It seemed remarkable to be flying the 777 a mere year after it was first unveiled, but such has been the pace of the programme from the start. Flight test hours have grown at twice those for previous models, in a schedule of certificating three engine types and early qualification for 180min extended-range twin-engine operations (ETOPS).
Enhanced computer-aided-design techniques allowed early definition of the aircraft, with wide participation of invited airlines and pilots. I flew the high-fidelity-engineering simulator cab nearly three years ago. Now, under Captain John Cashman, 777 chief pilot, I was to be amazed by how close the aircraft itself is to that early standard.
Aircraft N-7771 had a balanced look, sitting above its reflection in the water-covered parking area on a gloomy day at Boeing Field. The engines did not look so large from a distance, yet its Pratt and Whitney PW4077s, of 342kN (77,000lb) thrust, have a fan diameter of 2.845m. The fan of the Rolls-Royce Trent 800 series is slightly smaller, but that of the General Electric GE90-85B is 280mm larger.
The fuselage is 1.15m wider than the Boeing 767's - only 260mm narrower than that of the Boeing 747. The size of the belly area for containers, and loose cargo at the rear, became apparent in the long climb up the mobile steps to the front passenger-door. This, the first of three initial test aircraft, was fitted with test gear and water-ballast tanks. Partial furnishing showed the flexible location of galleys and toilets and the very large baggage bins above the window line.
The cockpit is the most spacious of any airliner yet. The electrically powered seats run very smoothly and rudder-pedal adjustment is in a handy raised position. Large, opening, windows give a wide-angle side view.
The 200mm-square display screens closely resemble the 747-400 electronic flight-instrument system: the primary flight display (PFD) and navigation display (ND) are side- by-side ahead of each pilot, while the engine-indication and crew-alerting system (EICAS) and multi-function display (MFD) are one above the other in the centre. The EICAS screen is normally dedicated to primary engine instruments, with blocks to the right and below for memos, warnings and supplementary-systems data. The MFD selector panel is easy to see and reach, to the right of the glare-shield autopilot-mode control panel (MCP). The centre console houses three flight-management control and display units and rugged integrated-navigation/communications-frequency selector panels. At its rear is a full-width printer, with a bin holding some 100m of paper directly underneath. The well up-sloped overhead-systems panel resembles that of the 767-300ER.
The liquid-crystal-display (LCD) screens in this aircraft were blotchy because they are the early variety. The production-standard screens, which I have seen on the simulator, however, are uniform, crisply clear, and brighten automatically to counter the strongest outside light. Cool light-emitting diodes give the panel lighting a faint green tinge. Back-lighting and LCDs can be adjusted quickly by single controllers.
Automatic fully monitored starting is deceptively simple: select secondary engine data on the MFD, and starter masters and fuel switches to "on". The auxiliary power unit (APU) supplies ample air to accelerate both engines simultaneously to light-off speed. The start was completed in 50s, with exhaust-gas temperature (EGT) peaking at under 400°C - the maximum allowed is 535°C. Airframe and engine anti-icing operates automatically in the air; if it is needed on the ground, engine anti-icing is selected manually. Once all checks are done, the overhead panel goes dark.
Idling parameters were, 20% N1 (fan speed); 56% N2 (high-pressure turbine); EGT (with all air-supply packs) 35°C; and fuel flow 775kg (1,700lb)/h. The departure weight was 202,250kg, with 61,500kg fuel; the centre of gravity was mid-range, at 28%.
The aircraft rolled easily at 25% N1. The brakes are as smooth in taxiing as they have proved to be effective in high speed abort - at weights up to 286,900kg from 185kt (350km/h) - but the long, flexible, fuselage can nod slightly at certain speeds. One pair of the six brake sets is cycled off at each braking, at speeds below 30kt.
Wheelbase and track are approximately 26m and 11m, as on the 747-400. The wing-tips cannot be seen by the pilots, unless they open their side windows and lean far over. Minimum pavement width for a 180° turn is 47.5m. When turning hard in confined areas, a nose clearance of 13m from obstructions is advised; the nose-path radius is 33.5m, but a wing tip will track at over 44m and the tail at 40m. A sense of both the span and of the long tail behind has to be developed.
The steering tiller is the usual lozenge on an arm. As on other large Boeing aircraft, it is turned almost full circle at full nose-wheel angle. The nose wheel is 3.5m behind the pilot and a large overshoot is needed before turning 90° onto a taxiway centre-line. At nose-wheel angles of over 13°, the rear axles of the main bogeys also steer, which assists the tug in pushback, too. The axle angle is shown on the EICAS, with tyre-pressure and brake-temperature data, and a take off warning sounds if the axles fail to lock ahead for take-off.
As the runway was approached, the EICAS showed confirmation that full auto-brake was selected, and a reminder that the APU was following its 2min cool-down. The flight director was armed in lateral and vertical navigation modes; they engage automatically at 50ft (15m) and 400ft. The autopilot can be engaged at 200ft. Engagements of the full-time triplex autopilot are quite free of transient jerks.
With the flaps set at 20°, take-off speeds were V1 (decision speed) 125kt, VR (rotation speed) 129kt, V2 (airborne safety speed) 137kt. At take-off, the power levers are set to 1.05 engine pressure ratio (EPR), to check symmetry in thrust. A mismatch between such high-thrust engines (the higher-thrust PW4084 is rated at 374kN and engines in this class have been run at over 400kN) could cause directional problems. With the auto-throttle armed, a touch on the triggers ahead of the power levers drove the EPRs up to 1.365. With acceleration at 7kt/s, time to rotation was just 18s.
The 777 is Boeing's first aircraft with fly-by-wire (FBW) primary flight controls. The untrimmed upload on the stick, however, increases conventionally with airspeed; a stick-load actuator creates about 1.5kg stick-load per 10kt error, which means that stick-release results in conventional-feeling speed stability. This contrasts with the mechanically simpler Airbus FBW philosophy, which abandons feedback actuators in favour of a spring-loaded side-stick with automatic pitch-trim. When released to central with wings-level, the Airbus side-stick delivers 1g flight, despite airspeed changes.
The well-balanced 777 controls allow the nominal pitch-rotation rate of 2.5°/s to be closely kept up to 15° target attitude; the flight director then held V2+15kt = 152kt. The PFD's speed scale shows successive speeds for flap retraction on acceleration and extension on slowing down. These target-manoeuvring speeds, at simple intervals of 20kt, are close to each minimum-drag/best-gradient speed - at which required thrust and pitch attitude for level flight remain nearly constant.
The route climb (and descent) speed is 310kt/Mach 0.84. EPR rose, from 1.20 to 1.40, during the climb to 35,000ft, to reach under-rated climb thrust at about 28,000ft. (Fixed levels of reduced take-off thrust, or variable assumed-temperature de-rates, are optional, up to a maximum reduction of 25%. Reduced climb thrust de-rates depend on whether take-off de-rate was less or more than 15%). N1 increased, from 85 to 92.5%, and EGTs ranged between 490°C and 470°C; individual fuel flows steadily reduced from 8,600kg/h at 5,000ft to 6,900kg/h at 17,500ft and about 4,500kg/h at level-off.
The climb to 35,000ft took 20min and 4,900kg of fuel. With the 777 now weighing 197,350kg, the optimum altitude in standard conditions would have been 38,000ft. Cruise figures at M0.84 were: EPR 1.125; N1 82%; EGT 370 °C; and fuel flow 3,300kg/h. The climb and cruise speeds of M0.84, instead of the originally promised M 0.83, are caused by lower-than-expected drag. The certificated ceiling is 43,100ft, achievable at end-of-cruise weights of around 175,000kg.
Even with the aircraft trimmed to a reference speed 50kt lower than actual speed, the force on the control column was modest. Speed stability is "relaxed". The phugoid period was 1min 20s, with 20° flap; after diving to 180kt and releasing the controls, with a 150kt trimmed airspeed, minimum speed at the top of the pitch-up was 135kt.
The pitch-compensation function keeps the nose up during turns up to 35° bank, while rudder is fed in automatically. Roll-control-surface angles are simply proportional to the angle of the wheel and wheel force is light up to 35° bank. At larger angles, the roll marker on the PFD attitude display turns yellow and there is a clear increase needed in wheel force and stick back pressure as bank angle increases further.
When the wheel was released at 50° bank, the 777 "flicked" back to below 30¡. The maximum roll rate is high for an aircraft of this size, at about 15°/s. Rapid roll entry causes a brief but clear instant of adverse yaw, with the nose jerking against the direction of the turn. Turns to 60° bank at M0.84 brought up the pitch-limit indicator on the PFD attitude display - with an extra pull on the control column there was a light burbling sound, maybe from airflow over the cockpit rather than the wing.
My basic rule for flying this aircraft manually would be to be light-handed and to keep the aircraft closely in trim. It has the inertia in pitch, which one would expect of a large aeroplane, but none in roll. I would, however, like a small marker on the speed tape to show current trimmed airspeed.
The aircraft has been designed for ETOPS from the outset; for example, the main AC system is served by main-engine and APU generators and also by standby engine-driven variable-speed/frequency generators with converters.
The AlliedSignal 331-500 APU can be started both electrically and, unusually, with bleed air - and can be relit at maximum altitude after a long cold soak. A ram-air turbine, which is unlocked either by main or APU-battery power, drives a generator and hydraulic pump. The pilot has full flight and main-engine instruments, even on battery power.
CONTROL LAWS AND MODES
Direct mode was engaged by a switch on the overhead panel (secondary mode cannot be selected manually). There were now no protections or trim bias for configuration changes. Handling in this mode is conventional, with the aircraft feeling slightly looser in roll and yaw. Dutch Roll is slow to damp and roll response to rudder inputs is high. Switching back to Normal Law was kick-free, apart from the canceling of some residual yaw, which was immediately countered by yaw damping in the PFCs.
In Direct mode, the controls are still driven by analogue electrical signals, rather than by rods, and cables are not generally fitted. There are mechanical connections from the control wheels to operate Nos 4 and 11 spoilers and from the standby pitch-trim handles to the stabiliser. This "mechanical" system, is intended to be used only while electric power is being restored. This is another area in which Boeing differs from Airbus: on Airbus aircraft, the mechanical back-up controls yaw via the rudder, rather than roll via spoilers.
In a 55° banked spiral dive at idle thrust, the speed started to edge up above M0.84; bank angle is restricted to 15° at high speeds and the wheel force becomes much heavier. When I released the controls completely at M0.86, the bank swung rapidly back, nearly to wings-level.
The rate of descent with full speed brake is around 6,000ft/min (30m/s), which should be ample for an emergency descent. A slight delay in the response to a sudden full movement of the speed-brake lever damps out any "jump" at extension. At high speeds, a gentle rumble is heard, rather than felt, when the speed brakes are deployed. At low speeds, it is felt slightly more - a subtle warning of low speed. The speed brakes can be deployed at all flap settings, but Boeing recommends against their use with landing flap. TA warning is given if thrust is increased while the speed brakes are extended. The flaps cannot be extended above their 20,000ft maximum certification altitude, nor above 250kt.
During flight development, slight changes were made to the trailing edges, inboard slat gap and aileron rigging, to improve the stalling characteristics - in early tests, 110° of roll had been reached when stalling in the landing configuration. Other stall protections are typical of modern practice: auto-thrust, when left armed during manual thrust control, "wakes up" as stick-shaker speed (Vss) is neared. The slats also open fully if "Flaps 1" (20°) is set, and the pitch limit indicator appears on the attitude display.
The auto-throttle was now disarmed and thrust set at idle. The autopilot then relaxed its altitude-lock at stick-shaker speed and the flight director disengaged; the aircraft descended at a constant-Vss angle-of-attack (alpha) of 9-10°.
At speeds within the PFD's amber scale (below minimum manoeuvring speed) upward pitch-trim is not available. Control-column load increases to a 7kg/10kt gradient by Vss and the load rises sharply, from 11kg to 41kg at stall speed (Vs) - almost acting as a "pusher". Similar protection is applied above maximum operating speeds.
There is natural buffet at Vss in all configurations; at 158kt clean, with weight at 190,000kg, it was heavy. Recovery from the stall incurred a 1,500ft loss of altitude. Vss with flap at 5° was 122kt and Vs 110kt: there was a 500ft loss during recovery. A stall with 5° of flap in a 20° bank was achieved only with difficulty; the 777 fights its way out of it. The clean stalling speed at maximum take-off weight is 177kt and, at 272,000kg in future B models, it would be 192kt.
In a 20° banked side-slip, with nearly full rudder deflection, close to minimum manoeuvring speed, control-wheel loads lightened very slightly, but the handling was otherwise completely conventional.
CIRCUITS AND LANDINGS
With the weight now at 185,000kg, we headed for the Moses Lake test area. At this weight, reference speed (Vref) was 135kt (Vref is within a knot or two of the V2 take-off speed, which is convenient for circuit work). The side-window size is excellent for low-level circuits and the pitch stability and responsive roll disguise the mass being manoeuvred. The control wheel feel is a compelling reminder if you over-bank in the less-forgiving area close to the ground.
The cloud base at Moses Lake was 1,200ft, with scud below. An instrument-landing system automatic landing was first demonstrated. The automatic-landing flare mode engages at about 50ft and the auto-throttle retards at 25ft, this seemed drawn-out, compared with manual landings.
I made three circuits at 900ft, with touch-and-goes: a standard visual; a single-engined approach with 20° flap; and a deliberate fly-though of the centre-line, with an S-turn back to correct it by 300ft. The first circuit and landing were precise, but my height holding was sloppy when flying manually on one engine. I was low - "in all the reds" - and slow, too, against a target speed of Vref + 6Kt at 400ft on finals, but a late correction was not difficult to make with the high thrust available.
I still find airspeed tapes odd, although the automatic speed-limit markings are invaluable. A difficulty is that, when maintaining Vref closely, there is a lot of data in a small area of the tape. Cashman says that he monitors the roll of each numeral in the digitally displayed speed box alongside. The last landing was stabilised very late, but I did not feel rushed - one quick check to both attitude and power was enough. Having the auto-throttle in manual flight creates no adverse interaction between pitch and thrust.
Alpha in the landing configuration is 4-5¡, so the view down over the nose is good. Eye-height at touchdown, is less than 10m; with 20° flap, on take-off at V2 or in go-around at Vref, it is about 8.5°. There is no procedure for a flapless landing as, with loss of all hydraulics, flaps and slats are transferred to electrical power. A further, slower, switch-controlled alternative system, is limited to 20° flap extension.
The 777 is the best-landing large aircraft that I have experienced, mostly because of the massive ground-effect of a 747-span wing closer to the ground. Other wide-body flare-heights vary between 30-25ft; for the 777 a height of 20-15ft is fine. On raising the nose a couple of degrees, the aircraft just sinks through its own ground-effect to a soft touchdown.
At altitude, I flew down an imaginary 3° glideslope at Vref -5kt, selected full go-around power (two touches on the triggers, instead of one for de-rated power) and pulled back on the stick. Pitch-up is gradual, as the large fans wind up at a normal rate, (full power in just under 8s at sea level and about 10s at altitude). I let it continue, allowing the airspeed to decrease gradually to Vref -15kt before taking action; there was ample time to assess the situation. Minimum control speeds lie between 105kt and 110kt - commendably low with the high-thrust engines. At light take-off weights however, when full power is not needed for runway performance, the use of de-rated power is advised.
The flight back to Boeing Field, at 280kt and 16,000ft, with autopilot engaged, showed the effective gust damping on elevator and rudder. Turbulence penetration speed is 280kt/M0.83. The ND lacks the range rings seen on the 747-400; this made the judgement of relative distances off-track less clear.
Being high and fast on descent, I saw how low the drag is until 20° flap is selected: without flaps, the 777 takes nearly 2min to slow to flap- selection speed at maximum landing weight. Using the speed brake halves these times, but it is naturally less effective below 250kt. The flaps and slats are retracted automatically under load relief if limit speeds are exceeded.
Rather than select the landing gear early, I persisted with a normal flap sequence and settled on the profile once airspeed was below 210kt with 5° flap. Once landing flap was selected, I could quickly reduce speed to Vref + 5kt from the maximum approach speed of Vref + 20kt. Cashman confirmed the last inbound checkpoint on the global-positioning system and descent was started using flightpath-angle mode on the flight director. The ground was visible from 3,000ft, but scud persisted down to 700ft in rain.
On some large, long, aircraft, spoilers are used for "direct-lift control" at full flap. Boeing does not favour this, as pilots continue to pull into the flare, to avoid sudden spoiler extension with forward-stick movement, with a risk of a firm high-angle touchdown. The attitude limit for tail contact on the 777 is 12° with extended oleos and 10° with them compressed and all tyres in ground contact. This is adequate at speeds as low as Vref -10kt, but a long float should be avoided. The larger take-off flap settings provide more aft body clearance - normally 1m - at Vr.
After early trials, the elevator rate was adjusted, to avoid occasional pilot-induced oscillation with 20° flap and spoiler extended, and a 3Hz filter was added to counter in-phase stick/control coupling in the landing configuration. Ground effect is also countered slightly with artificial pitch-down. The handling now seems perfectly matched to the aircraft.
I made a minimum hold-off on the wet runway, still to a passable touchdown and the spoilers deployed automatically. The feel of the reverser arms, moving back over the throttle levers, was smooth: as the engine cowls ran back, the arm-baulks were quickly removed and I selected full reverse thrust. In the cockpit, the sound of reverse thrust is faint, so the prompt deceleration seems effortless.
Cancellation of reverse thrust can be started as late as 60kt in the rollout, to reach idle thrust by 30kt, Cashman says, that intake-debris-ingestion problems have been minimal. While taxiing in, I was reminded that awareness of engine-exhaust flow volume and risk of blast effect when manoeuvring near vehicles and aircraft are important. With the low engine noise and slow acceleration from idle, it is easy to move the power levers too far ahead.
At times the 777 can be almost too agile; it responds best to a light touch on the controls, rather than a grip, taking full advantage of pitch trim. It is sweet in normal operation at scheduled speeds, with pilot-friendly reminders of speed and attitude limits. You always feel in contact with the aeroplane, mainly by virtue of the actuator-produced speed stability. In contrast, the Airbus feels a little more distant. I would not say that the 777 is easier to fly than the Airbus FBW types, but perhaps easier to learn to fly if you come from a conventional type.
Source: Flight International