From the archive: Boeing 747 flight test

When a pilot gets the chance to fly a Boeing 747, he is unlikely to pass up the offer, even if it means a journey to Roswell, New Mexico where the Pan Am training base is situated. I meet Captain John Walker who is the chief pilot, and Captain Jim Waugh, who is in charge of checking and standards; their benign enthusiasm reflects that of the whole team.

It was very dark when we arrived on the ramp at 04:30 and, although one has heard many superlatives attributed to this aircraft and perhaps seen it at a distance, it is not until one arrives at the foot of the steps and looks up that its size is really apparent. A double flight of mobile stairs is needed to reach the passenger deck, and then a further climb up the spiral staircase from the first-class lounge takes one up to the flightdeck. The pilots’ eye level is 29ft (9m) above the tarmac, equivalent to the height of a three-storey building, and the engineers appear like midgets on the ground below.

The flightdeck is not unusually large considering the size of the aeroplane – in fact I would call it compact – but it is quite scientifically laid out. The pilots’ panels are relatively uncluttered, the captain’s being dominated by the attitude director indicator (ADI). A large 5in (12.7cm) instrument with an expanded pitch scale and flight-director elements superimposed. Below that is the horizontal situation indicator (HSI), an electromagnetic compass repeater having readouts above it in large illuminated figures which give distances to go to the next two way-points. Information for the HSI comes from the inertial navigation systems.

At the top right-hand corner of the panel is the flight mode annunciator, which has amber and green sectors. When a desired height is dialled up, the letters “ALT” are illuminated in the amber sector, and when that height is subsequently acquired, the amber light goes out and the green light appears. There are similar sectors associated with acquiring a localiser and glidepath, and for the flare and “go-around” modes.

The airspeed indicator on the left has a red “bug” which is moved by the speed selector knob on the autothrottle control, situated on the left of the mode-selector panel on the glare shield. Thus, when for example one dials up approach speed, the “bug” runs round the face of the instrument to the appropriate point on the scale. In the middle of the ASI there is also a numerical readout, and it has been suggested that this gives the pilot a good indication of the airspeed trend.

At the bottom left is an annunciator panel with amber windows which illuminate when there is a malfunction in the engines or ancillary systems. Reference to the engineer’s panel is then necessary to pinpoint the problem more precisely. The co-pilot’s instrument panel is a mirror image of the captain’s.

On the central pedestal are the control panels for the three separate inertial-navigation systems (INS). Nos 1 and 2 INS sets feed in roll and pitch information to the captain’s and co-pilot’s ADIs respectively; the third set is used for general navigation, and can be called into use as a standby should another set fail.

The normal drill for a cold start is for the pilot to open the cut-off lever to the “rich” position when 22% N2 (high-pressure spool rpm) is called by the flight engineer. When the starting cycle is complete, the fuel controller reverts to “idle” flow automatically, and the cut-off lever can be moved to the “idle” position at any time subsequently. Throughout the flight the EGTs were rather higher than I was used to seeing, but then they are picked off further upstream in the turbine system than they are in British engines, and the figures are, of course, purely relative.

STARTING UP

The starting order was 3, 4, 2, 1, and N2 speeds were called at 20%, 18% and 16%, resulting in EGTs which never exceeded 400°C. No. 2 was the only engine that gave a “pop” or “rumble” start that I heard, but then I was busy with my tape recorder at the time and engine noises are heard but faintly on the flightdeck at this stage. There was much more noise coming from the air-conditioning system.

The 747’s Pratt & Whitney JT9Ds are mighty engines of course, developing 43,500lb (191kN) thrust dry, and 45,000lb wet. The fans, which are nearly 8ft in diameter, give a bypass ratio of 5:1. and as the mean velocity of the exhaust is relatively low, the engine noise is no more than powerplants of half the size. Some 75% of the total thrust comes from the cold low-velocity bypass air.

The weight of the aircraft for this flight was 539,000lb (245,000kg) and the centre of gravity was at 18.8% standard mean chord. These figures could be read off a gauge on the flight engineer’s panel, the information coming from sensors on each of the undercarriage legs. The fuel load was 190,000lb.

The aircraft moved off chocks within 5min of the planned time, and quite a push of power was needed to get rolling, but this was pulled off as soon as the aircraft started to turn in order to avoid blasting the area. The undercarriage gave a slightly hard ride on the flightdeck, and on one part of the taxi track we hit a harmonic which gave a low-frequency vibration for a time.

As this was the first run of the day, the aircraft was turned on to the runway, the brakes applied and the engines run to full power. At this setting the EGTs went to 700°C, the EPR was 12.5% at 79% N1 and the fuel flow was 9,500lb per engine per hour. The engines developed a powerful buzzing noise at this stage, but the predominant noise was still the rush of air coming from the pressurisation.

At this weight the take-off speeds with a 10° flap setting were: V1, 127kt (235km/h); Vr, 135kt; and V2, 145kt.

During the take-off run the acceleration was fairly modest up to 100kt and Vr was reached in a little over 8,000ft. The rotation is made in two stages in order to avoid the risk of hitting the tail; the first attitude struck is 10° nose-up, and then, as the aircraft lifts off, a further rotation is made to restrain the speed at V2 plus 10kt with all engines operating.

By the time we reached the VOR station at 6,000ft (1,830m) on the QNH, the air was very turbulent, although it was still dark. This was probably due to wind shear, for the surface wind was 4kt but at this height, a little over 2,000ft above the terrain, it was 40kt. I would say that the turbulence was moderate, but the side whip on the flightdeck was most marked. One expects this on a long-bodied aeroplane, but I suspect that it is accentuated on the 747; the pilots sit not only a long way ahead of the wings but also well above the centre of pressure. There was another peculiar motion too under certain conditions, rather like a speedboat when planing and slapping the top of a short chop.

I made my way down to the passenger deck to see what the motion was like there, and it appeared to be very much reduced, particularly around the centre section. I think that any effects of turbulence appear to be exaggerated up on the flightdeck.

Among the many superlatives that can be applied to this aeroplane is the fact that there are 53 structurally independent movable control surfaces, and the operating power sources are so arranged that there is a high degree of redundancy.

The 747 has greater wing sweep, lower wing thickness ratio and relatively shorter wingspan than previous aircraft of its category, yet it takes off in a shorter distance than a proportionately loaded 707. The low-speed handling qualities are enhanced by four triple-slotted trailing-edge flaps, and no fewer than 26 leading-edge flaps, the outboard 20 of which alter shape during extension, flexing from flat to curved. The resulting camber gives stable laminar flow down to low airspeeds and to high angles of attack. With all the flaps extended, the wing area is increased by 21% and the lift by approximately 90%, a fact which is evident when one examines the stalling graphs; at 540,000lb the Vs is 106kt with full flap and 167kt with flaps retracted. The stall itself is conventional with a classic g break and a normal recovery; one is conscious of the smooth decay of lift from this very sophisticated wing.

After three hours’ flying, I was invited into the left-hand seat and given control of the aircraft. Having been burning fuel at the rate of 25,000lb/h, the gross weight was down to 464,000lb – relatively light compared with the all-up weight of 710,000lb. The stability in pitch was excellent, and the elevator control was positive and, from what I remember of it, slightly lighter than the 707. Certainly the trim change with speed was negligible around the 200kt mark and only occasional use of the stabiliser trim switches on the control column was needed as the speed was reduced.

The rudder was heavy and the directional stability was good, necessitating a slight touch of rudder as one entered the turn. The rudder circuit has a g programme; this limits the rudder deflection which one can achieve under extreme conditions, thus protecting the integrity of the structure. The aileron break-out force was around 8lb and, as the control loads were also light, one needed to guard against a tendency to over-bank at first. With four surfaces working for me in the flap-down configuration, the rate of roll was brisk. In the clean-cruise configuration the outboard ailerons are faired. For large roll demands the spoilers also come out on the down-going wing.

The spiral stability appeared to be about neutral and the self-centring was good. The aileron control is so light and yet so effective that if one gives the spectacles a twitch a shiver goes through the aeroplane starting at the wingtips, passing through the centre section and running up the captain’s spine.

From the outset I found no difficulty in flying heading and height reasonably accurately, though I don’t get along too well with the American habit of flying the QNH. Roswell is 3,669ft above sea level, and so the circuit height worked out at around 5,200ft.

It was decided to carry out an auto-approach and automatic landing in the first instance, not only to see the performance of the autopilot but also to give me a chance of assessing the approach and the height for the landing flare.

At 60ft radio height, with the multi-position rotary switch on the mode selector panel put to “land,” the “flare” and “go-around” windows on the annunciator panels turned green, the throttles closed smoothly and the aircraft described a fair curve of pursuit until touchdown. This was good, with, I estimate, a rate of descent of no more than 3ft/sec just before touching.

I tripped out the autopilot with the switch on the control column, and tripped it again to mute the audio warning, at the same time lowering the nosewheel on carefully. As we had agreed to carry out a touch and go, Captain Monan dealt with the flap retraction while I opened up the throttles to the central position, waited for the engines to spin up and then pushed them through to achieve the bugged EPRs of 13.9.

The aircraft ran true down the runway and only slight corrections with the rudder were needed to hold the centreline. As Vr was called I found that the control load involved in rotating was much less than I had expected, and the aircraft came off the ground very readily as the requisite 10° nose-up attitude was reached. The speed then quickly accelerated through V2; further rotation was needed to contain it at V2 +10kt and the power was reduced to keep the rate of climb at around 1,000ft/min.

GO-AROUND DRILL

If one is to abandon an automatic approach, the go-around drill involves disengaging the autopilot and hitting the “palm” switches behind throttles 2 and 3 before opening up to full power. This switches out the localiser and glideslope signals from the zero-reader-type flight director, and substitutes demands for wings-level, 6° nose-up, climb. However no heading information is fed in, and this, to me, seems a shortcoming. If one had an engine failure at this stage, compliance with the demands would give no guarantee that the aircraft was not swinging off heading.

Flying round the circuit once more, I was struck by the fact that the 747 just does not feel like a big aeroplane. Because of the responsiveness of the controls, one is not made conscious of the inertia involved.

Turning on to base leg at 5,200ft QNH, I started a drift down and a turn in to the runway. The speed stability was good, but I found it necessary to make rather large power corrections for relatively small speed adjustments. This caught me out at first, and I became rather high on the approach at one stage. I found the 3° slope rather difficult to judge initially without the aid of VASIs, and I fell back on the ILS glidepath for guidance. The slot was regained by about 400ft and after the threshold was crossed the captain called out: “30ft”; this was by reference to the radio altimeter, which is corrected to wheel height and is the cue for the commencement of the flare.

I pulled off the power progressively and found that I could flare the aircraft comfortably with one hand. The rate of sink over the threshold had been a little high, and it was necessary to quicken the rotation a little at the end in order to rumble it on.

Another touch-and-go was initiated, but this time number 4 engine was cut at V1. The swing was progressive and smooth as the engine spooled down, but the effective rudder control and inherent stability in yaw made it easy to keep straight. The aircraft came off a little less readily this time, but then considerable further rotation was required to contain the speed at V2. In fact a nose-up attitude of 16° was achieved, and all visual references were lost, but on reverting to instruments it was found that a straight climb-out was being made.

On the next approach much coarser power changes were made to control the speed, and in the absence of glide-path guidance the aircraft was tracked towards the approach aiming point at a rate of descent of some 600-700ft/min. This resulted in a more stable approach, and after the 30ft point was reached a smooth continuous flare was made. Surprisingly it didn’t seem difficult to assess the correct height, and the smooth decay of lift over the wing plus, I suspect, considerable ground-cushion effect resulted in the aircraft touching down gently and without fuss.

GROUND HANDLING

The nose was lowered on with a slight forward movement of the control column, the spoilers selected out and reverse thrust applied without its causing the control column to thrash. The brake pedals had a reasonable breakout force, but with some feedback it was not difficult to brake smoothly.

The rudders, which steer the nosewheel over a 15° range, were used with restraint. It was only when we were back on the ground that it became apparent that this was a large aeroplane, for, being nearly 30ft up, it is difficult to appreciate one’s groundspeed. I was about to take the first turnoff when a glance at the ASI showed a reading of 100kt. The aircraft was allowed to roll on to the end, by which time I had brought it almost to a stop – or so I thought, until it was pointed out that the INS equipment showed a groundspeed of 30kt.

A tiller with a handle on the end is used for the nosewheel steering control, and, with its positive action, smooth turning control is easy to achieve, though one must aim to overshoot a turning point with the flightdeck if the main gear is not to cut the corner. With the throttles at idle, continuous brake application was needed to contain the speed to a reasonable figure and reference was made to the inertial navigation system for this purpose.

Continuous turns into a parking position have to be made with considerable care, for speeds in excess of 12kt result in serious scrubbing of the tyres; but if the speed is allowed to fall below 5kt then the machine tends to settle in and considerable power is needed to get it under way again. Once again I became conscious of the size in relation to the people down on the tarmac, and even with the good cutoff angle from the flightdeck the ground can only be seen more than 80ft from the nose.

I do feel that Boeing must be congratulated on building a remarkable aeroplane. It’s not just a bit bigger than existing types, it’s a whole order bigger – and yet it has the most delightful handling characteristics a pilot could wish for.

To achieve a Vmcg (ground minimum control speed) of 103kt with an outboard engine stopped, and a two-engined Vmcg of 138kt when those very large engines are set well out on the wing, says a very great deal for the directional stability and control in yaw. The multi-leg undercarriage obviates rolling problems on the ground, and the soft tread, coupled with aerodynamically efficient wing and positive pitch control, makes landing the 747 very easy despite its size.

However, I think that pilots will have to watch their step when flying into fields with inadequate aids, particularly in conditions of strong wind shear. A jumbo VASI is a very real requirement, and I don’t really see that Cat 2 weather minima can be tackled manually. For decision heights below 200ft automatic landing would seem to be the safe answer.