The US Defense Advanced Research Projects Agency is continuing its tradition of looking into some weird and wonderful designs with three "new" concepts
They look familiar because they are concepts that have been around for decades – shape-changing aircraft, oblique flying wings and ultra-large transports. They are back on the drawing boards because, this time around, technical maturity and operational necessity might actually coincide.
The three concepts are being pursued by the US Defense Advanced Research Projects Agency (DARPA), which prides itself in taking on challenging – sometimes baffling – ideas and demonstrating their feasibility and utility. The Department of Defense agency’s successes include the internet and stealth. DARPA has also had its failures, or dead ends.
Shape changing is being pursued under DARPA’s Morphing Aircraft Structures (MAS) programme, now coming to the end of its second phase with windtunnel tests under way at NASA Langley. Phase 1 studies of the Walrus ultra-large aircraft have just begun, and industry proposals for the first phase of the Switchblade oblique flying wing (OFW) project will be received by early next month. All three could result in X-plane flying demonstrators.
The programmes are part of a resurgence in aeronautics research at DARPA at a time when NASA funding for air vehicle technology is in decline. Other aeronautics-related programmes under way include the Boeing A160 Hummingbird long-endurance unmanned helicopter and X-50 Dragonfly canard rotor/wing (CRW) vertical take-off and landing demonstrator; Lockheed Martin’s Cormorant submarine-launched and -recovered multipurpose unmanned air vehicle and Falcon hypersonic cruise vehicle; and Vought’s Kingfisher sea-going UAV (see P50).
What these programmes have in common is that each is built around an idea that, if proved feasible and useful, could transform aircraft design. For the morphing aircraft it is flexible, adaptable structures that enable vehicles to change their form and function. For the ultra-large aircraft, it is the ability to control a combination of aerodynamic, buoyant and propulsive lift. And for the oblique wing it is the unique combination of supersonic speed, long range and endurance.
“DARPA’s task is to take the technology argument off the table,” says Dr Arthur Morrish, director of the research agency’s Tactical Technology Office. “If we don’t, we go back and try again.” The three programmes represent “DARPA-hard” technical challenges, with no guarantee that they will produce viable aircraft.
The first thing Walrus programme manager Phil Hunt says is: “This is not an airship. This is a heavier-than-air vehicle.” Large transport airships have a long history dating back to the Zeppelins of the early 1900s, but advances in fixed-wing aircraft quickly left them behind. The last and largest US rigid airship, the US Navy’s 239m (785ft)-long USS Akron, crashed into the sea in April 1933 during a violent storm. Smaller non-rigid “blimps” continue to be used today and there is a resurgence of interest in high-altitude airships for surveillance missions.
The concept of a large hybrid aircraft, combining buoyant and aerodynamic lift, emerged in the 1990s, and in 1999 NASA announced plans to demonstrate a piloted subscale model of a partially buoyant cargo airship – the Lockheed Martin Skunk Works Aerocraft – under its Revolutionary Concepts (RevCon) programme. But the RevCon programme was cancelled soon after.
The hybrid aircraft concept then arrived at DARPA, which decided it had too many shortcomings and not enough military utility, says Hunt. Instead the agency decided to take the next step by demonstrating the technology for an ultra-large aircraft combining aerodynamic, buoyant and propulsive lift.
Lessons of war
The operational vision behind the Walrus – the ability to mobilise and manoeuvre forces quickly – is based on the lessons of previous wars, says Hunt. To determine whether the concept is reasonable, DARPA employed an analysis tool used by the Joint Strike Fighter programme to look at the needs of multiple customers and identify operational tasks that can be performed by the aircraft.
The four notional tasks identified for Walrus are strategic lift from the continental USA to the theatre of operations; theatre lift to move forces closer to the front; support of sea-based operations; and missions requiring persistence, including mobile command and control, airborne hospital, aerial refuelling and UAV launch and control.
“We do not have formal service customers, but we have a notion of where to go and something for each of the services,” says Hunt. Deciding the operational tasks allowed DARPA to identify goals for the Walrus programme, the principal one being “to control lift at all times, in the air and on the ground”. If successful, the programme will address the major failings of airships in the past, including their inability to operate in adverse weather and the need for ground infrastructure to handle loading and unloading. “The Walrus will be able to fly in, set down, unload and not get blown away,” he says.
As envisaged by DARPA, the operational vehicle would be able to carry a 500t payload 22,000km (12,000nm) in less than seven days at a competitive cost, operating without significant support infrastructure from unimproved landing sites and deploying the components of an army combat unit that can be ready to fight within 6h of disembarking. The vehicle would be capable of vertical, short or conventional take-off and landing “not just from an airfield, but from an open field”, says Hunt.
“The vehicle is almost as big at the Nimitz [aircraft carrier] and a darned sight fatter,” says Hunt. Whether the structure will be rigid, semi-rigid or non-rigid has not been decided. “The jury is out. It’s a real challenge,” says Hunt. “The propulsion system is also a challenge, as is the space available for storage. Life-cycle cost and survivability are also issues, but these will be dealt with later.”
The first critical challenge is the control of lift, which will be generated in multiple ways, says Hunt. Much of the lift will be provided by lighter-than-air gas, such as helium, which could be superheated to increase buoyancy for take-off and supercooled for landing. “If we raise the temperature 35°C we get an extra 15% lift,” he says. Other ways of controlling buoyancy include ballonets inside the envelope, which can be filled with offboard air and then superheated or supercooled.
The second source of lift will be the aircraft’s body and aerodynamic surfaces such as canards. Techniques to change aerodynamic lift and reduce boundary-layer drag will be required, says Hunt. The third source will be direct lift, either by vectoring the propulsion engines or by embedding thrusters in the airframe. Each lift-producing mechanism has a different frequency of response, and they must be integrated to provide a “fly and forget” control system, he says.
Two contractors have been selected for the 12-month first phase of the Walrus programme: Lockheed’s Skunk Works and small US airship manufacturer Aeros Aeronautical Systems. “They have significantly different approaches,” says Hunt. During Phase 1 they will develop conceptual designs for the operational vehicle and define a technology demonstration plan for the three-year second phase.
DARPA plans to award one Phase 2 contract to fly a subscale demonstrator and take the operational vehicle to a preliminary design review. Although the agency does not need a sponsor to enter the second phase, it will need interest from the services. A go/no-go decision on the demonstrator is due midway through Phase 2, by which time DARPA hopes to have a service sponsor for Walrus.
DARPA’s morphing aircraft programme has as its goal the design of a wing that can change its shape drastically to meet the conflicting mission requirements of efficient loiter and transonic dash. “UAVs fit two classes,” says MAS programme manager Terry Weisshaar. “Strike aircraft are fast and survivable, but have inefficient aerodynamics and can’t loiter if the target moves. Surveillance aircraft are aerodynamically very efficient, but can’t dash.” Morphing would allow the roles of long-endurance surveillance platform and high-speed unmanned combat air vehicle to be combined in a single hunter-killer UAV.
The fact that shape-changing is not new is best illustrated by variable-geometry aircraft, such as the Grumman F-14 Tomcat carrier-based fighter, which use a variable-sweep wing to provide good low-speed and high-speed performance. But the idea goes back much further, says Weisshaar, citing a morphing-wing design, the Eole, proposed by French aviation pioneer Clement Ader in 1890. Ader proposed a bat-type wing that could reduce its size by a half to a third. “We will get down to about half with MAS,” he says.
Among the first practical “polymorphous” aircraft was Westland’s Pterodactyl IV, a tailless monoplane designed by Geoffrey Hill and first flown in 1931, which had a wing sweep that could be varied manually by just under 5° to accommodate shifts in centre of gravity. Another early example cited by Weisshaar is the Russian Nikitin-Shevchenko IS-1 fighter, first flown in 1940, which converted from manoeuvrable biplane to faster monoplane by pneumatically folding most of its lower wing into the fuselage.
The Pterodactyl, and later variable-geometry aircraft such as the F-14, are examples of in-plane morphing, involving a two-dimensional surface movement to change wing sweep or span. The IS-1 was a three-dimensional morphing design, using out-of-plane wing segment folding to change area. Another form of three-dimensional morphing was used by the North American XB-70 bomber, first flown in 1964, in which the outer panels of the delta wing rotated downwards by up to 65° to increase directional stability and aerodynamic efficiency at supersonic speed. At Mach 3, the wedge-shaped lower fuselage and drooped wingtips created an air dam that slowed the airstream and generated compression lift that carried 35% of the aircraft’s weight.
Shape changing has always incurred penalties in cost, complexity and weight, but in the cases of the F-14 and XB-70, these were outweighed by the advantages to the overall design. “Grumman chose a variable-geometry aircraft over a fixed-wing design that was 5,000lb [2,300kg] heavier because it needed a bigger wing, bigger engines and more fuel,” says Weisshaar. The XB-70’s wing joints and actuators weighed 10,000kg, but drooping the wingtips improved the bomber’s supersonic performance. “There will be cost and weight penalties, but you will get a better aircraft,” he says. “You have to think at the system level.”
The two designs chosen for the 18-month second phase of the MAS programme are examples of in-plane and out-of-plane morphing. Lockheed’s Agile Hunter design folds the inner wing sections upwards against the fuselage to “hide” a substantial portion of the area to reduce drag during the low-altitude transonic dash. NextGen Aeronautics’ design has an articulated wing structure that can transform from a low-sweep, long-span loiter shape to a high-sweep, reduced-area dash shape. During Phase 2, both contractors have built small-scale radio-controlled flying models and full-scale semi-span windtunnel models of their morphing UAV designs.
Lockheed’s wing folds in two places, the outboard section rotating downwards to remain level while the inner section rotates upwards and inwards to lie against the side of the fuselage. The windtunnel model uses F-16 flap actuators, and the joints are covered by flexible, silicone-based skins to provide a seamless, aerodynamically clean surface. A slight vacuum is applied to the inner joint during folding to prevent binding. An operational vehicle could use a shape-memory polymer, says Weisshaar. When heated, the material would become elastic, allowing the joint to rotate, then “remember” its original, rigid shape when the heat is removed.
When the wing is folded, a flap on the inboard leading edge folds against the fuselage side to stop air flowing through the gap. This is the first use of a thermopolymer actuator, says Weisshaar. Compact enough to fit inside the small space available, this uses a material that expands when heated to drive the flap then locks the surface in place when the plastic cools and solidifies. “Bigger thermopolymer actuators could be used to move the wing,” he says.
NextGen’s wing design is based on a jointed endoskeleton that is moved by a distributed array of actuators to adjust span, area and shape. The semi-span windtunnel model uses 10 small hydraulic actuators to provide redundancy, says Weisshaar. The metal wing structure is covered with stretchable and compressible elastomeric silicone skins that are reinforced with titanium or steel mesh to prevent out-of-plane “puckering” that could affect aerodynamics. Attaching these skins is a challenge, “but they think they have solved it”, says Weisshaar. An alternative approach is to use sliding skins.
Windtunnel testing of both wings up to Mach 0.85 and 50,000ft (15,000m) is due to be completed in October. The tests, in NASA Langley’s transonic dynamics tunnel, are focusing on the aero-elastic behaviour of the morphing wings, and on measuring the actuator loads, which are hard to predict, says Weisshaar. The wings take about a minute to change shape, a time set by DARPA, but faster morphing could be used for flight control or to compensate for battle damage, he says.
The third phase of the programme, during which one of the contractors would build the MAS-X flying demonstrator, has yet to be funded, and DARPA is seeking a service sponsor. “When we are done with windtunnel testing, we will be at a TRL [technology readiness level] of 5. Then we will need to fly,” he says. “We need a decision soon on whether they see value in advancing to a higher TRL.”
The oblique wing was first proposed by German designer Richard Vogt in 1942, for the Blohm & Voss P.202 variable-geometry jet fighter project. This had a pivoting oblique wing mounted atop the fuselage, a configuration that was to recur several times over the ensuing decades. Famous NASA aerodynamicist Robert T Jones developed the theory for oblique-wing aircraft in 1952, and concluded they would have significantly lower wave drag than traditional swept-wing aircraft. And, as early as 1961, Handley Page designer Geoffrey Lee proposed a Mach 2 oblique flying-wing airliner, the Sycamore.
Despite this brisk start, oblique wings failed to make much progress, although several experimental aircraft were tested over the years. NASA Dryden flew a small, remotely piloted oblique-wing research aircraft in the mid-1970s and followed this with the single-seat AD-1, a subsonic research aircraft that made 79 flights between 1979 and 1982 with its fuselage-mounted oblique wing pivoted up to 60°. This was intended to be followed in 1989 by a supersonic oblique-wing research aircraft based on NASA’s Vought F-8 fly-by-wire testbed, but the programme was cancelled.
NASA Ames, meanwhile, continued work on R T Jones’ oblique all-wing aircraft, completing a preliminary design study of a 500-seat, M1.6 supersonic transport in 1991. The 124m-span aircraft was designed to take off and land with 37.5° oblique sweep, increasing to 68° in the cruise. Because of the control challenges posed by a tailless, unstable, oblique flying-wing, NASA built a small-scale remotely piloted demonstrator. The 6.1m-span model flew once in May 1994, for 4min, but demonstrated stable, controlled flight over a range of sweep angles from 35° to 50°.
Despite their advantages over conventional supersonic airliner configurations, oblique-wing designs – with or without fuselages – were ultimately rejected by NASA’s High Speed Research programme, which was eventually cancelled, ending US work on a high-speed civil transport. DARPA is now rekindling work on the oblique flying wing, while acknowledging that challenges still have to be overcome.
Back in 1952, R T Jones showed that a variable-sweep oblique flying wing is the most efficient configuration over a wide range of subsonic and supersonic speeds, and DARPA sees the opportunity to produce an aircraft that will combine high speed with long range and endurance. Challenges range from ground manoeuvring, and the integration of fully embedded engines into the airframe, to the control of a tailless, unstable, variable-sweep flying wing at speeds up to Mach 2.
The oblique wing’s problem over the years has been the control challenge posed by the unique coupling between the asymmetric aircraft’s aerodynamic and aerostructural modes. “Any time it came close, another configuration that was nearly as good would be selected because of the perceived risk for a manned aircraft,” says Stephen Morris who, as a graduate student at Stanford University, built and flew the small oblique flying-wing model for NASA. “It’s ideal for an unmanned aircraft.”
This early in the Switchblade OFW project, there is no service sponsor and no specific mission, so DARPA has drawn up two requirements for a conceptual, 2020-timeframe vehicle: a surveillance mission with a 4,600km radius and 15h subsonic loiter at 60,000ft carrying an 1,800kg payload; and a bomber mission with a 4,600km radius, maximum speed of M2, cruise of M1.6 and 6,800kg payload. The vehicle must be tailless in the subsonic loiter and supersonic cruise.
Because of the technical challenges, DARPA has not set the funding or timescale for Phase 1 of the Switchblade programme. Instead, bidders will propose a baseline programme to mature the OFW to a TRL of 4 or 5 by the end of the first phase. A supersonic unmanned X-plane demonstrator could fly by 2010 under a planned Phase 2, with the goal of demonstrating the concept’s feasibility so that a variable-sweep flying wing can be considered for a surveillance or bomber aircraft in the 2020 timeframe. “Of all the crazy ideas in aeronautics, this is one with a relatively large payoff,” says Morris.
GRAHAM WARWICK/WASHINGTON DC
When it crashed last month in the final moments of a 12h flight, Boeing’s A160 Hummingbird set two milestones: it established a new endurance record for an unmanned helicopter; and it was the third of the original DARPA-funded vehicles to crash. It was also the second DARPA-sponsored Boeing experimental unmanned rotorcraft to crash – the first X-50A Dragonfly canard rotor/wing (CRW) having met a premature end on its third flight in March last year.
Such unforeseen events underline the fact that DARPA programmes may be potentially high-payoff, but are also high-risk. But the research agency seems to have a higher tolerance for failure than NASA, and it is continuing to back both the A160 and the X-50 while taking on new high-payoff/high-risk projects.
Both programmes are typical in embodying an enabling concept which, if proved feasible, could transform aircraft design. The A160 features a variable-RPM “optimum-speed” rigid rotor that offers efficient low-power loiter and long endurance. The goal is an unmanned aircraft with an endurance of 24-32h, rivalling fixed-wing UAVs.
The CRW has a reaction-drive rotor, and is designed to take off vertically like a helicopter then transition to high-speed forward flight supported by the canard, stopped rotor and aft wing. The same powerplant that drives the rotor via tip jets provides propulsion in forward flight. The low-speed X-50 crashed before transition could be demonstrated, but a second unmanned vehicle is being prepared for flight testing.
Much of DARPA’s aeronautics-related research is focused on unmanned aircraft because of affordability and military utility. New programmes include Lockheed’s Cormorant multipurpose MPUAV, designed to be launched from the sea surface and submarines to provide close air support for vessels such as the planned Littoral Combat Ship and the SSGN, a Trident ballistic-missile submarine modified for covert strike and special operations.
The folding-wing MPUAV is housed in the SSGN’s 2.1m (7ft)-diameter ballistic-missile launch tubes. After release from the submerged submarine, the UAV is boosted out of the water by expendable solid-rocket motors, starts its turbofan and flies a 1,100-1,300km (600-700nm)-radius reconnaissance or strike mission, returning to a designated retrieval point, shutting down its engine and splashing down to be recovered by the submerged Trident using an unmanned underwater vehicle.
Under a small DARPA contract, Vought Aircraft Industries is working with Geneva Aerospace to fit a Dakota UAV with floats and a flight control system to demonstrate the feasibility of a sea-going unmanned aircraft. This will support Vought’s proposed Kingfisher II, a jet-powered UAV able to take off and land on the open ocean.