March, 1939 AIRCRAF T ENGINEERING A Review of Existin g Knowledge with Some Important Matter s Still Requiring Attention B y E. F . Relf, F.R.S., A.R.C.S., F.R.Ae.S., F.I.Ae.S. H E great advance in aerodynamic effi two surfaces at about 20 per cent of th e chord actually be attained in some way in th e future. ciency of aircraft achieved during th e in this tunnel and, on full scale, transition as Such a transition will certainly be in a sensitive last few years has largely been due to far back as 40 pe r cen t of th e chord has been condition, and no great disturbance will be th e continued efforts that have been made t o observed. It seems likely that the farthest necessary to cause the transition point to reduce drag to a minimum. It is of interest, a t th e transition could possibly move back is t o move forward. Wings or bodies on which it th e present stage of development, to review th e point at which a laminar boundary layer is to be attaine d will therefore have to be very broadly the present state of knowledge on these would separate from the surface. This point is well made as regards absence of waviness matters, and t o see where vital information is calculable to a fair degree of accuracy, though or roughness. Moreover, a very far-back still lacking. It is no t so long ago tha t we were th e calculation has no t ye t been made for an transition is certainl y not attainable behind a n greatly puzzled by th e large differences in drag aerofoil. It has been made, and th e laminar airscrew, so that for large machines with tha t were sometimes observed in different wind separation observed experimentally, for an multiple power units it would seem that tunnels and in comparisons between results elliptic cylinder with axes in the ratio 3 to 1, advantage of th e effect could only be take n b y from tunnels of th e compressed air typ e and and here it occurs at 63 per cent of th e chord. using pusher airscrews, a matter that raises those obtained in flight. It can now be said that The inference is tha t there is some hope that in many other difficulties. No doubt, if a sub the reason for these differences is well under non-turbulent air it may be possible to find stantial drag reduction could be proved to bo stood, but that it is no t yet possible to account wing sections for which transition is delayed possible, some means would be found to over for them, or t o predict them at all accurately, beyond the points hitherto observed, though come difficulties of application such as these. from the theory of boundary layer flow. how near it may prove possible to get towards The whole argument really amounts to saying Considering only wings with smooth surfaces the th e laminar separation point it is not possible tha t the ideal low drag aeroplane, first envisaged dra g depends on th e shape of th e section, which even to guess. b y Professor Melvill Jones, in which the dra g defines the pressure distribution over its is merely the turbulent skin friction of th e surface, and o n th e point at which transition surfaces, may not represent the final irreducible from laminar to turbulent flow occurs in the minimum, since it may prove possible t o retain Transformed Wing Sections boundary layer. If the pressure distribution much more laminar boundary layer flow than The kind of section which is likely to lead and the transition point are known it is now was previously thought practicable. t o far-back transition is one in which the possible to calculate quite closely what the Apart from these fundamental problems of pressure minimum occurs well back, such as profile drag will be. But theor y has no t yet skin friction drag, there are others of much th e symmetrical aerofoils with maximum thick provided a means of determining where this practical importance of which the most out ness at 40 per cent of the chord recently transition point will be on surfaces of different standing is the drag connected with the in discussed in these columns by Dr. Piercy.* shapes and in air stream s whose turbulence is stallation and cooling of the engine. Here Such sections happen also t o be favourable from characteristic of the wind tunnel on th e one again, a great deal of knowledge has been the point of view of delaying the inception of hand, and the free atmosphere on the other. The gained recently, especially from tests in th e shock waves and consequent "compressibility" experimental determination of bot h profile drag 24 ft. wind tunnel at the R.A.E. on engine drag to th e highest possible speed. The idea of and transition point in flight has done much to installations. The avoidance of leaks in th e using symmetrical sections at all might have clarify ideas on the whole question of wing drag. surface, through which jets of air can disturb been severely criticised some years ago, on I t is now definitely known that the transition th e boundary layer, has been emphasised, an d account of their low maximum lift, but the point may occur much further from the leading th e importance of taking the cooling air in, advent of the flap and the accumulation of dat a edge in the non-turbulent atmosphere than it ejecting it , an d controlling its passage through a t high Reynolds numbers suggests that at any does in a wind tunnel at the same Reynolds the cooling passages rightly, has been thoroughly rate for th e major part of a wing th e symmetrical number, so tha t the drag is lower in th e former established. Yet there is still a long way to go section with a flap is a t no disadvantag e when case. It is also known that this far-back before the drag of actual engine installations compared with the usual cambered sections. transition can occur up to Reynolds numbers as approaches the value believed to be theoreti I t might be desirable however to avoid thin high as 17 × 106 in flight, and that its occurrence cally possible. The difficulties are particularly symmetrical sections near the wing tips from is related to th e nature of th e presssure gradient acute with the smaller high-powered aircraft the point of view of wing dropping. In view of over the surface, that is, t o the shape of the on account of th e relatively large size of engine these ideas, a very interesting line of research wing contour. Two important questions at nacelles in this case. On large machines th e is t o discover how far it is possible t o go in the once arise. Will this far-back transition still da y can hardly be far distant when means direction of wing thickness combined with far occur on suitable sections no matter how far the will be found to enclose the engines in th e back maximum thickness before the con Reynolds number is increased, and is it possible essential shape of the aircraft without adding ventional high lift devices show signs of failure. to encourage a still farther back transition by undesirable bulges to house them, but it is This problem is being attacked in the Com an y alteration of section that is practically difficult to see how such a means can be applied pressed Air Tunnel, and ma y lead to results of possible? The answer to these questions may t o the small twin-engined type of high per much importance in th e near future. If pro well have a profound effect on design. To answer formance. mising results are obtained, it would be neces the first we must have an aeroplane capable The two problems above mentioned are sary to measure the drags in flight, in order to of attaining the desired high Reynolds number interconnected, since the further we progress see tha t any increase in the area of surface under and having a sufficient spanwise length of wing, towards low frictional drag, by keeping the laminar flow in the turbulent tunnel was re undisturbed by airscrew slipstreams or dis transition point back, the greater will be the flected by a corresponding increase in th e free continuities such as ailerons, to enable the difficulties of adding necessary departures from atmosphere. So far only the transition measurements to be made. Such a machine docs th e ideal outline, such as nacelles and cabin phenomena on wing surfaces have been studied not exist a t th e moment. There is some indica in flight. The question of transition points on tops, without upsetting those transition condi tion that the second question may admit of a the forward part of a streamline fuselage is of tions and so introducing what is in effect an favourable answer in the range of Reynolds interference drag. The whole question is one nearly equal importance, at an y rat e for the number now available . The transition point of absorbing interest, to which future research smaller machines. would almost certainly be well forward on a may contribute some important and perhaps flat plate or very thin wing section, unless, There are one or two known facts that unexpected results. The most encouraging possibly, the turbulence in the air was incredibly render the application of these ideas difficult, feature of it is that modern boundary layer small. In a strea m such as tha t of th e Compressed supposing that the far-back transition can theory, though still far from complete, has now Air Tunnel it occurs practically a t th e leading provided the fundamental ideas which enable edge at Reynolds numbers above 5 × 106. On us to thin k much more clearly than heretofore, the other hand, tests on a 25 per cent thick *"Th e New Transformed Wing Sections," b y X . A . V. Pierey, and therefore to interpret correctly the facts wing indicate a mean transition point for the R. W. Piper and L. G. Whitehead. AIRCRAFT ENGINEERING, tha t are observed in our experiments. Vol. X , November, 1938, pp . 339-343.
Aircraft Engineering and Aerospace Technology – Emerald Publishing
Published: Mar 1, 1939