John Richard Morgan was born at Barry on the 13th July 1904, and received his initial education at Barry Boys’ County School. In 1920 he began an indentured apprenticeship with the Barry Graving Dock and Engineering Co Ltd, and in 1923 was awarded a Glamorgan County Scholarship to the University College of South Wales and Monmouthshire where he read for an Honours Degree in Mechanical Engineering under Professor A. J. Sutton Pippard. He was permitted by his firm to attend the University from September 1923 provided he spent all the vacation time at the works, so that he can justly claim to have been an even earlier example of a “sandwich student” than a previous recent President.

The series of Chadwick Memorial Lectures was started in 1956 by Harold Rogerson who gave a very complete picture of Roy Chadwick's life and experience as an aircraft designer. Since then lectures in this series have mainly concerned problems on future projects, which would certainly have excited his lively interest, rather than nostalgic reviews of the past.

However, when I was honoured by being asked to give the 14th Lecture of the series it occurred to me that nothing would be more appropriate than a paper on the History of the Vulcan, since it was not only a major Avro enterprise in its own right, but it was also the last project with which Roy Chadwick was intimately concerned. It had a peculiar fascination for him, because he had, all the time I knew him, an enthusiasm for the all-wing transport as the ultimate in aerodynamic and structural efficiency

It was quite a few years ago that BLEU showed the feasibility of automatic landing. In the intervening period, considerable development has taken place. Before we look at the recent developments, it is worthwhile reiterating the early findings of BLEU. They showed that an aircraft could be landed automatically using an extension of a standard autopilot plus radio altimeter and an autothrottle system. In the pitch plane the aircraft is controlled on the glide slope as usual, but continues to a low height (100ft-150 ft) (30-36 m); a short period of constant attitude flight follows to about 50 ft (15 m) when the autopilot switches to flare and the throttles are closed at a steady rate. The control for the flare comes from the radio altimeters and an exponential flight path is aimed for. The rate of descent is kept to one-fifth of the height. In the lateral plane the control is purely an extension of the usual localiser control. Originally it was intended to use leader cable, but this had added complications of installation, and as techniques developed the localiser was found to give adequate performance.

When Mr. McManus and I first agreed to give this lecture, we did not realise that the audience would be likely to contain so many pilots. It could be construed that our job is to automate pilots out of business. However, I wish to assure you all that this is far from the truth. Our job is to completely automate the whole flight envelope so that pilots become “Systems Managers”. The result will be more accurate, reliable and safe flying, with the pilot assuming manual control on a minimum number of occasions consistent with keeping in practice.

This paper is a revised version of a contribution towards the work of the Society's Future Policy Committee's Working Party on Effectiveness in R & D. An abbreviated version of the Working Party's report was published in the September 1969 Journal.

The R & D process in any sphere exhibits features of uncertainty in many of its phases and it therefore seems appropriate to examine aeronautical R & D from a probabilistic viewpoint. Before discussing the origins and treatment of these uncertainties it is necessary to define Research and Development in the aeronautical context and to state its aims as seen by the author.

I am delighted as a test pilot, to have been asked to talk to the Test Pilots’ Group on the subject of light aviation. Originally I was asked to talk about the design of homebuilt aircraft; but, since, leaving the Service and joining the Air Registration Board, to look after all aeroplanes weighing less than 6000 lb, it seems appropriate to take a look at a wider field at this time, without focusing on any specific class of light aircraft.

During the past decade, the theory of optimal control associated with linear multivariable systems having state and output equations of the respective forms

and

has been applied extensively to the design of controllers for a variety of aerospace systems. Such applications of this theory have been mainly concerned with the design of controllers that generate optimal control vectors, ẑ, which minimise performance indices of the form.

where Q and R are positive-definite matrices.

Spalding and Chi have shown that their method of predicting the skin friction and momentum thickness for the compressible turbulent boundary layer with and without heat transfer gives good agreement with experimental data. For practical purposes a knowledge of the shape parameter, displacement thickness and boundary-layer thickness are also often required. In this note it is shown that the predictions of these additional boundary-layer parameters by extending the work of Spalding and Chi do not compare so favourably with experiment. New correlations have therefore been developed and are reported in this note which provides the basis of Item No. 68020 in the Royal Aeronautical Society Aerodynamics Sub- Series of Data Items.

Many of the flow patterns associated with experiments in fluid mechanics include vortices, or vortex sheets, and frequently knowledge as to the position and strength of these vortices is desirable. Wool tufts and spinners are useful in determining vortex position, but can give no indication of the vortex strength or circulation. Yawmeters cannot readily give vortex position, although (with a fair amount of labour) the circulation can be determined.

A convenient aid to the investigation of the aeroelastic properties of aircraft is provided by wind tunnel tests on scale representative aeroelastic models. A wide variety of tests is commonly made ranging from tests on models of component parts of an aircraft (wing, tailplane, fin, etc.) to tests on models of the whole aircraft. In many of these tests it is adequate to mount the model structure rigidly to the tunnel wall or sting, but in other cases the model must be provided with rigid body freedoms for representative results to be obtained. This is generally necessary for aeroelastic models of complete aircraft, and particularly for models of slender configuration.

In all of the experimental and theoretical work done on wall jets(1-3), the flow always passed over a stationary surface. An exception to this is the theoretical analysis reported in ref. 4 in which a laminar axi-symmetric jet impinged on a rotating disc.

The present investigation aims to provide information on the flow of a turbulent jet over a rotating cylinder. The curvature of the wall, the jet slot width and the initial speed of the jet are kept constant. The speed of rotation is the important variable.