Diamond Aircraft's DA42 Twin Star is high on technology and economy - and proves itself a delight to fly
A quiet revolution in European general aviation is taking place in Austria. Diamond Aircraft, based at Wiener Neustadt, south of Vienna, has just certificated its DA42 Twin Star under the new European Aviation Safety Agency (EASA) procedures. US certification is expected by September.
What makes this four-seat touring and training aircraft different is the combination of advanced features within the first twin-engine design in this category seen in more than 25 years on either side of the Atlantic.
The airframe is carbonfibre-reinforced plastic for lightness and strength, giving the DA42 passive safety levels that meet the new EASA 21 rules, as well as efficient aerodynamics and essentially unlimited airframe life. The engines are turbocharged direct-injection diesels, 135hp (100kW) Theilert Centurion 1.7s, with dual-channel, full-authority digital engine control (FADEC) allowing powerplant and propeller settings to be governed by a single power lever and allowing carefree handling. Finally, the DA42 features Garmin G1000 integrated digital avionics based around two large flat-panel screens that replace conventional instruments and include a crew alerting system (CAS).
This package of features has persuaded customers to place deposits on more than 410 DA42s at about €390,000 ($475,000) a copy in its basic version - and all this before delivery of the first production aircraft, set for September.Flight International was invited to flight test the DA42 at the Diamond factory at Wiener Neustadt Ost airfield, where new facilities will allow the company to manufacture up to 600 aircraft a year. The DA42 is the latest addition to a product range that includes two piston singles - the two-seat DA20 and four-seat, turbo-diesel DA40, from which the DA42 has been developed.
With the production-standard DA42 at an air show in Lyon, France and the other DA42 at Neustadt rigged with Lycoming engines for the US market option, I was to test OE-VDA, a diesel-powered prototype. The aircraft is involved in autopilot certification and had early versions of several cockpit items and additional cables and wires for test measurements.
At the moment, the DA42 is certificated only for visual flight rules conditions, but certification to full instrument flight rules (IFR) should coincide with the first production delivery in September. The aircraft is also only cleared to use paved surfaces, because Diamond had not yet demonstrated the DA42 on grass and unprepared surfaces. This should be a formality.
When fully certificated, the DA42 will be IFR compliant and able to be flown into forecast icing conditions or areas with predicted thunderstorms. The maximum operating altitude is 18,000ft (5,490m) and, although the aircraft is unpressurised, a cabin oxygen system is fitted to supply all four occupants for extended flights above 10,000ft.
Accompanying me on the test flight was Josef Trattner, who delivers and demonstrates DA40s and DA42s for Diamond. We were to use runway 10. The weather was warm and sunny, 24°C (75°F) with a 15kt (28km/h) wind from 180°.
Two pilots, no baggage and full fuel gave a take-off weight of 1,555kg (3,425lb), at mid centre of gravity, compared with the maximum 1,700kg. All limits were to be adhered to, including a never-exceed speed of 197kt and a generous positive g limit of 3.8 flaps up, or 2 flaps mid/down, and a negative g limit of -1.5. The prototype was fitted with a g meter.
Ground inspection was simple and straightforward, but the single locking catch for each baggage compartment door in the nose needs to be checked thoroughly to prevent inadvertent opening in flight. Production aircraft will have an inspection panel in each nacelle to allow the pilot to check engine oil level.
Cockpit entry is via a small step just under the wing root and grab handles above the cabin windows to climb on to a non-slip walkway strip on the wing. Each pilot can enter from his own side, but passengers only from the left side. The forward-hinged clamshell cockpit canopy is enormous and you tend to step into the seat from above rather than from the side. It seemed a pity that, with a canopy that would put an F-16 to shame, Diamond has chosen to paint a large fixed sunscreen on to the perspex to shade the occupants. This restricts visibility when manoeuvring.
Field of view
From the left seat, field of view was excellent all the way around to the horizontal stabiliser above the fin, with the winglet providing an excellent reference. But the view over the nose was restricted, for a number of reasons. Compared with the DA40, the DA42 has a higher glareshield to accommodate a horizontal row of standby flight instruments above the Garmin screens. The seats, while comfortable, are fixed to the floor for crashworthiness and have no vertical or lumbar adjustment. Seat cushions were available, but I find that an inelegant solution for such an advanced aircraft and hope the seating position in the production DA42 will be as good as in the DA40. Rudder-pedal position was adjustable and the car-type three-point seat belts were comfortable to wear and easy to use. Storage nets on the cockpit walls secure small personal items.
Each pilot has a fighter-like central stick, about 250mm (10in) high, positioned at the front face of the seat. This makes for a clean and uncluttered cockpit floor and avoids the pilot's hand obscuring any of the instruments.
Pre-start checks were simple. Inserting the ignition key brought up a caption on the screen confirming the diesel glow plugs were energised. Engine start and propeller rotation was instantaneous, with the "glow" caption disappearing, accompanied by a distinct change in engine tone, after about 5s. The key could then be released, with the propeller stabilising at around 900RPM. Total time to start is 7-8s. The other engine sequence is identical, but care is needed to avoid turning the key inadvertently towards the already-running engine because there is no logic protection to the starter.
With both engines at idle, the avionics were turned on, bringing up the flight and navigation displays on both screens. The FADEC channels were checked and it was time to taxi. The canopy was locked in the cooling gap position, the parking brake set to off using one of three identical small levers in the centre console (the others being cabin heat and cabin defrost - more about this later), power increased to 1,200RPM and we moved forward quickly. Once moving, power was reduced to 1,000RPM to maintain a fast "walking pace" taxi speed. The wide-track undercarriage gave a solid and stable feel to the aircraft, especially when manoeuvring within confined parking areas.
Pedal and brake
The rudder pedal and toe brake arrangement is similar to that in gliders with the pedals pivoted from a central point at their base. This took a little getting used to. Rudder-pedal breakout and friction for nosewheel steering was higher than expected, possibly unique to this test aircraft, but accurate taxiing could be maintained with small pedal deflections every several metres. The disc brakes were effective.
I checked the parking brake could hold the aircraft at full power. The two power levers are the only engine/propeller controls, so the centre console is uncluttered, unlike other twin- engined aircraft. But the power levers were short and had to be operated with my hand below leg level. The production aircraft has longer power levers to make their operation easier and position in the console more distinct.
We refuelled before taking off. The overwing filler caps near the tips are easy to reach. The filler cap is surrounded by painted warnings stating "Jet A-1 or Diesel EN 590 only". But there is no physical interlock to prevent someone filling the tank with avgas, which would be disastrous for the engine. This has occurred at least three times with the DA40. I find it almost inconceivable that no one has yet designed a simple means to prevent incorrect fuel nozzle entry for any GA aircraft.
Refuelled, we restarted and taxied to the holding point. Pre-take-off checklists were again short, logical and simple. Trattner showed me how to input V-speed bugs on the vertical airspeed tape of the G1000: VR was set at 74kt and V2 at 83kt. We taxied into position, set full power against the toe brakes, and checked each engine was stable at 2,300RPM and 100% load and the CAS field in the G1000 screen was empty.
Brakes were released and acceleration was brisk. Despite the 10-15kt crosswind from the right, weathercock effects were slight and the centreline could be accurately tracked throughout the roll. Rotation off the runway and into an 8° nose-up climb attitude was done in one easy movement with a small 3-4cm of back stick deflection and light 1-2kg of force. Time from brake release was 8-9s and take-off distance about 350m. Gear was retracted almost immediately after lift-off and no trim change was felt. The DA42 accelerated to a climb speed of 100kt and I found the aircraft immediately responsive and easy to fly. Any nose-up forces away from the takeoff trim setting were neutralised by half a forward turn of the elevator trim wheel.
The climb at speeds from 100kt to 190kt allowed me to assess the aircraft's control characteristics and trimmability. For small turns, the controls seemed well harmonised but, as in any small aircraft, I had to learn to use my feet again and actively use the rudder for turn co-ordination and minimising sideslip. Breakout and friction for aileron and elevator were low. Centring of the controls (when released from trim position) was adequate at 140kt but much better at 190kt. The elevator trim wheel had about a quarter-turn of freeplay but was effective, but the rudder trim wheel was difficult to turn and its effect seemed minimal - the system is being updated for production.
The engine power levers were a delight to use, despite their small size, allowing virtually carefree operation. Diamond recommends the levers are advanced from idle to full power over about 4s, rather than slamming them open. Propeller RPM at idle (0% load) is 2,175, at 20% load 1,750 and at 100% load 2,300. As power is increased rapidly from idle to full, there is a disconcerting momentary, but distinct, change in engine/propeller note and slight loss of thrust. Once experienced, it then becomes completely normal.
We climbed quickly to 6,000ft, where bright sunshine allowed me to assess the readability of the G1000 screens. The flat panels, with a viewable area 220mm wide and 160mm deep, are identical but with differing functions for the menu buttons along the side and base of each screen. I suggest wearing gloves - the glareshield overhang is not that great and at some angles the screen was not easy to read if there was any finger grease contamination. Otherwise the flight and navigation information was easy to read. I found the moving map on the right-hand screen useful.
Both screens can show the same basic information, or flight data can be shown on the left display and navigation on the right. I feel the vertical tape presentation of airspeed, altitude and vertical speed will require careful attention when converting from an aircraft with dial-type instruments. The G1000 offers an incredible array of features, but it is an advanced display and flight-management system combined and pilots will need almost as much training on this as the aircraft itself.
Longitudinal and lateral/directional stability was checked next around a trim speed of 140kt and level-flight power with 50% load on each engine. At 190kt, it took only about 3kg forward stick force to hold the elevator out-of-trim force. This means you are not constantly trimming when accelerating. The phugoid was well damped and, with a period of around 60s, would not be obtrusive in cruise flight. The short-period check showed the aircraft was quick to respond to stick input and well damped. Turns, flaps up, up to 3.5g were made in both directions with no stick-force discontinuities and no buffet.
Spiral stability was neutral or slightly positive, which is a good safety feature for inattentive pilots. Dutch roll was quickly damped within half a cycle without the need for pilot input. Roll rates at 140kt with full stick deflection but without rudder were about 35°/s and about 50°/s when half rudder was added. At full rudder deflection, at 140kt, the opposite bank angle was 30°, showing the aircraft has good sideforce characteristics to kick off drift during a crosswind landing.
Aileron-only turns, using full lateral stick, resulted in an immediate, opposite heading change of about 5° due to adverse yaw, but was neutralised by the aircraft's directional stability after 2-3s. In rudder-only turns using full pedal deflection, the aircraft yawed in the direction of the applied rudder and 3-4s later rolled positively in the same direction at about 10-15°/s.
Throughout the flight I found the DA42 to be "lively" directionally and did my best to use the rudder to counteract the slip. But the only slip indicator is a 1 x 5mm grey bar that sits horizontally under the small grey bank-angle indicator, on a blue background, at the top of the horizon display on the G1000 screen. Given the slip indicator's importance, it did not exactly stand out. I hope Diamond will consider fitting a chunky, old-fashioned "spirit level" type slip ball in the production aircraft.
Unaccelerated stalls were evaluated with zero, take-off and landing flap deflection, all at idle power, and were probably the most docile of any aircraft I can remember. Controls in all axes were effective approaching and in the stall. The warning horn sounded 5-7kt above stall, which was indicated by increasing airframe buffet, distinct nose drop of 5-10° and vertical descent of about 2-300ft/min. There was no wing drop. Stall flaps "up" was around 65kt, flaps "take-off" 55kt and flaps "landing" 45kt. I tried a stall in approach configuration, with take-off flaps and 30% engine load, but with the nose about 20° above the horizon and distinct airframe buffet, horn sounding and virtually no sink rate, I gave up. I believe any pilot or student would have to more than grossly mishandle the aircraft on approach to get it to stall.
My final check was to assess the DA42's behaviour after a simulated engine failure at V2. I stabilised at 80kt, full power, nose about 10° above the horizon and rapidly brought the left-hand power lever back to idle (the propeller was not autofeathered). Swing towards the dead engine was instantaneous, but could be held with rudder, although the forces were high at 80kg. Once gear was retracted, these reduced to about 50kg, with the nose gear doors no longer acting as a forward fin. With autofeather, these forces would reduce even further, as they would with rudder trim if it was not so difficult to grip.
On the return leg, at 75% engine load, in level flight at 165kt and 5,000ft, fuel consumption was 21 litres/h (5.5 US gal/h) per engine, with the computer showing a 1,110km (600nm) range and 4h endurance. In a long-range cruise at 50% engine load and 140kt, fuel flow was 14 litres/h, and calculated range 1,480km and endurance 6h.
We dived back towards the circuit at 190kt. There is no overspeed warning but the airspeed tape shows a red/white "barber's pole". Another small niggle - at power settings below about 30% load, the aircraft has a non-cancelling gear-up warning horn, which sounded for the whole high-speed descent until we were in the circuit and power was increased on the downwind leg. I suggest the warning be linked to flap position so that if the gear is not down when the first stage of flap is set on the approach, then the gear-up warning sounds.
Various normal and flapless circuits and roller landings were flown on to runway 10, with the crosswind still about 10-15kt from the right on touchdown. They were easy to fly. The low longitudinal static-stability gradient combined with the vertical-speed tape display meant I had to pay particular attention to speed control. Initially, my crosswind landings were untidy, with too much "into wind" wing down. I was kicking off drift in the flare a few feet too high, giving the delayed roll with rudder effect time to kick in, and requiring me to compensate. By waiting until the final flare phase before kicking off drift, I could arrive neatly.
The final full-stop landing required my first touch on the wheel brakes. As we rolled along the runway after touchdown, at about 60kt, I applied the toe brakes lightly and the aircraft suddenly decelerated sharply and veered towards the side of the runway. I kept it under control but the aircraft's reaction was slightly disturbing - until we understood the problem. During the test flight we had moved the three identical centre-console levers backwards - and one of them was the parking brake. But we had forgotten because the brake lever's design and position are not unique. The brake remained dormant until energised by application of the toe brakes. I recommend Diamond redesign the parking brake lever to make it more distinctive.
We had been airborne for 1h 20min and used 55 litres of fuel. Diamond has taken a bold and successful step in combining so many new technologies in such a fine aircraft. It provided its prototype at short notice and many of the shortcomings I noted will be rectified in production.
I found the aircraft a delight to fly and easy to operate, combined with high performance, modern technology, excellent economy and the latest and highest levels of passive safety. The DA42 has no natural competitors in its class and sets a benchmark for European general aviation.