1168 An early look at Fairchild Dornier's 328JET suggests a successful change to jet power

Peter Henley/Oberpfaffenhofen

Fairchild Dornier's 328JET, a derivative of the company's twin turboprop 328, is aimed at the regional and corporate markets and will be offered as a 34-seat commuter aircraft or as a business jet (recently named the Envoy 3) with a cabin configured to suit individual customers. So far, orders and options total 79 - including 51 firm sales, of which 11 are for the Envoy 3.

Apart from the fundamental concept of changing to turbofans, alterations to the 328 turboprop have been relatively few. The basic wing is retained, with its characteristic swept tips - designed to reduce drag without recourse to a winglet. The fuel capacity within the wing has been slightly increased and its structure strengthened where required. The wing is of modern construction, with upper and lower skins milled from the solid to give a smooth surface relatively free from riveting. The wing section was originally developed by Dornier for its 228 and retained for the 328 turboprop. Thickness-to-chord ratio has been reduced for the 328JET by the simple expedient of increasing the flap chord by 100mm.


Similarly, the fuselage remains essentially the same as the 328 turboprop's. Frames 24 and 26, which carry the wing and undercarriage attachment points, have been strengthened. The jet is designed to operate at up to 35,000ft (10,700m), whereas the turboprop was limited to 31,000ft, meaning an increase in cabin pressure differential from 0.483bar (7lb/in2) to 0.517bar. No modification of the pressure hull was needed.

A stronger undercarriage has been developed for the heavier 328JET, incorporating dual braking and anti-skid systems and carbon brake discs for better heat dissipation - all manufactured by Dunlop. Part of the motivation for more powerful brakes arose from the decision not to fit thrust reversers - although they are planned as a possible future option.

The airframe de-icing system - using BFGoodrich inflatable boots for the leading edges of the wing, fin and tailplane - is to be retained, although bleed air anti-icing is being considered as a replacement on future variants. The turboprop nacelles have been replaced by pylons for the Pratt & Whitney Canada PW306B turbofans, and the panels fitted to the turboprop fuselage sides - to protect them from flying ice shed by the propellers - have been removed. Cabin sound suppression is expected to result in internal noise levels 2-4dB quieter, zone for zone, than those in the turboprop - which, Fairchild Dornier claims, is itself exceptionally quiet.

The result, says Earl Robinson, Regional and Business Aircraft division president, is "-a jet aeroplane with field performance as good as that of a turboprop - maybe better - which is more economical to own than a turboprop".

The 328JETis intended for use on relatively short sectors - ideally of about an hour, but up to a maximum of 2h 20min. Its main competitor is seen as the Embraer RJ-135 and Robinson claims some performance superiority for the 328JET over the Brazilian aircraft, notably in the climb rate at high ambient temperatures.



Eventually, there will be four development aircraft. The first to fly (and the aircraft flown by Flight International for this evaluation) was Jet 1, an adaptation of a 328 turboprop. It was a development airframe for the turboprop, so was far from a new aircraft and, in several significant respects, differs from the production 328JET specification (the undercarriage, for example, is as for the turboprop). Its first flight after rebirth as a 328JET was in January this year.

The second prototype first flew in May and was built from scratch as a 328JET. Each aircraft will have a different role in the development programme, covering handling, performance, reliability, demonstrations and avionics. Certification is targeted for February 1999 in Europe, and March 1999 in the USA.

Already the major milestone of flutter testing has been passed. These tests revealed a tailplane shockwave at high speed, which has been ameliorated using a fence fitted to the fin below the tailplane. There is a large fairing on top of the pressure hull, which blends the wing-to-fuselage joint. This also houses some equipment such as the environmental-conditioning system packs. The portion of this fairing behind the wing was found to provoke Mach buffet at M0.72 and 1.9G; it will be reshaped on the production aircraft.

Finally, in comparing the 328 turboprop with the 328JET, the latter has more powerful generators and a dual hydraulic system, while both versions share the Honeywell Primus 2000 digital avionics system. The two ventral strakes, or fins - required to help the 328 turboprop directionally - are not so important to the 328JET in the absence of propeller slipstream effect, but they are to be retained to avoid the cost of re-certification without them. Their retention carries no significant penalties. Although the 328JET is being certificated as a derivative of the turboprop (not as a new type), the flight test programme is clearly extensive and is expected to take 1,200h of flying.

Fairchild Dornier's facility at Oberpfaffenhofen, near Munich, Germany, seemed well suited to the development task, having a 7,800ft runway (not routinely shared with other users), a well equipped flight test centre, including the use of telemetry, and enjoying access to adjacent military airspace, also relatively unimpeded by other users.

The assembly line for both the 328 turboprop (which will continue to be made for as long as demand remains) and 328JET are on-site, thereby helping the management of development, production and marketing functions.


The 328JET has been designed for easy turn-round servicing. The door for the cabin and cockpit is on the left side, just behind the cockpit, and folds down to provide steps and a hand-rail. The baggage compartment door is on the same side, at the rear of the fuselage, and there is a service door on the right side of the rear fuselage, next to the usual site for the galley. The hydraulics bay is in the right undercarriage pontoon, the lavatory is serviced at the rear right fuselage and the single-point refuelling panel is on the right wing, although the required uplift has to be selected on a panel in the cockpit. The unfurnished cabin of the test aircraft provided no opportunity to form an opinion on seating or baggage accommodation, but the headroom is clearly good at 1.89m at the point where the aisle will be.

The cockpit provides a pleasant working environment for an aircraft of this size. Working oneself into and out of the seats is reasonably easy, with room between the seat and centre console to slide one leg through at a time. There is room to stow a standard flight bag outboard of each seat. The seat has a five-point harness, is adjustable fore and aft for height and recline, and has stowable armrests adjustable for height. The rudder pedals are adjustable for reach, using a release lever between the pedals. The field of view is good, each pilot being able to see the outer wing and engine intake on his side of the aircraft. A movable sun visor on a rail along the top of the windscreens is provided for each pilot.

None of the cockpit windows opens, but an emergency hatch with escape ropes is provided in the cockpit roof. The windscreen panels have electric heating, which needs to be switched on throughout the flight to achieve design impact strength for the glass. The cockpit is finished in an attractive combination of dark and lighter blue, rather than the more common grey or beige colour scheme.

The AlliedSignal 36-150(DD) auxiliary power unit (APU), which lives in the fuselage tailcone, was started using a ground power unit, but could equally well be started using the aircraft batteries. Once on-speed, it provides bleed air for cockpit and cabin conditioning and for engine starting, plus electric power until the engine generators come on line.

During normal use, the APU would be shut down after take-off and restarted before landing, and would be available during flight for emergency use. On this development aircraft it was kept running throughout the flight, with a slightly detrimental effect on the overall fuel consumption.

The 26.9kN (6,050lb)-thrust PW306Bs have full authority digital engine control (FADEC). Starting is automatic, using bleed air, with parameters monitored by the FADEC; any malfunction would provoke a crew alert - but not automatic shutdown, which remains a crew responsibility.

The Honeywell avionics and displays are becoming increasingly well known to pilots because of the extent of their market penetration. This Primus 2000 installation has the customary five displays - a primary flying display (PFD) and a multifunction display (MFD) for each pilot and a central display for the engine instrument and crew advisory system (EICAS).

During the process of APU and engine starting and when checking the flying controls before take-off, for example, it is possible to select for display on an MFD a page from a menu of systems schematics on which to monitor the process. Such schematics can also be selected at will in flight, to monitor a system operation or malfunction and the effect of corrective action. Because this was a development aircraft, it was fitted with a further flat screen display attached to the glareshield and using sensors entirely separate from the normal avionics.

A long nose probe provided a free-air pitot source and sensors for measuring angle of attack and sideslip required by the development and certification programme. Fairchild Dornier's telemetry equipment allowed the crew instant access to many other parameters through a VHF radio link direct to personnel in the telemetry room.



Fairchild Dornier test pilot Wolf-Dietrich Havenstein, a graduate of the UK's Empire Test Pilots' School at Boscombe Down, Wiltshire, was in command of our aircraft, Jet 1, and took the left seat because there was only one nosewheel-steering tiller - on the captain's side console. I flew the aeroplane from the right seat.

A flightplan had been fed into the flight management system before engine start. Although the power interrupt between ground power and aircraft power after starting could be detected as a "blink" on the displays, none of the information was lost or corrupted. For take-off, I had the standard attitude display, plus airspeed, altitude, vertical speed and cleared altitude on my PFD, and the departure track on my MFD. Meanwhile, taxiing from the parking area to the runway holding point from the right-hand seat took a little crew co-operation.

For sharp turns, Havenstein had to use his tiller to steer the aircraft, but otherwise I was able to negotiate the normal curves of a taxiway using the rudder pedals, which move the nosewheel steering up to 10¼ either side of centre.

Differential brake and power can be used, if needed. This aircraft retained the turboprop's steel brake discs; the toe brakes had to be applied gently to avoid fierce responses, but, once applied, the brakes were powerful and progressive. It was not possible, of course, to experience the definitive carbon disc brakes.

Jet 1, registration D-BJET, had been prepared for flight with 2,268kg (5,000lb) of fuel; the cabin was untrimmed and was fitted with water tanks for weight and centre of gravity (CG) adjustments. No ballast was carried for this flight and, with a crew of three (including Fairchild Dornier safety pilot Klaus Scholz in the jump seat), the all-up weight (AUW) for take-off was 12,500kg, with a mid-envelope CG. The maximum take-off weight for production aircraft will be 15,200kg. The airfield altitude was 1,900ft and the temperature +15¼C. Runway 22 was used and there was a crosswind of about 4kt (7km/h) from the right.

Before take off, the flap had been set to 12¼ (the other flap settings are approach/20¼ and land/32¼) via the flap lever to the rear of the power lever quadrant on the centre console. The take-off reference speeds had been calculated and entered into a panel at the bottom of the EICAS display. The nosewheel-steering tiller had been isolated via a button, also on the rear of the power lever quadrant (rudder pedal control of the nosewheel is used on take-off).


The parking brake, on the co-pilot's side of the quadrant, was released and the power advanced for take-off power, placing the engines under the authority of the FADEC system for the entire flight. The engines spooled up and the aircraft accelerated smartly. Keeping a straight course on the runway was easy, despite the slight crosswind. Rotation was at 117kt, after which the aircraft climbed initially at the final segment speed of 131kt.

Havenstein retracted the undercarriage, which produced no pitch change, and the take-off flap was retracted at 400ft above the ground. The best rate of climb seen was just over 5,000ft/min (25m/s) at a pitch angle of about 15¼, giving credence to Robinson's earlier claim about the 328JET's prowess in the climb, even allowing for the lower than maximum AUW and the relatively cool day. Certainly, there was an abundance of thrust available.

The en route climb was at the discretion of Munich air traffic control (ATC) and did not allow for much manoeuvring. It was immediately clear, however, how high the roll control forces were. All the primary control surfaces of the 328 (JET and turboprop) are mechanically operated. The turboprop uses, in addition to the ailerons, a pair of hydraulically activated roll spoilers. Jet 1 has inherited this roll control, whereas the second prototype and all 328JETs thereafter will have ailerons only, without roll spoilers, but with much closer gaps between the control surface and the adjacent wing. This, I was assured, has brought about a great improvement in roll control, already evident during the early flights of the second prototype.


Jet 1 was not equipped with an autopilot, so had to be flown manually throughout. The heaviness in roll was confirmed with exposure to handling the aircraft, becoming increasingly more apparent (as to be expected) at high airspeeds and progressively more tolerable at speeds below about 200kt.

While manoeuvring at about 15,000ft and 200kt, roll reversals from 45¼ bank to 45¼ of opposite bank not only endorsed the high control forces needed, but produced disappointing roll acceleration and rate of roll. Also, the 328JET is a relatively close-coupled aeroplane and a Dutch-roll tendency was evident - the yaw damper to be fitted to production aircraft is part of the autopilot servo function and was not fitted to Jet 1.

These comments highlight the numerous dilemmas that face manufacturers during the early development of aircraft - in this case, should outsiders (including journalists) be allowed to see the problem areas before they are rectified in order to have an early impression of the aircraft overall?

On balance, there is no doubt that they should; the cures both for roll control and the Dutch roll mode are known and will be built into subsequent 328JETs and there is no reason to believe they will not be successful. Meanwhile, the overall impression gained from flying the 328JET hybrid development aircraft was entirely favourable.

A particularly impressive area of the 328JET is the versatility of the wing. This is a straight (unswept), relatively high aspect ratio wing, designed to give a turboprop the best compromise between field performance and economical en route performance. Now it is being asked to be the wing on a 400kt (true airspeed) jet aircraft. Nevertheless, at 35,000ft and M0.7 and, with an AUW of 11,800kg, it was possible to detect only the slightest trace of Mach buffet and, in turns with up to 45¼ bank at M0.68, no stall buffet was provoked. The aircraft remained pleasant and easy to fly at this level.

Still at 35,000ft, where the ambient temperature was ISA +10¼C, a cruise was flown with the engine N1(fan RPM) set at 97.6%, resulting in a stable M0.69 and a fuel flow of 817kg/h. The 100mm increase to the flap chord (thereby decreasing the thickness-to-chord ratio of the wing, while slightly reducing its aspect ratio) appears to have produced a surprisingly efficient, comparatively high speed, high altitude wing.

A descent from 35,000ft to 20,000ft was started next. Here, the 328JET's clean airframe and lack of airbrakes becomes apparent. Closing the power to flight idle and flying at M0.61 gave a maximum rate of descent of more than 5,000ft/min, but an average for the 15,000ft of descent of 4,000ft/min. Slowing the 328JET is not easy and its rate of descent could be made greater with airbrakes. Their absence is likely to be noticed most when making steep approaches to airfields where local terrain or noise abatement techniques require this. Fairchild Dornier is studying this issue.

A possibility would be to use the redundant roll spoilers symmetrically as airbrakes, but, from their size alone, it seems that they would not necessarily be effective enough. A common location for airbrakes on modern commercial jets is in the tailcone, but, in the 328 JET's case, this is where the APU is housed. The undercarriage does not create much drag when extended and is limited to a maximum speed, extended, of 200kt. An obvious solution would be to fit airbrakes to the rear fuselage sides, but the cost, weight and engineering feasibiIity will all need careful consideration.



Another cruise, at 20,000ft, was flown using 92.4% N1, giving a stable M0.66 or 405kt TAS, and 1,300kg/h fuel flow. Then a further descent into a block of airspace was made, for general handling. An acceleration-deceleration was flown at 15,000ft, starting at 135kt (clean), with power to maintain speed and a pitch angle of 6¼ nose up. Applying full power and accelerating to 300kt (VMO) - where the pitch angle was 0¼ - was impressive, taking a mere 1.09min. The longitudinal pitch change with power and speed was easy to trim using the twin control-wheel switches. When the power was then reduced to flight idle, the "slipperiness" of the 328JET was again evident, the deceleration to 135kt taking 2.20min, and the pitch changes once more being easy to trim.

The behaviour of the 328JET during flap extension at limiting speeds was examined next. Selecting 12¼ of flap at 200kt, 20¼ at 180kt and 32¼ at 145kt all caused some ballooning, but the effect was readily countered with elevator and the resultant pitch force easily trimmed.

Stalling, at 16,000ft, was next on the agenda. The 328JET has both a stick shaker and a stick pusher. The natural stall was not a problem and is better still in the 328JET without propeller slipstream effect. The shaker and pusher were fitted to the turboprop merely to ensure trouble-free certification, and have been kept on the jet.

Two power-off (flight idle) stalls were tried. For the clean stall, the trim speed was 135kt, the shaker speed 107kt and the pusher 103kt. With flaps at 32¼ (land) and undercarriage extended, the trim speed was 100kt, the shaker 82kt and the pusher 80.5kt. The stick shaker was compelling and the pusher firm and effective. Handling of both during deceleration and near the stall was good, with reassuring roll control down to the shaker speed. Recovery from the stall was equally normal and benign. I was told that the natural stall (without the shaker or pusher) is similarly benign and is identified by a G-break.

The AUW was now down to 11,250kg, and an engine failure at safety speed (V2) was simulated at 15,000ft. When the aircraft was on condition at take-off power (102.8% N1), undercarriage up, flaps at 12¼and at 117kt, Havenstein shut down the right-hand engine. The asymmetric thrust was easily contained using rudder, with the wings level, and applying a foot force of 42kg - the figure being helpfully provided by the on-board test equipment. The foot force was then easily trimmed using the small round trim switch on the centre console. The single engine rate of climb was 1,100ft/min.

The final phase of the flight comprised a descent and return to Oberpfaffenhofen for an instrument landing system approach to a low go-around and entry into the visual circuit, or traffic pattern. Here, the versatility of the Honeywell flight management system (FMS) asserted itself. The route, as required by Munich ATC, had been entered and was displayed on my MFD, in lateral and vertical format, so that situational awareness was good throughout. Munich ATC was then obliged to change the route and give radar headings for traffic avoidance, while we had to make several requests for deviations to avoid weather build-ups.



The FMS was simple to reprogram (by the non-flying pilot) and the resultant flight director demand information was easy to follow. Because of these changes, a visual intercept at about 2km on finals had to suffice. Maintaining the centreline and capturing the glideslope were both straightforward to accomplish manually in the absence of an autopilot, the 328JET being pleasant and easy to fly throughout.

At this early stage of the development programme, the landing reference speeds had deliberately been kept to about 10kt higher than those predicted for in-service flying. Nevertheless, the elevator control force in the flare was a little higher than I had expected and the aircraft settled firmly, but precisely, on to the runway. Automatically triggered lift spoilers deployed on touchdown and caused the nosewheel to be brought positively on to the runway - after which, wheel braking could be applied as required.

There was no difficulty in keeping straight, using the rudder pedal operation of the nosewheel steering, and the light crosswind required no countering despite the 328JET's high wing and narrow undercarriage configuration. The standard operating procedure for taxiing to the stand from the runway will be to shut down the left engine and use the right only. This is to allow safe and quiet disembarkation of passengers and unloading of baggage once the aircraft has stopped.

Although D-BJET is different from the production standard for the 328JET, it was obvious that this test aircraft has plenty of power, giving impressive and economical performance. Engine response is good throughout, particularly at low engine speeds and low power lever angles. The wing is remarkable for its high altitude, high speed performance and for its benign behaviour at low speed and approaching the stall. The cabin seems roomy for the overall size of the fuselage, baggage space is good and the cockpit noise level impressive, there being little aerodynamic noise and little intrusion from the engines, even when they are working hard.

This was an enjoyable and interesting flight at an early point in the 328JET's development and certification programme.

Source: Flight International