The success of Northrop Grumman's Global Hawk is helping transform opinion on the military capability of unmanned air vehicles


Soaring in from the vast wastes of the Pacific Ocean like a 21st-century Wandering Albatross, the Global Hawk points its bulbous nose towards the Australian coastline and steers a course across the country for South Australia and an air base near Adelaide.

All being well, this will be the scene in mid-April, when one of the most unusual aircraft ever built will complete one of the most ambitious flights ever flown - the crossing of the Pacific by an autonomous unmanned air vehicle (UAV). But this dramatic flight by a Northrop Grumman RQ-4A Global Hawk is not about breaking records, it is about proving the tactical and strategic value of long-range UAVs.

With a string of similar militarily useful achievements to its credit already, including a trans-Atlantic reconnaissance sortie to the Portuguese coast and back in 1999, the Northrop Grumman Global Hawk team is determined to try to make the Australian exercise a showcase of UAV maturity.

Global Hawk is expected to participate in up to 12 exercises in Australia, at least six of which will be part of Operation Tandem Thrust, a joint US/Australian exercise involving special operations forces and Pacific Command on the northeast coast. The UAV team will work with the Australian Defence Science and Technology Organisation (DSTO), which is providing a prototype ground station for image exploitation over the duration of the visit. This will also include coastal patrol missions over the country's remote northern littoral zones.

In some ways, the Australian mission represents a coming of age for the Global Hawk, which traces its lineage back to 1993 and pioneering work by UAV specialist Teledyne Ryan Aeronautical (TRA). "It was then we had a dream about developing a high-altitude, long- endurance [HALE] UAV," says Northrop Grumman vice-president for Global Hawk business development and Norm Sakamoto, an acknowledged father of the UAV.

Using the twin-boom, pusher-propeller Model 410 as the basis for its bid, Ryan entered a competition that year which was ultimately won by the General Atomics Predator. "We lost the competition, but we didn't lose the aircraft, which took some good imagery," says Sakamoto. More importantly, the event prepared Ryan for the request for proposals (RFP) issued in March 1994 by the US Defense Advanced Research Projects Agency (DARPA) for a HALE UAV.

The RFP was prompted by glaring shortfalls in real-time and consistent reconnaissance data exposed in Operation Desert Storm, and was responded to by 14 contractors. Bids called for a remarkable machine capable of carrying a 1,000kg (2,200lb) payload for more than 40h at altitudes of up to 65,000ft (20,000m).

Known as Tier 2 Plus, it would be one of a range of UAVs planned by the US Defense Airborne Reconnaissance Office (DARO). The first of these, the Tier 1 General Atomics Gnat 750, was already in service with the Central Intelligence Agency, spying on troublespots such as Bosnia. A contract for the Tier 2 UAV, which could monitor a small, 34,200km2 area for up to 24h from altitudes as high as 25,000ft, had also been won by the General Atomics Predator.

The new UAV would work closely with a third type of aircraft dubbed Tier 3 Minus. This stealthy, 12h endurance UAV was planned to operate up to 50,000ft carrying a 500kg payload. Developed to an unusual, low-observable (LO) design by Lockheed Martin and Boeing, the Tier 3 Minus UAV was called DarkStar and was optimised to gather intelligence before suppression of enemy air defences (SEAD) had taken place.

In May 1995, a TRA-led team including E-Systems as the sensor package supplier won the Tier 2 Plus competition and launched into development of what was to be called the Global Hawk. Even at this stage, Congressional plotting looked set to kill the fledgling bird before it had even taken its first flight.

Proposals within Congress threatened to combine DarkStar and Tier 2 Plus in fiscal year (FY) 1995, and kill the Global Hawk in FY96. Following mishaps and accidents to the more glamorous looking DarkStar, however, Tier 3 Minus was axed and its role modified and absorbed by Global Hawk.

In October 1998, the programme transitioned from DARPA to the US Air Force's Aeronautical Systems Command at Wright-Patterson AFB in Ohio, under the continuing auspices of a fast-track advanced concept technology demonstration (ACTD) effort.

This had been launched as a way of rapidly developing a system and bringing it to military users for hands-on experience before production was committed. It also contained a crucial target - a unit flyaway cost of $10 million (FY94 dollars) for the second production batch of 10 or more vehicles, including payload. Since then, Sakamoto says, "the ground rules have changed. But we are managing to keep costs in the region of $15.3 million per copy according to a report from the General Accounting Office [GAO]."

The report by the Congressional watchdog acknowledges Northrop Grumman's achievement in keeping the cost relatively low despite changing procurement conditions, principally a much lower acquisition rate. The plan originally called for the USAF to buy two Global Hawks, followed by eight then a batch of 10. "They've actually bought two, then three and then another two. The GAO says under the circumstances we've done really well," he adds.

Successful compromise

As with every aircraft, the Global Hawk emerged from a complex matrix of competing requirements. The winning design is the 12th of 13 concepts studied, ranging from multi-engined, multi-tailed designs to a flying wing (the 13th).

All were aimed at meeting the principal target of flying at 65,000ft for 24h after travelling 1,600km (3,000nm), with the ability to fly a further 1,600km to return to friendly territory.

Low observability was not a design driver, simplifying the task for TRA, which saw its loiter profile above 60,000ft as its best protection against most ground-launched weapons and subsonic air-launched missiles.

Another key driver was the relatively large sensor payload which included electro-optical (EO), infrared (IR) and synthetic-aperture radar (SAR). Although the weight of the package constituted a formidable challenge in its own right, the main design focus was on finding suitable locations on the airframe where the sensor apertures could function the best. "So we took the sensor sizes and scrutinised the industry to see who had what, and we figured out how much fuel you needed to get it all around," says Sakamoto.

This drove the overall size of the Global Hawk, which has a large, 35m (116ft)-span wing and a relatively stubby 13.5m-long fuselage. The thin, slightly swept wings have sufficient capacity, combined with fuel tanks in the fuselage, for almost 6.8t of fuel and have an area of 50m² (540ft²) giving the aircraft a glider-like aspect ratio of 25.

"We needed something that would give us good high-altitude performance, so our advanced development section went to NASA Ames where they let us test several [aerofoil] sections in the windtunnel," he adds. Some of the work built on experience with TRA's similarly proportioned, 25m-span Model 235 (Compass Cope R) developed in the mid-1970s, as well as later studies including the Model 275 and 329, though the latter two were never built.

The section finally chosen, with its laminar flow and supercritical characteristics, was designed to generate a lift/drag (L/D) ratio of 37. Subsequent flight tests showed an L/D of "more like 34 and maybe 33," Sakamoto says.

As a result, Northrop Grumman is examining revised aerodynamics for the leading edge, as well as potentially increasing inverse camber on the aft underside of the trailing edge. Leading edge changes could include reducing the bluntness and will be based on chordwise pressure data now being collected on test flights using specially mounted pressure strakes. The revised design will retain the existing 5.9¹ sweep angle measured at the 25% chord point.

Lightweight structure

The wing is also lightweight, being constructed entirely from carbonfibre-epoxy composites. Running through the wing, which is designed in sections (a 15m-span centre, two 10m outer panels and wingtip assemblies), are four shear spars which bear a closer resemblance to shear webs than conventional spars. Together with the high-modulus composite skin, the result is structure capable of carrying loads from the 4,000kg (8,800lb) of fuel stored in the two main wing tanks. When the Global Hawk is fully fuelled, the tips sag down by 0.3m.

While weight and performance characteristics meant an emphasis on composites for the wing, the cost implications determined that the fuselage structure should be a conventional aluminium monocoque.

"We looked at designing the whole thing out of composites to save weight, but one of the problems with that was a lot of non-recurring costs to develop the tooling. So we judged it was too expensive and would be at about the 12 to 15 airframe point before it got cost effective," says Sakamoto. The mostly aluminium fuselage accounts for around 35% of the structural weight of the entire aircraft, adds deputy chief engineer Alfredo Ramirez.

The fuselage also supports the aircraft's single Rolls-Royce AE3007H turbofan, which is mounted above and aft. "We looked at multiple engines, a lower engine location and bifurcated inlets before settling on this," explains Sakamoto. "We liked the engine on top because, if it is single engined, it had to be either inside the airframe or outside.

"If inside, it occupies around a third of the internal volume and you lose all that space for equipment. The other option was to put it in a pod underneath, and we didn't want to do that in case we had to belly land - and engines are very expensive." The only major drawback with the upper location was maintainability, but engine reliability levels are generally described as "excellent" with few reasons to make the journey to the upper fuselage.

The engine was chosen because of its "good heritage" (it is developed from a common core used in powerplants for the Lockheed Martin C-130J and Bell-Boeing V-22, as well as being virtually identical to the AE3007 turbofans powering the Embraer ERJ-135/140/145 regional jet family and Cessna Citation X business jet) and "because it had the best specific fuel consumption/thrust performance at altitude we could find at the time. Plus they had a lot of experience that gave us confidence," Sakamoto says.

Before flying on the Global Hawk, the engine was tested to 71,000ft equivalent altitude at the USAF's Arnold Engineering Development Center, Tennessee. "However, it showed a disconcerting tendency to surge at the highest altitudes, so we decided to limit it to 65,000ft," Sakamoto says. The aircraft is programmed to throttle back the engine in a slow "roll back" as it approaches this altitude. "The highest we've been is 66,400ft," he adds.

Aft of the engine, mounted either side of the exhaust, are the Global Hawk's distinctive V-tails. Set at a dihedral angle of 50°, the 3.5m-span tails have an aspect ratio of 3 and each covers an area of 4m². The design was chosen because two surfaces were "cheaper and lighter than three", adds Sakamoto. "Although vertical tail volume is good for directional control, we needed to keep it away from the ground and if we put in a conventional single tail, we needed some sort of fancy S-shaped exhaust. That was too heavy."

Loads from the glassfibre-composite V-tails, made by West Virginia-based Aurora Flight Sciences, are absorbed into the aircraft's broad "boat tail" aft fairing. This is also constructed of lightweight glassfibre like the "turtleback" fairing covering the large, 1.2m-diameter wide-band satellite communications (satcom) antenna mounted over the forward fuselage.

The satcom antenna was sited forward to provide a counterbalance to the engine, as well as to give it the best uninterrupted sky-viewing angles. "The other reason was that we don't have to worry about distortion effects on the inlet, and we did a lot of windtunnel work [at a former Rockwell North American site in Los Angeles] to prove it," Sakamoto says.

Experience shows

Compass Cope experience also helped determine the tricycle gear configuration, which was chosen over an alternative solution involving podding the main gear in the wings. The main undercarriage is an off-the-shelf unit taken straight from the Bombardier Learjet 45, while the nose gear is a two-position unit from the Canadair CF-5F. The nose strut can be extended to provide a higher angle of attack for take-off.

Compass Cope also provided the foundation, in principle at least, for many of the key systems found in Global Hawk: "We learned a lot of things on that programme, like redundant flight control system [FCS] architecture, hydraulics and pneumatics," says Sakamoto.

The dual-redundant FCS is controlled by two onboard flight control computers which receive constant input from the aircraft's suite of navigation and air data sensors. This includes an inertial navigation system, inertial measurement unit and global positioning system (GPS). All provide input to update the flight control computers, which are pre-programmed with a flight plan before departure. The UAV navigates via GPS waypoints, and has several built-in default modes which are activated if necessary.

No flight control commands are accepted by the Global Hawk until after take-off. Once airborne, the flight is controlled and monitored by the launch and recovery element (LRE) - part of a suite of command and control (C2) and mission planning equipment developed especially for the HALE UAV by Raytheon. Communicating initially to the UAV via a line-of-sight (LOS) common datalink (CDL), and then by Ku-band and UHF satcom, the LRE hands over to the Mission Control Element (MCE) for the actual task of completing the mission. Ku-band and CDL are mostly used for data transmission, including threat information and UAV status, while UHF is mostly used for command and control.

As the UAV ascends and crosses controlled airspace, the LRE and MCE crews communicate with air traffic control via VHF/UHF radio. Unfamiliarity in dealing with high-flying, autonomous UAVs has led in at least one case to a controller asking the unmanned aircraft "what is it like up there?" The duty crewman, stationed on the ground thousands of kilometres away, simply replied: "I don't know - I'm not up there!" The Global Hawk is otherwise treated the same as any other aircraft transiting through controlled airspace, possessing an identification friend or foe system (IFF).

Should a problem occur, the Global Hawk FCS has a built-in safety net designed to cope with four main contingencies. If it loses the C2 link, the aircraft is programmed to continue on course for 1.5min before returning to base if no signal is picked up. If a critical system fails, or the diagnostic system detects an imminent unacceptable failure such as a generator going off-line or an inertial measurement unit failing, then the vehicle is also programmed to return to base.

A third failure scenario is an engine flame-out. In this event, the Global Hawk is programmed to search its memory for the nearest "friendly" alternative runway. A windmill restart is out of the question because the slow-flying UAV could never attain the required dive speed. Alternative relight options such as a ram-air turbine, air bottles or auxiliary power unit were ruled out on grounds of cost, weight and complexity.

If no suitable friendly field is judged to be within range, the Global Hawk is programmed to glide to one of several pre-determined optional points and "die", says Sakamoto. Reactive programming is also embedded to cope with a suite of landing and take-off failure scenarios. Global Hawk will abort its own take-off if it begins to deviate too far from the centreline, or it fails to reach V1 (decision speed). On landing, the aircraft will automatically initiate a go-around if it is incorrectly lined up with the runway.

Monitoring status

The status of the vehicle is monitored by a fault log computer that records any problems which are detected during a mission. The results can be downloaded for analysis after landing.

For real-time monitoring, the ground crew watches discrete and continuous signal comparitors. These provide pre-set upper and lower operational limits for every onboard system ranging from engine pressures, temperatures and RPM, to hydraulic pressures, electrical voltage levels and airspeed. "If any of the levels are exceeded, a red light comes on and, if the system is critical to the vehicle, it will start flashing. This tells you it will do something whether you are going to or not - the vehicle will come home," says Sakamoto.

The sensor suite at the heart of Global Hawk requires extensive monitoring and consists of synthetic-aperture radar/moving-target indicator (SAR/MTI) and EO/IR systems. The SAR/MTI antenna is housed in a bulged fairing immediately aft of the nose gear, and provides real-time imagery of the ground in several formats. With a field of regard of ±45° either side of the aircraft, the Raytheon X-band radar can cover up to 138,000km² per day in search mode from a range of 200km.

In ground MTI (GMTI) mode the radar can search up to 15,000km²/min, detecting any targets with a ground velocity of 4kt (7.5km/h) or more from a range of 100km. With a 10m range resolution, the GMTI mode scans a 90° sector, and can be used to cover zones between 20km and 200km either side of the aircraft.

Operators can hand off detected targets to the SAR spot mode for more detailed viewing. Up to 1,900 images per day can be made in spot mode. Each spot covers 4km² and has a range resolution of around 0.3m. In SAR strip mode, the 600MHz bandwidth radar can cover a swath 10km wide with a resolution of 1km.

The Raytheon-supplied EO/IR system, mounted in the chin of the Global Hawk, combines a Recon/Optical camera with a Raytheon IR sensor. The EO system uses a commercial, 1,024 x 1,024 pixel Kodak silicon charge-coupled device (CCD), while the IR sensor has a 640 x 480 pixel 3-5um indium antimonide detector derived from Raytheon's common-module forward-looking infrared (FLIR) system.

Both EO and IR sensors are fed by a fixed focal-length reflecting telescope with a beam splitter. Neither of the systems has the 6,000-plus pixel width needed to provide the required 1m resolution in a single exposure, so the telescope scans continuously sideways while an internal mirror back-scans to freeze the image on the sensor. This "step-stare" approach means the mirror returns to the start point every one-thirtieth of a second, while the small patches are assembled to generate a larger picture.

The entire system uses a gimbal mount derived from a Raytheon AAQ-16 FLIR turret which can roll ±80° or move ±15° in pitch and yaw. Stabilised to 3mrad, compared with the more normal 20mrad, the system can cover up to 104,000km² per day in wide-area search mode, or generate up to 1,900 4km2 spots in spot mode. Dual-band coverage is provided in the visible (0.4-0.8um) and IR (3.6-5um) wavebands.

Imagery from the sensor suite is sent via the Ku-band link originally developed by Loral. The original goal was to transmit at up to 50Mb/s, though projected performance now ranges between 1.5Mb/s and 47.5Mb/s.

Radar imagery output is around 30Mb/s, and can be compressed to 8Mb/s at 2bits per pixel. The EO has a raw data rate of about 40 million pixels/s at 8-10bits/pixel, or up to 400Mb/s. This can be compressed to about 40Mb/s using JPEG techniques. The IR sensor data, which has an output of around 13 million pixels/s, can also be compressed by a factor of 8:1.

The SAR/MTI radar, with a peak power output of 3.5kW and weighing 290kg, requires 4.7kW of 400Hz power and 1.3kW of 28V DC power, while the EO/IR system, weighing 100kg, requires more than 0.58kW of 28V DC power. This relatively large power requirement, added to the needs of other systems on the aircraft, is met by a 28V DC generator mounted on the engine. This develops around 10kW of DC power. A further 8-10kW of power is generated by a hydraulically powered AC generator.

"We have an inverter on board to go from DC to AC, and a transformer/rectifier to go from AC to DC. It normally goes from AC to DC because of critical things like the flight control system," says Sakamoto. Three batteries are also carried, providing back-up power for up to 1h.

"We never have enough power," says Sakamoto, who adds that a power enhancement study is currently under way: "We will need more for future payloads, particularly if we extend the tether and have to fly further, and deeper and higher." Rolls-Royce is studying an engine core- driven generator as one potential new power source for the Global Hawk, which is to be one of the platforms to carry the power-hungry active-array SAR/MTI radar being developed by Northrop Grumman and Raytheon under the USAF's new Multi-Platform Radar Technology Insertion Programme.

To help keep the avionics and sensor system modules warm at high altitudes and cool at lower altitude, air temperature is carefully controlled in a pressurised section of the fuselage. Monitored autonomously by a Honeywell environmental control system built to Northrop Grumman specifications, the system uses the aircraft's own fuel as a heat sink.

Fuel is fed though tubing in the leading edge to the outboard tanks and gravity-fed back to the centre fuselage tank. Two pumps feed the fuel to the engine and excess fuel, which is pumped around the equipment, goes to a fuel/air heat exchanger. "At altitude we need to warm it, and we do this by dumping bleed air to warm the fuel itself. This then pumps around the compartment and warms it up," adds Ramirez.

Northrop Grumman hopes the performance of the Global Hawk in Australia will underscore the positive verdict already conferred on the aircraft after its thorough military utility assessment in the USA. After its Pacific venture, the UAV is set to complete the engineering, manufacturing and development phase before launching into low-rate initial production (LRIP) later this year.

Extending to the end of FY2005, the LRIP phase is expected to cover production of a further eight RQ-4As. Beyond this lie further plans to develop and demonstrate signals intelligence and communications relay versions, as well as potential "Eurohawk" and "Gulfhawk" surveillance variants. By the end of the decade, Northrop Grumman hopes to build its 20th Global Hawk, with plenty more eggs still to hatch.

Source: Flight International