Analysis of 30°-90° Blended Winglet Additions to Cessna 172, Piper Malibu, and Boeing 737 Wings

NSF DoD REU Milestone. Award # 1950207 Under Supervision of Dr. Soloiu and Dr. Rahman

John CroweFernando Ramos

Contents1. Objectives

2. Background Information

3. Experimental Methods

4. Simulations

5. Results

6. Conclusions

7. Acknowledgments

Objectives● Gather information that will be helpful for winglet design and optimization.

● Verify that a blended winglet addition can increase aerodynamic efficiency.

● Use Computational Fluid Dynamic (CFD) software to find the respective coefficients of lift and drag for each wing-winglet configuration.

● Calculate the coefficients of lift and drag experimentally for comparison.

● Find the cant angle that creates maximum aerodynamic efficiency for each wing.

● Discover the structural consequences of winglet additions.

Background Information

Aerodynamic Forces

1. There are only 4 forces acting on a plane at all times.

2. The aerodynamic efficiency of an aircraft can be determined by the lift to drag ratio.

3. 80% of total drag is caused by friction and induced drag [2]. ○ Frictional drag: Due to skin

friction on the boundary layer.○ Induced Drag: Caused by

wingtip vortices.Figure 1: Main Forces acting on aircraft [1].

How is lift generated?

1. The Bernoulli’s Principle says that:

“A decrease in pressure results in an increase of velocity”

1. Particles of air are split at the leading edge of the airfoil. The particles following the path at the top move faster than the ones at the bottom.

Figure 2: Airflow speed around airfoil [2].

● The faster air stream creates low pressure at the top of the wing and the slower air stream creates a high pressure below it. This difference creates lift.

How are vortices formed?

1. Formed at the tip of a wing due to pressure differences converging.

2. Cause increased induced drag due to vortex hitting wing.

3. Only exist in 3D, real world cases○ Do not exist for 2D, making

them near impossible to analyze by hand.

Figure 3: Wingtip vortex [3].

Visualization

Figure 4: Vortex Formation [4].

How do vortices create induced drag?

1. Vortex causes a downwash flow that tilts the lift vector backwards creating induced drag.

2. This lowers the fuel efficiency of the aircraft.

Figure 5: Induced drag due to vortex [5].

How do winglets reduce induced drag?

1. The winglet creates lift that is

perpendicular to the local relative wind.

2. The generated lift force has a

component that opposes the induced

drag.

Additionally, it reduces the vortex strength

and it moves it away from the wing. Figure 6: Reduction of induced drag due to winglet [5].

Wind Tunnel Experimental Method

● Each wing and winglet combination was attached to a strain transducer setup and pressure tubes

were attached to the surface of each wing.

● Setup was placed in the wind tunnel. The speed of the airflow was 10 m/s.

● Lift and drag force from transducer were recorded.

● Pressure difference from top and bottom of the wing was recorded to gather lift force.

Selected wing geometries

Technical and wing specifications were based on real modes [7-10].

Winglet Geometries

1. This research aimed to determine the aerodynamic/structural effect of the change of cant angle of the blended winglets.

2. Winglets with cant angles of:○ 30°○ 60°○ 90°

Figure 7: Cant angle of a winglet [11].

Test Models 1. SolidWorks was used to design

the wing and winglet models.2. This models were transferred to

ANSYS for airflow simulation.3. Models were also 3D printed and

assembled together using crazy glue.

4. Primer was sprayed over the wing and then they were smoothed out using sandpaper.

Figure 9: SolidWorks blended winglets models.

Figure 8: SolidWorks model of Cessna 172 wing.

● SolidWorks was used to design the wing and winglet models.This models were transferred to ANSYS for airflow simulation.

● Models were also 3D printed and assembled together using crazy glue.

● Primer was sprayed over the wing and then they were smoothed out using sandpaper

Figure 11: Piper PA-46 3D printed model. Figure 10: Cessna 172 3D printed model.

3D printed Wing Models

Figure 12: Boeing 737-300 3D printed model.

3D Printed Winglet Models

Figure 13: 3D printed blended winglets.

● Frequency controlled wind speed.

● Air speed range is 1 to 13 m/s.

● Inlet has a mesh to homogenize the

flow.

● 4 ft2 wind outlet.

Figure 14: Strain Transducer Setup.

Wind tunnel

Wind tunnel Experimentation -

Reynold’s Number

● Re = Reynold’s number.

● Dimensionless number to determine flight

conditions.

● Reynold’s number for each wing test:

Wing Type Re

Cessna 172 7.29e6

Piper Malibu PA-46 5.16e7

Boeing 737-300 5.53e7

Equation for Reynold’s number:

Table 1: Reynolds Number of each Wing .Where:● ρ = Density of fluid.● u = Speed of fluid.● L = Characteristic length.● μ = Viscosity of fluid.

1. Two S-type load cells were used to determine the overall lift and drag acting on the wing.

2. Vertical sensor determines overall lift and horizontal sensor determines overall drag.

3. Arduino was used to get a calibrated force reading from both S-type load cells.

4. An LCD was used to display the readings at all times without the need of a computer.

Figure 15: Strain Transducer Setup.

Strain Transducer

Wind tunnel Experimentation -

Strain Transducer - Arduino1. Arduino Mega 2560 was used for

data processing. 2. The load cell output voltage signal

is too small for the Arduino to read (~10 mV).

3. Voltage signal from load cells was amplified using HX711 breakout board. It has a 24-bit Analog to Digital converter (ADC) [12].

4. It allowed for an accuracy of 5V/224 = 0.298µV (on a 5V source).

Figure 16: Strain Transducer Arduino sketch [13].

Strain Transducer - Calibration1. A code written by Nathan Seidle [14]

was used to determine the coefficient factor that relates the voltage signal to the mass placed over sensor.

2. Once the calibration factor was determined, it was input to a program that determines the force respective to the mass reading with the following equation: Figure 17: Graph that shows accuracy of program used.

Strain Transducer

Wind tunnel Experimentation -

1. The strain transducer determines

the overall Lift (L) and Drag (D)

caused by the wing. The values

recorded are then input in

Equations 1 and 2 to determine

the coefficient of lift and drag.

Where: - CL = Coefficient of lift- CD = Coefficient of drag- L = Lift- D = Drag- ρ = density of air- V = velocity of air- A = wing’s surface area

● Multiple Airspeed Sensor Breakout Board MPXV7002DP sensors were used to determine the pressure at the surface of the wings.

● Arduino Mega 2560 was used to receive data from the sensors and output the pressure readings on computer.

● Plastic tubes were placed inside wings. One end was cut at the surface of the wing and the other was connected to the pressure sensors.Figure 18: Strain Transducer Setup.

Pressure Sensors

Windtunnel Experimentation -

1. Arduino Mega 2560 was used for data processing.

2. It has a 10-bit Analog to Digital Converter (ADC).

3. Allowed for an accuracy of 3.3V/210 = 3.22mV.

Figure 19: Pressure Sensors Arduino sketch [15].

Pressure Sensors - Arduino

Pressure Sensors - Calibration Curve

Figure 20: Graph that shows accuracy of program used.

The slope of the calibration curve allowed to find the factor value that correlates the voltage signal to its respective pressure.

Equation 3 was implemented in the code to determine the average pressure difference.

● Average top and bottom pressure allow

to calculate the average pressure

difference that is acting on the wing.

● This relationship can be solved for

average force:

● Substituting Lift in Equation 1, the

coefficient of lift can be determined:● It is known that:

- Pressure SensorsWindtunnel

Experimentation

Figure 21: Experimental Setup.

Computational Fluid Dynamic Simulations

Simulation Equations

● The Navier Stokes governing equations for flow calculation (Eq. 1-4.)

Simulation equations (cont.)

● For the simulations, the k-omega Shear Stress Transport (SST) model within ANSYS

fluent was used to model the turbulence of the airflow as it has high accuracy for

predicting boundary layers and flow separation [5].

CFD Simulations Using ANSYS Fluent

1. Large, meshed domain to capture vortex formation around wing & winglet.

2. Angle of attack simulated at 0°, 5°, and 10° for all wing/winglet configurations.

1. Observed outputs:○ Coefficients of lift and drag.○ Pressure and velocity contours.○ Overall vortex formation.

Figure 22: Flow domain for Cessna 172 with 60° winglet. Figure 23: Mesh for Cessna 172 with 60° winglet.

FEA Simulations1. Surface pressure imported

from CFD Simulations2. Determined maximum stress,

strain, modal stress, and transient stress for each model

3. Al 7075-T6 used for material of model

Material Young’s Modulus

Poisson's Ratio Density

Aluminum Alloy 7075-T6

71.7GPa 0.33 2.81g/cm3

Figure 24: Imported pressure, gravity and fixed support for Boeing 737-300 wing

Table 2: Properties of material used in FEA[16].

Experimental and Simulation Results

Cessna 172: Wind Tunnel

Cessna 172: SImulation

Cessna 172: Example velocity contours

● No winglet at α = 5 ● 60° winglet at α = 5

Piper Malibu PA-46: Wind Tunnel

Piper Malibu PA-46: Simulation

Piper PA-46: Example velocity contours

● No winglet at α = 5 ● 60° winglet at α = 5

Boeing 737-300: Wind Tunnel

Boeing 737-300: Simulation

Boeing 737-300: Example velocity contours

● No winglet at α = 5 ● 60° winglet at α = 5

Finite Element Analysis Results

Cessna 172

Figure 25: Deformation for Cessna 0° (top) and 90° (bottom) winglet at 5° angle of attack.

Piper Malibu PA-46

Figure 26: Deformation for Piper PA-46 0° (top) and 90° (bottom) winglet at 5° angle of attack.

Boeing 737-300

Figure 27: Deformation for Boeing 737 0° (top) and 60° (bottom) winglet deformation at 5° angle of attack.

Conclusions

● All winglets tested more beneficial over wing without winglet

● The 60° winglet was the best for the Cessna 172, the 30° winglet was the best for the Piper PA-46, and the 90°

was the best for the Boeing 737-300.

● As airspeed increases, the positive effect of a winglet reduces

● Lower cant angles more beneficial for lower airspeed, higher cant angles for higher airspeed.

● Winglets do increase stress on wing, but not by large margin.

● Results mostly comparative, further research would allow for exact results for each case for all operating

conditions

Acknowledgements

● Thank you to Dr. Rahman and Dr. Soloiu for allowing us to use your instruments and facilities, as well

as allowing us the amazing opportunity to be a part of this program.

● We also thank the Department of Defense and National Science Foundation for funding the research

involved in this study under Assure REU Site Award 1950207

● Thank you to Charles Fricks, Evan Cathers, and Eric Pernell in assisting us hands-on throughout the

research with preparing simulations, as well as preparing the wind tunnel and models for these

experiments.

● Thank you to everyone involved in the REU program for all of your help and advice.

References1. Toronto, P. (2020, June 22). Forces of flight. Pearson Airport. https://www.torontopearson.com/en/whats-happening/stories/whyyz/forces-of-flight.2. SKYbrary Wiki. AP4ATCO - Aerofoil Terminology - SKYbrary Aviation Safety. (n.d.). https://www.skybrary.aero/index.php/AP4ATCO_-_Aerofoil_Terminology.3. This is how winglets work. Online Flight Training Courses and CFI Tools. (n.d.).

https://www.boldmethod.com/learn-to-fly/aerodynamics/how-winglets-work-to-reduce-drag-and-how-wingtip-vortices-form/.4. “Induced Drag.” Tangvald, November 9, 2013. https://tangvald.wordpress.com/2013/11/09/induced-drag/.5. Guerrero, J., Sanguineti, M., Wittkowski, K. (2018). CFD Study of the Impact of Variable Cant Angle Winglets on Total Drag Reduction. Aerospace, 5(4), 126. 6. Haddad, Nicolas El. “Aerodynamic and Structural Design of a Winglet for Enhanced Performance of a Business Jet,” 2015.7. Cessna Aircraft Company. (1998). Cessna Model 172S Information Manual (5th ed.). Cessna Aircraft Company.8. Estaff, and Editorial Staff. “Piper Malibu-Mirage.” AVweb, April 21, 2016.

https://www.avweb.com/features/piper-malibu-mirage/#:~:text=Malibu%20performance%20puts%20the%20airplane,at%20FL250%20at%2075%20percent.9. Brady, Chris. “Detailed Technical Data.” The Boeing 737 Technical Site. Accessed June 30, 2021. http://www.b737.org.uk/techspecsdetailed.html

10. UIUC Airfoil Data Site. UIUC Applied Aerodynamics Group. Accessed June 30, 2021.11. Guerrero, J. E., M. Sanguineti, and K. Wittkowski. “Variable Cant Angle Winglets for Improvement of Aircraft Flight Performance.” Meccanica 55, no. 10 (2020):

1917–47. https://doi.org/10.1007/s11012-020-01230-1.12. Avia Semiconductor Co. “24-Bit Analog-to-Digital Converter (ADC) for Weigh Scales”13. “Load Cell Straight Bar 0-1/5/10/20/50kg Force Pressure Weight Sensor / HX711 Module for Arduino IoT.” Shopee Malaysia. Accessed July 23,

2021.https://shopee.com.my/Load-Cell-Straight-Bar-0-1-5-10-20-50kg-Force-Pressure-Weight-Sensor-HX711-Module-for-Arduino-IoT-i.33091591.568966941?__hybrid_pc__=1&stm_referrer=https%3A%2F%2Fwww.google.com%2F

14. Load Cell Amplifier HX711 Breakout Hookup Guide. Sparkfun. Accessed July 23, 2021. https://learn.sparkfun.com/tutorials/load-cell-amplifier-hx711-breakout-hookup-guide/all.

15. Hrisko, Joshua. “Arduino Pitot TUBE Wind Speed and Airspeed Indicator - Theory and Experiments.” Maker Portal. Maker Portal, April 2, 2021. https://makersportal.com/blog/2019/02/06/arduino-pitot-tube-wind-speed-theory-and-experiment.

16. Metals Handbook, Vol.2 - Properties and Selection: Nonferrous Alloys and Special-Purpose Materials, ASM International 10th Ed. 1990.