v2.0.2 / 01 nov 04 / greg goebel / public domain
* While the US military's current first-line aircraft remain formidable weapons, their basic designs are decades old, and the services would like to obtain more modern aircraft to fit their future needs. The US Air Force, the US Navy (USN), the US Marine Corps (USMC), and the British Royal Navy (RN) have now committed to a new, advanced attack aircraft, the Lockheed Martin F-35 "Joint Strike Fighter (JSF)". This document outlines the history of the JSF effort.
* The Joint Strike Fighter program began with defense reviews conducted by the Clinton Administration after taking office in 1992. At the time, several government organizations were working on next-generation strike aircraft.
The US Navy had been working in secret on an advanced stealthy strike aircraft named the "A-12 Avenger II", but the program ran into financial trouble, and was cancelled shortly after going public in 1991. The Navy requirement remained open and evolved into a new effort designated "Attack / Fighter - Experimental (A/F-X)". The USAF was also considering a next-generation strike aircraft and a replacement for the F-16, with the designation of "Multi-Role Fighter (MRF)".
On another track, the US Marine Corps had been interested in a follow-on for their AV-8B Harrier II "short takeoff, vertical landing (STOVL)" attack aircraft, while the British Royal Navy wanted a next-generation STOVL fighter to replace their Sea Harriers. The two services collaborated on STOVL research in the late 1980s, with control of this effort finally picked up by the US Defense Advanced Research Projects Agency (DARPA) in 1989.
DARPA conducted conducted studies on new STOVL concepts over the next few years. By this time, the USMC had expanded their vision from a STOVL aircraft to replace the Harrier to one that would also replace their F/A-18 Hornet fighters. DARPA began to see the advanced STOVL fighter as a basis for a "conventional take-off and landing (CTOL)" aircraft as well, and the concept became known as the "Common Affordable Lightweight Fighter (CALF)".
In September 1993, the Clinton Administration killed the Navy A/F-X and Air Force MRF projects. The Pentagon was allowed to set up a research office to investigate "Joint Advanced Strike Technologies (JAST)" applicable to fulfilling these requirements at some unspecified time in the future.
A few months later, the US Congress mandated that the DARPA-led CALF research effort be merged into the JAST office. At first, the JAST office was generally regarded as just another inept defense boondoggle, but under the direction of Air Force Major General George Muellner it quickly attained a critical mass for actually building a next-generation strike fighter. Muellner wanted to build a "universal fighter" that would fulfill the needs of all the participants, and was able to get everyone pulling in more or less the same direction.
One of the primary goals of the JAST effort was "affordability". In the wake of the end of the Cold War, procurement funds for new combat aircraft were hard to come by, particularly because the Clinton Administration was determined to balance the budget while simultaneously conducting expensive military interventions all over the world. Although the US economy was in an extraordinary boom at the time, this ironically led to further constraints on the military budget, since the services had to increase pay to keep their people.
* The JAST concept that emerged did not define a single aircraft, but three different aircraft based on common technology:
The JAST also had to perform a secondary air-defense mission, using air-to-air missiles (AAMs) to defend itself, or to protect fleet assets from airborne intruders. High performance was not a requirement, though of course it was desireable. Performance was specified to be comparable to existing F-16s and F/A-18s operating in the strike role, though any incidental improvements in performance were welcome. Nominal top speed was specified as Mach 1.5, or about 1,600 KPH (1,000 MPH).
The answer to all these requirements was to develop a baseline CTOL aircraft at a base price for the USAF requirement, and variants at incrementally higher prices with the features needed for CV or STOVL operations. The STOVL variant was to have essentially the same performance as the other two variants. This was an ambitious requirement, but the DARPA STOVL studies had indicated that STOVL technology had finally matured to the point where such a thing was possible.
The JAST concept that emerged envisioned a stealthy, single-seat aircraft, with a high degree of cockpit automation to make life easier for the pilot. It would accommodate sensors and avionics adequate for its its mission. It would be fitted with an advanced radar system, working in conjunction with other electronic and infrared systems for defense and all-weather attack, but would not feature a highly optimized integration of systems in order to lower costs and preserve flexibility.
The stealth requirement meant that JAST would be able to accommodate a small warload internally, consisting of two guided munitions and a pair of AAMs, along with a much larger warload on underwing stores pylons. The JAST would operate strictly with internal weapons during initial phases of an air campaign, allowing it to perform stealthy strikes to suppress air defenses or hit heavily defended targets, and then carry heavier external loads in later phases of a conflict. The JAST office referred to this operational concept as "first day stealth".
* Major aircraft manufacturers began to consider JAST designs in 1994. The JAST program office issued a request for proposals in March 1996. A short time later, the project name was changed to "Joint Strike Fighter (JSF)" to indicate that they were working on real flying hardware, not blue-sky design concepts. Three companies offered proposals to meet the JSF request:
The STOVL version of the McDonnell Douglas JSF proposal featured a main engine with side nozzles under the rear fuselage that could be opened to provide vertical thrust, as well as a vectored exhaust that could be rotated 20 degrees in any direction, both to assist in short takeoffs and for enhanced flight maneuverability. The STOVL variant also included a a separate "lift engine" installed vertically behind the cockpit for vertical thrust.
The CTOL variant for the USAF deleted the lift jet and replaced it with a fuel tank for longer range. The CV variant for the USN had more wing area for carrier landings, and other appropriate gear for carrier operations.
The STOVL version of the Boeing JSF proposal drove engine thrust forward to a pair of vectored lift nozzles under the aircraft's center of gravity. The nose intake scoop hinged forward to provide greater airflow for a short takeoff. To achieve the thrust needed for STOVL operation, Boeing began work with Pratt & Whitney to uprate the thrust of the already powerful F119 afterburning turbofan, designed for the Lockheed Martin F-22 Raptor.
Boeing's JSF configuration for the CTOL and CV variants deleted STOVL features in favor of greater fuel capacity. The stubby wing, which had a span short enough to allow it to be conveniently stowed without folding even on the smaller British carriers, also provided large fuel capacity, giving the Boeing design considerable range.
The lift fan approach had the advantage that it minimized hot exhaust ingestion back into engine, a common problem with STOVL designs that robs them of vertical thrust. The configuration had some resemblance to that pioneered by the Russian Yakovlev Yak-41 "Freestyle" STOVL fighter, which did not enter production.
For the CTOL and CV variants, the lift fan was deleted and replaced with additional fuel tanks. The wings and tail were smaller for the CTOL version.
The STOVL designs from all the manufacturers had the same infrared signatures in normal flight as the CTOL and CV variants. This was an improvement from the Harrier, whose four rotating exhaust nozzles arranged around the aircraft's center of gravity were regarded by critics as providing a perfect "bulls-eye" for heat-seeking missiles.
* In 1996, the Boeing and Lockheed-Martin concepts were selected as finalists, and both companies then began formal development of demonstrator aircraft. Being dropped from the competition was a major blow to McDonnell Douglas, and one of the contributing factors to the company's subsequent buyout by Boeing. The Boeing demonstrator received the designation "X-32", while the Lockheed Martin demonstrator received the designation "X-35".
In 1998 Boeing conducted a major redesign of its JSF concept to reduce cost and weight, changing its pure delta-wing configuration by adding a conventional horizontal tailplane. The Boeing demonstrator was not changed to the new configuration.
In production, Boeing envisioned its USAF JSF CTOL variant to have an empty weight of 10 tonnes (22,000 pounds), a wingspan of 11 meters (36 feet), and a length of 13.7 meters (45 feet). The USN CV variant would be similar, except it would weigh almost a tonne more, due to the need for stronger landing gear, arrester hook, and other equipment required for carrier landings. As with the original Boeing JSF demonstrator design, the wings did not fold. The USMC and Royal Navy STOVL variants were to weigh 10 tonnes, though the wings were to be clipped, with a span of 9.15 meters (30 feet).
* First flight of the Boeing "X-32A" CTOL variant was on 18 September 2000. First flight of the "X-35B" STOVL variant was in March 2001, with vertical flight testing beginning in June 2001.
First flight of the Lockheed Martin "X-35A" CTOL variant was on 24 October 2000, with first flight of the "X-35C" CV variant on 16 December 2000. The X-35A was then updated the "X-35B" STOVL configuration by addition of the lift fan system and other hardware, with initial vertical flight testing in June 2001.
In October 2001, the Lockheed Martin X-35 was selected as the winner of the competition. Boeing was perceived as having the edge in management, while both companies were rated equally on cost and support. However, the Lockheed Martin design was seen as involving lower risk, with the lift-fan concept for the STOVL variant scoring particular points on the win.
The loss of the contract was seen as a blow to Boeing, but the company has plenty of other defense work, and nobody is seriously considering any sort of bailout. It is possible that Boeing will leverage technology developed for the X-32 into robot strike aircraft known as "uninhabited combat aerial vehicles" that the company is now developing.
Lockheed Martin expects to build 3,000 JSFs, including:
In 2004, the Air Force also expressed interest acquiring hundreds of the F-35B STOVL variant for the battlefield close-support role. This will no doubt affect the total numbers of the USAF JSF buy, almost certainly downward, but Air Force brass say they can't give clear numbers until a US Army force reorganization is completed. A formal decision isn't expected until 2006. The USAF is interested in a range of options for their STOVL variant, but JSF program officials have insisted that they do not want to develop a fourth JSF variant strictly for the Air Force.
The actual mix of British aircraft similarly remains a bit up in the air. While the British have committed to the F-35B STOVL variant as a Harrier replacements, the requirement in that case is from 60 to 90 aircraft. The remainder are likely to be the F-35C CV variant. Two new British carriers are being built to go into service in the next decade. They will be built to handle STOVL aircraft, but will be designed to accommodate catapult and arresting gear in case the decision is made to operate fixed-wing aircraft as well.
Estimated cost for the CTOL version will be about $45 million USD, with the price climbing to $55 million to $60 million for the STOVL version. There are those who wonder if the expected production quantities of F-35s will actually meet projections, since the US is now moving forward enthusiastically on development of a "unmanned combat air vehicle (UCAV)" that would be cheaper and capable of taking on many of the JSF's expected roles. However, the JSF is more or less a present capability while the UCAV is a future capability. Speculations on future possibilities are just that, speculations.
* The JSF program is now in the "system design and development (SDD)" phase. The SDD plan involves the construction of 14 flight-test aircraft and eight static-test airframes. First flight of an SDD aircraft, an F-35A CTOL variant, is expected in 2006, with an F-35B STOVL variant flying in 2007. The F-35B ran into problems with "weight creep", with Lockheed Martin performing substantial redesign of structural elements and other features to trim well more than a tonne off the fighter's weight, allowing it to return to a carrier without having to dump heavy munitions.
Low-rate initial production is expected in 2007. The USMC and the USAF will get their first evaluation aircraft in 2008, with the Marines reaching operational capability in 2010 and the Air Force reaching operational capability in 2011. The USN and the UK will get their first evaluation aircraft in 2010, leading to operational status in 2012 for both. Of course, schedules tend to change during the course of a program.
Australia, Canada, Denmark, Italy, the Netherlands, Norway, and Turkey are all contributing to the JSF SDD program to allow them to observe progress and provide inputs for possibly future purchases of the aircraft. Lockheed Martin believes that international sales may double the ultimate production quantities of the JSF.
* Lockheed Martin released a "finalized" design for the production F-35 in the summer of 2002. The following discussion describes most of the details in the present tense for readability, though of course no production aircraft has flown yet, and there still may be some changes in the configuration before that happens.
The F-35 has a nose 12 centimeters (5 inches) longer than the X-35 demonstrator, while the tailplane has been moved back 5 centimeters (2 inches), and the tailfins have been rearranged a bit. All flight controls are electric, in principle providing easier maintenance and greater combat survivability than hydraulic systems.
Compared to the USAF F-35A CTOL variant, the USN F-35C CV variant has a larger wing and tail, giving it better range and good low-speed carrier landing characteristics. The wing features folding wingtips. Of course, the F-35C has stronger landing gear and an arrester hook. The F-35B STOVL version has shorter tailfins, implemented as part of the weight-reduction redesign.
The Air Force F-35A has a refueling-boom socket behind the cockpit, while the F-35B and F-35C have a retractable refueling probe on the right side of the nose. The tricycle landing gear, with a forward retracting nosewheel and inward-retracting main gear, has single wheels on all assemblies in the F-35A and F-35B. The F-35C differs in having twin wheels on the nose gear to handle hard carrier touchdowns.
The F-35's airframe makes heavy use of composite materials, with much work placed on reducing the cost of composite assemblies, which have traditionally been extremely expensive. In fact, the F-35 has been designed to be as cheap to manufacture as possible, using the latest computer-aided design and manufacturing tools.
The F-35 is powered by a modified version of the P&W F119 engine, designated the "F135". While it is as powerful as the original F119, it is much cheaper, as it uses lower-cost components at the expense of greater weight. It has the same thrust levels as the F119, with 151 kN (15,420 kgp / 34,000 lbf) dry thrust and up to 222 kN (22,675 kgp / 50,000 lbf) afterburning thrust. The engine intake ducting is arranged in a "serpentine" fashion to eliminate radar reflections from the compressor blades.
Although the P&W F119 engine was selected as the basis for the different engine options of the JSF, in 1995 the US Congress indicated a need for an "Alternate Engine" as a backup plan. The GE F120, originally designed for the F-22 Raptor program in competition with the P&W F119, was selected as the Alternate Engine, and refinements to the F120 for F-35 are under development by a collaboration of GE, Allison, and Rolls-Royce. Thrust levels will of course be similar to those of the F119.
The shaft-driven lift fan for the STOVL F-35B is built by Rolls-Royce / Allison, and provides up to 80 kN (8,150 kgp / 18,000 lbf) of lift thrust.
* The F-35 has two weapons bays, each of which can accommodate a single "Joint Direct Attack Munition (JDAM)" GPS-guided bomb and an AIM-120 Advanced Medium-Range Air to Air Missile (AMRAAM). The F-35A and F-35C can carry two 900 kilogram (2,000 pound) JDAMS internally, while the STOVL F-35B is limited to internal carriage of two 450 kilogram (1,000 pound) JDAMs. The F-35A and F-35C variants have bulged weapons bays to accommodate the larger munitions; the F-35B's weapons bays also have less internal volume. The two bays have two doors each, with the AMRAAM fitted on a launch rail on the inner door.
Four stores pylons can be attached to all variants to provide a much larger warload, at the expense of stealth. The inner pylon on each wing is rated for up to 2,270 kilograms (5,000 pounds), while the outer pylon is rated for up to 1,135 kilograms (2,500 pounds).
Only the USAF F-35A has a built-in gun, with an "Advanced 27 Millimeter Cannon", an improved version of the Mauser BK-27 revolver-type cannon, in the left wingroot. The other variants do not have a built-in gun, but can accommodate a cannon pack plugged into one of the weapons bays.
* Northrop Grumman is developing the sensor suite for the F-35. The initial design assumption was that the JSF would be a consumer of sensor data, obtaining information from specialized intelligence-gathering aircraft, satellites, and other sources. This approach promised to keep costs down. However, as the pieces began to fit together, something different emerged. This was partly due to the "bottom-up" realization that the new technologies being developed for the JSF were far more powerful than had been considered; and to the "top-down" realization that the numbers of expensive specialized intelligence-gathering aircraft would be small, while there could be thousands of JSFs.
Now the F-35 is seen more as a producer of sensor data, with each aircraft interacting through high-speed data links with other aircraft to provide greater "electronic domination of the battlespace". If the other aircraft are F-35s, they will be able to cooperate to provide a capability greater than the mere sum of the parts.
The heart of the F-35's sensors is the Northrop Grumman AN/APG-81 radar, based on the AN/APG-77 "active electronically scanned array (AESA)" developed for the Lockheed Martin F-22 Raptor. An AESA consists of an array of "transmitter-receiver (T/R)" modules linked by high-speed processors. Different T/R modules in the array can be allocated to different tasks, with more modules allocated to tasks that require greater power or sensitivity.
The F-35's AN/APG-81 provides a range of functions, acting as a multimode radar; active jamming system; passive electronic defense system; and communications system. The system generates signals over a wide range of frequencies and pulse patterns in an unpredictable fashion to ensure "low probability of intercept", allowing the F-35 to "see but not be seen." The AN/APG-81 uses improved technology compared to the F-22's AN/APG-77, but airframe constraints mean that it has fewer T/R modules, limiting it to about two-thirds the range (165 kilometers / 90 nautical miles) of the AN/APG-77.
The F-35 is also fitted with additional sensor systems, including a an "infrared search and track (IRST)" system for defense and air-to-air combat, and a targeting system for precision attack on ground targets.
The IRST system is known as the "distributed aperture infrared system (DAIRS or DAS)". DAS includes six IR sensors mounted on different points of the fuselage to provide full-sphere IR detection and tracking. DAS can identify and pinpoint both incoming missiles and airborne targets.
Targeting is performed by the "electro-optical targeting system (EOTS)", featuring a forward-looking infrared (FLIR) imager; a CCD TV camera; a targeting laser; and a laser spot tracker. Unlike typical contemporary targeting systems, EOTS is not turret-mounted. It has a wide aperture that is blended into the aircraft's nose contours, covered by a window that is opaque to radar, and remains operational through the entire mission. It is derived from technology developed for the Lockheed Martin "Sniper" targeting pod.
Other avionics include a Northrop Grumman "communication, navigation, and identification (CNI)" system and a countermeasures suite provided by BAe Systems.
* The F-35's software collects the inputs from all the sensors, as well as inputs relayed over a high-speed datalink, to provide sensor fusion and seamless data display. The software is executed on an "integrated core processor (ICP)". The ICP serves as a central "brain" for the aircraft, integrating all the other electronics systems and coordinating them for display to the pilot, and also executing the pilot's commands. This system is vitally important, since the F-35 is a single-seat aircraft, and the pilot needs help to carry out his or her mission.
Northrop Grumman selected a "commercial off-the-shelf (COTS)" processor system for the ICP. The F-35 ICP is cheaper than the F-22's "Integrated Core Processor", which was designed a decade ago, but is an order of magnitude more powerful.
One of the functions of the central processing system is to provide "automatic target recognition and classification (ATRC)". It can often identify specific targets, and if it can't say exactly what a target is, it can at least show which targets are different.
The processing power of the F-35 has presented the electronics system developers with a formidable software challenge. The F-22 Raptor uses about 2.5 million lines of software, but the F-35 will use 5.6 million lines of code. The F-35 not only has a more advanced electronics system, but it operates in both air-to-air and air-to-ground modes, and is being built in three different versions. The software design strategy is focusing on modularizing the code so that the portions that are unique to each F-35 variant can be isolated, and the remaining code used as-is on all three variants. The portions that are unique to each variant are a minority, about 1.1 million lines.
In addition, the code is largely executed by an interpretive software layer known as "middleware" that isolates the code from the specific details of the processor used. In principle, this will allow software to be ported to new processors as they become available, requiring only new middleware and maybe a few software tweaks. Interestingly, the code is written in C/C++, strongly suggesting that the military's effort to create Ada as a standardized programming language for defense projects was a dead end.
The current plan is to have a comprehensive but minimal software suite for F-35 operational introduction, and provide improved releases to bring the F-35 up to full combat capability. F-35 electronics system designers hope to leverage off work done for the F-22 Raptor.
* The pilot receives inputs from the F-35's electronic systems using an advanced cockpit layout, featuring a full-panel-width display, with dimensions of 20 by 50 centimeters (8 by 20 inches), plus a secondary flight display array, along with "hands on throttle and stick (HOTAS)" controls. It does not have a "head-up display", however, with this function taken over by a "helmet-mounted display" being developed by Visions Systems International, a collaboration of Kaiser Electronics and Elbit of Israel.
The "smarts" of the F-35 will be particularly appreciated by pilots flying the F-35B STOVL version. Short takeoffs in the Harrier are a troublesome affair that require the pilot to have "three hands": one for the throttle, one for the stick, and the third for the lever that controls the direction of the Harrier's swiveling exhaust nozzles. An F-35B pilot, in contrast, simply flies the plane with stick and throttle, with the software handling the fine details of short takeoff.
While the Harrier has reaction control thrusters driven by engine bleed to provide low-speed maneuverability, the F-35B simply modulates the four points of its vertical-lift system -- the pivoting exhaust, the two wing exhaust ducts, and the lift fan -- to provide control. This trick would be difficult or impossible to do manually.
The X-35 prototypes are fitted with a Martin-Baker Mk.16E ejection seat. Production F-35s are supposed to use a new seat from the "Joint Ejection Seat Program".
* The NATO air campaign against Yugoslavia over Kosovo in the spring of 1999 revealed a shortfall in electronic warfare (EW) capabilities. EW missions during the Kosovo campaign relied heavily on the venerable EA-6B Prowler, and Prowler crews were stretched to the limit. The F-35 is now being seriously considered as a EW aircraft to supplement and eventually replace the Prowler.
Manufacture of the "EF-35", to give it a plausible name, would certainly even further increase production quantities for Lockheed Martin. Since the EW mission is regarded as requiring at least two aircrew, the EF-35 would have to have tandem seats, with the back-seater sitting where the lift fan / additional fuel tanks are in the current variants. However, the EF-35 is strictly a future prospect at the present time, with no commitment even to a formal investigation phase. Other future options for the F-35 are also under consideration, but any commitment to them seems a bit premature until the JSF actually goes into service.
* The v1.0 version of this document was released in December 1999, but at the time the JSF hadn't even flown and details were sketchy, so in May 2000 I merged it with an earlier document on the F-22 Raptor. By late 2001, the JSF prototypes had all flown and a winner had been selected, so I split them back out again and came up with the v2.0.0 version of this document.
Elimination of the Boeing entry from the competition made documenting the program much simpler. Confusion between systems developed for the Boeing and Lockheed Martin aircraft was a particular problem, and in some documents it wasn't clear if the confusion was just in the reader or the writer as well.
* Sources include:
The JSF Program Office website was a very useful source of many details.
* Revision history:
v1.0 / 01 dec 99 / gvg / Merged into F-22 document in May 2000.
v2.0.0 / 01 jan 02 / gvg / Reconstruction & rewrite.
v2.0.1 / 01 nov 02 / gvg / Minor update.
v2.0.2 / 01 nov 04 / gvg / Minor update.