It is often supposed that the flow of gas from a compressor is, or should be, stable if the pressure-flow characteristic has a negative slope. It is shown here that this is only true if the Mach number is zero, i.e. if the machine is pumping a liquid. As the Mach number is increased towards one, a third regime becomes more and more important, wherein disturbances give rise to damped oscillations. The flow in this regime is stable in the mathematical, but not in the physical sense, since disturbances can occur all the time.

Two conclusions can be drawn. First, although compressors with “flat” pressure-flow characteristics have a wide range of stable flow at low speeds, they are poor at high Mach numbers, where the flat part cannot be used. Next, the actual points on the characteristics at which the flow becomes unstable are not fixed, but depend to a large extent on the steadiness of the entry flow.

The title of this paper is not strictly correct since it is not really concerned with “efficiency” in the accepted sense of the word. It has in the past been considered that the sole duty of an engine air intake is to deliver air to the engine at a high efficiency, that is at a mean total pressure as nearly as possible equal to the aircraft pitot pressure. More recently other considerations have arisen, such as the necessity of avoiding asymmetric flow in twin duct air intakes. These aspects of air intake design have frequently been discussed, for example in Dr. Seddon’s lecture to the Society in December 1951, but the particular subject of this paper has been rather neglected—the velocity distribution of the air delivered to the engine.

I have now reached the end of my task. The theme which I have sought throughout is that operating economics, in the broadest sense, are the essence of commercial Air Transport. The purpose of Air Transport is to improve communications—and to improve them economically. Although Air Transport the World over is losing money at present, through a combination of unfortunate circumstances, the facts of the present situation and of developments in train will, I am sure, confound the prophets of gloom. Air Transport, provided with adequate tools and run on the right methods, can be made to pay—furthermore, air travel can be provided economically at fares which the average man will be able to afford.

A summary is presented of information collected on coefficients of friction (rolling and sliding) between rubber tyres and road or runway surfaces. Nearly all the data collected are from tests on automobile tyres and are limited to speeds of about 40 m.p.h. In some cases the primary object of the tests was to distinguish between good and bad roads, and not to determine absolute values of friction coefficients.

The reasons for the importance of the information for aircraft design use are discussed and distinctions are made between coefficients measured in different ways.

The results show the variation of sliding friction coefficient with various parameters at speeds up to 40 m.p.h. There is a set of American tests at speeds up to 110 m.p.h., and some estimates made by the Dunlop Rubber Company up to 120 m.p.h.

It is concluded that for sliding friction coeficients differences in surface texture (as distinct from surface material) are significant, and that information can be given only in the form of limits within which the values can be expected to lie.

Few results are available on the variation of rolling resistance coefficient, but it is felt that the approximate values quoted for different types of surface are sufficiently accurate.

These notes are an attempt to examine what may be achieved in the way of performance, with existing materials and knowledge of aerodynamic engineering.

No claim is made for novelty, either in the data used or in the methods of working from such data to obtain the results arrived at.

Only the simplest engineering mathematics are employed so these results cannot claim academic accuracy, in fact they are most probably considerably inaccurate. But though inaccurate absolutely, the results should be accurate enough relatively for comparative purposes.

Full details of the methods employed are given in the various Appendices. This appears advisable, as it should make quite clear how the results are obtained and leaves it open for anyone to work out further examples by the same methods or to modify the methods themselves.

In the past the usual method of dealing with the articulated rod has been to treat the problem graphically. Although this method may be sufficiently accurate for a first approximation the inherent difficulties often lead to serious errors after the first few stages. A graphical method is always sufficiently accurate to determine the travel of a piston operated by an auxiliary rod ; but, the double differentiation of the displacement curve by graphical means (to obtain the acceleration curve), is not only inaccurate but a long and tedious operation.

To remedy this defect a Fourier's series has been developed which can be differentiated at sight and which gives a remarkable accuracy.

The geometrical properties of the mechanism have been analysed and thus equations are formed giving the resultant radial and tangential components at the crank pin, the bending forces and tensions on the master rod, etc.

We have published elsewhere a considerable amount of work on the development of the cathode ray engine indicator and, while most of the pressure elements and circuits which have been developed are perfectly satisfactory on normal petrol and Diesel engines, it has been impossible up to the present time to obtain a really satisfactory pressure-time card on the high duty boosted aircraft engine, the main difficulty being due to the super position of mechanical vibration on the pressure-time diagram. A satisfactory knock indicator has, however, been produced and a few of the results are here discussed, in the hope that they may be of general interest.

Recent contributions to the theory of structures with particular reference to the design of aeroplanes have dealt with the theory of frames in which the members are rigidly connected together at the joints; the object being to ascertain the degree of increase in strength of the individual members due to the restraint at their ends supplied by the contiguous members.

Briefly, when a member forms part of a rigid-jointed structure it acquires a rigidity arising from the rest of the structure. The amount of rigidity given by the rest of the structure may vary from zero end fixation to full encastreing.

Although the problem of the stability of a fuselage that is subjected totorsional force is important in connection with the case of light metalconstruction of an aeroplane in flight, the theoretical side of the problemdoes not seem to have received much attention. This is probably because ofthe difficulty of obtaining its mathematical solution even in the case wherethe fuselage is assumed to be a circular hollow cylinder.Thus, I studied the problem in assuming that the fuselage is a cylindricalshell for the first approximation. Southwell and Skan have dealt with thestability of a plane elastic strip due to edge shearing forces, but theirresult cannot be applied to the problem of the cylindrical shell unless itslength is very short. Schwerin seems to be the only one who has written onthe torsional stability of a cylindrical shell. Although his method ofconstructing the differential equations of the equilibrium of a cylindricalshell has been chiefly derived from Love's text book, and appears correct inthe main, yet owing to the certain apparent particularities on his part hissolutions of equations are open to grave doubts.

Test Flying should not be confused with Flight Testing. There is a very large difference between the two. Flight Testing covers the work of an infinite variety of people, while Test Flying is done by pilots and observers only. It is with the work of the pilots that this paper will deal.

A Test Pilot is presumably a pilot who tests aircraft, but it is a very loose, term. Out of the multitude of pilots engaged on test work in aircraft, only a small percentage actually test the aeroplane itself, while the remainder may be divided into two classes: those who do flight inspections, and those who use the aircraft merely as a vehicle for the testing of equipment and engines.