Chris Yeo/ISTRES FLIGHT TEST CENTRE
Prototype fighters are usually first seen by the public at air shows. They are presented in an optimum configuration without large, heavy, drag-producing external stores and with a reduced fuel load to minimise gross weight.
Recognition of the multirole nature of modern combat aircraft has resulted in more and more air displays with external stores fitted. These performances show what the aircraft's capabilities are in close combat near the end of a sortie.
This is only part of the story, however. Front-line fighters are expected to be capable of more than one role and will be certificated to carry a wide range of external stores including bombs, missiles and pods. Even a pure fighter will normally be loaded with external fuel tanks to extend the basic aircraft's range and endurance. The aircraft and pilot will then be required to fly and fight in these heavy configurations.
To be used effectively, the flight control system (FCS) must still allow the aircraft to be flown aggressively, with as little restriction on the allowable manoeuvre envelope as is practical. Fighter pilots have long known that, if they can force an enemy to drop its external stores, the attack has been neutralised - whatever the outcome of the resulting combat.
The aircraft in this test, the tandem-seat Rafale BO1 powered by two 16,850lb (75kN) thrust (with reheat) Snecma M88-2 engines, was at the time of my flight being prepared for display at the Paris air show with an impressive range of fuel tanks and armaments fitted under the wings.
The test configuration included three 2,000litre (530USgal) fuel tanks, two SCALP cruise missiles weighing 1,300kg (2,860lb) each and four Mica fire-and-forget air-to-air missiles. For a small aircraft such as the Rafale, this is a heavy load - but one that promises to allow a thorough evaluation of the aircraft's flight control system and performance.
Prototype weight limit
Prototype Rafales are limited to a maximum take-off weight of 19,500kg, although, after a modification to the undercarriage, the production aircraft will be cleared to 22,500kg. Eventually, the maximum take-off weight may be further increased to 24,500kg. The empty weight of the aircraft was 10,000kg, with the external stores contributing a further 4,500kg to give a weight without fuel of 14,500kg. The maximum fuel load that could be carried was therefore 5,000kg, of which 800litres, weighing 680kg, was in the centreline fuel tank.
The prototypes have been tested to 22,500kg using air-to-air refuelling (AAR) to top up the external tanks. This is commonplace because it extends the duration and productivity of test sorties and allows several test points to be achieved in one flight at maximum weight. For this flight, the aircraft had a fixed refuelling probe fitted to the right side of the nose ahead of the cockpit. The Rafale FCS has a sub-mode tailored for AAR.
BO1 is the only two-seat Rafale prototype and has performed 990 flights since it first flew in mid-1992. The four Rafale prototypes (two air force versions and two navy) are fitted with test equipment that allows them to be used in any part of the test programme. However, individual aircraft have tended to be specialised to some degree.
BO1 has been used for flight envelope expansion and weapons systems tests, including Mica firings and SCALP separation demonstrations.
It has also been used to examine systems specific to the two-seat aircraft. Of these, perhaps the FCS is the most interesting to pilots. Because the control columns and rudders are not linked, some form of priority protocol must be established, otherwise an instructor and a pupil with diametrically opposed views of the best input for the current situation will achieve only a neutral result.
In the Rafale, the front-seat pilot has a switch that gives priority to either the front or rear cockpit. The rear pilot can use a paddle on his control column to totally override the front pilot. A small light on the left glareshield confirms the instructor's actions to the pupil.
This test flight was made from the Istres flight test centre, near Marseilles. Yves Kerherve, Dassault Aviation's chief test pilot, occupied the rear cockpit. The flying clothing used included a conventional g suit, without leg restraint garters - these are built into the Martin Baker Mk16 seat - and an upper body jacket (still called a Mae West, even in French), with arm restraint.
This level of equipment is fairly hot to wear, but was surprisingly comfortable and easy to put on. The helmet was close-fitting and light, an absolute essential for an aircraft capable of long periods at 9g, as the clean Rafale is.
The Rafale's appearance was dominated by the external stores, particularly the large, black SCALP missiles mounted under the wings outboard of the external fuel tanks. The cockpit is ahead of the canard and is boarded by an external ladder (integral on the naval version). It is small, but laid out logically, with essential operational functions on the side-mounted single throttle and control column.
Both the throttle and stick are mounted unusually high on the side of the cockpit, just below the canopy sills, with an adjustable wrist rest in the case of the stick. This arrangement releases more space on the side panels for switches and helps alleviate the problem of blood pooling in the pilot's arms at high g.
In front of the pilot are three liquid-crystal electronic displays, a wide-angle head-up display (HUD), two small keyboards and several switches. The HUD is the primary flight instrument display, using information supplied from the twin inertial navigation/global positioning system. There is a small liquid crystal panel low on the right glareshield with a conventional attitude display that includes heading, airspeed and altitude. This is driven from the FCS and is used as a "get-you-home" display in the event of a major system failure.
The left and right display surfaces are touch-sensitive. The pilot can use this method or a cursor to change modes or display settings. Touch control was easy, even with gloved hands. The cursor was driven from a small thumb-operated joystick control on the stick. It could be moved smoothly across all four electronic displays and would be used under high g to designate the function.
Different system formats are called up on the displays by small levers which the pilot can reach with his fingers from the stick and throttle. All the head-down displays were well shielded from the bright sunlight and were clear and easy to read. The centre (eye level) display is collimated to infinity and is used primarily for tactical data. Collimation allows the pilot to look down from the HUD and the outside world to the centre display without refocusing his eyes. In practice, I found this worked reasonably well, but I did have to refocus to some extent. The prototype system shows route and waypoints. The production system will also have a full geographic map.
The starting procedure for the Rafale was simple and comprehensively automated. Having closed the canopy, I set two small auxiliary throttles to idle and selected the engine starting order, left then right (there is no preferential order) with a single rotary control. The auxiliary power unit (APU) started after 20s, followed by both engines (1min 50s). The FCS pre-flight test was equally as easy - a guarded switch was selected aft and, after 8s, a green light indicated GO.
The development APU is used only for starting, but the production unit will also provide air conditioning, although not electrical or hydraulic power. This will be a welcome improvement because, once the canopy was closed, the cockpit was warm until the engines began to provide a cool airflow.
With clearance to taxi and confirmation from telemetry that all was well, I set the small auxiliary throttles to normal to put both engines under the control of the single main throttle, engaged the nosewheel steering via a stick-mounted switch and waved the chocks away. A small amount of thrust was needed to get the Rafale moving and around turns, but for the most part it maintained a steady pace with the engines at idle.
The nosewheel steering, controlled via the rudder pedals, was smooth and accurate. The wheel brakes, operated by toe pedals, were progressive and smooth. Throughout the long taxi route to the threshold of Runway 15, the ride was good and nosewheel strut damping excellent. Like most fly-by-wire aircraft, the Rafale FCS was active once engaged and the canards could be seen in the mirrors moving in response to aircraft pitch movements.
After lining up behind a departing Boeing KC-135, we were reminded that even flying at a test airfield is never straightforward as the tanker returned for a low pass by the tower to check for a fuel leak. Nothing untoward was seen and we were cleared for take-off.
At Kerherve's suggestion, I abandoned my normal technique of checking the engines at maximum dry power and simply released the brakes and slammed the throttle to full reheat, trusting the automatic fault detection system to warn of any failures and the longitudinal g monitor in the centre of the HUD to verify that the expected thrust was being developed.
The engines took 5s to achieve full thrust, at which point the aircraft was accelerating briskly at 0.56g (longitudinal). The stick back speed was 150kt (277km/h) and the aircraft lifted off at about 165kt after 15s, including the time taken for the engines to accelerate. It was easy to keep straight in the 7kt crosswind component from the right using nosewheel steering up to 60kt, at which point it automatically disengaged.
Rotation to a suitable take-off attitude was easy, although I initially underestimated the pitch responsiveness and corrected forward rather more aggressively than I would have liked to maintain the initial climb flightpath angle (FPA). The undercarriage retracted in 5s, with no trim change. The FCS laws change from angle-of-attack (alpha) to g demand with the undercarriage retracted.
The initial climb to 10,000ft away from the Istres circuit area and westwards across the serene Carmargue countryside was made at full dry thrust at 340kt (a nominal rather than ideal speed) at an FPA of 15º.
First impressions are important to a test pilot - the human body adapts to a new environment quickly, so I used this period to start to become acquainted with the Rafale. Despite not having flown a fighter for several months, I quickly felt at home with the FCS and the cockpit. The aircraft responded quickly and positively to control inputs and could be placed accurately. The control forces were pleasant and well balanced.
The Rafale FCS automatically trims the aircraft in all three axes. In pitch, it trims for 1g flight, so speed changes are made without the pilot needing to retrim manually. The only time that conventional static stability is introduced is above 16º alpha, the normal approach incidence, with the undercarriage down. Throughout this flight, the autotrim system worked well and unobtrusively. I was briefed, but could not check, that the system copes with asymmetric loads, such as a hung-up bomb.
Once level at FL100, at 300kt and 85% engine core speed (NH),I made a few turns before manoeuvring the aircraft more aggressively. Immediately I was reminded how useful a well-sorted HUD is for accurate flying. Level turns at 45º (90% NH) and 60º (93% NH) angle of bank merely required keeping the aircraft symbol on the horizon line and adjusting engine thrust to keep the energy markers at neutral. This was an easy, straightforward and intuitive process that gave level turns accurate enough for any instrument-rating examiner.
Peak roll rate
Four full-stick rapid rolls through 360º were made at 1g and 2g at 300kt. The roll acceleration was good and, in each case, the roll was completed in 3.5-4s. The peak roll rate was about 150º/s. Without the heavy external stores, the FCS would have allowed a higher roll rate of 250-270º/s. The aircraft was inverted briefly in level flight - something only a test pilot would attempt with two large cruise missiles and three external tanks on board - and remained easy to fly accurately. The FCS limits negative g as well as positive g, although I did not bring in the g limiter during this test point.
Finally, before climbing to high level, a hard turn was made, starting at 330kt using full reheat, principally to test the behaviour of the FCS. This was the first moment in the flight for controlled aggression. I simply rolled the Rafale into a nose-down steep turn and, as the reheat became effective (about 2s), moved the stick quickly to the aft stop. The aircraft responded by rapidly achieving 5g at 17-18º alpha, turning smoothly and without buffet with the stick held on the aft stop.
Heading north towards the mountains, we climbed to 25,000ft. I wanted to explore the Rafale's handling at high level at the Mach limit in the current configuration - 0.9 indicated Mach number - and in a typical long-range cruise condition. The climb also gave me an opportunity to look more closely at the autopilot.
A few years ago, the best that a fighter pilot could expect was a basic attitude mode plus heading/track and height holds. A sophisticated autopilot has now become recognised as an important facility in a modern fighter, because it allows the single pilot to devote more time to managing the wide variety of electronic systems that demand his attention. The Rafale autopilot was engaged by a small paddle switch on the stick, and the autothrottle was engaged by a similar switch on the throttle. After that, the aircraft could be "flown" using a small thumb switch on the main control column.
The autopilot could also be linked to a heading and height set via the left display screen and/or coupled to a flightplan pre-entered in the navigation computer. The flightplan was shown diagrammatically on the central eye-level display. Although I did not have time to test all the autopilot facilities, Kerherve explained that the system is capable of flying the aircraft through a four-axis flightplan, which includes timing.
The autopilot was remarkably easy to use and could demand rapid but smooth changes of FPA and heading. One novel feature of the system which was simple and logical was that a turn to a new track could be demanded by pressing either of the toe brakes. Pressing both toe brakes levelled the wings and held the track constant.
The autothrottle holds the airspeed at engagement or at a speed set by the pilot via the left-hand multifunction display and has authority over the unreheated thrust range only. Once engaged, the autothrottle controls the engine, but does not drive the throttle position. To my mind, this system - also used in Airbus airliners - deprives the pilot of useful information. I would prefer to know by tactile feedback how much thrust is being demanded from the engine. However, I concede that a mechanism to backdrive the throttle would be complex and that it might be difficult to install on the high-set throttle in the Rafale cockpit.
At top of climb, the autopilot levelled the aircraft smoothly at the pre-set altitude of 25,000ft. Visibility through the canopy across southern France was superb. The canopy is covered by a gold film to make it opaque to radar energy (and thereby reduce the aircraft's radar signature). This coating gave the effect of looking at the world through pale sunglasses.
At a typical cruise speed of M0.82/347kt, the aircraft could sustain a 60º banked turn at maximum dry power. Slamming the throttle to maximum reheat and rolling quickly into a full stick-back hard turn to simulate a break away from a threat gave a rapid response, automatically limited initially to 18.8º alpha and 4g. As the turn progressed, the FCS allowed the incidence to increase to 19.2¼ alpha as the airspeed decayed. Again, I was impressed with how easy it was to extract the maximum performance from this heavily loaded aircraft.
Rolling back to wings-level flight, I re-engaged reheat for a level acceleration from M0.6 to M0.9 in 35s. As the aircraft approached the Mach limit (M0.9) for this configuration, there was a slight airframe or aerodynamic rumble from the external stores. At M0.9, I quickly retarded the throttle to idle, without any noticeable trim change, and selected the airbrakes out to decelerate the aircraft rapidly back to M0.6 in 37s. There was no trim change with engine thrust variations at this or any other time during the flight.
The Rafale airbrake is worth more attention than such a mundane control would normally be given because it illustrates the flexibility of a digital FCS system. Usually, airbrakes are large panels forced into the airstream by hydraulic rams. In the Rafale, airbrake selection deploys the two elevons on each wing in opposite directions, while the canard adopts a leading edge-up attitude - a remarkable sight in the cockpit mirrors. This causes four of the six large control surfaces to move significantly, and the only noticeable effect, apart from the aircraft slowing down, is a very slight lift increase.
The advantage is a saving in weight and system complexity - although I use "complexity" in terms of the hydraulic and electrical systems needed to drive conventional airbrakes. I imagine the mathematics within the FCS to achieve all this are eye-wateringly complex. The system is effective and unremarkable to the pilot, however, apart from a mild aerodynamic rumble.
Simulating the end of the cruise portion of a high-low attack profile, we descended at idle thrust into the low-level part of the flight over the picturesque ridges and gorges of southern France. On the way down, I reselected the autopilot and the flightplan route and engaged the terrain following (TF)system at a set clearance height of 500ft above ground level (AGL). The aircraft descended at 14º FPA and began to level off automatically as it passed 1,500ft AGL. I engaged the autothrottle at 400kt and sat back with my hands off, but close to, the controls as the aircraft followed the scheduled route across rugged terrain.
Despite many years' experience of flying TF systems, I still found it impressive to allow an aircraft to fly hands-off close to the ground through mountain passes and across ridges. The autopilot used from 3g to 0.2g and accurately crossed ridgelines at 500ft AGL. It is anticipated that the TF system fitted to the Rafale BO1 will eventually be cleared to 100ft AGL over land and 50ft over water.
The Rafale TF system uses a radar altimeter as the primary sensor and a digital data map of the earth, rather than a radar system. This has two advantages. Firstly, it eliminates the radar emissions that can be detected and jammed by an enemy. Secondly, the TF system has information about the terrain profile all around and can manoeuvre the aircraft to the maximum allowed either via the autopilot or manually by the pilot. During this flight, the aircraft, under autopilot control, crossed a ridgeline in an 85º banked turn at 3g - a manoeuvre that would surely get a pilot's attention at night or in cloud.
The aircraft was also easy and straightforward to fly manually at low level, with or without the autothrottle engaged. The ride quality through the light low-level turbulence was good - helped, no doubt, by the heavy gross weight. Originally, the Rafale FCS was designed with a turbulence suppression mode, but this has not been found necessary in practice.
The cockpit environment at low level was comfortably cool and quiet. I had no difficulty in communicating with Kerherve or listening to the two radios, although the selectors and volume controls were slightly hidden by the throttle mounting platform.
Much as I was enjoying flying at low level, I wanted to finish my investigation of the aircraft's handling before recovering back to Istres, so, with maximum dry power applied, a climb at 16º FPA was begun, initially to 5,000ft. This brief check in the climb profile allowed time for a dry power level acceleration from 309kt to 460kt in 35s.
When cleared by air traffic control, the Rafale was further climbed into the height block between 5,000ft and 10,000ft. Once level, it was accelerated to M0.88 for a hard turn using full reheat to the FCS g limit. Although I entered the turn quickly, the voice warning (female) informed me that I had slightly exceeded the configuration limit of M0.9 (it was M0.91). The FCS limited the aircraft to 5.2g.
Once in the turn, I adjusted the roll and pitch attitude so that the aircraft decelerated, still turning with full back stick, so that, at 330kt, the FCS transitioned from the g to the alpha limit of 20.8º, an incidence that was maintained until I rolled out at 200kt. Finally, to give the FCS a further hard test, I made full-stick rapid rolls with the stick held fully back. At the incidence limit, the aircraft took 6s for a 360º roll and 5.5s at the g limit of 5.4g. The rolls were smooth and the roll rates even. Given the configuration, this is an excellent performance.
Return to base
On the way back to Istres, the Rafale was slowed down with the undercarriage lowered (taking 5s) to full back stick. With 1,410kg of fuel remaining (aircraft weight 15,900kg), the minimum speed was 120kt at 18º alpha. As noted before, the ideal approach incidence is 16º alpha and, above this incidence, the control column must be deflected aft of neutral. At 18º, the voice warning reminds the pilot to reduce incidence.
The final manoeuvre before entering the circuit was to loop the aircraft from 3,200ft. As with inverted flight, I suspect only test pilots would expect an aircraft to loop while fitted with two cruise missiles and three fuel tanks. The minimum entry speed was 360kt, but I elected to use 390kt to give myself a slightly wider margin in view of the aircraft's heavy configuration. Using 4.5g at the entry to the loop and full reheat until pointing vertically down, the manoeuvre was easy to fly and totally undramatic. Without trying to minimise the size of the loop, the maximum altitude was 9,500ft and the aircraft was back in level flight, having gained 1,000ft on the entry height.
Compared with simple general aviation aircraft, modern automated fighter systems are easy to manage in the circuit. Two right-hand circuits were flown to Runway 15 with a wind of 190º/16 kt, gusting to 20kt. At Kerherve's suggestion, during the first circuit I used the autothrottle to maintain 16º alpha from midway along the downwind leg all the way to touchdown - this constant incidence technique is favoured by naval pilots.
Although you could invent more checks, the only really important actions are to put the undercarriage down and check the fuel. The aircraft remained easy and straightforward to fly, the HUD helping considerably with the approach. The aircraft symbol was displayed on the HUD velocity vector and there were 3º descent markers. The optimum approach incidence was shown by two bracket symbols, which were either side of the aircraft symbol at the correct incidence. All the pilot has to do is adjust the flightpath so that the 3º markers are beside the touchdown point, put the velocity vector on the threshold and control the speed to achieve the approach incidence. Perhaps that sounds difficult, but in practice it is straightforward in the Rafale or any other aircraft.
Shortly before touchdown, the velocity vector was raised from the touchdown point to about halfway down the runway to give an easy flare into a perfect touchdown at 132kt. Applying full throttle to execute a touch and go tripped out the autothrottle.
The final circuit to land was flown using manual throttle control, which I found no more difficult than with autothrottle and allowed me to fly the final turn at a slightly higher speed and lower incidence. After landing, the gusty crosswind held the right wing up until I positively de-rotated the aircraft on to the runway.
The total flight time was 1h and the landing was made with 690 kg of fuel remaining. As I taxied back, a Mirage 2000 was beginning another rehearsal for the Paris air show.
A modern fighter is a means of bringing sophisticated weapons and defensive systems to bear on the enemy. To be operable by a single person, the aircraft must be easy to fly, not only in the cruise but also at the limit of performance in any configuration. The automatic systems must be easy to manage and understand so that they help the pilot to achieve the primary mission objectives rather than become an end in themselves.
To give a practical example, perhaps the best compliment I can pay to Snecma is that I never gave the engines a second thought after starting them. When I wanted more thrust, I put the throttle forward; when I wanted less, I throttled back. While I was not particularly heavy handed, I was not gentle either. The engines responded quickly and without fault, and as an ex-fighter pilot, that is what I want.
As can be seen from this test, the Rafale FCS is well designed, developed and built. Even in this heavy configuration, the aircraft is easy to fly and to manoeuvre to its limits. The short displacement controls were easy and natural to use, so that, within a few minutes of take-off, I was not aware that I was using a side stick. The autopilot is well integrated and will greatly reduce the operational pilot's workload.
During this short test, I could not complete a thorough evaluation of all of the displays and automatic system functions. However, those parts I did use were logical and quickly understandable. The ergonomics of the cockpit were generally good and, although some of the side panels were hidden, I could usually find what I needed fairly quickly.
Of course, this test addressed only part of the weapons system that makes up a modern fighter such as the Rafale. To a pilot, the aircraft is arguably the most important part, and must inspire confidence that it will function reliably and perform better than the opposition.
In the heavy and high-drag configuration tested, the Rafale performed and handled exceptionally well. I am certain that the French navy and air force pilots who will soon fly the production aircraft will enjoy the experience as much as I did.