Print friendly version of this article (text only)

An engineer brings the JSF119-611 test engine to life by turning a valve that sends compressed air to an air starter. The air starter is attached to a gearbox on the underside of the engine. The gearbox turns the engine’s high-pressure compressor and high-pressure turbine. The various stages of the engine spin faster and faster. When the rotation reaches about 3500 rpm, fuel pumps send JP8 to fuel injectors in the combustion chamber. An igniter spark sets off a continuous reaction of air and fuel. The air starter shuts down. The engine is running. As a throttle lever in the control room is advanced beyond military power setting, fuel is injected into the exhaust stream in the engine augmentor. A purple-orange flame shoots out the exhaust signaling the first stage of afterburner. The flame increases in intensity as the engine throttle advances to the second and third afterburner stages. The rumble of the raw thrust can be heard miles away.

A complex web of multicolored wiring covers this ground-based engine. The wiring carries signals from a variety of sensitive test instrumentation installed throughout the engine to monitoring equipment in a nearby control room. Inside the control room, dozens of technicians and engineers watch over a variety of displays and track thousands of parameters during these engine runs. "We may have up to fifty people in the control room monitoring these parameters during the initial runs for these developmental engines," explains Ed Clark, Pratt & Whitney’s development manager for the Lockheed Martin Joint Strike Fighter engines. "We are looking at clearances, pressures, temperatures, and vibration levels. We use lasers for measuring clearances. We use specialized devices called pyros for measuring temperatures in the turbine."

Clark admits that the ability to measure more engine parameters often leads to the need to measure more parameters. "We are monitoring about 2,000 parameters on the engine alone," he says. "That’s at the high end for test engines. If we had put instrumentation on it to satisfy all the original requests, we would be measuring 4,000 parameters." The vertical lift systems associated with the short takeoff and vertical landing, or STOVL, engine for the Lockheed Martin JSF adds about another 700 parameters to the total.

Pratt & Whitney has built two of the six JSF119-611 engines for the current concept demonstration phase of the Lockheed Martin JSF program. These first two engines—one conventional takeoff and landing, or CTOL, and one STOVL—are dedicated to ground tests. The first CTOL test engine is designated FX661. The first STOVL test engine is designated FX662. The third engine (FX663), being put together now, will be used for software development and qualification testing. It will also serve as a spare for the flight test program. The fourth engine (FX664) is for accelerated mission tests. In these tests, the engine is taken through duty cycles it will experience in typical flight test missions. The total number of cycles will be twice what the engines will go through in the entire concept demonstration program. The cycles specify throttle transients and STOVL operations as well as sustained operations at high temperatures. The cycles themselves are derived from the flight test plan. The fifth engine, the first flight test engine (designated YF001), will power the CTOL demonstration aircraft. The sixth engine (YF002) will power the STOVL demonstration aircraft.

Testing developmental engines like the JSF119-611 begins with collecting aeromechanical data. This data includes parameters measured by strain gauges placed inside the engine. After aeromechanical tests, the engine is "run-in," which involves operating the engine for sustained periods at full speeds and operating temperatures. The run-in process for a jet engine is much like the break-in period for an automobile engine. Piston rings in an internal combustion engine form seals as they rub against cylinder walls in the first hours of operation. Similarly, turbine blades and other moving parts in a jet engine create their own clearances as they rub against softer materials called abradables. This run-in process applies to production engines as well as to developmental engines.

"Everything we do on developmental engines early on is driven by strain gauge life," Clark explains. "If we did the engine core run-in right up front, the material that comes off the abradables would eat up the strain gauges. That’s why we get strain gauge data first. Then we run-in the core. After that, we measure performance."

Developmental testing leads to qualification testing, which involves high cycle fatigue tests, the accelerated mission tests mentioned earlier, and altitude qualification tests. The fatigue and accelerated mission testing is conducted at Pratt & Whitney facilities in Florida. The altitude tests are conducted at Arnold Engineering Development Center in Tennessee. The goal of all the testing is to clear the engines for the JSF concept demonstrator aircraft, the first of which is scheduled to fly next year.

The developmental engines, however, must pass certain conditions before any demonstrator aircraft flies. A functional characteristics document sets the specifications for the engine. The Allison lift fan has its own functional characteristics document as well. The specifications include a flight envelope (altitude and speed) that goes a little beyond the one described in the flight test plan so that the demonstrator aircraft are not limited by the engine during flight test. The engines are essentially cleared for flight after the accelerated mission tests are completed. "The fact that we have four new engines running twenty-three months after the contract was awarded is amazing," says Walt Sirmans, Pratt & Whitney’s model manager for the Lockheed Martin JSF engines. (The four total also includes two engines for the Boeing JSF design.)

Pratt & Whitney received the propulsion contract for the JSF in January 1997. Fabrication of the first engine began that March with the machining of the titanium billet for the first-stage fan of the CTOL engine. The first JSF compressor section was completed in September 1997. The first JSF engine core, which came directly from the F119 production line in Connecticut, was completed in January 1998. Engine testing began in June 1998.

"With all the parts in hand, we can build a JSF engine in thirty days," says Milt Jepson, manager of assembly operations for Pratt & Whitney. "Instrumentation requirements for these developmental engines, however, can more than double that time. We typically take anywhere from sixty to ninety days to build a developmental engine."

While instrumentation complicates the construction of these initial JSF powerplants, the engine itself is designed to be more easily assembled, repaired, and maintained. "This engine is much easier to work on than previous engines," Jepson adds. "Maintainers can replace any component on the engine with a minimal set of commercially available tools. All controls and removable components are located on the bottom surface of the engine where the access bays are so that maintainers don’t have to pull the engine to replace these components. The engine has split fan cases so that a fan blade can be repaired without removing it from the engine. The same is true for the compressor. This engine is a lot more maintainable."

Engines normally play a relatively small part in the flight control system of a fighter. In vertical flight modes for the JSF, however, the engine will be the primary source of controlled flight. The vertical lift system of the STOVL aircraft, therefore, presents a unique challenge for the JSF propulsion team. The vertical lift system on the Lockheed Martin JSF includes a lift fan in the front of the aircraft behind the cockpit, roll ducts on the wings, and a three-bearing swivel nozzle on the tail end of the engine.

The lift fan is driven by a drive shaft connected to the face of the main engine. As the clutch engages the vertical fan at one end of the drive shaft, doors above and below the fan open as it spins up. The rest of the vertical thrust is provided by the three-bearing exhaust nozzle on the main engine and by the two roll ducts on the wings. The exhaust nozzle can swivel 110 degrees downward from the horizontal. In addition to providing lift and pitch control, the nozzle can swivel side to side for yaw control. Thrust for the wing ducts, used for roll control, is supplied from the fan section of the main engine. This thrust comes from cooler air that normally bypasses the engine’s turbine section.

The Lockheed Martin propulsion approach circumvents hover problems associated with high-temperature and high-velocity air by providing much of the downward thrust with cool air from the lift fan. The lift-fan approach also removes energy from the hot turbine section of the main engine, which, in turn, lowers the main engine’s exhaust temperature, producing an even cooler footprint. Engineers selected the lift fan approach because STOVL thrust requirements could be decoupled from the cruise engine. The approach allows the engine to be optimized for conventional up-and-away flight. Further, the amount of thrust from the lift fan greatly exceeds the weight penalty of the lift fan system. Finally, the exhaust jet temperatures and pressures produce an acceptable ground environment during hover modes.

Components for the STOVL lift system were tested separately before they were integrated with the second JSF engine at Pratt & Whitney facilities in Florida. Allison tested a model of its lift fan nozzle at NASA-Lewis Powered Lift Facility in Ohio in the summer of 1997. The tests validated predictions of exhaust nozzle performance. B.F. Goodrich demonstrated high-speed clutch engagements representative of STOVL operating conditions for its lift fan clutch in 1997 as well. The clutch, gearbox, and lift fan went through additional tests in June 1998. The first lift fan test article was then shipped to Florida in October 1998. The fan was driven by the JSF119-611 engine for the first time a few weeks later and to 100 percent speed in November 1998.

"These engines are running extremely well and are meeting all their performance goals," says Sirmans. "We’ve already done military and maximum power testing in the up-and-away flight modes and STOVL work, including conversions from vertical takeoff modes, to conventional flight, to vertical landing modes."

The STOVL tests so far have been conducted in Pratt & Whitney’s A-9 test stand in Florida. The engine is held in the stand from above and mounted on two large tracks. Three large metal mesh baskets, one above and two facing forward, protect the inlets from ingesting foreign objects. Between engine runs, a large work stand called the Enterprise (a reference to the control deck in Star Trek) is rolled into place from the front to provide technicians access to the engine and all of the associated test wiring. For engine runs, thrust collectors are rolled into place below the lift fan and behind the engine. The front collector diverts air from the lift fan to a large diameter pipe that leads out the back of the facility. The huge back collector, made of quarter-inch cold-rolled steel, has a series of internal vanes that distribute the brunt of the hot exhaust gases when the engine is in hover mode. In cruise mode, thrust exits through a huge pipe directly behind the engine. The back collector is surrounded by large sound-absorbing panels that form the back wall of the building. The panels lower the noise levels and prevent exhaust gases from getting near the inlets.

Subsequent STOVL engine tests will be conducted in a separate outdoor stand called C-12. This stand, also in Florida, was originally designed to accommodate Pratt & Whitney commercial engines. It was recently modified with a fixture for measuring the performance of the JSF STOVL engines in three axes. The stand, which places the engine well above the ground to avoid ground effects, can be rotated to deal with changing wind directions.

Rated in the 40,000-pound thrust class, the JSF engine will be one of the most powerful engines ever put in a fighter. "Our commercial engines put out up to 90,000 pounds of thrust," notes Sirmans. "But we can’t have a fighter running around with a 112-inch diameter fan on the front of it. Size and supersonic requirements tend to determine a fighter engine. Supersonic drag for a fighter is driven by cross-sectional area, with the engine the biggest contributor to that area. Thrust itself is a function of airflow, pressure, and temperature. We can get more thrust by increasing temperature; however, temperature can be the limiting factor in such an engine."

At temperatures exceeding 3000 degrees Fahrenheit, the turbine section of an engine runs at temperatures well above the melting point for any metal known. Metal parts in these hotter sections are kept solid by bathing them in cooling air so that the hot gas path never touches them. The cooling air is carried through hollow passages inside the turbine blades and exits through hundreds of tiny holes strategically placed on the surface of the blade.

Higher temperatures are also addressed by advances in materials. While front stages of the compressor are titanium, hotter fan stages in the back of the engine are made of nickel alloys. The hottest part of the engine, the high-pressure turbines, are made of single-crystal material, which refers to a unique process for manufacturing high-temperature metal alloys.

"We’re leveraging all the resources we can to get the best technologies that we can use in these engines," Sirmans says. "Many of these technologies are coming from the US government’s Integrated High-Performance Turbine Engine Program. We will fly with the engines we’re testing now. But we plan to pursue even more advanced technologies in the next phase of the program."

These advanced technologies include supercooling, a cooling technology that makes more efficient use of the cooling area in the turbine section. The technology involves advances in computational fluid dynamics. "Cooling air in an engine is like gold," adds Sirmans. "We don’t want to extract any more cooling air than is needed. Every bit of air used for cooling reduces engine performance."

Many of these improvements are a direct result of recent advances in the tools used to analyze and design engines. Advances in computational fluid dynamics are improving airfoil designs. Digital manufacturing techniques allow intricate designs to be produced with high fidelity. Digital engine controls improve response times and efficiencies. Many of these improvements also address affordability, a major driver on the JSF program.

"A lot of our risk-reduction activities are working on making supercooling more affordable," says Sirmans. "We want the concept demonstrator engines to be a minimum change from the F119 core engine, which powers the F-22 Raptor. We share engine cores on the commercial side all the time. Our family of PW4000 engines, for example, covers a thrust range from 52,000 pounds to 98,000 pounds. The core represents one-third to one-half of the development effort for an engine. Furthermore, the engines that power the STOVL and CTOL aircraft are entirely interchangeable."

With a first flight date scheduled for next year, the 400 or so people on the Pratt & Whitney JSF program are themselves hitting the afterburners. "Every program is a challenge in its early stages," says Jepson, who is in charge of the engine build teams for the JSF program in Florida. "We’ve been here before. The F119 was a challenge. The -229, the latest engine that powers the F-16 and F-15, was a challenge. We can’t let challenges impact our delivery dates. We hustle."

Article by Eric Hehs; photos by Greg Roberts (Pratt & Whitney)