As technical progress has improved the airborne performance of aircraft, so the difficulties associated with their take-off and landing have increased.

The grass airfields which were still commonplace at the end of the Second World War have now developed into complex arrangements of concrete runways of up to three kilometres (two miles) long, taxiways, dispersal areas, and so on. The main disadvantages of these large airfields, both for military and for civil aviation, soon become obvious: —

1. (i) The enormous expenditure in either case; for military aviation this expenditure has to be approved, but it is at the expense of the production of actual aircraft—like Ugolin who ate his children so that they would not be fatherless;

   for civil aviation, the high cost of airfield construction has hampered the development of aviation in the remote areas where it would be of particular use in raising the standard of living.

2. (ii) The lack of flexibility of air forces operating from fixed bases, thus reducing their efficiency.

3. (iii) The increase in vulnerability as destructive weapons become more highly perfected.

The present preoccupation of aircraft designers and users with fatigue problems, brought to a head by the Comet disasters, has rendered the notion of the finite life of an aeroplane a commonplace to the man in the street, and has made the assessment of probabilities of failure and the collection of statistics regarding loads and structural properties a matter of detailed concern for all designers. This was not always so. In 1939, when Fairthorne and I essayed a first paper on airworthiness statistics it received practically no attention and publication was delayed for years until after the 1939–45 War. Even in 1942, an exploratory “ Philosophy of Aeroplane Strength ” brought smiles to the faces of practical designers rather than the real interest that the “new slant,” as some said, of its probability approach raised in a few scientists.

Knowledge of the motions of aircraft control surfaces due to oscillating hinge moments is of great importance for flutter investigations. While the impedance of the power control unit may play a large part in determining these motions, it is not their only cause. The elasticity of the power unit mounting and the effect on the input of distortion of the aircraft structure, an effect which may depend on air speed, may also be important. Nevertheless the influence of the power unit itself is of major importance. No very satisfactory methods appear to have been devised yet, however, for making such tests. The difficulty lies chiefly in applying the large oscillatory loads to the output of the power control unit, which are necessary if the test is to be realistic. The facts that the oscillating loads must be applied to the jack while it is moving, and that a non-oscillatory opposing load is also desirable, add to the difficulties. For convenience in assessing the characteristics of the unit, it is desirable that the oscillating load should be superimposed on a constant opposing load and that the basic jack motion should be of constant velocity.

In a recent paper, Yates has shown that it is convenient to distinguish between two main types of stimulus which induce torsional vibration in geared-shaft systems. The first, which is dealt with in detail in the literature, is “force excitation” by which periodic torques act upon the system and throw it into oscillation; for the purpose of this note, it will be convenient to refer to a previous article as dealing with this problem.

The second type of excitation to which Yates refers is “displacement excitation.” It may be thought of as arising from errors in the cutting of gears which cause periodic variations in the velocity ratio of mating components. A Hooke's joint which couples two rotating systems which are capable of torsional oscillation would give rise to this form of stimulus.

The ability to manoeuvre at high altitude has become one of the main requirements in the design of modern military aircraft. Ability to manoeuvre enables the fighter pilot to execute tight turns during the attack on a bomber and the bomber pilot to perform evasive manoeuvres and so make the task of the fighter more difficult.

Normally the main manoeuvre considered is the application of increased normal acceleration, or g, and if we assume that (a) there is sufficient pitching power available from the longitudinal control surface to increase the wing incidence to give the necessary increase in normal force, and (b) that the new wing incidence is below the stalling incidence of the wing, then the degree of manoeuvrability available is a function of thrust and drag alone. In the past the manoeuvrability required at a given altitude has been expressed in terms of the rate of climb available at that altitude and thus we have such definitions as “ the 1,000 ft./min. ceiling ” and so on.

Reinforced rectangular plates find frequent application in the field of aeroplane structures, ship structures and allied fields. The stability of a structural element, if its critical load in compression is rather low, is normally influenced by its own weight. Usually the ratio of the body forces of the stiffened plate to the critical load is a considerable amount and the influence of the body forces of the plate and the ribs on the critical load of the stiffened plate is discussed here.

In this note (following Timoshenko), the extensional deformation of the middle surface during buckling is neglected, which greatly simplifies the analysis. Although it leads to an under-estimation of the critical load (of the order of 10 per cent at most) it is not likely to influence the general conclusions. Strain energy method of analysis is used.

It is gratifying to learn from Mr. Nicholls (May 1955, p. 364) that he is favourably impressed with the exponential function, and it is to be hoped that he, and others interested in stress-strain relationships, may find it to be of practical value.

While the modifications which he suggests are quite legitimate, they tend to mask the essential simplicity of the function. It would be difficult to obtain the asymptotic stress from his first equation. The second can be easily solved for strain, but the converse process is complicated by the fact that the current stress appears both inside and outside the exponential. It is surely much more straightforward, and infinitely quicker, to plot the stress-strain curve from the proof stress data and analyse it in the normal way.