There is still life left in Bombardier's CRJ family of aircraft, a test flight of the CRJ1000 NextGen reveals. The future may belong to larger aircraft, such as the airframer's CSeries, but the NextGen family of CRJs delivers a comfortable passenger experience - plus fuel savings of up to 12% on a 1,000nm (1,850km) stage length, compared with competitive aircraft.
Bombardier senior engineering test pilot Andy Litavniks was the host for my flights - one in the CRJ1000 and one in the CRJ700 - conducted from Bombardier's Mirabel CRJ production site north of Montreal. Flying the aircraft back to back allowed me to note how the CRJ1000's fly-by-wire rudder compares with the CRJ700's conventional one.
From a pilot's perspective, the NextGen line offers several advantages. While the flightdeck is still a Rockwell Collins Pro Line 4, it incorporates several Pro Line 21 units.
The altitude heading reference system has been upgraded to laser ring gyros from the previous flux valve units. Optional NextGen cockpit items include a Class 2 electronic flight bag, sourced from CMC Electronics, and a Collins head-up guidance system. With proper training and authorisation, the head-up guidance system can allow for reduced visibility take-offs as well as a Category IIIa landing capability.
Unfortunately, neither the EFB nor the head-up system were installed on the test aircraft, which was a production example in Brit Air colours.
With the outside temperature hovering around -15°C (5°F), the pre-flight walk-around inspection was conducted in the warmth of a hangar. The inspection itself was straightforward, and the longer fuselage and larger wing were obvious.
I used the seat alignment balls on the centre pillar to gain a comfortable seating position. With a few minor exceptions, the flightdeck is identical to that of the CRJ900 I had flown earlier.
The six 6 x 7in (15 x 18cm) cathode ray tube displays - leading edge when the Pro Line 4 avionics suite was first deployed - seem merely adequate in today's marketplace. Aircraft systems are primarily controlled by switches and push-button lights on the well-arranged overhead panel.
The flight guidance control panel on the glareshield provided ready access to autopilot and flight-director modes, and the forward instrument panel was arranged in the conventional manner, each pilot having a dedicated primary flight display and multifunction display.
Two flight management system control units are at the head of the centre pedestal. Unlike the CRJ700/900, which has a manual alternate gear-release lever aft of the pedestal, the CRJ1000's alternate gear system is electrically actuated and controlled by a switch next to the gear lever.
Flight's test pilot Mike Gerzanics infront of the CRJ1000
After completion of pre-start flows, both engines were started using bleed air from the tail-mounted Honeywell auxiliary power unit. The automatic starts are triggered by the full-authority digital engine control, the only pilot action being to move the respective throttle from "shut off" to "idle" at 20% N2.
Starter cut-out was at 50% N2, and the peak inter-state turbine temperature of 475°C for both engines was well below the limit of 815°C. Post-start flows were easily accomplished and the flaps were set to position 8 for our take-off. Once the aircraft was rolling, idle thrust was sufficient to move at a comfortable taxi speed. Rudder-pedal-actuated steering gave +/-7° of displacement and was used on the straight portions of the taxi, while the tiller, with +/-70° of displacement, was used for large turns.
During the taxi to Runway 06, I found the steer-by-wire nose wheel steering allowed me to precisely track taxiway centrelines. Mirabel is an uncontrolled field and, after visually clearing final, we lined up on Runway 06. I pushed the throttles to the TOGA detent, N1s settling at 85.9%N1.
On that cold day the aircraft raced down the runway, Litavniks calling "rotate" at an indicated airspeed of 133kt (245km/h). Yoke forces were light as I pulled the aircraft into a 12° nose-high take-off attitude. Had I been overly aggressive during the rotation, an enhanced pitch director mode of the flight director would have commanded a lower rotation rate/pitch attitude in an effort to prevent a tail strike.
Bombardier lists a maximum take-off weight of 40,825kg (90,000lb) for the base line CRJ1000. With 7,785kg of fuel (8,820kg maximum) our take-off weight of 31,300kg was not representative of an actual airline operation. The aircraft cleared the runway after a 1,000m (3,280ft) ground run, about one half of the certificated take-off distance, which includes clearing a 35ft obstacle at maximum take-off weight.
After cleaning up the aircraft, and when passing 1,000ft above the ground, I retarded the throttles to the "climb" detent.
A turn was made to the north, air traffic control clearing Canadair 112 into one of Bombardier's local test areas.
I hand-flew the aircraft until we passed 10,000ft, where a climb speed of 290kt was captured and held by the autopilot. With the throttles in "climb" detent, the full-authority digital engine controls kept the optimum N1 setting. When the aircraft passed 32,000ft, Mach 0.77 was captured and held until we reached our cruise altitude of FL350. Time from brake release to level-off was only about 18min.
Once level, I established M0.78 cruise point, Bombardier's recommended normal cruise speed. I found the airspeed tape's trend arrow allowed me to expeditiously set and hold the desired speed. A total fuel flow of 1,480kg/h was needed to hold M0.78/262kt with a resultant true airspeed of 454kt.
Next, the power was increased and an M0.80 cruise speed was established. Total fuel flow increased to 1,740kg/h with a resultant true airspeed of 473kt.
To a working airline pilot, the prospect of being stuck behind slower traffic is never appealing. Bombardier's CRJs can definitely keep up with traffic - and even outpace aircraft such as the Boeing 737 Classic.
The ambient noise level was reasonably low, on a par with the flightdeck of a 737. For the flight we used David Clark headsets, which further reduced background noise. At altitude I found the flightdeck a pleasant environment to work in. Given the CRJ1000's 1,345nm range, with 100 passengers and IFR reserves, legs with three-plus hours of block time will not be out of the question.
While still at altitude, Litavniks put the flight-control page schematic up on one of the centre multifunction displays.
The rudder schematic showed actual rudder displacement against a lateral scale. Nothing unusual in that, but what was interesting was the two sets of lateral limits displayed. Large hash marks indicated the ultimate displacement limit, while smaller moving ones inside them showed the displacement allowed given the current flight conditions and aircraft configuration. At FL350 and cruise conditions, only about a quarter of maximum rudder displacement was available.
The CRJ1000 is the only CRJ to have a fly-by-wire rudder. While almost all transport aircraft have some sort of rudder travel limiter, what distinguishes the CRJ1000 is how the FBW scheme is implemented. In the CRJ1000, the pilot always has full rudder pedal travel, only the resultant rudder displacement is reduced.
Reducing rudder pedal travel is another method of limiting rudder displacement. Reduced rudder pedal travel has, however, been cited as a contributing factor in several accidents. The crash of an American Airlines Airbus A300 after take-off from New York Kennedy in 2001 is one salient example.
Bombardier's decision to develop a fly-by-wire rudder for the CRJ1000, instead of using the proven existing design, may have been driven more by a corporate strategy to gain fly-by-wire experience for the forthcoming CSeries than by any pressing technological issues unique to the CRJ1000.
During the descent to medium altitude, I accelerated the CRJ1000 to its maximum of M0.80 at 28,000ft. At MMO, small, sharp control inputs in all three axes elicited a well-damped response.
To slow the aircraft, I retarded the throttles and extended the speedbrakes to their full open position. There was a noticeable burble but no untold pitching motion as the aircraft slowed.
I stowed the speedbrakes and continued the descent to 18,000ft, where I would be able to evaluate the CRJ1000's handling characteristics. At an indicated 300kt, I accomplished a series of 45° angle-of-bank steep turns.
I did these with my feet on the floor, allowing the digital yaw damper - part of the rudder control scheme - to co-ordinate the turns. Roll in and out of the turns was crisp, with good harmony between the pitch and roll axes.
Slowing to 250kt, I again did 45° steep turns. At the higher speed, aircraft response was crisp and the turns well co-ordinated.
The flight-control schematic was put up on a multifunction display where, in a clean configuration and 220kt, a maximum available rudder deflection of 1/4th was indicated.
Next, I did a wings-level sideslip and found breakout forces for the rudder to be about 7kg. Full travel required about 20kg of force, with rudder deflection roughly proportional to pedal forces. Allowable rudder deflection increases as speed decreases and the flaps are extended. Speeds below 150kt and flaps to position 30/45 allow full rudder deflection of 33°. Loss of an engine roughly doubles the available rudder deflection at any given speed and flap setting.
Two stalls were accomplished next. The first was in a clean configuration and with a fuel load of 6,470kg. Slowing at 1kt/s in level flight, trim in the pitch axis was halted passing 166kt. Ignoring the slow speed warning on the primary flight display's airspeed tape, the stickshaker actuated at 142kt.
The wing remained steady and level as the airspeed continued slowing. At 131kt, the stick pusher fired, prompting me to recover the aircraft by lowering the nose to break the stall and advancing the throttles.
Adding power caused a slight nose-down pitching motion, further aiding the recovery. The last stall was in the landing configuration, flaps set to "45" and gear extended. The shaker activated, slowing through 128kt with the pusher firing at 106kt.
As with the clean stall, control in all three axes was good at these very slow speeds. The wings remained level and displayed no tendency to oscillate or drop.
After recovering to normal flight conditions, the gear was retracted and flaps retracted to position "8". At 150kt, a half-amplitude rudder doublet was used to excite the CRJ's Dutch roll mode.
With the yaw damper disengaged, aircraft response was lightly damped with a roll to yaw ratio of about one to one. Engaging the yaw damper served to quickly damp out the motion. Aircraft motion from another rudder doublet, with the yaw damper engaged, was quickly damped out without pilot intervention.
Return to Mirabel was via radar vectors to an instrument landing system approach to Runway 06. Litavniks installed the approach procedure in the flight management system, which auto-tuned the navigation radio to the ILS localiser frequency - a handy feature.
Nearing the final approach course, I disengaged the autopilot and hand-flew the aircraft. Extending the landing gear and the initial flap extension to position 8 caused minor changes in pitch force, easily countered with pitch trim.
Flight-director guidance allowed me to roll out on course, as I continued to slow the aircraft. The flaps were extended to position 45 and aircraft slowed to a target speed of 126kt as we approached glideslope intercept. A good amount of pitch trim was needed to null out the change in pitch forces due to flap extension. Again, the flight director provided excellent cues, allowing me to easily capture and track the glideslope.
At about 30ft above ground level I started a slight pre-flare, and retarded the throttles to idle passing about 15ft above ground level. Main-wheel touchdown occurred in a 5° nose-high attitude. After lowering the nose wheel to the runway, Litavniks raised the flaps to position 20 and reset the pitch trim in preparation for the go portion of our touch and go.
With flaps and trim set, I advanced the throttles to the TOGA detent. Litavniks called "rotate" at 126kt, our previous approach target speed.
With the landing gear retracted and safely on a crosswind leg for a visual circuit, Litavniks rapidly pulled the right engine to "idle", to simulate an engine failure.
The CRJ1000's rudder authority is increased in the event of engine failure, and referencing the flight control page on the multifunction display confirmed the FBW system was working as advertised.
Less than 10kg of pedal force was needed to maintain co-ordinated flight at pattern altitude. Rudder trim was more than sufficient to zero out pedal forces, but I elected to fly the approach and landing with rudder trim centred. The simulated single-engine approach and full-stop landing was flown with flaps set to 20.
I found the FBW rudder allowed me to maintain co-ordinated flight on final approach, as I used the left engine to maintain a target speed of 138kt.
As with the first approach, I started a flare at about 30ft above ground level. I centred the rudders as I reduced power on the good engine passing 20ft above ground level.
Just before touchdown, I slightly reduced yoke back-pressure and the CRJ1000 settled down nicely on the runway. With the nosewheel on the runway centreline, I used reverse thrust on both engines and moderate wheel braking to sharply slow the aircraft to a stop.
Bombardier lists a landing field length of 1,754m, a believable number given the short ground roll I observed for the 29,000kg test aircraft.
After my 1h 37min block-to-block flight in the CRJ1000, I was able to jump in the left seat of a CRJ700 Next Gen for a short familiarisation hop with Litavniks.
During the flight we sampled a number of the events we had undergone in the CRJ1000. Aside from gaining a better feel for the CRJ family of aircraft, I was also able to subjectively compare the FBW rudder to the conventional one. In all sampled flight regimes, aircraft responses in the lateral directional axes (roll and yaw) were similar.
Had I not known of the two different control schemes, my brief flights in the CRJ700 and CRJ1000 would not have alerted me to the fact.
MORE THAN A DC-9
More than nine years ago, I had the opportunity to fly the then new CRJ900 from Bombardier's Wichita, Kansas flight-test site.
It was my first exposure to the cockpit of a CRJ, other than the several times I had ridden on cockpit jump seats. I stated then that the CRJ900 blurred the distinction between regional and mainline jet and, in fact, could be considered a mainline jet.
On reflection, I find delineation had more to do with who was flying the jet than what the jet was capable of doing. Several regional jets in current production have seating capacities dictated/limited by labour agreements, not actual aircraft capabilities.
The original Douglas DC-9-10 airliner sat 80 in dual-class and 90 in single-class configuration. The CRJ900/1000, as well as the Embraer 190/195, all have seating capacities equal to or greater than the original DC-9-10. While the CRJ1000 may well be called a regional jet, it is a capable aircraft that can do more than a DC-9-10.