Flight International's test pilot is the only person to have flown all three of Europe's fourth-generation fighters, the Eurofighter, Dassault Rafale and now, the Saab Gripen

Chris Yeo/Linköping


The problems confronting a fighter design team are broadly the same, whatever their nationality. The best solutions to maximise aerodynamic performance while minimising weight and cost depend on the technology available at the design freeze. It is no coincidence that Europe's three fighter aircraft, the Dassault Rafale, the Eurofighter and the Saab Gripen, are unstable canard delta designs with fly-by-wire flight-control systems (FCS), airframes with a high percentage of advanced structural materials and electronic display-dominated cockpits. The approach to detailed problems, however, such as integration as a weapon system, the pilot/aircraft interface and the incorporation into the national defence structure, varies considerably.

The Gripen, the latest in a series of Saab fighters for the Swedish air force, was specified and ordered into development before the end of the Cold War when Swedish national policy was one of neutrality. The systems' design addressed Swedish defence priorities in a logical way and was not influenced by NATO compatibility considerations.

The aircraft flown for this evaluation, a Swedish air force two-seat JAS39B, is the product of this design philosophy. After the Cold War ended, Swedish policy became one of co-operation with NATO, and Gripens will be modified to be interoperable. Saab, with BAE Systems, is pursuing export sales. These factors will markedly change the equipment and displays fitted to aircraft delivered to the Swedish air force after April 2002 and those destined for export customers.

Flying any high-performance fighter safely and efficiently requires careful preparation and thorough briefing. This was particularly so for this flight, as the Gripen evaluation was to be flown from the front cockpit, which had controls not available to the Saab test pilot, Magnus Ljungdahl, in the rear seat. Since these include the back-up throttle, engine fuel-control computer mechanical and electrical over-ride selections and emergency undercarriage lowering, all denoted by labels using Swedish acronyms, it can be appreciated that both pilots wanted to be assured that the evaluation pilot could react to an emergency quickly and without error.

The evaluation started in the simulator, where flight profiles and emergency procedures were rehearsed. As with any modern fighter, the Gripen makes use of hands on throttle and stick (HOTAS), and the simulator quickly allows familiarity, if not infallible dexterity, with the switches. The simulator provides a faithful mimic of the aircraft's control characteristics other than in-ground effect during landing.

Preparation also included flying clothing fitting, necessary for the pilots to communicate, to operate effectively at sustained high g and to survive an ejection into a cold sea. In the Gripen, the clothing includes immersion gloves and suit with air ventilation, calf-length boots incorporating leg restraints, a full-coverage lower body g suit, a jacket with arm restraint, flotation and survival aids, as well as an upper body g suit (counter pressure), a helmet (without chin strap) and mask. The equipment can be donned unaided in about 10min and is reasonably comfortable.

During the evaluation, ceiling and visibility were good, with an air temperature of 2°C, and a wind of 230°/14kt (25km/h) at Saab's Linköping test facility. The runway surface was dry, although there were some frost and ice patches on the parking apron. The aircraft's configuration was two dummy wingtip missiles, a full centreline 1,100 litre (290USsgal/ 880kg) fuel tank and full internal fuel (2,280kg). The empty mass was 7,100kg, making take-off weight around 10,600kg. The single-seat JAS39A Gripen is 0.7m (2.29ft) shorter, weighs 600kg (1,322lb) less and carries 120kg more internal fuel than the two-seater.

Compared with other fighters, the Gripen is a small aircraft and the whole airframe is visible during the walk-round inspection. The cockpit is also small, but well laid out, with essential operational controls readily to hand. Several switches at the back of the port console are difficult to access since downward vision is partly restricted by the flying clothing. These switches, however, are different in shape and action and are guarded, so there is little risk of inadvertent operation, although care is required when operating the seat raise and lower switch, which is tucked away at the rear of the console. Once strapped in, the cockpit is comfortable, with the three electronic head-down displays (HDDs) and head-up display (HUD) clearly visible.

Ground independence

Pre-start checks were brief. The electrically powered, sideways hinged canopy was lowered and locked down and the auxiliary power unit (APU) started. The APU provides hydraulic and electrical power, as well as air conditioning, and makes the aircraft independent of ground support for road operations. After bringing the electronic system and FCS on line, there was a 56s interval while the inertial navigation system initialised and aligned. After checking the throttle for full and free movement it was set to ground idle and the single Volvo RM12 engine started.

The start was smooth and rapid, with the engine soon idling at 60%/400°C. The after-start checks were simple as the utility systems are highly automated and little, if any, pilot intervention is either possible or necessary. There is, however, an extensive range of automatic built-in tests that are reported to the pilots, showing system functionality and serviceability. After a short delay while a final system finished its self-check, the aircraft was taxied to the runway 11 threshold. A small throttle movement to just above flight idle was necessary (ground idle opens the reheat nozzle to reduce thrust) to get the aircraft moving, but it taxied at ground idle. The nosewheel steering is scheduled with ground speed and is powerful and accurate; below 11kt, full nosewheel deflection is available and the aircraft can be turned in a small radius. The toe pedal-operated wheelbrakes react slightly abruptly to small movements, but were satisfactory in other respects.

Once lined up on the threshold, full dry power was set, with the wheelbrakes firmly applied. The left-hand electronic display includes a vertical scale indicating throttle angle and, on the ground only, actual thrust achieved as a percentage of the maximum that should be available. This is useful since the pilot can rapidly determine engine serviceability from a single display. The minimum acceptable achieved thrust for take-off was 90% on this occasion, and the engine stabilised at about 100%. The brakes were released and, as the aircraft moved forward, full reheat was selected by moving the throttle fully forward through a soft detent.

Reheat light-up was smooth and progressive and the Gripen accelerated rapidly. A nozzle area gauge is not fitted but, given the power-to-weight ratio of the aircraft, there was little doubt that reheat was functioning. Rotation speed is indicated in the HUD by a small vertical marker moving up towards the aircraft symbol. The stick was moved back at 134kt, almost full back stick was used to raise the nose and the aircraft left the ground at 154kt. It was hard to estimate the ground roll accurately, but it was around 600m. Once airborne, the aircraft was easy and natural to fly, with no sign of over-controlling. The undercarriage was retracted and reheat cancelled - both without any trim change. Altitude was gained at the best climb speed of 380kt to 9,000ft at a 15° climb angle. Start up, take-off and the climb to the cruise altitude used 14% internal fuel. Unfortunately, the HUD displays either indicated airspeed or Mach number, but not both. The automatic change-over point is close to the normal climb speed and it was necessary to refer to an HDD to obtain either Mach number or airspeed. It is usual to display both on a HUD, and this would improve the display and reduce the need to scan regularly across two instruments.

Display quality

En route to a rendezvous with another Gripen, a brief check of the FCS was made and the electronic displays viewed in bright sunlight. All displays were adequately bright and the writing quality excellent. The test aircraft's displays were green monochrome, although export and batch three aircraft will have colour displays - which will make it easier to assimilate the quantity of data available. The map alphanumerics remained north orientated, although the map was track orientated, making it more difficult than necessary to read the data. While this may not be a problem for a pilot flying over a familiar area, it did not aid rapid assimilation of map or datalink information.

The control column is a small centrally mounted "mini-stick" pivoted just below the pilot's hand. The aircraft was flown by wrist action for small movements, with the addition of some arm action for larger movements. It was comfortable and there was no tendency towards overcontrol or pilot-induced oscillation. The FCS provides neutral longitudinal static stability with the undercarriage up and positive static stability undercarriage down.

An examination of the air-to-surface and air-to-air weapon systems was planned. The first exercise was a simulated attack on a ship in the Baltic, using a Saab RBS15 sea-skimming missile targeted using information datalinked from another Gripen. The Swedish air force makes extensive use of air-to-air and air-to-ground datalinks - a key part of the strategy to share information widely between friendly forces. As the aircraft was positioned for the ship attack and throughout the rest of the flight, the map display showed the heights and tracks of traffic flying along an airway across southern Sweden, as well as two other Gripens. The position of the target was passed from the other aircraft and shown on our map display. The anti-ship missile was selected by a single switch action on the stick. Immediately, the missile's flightpath directly to the ship was marked on the HUD and HDD with maximum and minimum firing ranges and the start of sea-skimming mode. The missile could have been fired immediately as the maximum range was passed, but other options were provided such as introducing a waypoint into the missile's flightpath and adjusting the sea-skimming start point. The attack was set up with little radio discussion between the two aircraft. Such an attack poses a considerable problem to a ship's self-protection team as the missile is fired from a passive aircraft with an accurate target position and the missile can fly a non-direct track to the ship.

All Gripens are multirole. After completing the ship strike, the aircraft was reconfigured for the air-to-air role by pressing a single button. Our friendly wingman became hostile and was shown as such on the datalink display, although the attack system recognised the transponder as uniquely Swedish and indicated this. The target was designated as something of interest to the track-while-scan radar using a hand controller on top of the throttle.

The HOTAS is configured with weapons selection functions on the stick, while the throttle has two pistol grips. The upper has the sensor controls (radar range, scan angle and, in the future, forward-looking infrared) and the lower grip some of the navigation functions and the airbrake selector. This may sound complex, but it was not. While I did not become totally adept at manipulating the various buttons quickly, I could select most modes and functions when required. To engage the target, full dry thrust was set, the aircraft taken to 13,120ft and accelerated to a more tactical speed of M0.88. The target aircraft was clearly visible on the radar ahead and 1,640ft above, cruising at M0.7.

The first attack was made head-on with a Raytheon AIM-120 AMRAAM. The engagement zones, including the no-escape zones, were shown on the display ahead of each aircraft. The first missile was "fired" at an unrealistically close range of 15km (9 miles) and the aircraft was turned rapidly away from the target before reversing into a hard turn to astern of the target for a Raytheon AIM-9L Sidewinder shot. Changing missile types was a single switch action on the stick top that also reconfigured the HUD and radar. Pulling up towards the target at 4g/10°a (angle of attack) the missile attack was easy (against a quiescent enemy), but seeing the small Gripen with a commendably smoke-free exhaust was rather more difficult.

Gun attacks

Finally, the aircraft were closed for gun attacks (although only the single-seaters are equipped with a 27mm Mauser cannon) and the target ordered to manoeuvre. At this stage, the fuel remaining had decreased to 105% of internal fuel. At take-off mass, the aircraft was limited to 5g by the full external fuel tank. As the tank emptied below 200kg, the g-limit started to increase progressively to a maximum of 7.5g of the FCS standard tested. The two-seater's limit will be increased to 8.5g in the near future. As the gun tracking exercise began the g-limit was 7g. The g and incidence limits are regulated automatically by the FCS. A tracking exercise is a good test of an FCS since very small and precise changes of flightpath are necessary to keep the gun-aiming reference over the target - particularly when it is manoeuvring. An accurate sight picture was easy to maintain.

The final exercise with the other Gripen was to join for close formation in echelon starboard, echelon port and line astern, the latter being a reasonable simulation of that used for air-to-air refuelling. The lead aircraft flew through weaving manoeuvres of up to 90° of bank and 2-3g. The combination of the excellent flight control laws and rapid engine thrust response made position keeping straightforward.

The formation separated and, for the next few minutes, the other aircraft's progress back to the airfield could be monitored on the datalink display. The evaluation aircraft was taken to 26,230ft for a handling and performance examination. Some performance data normally given in a flight report is classified and was not recorded. Subjectively, the Gripen's performance is comparable to its peer group, with roll acceleration and roll rate perhaps superior.

The displays include many useful features. Two in particular were used at high level. The first is economical cruise speed shown on the HUD and HDD speed scales when selected by the pilot. The second is the corner velocity where the FCS will change from g-limiting to angle-of-attack limiting, which is always shown on the same speed scales. In this case, with 92% fuel remaining, the economical cruise speed was M0.78M/313kt and the corner velocity was M0.9, at which speed the airframe limits were 20°a and 7.3g. The HUD, with several differences from other Western displays, was clear and easy to interpret.

A steady climb and level-off, followed by level turns at 45í angle-of-bank (AOB), were easy to fly accurately. At full dry thrust (below the optimum speed), the aircraft comfortably sustained a 60° AOB turn. As the first FCS test at a limiting condition, the aircraft was accelerated to M0.8 and then aggressively turned, using a fairly rapid application of full back stick and full reheat. The aircraft was limited positively at 19°a/4.5g without overshooting. The final test at this height was to accelerate in full reheat to M0.95 and then close the throttle rapidly to idle while simultaneously extending the airbrakes. There were no trim changes and the aircraft remained easy to fly accurately.

The aggressive handling tests continued at 6,560ft, where the aircraft (with 73% fuel) flying at M0.9 was put into a full back stick turn, using maximum reheat. The FCS limited the aircraft to 7.5g at 9°a. The turn rate was impressive and, although there was some speed decrease, it was not excessive. The turn was repeated using less thrust and the stick held fully back so that, as the airspeed decreased, the FCS changed from g to incidence limiting. This occurred smoothly at 335kt. The g and incidence limits will be increased to 26°a and 8.5g (the single-seat aircraft will be limited to 9g) by updated FCS software due for release next year.

Full stick rolls through 360° were made at 1g, 3g and 5g at M0.8. The aircraft's roll acceleration and roll rate were unusually and impressively rapid at 1g and 3g, subjectively more than 200°/s. The roll rate decreased somewhat at 5g/7°a, but was still good. The position of the airbrakes can adversely affect directional stability of an aircraft, even if it does have an FCS, so a 360° roll was made with the airbrakes extended. There was no noticeable change in roll rate or acceleration. Rolls were not made at the incidence limit since the FCS software used during this flight was not cleared for rapid rolling at full back stick (ie, carefree handling). This restriction will be lifted with the next software release.

While still at 6,560ft, the aircraft was slowed in level flight with the undercarriage down to the minimum approved speed of 124kt; this equated to 17.5°a with 61% of fuel remaining. With the undercarriage down, the FCS introduces pseudo longitudinal static stability and it was necessary to trim the aircraft in pitch as one would do with a conventional aircraft. This is an excellent feature since the out-of-trim forces give a good tactile cue if the aircraft's speed varies from that trimmed for. This characteristic is a valuable aid during a high workload instrument approach. Handling during turns with the undercarriage down remained easy and pleasant and the aircraft could be flown accurately without difficulty. In this configuration, the incidence limit remains 19°a and continues to be automatically controlled, but a 3g limit must be observed by the pilot.

Finally, at 6,560ft, the aircraft was inverted at 350kt several times and a push-up made to -2.5g/-3°a, just below the -3g negative limit. A slight forward pressure on the stick was required for level inverted flight, but in all respects the handling remained excellent.

No fighter flight would be complete without a few aerobatics, so a couple of loops were flown using maximum dry thrust. Entering the manoeuvre at 384kt with 51% fuel and without seeking to maximise performance, the looping radius was 5,900ft, with a minimum speed over the top of 154kt. As only 4g was used and the incidence limit was not approached, this was a relaxed manoeuvre and the radius could easily have been reduced significantly.

It may be a surprise that remaining fuel quantity is quoted in percentage of full internal fuel, the measure used by the Swedish air force. I found this unusual system reasonably practical and fairly logical for a fighter that will almost always be completely refuelled. The Gripen, however, does not have a fuel-flow gauge, a facility I would miss in service, particularly during long sorties close to the range limit. To counter this argument, the navigation system continually calculates the fuel to complete the planned mission, perhaps rendering a flow meter desirable rather than essential.

The final part of the evaluation was at low level, 656ft, at a typical low-level cruise speed of 450kt. The conditions were clear, but there was moderate turbulence. The ride quality was good and it remained easy to fly in straight flight and in turns to 3.5g. The view from the cockpit is excellent and the environment comfortable. The HUD has a declutter mode to remove the less relevant symbology while at low level. A simulated pop-up rocket attack was made, turning left towards a small island to test the aircraft's tracking qualities through the low-level turbulence. These were found to be good.

We returned to Linköping, initially for an instrument approach, using the Tactical Instrument Landing System (TILS). From the pilot's point of view, this unique-to-Sweden system is much like a standard ILS with the localiser offset 3í from the runway centreline. The first part of the approach was made to a standard initial point using the aircraft's navigation system. Thereafter, the aircraft's receiver used the localiser and glidepath signals to drive a flight director in the HUD to guide the aircraft to a decision height of 197ft.

In addition, the computer used the localiser and glidepath to calculate that there was a small error in the navigation computer position and inserted a fix. The approach to the tactical instrument landing system initial point (TILS IP) was made at 297kt, reducing to the normal approach incidence of 12ía (154kt) on the glidepath. The autothrottle held the approach incidence within ±1° despite the moderate turbulence. The TILS provided accurate guidance to the decision height, at which point the aircraft was turned to the right to align with the runway and a touch-and-go made.

Strong ground effect

Three more circuits were flown. The final two used 14°a on the final approach, the incidence used for road landings. This reduced touchdown speed by 16kt. The aircraft was easy to fly around the visual pattern and could be placed accurately on the runway. In common with other delta aircraft, there was a strong ground effect when the aircraft was at a height of about one wingspan above the runway. This flattered the pilot by making each touchdown quite gentle, even when a firm landing - rather than a runway consuming float - was required. Clearly, with a little experience, accurate touchdowns at a good sink rate would be achieved consistently and without difficulty. The final landing used maximum brake effort from a 14ía approach. The brakes have touchdown protection, so the wheelbrakes were fully applied before landing. The landing distance was commendably short although the wheelbrakes snatched quite heavily just after touchdown, perhaps because the touchdown was not hard enough, and once again when reapplied during the landing roll. The Gripen is equipped with brakes on the nosewheels as well as the main wheels, the operation of which was transparent.

The final landing was made with 19% fuel remaining after a flight time of 1h 10min.

In summary, the Gripen was easy and very pleasant to fly. It has good performance and provides a comfortable working environment. The high level of utility systems automation has been well thought through and frees the pilot to get on with operational tasks. During this flight the aircraft and engine performed faultlessly. Although the tests of the operational systems were of necessity superficial, they served to show that the aircraft has a logically organised, competent and sophisticated weapon system that will allow use in the fighter, attack and reconnaissance roles. In particular, the datalinks are fully integrated into the weapon system and multiply the effectiveness of the aircraft.

The Block 3 developments should enhance the aircraft's capabilities and will make it a serious export contender. No aircraft is fault free, but the items that have been commented upon are either personal preference or minor and could be easily corrected. They do not detract from the favourable conclusion.

Source: Flight International