Active Winglets as Multi-Axis Effectors

1. Numerical Analysis

> Application to Pitch Control:

When moved in tandem, active winglets allow to nose up or down the flying-wing airframe and/or to adjust it to a new longitudinal equilibrium state (i.e. flying at a different speed or climb/glide angle). This has been ascribed to the relocation of the aerodynamic centre (blue spot in the figures below) that occurs whenever winglets mounted on a swept planform are moved around their root axis, while the chordwise position of the center of gravity (red spot in the figures below) remains the same. Hence pitch control with active winglets is achieved through a dynamic static margin, which is in contrast with elevators which generate a control moment by altering the zero-lift pitching moment. Pitching moments predicted via an enhanced VLM method are plotted below for the active winglets moved in tandem: the baseline, trimmed states corresponds to CL = 0.2 with both winglets deflected either upwards or downwards at 45 degrees from the wing plane; one winglet represents 33% of the overall semispan.

> Application to Roll Control:

Further VLM results also show that the active winglets, when moved differentially (say one is set planar and the other is folded up or down), allow to roll the aircraft because one wing half produces then more lift than the other. A proverse coupling between roll and pitch has been found out when controlling the lateral balance in that manner (a nose-up pitching moment appears along with the intended roll, which is desired to achieve a level turn). Some coupling as well between roll and yaw appears due to the sideforce acting on the folded winglet: with up-winglet the yaw is in the same direction as the intended roll, which is desired to achieve a coordinated turn, the opposite is true with down-winglet. Figure below shows the attainable steady roll rate plotted against the deflection angle of the controlling winglet (1 rotational DOF assumed for the airframe, damping-in-roll derivative and rolling moment predicted with an enhanced VLM).

> Preliminary Conclusions:

So far the concept of active winglets appears to be a promising alternative to conventional control surfaces such as ailerons, elevators and rudders as far as basic manoeuvres are concerned. Active winglets enable control moments about multiple axes, forming then a highly coupled flight control system; this is in contrast with conventional control surfaces which form a decoupled control system. However, a single pair of active winglets cannot substitute for all the conventional control surfaces at the same time if one wants to get a full control envelope. Indeed, numerical simulations showed us that one can achieve a trimmed level turn (i.e. pitching, rolling and yawing moments are zeroed out while in banked flight) with a single pair of active winglets as sole control effectors, but only for a specific turn radius. To access a continuous range of turn radii with active winglets as control effectors, one has then to combine their action with some elevators for instance.

2. Experimental Analysis

> Wind-Tunnel Testing:

The model (Anakin-1) was installed inside a closed test section (2.1 m * 1.5 m), closed circuit wind tunnel whose maximum operating freestream velocity was 60 m/s. The freestream velocity chosen for this investigation was 10 m/s, giving a Reynolds number of 230000 based on the wing root chord. The freestream turbulence level at the model station was approximately 0.2%. The model was mounted at mid-height in the test section, on top of a support strut connecting the model to a high-frequency, dynamic load cell mounted to the underside of the floor of the test section. Access to the wind tunnel test section for the support strut was provided by a cutout in the wind tunnel floor which was covered, during wind tunnel testing, by two thin sheets of fibreboard. Each sheet was constructed to ensure no contact between the supporting strut and the test section was possible. Four high-tension wires were also installed between the active balance plate and the top of the support strut to increase the stiffness of the entire support system thereby improving the natural frequency characteristics of the combination.

Force and moment data obtained from the model were acquired using a JR3, multi-axis load cell in combination with conditioning electronics and a DAQ card installed in a PC. Calibration of the load cell conducted prior to wind-on test conditions indicated a percentage error in the reading of all forces and moments to be less than +/- 5%. Estimates of the static support tare for all six degrees of freedom were also obtained prior to wind tunnel tests with all data presented hereafter corrected for these results. No blockage correction was applied however, since the frontal wing area to test cross-section area ratio was less than 5%.

The observed moments attainable by folding up or down the right winglet while the left one remains planar are plotted above versus the right-winglet dihedral angle (long-winglet case; from top to bottom: Cl, Cm, Cn; left column: CL = 0.2; right column: CL = 0.4; Re displayed for experimental data is based on root chord; for comparison the VLM results are also shown -- red solid lines). The agreement between predicted and observed moments is satisfactory enough to demonstrate the viability of the concept.

> Flight Testing:

Anakin-1 has been fitted with an electric motor driving a propeller (pusher configuration) and flight-tested. The only control surfaces are a pair of elevators (main pitch effector) and a pair of active winglets (main roll and yaw effectors). The throttle and the elevator and winglet actuators were radio-controlled by a ground operator.

The long active winglets (each accounting for 33% of the structural semispan) provided too much roll and lead to spiral divergence. We therefore decided to chop them off in halves. The new configuration (codename Anakin-1.5) featured then short winglets, each representing 20% of the structural semispan. Still too much roll was provided when deflecting one winglet off the wing plane, so it was decided to limit the amount of deflection to approximatively 10 degrees upward off the winglet neutral plane. The winglet neutral position corresponds to 10 degree up-winglet w.r.t. the main wing plane, so that the lateral stability of the sweptback configuration is increased in straight flight.

The following movies show successful level turns with Anakin-1.5, initiated by deflecting upwards the winglet inside the intended turn, and sustained by combining up-elevator and nearly-symmetrical winglet deflections: