The position is somewhat simplified, as already pointed out, by the fact that though the lift is decreased by the decrease in span, it is to a great extent compensated by the increase in speed. Also another compensation is offered by the fact that fore and aft stability requires a lifting tail.
Lift is largely in proportion to the length of the entering edge of the plane, but it does not always follow that this entering edge must be at right angles to the direction of flight. The Dunne machine obtains its lift with an entering edge that is entirely at an angle of some 45 degrees, and its shape is an exact replica of the arrow head of prehistoric man and the paper darts of our schooldays, a design, by the way, that was patented in 1860.
At first sight it would seem that the lift on a plane shaped thus (Fig. 18), would only be equal to the lift given by a plane with an edge as long as the distance between A and C, thus (Fig. 19), but this is not so. Although the lift is not so great as it would be if the edge was straight in one line (Fig. 20), it is very much greater than it would be on Fig. 19. The probability is that it is about half-way between (19) and (20), but probably nearer to (20) than (19). There are no exact data to go on, but the efficiency of the Dunne machine would seem to show this.
Fig. 18.
Fig. 19. and Fig. 20.
Again, in seeking for planes that offer the least resistance to the air, one of the best that suggests itself is the T-shape (Fig. 21), and this may be improved by cutting off useless corners (Fig. 22). A plane of this shape lends itself to great strength of construction owing to its small extending parts. It is compact, it gives an entering edge half as long again as its span, and gives a lift in proportion to that edge, and it is in itself stable. Having thus evolved a suitable plane for the front of the machine, the best thing to do is to base the back plane on the same design, and join the two planes together to form the supporting surface of the machine, allowing sufficient space between them to avoid any interference or overlapping. The design then stands thus (Fig. 23), when the back plane is a slightly smaller copy of the front one. The position of the centre of gravity in this design would be coincident with the centre of pressure longitudinally and laterally, and would be situated about at A. A paper model on these lines with a low centre of gravity may be easily constructed and will prove useful in illustrating the different points here stated. The paper should be cut out sufficiently wide to allow of a central longitudinal fold (Figs. 24 and 25), and a roll of paper should be made for ballast and pushed through the fold as shown in Fig. 26 at the point marked A.
Fig. 21. and Fig 22.