ANALYSIS: Flight test of the Boeing 787 Dreamliner

This story is sourced from Flight International
Subscribe today »
<="" p="">

Two years ago, Flight International was offered a preview of the Boeing 787’s flying qualities when Mike Gerzanics flew and reported on CAE’s full-flight 787 simulator. The day before my evaluation flight of the aircraft itself, I spent time in one of Boeing’s simulators, which would prepare me well for the real thing. The preview Dreamliner was flight-test aircraft ZA005 (N787FT), configured with General Electric GEnx-1B engines. The flightdeck was production representative, while the passenger cabin was configured for flight-test operations.

Aviation buffs the world over would agree that if it looks right it will probably fly right, and as I walked to the preview aircraft I was once again reminded of how pleasing the 787 is on the eye. The gentle sweep of the wings, the smooth composite surfaces and the unique noise-reducing engine nacelles paint the picture of a bird yearning to fly. Aesthetics aside, what makes the Dreamliner’s design so intriguing is how Boeing has balanced the needs of its three main constituencies: the paying passenger, the purchasing airline, and the operations and maintenance crew.

My experience in the Joint Strike Fighter programme taught me balancing requirements is the chief engineer’s main task. Boeing has successfully balanced these requirements in the Dreamliner by retaining sufficient commonality with its 777 line to keep pilots and mechanics happy, and by using composites and alternative aircraft systems to cut manufacturing and operating costs and improve dispatch reliability.

While changes affecting pilots might be considered evolutionary, systems design considerations are revolutionary, reducing the need for engine bleed air and hydraulic systems power by relying more on electrical power and the large-scale use of composites in the structure.

There are many reasons behind Boeing’s composite design, including significantly less touch labour in manufacturing and an overall lower gross weight, with corresponding cost savings. Another justification for a composite design was to improve passenger comfort. Boeing conducted a vigorous analysis early in the design process to determine exactly what causes passenger fatigue. Cabin altitude and humidity play a part, but even more important is the off-gassing of volatile organic compounds. We may enjoy the “new car” smell associated with these VOCs, but they are subtly nauseating. They even emanate from the luggage and personal affects brought onboard by passengers.


In addition to a cabin filter, the Dreamliner has a gaseous filtration system to further purify the air. The larger window design permitted by the composite structure creates a calming effect for passengers as it provides more horizon references to reduce flying stress. The smooth composite fuselage skin reduces cabin noise – a subtle stressor.

The fuselage shape provides greater room at the cabin edges. For this 193cm (6ft 4in) pilot, the increased cabin height gave an airy feel to the cabin. This extra volume also extended to the cockpit, where I could easily walk in and sit at the controls, without being a contortionist.

Other revolutionary design considerations are not so apparent to the passenger but become a major consideration to the operating airline and mechanics who keep it flying.

Electrical power production is much more efficient than conventional engine-driven accessories and the Dreamliner’s electrical system has six generators: two 250kVA ones on each engine and two 225kVA units on the auxiliary power unit. The engine generators combined are starter/generators.

One primary benefactor of this electrical vice hydraulic power is the electrically driven brake system. Brake-system health is continuously monitored and alerts are set as needed if malfunctions or failures occur.

Brake-wear measurements are periodically checked after gear extension and this information is communicated to maintenance for tracking and replacement planning. In terms of dispatch reliability, the aircraft can have up to 25% of the brakes inoperative on each side. During slow-speed braking, only half of the brakes are used: two of the four on each side.

This saves on brake life, since carbonfibre brake wear is a function more of the number of brake applications than of braking force. The system rotates which four brakes are working as the pilot performs each brake application. At any time, full braking capability is available if the pilot needs it. From my perspective, the braking effectiveness was adequate and, more importantly, easily controlled, preventing sudden stops which might disturb passengers.

As with the 777, the Boeing 787 has three hydraulic systems but they operate at 345bar (5,000lb/in2) versus the 777’s 207bar. The higher operating system pressure allows for a lighter and more compact system.


The electrical system powers typical pneumatically operated ones such as air-conditioning and pressurisation which, in conjunction with the electrically powered hydraulic system, results in a significant reduction in wasted engine horsepower. For extended operations, the electrical system also provides back-up power and it is protected by electronic circuit breakers, accessible via MFDs. Each circuit breaker status is depicted in an easy-to-understand format which mimics actual physical circuit breakers.

I accompanied Boeing test pilot Mike Bryan as he performed the pre-flight walk-around inspection. The pre-flight was perfunctory and not much different from that of any other transport aircraft.


When our test pilot Mike Gerzanics had his hopes of flying the Dreamliner dashed by a car accident, his long-time colleague Paul Smith volunteered to step into the breach.

Smith and Gerzanics attended US Air Force test pilot school together. While Gerzanics left the USAF to work as a test pilot for United Airlines, Smith continued his service, finishing his air force career as lead government X-35 test pilot and commander of the Joint Strike Fighter joint test force.

He now flies Boeing 757s and 767s for a US-based airline. Additionally, he has developed an aviation consultancy that specialises in pilot-vehicle interface issues.

The preview aircraft was equipped with a trailing cone pitot-static system, reminding me this was not a passenger flight. The brake system was extremely clean – no flammable hydraulic fluid near hot brake stacks. There are no bleed-air holes in the leading-edge devices as the slats are electrically anti-iced. Powerful copper heating elements embedded in the composite make it a fully evaporate system. The only parts of the aircraft de-iced by more conventional bleed air are the engine nacelle inlet lips.

Boeing has been touting the 20% fuel savings the Dreamliner offers over legacy aircraft, and 787 vice-president and chief project engineer Mike Sinnett says about 40% of the fuel-efficiency improvement is down to the engines. The large diameter fan with its wide chord and wildly curved blades lend credence to this claim on a visceral level. Major design effort has also been spent in optimising engine nacelle aerodynamics, attaining laminar flow over most of the nacelle

Entering through door 1 Left, it was immediately obvious we were on board a test aircraft – instrumentation racks and water tanks filled the cabin area. For the first half of the three-hour flight I was seated at a flight-test station in the cabin, with a great view of cockpit repeaters, while also being able to monitor cockpit communications.

I looked on as another preview pilot flew a similar flight profile. I could see the flightdeck displays are well arranged, with the primary flight display in a prominent position. The engine display is unobtrusive and can be moved from one pilot’s display to the other.


A benefit to sitting in the back while another pilot flies a similar profile is gaining the perspective a passenger has during manoeuvring flight. It is evident when the flight-control system’s “normal” mode is engaged or disengaged. The system provides pitch compensation by utilising control surface commands to minimise pitch responses to thrust changes, configuration changes – gear, flap, speedbrake – turbulence, and turns up to 30° of bank.

The system essentially smoothes pilot inputs and effectively washes out the adverse yaw, with the resulting kick one occasionally feels when a pilot is a little more aggressive on the controls.

We encountered no bumpy air, but the Dreamliner’s gust-suppression system is designed to alleviate those uncomfortable ups and downs a passenger might feel. It utilises symmetrical deflection of the flaperons and elevators to alleviate gust acceleration. This function is active only with the autopilot engaged in “altitude hold” or VNAV level flight modes. Lateral gust suppression improves ride quality and can reduce pilot workload on approach by automatic application of discrete yaw commands in response to lateral gusts and turbulence. Operation of vertical and lateral gust suppression does not result in control yoke or rudder-pedal movement.

After watching the full stop landing on runway 32R at Moses Lake’s Grant County International airport from the cabin, I took my place in the cockpit’s left seat while Bryan configured the flight-management system (FMS) for our take-off.

We used the onboard performance tool to calculate take-off data, information which must be verified by both pilots before loading into the FMS. Being a test aircraft, we were able to read the centre of gravity immediately off the instrumentation display mounted on the top of the mode-control panel.


The long taxi back to the approach end of runway 32R provided ample opportunity to evaluate the nose-wheel steering. At our light weight, breakaway thrust was minimal and idle thrust resulted in a slight acceleration throughout the taxi. The tiller and rudder steering are well integrated and I was able to gently nudge the Dreamliner around the field.

Toe braking was intuitive. The head-up display (HUD) shows groundspeed, allowing me to manage deceleration rates and minimise disturbance in the cabin. Dialling down the range scale on the horizontal situation display (HSD) gives a detailed airfield schematic, greatly improving situational awareness on an unfamiliar airport. This feature should help prevent runway incursions, especially in low-visibility conditions. Bryan tells me Boeing is studying future improvements to the HSD and HUD to enhance safety regarding runway incursions and excursions.

Holding short of the runway, Bryan reminded me we were going to carry out a simulated engine failure by pulling the right thrust lever back at V1. After ensuring final checks were accomplished using the electronic checklist, I pulled on to the runway and began the rolling take-off.

Of particular note is the ground-roll guidance provided by the HUD for low-visibility take-offs. The cues give the pilot steering guidance when transitioning from visual to instrument conditions. This was extremely useful during the V1 cut. Bryan pulled the right engine to idle shortly after reaching V1.

While forewarned, the subsequent yaw was surprising. We purposefully left the moderate yawing motion in the flight-control logic to cue the pilot as to which engine had failed. A small rudder input stopped the drift and, once airborne, the flight-control system fully kicked-in to zero-out yawing forces.

With the gear retracted we continued the climb to level-off altitude. During the entire V1 cut, the aircraft was stable. The back-driven rudder pedals assured me the aircraft was making the right corrections and helped me smooth my bank-angle rates in a surreptitious way. When I flew the same take-off in the simulator, I asked the operator to dial in a 15kt (28km/h) right crosswind and fail the right engine. The crosswind effects made the take-off more critical and in the simulator, the take-off was remarkably easy to fly. I was unable to dial in the wind for the actual take-off at Moses Lake, but I did feel confident the simulator modelled the single-engine effects well.

Once safely airborne, Bryan advanced power on the right engine to terminate the exercise. At this point, with only a few minutes under my belt in the left seat, I felt at home.

The Dreamliner shares a common type rating with the 777 – Boeing proposes only a five-day differences course for 777 pilots. While I have no time in the 777, there are many similarities between the 757/767, which I currently fly, and the 787.

We left the landing gear extended to cool the wheels and brakes for the pattern work. I looked on as Bryan accomplished all the FMS work, such as loading approaches, setting speeds and frequencies. After completing the landing checklist, the item “speed brake” was annunciated, as we left them unarmed for the touch and go. Bryan acknowledged and overrode it to close out the checklist.

Levelling off on downwind at 3,000ft (900m), Bryan set me up for an instrument approach. In the legacy 767, a non-precision approach must be hand-built, in a fashion, to provide vertical guidance to stay on a constant descent path.

Levelling off at an intermediate altitude, a technique called “dive and drive” is no longer an appropriate way to fly a non-precision approach. The autopilot flight director system (AFDS) in the integrated approach navigation (IAN) mode takes available guidance and builds an approach with a common format.

The Dreamliner’s AFDS IAN mode allows the use of consistent procedures for all types of instrument approaches and provides vertical and lateral guidance. To the pilot, all approaches look like a precision approach. The IAN mode does not support automatic landings, however, and the pilot must disengage the autopilot and complete the landing manually.

I liked the vertical situation display on the lower portion of the HSD, which presented location on the vertical path and pointed to slow and configure in order to fly an efficient constant descent approach. Turning on to final fully configured and on the vertical path for the approach, I easily kept the aircraft on both the lateral and vertical path using the flightpath marker in the HUD.

The approach’s glidepath angle was referenced on the same display allowing me to visually ensure I was on the correct glidepath. Lining up the flightpath marker with the end of the runway kept me on glidepath.

More importantly, had I been in the weather and broken out at decision height, seeing the flightpath marker on the landing zone and at the right glidepath angle would have immediately confirmed I was in a safe position to continue the approach to landing.

Although the HUD is not certificated as a primary flight display, the large display provides excellent situational awareness while allowing the pilot to keep his eyes outside where the real threats are. Speed cues on the airspeed tape made airspeed control a snap. Particularly useful were the flight-mode annunciators showing the commanded flight modes in the HUD.

The next pattern was a planned single-engine approach to a single-engine missed approach. On downwind, with the right engine pulled back, the P-beta feature of the flight-control system once again seamlessly manipulated the rudder to keep co-ordinated flight, in spite of asymmetric thrust levels. This inherent capability of the flight-control system significantly reduces pilot workload. Additionally, the 787’s autothrottle can be engaged on the good engine alone, allowing for automated speed management even during a single-engine approach.


Cumulative time spent with the right engine pulled back resulted in a slight fuel imbalance. While on final, Bryan pulled up the fuel synoptic to confirm the imbalance. Pushing a button on the overhead enabled the auto fuel-balancing feature – no more putting the checklist between the thrust levers to remind me I was balancing fuel. Once balanced, the system repositioned the valves and secured the fuel-balancing procedure.

Throughout the single-engine approach the aircraft demonstrated well-behaved flying qualities, even when turning into the simulated dead engine. I could feel the back-driven rudder pedals comfortably reminding me the system was performing as desired.

Reaching minima, I pressed the thrust lever-mounted TOGA switch to initiate a go-around. The autothrottle smoothly increased power on the good engine, while the P-beta scheme of the flight-control system zeroed out the yawing moment. With positive climb rate, I asked Bryan to raise the gear and we retracted the flaps on schedule. The weather had closed in so Bryan co-ordinated with air traffic control to fly towards the Tatoosh waypoint.

Shortly after reaching FL200, we broke out into clear air. After obtaining a block of airspace, I set up for a couple of turns in the flight-control system’s “direct” mode. Control doub-lets in all three axes elicited an aircraft response similar to one I would expect in the 767 or 757. I could feel the tail kick out during the rudder doublet as well as during the bank doublet. With “normal” flight-control mode selected, aircraft response to doublets was not degraded but the associated tail kick was gone.

Next came a demonstration of the Dreamliner’s flight envelope protection features. Boeing incorporates active protections in all three axes as well as overspeed and stall protection. In the pitch axis, the Boeing 787 uses a scheme called C*u, a blended g rate and pitch rate command system. At high speeds, the control column commands a g rate, with neutral being 1g. At slow speed, the control column commands a desired pitch rate. For lateral-directional (roll-yaw) control, movement of the yoke and rudder pedals commands proportional displacement of the ailerons – plus roll spoilers – and rudder to achieve the desired roll rate and sideslip angle.

Control surface deflection is a function of aircraft speed and yoke/rudder input magnitude. On the ground with speed below 60kt, the rudder-pedal movement commands rudder deflection. Above 60kt, rudder-pedal movement commands a yaw rate.

Once airborne, rudder-pedal movement commands angle of sideslip. In a conventional-control aircraft, sideslip will generate a roll in the same direction as the pedal input: left rudder will cause the left wing to drop. Pilots expect some roll caused by yaw and the 787 exhibits this characteristic while airborne.


At bank angles less than 35°, the aircraft will maintain the desired bank angle and pitch attitude with no additional aft yoke input required. If the pilot commands a larger bank angle and then releases the yoke, the aircraft will return to – and hold – an angle of bank (AoB) of less than 30°. To evaluate this feature, I established a 30° AoB and continued to overbank until reaching 45°, which activated an aural warning. When I released the yoke, the aircraft corrected back to about 25°.

Design of the Dreamliner’s flight-control laws in the lateral-directional (roll and yaw) axes significantly enhances flight safety. At bank angles of 35° AoB or less, it demonstrates neutral spiral stability.

That is, when roll inputs are released, the aircraft maintains the current bank angle. Above 35° AoB, releasing roll inputs causes the Dreamliner to roll to, and maintain less than, 30° AoB. This feature prevents the pilot from inadvertently entering a descending spiral after rolling into a bank.

More impressive is the system’s ability to provide this stability even with the adverse yaw caused by an inoperative – simulated in this instance – engine. As demonstrated during my flight, the features designed for the flight-control system will reduce pilot workload and chances for error during a demanding engine-out event.

In the simulator, I flew the Dreamliner to its speed extremes to test the overspeed and stall protection features. However, in the aircraft I was a little more conservative in my approach.

At maximum velocity, the trim reference speed is limited by inhibiting trim in the nose-down direction. To overcome this, the pilot must apply twice the normal forward column pressure.

Likewise, in a slow-speed condition the pitch-trim function is inhibited in the nose-up direction. The pilot must apply continuous aft column pressure at twice the normal force to maintain airspeed below the minimum -manoeuvring speed.

The autothrottle supports stall protection if armed but disconnected. If speed decreases to near stick-shaker activation, the autothrottle connects in “speed” mode and advances thrust to maintain minimum manoeuvring speed – approximately the top of the amber band – or the speed set in the mode-control-panel speed window, whichever is greater. Another good safety feature, especially at low speeds, is the automatic retraction of the speedbrakes when a thrust lever is advanced to about three quarters of the way to “firewall”.

The icing on the cake to this flight was the opportunity to fly a typical airline descent and approach back at Boeing Field – giving me the opportunity to see how the aircraft behaved in normal service.

Having flown the ILS runway 13R approach in a 767 many times during my airline career, I felt familiar with the arrival. The FMS provided a good constant descent approach vertical navigation plan, which was disturbed by air traffic control’s command to fly heading 100° for spacing and then to maintain 170kt to the marker.

V-speeds presented in the HUD allowed me to stay “head’s out” and search for the numerous traffic calls from air traffic control as I sequenced the flaps for landing. Navigating off lateral path did not prevent me from flying an efficient approach as the displays provided excellent situational awareness and energy state planning tools.

At TOGAE – the final approach fix – we dropped flaps 30 and quickly slowed to final approach speed. Fortunately, I was easily able to stay on vertical path, but had I been high the Boeing 787’s auto-drag function would have provided essentially what US Navy Lockheed Martin F-35 pilots will use to land on a carrier – direct lift control.

I observed the auto-drag feature provide direct lift control during the simulation sortie I flew. In this condition, the ailerons are deflected upward and the two most outboard spoilers are raised with no corresponding yoke deflection.

Auto-drag is only active in a landing configuration with flaps 25 or 30 and the thrust levers at idle, effectively increasing descent rate while maintaining the desired reference approach speed. These control deflections are gradually washed out below 500ft above ground level so flare and touchdown are not adversely affected.


The HUD gave me well-tuned directional guidance, which naturally dampened any excessive inputs from my hands. During the latter part of the approach to flare and landing, the flight-control system demonstrated its outstanding ability to alleviate excess motions during the slightly bumpy final approach.

Use of HUD symbology, including radar altitude and listening to the altitude call-outs, allowed me to smoothly round out for the flare manoeuvre.

I placed the flightpath marker at the end of the runway and the aircraft smoothly touched down. Workload during the approach was low, even given the busy traffic around Boeing Field and the need to interact with heavy traffic flow into Seattle-Tacoma International airport.

The large HSD put myself and Bryan in the loop with respect to the numerous traffic calls as well as the approach’s lateral and vertical paths. The key to a successful approach and landing is being stabilised at 1,000ft, and every system on the Dreamliner worked to make my final landing a smooth one.

Many decisions factor into fielding a safe and cost-efficient aircraft, and a manufacturer must balance the requirements of the owners, travelling public and operators. My flight experience in the 787 demonstrated to me the commitment Boeing has made to satisfy those three main constituencies.

Obviously, the composite design and more electric system architecture will benefit the airlines from in-service rate and maintenance cost standpoints. Likewise, the improved comfort inherent in the Dreamliner’s larger, quieter and cleaner cabin will cause passengers to select it over legacy aircraft when given a choice.

At the business end, up front in the cockpit, the improved fly-by-wire flight-control system, larger, better-designed displays, and more intuitive systems management promote improved situational awareness and crew resource management. Pilots transitioning to the Dreamliner will no doubt quickly embrace the new technology to decrease workload and increase safety.

The Boeing 787 Dreamliner should offer improved dispatch reliability and operating cost performance, while enhancing passenger experience and the airline’s bottom line. However, perhaps a more important observation – at least from this pilot’s perspective – is that the Dreamliner is truly a dream to fly.