SCIENTIFIC NEWS AND
INNOVATION FROM ÉTS
Artificial Intelligence Methods for a Novel Morphing Wing Actuator - By : Shehryar Khan, Ruxandra Botez, Teodor Lucian Grigorie,

Artificial Intelligence Methods for a Novel Morphing Wing Actuator


Shehryar Khan
Shehryar Khan Author profile
Shehryar Khan is a research assistant and PhD student at ÉTS. His research interests lie mainly in the design of electromechanical system models, and in the development of control laws.

Ruxandra Botez
Ruxandra Botez Author profile
Ruxandra Mihaela Botez is a professor in the Systems Engineering Department at ÉTS. She specializes in modelling and simulation for aircraft, helicopters, aerial systems and morphing wings.

Teodor Lucian Grigorie
Teodor Lucian Grigorie Author profile
Teodor Lucian Grigorie is presently senior researcher at the Military Technical Academy in Bucharest, Romania, and PhD supervisor at University Politehnica of Bucharest, Faculty of Aerospace Engineering and an associate professor at ÉTS.

Aircraft in flight

From julesfuel.com. No known copyright restrictions.

SUMMARY

New research on morphing wing technologies was carried out by the LARCASE team in a second major research project, “Multi-Disciplinary Optimization 505” - MDO 505, intended to obtain fuel consumption economy using these technologies on a real aircraft wing. The wing was equipped with an aileron, and morphed using a new actuation mechanism, based on BLDC electric motors, manufactured in-house at ÉTS. A key milestone in the project was modelling the new morphing actuator and developing related intelligent control methods.

INTRODUCTION

Air travel statistics from International Civil Aviation Organization (ICAO) indicate that 4.3 billion passengers opted for air travel in 2018. This reflects a 6.1% rise in the number of passengers compared to 2017.  Aviation counts for 2% of the global carbon emissions from all other sources, and compared to road transportation, which produces 74% of carbon emissions, aviation accounts for 12% of emissions in the transportation industry.

Also, according to the New York Times, a single trip from New York to California creates the same amount of carbon emissions that a car can emit over an entire year. According to some estimates, 20,000 planes fly annually and serve around 3 billion passengers. It is forecasted that approximately 50,000 planes will be in service and flying by 2040 [13]. This increase in flight operations will contaminate the earth climate and will endanger living beings, both on earth and in the oceans.

If you are one of those people who fly frequently, you are a contributor to the global carbon emissions. The million-dollar question is: how to cut down on global carbon emissions? In fact, for longer trips, air travel is better since aircraft consume less fuel in the cruise phase of the flight, compared to take-off and landing. According to NASA, 25% of aircraft CO2 emissions come from the take-off, landing, and taxiing phases [13].

Carbon emission during flight

Figure 1. Aircraft carbon emission profile [1]

As an air traveller, buying carbon offsets is another way to help bring about carbon neutral growth. Buying carbon offsets will help various initiatives around the world to replant trees, hence reducing existing CO2 emissions.

Based on the technological road map laid down by the ICAO, various research organizations and industries around the world initiated collaborations to develop aircraft technologies that can contribute to carbon emission reductions. One such initiative, CRIAQ MDO 505, was launched at the Laboratory of Applied Research in Active Controls and Aeroservoelasticity (LARCASE).

MORPHING WING EXPERIMENTAL MODEL

The project started with the aim of developing morphing technology to improve laminar flows over the wingtip. The experimental model, manufactured at ÉTS, is a full-size scale wingtip for a real aircraft, which includes an aileron (Figure 2). The resulting model had the same structure and stiffness as that of a real aircraft wing.

To morph the model, the choice went with a flexible upper surface, manufactured with composite materials. Its actuation is performed by a system integrating four similar electric actuators, arranged in two lines placed respectively at 32% (Act. #1 and Act. #3) and 48% (Act. #2 and Act. #4) from the chord (Figure 3).

To monitor airflow over the wing’s upper surface, 32 high-precision Kulite pressure sensors were installed on the flexible skin. The resulting pressure data were processed in real time in order to provide information related to the laminar-to-turbulent transition location; the Fast Fourier Transforms (FFT) for the acquired pressure data were visualized in real time. Infrared (IR) thermography was used as an additional method to evaluate the laminar-to-turbulent transition location, but this time, over the entire wing upper surface, not only in the pressure sensors section.

Experimental model of a morphing wing

Figure 2. Morphing wing experimental model in the NRC wind tunnel testing

In the previous LARCASE CRIAQ 7.1 project, morphing actuators were made up of shape memory alloys (SMA) [2]-[6]. Although an improvement in laminar flows was realized both in simulations and experimentally, the actuation response was slow due to the innate nature of shape memory alloys (SMA). Experience from CRIAQ 7.1 led the LARCASE team to follow the trend of more electric aircraft technologies. It was then decided to design a morphing actuator based on the brushless DC motor, the first of its kind, as shown in the following figure.

Morphing wing actuator

Figure 3. (a) Morphing wing actuator (b) Four actuators on both actuation lines

CONTROL STRATEGY

2D and 3D aerodynamic analyses and optimizations were carried out using the Fluent and X-Foil software for various flight conditions obtained as combinations of Mach numbers (M), angles of attack (α) and aileron deflection angles (δ). For each flight condition, optimal values for the four actuation distances were established. The aim of the aerodynamic analysis was to find the actuator positions for each flight case that can improve laminar flows over the wing. Figure 4 [7] presents a Monte Carlo map for one of the cases with the optimization results plotted in. The Monte Carlo map shows all possible combinations of two actuator displacements, essentially all possible results, and optimized results can be plotted in to estimate how close was the optimization code.

Optimal values for α, M, and δ

Figure 4. Monte Carlo map with optimization results for α=2°, M=0.2, and δ=4° flow case

A key milestone in the project was modelling the new morphing actuator and developing related intelligent control methods (Figure 6.a) in order to perform the position control in the desired range of -3 mm to 3 mm [8]-[12]. Neural networks take their inspiration from the human brain, where millions of neurons are interconnected and perform complex decision-making processes (Figure 6.b).

Morphing actuator

Figure 5. Morphing actuator Matlab/Simulink modelling and control

Neurons in the human brain

Figure 6. Neurons in the human brain

A method called ANFIS was implemented from the branch of AI known as supervised learning technique. The necessary data was logged from the conventional controller to design an artificial neuro-fuzzy controller. The ANFIS method benefits from the learning ability of neural networks coupled with the reasoning ability of the fuzzy control. Figure 7 presents the flow chart of the fuzzy controller design using ANFIS. Five membership functions were used during the ANFIS training for each set of training data positions and current controls. Figure 8 shows the training of the fuzzy position controller using ANFIS.

ANFIS flow chart

Figure 7. ANFIS flow chart

Training data of the fuzzy position controller

Figure 8. Training of the fuzzy position controller using ANFIS

CONTROL RESULTS

Figure 9 presents the experimental testing system architecture of the morphing wing actuation mechanism, while Figure 10 shows the actuation results obtained for one of the four actuators for repeated step inputs in the range -3 mm to 3 mm.

Testing system of a morphing wing

Figure 9. Architecture of the experimental testing system

Actuator position

Figure 10. Actuation results obtained for an actuator between 3mm and -3mm

The infrared (IR) thermography method was used to estimate the laminar to turbulent transition location over the entire upper surface of the morphable wing. An IR evaluation of the transition position displacement through wing morphing is presented in Figure 11. It can be observed that the gain related to the transition position is about 6% of the chord for the presented flow case.

Infrared evaluation of a laminar to turbulent transition displacement

Figure 11. An IR evaluation of the transition position displacement through wing morphing

The next IR video shows how the colour of the analyzed wing changes in real time when it is morphed.

Real time evaluation of the transition position using IR techniques

CONCLUSION

For the great majority of wind tunnel flight test cases, the project research team observed that the morphing technology improved the average position of the laminar to turbulent flow transition over the whole wing with more than 2.5% of the wing chord. Therefore, it seems likely that for future generations of aircraft, the morphing wing technology will be a serious alternative to the current rigid control surface.

Additional Information

Shehryar Khan, Teodor Lucian Grigorie, Ruxandra Mihaela Botez, Mahmoud Mamou, Youssef Mébarki, “Fuzzy logic based control for a morphing wing-tip actuation system:design, numerical simulation and wind tunnel experimental testing” published as a special issue in the “Morphing Aircraft Systems”, journal of biomimetics, 2019, 4, 65.

Shehryar Khan

Author's profile

Shehryar Khan is a research assistant and PhD student at ÉTS. His research interests lie mainly in the design of electromechanical system models, and in the development of control laws.

Program : Automated Manufacturing Engineering 

Research chair : Canada Research Chair for Aircraft Modeling and Simulation Technologies 

Research laboratories : LARCASE – Aeronautical Research Laboratory in Active Control, Avionics and Aeroservoelasticity 

Author profile

Ruxandra Botez

Author's profile

Ruxandra Mihaela Botez is a professor in the Systems Engineering Department at ÉTS. She specializes in modelling and simulation for aircraft, helicopters, aerial systems and morphing wings.

Program : Aerospace Engineering  Automated Manufacturing Engineering 

Research chair : Canada Research Chair for Aircraft Modeling and Simulation Technologies 

Research laboratories : LARCASE – Aeronautical Research Laboratory in Active Control, Avionics and Aeroservoelasticity 

Author profile

Teodor Lucian Grigorie

Author's profile

Teodor Lucian Grigorie is presently senior researcher at the Military Technical Academy in Bucharest, Romania, and PhD supervisor at University Politehnica of Bucharest, Faculty of Aerospace Engineering and an associate professor at ÉTS.

Program : Aerospace Engineering  Automated Manufacturing Engineering 

Research chair : Canada Research Chair for Aircraft Modeling and Simulation Technologies 

Research laboratories : LARCASE – Aeronautical Research Laboratory in Active Control, Avionics and Aeroservoelasticity 

Author profile