca 0012 airfoil and optimization of the
blowing slot geometry. J. Theo. Appl. Mech.
52, 165-179 (2014).
6. V. Kumar, M. Singh, M. Thangadurai, Effect
of free stream turbulence on flow past a
circular cylinder at low reynolds numbers. J.
Inst. Eng. India Ser. C. 100, 43–58 (2019).
7. V. Modi, F. Mokhtarian, T. Yokomizo, Effect
of moving surfaces on the airfoil boundary-
layer control. J. Air. 27, 42-50 (1990).
8. V. Modi, S. Munshi, G. Bandyopadhyay,
High-Performance Airfoil with Moving
Surface Boundary-Layer Control. J. Air. 35,
544-553 (1998).
9. V. Modi, F. Mokhtarian, M. Fernando, T.
Yokomizo, Moving surface boundary-layer