Since the advent of the aircraft gas turbine engine, various ways have been employed for feeding the fuel into the combustion chambers in the most advantageous manner. Perhaps the most commonly employed method has been the use of swirl atomising nozzles, and in this group is the spill type which has been used successfully for many years in industrial oil burning furnace equipment.

Arising from the general review of burner nozzles made for the aircraft gas turbine engine in its early days, spill burners were used experimentally on the W2/700 engine. Among the problems encountered was that of achieving suitable flow matching characteristics between the several burners of a set and also, that of arranging a satisfactory control system. The knowledge gained by this early work has been made available and has proved helpful to those who chose to carry on with this particular type of fuel nozzle.

In the design of high-speed aircraft the problem of relating structure loads to a prescribed flight envelope leads very often to the necessity of airframes being considerably stronger than past experience would tend to suggest, due to the methods used in calculating the loading.

Hitherto, when dealing with subsonic flight, it has usually been sufficient to determine structural loads in manoeuvre on the basis of a rigid aircraft. That is to say, the influence of structural distortion upon the aerodynamic derivatives, which enter the load calculations, is neglected. That this assumption becomes untenable as the speed is increased is clear when it is realised that, for a given fixed loading on the wing, the distortion is fixed; whereas the incidence necessary to produce the loading is approximately inversely proportional to the square of the flight speed.

A method is developed enabling rapid determination of the optimum cross-sectional dimensions of compression surfaces having unflanged integral stiffeners, and consideration is given to the effects of practical limitations on the design. The theoretical efficiency of the optimum integral design is found to be only 85 per cent, of that of optimum Z-stringer design.

In view of recent adverse publicity I must begin this paper by stating that anything I say must not be taken to imply any criticism of the Aircraft Industry. Any conclusions I draw, probably from insufficient evidence, must be regarded as purely intentional!

It has been difficult to decide how to prepare this paper. Should I confine myself to technological matters or should I range wide over the Industry's past, present and future problems, and run the risk of displaying my ignorance? I decided to range wide.

In a recently published book, The Aerodynamics of Propulsion, by Küchemann and Weber, there is included a brief discussion of heat flow in an air-air heat exchanger. The treatment given rests on an assumption which appears to be generally accepted, but which seems to the present writer to be physically unsatisfactory and to yield at least one misleading result. It is the purpose of the present note to show that a solution of the heat transfer equations can be obtained using an alternative assumption which seems physically more acceptable.

The only simple method of estimating the tip speed limitations imposed on a rotor by retreating blade stall on the one hand, and compressibility losses on the other, is that given by Hafner. It suffers from the lack of accuracy always inherent in a “standard radius” method, and for present-day use a more refined approach is obviously desirable, provided that it still enables rapid estimates to be made.

The Integral I occurs frequently in aerodynamics. To cite a particular case, Ward has shown that if S(x) is the cross-sectional area distribution of a slender body its wave drag D is given to a first approxition by D = qI/(2π), where q is the kinetic pressure. In this case, S(x) is often defined numerically, and the direct evaluation of I is then complicated by the presence of the logarithmic singularity. Several methods may be employed to avoid this complication. Legendre has recently suggested rewriting I in a non-singular form, which unfortunately is not well suited to numerical work. Here, a method for the evaluation of the unmodified integral is presented.