been developed beyond carrying the most forward lamination to the tip of the blade and succeeding laminations to smaller radii, owing to the method of construction and the nature of the material used. It has recently been stated that this forward tilt renders the blade liable to twist under load.
The stresses in the blades have been calculated by crude methods which give an approximation to the stress along the grain. Fracture has, however, almost invariably occurred across the grain, in the ear- lier airscrews, by failure of the glued joints. Workshop practice has now so far improved that the strength of glued joints is equal to the strength of even hard woods across the grain. The evident need for knowledge of torsional stress in an airscrew blade led to the practical solution by G. I. Taylor and A. A. Griffith in 1916 of the problem of torsion of prisms of any section. The mathematical equations had already been stated and the new development was the provision of an experimental method of solution. Theory can now indicate the shape of blade required to avoid twisting under the loads imposed in flight. Apart from the reduction of stress, this is of great value to the designer, who cannot with any certainty predict the performance of an airscrew if the blades twist in an unknown manner in flight.
In order to protect the blades from moisture the airscrew is var- nished, or painted, and to protect against sand on land and spray on the sea, the tips have in some cases been sheathed in metal, but the practice of covering with fabric (dating from 1912-3) has re- . cently found more favour. Japanese lacquer has also been used as a protective covering.
Several early airscrews (e.g. Breguet's) were entirely of metal, commonly aluminium blades bolted to a steel tube, a method only possible with the low powers and speeds of rotation of the period. Bleriot crossed the Channel with a small, high-speed, laminated wood screw. Experiments with steel construction have proceeded slowly and steel may in time come into common use. Failure has been largely due to the unreliable nature of welding, and to brittleness produced in the process. For production in moderate quantities, wood requires far less outlay. A modern development is the air- screw with detachable blades, so far in a purely experimental stage. It allows of adjusting the pitch of the blades, if the airscrew has been imperfectly designed or the conditions of operation be altered, and of replacement of a damaged blade without renewing the whole. If the blades are of wood, shorter lengths of timber may be used, but it is doubtful if this can be regarded as an inherent ad- vantage of the system, because the difficulty of attaching wood blades to a centre are probably as great as the difficulty of making a satis- factory joint at the centre of an airscrew constructed entirely of wood. The airscrew whose pitch is variable in flight is a particular case of the detachable blade screw, and the chief difficulty in the design of such a screw for high speeds of rotation is that of making the joint between the blades and the centre.
In Britain and in America airscrews have been tested before use in flight by " spinning " by means of an electric motor. This test has been applied to new designs, to airscrews whose strength has been suspected by an inspector, and to samples taken from batches. The practice was in force in this country in 1914 and has been con- tinued. Flight conditions are not reproduced by the test, because the airscrew is not advancing through the air, and because the crank- effort variation and vibration of the engine are absent. The thrust loading is more severe, the centrifugal loading less severe. Experi- ence has, however, given considerable confidence in the test. In France the only test applied has been a loading of the blades to counterfeit the air forces, without rotation.
BIBLIOGRAPHY. British. -Reports and Memoranda of the Advisory Committee for Aeronautics (1909-19) and the Aeronau- tical Research Committee (H. M. Stationery Office); L. Bairstow, Applied Aerodynamics (1920); G. P. Thomson, Applied Aerodynam- ics (1920); A. I. S. Pippard and J. L. Pritchard, Aeroplane Struc- tures (1919); H. C. Watts, Design of Screw Propellers for Aircraft (1920); E. C. Vivian and W. Lockwood Marsh, A History of Aero- nautics; technical periodicals: Aeronautical Journal; Flight; Aero- nautics; The Aeroplane.
American. Reports of the National Advisory Committee for Aeronautics (Government Printing Office, Washington) ; technical periodicals: Aviation; The Aerial Age.
French. -G. Eiffel, Nouyelles recherches sur la Resistance de I' air et I' Aviation (1919) ; technical periodicals: L'Aerophile; L' Aviation.
German. Technical periodicals: Zeitschrift fur Flugtechnik und Motorluftschifffahrt. (R. McK. W.)
III. AERODYNAMICS
Experiments and Calculations on the Principles of Flight. The recent history of the development of aeronautics rests largely on experiments on aircraft or models of aircraft and their parts. That branch of investigation whicn is least related to any other subdivision of engineering is the study of the forces which are experienced by a body when moving through the air. The air forces due to motion are dealt with under the general head of " Aerodynamics." A knowledge of air resistances is a primitive necessity in connecting the subject with the much older and well-established subject of " Dynamics."
In dealing with dynamics, the forces acting are frequently given by a simple fundamental law such as the theory of gravita- tion when accounting for the motion of planets and comets, and very many of the more complex reactions have been worked out. The corresponding fundamental theory of fluid motion has been known for more than half a century, but application to the determination of air resistances has proved to involve mathe- matical problems beyond the capacity of the times. Recourse has therefore been made to direct experiment , and in the early stages of aeronautical development almost every new idea could be tested. The number of variables under review has now grown so greatly as to exclude such a method on the ground of cost, and a period of fundamental experiment is being entered on. The object of such experiments is to find out what is happening to the air disturbed by the passage of a body in such a way that the results can be applied, with a reasonable degree of approxima- tion, to a large number of related problems. Some success has been obtained in the case of airscrews, where the experimental data are so used that it is unnecessary to test every new design of airscrew. Extension to the aeroplane is gradually taking place.
For the same reason expense experiments on models have been used to cover the main field of inquiry, and the costly and frequently dangerous experiments on the full scale have, on the whole, been directed to crucial tests of the validity of the use of models. ' There has, of course, been a great amount of testing of aircraft in connexion with their value as fighting craft. At the present time, the value of such testing as an aid to design is very limited, detailed analysis being required to indicate lines of progress.
It then happens that the most comprehensive view of the subject of aeronautical principles is obtained from those aero- dynamical laboratories which deal with experiments on models, experiments carried out under almost ideal conditions in the artificial air current of a wind tunnel. The theory of the use of models 1 becomes of great importance in aeronautics and has been studied extensively. When the maximum possible use has been made of the theory the position remains one for experiment, and full-scale cooperation is found to be essential for estab- lishing a sound position. The theory of models has great value in showing the correct type of experiment and the method of comparison with the full scale. Finally, it is now known that when certain precautions are observed in model tests the applica- tions to full scale have an accuracy sufficient to give them a high value as an element in progress. 2
A. Air Intake B. Working Section C. Aerodynamic Balance
D. Position of Airscrew E. Distributor
FIG. 1 6. Wind Tunnel.
Laboratory Experiments. (a) The Wind Tunnel. The num- ber of first-class wind tunnels in existence in the world in July 1921 was probably between twenty and thirty. Of these, seven were at the National Physical Laboratory at Teddington, three
1 Report, Advisory Committee for Aeronautics, 1909-10, p. 38.
1 Report, Scale Effect Sub-Committee A. C. A., 1917-8, R and M, 374-
