Category Aircraft Flight

For safety in the event of power failure, duplicate, triplicate or even quadruple systems are provided on aircraft with power controls. Davies (1971) gives an excellent diagram showing the complex arrangement used in the Boeing 747.

On very large aircraft there may be as many as four separate systems, each driven by a separate engine, and each with an auxiliary air-turbine driven backup supply. Most surfaces are powered by more than one system, and there may be alternative surfaces for each function. Unfortunately, no system is totally fail-safe and one fatal accident occurred when a fin was lost, because all four systems were used to operate rudder surfaces. Although the aircraft was flown successfully for some time with no fin, a total loss of hydraulic fluid finally resulted in disaster. Such fluid leakage can be prevented by means of limiter valves, which seal when excessive fluid flow occurs.

Mechanical devices are not necessarily any safer, however, and many accidents have been caused by control cables breaking or jamming. Hydraulic tubing and electrical wiring have the added advantages that they can follow much more tortuous and convenient paths.

Because the damping of the Dutch roll depends mainly on the moment pro­duced by the fin of the aircraft in response to the rate of yaw of the aircraft, it will be altered by changes in the fin effectiveness for a given amplitude of the motion. The fin effectiveness in response to yaw rate is reduced by in­creased altitude. This occurs for exactly the same reason that the effectiveness of the horizontal tail surface is reduced in response to pitch rate, as we saw earlier.

The Dutch roll therefore tends to become less stable for a particular aircraft attitude as the altitude increases. This effect can be very marked. D. P. Davies in his excellent book Handling the Big Jets (1971) states that in a landing configuration, Dutch roll can pass from a stable to an unstable

condition between altitudes of 1000 ft to 8000 ft for a typical airliner. A similar deterioration in cruise configuration can take place between 18,000 ft and 22,000 ft.

Fortunately there is some slight compensation due to the fact that the same effect improves spiral stability with altitude. This, however, is small comfort as we are usually far more concerned with Dutch roll behaviour.

A further factor which works against us is caused by the fact that many high altitude aircraft, such as airliners, use some degree of sweepback. As we saw previously, this acts in the same way as dihedral, making the Dutch roll worse. Sometimes a degree of negative dihedral (anhedral) may be employed to counteract the effect (Fig. 12.13), but frequently it is better to avoid this and become resigned to artificially enhancing the stability characteristics, as will be described later.

Effect of structural stiffness

In the above arguments we have treated the aircraft as though it were a rigid body, but in reality there will be a considerable degree of flexibility both in the airframe and in the control systems. This will clearly influence the stability of the aircraft. In general, flexibility in, for example, the rudder will reduce the damping and make the Dutch roll less stable than before. Thus, when all factors are considered, it is difficult to design an aircraft, particularly a swept – wing aircraft which is required to cruise at high altitude, which naturally com­bines satisfactory spiral and Dutch roll behaviour. Nowadays the problem can be overcome electronically, as we describe in the following section.

For most wing sections, the amount of lift generated is directly proportional to the angle of attack, for small angles; the graph of CL against angle of attack is a straight line, as shown in Fig. 1.17. However, as illustrated, a point is reached where the lift starts to fall off. This effect is known as stalling. The fall-off may occur quite sharply, as in Fig. 1.17 which shows the variation of lift coefficient with angle of attack for a wing with a moderately thick aerofoil section (15 per

cent thickness to chord ratio). You can see that the stall occurs at an angle of attack of around 12°. A thin uncambered wing may stall even more sharply, and at an angle of attack of 10° or less. A sudden loss in lift can obviously have disastrous consequences, particularly if it happens without warning.

The stalling characteristics of an aircraft wing depend not only on the aero­foil section shape, but also on the wing geometry, since not all of the wing will stall at the same angle of attack.

Stalling occurs when the air flow fails to follow the contours of the aerofoil and becomes separated, as illustrated in Figs 1.18 and 1.19. The causes of this flow separation are dealt with in detail in Chapter 3.

Once the flow separates, the leading-edge suction and associated tangential force component are almost completely lost. Therefore, the resultant force due to pressure does act more or less at right angles to the surface, so there is a significant rearward drag component. The onset of stall is thus accompanied by an increase in drag. Unless the thrust is increased to compensate, the aircraft will slow down, further reducing the lifting ability of the wing.

After the stall has occurred, it may be necessary to reduce the angle of attack to well below the original stalling angle, before the lift is fully restored. An aircraft may lose a considerable amount of height in the process of recovering from a stall, and trying to prevent its unscheduled occurrence is a major con­cern of both pilots and aircraft designers. Later on, we shall describe some of the preventive measures and warning systems that may be employed.

Apart from the problem of wing stalling, there are several areas in the flow where we wish to prevent flow separation, or inhibit the build-up of thick low-energy boundary layers. Flow separation in air intakes of gas turbine engines is a particularly serious problem, since it is most likely to occur at high angles of attack on landing and take-off; just the time when it is least wanted. Stalling of the intake flow can cause the engine to lose power, or flame-out (switch-off) altogether, with potentially disastrous consequences. Some air­craft are even fitted with a device that automatically operates the starting igniters at high angles of attack.

In high speed flight, flow separations may also be caused by the interaction between shock waves and a thick boundary layer. Notice how the air intake of the supersonic Tornado shown in Fig. 3.15 is separated from the fuselage, to form a slot through which the fuselage boundary layer can pass, preventing its interfering with the intake flow.

In addition to the problem of air intakes, it is also important to prevent separation in the vicinity of control surfaces, since the last thing we want to lose in the approach to a stall, is the ability to control the aircraft.

One way to prevent local flow separation is to apply engine generated suction via small slots or openings in sensitive areas. An alternative pass­ive measure is the attachment of small tooth-like vortex generators on the surface. These are designed to give a highly turbulent surface flow, thus inhibiting separation. Figure 3.7 shows the vortex generators on the wing of

a Buccaneer. This type of vortex generator may be seen on many early swept – wing aircraft, where they were used to try to overcome the problems described below.

We have spent some time in looking at the differences between flows at high and low speed. It is worth emphasising that, for both the duct flow and the ‘external flow’, although Bernoulli’s equation becomes inaccurate as speed increases, it is still true that an increase in speed is accompanied by a decrease in pressure, irrespective of whether the flow is sub – or supersonic.

We find that our criterion for high speed, introduced above (the speed at which density changes first become apparent) is related to the Mach number. For an aircraft this usually occurs at flight Mach numbers above about 0.5. Rather than a single measure of what constitutes a ‘high speed’ we can now begin to identify Mach numbers at which distinguishing features of high speed flow begin to appear (Fig. 5.5).

This figure shows the Mach numbers at which we will obtain our typical low subsonic and fully developed supersonic flows. It also shows a number of other features, which we will discuss shortly, such as the intermediate stage between these flows, the transonic speed range. The advent of important heating effects caused by the passage of the aircraft through the air is also shown.

The overall efficiency of a gas turbine propulsion system depends on two major contributions, the Froude efficiency which, you may remember, is related to the rate at which energy is expended in creating a slipstrean or jet, and a thermal efficiency, which is related to the rate at which energy is wasted by creating hot exhaust gases.

As noted earlier, the Froude propulsive efficiency of the pure turbo-jet is low, because thrust is produced by giving a small mass of air a large change in velocity. However, for a fixed amount of thrust, as the speed of a jet or gas – turbine-propelled aircraft increases, the air (mass) flow rate through the engine also increases. A smaller change in velocity is needed for this larger mass of air, and the Froude efficiency thus improves. However, for an aircraft in steady level flight, the thrust required is equal to the drag. Since the drag varies with speed, the thrust required must similarly vary, so the overall efficiency of propulsion depends on the drag characteristics of the aircraft. This inter­dependence between the propulsion device and the aircraft aerodynamics is an important feature of aircraft flight and is described further in Chapter 7.

In general we may wish to design for one of two goals as far as the climbing performance is concerned. Firstly the climb angle rather than the rate of climb may be of primary consideration. This will be true, for example, if we are con­cerned with the take-off performance. The primary concern will be to avoid hitting high structures in the vicinity of the airport and for this it is the angle of climb that is critical. In other circumstances it may be the rate of climb that is the factor of most interest. This would be true, for example, for an interceptor aircraft. It is important at this stage to realise that the maximum angle of climb and the maximum rate of climb do not occur together, but as we shall see, at two distinct operating points.

For aircraft designed for cruise in the transonic range the use of swept forward, rather than swept back, wings offers some advantages. An optimum spanwise load distribution can be obtained with conventional taper towards the tip. The problem of boundary layer drift towards the tip, which encourages tip stall is also alleviated. Because the velocity component along the leading edge is now directed inboard, the boundary layer tends to thicken towards the root rather than the tip.

With this catalogue of virtues the reader may wonder why forward sweep has not been employed exclusively. The main problem lies with the struc­tural behaviour of the wing. When the wing is loaded the angle of attack increases unlike the swept back wing (Fig. 9.13). Because of this the tip lift is increased and the deflection worsens. The wing can then suffer progressively increasing twist. This condition is known as divergence and is encountered again in Chapter 14. The problem is made worse because an increase in load at the tip of a swept forward wing will produce a nose-up pitching moment thus increasing the angle of attack over the whole wing and again increasing the load.

After early attempts at using forward sweep (e. g. the Junkers 287 in 1942) the structural problems led to the virtual disappearance of the idea. Recently, however, advances in structural materials have led to renewed interest in the concept. Modern composite materials (Chapter 14) allow suitable flexural behaviour to be designed into the wing to prevent the occurrence of divergence. Another technique which can be employed is to automatically sense the twist as it occurs and to use a computer-driven system to deflect the ailerons down­ward simultaneously in order to cancel the twisting effect.

The X-29 aircraft (Fig. 9.20) is an example of an experimental forward – swept configuration. Radical new design features, such as forward sweep, are, however, expensive and commercially risky to introduce. It is likely to be many years, therefore, before the conventional swept back configuration is seriously challenged.

In the discussions above, we have assumed that the control stick is being held firmly, so that there is no movement of the control surfaces. However in real­ity, changes in aircraft attitude will alter the loading on the control surfaces; trying to move them.

Some power-operated controls are irreversible; that is, the aerodynamic loads cannot drive them. On most aircraft, however, and invariably on those with servo-assisted or manual controls, the elevators will move in response to changes in pitch, if the stick is not held firmly. The influence on stability of allowing the stick to move freely depends on the design of the surfaces and the control mechanism. The degree of aerodynamic balancing, the stiffness and the inertia of the system components are all important factors. In general the effect of leaving the stick free is to reduce the static stability.

Because of the very real difficulty and high pilot workload during landing this phase of flight has been the subject of rapid development in systems to aid the pilot in his task. As well as improving safety these systems allow for better aircraft utilisation because one of the obvious limitations for operation under

This wing low

Compensating rudder

causing drift

Sideslip relative to wind

Fig. 13.10 Use of sideslip to correct drift

Aircraft is rolled slightly to induce sideslip into wind Aircraft axis is kept aligned with runway by use of rudder

purely ‘visual flight’ is that low cloud may make the approach impossible even to attempt.

Here we shall very briefly describe some of these aids, the proper study of which is a separate discipline in its own right.

It is apparent from the above discussion that one of the main problems in landing is that of following an accurate glide path to the runway threshold. This can, of course, be particularly difficult in conditions of poor visibility par­ticularly for large aircraft where the glide path needs to be established several miles from the touch-down point. The main purpose of any landing aid is thus to aid accurate flying during this phase of the landing. For modern aircraft a number of such aids is available. Among these are radio beacons which can be used for general navigational purposes as well as landing aids such as the non-directional beacon (NDB) which supplies the aircraft with a directional ‘fix’ on a known ground location or the very high frequency omni-directional beacon (VOR) which supplies both directional and range information. The most common aid dedicated solely to landing is the Instrument Landing System (ILS). In this a pair of radio beams are arranged to cross on the glide path. Deviation from the glide path is then indicated by a cockpit instrument, the ILS indicator (Fig. 10.2). Satellite-based GPS systems are also used.

Automatic flight along the glide path can be achieved by adding an auto­matic throttle and flight control system, with accurate height information being obtained from a radio altimeter. Automatic flare is provided to bring the air­craft on to the runway. Fully automatic landing systems of this type have greatly increased the range of conditions under which safe aircraft operation is possible.

A newer alternative to the ILS system is the more accurate Microwave Landing System (MLS). This system is however being challenged by an even newer technology, the Global Navigation Satellite System (GNSS). The advan­tage of the latter is the that it relies on signals from satellites and does not require expensive ground installations at airports. After lengthy trials, a satel­lite landing system has been cleared for use at a small number of selected airfields.