Category Aircraft Flight

The autogyro differs from the helicopter in that the rotor blades are not driven directly by an engine. A conventional aircraft propulsion system, such as an engine and propeller, is used to propel the aircraft forwards. Unlike the helicopter, the rotor axis is tilted backwards during flight, and the blades are

blown round by the relative air flow. As they ‘autorotate’, they generate lift, like the blades of a helicopter.

The autogyro requires a certain amount of forward speed in order to main­tain sufficient autorotation to lift the aircraft. Thus, although it has the advant­age of being able to fly very slowly and make a short vertical final descent, it can not hover or take off vertically. Although mainly of historical interest, there has been some renewed interest in the autogyro for recreational flying, as illustrated by the example shown in Fig. 1.31.

As on the helicopter, the blades of an autogyro are normally free to flap up and down. In flight, the difference in the relative velocity between the advanc­ing and retreating sides again tends to cause the blades to flap up towards the front and down towards the rear, thus tilting the axis of blade rotation rearwards. For an autogyro, this rearward tilt presents no problem, as the forward thrust is provided by a separate propeller. Rearward tilt is in fact necessary to sustain the rotor rotation. As the blades are hinged freely to the hub, they cannot produce a strong rolling moment, and for an autogyro, cyclic pitch control is not essential.

A recent development is a flexible-skinned wing where the aerofoil section can be bent by internal jacks to give varying degrees of camber. The purpose of this arrangement, is not primarily to provide a high lift coefficient for landing and take-off, but to enable the camber to be matched to the flight conditions. With this mechanism the aircraft can be flown with a high aerodynamic efficiency over a wide range of conditions; a feature that is particularly desirable in a combat aircraft which may be required to fly at subsonic, transonic and super­sonic speeds during different phases of a mission. Figure 6.35 shows a F-111 fitted with a NASA experimental ‘mission adaptive wing’ of this type.

So far we have discussed supersonic flow at a relatively high Mach number but ignored the complicated processes necessarily involved in accelerating from subsonic to supersonic speed. We now return to our aerofoil problem in order to illustrate some of the important things which occur during this process.

Figure 5.18 shows photographs of an aerofoil at various Mach numbers from fully subsonic to fully supersonic. The photographs were taken using an optical system which shows shock waves as a dark band and expansion waves as a light coloured area. This system, which is extensively used in high speed wind-tunnel testing, is known as a schlieren system.

Because of the thickness of aerofoil, the flow is speeded up over the top and bottom surfaces. Thus the flow will eventually become supersonic in these regions, although the free stream is still subsonic. The flow is decelerated from its locally supersonic speed by a shock wave (Fig. 5.18(b)). The local super­sonic patches on the top and bottom of the aerofoil grow in extent as the free stream speed is increased, and the strength of the shock wave also gets greater.

It can also be seen from the schlieren photographs (Fig. 5.18(b)) that the pre­sence of the shock waves leads to boundary layer separation, about which more will be said in the following section.

As the free-stream Mach number is increased further, the shock wave moves further back as well as increasing in strength. As the free-stream flow just becomes supersonic, another shock wave starts to appear upstream of the aerofoil forming the bow shock wave mentioned previously. This bow shock wave gets progressively nearer to the nose of the aerofoil as the Mach number is increased and the typical flow for a fully supersonic aerofoil shown in Fig. 5.18(c) is obtained.

Further improvements in efficiency are obtained by increasing the by-pass ratio; the ratio of the amount of air by-passed around the core engine to that which passes through it. Increasing the by-pass ratio requires making the lowest pressure stage larger in diameter.

Шлм:

Three independent sets of turbine discs

Fig. 6.24 A triple-spool turbo-fan engine based on the Rolls-Royce RB-211

This high by-pass ratio engine has three concentric shafts or spools running at different speeds. A large proportion of the thrust is provided by the front low speed fan

All of the more recent designs of civil jet transport aircraft use high by-pass turbo-fans

Figure 6.24 shows a high by-pass ratio Rolls-Royce RB-211 turbo-fan which uses three shafts or spools, one being dedicated to the fan drive. A significant proportion of the overall thrust comes from the pressure difference across the fan blades, as with a propeller.

The big primary fan results in an engine of much larger diameter than the earlier simple arrangements. The large diameter of the fan is evident in Fig. 6.25.

In turbo-fan engines, the speed of the air relative to the surrounding shroud walls is subsonic, but relative to the moving fan blades, it is supersonic. As explained earlier, however, losses due to shock wave formation are less severe for fans than for simple propellers. The shroud helps to suppress some of the noise from the fan, and because of the low jet speed, turbo-fan engines can be extremely quiet. Jet noise is related to the eighth power of the jet speed. The British Aerospace 146, shown in Fig. 6.26, is an outstanding example of a quiet turbo-fan-propelled aircraft.

The turbo-fan provides a practical means of propulsion at Mach numbers above the limiting value of about 0.6 to 0.7 for a conventional propeller. It also represents an alternative to the much less efficient turbo-jet.

Low by-pass engines are now normally used for all combat aircraft, even for types designed for flight at supersonic speeds. High by-pass engines are mainly used for subsonic transport aircraft, both civil and military.

We have spent some time considering the way in which lift is produced in subsonic flow (Chapter 1). There are some similarities in supersonic flow. The lift is produced by a difference in pressure between the top and bottom sur­faces, and this requires a high speed on the top surface and a reduced speed on the lower surface whether the flow is subsonic or supersonic.

However, although the two cases have this much in common, there are con­siderable differences between the flow patterns of the high and low speed cases. For example shock wave generation is an important factor at high speed, and suitable design to minimise the drag caused by the formation of these shock waves is extremely important. With these points in mind, it is likely that the aerofoil sections which perform best in supersonic conditions may look con­siderably different from their low speed cousins.

In Chapter 5 we examined the changing flow over a typical subsonic type of aerofoil as the upstream Mach number increases (Fig. 5.18). The flow is characterised by the development of shock wave systems at the leading and trailing edges. In the supersonic flow regime, the flow field is entirely super­sonic, with the exception of a small patch of subsonic flow on the blunt lead­ing edge in the region of the stagnation point.

The wave drag associated with this type of aerofoil is high because of the strong bow shock wave. Such an aerofoil is therefore not suitable for use in supersonic flow unless the wing is swept to reduce the effective approach velocity (see Chapter 2).

In order to reduce the strength of the bow shock wave it is desirable to make the leading edge of the aerofoil sharp. This will remove the region of near nor­mal shock associated with the blunt leading edge, with a consequent reduction in wave drag. Figure 8.5 shows a particularly simple form of supersonic aero­foil, the ‘double wedge’ section. We met this briefly in Chapter 5 and now look at its suitability for practical application.

Figure 8.5 also gives a comparison of the surface pressure distribution on the double wedge aerofoil at subsonic and supersonic speeds for small angles of attack. In the subsonic case we would expect to get the typical suction peak near the leading edge on the upper surface followed by a recompression as we move towards the trailing edge. On the bottom surface we will obtain a

Pressure lower than surrounding atmospheric

Pressure higher than surrounding atmospheric

Pressure higher than surrounding atmospheric

Pressure lower than surrounding atmospheric

Fig. 8.5 Pressure distribution on double wedge aerofoil

(a) Subsonic (very low angle of attack) (b) Supersonic stagnation point, and the higher pressure on the undersurface will also con­tribute to the overall lift.

The pressure distribution on the aerofoil in a supersonic air stream is very much simpler, each of the four faces of the diamond cross-section experiencing virtually constant pressure. This follows from the fact that the flow over the two forward-facing surfaces is uniform as the bow shock waves simply deflect the entire flow until it becomes parallel with the surface direction (Chapter 5). Similarly the expansion fans generated from the apexes on the upper and lower surfaces turn the flow so that it is parallel to the rearward-facing surfaces. This results in a uniform pressure over these surfaces as well.

It is when we increase the angle of attack that the biggest surprise occurs, however. We already know that, for low speeds, thin aerofoils and, even worse, those with sharp leading edges, will stall at relatively low angles of attack. Even if the flow were to successfully negotiate the sharp leading edge we would not do all that well. The sudden change in surface direction at the junction between the front and rear surface would again lead to separation; this time over the rear part of the aerofoil (Fig. 8.6(a)).

SUPERSONIC AEROFOILS 221

When we look at the supersonic flow, however, we find that the flow deflection caused by the bow shock waves removes any problem at the sharp leading edge. The flow now divides right at the leading edge rather than at an undersurface stagnation point as is the case with subsonic flow.

The flow is also quite happy to negotiate the subsequent abrupt change in surface direction by means of the expansion fan (Fig. 8.6(c)), because, as we saw in Chapter 5, the local pressure gradient is favourable at supersonic speeds (i. e. pressure reduces in the direction of motion). At subsonic speed, however, there is a locally unfavourable gradient, and so the boundary layer would separate here, even if the angle of attack were low enough to prevent earlier separation at the sharp nose.

Thus we find that aerofoils with sharp leading edges and abrupt changes in surface slope, factors which would lead to disastrous performance at low

speed, perform quite well in the supersonic speed range. Compared to a typical low speed aerofoil, for which L/D ratios in the order of 40 can be obtained, their performance does not look all that exciting. The comparatively poor performance is, of course, due to the wave drag which has to be overcome. This penalty may, however, be acceptable in many military applications where speed is of prime importance. For civil transport aircraft, too, the poor lift-to-drag ratio may be acceptable. The increased cruising speed allows better utilisation of the aircraft and a better measure of overall efficiency may be the cruising speed times the lift-to-drag ratio (Chapter 7).

The only trouble with all this is that although these simple aerofoil sections are good at supersonic speed, their performance, as we have seen, is hopeless at low speed. There is, however, a class of ‘aircraft’ which is not called upon to fly at low speeds at all; air-to-air missiles. Such aerofoil sections are therefore very often employed for these devices.

The problem of poor maximum lift coefficient at low speed is not the only difficulty encountered in the aerodynamic design of high speed aircraft. In Chapter 5 we saw how the centre of pressure moves rearwards on an aerofoil as the supersonic flow pattern is established. This change in centre of pressure position results in a large change in longitudinal trim which must either be accommodated by the provision of large tail surfaces, or by other means. For example the Concorde used fuel transfer fore and aft to change the position of the aircraft centre of gravity as is mentioned in Chapter 10.

If we want to take off or land our aircraft from conventional runways and to have a reasonable subsonic performance, as well as operate at supersonic speeds, we need to employ a wing with acceptable low speed and high speed performance and which does not have any violent change in flow character­istics as the aircraft accelerates through its speed range. It is the precise nature of this compromise which is responsible for the large variety of solutions which are found in practice.

As the aircraft yaws to the right, the left-hand wing will move slightly faster than the right-hand wing. The faster-moving left-hand wing will, therefore, generate more lift, and the aircraft will tend to roll clockwise (left-wing up). However, the fin and rudder are normally mounted on top of the fuselage, and hence, above the centre of gravity. As the aircraft yaws, the sideforce on the fin therefore exerts an anti-clockwise rolling moment. The overall result of these two opposing tendencies depends on the aircraft design. The cross­coupling of yaw and roll movements is an important feature in aircraft stabil­ity and control.

Pitch control

Figure 10.4 also shows the pitch-control surfaces of a conventional aircraft. In the traditional arrangement, the rear portion on the tailplane (horizontal stabiliser) is hinged to form an elevator. The same arrangement may be seen on the old Auster in Fig. 10.5. By deflecting the rear of the elevator upwards, the tailplane is given a negative camber, resulting in a downward (negative lift) force. As the tail is pulled down, the angle of attack of the wing is increased, so that the final result of up-elevator is to cause a nose-up pitching moment, and an increase in overall lift.

Before the aircraft has had time to respond to the pitching moment, the initial effect is to produce a temporary reduction in lift, as a consequence of the downforce on the tail. On small aircraft with low inertia, the reduction may be so short-lived, as to be hardly noticeable. On large aircraft, and particularly on tailless types, the effect can be quite severe, and the aircraft may drop some distance before the increased wing angle of attack takes effect. On Concorde, the elevator control is linked to the throttle to alleviate this problem in low speed flight.

As the tailplane is required to produce both downward and upward forces, with little force during cruise, it is normally given a near symmetrical or un­cambered section.

For an inherently stable aircraft, the absolute forward limit of centre of gravity movement is determined by the maximum balancing moment that the tailplane can produce while still retaining adequate control. The maximum rearward movement is limited by the onset of instability or excessive control response. In practice, unless an automatic control system is used, it could be dangerous to fly with the centre of gravity near these limits, and designers and airworthiness requirements impose a more restricted range of allowable safe centre of grav­ity movement. Great care has to be taken when loading and fuelling aircraft to ensure that the centre of gravity will remain within the acceptable range throughout the flight.

Compressibility effects

As explained in previous chapters, in the transonic speed range, the centre of lift tends to move rearwards with increasing Mach number. This has a similar effect to moving the centre of gravity forwards. It has to be corrected by raising the elevator, which effectively increases the longitudinal dihedral, and hence improves the static stability. However, as we have seen, the trim drag then rises, and the elevator control forces are increased.

The problem of the rearward movement of the centre of lift is aggravated on swing-wing (variable sweep) aircraft, because the wings move backwards as the sweep angle is increased for high speed flight. Such aircraft therefore invariably have very large tail surfaces, as seen on the Tornado in Fig. 11.12.

For a canard configuration, the rearward movement of the centre of lift at supersonic speeds is corrected by increasing the foreplane lift. Any increase in foreplane lift means that the main wing lift can be reduced, so there is little or no overall increase in trailing vortex (induced) drag. Thus, there should be less trim-drag penalty on a supersonic canard.

In the chapter on transonic aircraft (Chapter 9) we dealt with the buffeting that occurs at transonic speeds. This unsteadiness is primarily the result of unsteadi­ness in the position of the shock wave at the end of any patch of supersonic flow. Buffeting of the aircraft, and shaking of the controls can also happen at low speeds, when flow separation occurs at the onset of the stall. This can be regarded partially as a benefit, since it gives the pilot a warning of the approaching stall.

A mild form of buffeting is often felt when the flaps are lowered, but this is rarely a problem since it is neither severe nor sustained for long periods. A more serious form of buffeting may occur when the wake from the wing inter­acts with the tail surfaces.

Resonances

A major problem in aircraft structural design is that of resonances, which occur when a source of vibration has a frequency that coincides with one of the nat­ural frequencies of the structure. Many of the sources of vibration are purely mechanical in origin. Engine vibration is one obvious example. However the forcing frequency can also come from aerodynamic sources such as propeller wash.

Flow separations can generate turbulence that is sufficiently regular and periodic to set up resonances. In particular, bluff (non-streamlined) shapes such as fuselage-mounted dive brakes can generate a periodic shedding of vortices known as a Karman vortex street. Vortices are shed alternately from either side of a component such as a dive brake, and this produces an alternat­ing force on the brake, and anything in its wake. The ‘singing’ of telephone wires is caused by this effect.

The preferred cure for resonances is to increase the stiffness until the nat­ural frequency of vibration is well above the forcing frequency. Alternatively, the mass distribution can sometimes be changed, so that the natural frequency is much lower than the forcing frequency. Moving the engines outboard on a wing will reduce the natural frequency of bending oscillations. Care must then be taken to ensure that the forcing frequency does not coincide with one of the harmonics of the structure’s natural frequency.

Noise from engines and propellers, whether airborne as pressure waves, or directly transmitted, can result in structural fatigue due to the fluctuating loads that it produces. Noise-induced fatigue is particularly likely to occur in heli­copters and with unducted fan propulsion.

For this fourth edition we have updated the text and a number of illustrations. During the twenty years that have elapsed since the first edition was published, there have been few significant outward changes in the shape of aircraft; most developments have been in the areas of electronics, systems and structural materials. Two relatively new classes of aircraft have however emerged: the low orbit space-plane, and unmanned air vehicles. These vehicles are dealt with in this edition. As in the previous edition, we have included an appendix giving the characteristics of three different aerofoils. This information should be particularly useful for project work.

This book is intended to provide a description on the principles of aircraft flight in physical rather than mathematical terms. There are several excellent mathematical texts on the subject, but although many people may be capable of reading them, in practice few will do so unless forced by dire circumstances such as an impending examination and inadequate lecture notes. As a con­sequence, a great deal of aeronautical knowledge appears to be handed on by a kind of oral tradition. As with the great ballads of old, this can lead to some highly dubious versions.

We would of course encourage our readers to progress to the more difficult texts, and we have given suitable references. However it is always easier to read mathematical explanations if you already have a proper understanding of the physics of the problem.

We have included in our account, some of the more important practical aspects of aircraft flight, and we have given examples of recent innovations, descriptions of which are generally only to be found scattered around in assorted technical journals.

Although we do not include any mathematical analysis, we have slipped in one or two simple formulae as a means of defining important terms such as ‘lift coefficient’ and ‘Reynolds number’, which are an essential part of the vocabulary of aeronautics.

In a book of affordable size, we cannot hope to cover every aspect of aircraft flight in detail. We have therefore concentrated on items that we consider to be either important, or interesting. We have also restricted the book to cover the aerodynamics and mechanics of flight, with only the briefest consideration of other important aspects such as structural influences.

We see the book primarily as a general introduction for anyone interested in aircraft or contemplating a career in aeronautics. Students of aeronautical engineering should find it helpful as introductory and background reading. It should also be useful to anyone who has an occupational concern with aero­nautics, either as flight crew, ground staff, or as an employee in the aerospace industry. Finally, we hope that it will be read by anybody who, like us, just finds the whole business of aviation fascinating.

It is assumed that the reader has some school background in elementary physical science, and is at least vaguely familiar with concepts such as energy, and momentum.

Many years ago, someone thought up a convincing, but incorrect explanation of how a wing generates lift; the force required to support the weight of an aircraft in flight. This explanation is, unfortunately, so widely known and believed, that it is probably true to say that most of the world’s aircraft are being flown by people who have a false idea about what is keeping them in the air. Correct descriptions do exist, of course, but they are mostly contained in daunting mathematical texts. Our objective is to give an accurate description of the principles of flight in simple physical terms. In the process of doing so, we will need to demolish some well-established myths.

When an aircraft approaches the speed of sound, the flow begins to change to the supersonic type described in Chapter 5. Although it is possible to accom­modate the consequences of these changes, it is difficult to design a wing that behaves well in both low speed and supersonic flow. Even if the lift and drag

Normal component

СЛ V cos ili

Free-stream velocity V (= aircraft speed)

Sweep angle ф

Spanwise component

Fig. 2.17 Velocity components on a swept wing

Only the normal component contributes to the generation of lift properties are acceptable, the change of flow drastically alters the handling, control and stability characteristics. Difficulties occur particularly at transition from one type of flow to the other.

One way to reduce these problems, is to sweep the wings backwards, or less commonly forwards. As illustrated in Fig. 2.17, the air flow may be considered as having two components of velocity, one at right angles or normal to the span (the normal component), and one along the direction of the span (the spanwise component). The spanwise component does not alter much as the flow passes over the wing, and changes of speed occur mainly in the normal component. If
the angle of sweep is sufficient, the normal component of velocity can be slower than the speed of sound even when the aircraft is flying faster than the speed of sound.

If we look at the flow past a section of a swept wing, we will see that as long as the normal component is less than the speed of sound (subsonic), the flow patterns and general flow features are similar to those for ordinary low speed flow. This is true even though the resultant of the normal and spanwise com­ponents of velocity may be supersonic in places. The explanation for this is given in Chapter 8.

The normal component of velocity is roughly equal to the relative air speed multiplied by the cosine of the angle of sweep. It follows that the amount of sweep required increases with increasing aircraft maximum speed.

Even if an aircraft is not intended to be flown faster than the speed of sound, the wings may need to be swept, since the air flow may become supersonic locally, particularly on the upper surface, where it is moving faster than the free stream (the approaching flow). This can occur at flight speeds of around 60 to 70 per cent of the speed of sound, and since medium and long-haul airliners nowadays fly faster than this, they invariably have swept wings.

The idea of using wing sweep was developed by a group of German engineers including A. Betz, at around the time of the outbreak of the Second World War. The first successful operational jet aircraft, the Messerschmitt Me-262 (Fig. 2.18)

used a modest amount of sweep. Allied wartime jet aircraft, such as the Gloster Meteor, which were designed without the benefit of Betz’s theories, used unswept wings. After the war, when the information became available, many designs were hurriedly changed. The straight-winged Supermarine Attacker design was developed to produce the swept-wing Swift.