During the last years. sailplanes like the ASVV 24, DG 800 and others have used winglets. These winglets are not as big as the ones used 15 years ago. To improve the low speed performance with no drag increase for high speeds, a special winglet design is necessary. Last year I spoke with Martin Heide from Schleicher (ASH...) about this item. He told me how they had created the design of the ASW 24 winglet (published in separate articles not included here). The German F5B team will use winglets designed by me in Australia in Nov 94. (We await the results as the contest is being run at the time of this writing.)

What are the advantages of small winglets?

- reduction of induced drag
- reduction of laminar/turbulent bubbles at the upper surface of the wing tip area
- much better stall behaviour at low speeds
- higher roll speed in turns
- lower speeds can be flown
- the speed behaviour is not negatively influenced with the right winglet design

The former very high winglet designs wanted to increase the span of the wing and to reduce the induced drag in the way. The current designs mainly influence the boundry layer of the wing tip. The low Reynolds number from the low depth of the wing planform at the tip area brings mostly separation of the airflow. Imagine the airflow at the wing tip as a circulation around the airfoil. When the winglet also has an airfoil with circulation around it, you can understand that these two circulations must influence themselves when the winglet stands nearly with 90 degrees to the wing (fig 1). You can understand, that, when this theory is right that a dihedral less tban 90 degrees also influences the boundry layer in a positive way. This must be the case in the original designs like the Nimbus 4 and the Duo Discus from Schempp Hirth. That means polyhedral and better multi dihedral forms are the best. But this means also, that a wing with a low bending stiffness (but must have a good torsion stiffness) works better than a stiff wing! This wing which could have an elliptical dihedral form is theoretically the best.
 

fig. 1

Back to the winglet. How must a winglet be designed that works in the correct way? Martin Heide could not give me an equation that works in every case, he said, that we must test it. We have done it, at first only on one side of the wing.
Some rules, however, have to be fixed:

- the winglet should be fixed in flight as illustrated in fig. 2
 

fig. 2

- the sweep angle should be 30 degrees (fig. 3)
 

fig. 3

- the ratio between tip chord and the chord of the winglet mean chord should be 0.6. That means, a tip chord of 150mm should get a winglet with with a mean chord of 90mm.

- the winglet should be mounted at the wing parallel to the direction of flight or airplane centerline (fig 4), or with an angle that a lift will be produced from the winglet. The lift should not be very high.
Normally an angle of  zero degrees are sufficient.

- because of the lift, the winglet should be twisted itself. With the drag/lift polars of the aerofoil used for the winglet, the twist should be taken that the lift at the tip of the winglet is zero (for example 2 degrees, fig 4).
 

fig. 4

- aspect ratio of the winglet should be 0.6, for example, when the mean chord of the winglet is 90mm, the chord of the winglet tip is 54mm.

- the winglet aerofoil should be for very low Reynolds numbers. We have taken the aerofoil that we have used at the wing tip.
 

References:
1) Waibel, G., Aerokurier 6/1993, pp. 99-101
2) Pletschnacher. P., Flug Revue + Flugwelt 7/1978, pp. 56-57
3) Albat, A., Idaflieg Berichtsheft 1992, pp. 97-111
 

Best of luck and good flights! Christian Baron