We belted into the cockpit of serial number 6013 on an afternoon in late January with G650 project test pilot Jake Howard in the right seat and senior experimental test pilot Tom Horne in the jump seat as safety pilot. The aircraft's BOW was 54,372 lb., giving it a potential 1,428-lb. full fuel payload. Gulfstream quoted the BOW at 54,000 lb. for BCA's May 2012 Purchase Planning Handbook. Early G650 operators say their aircraft actually weigh between 54,400 lb. and 54,922 lb., chock full of optional equipment and fully provisioned for long transoceanic missions with multiple meals. Thus, they only can carry four to seven passengers with full fuel. Each additional passenger, however, only costs about 35 nm of range.

Many of the G650's operating protocols, systems procedures and flow patterns are carried over from the G550. Gulfstream had hoped G550-qualified pilots would be able to use a common type rating for flying the G650. However, the FAA and EASA nixed that plan because the G650 has much less in common with the G550 than initially meets the eye, such as FBW flight controls, the secondary power distribution system and standby multifunction controllers (SMCs), among other significant changes.

Most checklists for the G650 are completed using “flow and verify” protocols, splitting responsibilities between left- and right-seat pilots shortly after starting the APU. The left seater flows the overhead panel, including running the systems tests, followed by sweeping from the left side panel, control yoke, left SMC, flight guidance panel and to the center console. The right seater only has to check the onside SMC, control yoke and right-side panel items and equipment.

As noted, the cockpit has much improved outward visibility because of its considerably larger windows. In addition, the EVS camera has been moved up close to the base of the windshield center post, thereby reducing parallax errors when using the EVS HUD imagery as an aid for taxiing at night or in low visibility conditions.

Fuel on board for our flight was only 15,600 lb., but it was enough to fly from Savannah to San Diego or Gander at Mach 0.85 and land with NBAA IFR reserves. Horne computed the ramp weight with safety pilot and other equipment at 70,022 lb., or about 70% of maximum ramp weight. Savannah's field elevation is 50 ft. OAT was 25C. Computed takeoff speeds were 108 KIAS for V1, 109 KIAS for rotation and 126 for the V2 one engine inoperative takeoff safety speed. En route OEI climb speed was 147 KIAS. TOFL was 3,400 ft.

Engine start was very similar to that with the G550. Switch on the boost pumps, turn on the start master and press a start button. With oil pressure and indication of fan rpm, turn on the fuel cock. The FADEC handles the rest. We noted that the air cycle machine packs automatically shut down during engine start to assure sufficient bleed air from the APU to turn the twin Rolls-Royce BR725s' air turbine starters. After start, the ACMs automatically come back on line.

After we completed the post-start checks, we rolled out of the chocks with very little thrust. Indeed, we needed to extend a thrust reverser from time to time to control taxi speed without riding the brakes. Howard pointed out that turning off the engine bleed air and using the APU for the packs causes a drop in engine ground idle rpm, thus reducing idle thrust.

Holding short of Runway 19, Howard ran through the pre-takeoff checklist and checked the flight controls for proper movement on the MFD flight controls page. Taxiing on to the runway, Howard armed the ground spoilers before we advanced the throttles, thereby causing them to pop up. Only by keeping one of the throttles slightly advanced above the idle stop could we make the ground spoilers retract. This functionality is carried over from legacy large-cabin Gulfstreams and seemed somewhat antiquated. We'd prefer to see the system updated with ground speed sensing so that the spoilers would pop up until the aircraft reaches a reasonable initial takeoff roll speed.

With a takeoff weight of about 69,600 lb. and 37,800 lb. of thrust, acceleration was sporty, even by Gulfstream standards. Shortly after rotation, the aircraft left the runway in about 3,000 ft. Control response was crisp and the aircraft was well damped in pitch, no doubt due in large part to the 36.6-ft. span, 439-sq.-ft. horizontal stabilizer that provides considerably more pitch control authority compared to those on the G450 or G550. But the high-level FBW control laws surely played a significant role as well in the G650's well-mannered behavior.

The aircraft also had pleasant artificial roll control feel and good roll response with adequate control yoke centering, but very little on-center breakout force. Engineers with Gulfstream and Rockwell Collins, which supplied the control yokes and rudder pedals, worked together closely to fine tune artificial feel and control response. Quite candidly, your fingertips might tell you this is a $200 million FBW Boeing even though the data plate says it's a Gulfstream.

The aircraft exhibited excellent short-period stability in all three axes, but it's difficult to tell how much was contributed by natural aerodynamics versus high-level FBW control laws. On the way up to initial cruise altitude, we had a couple of intermediate level-offs required by ATC and comparatively sharp turns. Yet, using a 250 KIAS/260 KIAS/Mach 0.85 climb schedule in mostly ISA conditions, the aircraft leveled off at FL 470 in 23 min. At ISA-7C, it cruised at Mach 0.85 or 480 KTAS on 2,400 pph at a weight of 67,500 lb.

Then we pushed up the throttles because high-speed cruise is the G650's forte. The 67,400-lb. aircraft smartly accelerated to Mach 0.90, resulting in 506 KTAS on 3,000 pph in ISA-7C conditions. There is only a 1.5-2.0 dB increase in cabin noise when cruising at Mach 0.90 instead of Mach 0.85 long-range cruise, Gulfstream claims. Horne also noted that the cabin altitude only was 4,300 ft.

We couldn't perform our usual long-period (phugoid) pitch stability check because the FBW system masks the natural, high-Mach aerodynamic stability characteristics of the aircraft. But we did check Mach buffet margins. A wind-up turn indicated the aircraft has robust high-speed buffet margins, albeit at a comparatively light weight. We didn't encounter buffet until about 1.4 g at Mach 0.88 in a 45-deg. turn.

Then we descended to FL 430, using idle thrust and speed brakes. We noticed only mild buffeting and a slight pitch change when the speed brakes were extended. Once level at that flight level, we again pushed up the thrust to check how fast the aircraft would cruise. Horne reported the cabin altitude was 3,380 ft.

As for speed, the G650 did not disappoint. At max cruise thrust, it accelerated to its Mach 0.925 redline at a weight of 67,000 lb. while burning 4,100 pph. Some operators doubtlessly will dash across North America or the North Atlantic at 530 KTAS or faster depending upon OAT. But BCA estimates that maximum range will be cut to 5,000 nm or less at that speed.

Down at 15,000 ft., we flew a series of standard air work maneuvers. Steep turns are easy to fly. The HUD's flight path vector (FPV) and velocity trend vector provide precision guidance cues. Stick force is moderately heavy, but that's appropriate for this class of aircraft. We also flew clean and dirty stalls, at least to the maximum AOA permitted by the FBW control laws.

We used a clean configuration for the first stall. At a weight of 66,800 lb., we trimmed the aircraft for a 156 KIAS Vref speed or 0.67 normalized AOA, reduced thrust and decelerated. “Normalized” means that 1.0 AOA is the maximum lift coefficient adjusted for high-lift configuration and local Mach number because of its influence on buffet and stall. At 15,000 ft., though, the effect of local Mach number on the wing is insignificant.

The pitch limit indicator appeared on the PFD and HUD at 0.75% maximum normalized AOA. During the approach to clean stall, the stall-warning stick shaker fired at 129 KIAS or 0.94 normalized AOA. At 0.97 AOA, the FBW system limited elevator and horizontal stabilizer pitch control authority to prevent untoward handling characteristics. Holding the control wheel full aft, the nose thus gently pitched down and we initiated recovery.

The dirty stall, with gear down and flaps extended to the full 39 deg., was equally non-dramatic. We trimmed for 122 KIAS or 0.67 AOA, commenced a normal glidepath-like descent and then leveled off without adding thrust, thus allowing the aircraft to decelerate. After the stick shaker fired, we continued to pull aft on the yoke until reaching the stops. At 0.98 normalized AOA, the nose gently dropped and we initiated recovery with only a slight loss of altitude.

The two maneuvers quite clearly demonstrated the G650's improvement in high AOA behavior compared to previous generations of large-cabin Gulfstreams. If both stick shaker and stick pusher are ignored in some of the legacy Gulfstream models, positive stick force gradient can be neutralized or even reversed. Stall recovery thus becomes much more challenging and there can be a substantial loss of altitude.

Returning to Savannah, we prepared for a WAAS LPV precision approach to Runway 19. Horne computed Vref at 120 KIAS for a 65,500-lb. landing weight and a non-factored landing distance at 2,873 ft. based upon 13-kt. headwinds. It's apparent that the G650's ref speeds at typical landing weights will be similar to those of the G550, even though it is a heavier airplane with more wing sweep.

We bugged the target airspeed at 125 KIAS and let the auto throttles maintain speed in gusting wind conditions. The G650's big airfoil with low wing loading doesn't provide as smooth a ride in turbulence as the more highly loaded airfoils of competitors' aircraft. But it does enable the aircraft to cruise higher where the air generally is smoother for most of the mission.

The HUD's azimuth and glidepath guidance cues, along with the FPV marker, made it easy to hand-fly the approach. We noted that the airport database used by the HUD needs a little updating. The synthetic runway outline displayed was 2 deg. right of the actual pavement borders. In addition, runway touchdown elevation and glideslope/glidepath data must be manually entered through the standby multifunction controller because the HUD isn't fully integrated with the FMS airport database.

The FBW system transitions from high level control law to direct law for takeoff and landing, so the G650's smooth handling behavior during approach reflects its aerodynamic refinement. At 50 ft., we pulled back the thrust to idle and continued to use the HUD until touchdown. The aircraft appears to float less than the G550, but touchdown behavior was nonetheless very smooth. We deployed the thrust reversers, but kept the throttles at idle, thereby allowing the aircraft to decelerate leisurely. A light touch of the brakes and we turned off at Taxiway B1 after a touchdown roll of about 5,200 ft.

Our standard profile calls next for a simulated one-engine inoperative takeoff and landing. But we previously performed those maneuvers in the G650 simulator at FlightSafety International's Savannah training center. This both reduced risk and added the realism of a complete engine failure rather than by reducing thrust to idle in the actual aircraft. Rudder pedal forces on the OEI takeoff were moderate and the aircraft was easy to control. For landing, though, we couldn't use the auto throttle because the system only works if both engines are operating. Managing the asymmetric thrust, however, wasn't much of a challenge. We appreciated the addition of the automatic rudder center function because it eliminates the need to retrim the rudder to neutral or hold rudder pressure during the flare.

The second takeoff in the actual aircraft on Runway 19 thus was a normal all-engine maneuver to a downwind VFR pattern. Downwind, we slowed to 180 KIAS, extended the flaps to 10 deg. and lowered the landing gear. Abeam the approach end, we started a shallow descent and extended the downwind leg. Turning base, we extended flaps to 20 deg. and slowed to 160 KIAS. We turned to a 3-mi. final while extending the flaps to 39 deg.

We aligned the 3-deg. nose-down pitch mark on the HUD with the runway touchdown zone and used the FPV to maintain desired trajectory. The technique worked well, but we flared with a touch too much speed causing a little float. Make a note. Aircraft with no leading edge devices tend to float more than those with slats if you carry excess speed.

We again turned off after a 5,000-ft. landing roll and taxied back to the Gulfstream ramp. Pulling into the chocks, Howard turned on the belly videocam and displayed the image on the MFD. This enabled us to track the parking alignment line within a couple of inches of centerline, shutting down the engines after the 1.7-hr. demo mission.

Conclusions? The G650 is the nicest flying large-cabin Gulfstream yet built. The fly-by-wire functionality is all but transparent unless probing the extremes of the flight envelope. Pilots might not know it's an FBW aircraft without being told. PlaneView II, the HUD and EVS, among advanced cockpit features provide unsurpassed situational awareness in the cockpit. The cabin environment, including increased volume, window size and pressurization, along with the redundancy and reliability of Gulfstream's Cabin Essential CMS, make it the most commodious and functional business aircraft yet built by the Savannah firm.