Graham Warwick in Cranfield, UK

BoeingX48BW445

Suspended from a yellow frame, the manta-like shape shakes almost imperceptibly as the second X-48B demonstrator undergoes final testing before being shipped to Edwards AFB in California. The ground vibration testing is being conducted by Cranfield Aerospace, which has built the blended wing-body (BWB) unmanned research vehicle for Boeing’s Phantom Works in an unusual example of transatlantic cooperation.

“We are providing Boeing with a research tool in which to test their flight control system software,” says D J Dyer, Cranfield Aerospace’s general manager UAV systems. The X-48B will allow the BWB’s low-speed characteristics and complex control system to be explored in flight. “We are giving them the complete thing – two 8.5%-scale aircraft, a ground control station, support equipment and spares,” says Prof Ian Poll, business development director.

K63682-08 W445
© Boeing

NASA and the US Air Force Research Laboratory are backing the X-48B project

Boeing, which has worked on BWB design for several years, planned to build a larger 14%-scale demonstrator, the X-48A, but NASA budget cuts killed the project. “Boeing decided we still needed to demonstrate the flight controls, so we looked for people to work with,” says Phantom Works’ X-48B chief engineer Norm Princen. “We did not want to build the vehicle in-house, so we chose Cranfield.”

Designing and building two X-48B BWB low-speed vehicles (LSV) fits with Cranfield Aerospace’s “concept to flight” strategy to reduce the cost and time to obtain high-quality aerodynamic and flight control data by combining its rapid prototyping and unmanned aircraft capabilities. “It is enabled by UAV technology and is a new way to look at aircraft design,” says Poll. “If you can get good data in the early stages of a programme you can reduce the risk enormously. BWB is right at the very beginning, and Boeing can get very high-quality data by flying this vehicle,” he says.

Formed in 1997 as a wholly owned subsidiary of Cranfield University, and located north of London, the company holds civil and military approvals to design, build and fly manned and unmanned aircraft. “We are a for-profit company, and all of our income is from commercial contracts,” says Poll. The company has access to the university’s intellectual and physical property, and the majority of its almost 100 employees are ex-university, but it gets no academic funding. “We only survive if we win contracts and make a profit,” says managing director David Gardner.

Dynamic scaling

Design of the 6.2m (20.4ft)-span, 9.3m2 (100.5ft2)-area X-48B was a challenge because it is scaled dynamically, rather than simply geometrically. Although much quicker than those of the full-size aircraft, the vehicle’s responses will scale perfectly, says Poll. “The moments of inertia are scaled, which is not normal. The frequency of the short-period oscillation, for example, will not be the same as full-size, but will be a function of the scaling parameter.”

Cranfield designed the X-48B with CATIA software, using the outer mould line of Boeing’s 451L BWB study configuration, for which an extensive windtunnel database has been developed. Dynamic scaling put a premium on weight saving and mass distribution, and the vehicle has a carbonfibre airframe built by UK subcontractor Lola Composites. “As you move away from the centre of gravity, mass is much lower than normal, and in places the skin is just one laminate thick,” says Dyer. “There was an enormous amount of finite-element modelling work done by Boeing with Cranfield.”

Funding the X-48B from its own resources, Boeing went to a smaller scale than the X-48A so it could afford two vehicles and a more robust test programme, says Princen. The smaller size also allowed Cranfield to tap into the model aircraft market, which supplied the X-48B’s three 50lb-thrust (0.22kN) micro-jet engines. These give the 178kg (392lb) vehicle a maximum airspeed of 118kt (218km/h), altitude of 10,000ft (3,000m), endurance of 60min and range of 218km (118nm).

The flying-wing vehicle has 20 control surfaces along its trailing edge, each driven by one or two electric actuators supplied by Kearfott and originally developed for the AAI Shadow tactical UAV.

Cranfield built the avionics system, which includes dual single-board processors – one for housekeeping and one for flight control, into which Boeing can upload its software. Cranfield is not allowed access to the flight control software because of US export restrictions.

“All the flight control software development is done by Phantom Works. The air vehicle is test equipment for flight control system research,” says Princen. “The big research challenge is control allocation.” The BWB’s control responses are non-linear and highly coupled, so all 20 surfaces are active all the time. Roll produces adverse yaw, so in addition to winglet rudders, the outer pairs of ailerons are split to act as drag rudders as well as speedbrakes.

Remotely piloted

X-48B K63682-09 W445
© Boeing

The first X-48B completed 250h testing in NASA windtunnel

Cranfield has also supplied a ground control caravan with pilot and flight-test engineer stations. The X-48B is remotely piloted, and will be flown by a Boeing test pilot using conventional stick and throttle and an out-the-window image downlinked from a camera in the vehicle’s nose and overlaid with a headup display. Normally the aircraft will take off from and land on the Edwards lakebed, but there is a recovery parachute and landing airbag for emergencies. There is also a spin recovery parachute in case the aircraft departs controlled flight during high-alpha stall testing.

Flying experience with the BWB configuration is limited to a 5.5m-span propeller-driven model built by Stanford University and flown in 1997, and tethered windtunnel tests of a NASA 3.7m-span free-flight model last year. Data from the Stanford flights was not good quality, says Princen, but the NASA tests proved useful. The X-48B will investigate stall characteristics, spin and tumble, asymmetric thrust and ground effects, plus dynamic interactions between the control surfaces, wing aerodynamics and inlet conditions that cannot be tested in a windtunnel.

The first X-48B has just completed wind­tunnel testing at NASA Langley in Virginia and will be used as back-up for the second vehicle, now en route to NASA Dryden at Edwards for flight testing. Complete except for its fuel system, the first vehicle underwent 250h of windtunnel testing to measure air loads and hinge moments. The flight control system was used to record air data during, and reposition control surfaces between, runs and resulted in “very efficient” testing, says Princen – “250h versus 1,000-2,000h without movable controls”.

Testing at Edwards will begin with low-risk parameter identification flights and work up to high-risk, high-alpha flying. The X-48B will fly first with fixed extended leading-edge slats. These will then be replaced with fixed retracted slats and the tests repeated, an approach Princen says is simpler. As the X-48B will assess the BWB’s behaviour at high angles of attack and speeds close to the stall, there is a risk of the aircraft spinning or tumbling end over end. Poll points out the irony in the tests being conducted at Edwards, named after a US Air Force test pilot killed in 1948 when his Northrop YB-49 flying-wing bomber began to tumble.

Although the X-48B will respond much more quickly than the full-scale vehicle, Boeing will be able to scale up the data for use in its BWB flight simulator. The control laws designed by the Phantom Works are intended to provide the control responses of a conventional aircraft, despite the BWB’s unconventional flying characteristics. “We don’t want to have to retrain the pilot to fly the vehicle. It should fly like a 777 or a C-17,” says Princen.

Source: Flight International