Flexible flyer

It's been called the military program of the century. Such hype is not far off in describing the capabilities and program goals for the U.S. military's Joint Strike Fighter (JSF). Nor is it exaggeration to say the Lockheed Martin-led team that won the design competition with its F-35 had to extend aerospace technologies to bring about a common aircraft that could meet the diverse but demanding needs of three services-and do it affordably. The three versions of the F-35 are: a conventional takeoff version for the Air Force, a larger wing, long range Navy plane, and an extra lift, vertical flyer for the Marines and UK.

Power lifting. Getting down to key technologies, Lockheed Martin VP of JSF Air Vehicle Product Development John Fuller says the Navy and Marine Corps unique basing needs had to be wrapped around good up-and-away (in-flight) performance of the basic airplane. "The biggest denominator and challenge was short takeoff/vertical landing (STOVL) propulsion," he notes. Here propulsive lift, engine, and flight controls have to all run smoothly together.

Team engineers decided to go with a more efficient and seemingly more complex system-a power takeoff (PTO) shaft from the propulsion engine's fan drives a lift-fan packed into the front of the airplane (see figure). The fan works in conjunction with a downward swiveling rear engine nozzle that can twist 90 degrees thanks to triple bearings around the nozzle's circumference. "Fueldraulic" geared actuators, which use the fuel feeding the engine as the working fluid rather than hydraulic oil to save weight, drive the bearings. For roll control in hover, there is a nozzle under each wing fed directly by engine air. The complete system produces about 40,000 lbs of lift.

On the previous generation Harrier, swiveling ducts leading directly from the engine blew air down for vertical flight. "The history of the Harrier was a struggle for thrust," Fuller states. "We could have stayed with steering thrust (and simpler ductwork) or make a bold move with the lift fan and get twice the payload of the Harrier. And we could do it within the same 'footprint,' that is stressing the deck or ground landing mats under the airplane with lower temperatures and erosion effects."

Without twisting flow paths in ducts, the fan produces 60% more lift. As for complexity, Fuller says, "Both lift systems are complex in their own way, having to both lift and control the airplane. And the Harrier has more small nozzles for control. Control wise, they are equivalent in complexity." He adds that as Harrier engines age, there is less lift available due to normal engine wear and seal deterioration in the duct system, a problem largely obviated on the F-35.

Clutch hit. While propulsively efficient (more so than expected because of lower pressure losses), lift-fan development also produced what was probably the program's biggest potential showstopper-initial tests showed efficient operation, but leaks developed in the fan-clutch lubrication and cooling system, accompanied by an alarming rise in clutch temperature. In transitioning to vertical hover, doors closing off the fan on top and underneath the fuselage open. Hydraulics then squeeze a stack of disks (like a multidisk brake) in the clutch connecting the shaft from the engine to the fan to bring it up to speed. Once stabilized, a locking spline engages for a positive mechanical coupling.

Jim Haney, who was chief engineer for lift fan development at subcontractor Rolls Royce Allison (Indianapolis, IN), says the problem was traced to the integration of the functions in the conventional-flight to propulsive-lift-flight sequence. The transition was to be within 10 seconds in order for the pilot not to linger in an interim handling mode. "The initial transition was too quick, with vibrations in the clutch pack breaking the seals," he notes. The phenomenon is similar to brake chatter in cars when force/torque levels combine with the speed of engagement (a combination of pressure and timing) and lead to vibration when contact is first made. "The solution was to go from open-loop control to closed loop with total feedback," Haney says. Sensors added for this control loop included rotary variable differential transformers (RVDTs) to provide position feedback and pressure sensors to monitor the squeezing forces.

Integration, flexibility. Lockheed Martin engineers and managers, along with contemporaries at design-team partner companies Northrop Grumman and UK-based BAE Systems, brought about a successful design not only with technical innovation such as the lift system but in team management as well. Designers were co-located by function rather than by which company they worked for. Engineers from suppliers were also on the design teams. Expert "greyheads" from elsewhere in the partners' organizations also reviewed key decisions and even retirees were consulted.

Common engineering productivity tools facilitated integration between the partner companies' facilities, according to Fuller. Dassault/IBM's CATIA was the primary solid modeling tool and Metaphase (now part of Teamcenter tools from EDS) for project data management tied things together. Fuller adds "Prime time was between 8 a.m. and noon Central Time when much collaboration was done between U.S. and European offices." Engineers on both continents were at work then, so phone calls and Internet conferencing using Microsoft NetMeeting could take place.

While some aircraft engineers wrestled with flying performance, others had to make sure the different JSF versions could be manufactured economically. Fuller says, "At one time commonality meant identical parts, now we say 'cousin,' not common parts. They can be the same design but dimensionally different, say beefier for higher landing loads, but be made of the same material, have the same part threads, and use the same assembly equipment for installation." The three versions share more than 80% identical parts. Cousin parts can be made from the same blanks and handled and assembled with common tools and fixtures. For example, bulkheads for each version differ only in thickness.

The team also borrowed flexible manufacturing methods from the automobile industry-the three JSF models will be assembled on the same line. And rather than have costly, precision jigs to join airplane components, "We put dimensional control in detailed manufacturing of the components," Fuller notes. In the past, precision was in the assembly tooling, where shimming the joined parts was required. "Precision parts now control the structure's joints," he adds.

In applying best practices, the three partners surveyed their procedures to see which company had the lowest cost for a given type of part. The others then adopted these methods. Finally Fuller notes, "In manufacturing aircraft, a lot of time is spent drilling holes and filling them with fasteners." So design engineers minimized the number of fasteners, and the costly time to assemble them, leading to a more unitized structure having stiffeners and features bonded to various parts.

Health maintenance. Making the fighter affordable also led to features cutting the cost of ownership in operational service. Once F-35 deployment starts beginning in 2008, it will have to be supported efficiently in the field to keep overall costs low. To ease the maintainers' tasks, most access panels are on the underside of the airplane or within the weapons bays to eliminate need for external stands. Even before an aircraft is serviced, Fuller says the technicians will know what needs to be done. "The philosophy is to continuously monitor systems and do preventative maintenance rather than wait for something to break," he notes. "There's a strong handshake between aircraft systems for prognostics and health management." As part of an automatic logistics system developed by Honeywell (Phoenix, AZ), maintenance actions required are downlinked to the ground before the airplane lands, allowing the quickest turnaround possible.

With engineers on its F-35 team evolving technologies, manufacturing processes, and systems integration, Lockheed Martin looks to have a good shot at making the aircraft both an aviation jack and a master.

Electrons for muscle and nerves
While aerodynamics shape the F-35, electronics allow it to function. Here are some of the significant systems on board:

The commonality conundrum

Up until now commonality has been the death knell of some aircraft designs. Significantly different mission requirements meant a common design had to be a jack of all trades but typically resulted in a master of none. Or else the design was optimized for one service at the expense of another-witness the F-111 tailored to the Air Force but scrapped by the Navy in the late '60s in favor of the specialized F-14. Needs more than speed. How high did the U.S. military and its British service partners set the bar for a single Joint Strike Fighter design to hurdle? Versions of the basic JSF would have to replace the U.S Air Force F-16 fighter and A-10 Thunderbolt attack plane, early models of the U.S. Navy and Marines F/A-18 Hornet dual fighter/attack aircraft, and the vertically flying AV-8 Harrier jump-jet used by the U.S. Marines and the British Royal Navy and Air Force. Aircraft in this aging fleet are reaching service life limits and facing increasingly sophisticated air defense threats to their survival in combat. The most recent of these existing aircraft, the Hornet, made its first flight back in 1978. Added to the equation was not only replacing the capability of these airplanes but doing it with an affordable aircraft that was easy to support out in the field or aboard ship. Cost goals (in 1994 dollars) were $28 million for the Air Force conventional takeoff and landing (CTOL) version, $31-$38 million for the carrier vehicle (CV) for the Navy, and $30-$35 million for the Marine and UK short takeoff/vertical landing (STOVL) variant. For comparison, the Air Force says the latest model F-16C/D models cost $18.8 million in 1998 dollars. The higher costs, and uncertainty, of the CV and STOVL models reflects the beefier structure and multirole missions needed for carrier operations and the complex propulsion for vertical flight, respectively. In addition, unlike its Harrier predecessor, the STOVL version will be capable of supersonic speed.