This article appeared in the July 1999 issue of Code One Magazine.
Print friendly version of this article (text only)
Building 602 dominates the landscape on the northeast side of Palmdale, California. The huge white building, part of Lockheed Martin Skunk Works facilities, was originally constructed for testing the L-1011 passenger aircraft. Since then, it has been used for a number of significant programs. The B-1, B-2, F-117, and the space shuttle orbiter all went through structural load tests in Building 602. The two YF-22 aircraft were assembled here. More recently, Argentinean A-4s were updated in 602. These days, the south end of the building bristles with activity related to the construction of the X-35 demonstrator aircraft for Lockheed Martins Joint Strike Fighter.
"We are in a race right now of building the airframes, completing the propulsion system, and testing the flight control software," says Gary Ervin, who leads the X-35 team at the Skunk Works. With a first flight planned for spring of 2000, the pace of the assembly process for the demonstrators is picking up considerably. "Last January, we had a few of the major bulkheads in place, and the inlets were in the fixtures," Ervin continues. "We had one shift and about ten people working in the assembly area. As of early February, the assembly effort has taken off. We added a second shift, and were working seven days a week."
The Plan
The effort in Palmdale centers on two concept demonstration aircraft that will be flown as three aircraft types. The first aircraft, called Ship 301, will begin life as a conventional takeoff and landing aircraft, or CTOL, which represents the US Air Force version of the JSF. This demonstrator will then be fitted with the short takeoff and vertical landing propulsion system to become a STOVL aircraft, which represents the US Marine Corps and the UK Royal Navy and Royal Air Force version of the JSF. The second aircraft, called Ship 300, is designed for demonstrating low-speed flying qualities suitable for landing on an aircraft carrier. Designated the CV variant, it represents the US Navy version of the JSF.
The plan varies from the demonstrator construction plan previously described in Code One. Originally, the STOVL demonstrator was the stand-alone aircraft, and the Air Force CTOL version was to be transformed into the Navy CV version by adding larger control surfaces and a larger outboard section on the wing. The change came about for several reasons.

"The new plan emphasizes STOVL work," Ervin explains. "We wanted to complete the STOVL demonstrator as soon as possible. Furthermore, by combining CTOL with STOVL, we eliminated one set of control surfaces since the CTOL and STOVL versions have the same basic planforms. This move saved some contract funds during our recent replanning effort.
"The externals of the STOVL and CTOL airplanes are exactly the same," Ervin continues. "The STOVL propulsion system has roll posts and doors for the lift fan bay. The bay contains a fuel tank in the CTOL variant. Some people may need a dance card to keep up with these changes. But they arent that complex. The fact that we swapped one-half of one airplane with the other highlights our emphasis on making these aircraft as common as we can."
The entire program philosophy of JSF is to manufacture several different variants of the same basic design on a single production line. The Lockheed Martin team is demonstrating that philosophy every day with the design and assembly of the X-35s. The demonstrator aircraft share one design feature that goes beyond the commonality of the basic design of the production aircraft. The lift fan bay is common between both X-35s. So, both X-35s will be able to carry and operate the shaft-driven lift fan STOVL propulsion system. Ship 301 will actually get the lift fan installed and flown, but Ship 300 could be retrofitted if the need arises.
"The history of STOVL flight testing is inherently risky, with the dangerous flight conditions close to the ground," says Ervin. "So we incorporated as much risk management as we could into the X-35 program. These aircraft bear the weight of thousands of production aircraft to follow on their success, so we did as much as we could to provide for every possible program twist and turn during the demonstration phase. That early planning is paying off already as we finish the assembly work on the two X-35s."
Aside from structural thickness and wing areas, the two airframes coming together in Building 602 look extremely similar. The most striking characteristics of the airframes are the massive engine compartment (the Pratt & Whitney JSF119-611 can swallow an F-16 engine) and the intricate machining on dozens of large, single-piece bulkheads. One bulkhead section, called the holy grail, stands out. Located just aft of the bay for the lift fan, this goblet-shaped piece of titanium joins the forward and aft halves of the inlets. The bulkhead has a webbed hole in its middle through which runs the shaft for the lift fan. This and several other complex titanium bulkheads represent the most sophisticated pieces of the demonstrator aircraft.
This sophistication streamlines the assembly process. "Some of the bulkheads would have been built from 100 or more separate parts in the past," Ervin explains. "Now we can design and fabricate them as single units. When these parts are delivered from the manufacturer, we set them in the fixture and tie them into the structure. We spend more time fabricating and less time assembling these bulkheads. However, they reduce our parts count significantly, they reduce our assembly time, and they simplify assembly. They also produce a stronger structure."

Build Process
The bulkheads comprise the starting points for the assembly of most of the airframe. The assembly process itself takes place in three large blue fixtures made of heavy-gauge steel and surrounded by scaffolded work areas. The fixtures are used to locate critical parts of the airframe precisely. One fixture is used for the cockpit section. A second fixture is used for the mid-fuselage. The third and largest fixture is used for the wing carry-through and the aft sections. This fixture is used to mate the front and rear sections of the airframe as well.
The airframe is made of three primary materialsaluminum, composites, and titanium. Bulkheads are fabricated from either aluminum or titanium. The wing skins are mostly composite. Titanium is used for frames and skins in portions of the airframe exposed to high temperatures or jet exhaust noise, particularly in the aft section.
The majority of the aluminum machined parts are made at Lockheed Martin Tactical Aircraft Systems in Fort Worth. All the composite parts and some of the titanium frames are being manufactured at the Skunk Works. "Since the era of the SR-71 Blackbird, the Skunk Works has been very comfortable in the use of titanium in aircraft structure," says Ron Pyles, the X-35 manufacturing manager for JSF. "In fact, a number of our machinists prefer to work with titanium. The composite parts for X-35 have also been some of the best first-article quality weve ever produced." Other suppliers fabricate the remainder of the titanium parts. The use of three-dimensional computer models in the design process streamlined the transfer of data from the designers to a variety of manufacturing sites. The sites were selected for their efficiency for the different types of manufacturing processes.
Assembling the cockpit and mid-fuselage sections of the airframe begins with the composite inlets, which go into the fixture first. The inlets were the highest risk structural assemblies for the X-35, since the rest of the mid-fuselage would be assembled around them. "The inlet program at Alliant Tech Systems is a major success story for JSF," says Pyles. "The detailed design, tooling details, and fabrication of the X-35 inlets using fiber placement technology was equivalent to an aerospace home run. The design and fabrication technologies of fiber placement have been demonstrated before, but never on such a large scale and never as an integral part of the fabrication and assembly process of an all-up aircraft. "The ability to get a sophisticated component like a jet engine inlet fabricated and assembled with nearly no joints and with the extreme precision required for robust inlet operation in flight will provide a huge payoff when JSF production begins."
After installing the inlets, keelsons the structural members that run from front to back supporting the landing gear bay are put in next. The rest of these two sections of the airframe are built out from the center. Skins are installed last. Some subsystems are installed along the way. The cockpit section is rolled into the mid-fuselage section before these two combined sections are joined with the wing carry-through and the aft sections in the mate fixture.
Assembling the wing carry-through, the section between the mid-fuselage and the aft section, begins with outer portions of the wing, which are held precisely in the fixture. The forward and aft spars are installed first. (Spars are structural members that span the wing from the fuselage out.) The major wing load carrying bulkheads are positioned in the fixture, then the rest of the spars are installed. The skins go on from the outboard wingtip to the center. The aft tail boom section is assembled separately and then attached to the back of the wing carry-through section.

Innovations
A laser positioning system finds the precise location of pieces within the fixtures. "Before we had laser-based positioning systems, we had to measure from a given bulkhead," explains Pyles. "Computer-aided design and manufacturing tools and the ability to position parts in space with lasers allow us to build the forward fuselage without measuring from a bulkhead."
Precision is the key attribute of these positioning systems. When the aft section and the wing carry-through section were first brought together, they lined up within three ten-thousandths of an inch. "Years ago, we would be off by forty or fifty one-thousandths of an inch," says Pyles. "We would have to shim the structure to make it fit. Ive never seen tolerances this exact. The airframe is going together like a jigsaw puzzle."
Downloading computer-based design information directly to numerically controlled machinery to fabricate parts is yet another manufacturing technique being put to good use on these demonstrator aircraft. "Today, we produce tubing directly from digital descriptions of tubing routes," says Pyles. "In the past, we used a mockup of the airplane and tack welded bits of tubing together to describe a particular tube. We would then use that tube as a master to build a weld fixture. The new method eliminates several steps."
Digital techniques are being applied to the manufacturing of composite parts for the demonstrator aircraft as well. Skunk Works is using laser projection systems for building up the more complex composite parts, like the inlet duct for the lift fan. The projections are derived from computer-based design files. The process is more efficient, and it reduces scrap. "We have scrapped only two composite parts out of dozens," says Pyles. "That is significant improvement over past programs." Even the prototype tools used to build the demonstrators are benefiting from digital techniques. Computer-based design is being combined with stereo-lithography to create hundreds of drill positioning tools. (Stereolithography is a process for producing complex three-dimensional shapes from a plastic resin.) These tools conform to the surface shape of a particular portion of the aircraft. They allow holes to be drilled perpendicular on curved surfaces so the rivet heads are flush with the surface. Historically, these tools would have to be fabricated individually from metal by hand.
"Many of these techniques have been a standard practice in our industry for the last ten years or so," says Ervin. "We used a lot of them on the YF-22 program. We used computer-based design and manufacturing tools and three-dimensional representations of the aircraft. The design tools used on YF-22, however, were in the earliest stages of three-dimensional design. We didnt have the level of integration between design tools and manufacturing methods like we have today. These tools have been improved dramatically since then. We are also taking advantage of entirely new processes, such as high-speed machining and superplastic forming, to a limited extent in this phase of the program. We are investigating other techniques in separate manufacturing technology demonstrations. These techniques will be used more extensively in the next phase of the program, the production JSF."
The JSF build team in Palmdale is experiencing an improving learning curve as it builds the second airframe. Installing and trimming the inlet ducts in the mid-fuselage took several weeks the first time. The same effort for the second airplane required only four days.
Once the sections of one airframe are mated, the airframe is removed from the mate fixture to make room for the second airframe. The remaining subsystems are installed next as are the final control surfaces, landing gear, cockpit controls, ejection seat, and canopy. The propulsion system is the last major item installed.

Borrowed Parts
As with most demonstrator aircraft, the Lockheed Martin JSF demonstrators borrow many items from existing aircraft. The environmental cooling system is from the F-18E/F. The airframe-mounted accessory drive is from the B-2. The Upco ejection seat is the same one used on the Harrier AV-8B. The multifunction displays, produced by Avionic Displays Corporation, are from the C-130J. The head-up display is a candidate display for the KTX-2 program. The flight test nose boom is from the X-31.
The nose landing gear is from the F-15E. The main landing gear is a 1990s update to the 1950s A-6 Intruder. "We had to redesign the main gear because the 1950-vintage structural analysis and testing for the A-6 landing gear were not up to the 1990s expectations for flight certifying landing gear," explains Rick Rezabek, the chief engineer for the demonstrator aircraft. "We basically had to reengineer and redesign the Grumman A-6 landing gear for these two aircraft. The gear on the production aircraft will look similar, but it will be a custom gear to be more compatible with the weapon bay."
Off-the-shelf items do not always provide the best solution. "We tried to select existing flight actuators," continues Rezabek. "But by the time we did all the design work, everything was essentially new. Each control surface has specific hinge moments and actuator flow rates to meet our flight control response requirements. The pistons, which determine the size of the actuators, are sized to our design requirements."
More to come
Once the CV variant is assembled and complete, it will be placed in a test fixture for structural proof testing to 100 percent of its design load. "Normally, we test airframes to 150 percent of their predicted loads and fly to 100 percent," Rezabek explains. "We also usually have a separate non-flying aircraft devoted to this testing. In demonstrator programs like this, however, we rarely have a structural test item, so we have to test one of our flying aircraft to more conservative levels. The engineers always debate about how high or how low we should proof test an airframe. We are doing the proof test to 100 percent load, which allows us to fly at eighty percent of the design loads."
The demonstrator aircraft must also pass a long series of checkout tests before they leave the ground. "We have to look at every joint of every hose of every line for leaks and then perform functional testing of all of the modes of operation of each system on the aircraft," Rezabek says. "This testing begins this summer and runs through early next year. The name of the game at this stage of the program is Integration. Each component of hardware is currently finishing its own qualification test program at the manufacturers facilities. Those tests confirm the pumps, actuators, heat exchangers, computers, generators, and valves that will breathe life into the X-35s all meet their specifications. Our upcoming sets of functional tests on the X-35s turn those individual tests into a symphony of sorts. Those tests will show us how good a composer we were during the detailed design stage. We are confident the results will be music to everyones ears as the X-35s roar to life."
Article by Eric Hehs
Photos by Denny Lombard, Marty Wolin, and Peter Torres

|