I count it a great honour to have been asked by the Royal Aeronautical Society to deliver the Seventeenth British Commonwealth Lecture. Reading down the list of previous performers in this rôle one feels that it is indeed an election to a very select company.

I have chosen as my subject for this lecture the post-war progress of air transport in Europe. The title may sound rather pedestrian, but I have two main reasons for choosing this theme. First there is the personal reason. I have been intimately associated with the development of air transport in Europe during the post-war period and I am very pleased to have an opportunity to set down a record of the developments which have taken place during those years. Moreover, my associations with European air transport go back to its very beginnings, as pilot of one of Europe’s—and the world’s—first scheduled air services in 1919. The second reason is broader. I welcome an opportunity to review the post-war progress of air transport in Europe because I think that this will show what a revolutionary change the industry has undergone within the relatively short space of fifteen years.

When I was invited to deliver the Lanchester Memorial lecture, I wondered why the great honour of being the first Frenchman selected for the celebration of the deeds of a great engineer was conferred on me. However, I am well aware that the Council of the Royal Aeronautical Society cannot make a mistake and there is good reason to choose, after so greatly famed professors, an engineer of the French Navy to illustrate, using a slightly different point of view, the influence of Lanchester on the development of aeronautics. This way, I do not have to apologise for a possible inadequacy which then would not be my own responsibility. I have only to do my best, so as not to betray the confidence of the Society.

Answering so plainly this first question, I had to find out how to fulfil my task. Possibly, my contribution to the intrepretation of the vortex sheets starting from the thin leading edges of a delta wing is more in the Lanchester line of thought than it is an extension of Prandtl’s inferences.

One of the main differences between manned aircraft and missiles is that in the former the “short period” dynamic characteristics should usually be at least moderately stable, while in the latter considerable instability can be tolerated.

However, this requires that the missile control system should have high gain “synthetic” feedback. Consequently, problems of stability associated with high gain are introduced, which do not occur with such severity in the manned aircraft. Typical missile stability problems of this kind, such as those arising from structural flexibility and propellant sloshing, are discussed.

Ballistic missiles, because they are required to operate in conditions where aerodynamic forces are sometimes negligible, need to utilise non-aerodynamic control forces. This introduces special problems, which have a parallel only in VTOL aircraft. The long duration attitude control of satellites in the presence of very small disturbances also presents a problem not encountered in the aircraft field.

The uncertainty of human pilot response is discussed in relation to the difficulties of adequate aircraft control system simulation. Some remarks are also made on the interaction between control about various axes. These interactions are often fundamentally different between aircraft and missiles.

The relative merits and possible applications of liquid and solid propellant rocket motors have long been topics for discussion and controversy. The arguments for and against have been sharpened by the fact that the development of the two types of motor are usually vested in quite different teams of people. At the same time, a fair comparison has often been confused because the ultimate choice has been made for reasons which are really not technical at all, but instead depend on availability and the utilisation of already deployed capital resources.

Nevertheless, the merits and de-merits of the “conventional” (as opposed to the packaged) liquid motor and the solid motor can be simply defined with reasonable accuracy. The liquid motor is complicated, demanding at least some, and often a great deal, of attention before use. At the same time, it has a much greater performance potential, can be used repeatedly and can be designed to give variable thrust. The solid motor is simpler (although is getting less so), is cheaper (although this is getting less true) and demands less preparation. However, it has definite performance limitations, is inflexible in its output, and is often quite sensitive to its environment, whether during storage, carriage, or operation.

The fundamental advantage of two-dimensional photoelasticity is its great simplicity. Neither the frozen-stress three-dimensional method nor the photoelastic layer technique share this simplicity as yet. Whereas the frozen-stress and layer techniques are most suitable for final investigation into the strength of a finished design, the simple plane photoelasticity can be used as a tool by the designer at every stage of his design, as a specialised slide rule, so to say, that gives sometimes faster and frequently more accurate answers than complicated stress calculations. The consequences of given changes in the outlines of a component can often be studied more conveniently and cheaply by this method than by almost any other.

When estimating the skin friction drag on bodies of revolution at the project design stage, it is normal practice to use an “equivalent flat-plate” method. This calculation is simple and requires little time to perform. However, in the final design stage, when more accurate estimates of the skin friction drag should be made the “equivalent flat-plate” method is still used. More exact methods, which include pressure gradient effects, are very laborious if they are evaluated by hand calculators. Even if a digital computor is used to evaluate the more exact methods, the time required to perform the calculations is considerably longer than that taken to calculate the other forms of drag. For these reasons, the simple “flat-plate” method has long held favour with designers.