
第46章
In a lecture before the Royal Society of Arts, reported in Engineering, F. W. Lanchester took the position that practical flight was not the abstract question which some apparently considered it to be, but a problem in locomotive engineering. The flying machine was a locomotive appliance, designed not merely to lift a weight, but to transport it elsewhere, a fact which should be sufficiently obvious. Nevertheless one of the leading scientific men of the day advocated a type in which this, the main function of the flying machine, was overlooked.
When the machine was considered as a method of transport, the vertical screw type, or helicopter, became at once ridiculous. It had, nevertheless, many advocates who had some vague and ill-defined notion of subsequent motion through the air after the weight was raised.
Helicopter Type Useless.
When efficiency of transport was demanded, the helicopter type was entirely out of court. Almost all of its advocates neglected the effect of the motion of the machine through the air on the efficiency of the vertical screws. They either assumed that the motion was so slow as not to matter, or that a patch of still air accompanied the machine in its flight. Only one form of this type had any possibility of success. In this there were two screws running on inclined axles--one on each side of the weight to be lifted. The action of such inclined screw was curious, and in a previous lecture he had pointed out that it was almost exactly the same as that of a bird's wing. In high-speed racing craft such inclined screws were of necessity often used, but it was at a sacrifice of their efficiency. In any case the efficiency of the inclined-screw helicopter could not compare with that of an aeroplane, and that type might be dismissed from consideration so soon as efficiency became the ruling factor of the design.
Must Compete With Locomotive.
To justify itself the aeroplane must compete, in some regard or other, with other locomotive appliances, performing one or more of the purposes of locomotion more efficiently than existing systems. It would be no use unless able to stem air currents, so that its velocity must he greater than that of the worst winds liable to be encountered.
To illustrate the limitations imposed on the motion of an aeroplane by wind velocity, Mr. Lanchester gave the diagrams shown in Figs. 1 to 4. The circle in each case was, he said, described with a radius equal to the speed of the aeroplane in still air, from a center placed "down-wind" from the aeroplane by an amount equal to the velocity of the wind.
Fig. 1 therefore represented the case in which the air was still, and in this case the aeroplane represented by _A_ had perfect liberty of movement in any direction In Fig. 2 the velocity of the wind was half that of the aeroplane, and the latter could still navigate in any direction, but its speed against the wind was only one-third of its speed with the wind.
In Fig. 3 the velocity of the wind was equal to that of the aeroplane, and then motion against the wind was impossible; but it could move to any point of the circle, but not to any point lying to the left of the tangent _A_ _B_. Finally, when the wind had a greater speed than the aeroplane, as in Fig. 4, the machine could move only in directions limited by the tangents _A_ _C_and _A_ _D_.
Matter of Fuel Consumption.
Taking the case in which the wind had a speed equal to half that of the aeroplane, Mr. Lanchester said that for a given journey out and home, down wind and back, the aeroplane would require 30 per cent more fuel than if the trip were made in still air; while if the journey was made at right angles to the direction of the wind the fuel needed would be 15 per cent more than in a calm. This 30 per cent extra was quite a heavy enough addition to the fuel; and to secure even this figure it was necessary that the aeroplane should have a speed of twice that of the maximum wind in which it was desired to operate the machine. Again, as stated in the last lecture, to insure the automatic stability of the machine it was necessary that the aeroplane speed should be largely in excess of that of the gusts of wind liable to be encountered.
Eccentricities of the Wind.
There was, Mr. Lanchester said, a loose connection between the average velocity of the wind and the maximum speed of the gusts. When the average speed of the wind was 40 miles per hour, that of the gusts might be equal or more. At one moment there might be a calm or the direction of the wind even reversed, followed, the next moment, by a violent gust. About the same minimum speed was desirable for security against gusts as was demanded by other considerations. Sixty miles an hour was the least figure desirable in an aeroplane, and this should be exceeded as much as possible. Actually, the Wright machine had a speed of 38 miles per hour, while Farman's Voisin machine flew at 45 miles per hour.
Both machines were extremely sensitive to high winds, and the speaker, in spite of newspaper reports to the contrary, had never seen either flown in more than a gentle breeze. The damping out of the oscillations of the flight path, discussed in the last lecture, increased with the fourth power of the natural velocity of flight, and rapid damping formed the easiest, and sometimes the only, defense against dangerous oscillations. Amachine just stable at 35 miles per hour would have reasonably rapid damping if its speed were increased to 60 miles per hour.
Thinks Use Is Limited.
It was, the lecturer proceeded, inconceivable that any very extended use should be made of the aeroplane unless the speed was much greater than that of the motor car.
It might in special cases be of service, apart from this increase of speed, as in the exploration of countries destitute of roads, but it would have no general utility.