Search results

The third term has been expressed as but in wind tunnel work it is often more convenient to measure were the omission of the dash signifies that the moment is now measured about a wind axis. The two quantities are very closely related and the measurement of one tells us almost as much as if the two were known. The latter, however, tells us either directly or indirectly what effect the addition of fin and rudder will have on the autorotation properties of the wings alone. The damping of fin and rudder being due essentially to the air flow meeting them at an angle on account of the rotation it should theoretically be possible to deduce this dynamic quantity from a simple static test of moment due to yaw angle. An experiment to test this was carried out several years ago but the static test did not give any approximation to the truth. This was ascribed at the time to the shielding of fin and rudder by the tail plane in the rotative experiment and subsequent work has amply confirmed this view. It is now known that shielding by the tail plane is by far the most important factor in determining the efficiency of the vertical surfaces at high angles of attack.

DURING the last year or two the construction of several new wind tunnels in this Country has been commenced, after many years of inactivity in this direction. The new tunnels are intended either to bring existing equipment up‐to‐date or to meet specific needs for researches which cannot be satisfactorily carried out in the older tunnels. In all cases the new tunnels are of types very different from those previously in use in this country, and it is interesting to trace the reasons for the change. In order to do this it would be well to review the history of the development of the existing tunnel equipment, in order to understand in the first place why the standard type of wind tunnel used in this country was entirely different from, and in some respects less efficient than, that developed on the Continent. When the study of aerodynamic problems was undertaken at the National Physical Laboratory in 1909, the question of a suitable design of wind tunnel was naturally one of the first to be raised.

THE increasing speed of modern aircraft has brought to the forefront the necessity for making a careful drag analysis of all aircraft in order to separate out the essential drag, that is to say the drag that is unavoidable, from the non‐essential drag. Most designers, we believe, now do this in order to see what progress is being made in the streamlining of their products. By this means we are enabled to see the relative importance of the drag terms and to arrive at a figure of merit. The ideally‐streamline aeroplane, though not at present a precise proposition, is like other ideals unattainable. It is the standard to which designers may aspire, but which they cannot achieve.

THE demand for aerodynamic research continues to grow, and there is difficulty in meeting it with the available equipment. The Compressed Air Tunnel will become available shortly, but will not very greatly alter the situation, as it will have a scries of scale‐effect problems of its own to solve, problems that have of necessity been shelved in the past, owing to the absence of any effective weapon of attack. The recent study of the design of open‐jet tunnels in this country has led to the realisation of the fact that a room which can contain a 7‐ft. tunnel of the N.P.L. type can quite easily house two open‐jet return‐flow tunnels of the same size or even slightly larger. A proposal has been made to increase very considerably the available tunnel equipment by replacing the 7‐ft. No. 1 tunnel at the Laboratory by two high‐speed open‐jet tunnels. In view of the feeling that some such step will be necessary in the near future unless research is to suffer delay, model tests have been made to provide design data so that the project could be rapidly carried into effect at short notice. There has been practically no expansion of wind‐tunnel equipment in this country since the War until the construction of the Compressed‐Air Tunnel was sanctioned and later the construction of a 24‐ft tunnel at Farnborough. While special tunnels such as these are absolutely necessary for the effective solution of certain problems, there is always a vast amount of research to be carried out in normal‐sized atmospheric tunnels of reasonable speed, and an extension of equipment on the lines suggested would undoubtedly be well worth the cost.

I CAN hardly say how much I appreciate the honour your Council has conferred on me in asking me to deliver the James Forrest lecture this year. I realise fully the difficulty of the task that has been set me, for it is no easy matter to follow in the footsteps of the illustrious men who have previously addressed you on aeronautical subjects. I will, however, do the best I can to give you some idea of the development of aircraft since Professor R. V. Southwell delivered his excellent review in 1930†, and in doing so, I will not fail to bear in mind the “leit motif” of this series of lectures, which is to trace, wherever possible, the interdependence of abstract science and engineering. My lecture is, in fact, arranged with that object mainly in view, for I shall endeavour to point out the advances that have been made in the technique of research methods, and the nature of the new knowledge of aerodynamic phenomena which has resulted from them, and to show, as far as I can, how this new knowledge has reacted on the practical design of aircraft.

DURING the past year the increase in performance shown by certain American and Continental aircraft has merited the attention of all aircraft technicians. The performance is, of course, partly due to increase in horse‐power, but this is a small factor compared with the advance shown in the reduction of drag. Now that it has been shown that it is possible to build aeroplanes of low form‐drag and with little interference, the drag due to skin friction is becoming of great importance, and it will probably be of advantage to outline the methods of estimating skin friction, and discuss what information is now available on the effect of surface finish and protuberances.

MUCH has been said earlier concerning the scale effect on drag. A few words must be added on the scale effect on lift near the stall, for it is here that ordinary small‐scale wind tunnel tests may be very misleading. The curves of Fig. 9 show the maximum lift coefficient of two wing sections plotted against Reynolds number, and the difference of behaviour is at once apparent.

THE great advance in aerodynamic efficiency of aircraft achieved during the last few years has largely been due to the continued efforts that have been made to reduce drag to a minimum. It is of interest, at the present stage of development, to review broadly the present state of knowledge on these matters, and to see where vital information is still lacking. It is not so long ago that we were greatly puzzled by the large differences in drag that were sometimes observed in different wind tunnels and in comparisons between results from tunnels of the compressed air type and those obtained in flight. It can now be said that the reason for these differences is well understood, but that it is not yet possible to account for them, or to predict them at all accurately, from the theory of boundary layer flow. Considering only wings with smooth surfaces the drag depends on the shape of the section, which defines the pressure distribution over its surface, and on the point at which transition from laminar to turbulent flow occurs in the boundary layer. If the pressure distribution and the transition point are known it is now possible to calculate quite closely what the profile drag will be. But theory has not yet provided a means of determining where this transition point will be on surfaces of different shapes and in air streams whose turbulence is characteristic of the wind tunnel on the one hand, and the free atmosphere on the other. The experimental determination of both profile drag and transition point in flight has done much to clarify ideas on the whole question of wing drag. It is now definitely known that the transition point may occur much further from the leading edge in the non‐turbulent atmosphere than it does in a wind tunnel at the same Reynolds number, so that the drag is lower in the former case. It is also known that this far‐back transition can occur up to Reynolds numbers as high as 17 × 106 in flight, and that its occurrence is related to the nature of the presssure gradient over the surface, that is, to the shape of the wing contour. Two important questions at once arise. Will this far‐back transition still occur on suitable sections no matter how far the Reynolds number is increased, and is it possible to encourage a still farther back transition by any alteration of section that is practically possible? The answer to these questions may well have a profound effect on design. To answer the first we must have an aeroplane capable of attaining the desired high Reynolds number and having a sufficient spanwise length of wing, undisturbed by airscrew slipstreams or discontinuities such as ailerons, to enable the measurements to be made. Such a machine docs not exist at the moment. There is some indication that the second question may admit of a favourable answer in the range of Reynolds number now available. The transition point would almost certainly be well forward on a flat plate or very thin wing section, unless, possibly, the turbulence in the air was incredibly small. In a stream such as that of the Compressed Air Tunnel it occurs practically at the leading edge at Reynolds numbers above 5 × 106. On the other hand, tests on a 25 per cent thick wing indicate a mean transition point for the two surfaces at about 20 per cent of the chord in this tunnel and, on full scale, transition as far back as 40 per cent of the chord has been observed. It seems likely that the farthest the transition could possibly move back is to the point at which a laminar boundary layer would separate from the surface. This point is calculable to a fair degree of accuracy, though the calculation has not yet been made for an aerofoil. It has been made, and the laminar separation observed experimentally, for an elliptic cylinder with axes in the ratio 3 to 1, and here it occurs at 63 per cent of the chord. The inference is that there is some hope that in non‐turbulent air it may be possible to find wing sections for which transition is delayed beyond the points hitherto observed, though how near it may prove possible to get towards the laminar separation point it is not possible even to guess.

THE method given in “The Stressing of Rigid‐Jointed Frames” published in the Journal of the Royal Aeronautical Society for June, 1936, may be applied to the case of frames embodying initially curved members, as for example, monocoque rings.

IN view of the exceptional dimensions of the structure shown in Figs. 1 and 4, which will be described in this paper, the author thought that it might be of interest to the members of the Institution to have a short account of the difficulties which we experienced in meeting the requirements of the National Physical Laboratory authorities, and the various methods which were adopted to overcome them.