![]() |
|
Glenn
|
At the end of 1901, the Wright brothers were frustrated by the flight tests of their 1900 and 1901 gliders. Neither aircraft performed as well as predicted using the design methods available to the brothers. Based on their measurements, the 1901 aircraft only developed 1/3 of the lift which was predicted by using the Lilienthal data for lift coefficients. During the fall of 1901, the brothers began to question the aerodynamic data on which they were basing their designs and decided to measure the lift and drag coefficients themselves using a wind tunnel. The brothers built two balances, one for lift, and the other for drag. and they would test models of their wing designs on both balances to determine the effect of design variables on lift and drag.
The brothers built models of their wing designs using materials available in their bike shop. Strips of 20 guage steel (1/32 inch thick) were cut, hammered, filed and soldered to produce various shapes. They made between one and two hundred models and made quick preliminary tests in October, 1901, to develop their test techniques and to investigate a wide range of design variables. Following the preliminary experiments, they chose about 30 of their best designs for more detailed testing. The models were numbered by the brothers and each was designed to be part of a parametric study of lift and drag by changing the value of only one design variable between models. On this web page, we show pictures of the actual models used in the tests and we have grouped them to highlight the design variable being tested. You can re-create the tests of the Wright brothers by using our interactive tunnel simulation. Data exists for all the models shown on this page, but there were many other models tested in the tunnel for which the data no longer exists. The major results from the wind tunnel tests are given on another page. The photos on this page are provided by The Franklin Institute Online where the actual model are currently kept.
Models 1, 2, and 3 are flat plates, having the same area (6 sq. in.), but different aspect ratios (AR), the ratio of the span to the chord of the wing. Model 1 has an AR = 1.0, Model 2 has an AR = 4.0, Model 3 has an AR = 6.0.
Models 4, 5, and 6 are circular arc airfoils, having the same area (6 sq. in.), and the same square planform (AR = 1.0). In this study, the camber (c), or curvature of wing is varied; the higher the camber, the greater the curvature. Model 4 has c = 1/12, Model 5 has c = 1/16, Model 6 has c = 1/20. You could include Model 1 in this study as well, with c = 0.0.
Models 7, 8, and 9 are circular arc airfoils, having the same area (6 sq. in.), and the same rectangular planform (AR = 6.0). In this study, the camber (c), or curvature of wing is varied; the higher the camber, the greater the curvature. Model 7 has c = 1/12, Model 8 has c = 1/16, Model 9 has c = 1/20. You could include Model 3 in this study as well, with c = 0.0. Comparing with Models 4, 5, and 6 would also show the effects of aspect ratio.
Models 10, 11, and 12 are parabolic airfoils, having the same area (6 sq. in.), and the same rectangular planform (AR = 6.0). In this study, the camber (c), or curvature of wing is varied; the higher the camber, the greater the curvature. Model 10 has c = 1/12, Model 11 has c = 1/16, Model 12 has c = 1/20. You could include Model 3 in this study as well, with c = 0.0. Comparing with Models 7, 8, and 9 would also show the effects of cross sectional shape (arc to parabolic). Model 12 is shown with a view from the bottom so that you can see the parabolic shape. The point of highest camber is near the leading edge (top).
Models 15, 16, and 17 are parabolic airfoils, having the same area (6 sq. in.), and the same square planform (AR = 1.0). In this study, the camber (c) is varied; the higher the camber, the greater the curvature. Model 15 has c = 1/12, Model 16 has c = 1/16, Model 17 has c = 1/20. Unfortunately, we do not a photograph of Model 17. Comparing with Models 10, 11, and 12 also shows the effects of effects of aspect ratio. Comparing with Models 4, 5, and 6 shows the effect of cross sectional shape.
Models 18 and 19 have the same area (6 sq. in.) and the same planform (AR = 4.0). Model 18 is a circular arc, while Model 19 has a parabolic cross section. Models 6, 18, and 9 constitute an aspect ratio study for circular arc foils at low camber (c = 1/20). In this study, Model 6 has an AR = 1.0, Model 18 has an AR = 4.0, and Model 9 has an AR = 6.0. Models 17, 19, and 12 constitute an aspect ratio study for parabolic foils at low camber (c = 1/20). In this study, Model 17 has an AR = 1.0, Model 19 has an AR = 4.0, and Model 12 has an AR = 6.0.
Models 18 and 19 were also used by the brothers in combination to study bi-wing configurations. The brothers called these tests Models 40, 41, 42, and 43; the difference being the gap between the wings. The gap was .55 chord for Model 40, .914 chord for Model 41, .4 chord for Model 42, and .2 chord for Model 43. The total wing area for these tests was 12 sq. in. which limited the range of angle of attack which could be tested.
Models 20 and 21 have the same area (6 sq. in.) and the same planform; a straight leading edge and curved trailing edge with an AR = 4.7. The cross section of Model 20 is termed a "bird wing" with a thicker section near the leading edge created by adding solder. Model 21 has a thin parabolic cross section. These models were built to determine the effects of the wing tip shape on drag and lift.
Model 25 has an area of 4 sq. in. and an AR = 8.0. The planform has curved trailing edges near the wing tips, with a straight leading edge. The cross section is nearly parabolic, with a slight thickening of the leading edge. A comparison with Model 12 shows the effects of the wing tip geometry and increased aspect ratio. Another identical model, Model 32 which is not shown, was used in combination with Model 25 to study a bi-wing configuration. The brothers called the bi-wing configuration Model 33. Model 33 had a total wing area of 8 sq. in. and the gap between the wings was nearly 1 chord length (11/16 inch).
Model 23 has an area = 3 sq. in., a rectangular planform with an AR = 6.75, a parabolic cross section, and 1/20 camber. There are two other models identical to Model 23, which are Models 22 and 26 (not shown). The brothers combined Models 22 and 23 to form Model 24 as part of the bi-wing tests. The spacing between the wings was nearly 1 chord length. You can compare results with Model 33 to see the effects of wing tip geometry. The brothers also combined Models 22, 23, and 26 to form a tri-wing configuration. They called this group Model 27. Comparing Models 23, 24, and 27, you can study the effects of 1 wing, 2 wing and 3 wing configurations for the same wing shape.
Model 35 has an area = 6 sq. in. and a rectangular planform with an AR = 6.0. The cross section was called a "bird wing" section and involved multiple curves and a thickening near the leading edge. Model 35 should be compared to Model 12 and Model 9 for a study of the effect of cross-sectional shape. While the actual wing shape for the 1902 aircraft was never modeled or tested, Wilbur wrote to Octave Chanute that the design was based on the average performance of Models 12, 9, and 35.
Models 30, 31, and 51 are special models which were not part of any parametric studies.
Model 30 has an area = 5 sq. in. and a rectangular planform with an AR = 4.1. The cross section shown in the figure is a "hook". This shape approximated a "Plin's curve" which had been tested in Europe and for which there was published performance data. The brothers probably built this model to check their work against published data. (The actual European geometry is called a "Pline's curve", but in writing to Mr. Chanute, a spelling error crept in).
Model 31 has an area = 8 sq. in., an elliptical planform with an AR = 4.64, and a parabolic cross section. This model approximated the configuration used by Lilienthal to generate the lift and drag coefficients which the brothers had used in the design of the 1900 and 1901 gliders. The brothers were able to verify that Lilienthal's data was accurate within the limitations of his measuring device, but that the data could not be applied accurately to a different configuration.
Model 51 is sometimes called Model 13. It has an area = 6.25 sq. in., a rectangular planform with an AR = 4.0, and a parabolic cross section. This model approximated the configuration used by Langley in the design of his Aerodrome. There was published data on this configuration so it was another check for the brothers on the accuracy of their test techniques. The brothers obtained the data to build this model from Mr. Chanute.
The wind tunnel tests were conducted from September to December of 1901. At the conclusion of the tests, the brothers had the most detailed data in the world for the design of aircraft wings. They used this data to design the 1902 aircraft which overcame the problems encountered in 1900 and 1901. They also used the data in the design of their propellers for the 1903 aircraft.
Navigation..
Last Updated Thu, Oct 14 09:00:38 AM EDT 2004
by Tom
Benson