Truth or consequences: testing the door on NASA’s new window on the universe

Washington DC
This story is sourced from Flight International
Subscribe today »

Anyone who has ridden in a four-door car or truck with the front windows closed and a single rear window slightly open will have experienced the surprisingly loud buffet when the fast airflow outside the vehicle interacts with the static air inside. Imagine how it might sound if a 4.3m (14ft)-wide "window" were open at the rear of a Boeing 747SP travelling at 45,000ft (13,700m) and Mach 0.85.

NASA is betting the occupants of a special 747SP will not even notice.

To that end, the agency has invested almost two decades and $500 million dollars on modifications to N747NA, a former Pan Am and United Air Lines mainline aircraft that is now a research platform called the Stratospheric Observatory for Infrared Astronomy (Sofia). Mounted in the rear of N747NA - in Section 46 - is a 19t telescope with a 2.5m (98.4in) primary mirror, 100mm larger than the Hubble Space Telescope's mirror, that scientists will use to study infrared and far-infrared light above 99% of the water vapour that interferes with such measurements from the ground.

The need for a large cutout in the side of the aircraft is self-evident when considering the needs of a telescope. Measuring 4.6m high by 4.3m wide, the opening roughly spans the 3 o'clock to 12 o'clock arc looking rearward. At any one time during operations at altitude, one third of that space - about 6.6 m2 (71ft2) - is open to the atmosphere, allowing the Cassegrain telescope to rotate from 15e_SDgr to 60e_SDgr above the horizon during observations.

747 telescope

The remainder is covered by an upper rigid door, a lower flexible door and a rearward "aperture" element, all moving synchronously to tame the slipstream. If the cavity were to become resonant, the equivalent of an aerodynamic whistle, the telescope performance and in the worst case the aircraft's structural integrity itself, could be compromised.

NASA will find out soon enough it the design was successful - first open-door flights of the highly modified aircraft are to take place starting as soon as February in preparation for limited science operations later in 2009 and full operational capability by 2014. The telescope is being provided by the German Aerospace Center, DLR. The system, aircraft plus telescope, is designed to have a 20-year life.

A key requirement for the scientists is a telescope unaffected by the open-door aerodynamics a ground rule that in part sent engineers to wind tunnels as far back as 1990. Critical to the effort was the input of NASA contractor Bill Rose, an expert in ground-based aerodynamic and full-scale flight testing and the former leader of the fluid mechanics group at NASA's Ames Research Center. Rose is president of Nevada-based Rose Engineering and Research.

Rose had experimented with methods to quieten NASA's Kuiper Airborne Observatory (KAO), a modified former military Lockheed C-141 that carried a 914mm reflecting telescope from 1975 until 1995. KAO's telescope looked through a square 130 x 130mm cutout near the top of the fuselage just ahead of the wing. KAO used a traditional method of wind control: a perforated fence at the front edge of the cavity, similar to the flap on a car sun roof. A fence is typically oriented between 30e_SDgr and 70e_SDgr to the flow in order to disrupt the boundary layer, add energy and deflect the air upward and over the opening. "It can stop cavity resonance, but it still leads to a noisy cavity," says Paul Fusco, one of the NASA Ames Research Center engineers responsible for the Sofia door. NASA's Dryden Flight Research Center, also in California, is managing the flight testing.


The problem with noise is that it bombards the telescope with pressure fluctuations, shaking the mechanism and blurring what should otherwise be clear observations.

Near the end of the Kuiper's operational life Rose tested a ramp structure at the rear of the C-141's cavity on the outer mould line of the fuselage. Though not an ideal configuration - a better solution would have been to install the ramp inside of the cavity at the aft end - flight tests showed that by using an external ramp the fence angle could be reduced to 10e_SDgr, quieting the cavity.

The purpose of the ramp, says Fusco, is to control the boundary layer and to give the airflow a "stable landing point". For Sofia, Rose ultimately developed teardrop-shaped raised contours on the forward and aft sides of the cavity to take the place of a fence, and created an articulating cavity with moveable top and bottom sidewalls and an internally mounted D-shaped composite ramp, also called the aperture, at the aft end.

All three components move in concert with the telescope as it changes elevation before and during observations. Mounted on air bearings, the telescope can also move in small increments in cross axes.

747 telescope

This seemingly simple solution took about seven years to develop. Rose says NASA originally wanted to upgrade KAO's light gathering capabilities by a factor of 10 in the new platform, which would have required a 3m telescope. "Reality set in and we went down to 2.5m," he says. Using the 4.3 x 4.3m transonic wind tunnel at NASA Ames Research Center, Rose and his team then went to work building and testing 7% scale models of the Boeing 747, initially with the cutout in the front of the aircraft, similar to the KAO. However, for a combination of "political and scientific reasons" NASA decided to move the telescope to the back, he says. The benefits included having a single pressure bulkhead in front of the non-pressurised telescope section as opposed to two (front and back) and from an aerodynamic standpoint, the new location would have thick and stable boundary layers and no shock waves to deal with.

Rose ultimately completed five wind tunnel test series with various cavity configurations on the 747-200, 747-400 and 747SP before settling on the two-part "partial external door" within a teardrop-shaped external fairing on the 747SP.

Then in 1996 NASA selected L-3 Communications to install the modifications on N747NA at its Waco, Texas facility. Given that Boeing did not bid for the modification work, L-3 found itself in the position of having to reverse-engineer major portions of the 747SP in order to accomplish the work, says Mark Henley, Sofia project engineer for L-3.


The work focused on Section 46 of the aircraft, the fuselage barrel behind the wing and forward of the tail. L-3's mechanical design department spent 370,000 man-hours designing the modifications and another 300,000 man-hours performing stress analyses to verify the design, says Henley. Before cutting metal on N747NA in 2000 the company purchased the Section 46 remnant of another 747SP, a former United aircraft (N141UA) and had the section flown to Waco in NASA's Super Guppy* to practise the modifications.

Along with reinforcing the fuselage to handle the telescope's weight, building a new forward pressure bulkhead, relocating the pressurisation outflow valves forward and rerouteing the flight controls and hydraulic lines, L-3 also modified the wing-to-body fairings and built Rose's teardrop-shaped fairing over the instrument cavity to cover an arc of 200e_SDgr centred near the top of the fuselage but favouring the left side. Before its first flight in April 2007, L-3 completed wind tunnel and computational fluid dynamics (CFD) tests to verify the final design.

As well as the dynamic door and ramp system, L-3 installed a dryer in the forward baggage compartment to take condensation out of the cavity and a liquid nitrogen cooling system behind the aft pressure bulkhead to help cool the heat generating elements of the telescope in flight to match ambient temperature at altitude and minimise heat-induced movement of the telescope.

Back at NASA Dryden, NASA began closed-door flight tests in early January this year, confirming the modified aircraft's structural integrity and performance before taking it out of service for telescope buildup.

The next phase of flight tests will be most crucial. NASA Ames's Fusco says in late February pilots will make several flights with the door closed before opening the door "just a crack" at 15,000ft to check for resonance using a suite of sound pressure sensors in the cavity. Engineers will progressively open the door wider during 10 or more test flights, from 10% open to 30% to 70% to wide open. After verifying acceptable performance at each step the test team will repeat the checks at different airspeeds and with minor yaw manoeuvres.

A portion of the airflow contacting the aperture is designed to flow into the cavity and circulate in a stable fashion, a process engineers are confident will occur at altitudes above 35,000ft and speeds of M0.8 or more. If an uncontrolled resonance turns up during flight tests at altitude and speed, Fusco says NASA is working on a fence design as a backup.

Possibly more concerning would be a door stuck in the open position, forcing pilots to fly through lower altitudes where resonances have been shown to be problematic in wind tunnel testing. Just in case, Fusco says the doors have been designed to handle 20min of the worst case resonance with a factor of four safety margin - enough to allow pilots to land the aircraft. "Somewhere before we get to the 10th flight we will land with the door open to prove that it is safe," says Fusco.