Sites Inria

Version française

Automatic control


Improving the aerodynamics of aircraft by controlling air flow

Maxime Feingesicht and the representative model of an aircraft wing used in the wind tunnel of the Onera, Lille.

As part of the regional partnership CPER (state-region planning contract) ELSAT 2020, the ContrATech sub-program focuses on new technologies in flow control able to reduce consumption and improve flight aerodynamics. Maxime Feingesicht is doing his thesis (co-funded by the graduate engineering school École Centrale de Lille and the Hauts-de-France regional council) on control issues. He has agreed to answer some questions on his research, which is supervised by Jean-Pierre Richard from École Centrale de Lille, Andrey Polyakov from Inria and Franck Kerhervé from the University of Poitiers.

Could you tell us about yourself, as well as the Non-A team?

After a specialization in electrical engineering and electronics, as well as a master's in automation, I began my PhD in January 2015. The Non-A (for Non-Asymptotic) team from the Inria Lille - Nord Europe center* is developing an estimation theory with mathematical approaches (algebra, non-linear analysis) that lead, in particular, to the estimation in finite time of the derivatives of noisy signals. These algorithms converge after a set period of time, unlike the "traditional" asymptotic approaches in automation.

What is the focus of your work?

Today we are working in one of the wind tunnels of the ContrAero regional platform, with our partner ONERA (the French Aerospace Lab), in Lille. We are using a representative model of an aircraft wing profile fitted with a flap, itself equipped with eight "hot film" type sensors: small, heated surfaces that are cooled by the air flow. The resulting variation in tension provides us with information on the friction on this surface, and therefore on the air flow regime (attached or detached) We can therefore calculate, in real time, an active control of upstream air jets, that modify the air flow to maximize tension - an indication of the reduction or even total elimination of the detachment, which leads to an increase in performances.

By actively modifying the aerodynamics of the wing, what is your objective?

The ONERA application aims to improve lift. In practical terms, this will make it possible to limit stall or "air pockets", these air swirls that disrupt flights and reduce passenger comfort; other applications concern the reduction of take-off or landing distances, or the lightening of the structures. Active aerodynamic control is also a promising technology in order to reduce aircraft fuel consumption and therefore their pollutant emissions, or to reduce noise. More generally, for all air or land vehicles; on a car or a train, aerodynamic resistance (the famous "drag") is considered a major source of energy consumption above 50 km/h. Estimations consider that a 25% reduction in this resistance would reduce CO2 pollution by 107 tonnes per year. We are starting work on these land vehicle aspects with LAMIH (Laboratory of Industrial and Human Automation control, Mechanical Engineering and Computer Science), another of the project's partners. Another advantage: reducing vibrations, and therefore improve the service life of the components.

What about your model? How is it innovative?

Our model has a shape that had not yet been used in air flow control. It has a non-linear component (to take into account the fact that the sum of two air flows does not cause the sum of each of their effects) and also contains lags (to represent air displacement times along the wing). In addition, it is very simple, as we are looking for the best lags and coefficients in order to minimize the number of variables of the model, whilst also retaining a proper representation of the data. Our goal is to integrate this into a micro-controller - we are currently using an Arduino - responsible for rapidly carrying out the calculations with a low computing power. As a comparison, in general we use 10 to 20 variables to obtain a level of reliability comparable to the usual fluid mechanics models that consist of several thousands, or even tens of thousands. Our solution is therefore easily implantable with its small size and virtually zero computing power.  Finally, our control algorithm is of the "sliding mode" type (the system is "slid" into a chosen direction by wedging it between two opposing controls, a bit like getting a drunk man to walk by propping him up between two strong men) well-known to automation specialists, but never used in fluid mechanics despite its relevance in terms of robustness and the fact that it corresponds very well to an all-or-nothing actuation. Those are innovative points.

Any prospects for vehicles other than aircraft?

For land vehicles, the application is quite straightforward and would enable a reduction in stalls and therefore consumption and pollution. Our technology could be applied to drones, however its size would have to be reduced beforehand, i.e. mainly the space taken up by the actuator valves. In water (ships and submarines), flows respond to different - but similar - hydrodynamic laws: there is nothing to prevent us from considering adapting our technology in the future, however this require further research.

ContrATech: A partnership between research and local government players

This promising project contributes to ELSAT2020, a project that brings together regional research on transport. It is the result of a collaboration between institutions such as the Hauts-de-France regional council, the state and the ERDF (Europe) that co-finance ContrATech. The other part is shared between the partner universities and schools (Centrale Lille, University of Lille, UVHC (University of Valenciennes and Hainaut-Cambresis)) and organizations (CNRS, Inria, ONERA).

Within these bodies and partners, several laboratories are working together: CRIStAL, IEMN, LAMIH and LML for the CNRS, the Non-A team at Inria, the aerodynamics team at ONERA. The partnership relies on the ContrAero regional wind tunnel platform, which brings together the major regional aerodynamic facilities. Finally, the researchers, research lecturers and engineers mobilized on this project are paid by ContrATech partner establishments.

Such an interdisciplinary consortium is rare, and even unique within a single geographical area.

A European cross-border Interreg 2 Seas 2014-2020 project application file has been submitted, opening up new possibilities. This European Territorial Cooperation program, which covers the UK, France, the Netherlands and Belgium (Flanders) could - if the application is validated - enable the continuation of flight testing on a large drone located at the University of Southampton: an essential building block towards the industrialization of the technology.

*(joint with École Centrale of Lille, the CNRS and Université Lille 1 - Sciences et Technologies) within the UMR 9189 CNRS-Centrale Lille 1 - Université de Lille - Sciences et Technologies, CRIStAL.

  • Wind tunnel of the Onera, where the model of the aircraft wing is located.

  • Wind tunnel of the Onera, where the model of the aircraft wing is located.

  • Wind tunnel of the Onera, where the model of the aircraft wing is located.

  • Model of an aircraft wing profile fitted with a flape, equipped with 8 sensors - Onera.

  • Wind tunnel of the Onera, where the model of the aircraft wing is located.

Keywords: Controle des écoulements Mécanique des fluides Aérodynamique Équipe-projet Non-A ContrATech ONERA Automatique