High-speed flight-and the future
Some of the problems facing the aeroplane designer as the speed of sound is approached have been mentioned as they have arisen in connection with their effect upon design. Now it is the behaviour of air in supersonic flight, and the effect of this upon the aeroplane, that is discussed—so far as current knowledge permits. Supersonic air flow has been studied for many years—it is necessary for the design of shells and bullets for instance—but it is only since the War that it has been of other than academic interest to the aerodynamicist. The first practical application was the V-2 rocket, which reached Maeh 5 at over 300,000 ft., where air resistance is negligible. Given the power, supersonic flight is feasible. High power is needed to overcome the sharp drag rise at the "barrier" as the shock wave is formed and this power must be sustained.
Probably the most striking difference between subsonic and supersonic aerodynamics is that in the latter the normal rules of streamlining no longer apply. Fineness ratio becomes the criterion for drag: frontal area is once more a vital factor, but its effect cannot now be overcome by coaxing the air round gentle curves; only by making the wing and body long so that the thickness does not exceed 5 per cent of the length can the drag be kept within bounds. Curves are actually at a premium in supersonic flight and the best results are obtained with straight, shallow tapers. In fact, the ideal supersonic wing would be a flat plate of negligible thickness, since camber is no longer required for the generation of lift. Such a wing is, of course, impossible to make, since it would have no strength. One suggested compromise is the double-wedge aerofoil of only 5 per cent thickness and low aspect ratio. With this wing would go a tubular fuselage of the same relative thickness, tapered off at each end like a sharpened pencil. In practice, the "points" would have to be modified for the jet, or rocket, orifice and the air intake for the former. For trim and control, a combination of the butterfly tail and the fins of a rocket, a 45 deg. cruciform would probably be better than the conventional fin and tail plane—by selective movement these fins might even be adapted to give rolling control.
In other words, the ideal supersonic aeroplane would look something like a guided missile, with its slender fuselage, stub wings of low aspect ratio and tail fins. Such an aeroplane could be pushed around the sky at Mach numbers of 2 and 3 quite efficiently, but it would be a "flat out or nothing" aircraft. The tiny wedge wing would give very little lift in subsonic flight, but the high power needed for supersonic speeds would almost certainly be enough to shoot the aeroplane into the air.
In fact, take-off may well be a relatively minor problem, since the engine thrust will be equal to, or greater than the weight of the aeroplane, in which case it could rise vertically like a V-2—at any rate with the assistance of booster rockets. Landing the thing is quite another story, however, even though the weight would be much reduced by fuel consumption.
There is an aerofoil with good supersonic characteristics which is of some use at low speed, and that is one composed of a double arc, again about 5 per cent thick. Such an aerofoil, by the use of both leading and trailing-edge flaps can be given quite a good camber for landing.
The plan form of a supersonic wing is another problem. To get a reasonable structural depth a low aspect ratio is necessary anyway, so the designer is left with a choice of a square
or delta shape. Sweepback reduces the violence of the sonic effects and also delays them to higher speeds. With a 60 deg. sweepback the drag peak does not occur until M=1-4, but the drag thereafter is rather higher than for a thin straight wing. Designers will probably choose the planform that gives the least inconvenient drag rise speed for their requirements, since either type of low aspect ratio wing is reasonably stable for landing. We are, however, a long way yet from practical supersonic flight. It is true that several types of fighter, including the Sabre and the Hunter, can exceed Mach 1 when full engine power is assisted by the energy of gravity, but the resulting dive lasts only a few seconds and can scarcely be called fully controlled. What these dives do prove is that, given sufficient power, a normal high-speed aeroplane can achieve supersonic flight and that under this condition the buffeting and other phenomena are greatly reduced. It also shows that in supersonic air flow conventional control surfaces become very sluggish, quite unsuited for combat.
Experience with supersonic research aeroplanes has been, if anything, even more discouraging. The first American aeroplanes , the Bell X-1 and the Douglas Skyrocket, which have flown at speeds of around 1,000 m.p.h. did so under freak, barely controlled conditions. Little has been said about results, but both aeroplanes were launched in the stratosphere from B-29s and then climbed higher still to get into the thinnest possible air. Powerful rocket motors gave a high thrust independent of oxygen, but of only two or three minutes' duration because of the enormous fuel consumption—about one ton a minute. Both these aeroplanes, although rather futuristic about the fuselage, were of clean, conventional, subsonic design, with small, thin, laminar-flow wings, that on the X-1 actually being straight, while the Skyrocket's was swept. In supersonic flight the violent lateral instability of the sonic zone persisted and was aggravated by the fact that the thin air provided little lift for the control surfaces.
Since it seems that we are still some years off from true supersonic flight, it might be as well to examine in more detail the venue of the immediate future, the troubled sonic zone. The picture is not just one of a set of three conditions—drag
rise, loss of lift, trim change—it is complicated by a crossing over and inter-weaving of various external factors.
The high speed of sound at sea level and the dense air give the best control conditions and the largest range of speed between the stall and "Mcrjt". At sea level, however, it is possible to overload and break the aeroplane with high g turns. Gusty air can cause dangerous turbulence and even fluctuations in Mach number. As the aeroplane flies higher conditions reverse . The speed of sound falls and with it the onset of compressibility , while at the same time the stalling speed rises, so that, instead of there being a range of perhaps 500 m.p.h. between the two extremes, there is only about 160 m.p.h. This speed range reduction is due both to the fall in air density and to the fact that the stalling speed is now entering the range at which compressibility begins to reduce the lift coefficient.
It is important to appreciate that, on the one hand, lower air density is making the air behave as if speed were reduced— in fact the indicated air speeds will be low since the A.S.I, is operated by air pressure. On the other hand, the speed of sound has fallen by 112 m.p.h. at 36,000 ft. so that the high true air speeds for the same I.A.S. readings represent much higher Mach numbers. What is called the manoeuvring envelope has been greatly reduced, for the aeroplane will be restricted to gentle turns of wide radius, otherwise it will stall. It is here that the question of wing loading enters the picture again. With a low wing loading (as on British types) the range of the manoeuvring envelope is maintained to a greater height. A concrete example lies in a comparison of the MiG-15 with the Sabre, the former's wing loading of about 45 lb./sq. ft. making it superior in all respects above about 25,000 ft., although below that height the cleaner, more powerful Sabre can climb and fly faster despite its 65 lb/sq. ft. wing loading. Three important factors affect sonic flight; assuming, of course, that there is ample power to achieve this speed.
Fuselage and wing must be thin and without sharp curves to cause local expansion and air speed increase. Surface finish must approach perfection, being of the order of a few thousandths of an inch tolerance. Attention to detail has in some cases shown remarkable results; on one fighter, the removal of the pitot head from a wingtip, where it had been causing air disturbances leading to aileron buffeting, to the fin made the aeroplane safe to dive supersonically. Snaking and misbehaviour of the tail have been traced to panting of the skin on the engine nacelle lip.
Skin stiffness sufficient to avoid panting and local deformation under aero-elastic effects is essential, even though the main structure may have to be highly flexible. Thick skin plating, either with machined integral stiffeners or forged in a single piece is the most certain way of ensuring local stiffness. Fundamentally, wing shape is, of course, one of the most vital features—but it is also the one over which most controversy rages. The straight wing gives good low-speed characteristics , but unless made almost impossibly thin, compressibility sets in early. Broadly speaking the more a wing is swept the better are its high-speed and the worse are its low-speed characteristics, while aero-elastic distortion is bad because of the geometry.
The air forces in sonic flight are very high and, in particular, the aero-elastic effects are acute. Especially troublesome is the plain geometrical fact that a swept wing loses incidence when
it bends upward, which means that when pulling out from a dive the tips will lose lift and the aeroplane will tighten into the pull-out, with a risk of overloading or a high-speed stall. (This occurred with the Skyrocket, in which Bill Bridgeman found he had to take off elevator control almost as soon as he had applied it.) Again because of flexibility, operation of the ailerons can bend the wing, instead of rolling the aeroplane, so causing control reversal. The Short S.B.5, a research type, has been built to test the effect of ailerons set chordwise at the wingtips as a means of overcoming reversal—and of getting better response at low speeds. Rotatable wingtips, as on the Pterodactyl, are another solution for overcoming both aeroelastic and tip stalling problems—spoilers are also a possibility. Up to medium sizes, at least, the delta wing seems to have most advantages and fewest drawbacks. It presents the deepest structure, yet can have the smallest thickness/chord ratio, and
so it is stiff and resistant to aero-elastic effects. Acute sweepback does not cause wingtip stalling and the large area ensures a low wing loading. Any loss in efficiency at high altitude due to high span loading causing higher induced drag is more than compensated by the reduced drag due to the lower incidence made possible by the low wing loading. It is only fair to add that in very large sizes the delta may suffer because of the skin friction of its very large area.
One American research aeroplane, the Bell X-5, suggests an interesting compromise, the variable-sweep wing. This is, at first sight, a startling proposition and it may well prove too cumbersome for everyday use. It is possible, however, that for a fighter the gain in operational flexibility would be worth the extra weight. Not only would the low-speed troubles be solved, but the pilot could shift his critical Mach number back, and forward according to the speed at which he wished to fly. Although this idea seems rather far fetched, it is worth recalling that twenty years ago there were many voices decrying the advantages of retractable undercarriages and explaining that the weight and complication made them valueless.
Direct aerodynamic problems are not the only ones confronting the designers of sonic and supersonic aeroplanes. Looking slightly ahead, overheating is a trouble of the immediate future. It is true that an aeroplane which is now almost too hot to touch on a tropical aerodrome may well be flying in the stratosphere at —56 deg. C. within 10 minutes— where, incidentally, its fuselage may have contracted as much as two inches. On the other hand, if this aeroplane flies at the
speed of sound at sea level the skin friction and compression of air by the formation of shock waves will raise its temperature considerably—to the point of being unbearable for the pilot. At a speed of 1,300 m.p.h. (that is Mach 2 in the stratosphere) the temperature rise is about 150 deg. C. and without refrigeration the pilot would be literally boiled alive. The fuel supply and much of the equipment must also be air-conditioned and there is the problem of cooling the structure, particularly near the engine. Already radiation from high-powered turbojets is causing trouble and with only pre-heated air available for cooling it will be almost impossible to protect the structure from the surplus heat of engine and jet pipe.
Since the heat rise is proportional to the square of the speed, it does not require a very great increase in speed to reach temperatures at which aluminium-alloys begin to lose their strength. Plastic hoods too, fail completely. To guard against these troubles two new American research aeroplanes, the Bell X-2 and Douglas X-3 which are intended eventually to reach 2,000 m.p.h., are built largely of stainless steel, nickel alloy and titanium, all materials that will withstand high temperatures. Actually, since neither of these rocket-propelled aeroplanes will be able to carry fuel for more than about four minutes and they will fly very high, there will scarcely be time for the airframe to soak up much heat.
Glass and asbestos plastic are other materials suggested for the structure of these very high-speed aeroplanes. If these do prove satisfactory, and it is a very big if, it will mean that wings and fuselages will be moulded. That, in turn, will affect the shape, for moulded parts are smoother and it is not easy to alter them.
The question of power for supersonic flight is an acute one. The ordinary jet engine is likely to have difficulty in breathing in supersonic air flow and there are complications in the formation of shock waves in the air intake. The troubles have been foreseen and may be overcome, but they will probably entail variable intakes to suit the two ends of the speed range, since the sharp lips needed for high-speed flow cause turbulence 'at lower speeds.
Rockets provide the best source of power, particularly the fully controllable motors now available which do permit cruising at very low power. The engine is small, light and relatively simple, but the propellant consumption is prohibitive except for special duties. At full power a rocket for a supersonic fighter would use at least a ton a minute. Even supposing
it can be throttled to give a cruising thrust of about ten per cent, it is difficult to foresee a duration of more than half an hour, without the aeroplane becoming too large. The rocket motor is the only answer for flying above 80,000 ft. and for the most rapid climb, but its more obvious use is to boost the ordinary engines for climb and speed. A tempting power unit is the ram-jet, or athodyd. This engine was first proposed in 1913 by a Frenchman, Rene Lorin, and is the jet engine in its most elementary form. Air from a small intake is compressed by converting its speed energy into pressure when it is slowed down by entering a wide chamber.
Fuel is added and the heat energy is converted into velocity by ejecting the gases through a venturi. The ram-jet obviously requires considerable speed to achieve sufficient compression to be self-propelling, but its simplicity is very tempting. Much work was done on it by the Germans and another Frenchman, Rene' Leduc, has managed to fly a machine, the 010, on this power alone. Since, however, his ram-jet needs an initial speed of 200 m.p.h., the aeroplane has to.be launched from a carrier plane. Estimates suggest that in supersonic flight, where compression ratios of six to one can be obtained with a welldesigned intake, huge thrusts can be obtained from slender tubes about a foot in diameter. It is also likely that a supersonic ram-jet could operate as much as 20,000 ft. higher than a turbo-jet, since the latter is limited by the stalling of the compressor blades when the air gets very thin. At very high speed the ram-jet is the most economical engine known—since it requires no internal power for a compressor—but it is very extravagant until sonic speed is reached because of the relatively poor compression ratio. The great failing of the ram-jet as an aeroplane power plant is that it can only be used at full speed. At present, there is little prospect of being able to throttle it efficiently, which makes it almost useless for aeroplanes save as a supplementary power unit—for guided missiles or helicopters the ram-jet is excellent.
Yet another problem exists for high speed, high altitude pilots, and that is vision. When flying eight or ten miles above the earth details of the ground are small, even if it is not obscured by cloud or haze. The sky itself merges into an indeterminate , monochrome vagueness without a horizon and it is difficult to focus the eyes without some cloud or other reasonably distinct object. A small aeroplane is visible at between five and eight miles, so that at 600 m.p.h. two approaching fighters will have come within sight of and passed each other in fifteen seconds. The angle of vision from the cockpit can make even simple navigation very difficult since the angle subtended by the fuselage obscures the ground for many miles beneath and ahead of the pilot—and at 600 m.p.h. small errors in maintaining direction soon put the aeroplane many miles off course. This factor may lead to more attention being paid to putting cockpits in the nose.
There is certainly a vision barrier to add to the sonic and thermal barriers when considering high-speed flight. Radar both for navigation and the locating of other aeroplanes seems to be a necessity in the future.
The foregoing is no more than an outline of the problems of high-speed flight which face the aeroplane designer of today and in the immediate future. To present a true, so far as it is known, and full picture would require not a short chapter but a larger volume. It would also be wrong to leave the reader with the impression that supersonic flight is the only goal after which aeroplane designers are striving. It is true that speed is the essence of aeronautics, but there is also much to be done in the direction of safer and more economical flying. The Handley Page Gugnunc and the Fieseler Storch have already been quoted as examples of what can be done in the way of coaxing lift out of the air at low speed. More recently the Prestwick Pioneer, designed by Scottish Aviation, has shown that slats and extension flaps can give the remarkable speed range of 38 to 162 m.p.h. Unfortunately, all such slow flying must be paid for either with extra engine power or a lower top speed due to higher wing drag.
An excellent invention toward slower and safer flying was the N.A.C.A. double-slotted flap, which was evolved after much wind-tunnel experiment. This type of flap has a slat interposed between the flap and the wing, so that flow over the flap is greatly improved. Unfortunately, the dimensions of the slot between wing arid slat have to be kept constant and this involves a complicated linkage; it has been used on several transports, but not, it is sad to relate, for the lowering of stalling speed so much as to increase the payload. Another promising line of development is the suction wing. For at least twenty years aerodynamicists have been toying with the idea of removing the sluggish, boundary layer of air that clings to the greater part of the wing and is responsible for most of the induced drag over the whole speed range and also initiates the stall.
Laminar-flow aerofoil sections, in which the slope of the front part of the camber is designed to have a steady pressure gradient, can keep the air flowing smoothly up to about 30 or 40 per cent of the chord. Such profiles are delicate in the extreme and must be accurate to within a few thousandths of an inch; dust, dirt and scratches can all lead to breakdown of the laminar flow. Any extension of the laminar area must be achieved by other means than mere shape and accuracy.
Before the War a special Miles monoplane was built on which areas of wing skin were perforated with small holes and, the boundary layer was sucked away by a motor and fan in the fuselage. One of the objects of doing this is that thicker sections with higher lift co-efficients may be used. These thicker aerofoils, of 20 or 25 per cent, are naturally much easier to construct and to make stiff than are the normal ones. Furthermore , they are more efficient as providers of fuel volume. Today, instead of perforated sheet for the suction areas, a special porous metal would be used.:
The early suction experiments led to further work with special aerofoil sections. Some of these new shapes were of exceptional thickness, up to 50 or 60 per cent, and of a most surprising cusp form, caused by cutting away the trailing edge. Many varieties were tested in the wind tunnel and that shown in the sketch was flown as a glider in Australia. This is less extreme than some of the shapes and it has very high lift and low drag characteristics. The theory lying behind these curious aerofoils is that the rear part of the upper surface, where the boundary layer is normally setting up drag, is cut away and by sucking in air at the lip so formed the flow is made to conform to the recessed surface. It had always been thought that one of the dangers of such a system would be failure of the suction motor, or blockage of the slots or the ducts. The Australian experiments proved that the wing did, in fact, behave quite
well without suction so far as lift was concerned, although the drag was naturally high.
There have been several projects for using these Griffiths suction wings for large long-range air liners. The wing is best suited to moderate cruising speeds of about 400 m.p.h and could be applied well to an airscrew-turbine aeroplane. Again, it might well be used to increase the endurance, by lowering the drag, of a patrol flying-boat. These applications would be to very high aspect ratio wings so that the lowest possible span loading would combine well with the already low drag of the wing. A French experiment on these lines is the ultra-high aspect ratio Hurel-Dubois monoplane. Originally flown in model form in 1949 as a single-seater with the almost incredible aspect ratio of 32, the ideas of M. Hurel showed themselves to be at least worth investigating. Finance having been forthcoming from the French Air Ministry, two medium-sized transports were built, the first being flown on January 27th, 1953.
The full-scale aeroplane, the H.D.31, has a less startling aspect ratio of 20-2, but even so it is more akin to the wing of a sailplane than that of an aeroplane. The idea behind this configuration is to get a large amount of lift from a given wing area with a reasonable structure weight, even so it is necessary to brace the wing at mid span to keep down the weight of the structure. Aerofoil-section struts, which contribute something to the lift, are used, but naturally these add considerably to the overall drag. Double-slotted flaps extend from the fuselage to the ailerons and are very effective because of their large span. Speed is not one of the attributes of the H.D.31, so a fixed undercarriage is used. The main characteristics are the ability to lift a large load on very moderate horsepower and also to get into the air with a much shorter run than is usual today. The first H.D.31, with two 800 h.p. radial engines, is designed to have a speed range of 64 to 168 m.p.h. and it is estimated that it can lift a payload of 7,150 lb., plus fuel for 620 miles, to 50 ft. in a distance of 600 yards.
An idea of the curious proportions of this unusual aeroplane can be had from the fact that its wing span of 147 ft. 7 in. is greater than that of the Stratocruiser, while fuselage length, 72 ft. 2 in., and cabin size approximate to that of the Bristol Freighter. Whether this aeroplane is just a fantastic ideal or a sound commercial proposition remains to be seen. A combination of the Hurel-Dubois principles with the thick suction wing presents fascinating possibilities for economical long-range freight carrying.
What then are the aeroplanes of the future going to be? Squat delta wings, slender, high-aspect ratios, diminutive knife-edge stubs, or shapes as yet unforeseen? It is always very dangerous to prophesy, as anyone can see by looking at forecasts made over the past forty years. In 1914 an eminent scientist told The Royal Aeronautical Society that the limiting weight of an aeroplane would be a little over 2,000 lb., and at about the same time another expressed the view that a span of a hundred feet would be best for economic flight. Predictions of future shapes have nearly always been far from the truth. Stronger materials have done as much to confound the prophets as have aerodynamic advances, while the enormous advances in power plants have brought in another unpredictable . The rise in wing loading has had as profound an effect on shape as it has on performance. Ahead lie, probably, even higher wing loadings for some duties, such as pure speed, and the imponderable effect of atomic power plants.