386 



EVENING DISCOURSES. 



important, this factor indicates the ratio of the strees in any part of the 

 structure under given conditions of flight to the stress under normal conditions 

 of horizontal flight. The load factor during recovery from a dive can be 

 determined in two ways. Starting with an aeroplane diving at a given speed 

 we could calculate the initial stresses on the parts, and if we knew the 

 aerodynamic forces, the values of ^-i and the resistances of the various parts, 

 and were able to make some aseumption as to the rapidity at which the 

 pilot altered his controls, we could determine mathematically the rest of the 

 path and the stresses on the machine at any point of that path. Fig. 1 shows the 

 results of that calculation for a machine flattening out from a dive, and Fig. 2 

 the results from a machine doing a loop. In both cases the machines were 

 moving at high speeds, and high load factors were required for safety. In the 

 former case the pull on the control levers was 160 lb., reached after 0-18 second; 

 in the latter case it was 100 lb., attained in about the same period. 



The other method of detennination relies upon the experimental measure- 

 ment of the forces on the machine as a whole by the aid of accelerometers. On 

 aeroplanes pendulums of heavy weights or springs of high inertia, which would 

 go on swinging, would be useless. The accelerometer used took the form 

 illustrated in Fig. 3. It consisted of a quartz fibre, 1 /2,000th in. in diameter. 



Fl^J. ACC£l£/fOM£rE/f 



jimn 



£2 



(b) 



bent to an arc A B C, and held, normally, in a horizontal position. The fibre 

 would be slightly deflected downwards by its weight, and would be still more 

 deflected by a vertical acceleration, and this additional deflection of the point C 

 to positions C^ or C^ was recorded photographically. The accelerometer camera 

 and the lamp were contained in a small box strapped to the observer's knee. 

 The curves thus obtained indicates only the acceleratinn.s of the centre of 

 gravity, without taking account of the twists and turns about this centre. The 

 kinematograph has also been utilised at the Royal Aircraft Establishment for 

 the determination of the stresses. A kinematograph was fitted on to the tail 

 of one machine rising steadily, while the second machine following close behind 

 performed the loop spin, or other evolution, it was desired to analyse. Series 

 of these photographs were exhibited. 



A commercial machine should be stable, as it would be impossible, without 

 a great expenditure of the pilot's energy, to fly through a continuous cloud. 

 Yet stability pushed too far gave the machine a will of its own and made it 

 sluggish, and military pilots, to whom rapid control meant life or death, differed 

 as to the desirability of stability. On the other hand, the stable machine would 

 fly on, if the pilot lost control, and, if the engine .stopped, the machine glided 

 down, while the unstable machine got into a .spin and crashed down to earth. 

 Stability may be avitomatic or inherent. Automatic stability, secured by con- 

 trivances which came into play when a deviation from the steady course 

 occurred, is unsatisfactory because the device necessarily takes some time to 

 operate, overshoots the mark, and hunts. An inherently stable machine would 

 be brought back to the steady state by the disturbing wind forces. The lecturer 

 indicated how, thanks mainly to the labours of Bryan and Bairstow. longitudinal 

 stability, at any rate, had been successfully secured; our knowledge of the 

 conditions for lateral stability is inifortunately far less complete. Longitudinal 

 stability is important in connection with looping. It is known that an unstable 

 aeroplane has a stable flying position on its back. If it got into that position 



