*Bear in mind that this aircraft (at the time of writing) is extremely new and many of its capabilities are classified. Information presented here may be based on outdated specifications.
Whether you view it as one of the most advanced strike fighters in service today, or the biggest waste of money to ever sprout wings and fly, the F-35 Lightning II certainly gained notoriety before it even entered service due to its protracted development as the most expensive weapons development project in U.S. history at over $1.5 trillion.
That’s right, the F-35 is one of the new age wonder planes that everyone keeps saying is incredible, but for some reason, it persists in having technical difficulties. So let’s take a look at the plane that cost so much money to develop that it just wouldn’t die.
History & Design Competition
The origins of the Joint Strike Fighter Program (JSF) that the F-35 developed from are rather murky. Much of this is seemingly due to the hostility that Short Take-Off Vertical Landing (STOVL) designs face within the defense establishment. While the famous British Harrier and subsequent designs have demonstrated that the concept works, with the most famous use occurring over the Falklands in the early-1980s, most of the designs for future jump-jets to supersede the Harrier II have been purely theoretical or seemingly some obscure department’s pet project to keep the concept alive.
In 1986, a joint U.S-U.K. program was signed called the Advanced STOVL (ASTOVL) project which aimed to ID candidates for a Harrier replacement that would be ready to enter service by 2000-2010. The plan was for the countries to develop a set of requirements and design an operational aircraft by 1995. In 1987 the U.S. Marine Corps (USMC) was identified as the most influential customer with the most money since they were potentially looking to replace both their AV-8B Harrier IIs and F/A-18 Hornets, thus converting to an all-STOVL force that was independent of the navy’s supercarriers. This also tacked on the additional requirements of air-to-air and all-weather ground attack capabilities. The Marines upgraded their AV-8Bs to the plus configuration, extending their service into the 2000s when the ASTOVL aircraft would be ready. Some of the key requirements were a maximum operating weight of 24,000 pounds, supersonic speed, stealth, internal weapons capacity, and an operating radius of 500 miles with 6,000 pounds of offensive weaponry. By 1987, the USMC and Royal Navy (RN) were committed to the basic requirements, but the problem was that neither service had enough money. Big surprise there! The Marine Corps isn’t exactly known for massive budgets and the Fleet Air Arm of the RN has been slowly gutted since the end of the Cold War. The secrecy surrounding stealth also presented a barrier to international cooperation. The program and lack of interest in STOVL were seemingly grinding to a halt in development hell (Sweetman, 2004, p. 23 – 27).
Defense Advanced Research Projects Agency (DARPA) took up the project and began a series of studies funded by Lockheed Skunk Works, McDonnell Douglas, and General Dynamics. These studies aimed to solve 2 problems of most STOVL projects:
- Hot, high velocity exhaust causing ground erosion and flying debris.
- Hot gas ingestion.
Thus, in 1993, Lockheed engineer Paul Bevilaqua patented a fan system which was evolved from earlier tandem fan designs. It was different because the forward fan was separate from the core airflow at all times. The fan shut down in cruising flight and was installed with its axis vertical. DARPA began looking into earlier Gas-Coupled Lift Fan (GCLF) and Shaft-Driven Lift Fan (SDLF) designs. Another contractor design earlier in 1990 looked at files detailing the Lift-Plus-Lift/Cruise (LPLC) idea with a vertical jet engine behind the cockpit. At the time in 1990, the future of fighters was uncertain given the aging F-14s and A-6s. The F-16 and F/A-18s were in production but had no direct successors. European countries were developing the Saab Gripen, Dassault Rafale, and Eurofighter Typhoon while the MiG-29 and Su-27 had entered service in the 1980s and remained credible threats.
While the A-12 was canceled in 1991, the navy eventually developed the F/A-18E/F Super Hornet, but there was no dedicated attack aircraft directly in the works. Various ideas were kicked around including a modified F-16 with a delta wing configuration, but with the idea of replacing the Air Force’s F-16s and the Marines’ F/A-18s and AV-8s with a common aircraft with STOVL capabilities, DARPA came to the idea of dual-service fighter under the project name of Common Affordable Lightweight Fighter (CALF). CALF was anticipated to save billions in research, development, and production. Lockheed Martin and McDonnell Douglas were awarded contracts and both proposed stealthy designs with internal weapons bays for 2,000-pound bombs and a pair of AIM-120 AMRAAM missiles. They also included auxiliary lift fans behind the cockpit (Sweetman, 2004, p. 31 – 38).
The difference was that Lockheed planned to use a modified P&W F119 engine with a clutch in front of the engine connecting the low-pressure shaft to the driveshaft driving another fan. McDonnell Douglas planned to use a modified GE F120 which would bleed high-pressure air from the compressor and move it to the lift-fan turbine (Sweetman, 2004, p. 38).
The prototype aircraft would be designated the X-32A for the CTOL version and the X-32B for the STOVL variant. In 1993, there was still no formal requirement for an advanced STOVL plane or a CTOL derivative, but the prospect presented an opportunity for Boeing. While Boeing hadn’t delivered a manned combat aircraft to the Air Force in 30 years, much less a fighter since the 1930s, it had a reputation as a defense contractor that handled difficult projects. Convinced that post-Cold War budgets would be tight, Boeing began designs for a low-cost, multi-service, multi-purpose, long-range fighter with STOVL capability similar to the Harrier and an empty weight close to that of an F-16. It came up with a delta-winged plane smaller than either Lockheed’s or McDonnell Douglas’ designs (Sweetman, 2004, p. 38 – 39).
Following the 1992 election, shifts in Pentagon policies changed the CALF program with a focus on the consolidation of projects. A new program called Joint Advanced Strike Technology (JAST) focused on developing aircraft, weapons, and sensor technology for future tactical aircraft using updated computer technology, low-cost designs, and the then-emerging internet as tools. Eventually, other programs like A/F-X, MRF, and CALF were absorbed into JAST under the direction of U.S. Air Force (USAF) Major General George Muellner. CALF would become the center around which JAST was formed (Sweetman, 2004, p. 39 – 41).
One of the issues with JAST was the idea that it could create a single fighter for every service. The last time this was tried under Pentagon direction, the resulting F-111 (particularly the B variant) suffered development problems. Another issue that arose was the debate over a single-engine versus a twin-engine design. The navy has long preferred twin-engine designs with the reasoning that if one of the engines failed, the second one would be able to get the pilot and aircraft back home. The JAST office thus sponsored an analysis done by Georgia Tech Research Institute and Johns Hopkins University. They concluded that a single-engined aircraft would have equal or better survivability than a twin-engined plane. The only time a twin-engine aircraft would survive a hit was if one engine was knocked out, but the other AND all other vital systems were left intact. The studies showed that this would hardly ever happen. It certainly wasn’t the case with Muellner who had 5,300 hours as a fighter pilot, 690 combat missions over Vietnam in F-4 Phantoms, and was shot down as well (Sweetman, 2004, p. 41 – 42).
In 1994, British Aerospace, long associated with the Harrier, and McDonnell Douglas was brought in. By the end of 1994, the four U.S. companies competing were Lockheed Martin, Boeing, McDonnell Douglas, and Northrop Grumman. In 1995, McDonnell Douglas abandoned the gas-driven lift fan in favor of the LPLC. They realized that the GCLF was too vulnerable to fragmentation damage. One of the benefits of the LPLC design over the GCLF and SDLF designs was that it eliminated the gas and mechanical transmission systems (and thus volume and weight). The problem was that it was essentially two separate engines, a lift engine, and the main engine. This presents a logistics difficulty with the need to care for two different systems. The USMC logistics people promptly started climbing the walls. Another disadvantage was that the LPLC system has 3 engine cycles. The life/cruise engine starts and stops once and the lift engine twice. Furthermore, if either one fails, the fighter can’t land on a ship (Sweetman, 2004, p. 43).
McDonnell Douglas’ design stemmed from the X-36 demonstrator declassified in 1996. However, both McDonnell Douglas and Lockheed Martin moved away from their canard designs due to the fact that, in order to land on a carrier with as big a weapons load as possible, the plane needed a bigger wing to keep the aircraft’s approach speeds down at higher weights. A bigger wing extends forward and competes for space with the fore planes. There was also an anti-canard prejudice in the establishment. F-16 designer Harry Hillaker was quoted as saying, “the optimum location for a canard is on somebody else’s plane.” At the time, in 1995, the Eurofighter Typhoon and Saab Gripen were also dealing with handling issues (Sweetman, 2004, p. 45).
Lockheed Martin’s design eventually took the shape of the F-22 given that it was the primary contractor for the design and could fall back on it. McDonnell Douglas was more radical. Their design had no separate vertical tails. Their horizontal surfaces were so shallow that it’s doubtful they would’ve been useful for yaw control. The main engine would be used for yaw and pitch maneuverability. Their design was dropped for being deemed too risky. Boeing also went ahead with its design (Sweetman, 2004, p. 45).
In March of 1996, the JAST office gave a request for proposals which was due in June. The project’s name was then changed from JAST to Joint Strike Fighter (JSF) reflecting its backing by operational requirements. Eventually, Boeing and Lockheed Martin were selected to compete for production contracts (Sweetman, 2004, p. 45 – 47).
The prototypes were known as Concept Demonstration Aircraft (CDA) and had to prove the performance characteristics, low-speed characteristics for carrier landings, and STOVL capability. However, the exact requirements were not set in stone and were subject to change. The tracking of progress was to be conducted by a series of Joint Interim Requirements Documents (JIRD). Instead of trying to meet a set of requirements with the cost being a by-product (often resulting in cost overruns), the JSF program was to have the contractor and customer agree on the design to be built for the money. Requirements had to be evaluated before changes were to be made in order to control costs. If a cost increase was the result, then savings had to be found elsewhere (Sweetman, 2004, p. 50 – 51). As you can see, this didn’t really pan out since the program went massively over budget.
The preliminary requirements of the CDA were based on 3 things.
- Other aircraft would handle more severe air-to-air threats. The F-35 didn’t need to replace the Air Force’s F-22 or the Navy’s Super Hornet. 70% of the requirements were weighted towards air-to-ground missions. Unlike the F-22, the X-35 had a higher wing loading, no in-flight vectored thrust (in other words, it wasn’t designed to be as agile as the F-22), and supercruise wasn’t required. The standard air-to-air missile would be the AIM-120 AMRAAM and not the AIM-9X Sidewinder.
- First day stealth. First day missions would be performed in a stealthy configuration with internal weapons. Later missions would carry more weapons externally to attack more targets.
- The JSF was designed around the idea that it would carry smart bombs and be just as effective as a plane with a larger load of dumb bombs.
All three services wanted 2x Joint Direct Attack Munitions (JDAMs), 2x AMRAAMs, and agility comparable to current fighters. The USMC and USAF wanted an unfueled radius of 500 miles and a range of 600 miles. The Navy wanted 600 miles at the minimum. Stealth requirements initially indicated the potential for a very low Radar Cross Section (RCS) with Radar Absorbent Material (RAM) and other features. In terms of STOVL capability, the P&W F119 was the only flight-rated engine powerful enough to do the job (Sweetman, 2004, p. 51 – 54).
Lockheed Martin began tackling the issue of how to design the STOVL version. The challenge was for the engine to have enough thrust to support the aircraft’s landing weight which would be directed vertically through the plane’s center of gravity while at the same time keeping the jet velocity and temperature at reasonable levels to avoid tearing up the ground or hot gas ingestion. The shaft-driven lift system did this by increasing the bypass ratio in vertical flight, moving twice as much air at lower than average velocity, and using the driveshaft to transfer much of the engine’s total energy to a point ahead of the center of gravity (Sweetman, 2004, p. 54).
The engine had a “three-bearing” vector nozzle where the tailpipe had two angled segments connected by rotating bearings. When the segments rotated in opposite directions the entire assembly bent through 100 degrees and provided a small degree of reverse thrust. Further adjustments could also move the tailpipe side to side (Sweetman, 2004, p. 55).
A clutch on the front of the engine linked a driveshaft to the lift fan. At full power, the lift fan generated 18,000 pounds of thrust. The fan had a folding nozzle and could direct thrust 60 degrees aft during transition flight. In vertical flight, the exhaust could be fed to lateral roll ducts which terminated in nozzles beneath the wing roots. This system was advantageous for pitch and roll because energy could be transferred among the four “lift posts” (the main engine, lift fan, lift ducts) rather than bleeding air and power from the engine. Roll was controlled via the roll ducts, pitch was controlled by shifting energy between the aft nozzle and lift fan, and yaw was controlled by moving the aft nozzle side to side. Since the Air Force and Navy versions did not require STOVL, a fuel tank replaced the lift fan, giving them greater range (Sweetman, 2004, p. 55 – 56).
In contrast to Lockheed, Boeing created a radical design which was essentially a flying wing. Known as AVX-70, later as the X-32, it was a thick delta wing with a leading edge sweep of 55 degrees. The wing could hold 18,000 pounds of fuel, more than twice the F-16’s internal capacity. The engine was attached to the underside of the wing in a large nacelle which included side-opening weapon bays (Sweetman, 2004, p. 56 – 57). In my opinion, Boeing’s design looks like some kind of freakish flying whale…from space.
The engine was forward of the center of gravity with the fan face behind the cockpit. A straight duct ran from the turbine to the nozzle. In the STOVL version, the duct was replaced by a lift module, incorporating two retractable vectoring nozzles, and a blocker built into the exhaust.
The inlet itself presented a challenge for stealth. Short and sharply curved, it partially blocked line-of-sight to the fan face and ensured radar signals would bounce off Radar Absorbent Material (RAM) coated duct walls before escaping, but that alone didn’t meet Radar Cross Section (RCS) targets. Therefore, the front had variable angle blocker vanes installed which would close in cruising flight and open for takeoffs and landings (Sweetman, 2004, p. 58).
The delta wing presented problems for carrier landings and had only been used by a Navy with the Douglas F4D Skyray for 8 years. Like other highly swept wings, it needed high nose attitudes at low speeds which meant poor over-the-nose visibility. No horizontal tail meant you couldn’t install lift-boosting flaps, and the trailing-edge elevons had to control both pitch and roll making quick responses for carrier landings difficult. To remedy this, Boeing installed a vortex flap above the inboard leading edge (Sweetman, 2004, p. 61 – 62).
The story between the competitors is quite long, but in the end, Lockheed Martin did better. In late 2000, Lockheed’s STOVL system was working properly, whereas Boeing missed its test goals and couldn’t perform a vertical landing with a complete airplane. Lockheed’s X-35 wowed audiences when it performed a short takeoff, went supersonic, and then landed vertically. While both Lockheed’s X-35 and Boeing’s X-32 met design requirements, in the end, the X-35 was deemed to have higher performance and Lockheed won the competition (Sweetman, 2004, p. 84).
Aside from all the differences, there’s of course the issue of what to do with the gun. Much of the debate over guns on modern fighter jets stems from the experience with the F-4 Phantom over Vietnam when air-to-air missiles were still somewhat unreliable and the F-4 had to have a gun hastily added to it. In the Arab-Israeli War in 1973, guns accounted for 70% of Israeli kills. However, by the late-1970s, the introduction of the AIM-9L (and similar missiles) offered a wider launch zone. In 1982 over Lebanon, 93% of kills were accomplished with missiles. Over the Falklands that same year, there were no air-to-air gun kills. By 1998, the issue was settled. The Marine and Navy aircraft would have provisions for a gun pod, and the USAF version would have an internal gun (Sweetman, 2004, p. 70).
After Lockheed’s X-35 was chosen as the winner, it remained to be decided how to designate the aircraft since the previous fighter was Northrop’s YF-23A. Therefore, conventions would have the JSF as the F-24 or F/A-24 given its air-to-ground role. While some in the air force recommended the designation of F-24, the number 35 was already cemented in people’s heads, and the designation of F-35 was confirmed in June 2002 (Sweetman, 2004, p. 94). Yeah, it was basically decided on a whim.
The F-35A made its first flight on 15 December 2006 over Fort Worth (Newdick, 2015, p. 219).
Specific Design Considerations of the F-35
Interestingly, due to the requirements of STOVL and the need to operate aboard navy carriers, the Marine and Navy variants drove the design of the internal layout and the wing and tail configurations respectively.
The software for the plane was released in several blocks. Block 0 included basic vehicle management and flight controls used for initial testing. Block 1 included avionics and allowed for single ship operation. Block 2 included multi-ship operation and interoperability with JSTARS and AWACS. It served as the Initial Operating Capability (IOC) standard for the USMC. Block 3 aimed for interoperability with the UK and European systems. It serves as the IOC for the USAF, USN, and the UK (Sweetman, 2004, p. 99).
The F-35 was originally meant to be delivered with a choice of engines between the P&W F135 or the GE/Rolls Royce F136. Both of which are derived from the F119 that powers the F-22. However, funding and interest for the F136 engine stagnated and its development stopped. While the F135 serves as the engine in the production variants, it does seem that in recent years, P&W has had difficulty with quality control (Sweetman, 2004, p. 106 – 107).
In case you didn’t know, there are three versions of this aircraft. The main differences are in their takeoff and landing modes. The F-35A is the CTOL version used by the USAF. The F-35B is the STOVL variant used by the USMC. It’s the one with the rotating engine nozzle and big fan behind the cockpit. It’s also the heaviest version. It’s estimated that the B variant’s STOVL system adds 4,000 pounds to the empty weight and it has 5,000 pounds less internal fuel than the other two variants. The F-35C is for the Navy. It has a 620 sq. foot wing which is 35% larger than either the A and B variants. It’s actually a bigger wing than the F-15 Eagle. In terms of empty weight, it’s slightly less than the B version. It has the largest fuel capacity with over 1100 pounds more than the A, giving it a hi-lo-hi combat radius of 800 nm with internal fuel and weapons. Unlike the A, it does not have a boom receptacle in the upper fuselage and instead has a retractable refueling probe (as does the B variant). That being said, the C is likely to be slower in acceleration at transonic speeds when compared to the A due to its larger wing (Sweetman, 2004, p. 107 – 108).
While it originated in the Common Affordable Lightweight Fighter program, the B and C versions have empty operating weights comparable to an F-15E. Their wing areas are larger than a F/A-18 Hornet but they’re significantly heavier. Given the weight of these aircraft and the fact that the engine delivers 28,000 pounds of thrust without using an afterburner, it reinforces the notion that the F-35 is a bomber (Sweetman, 2004, p. 108).
While the initial concept was to have the three versions be similar enough (around 70% commonality) to simplify construction, in reality, there’s only about 25% commonality in the parts between the different variants. This also means that it doesn’t benefit from economies of scale since it’s essentially 3 different aircraft (Barrett, 2017, para. 11).
One of the reasons for the sheer size of the F-35 is due to the internal weapons bays. Weapons carried on external pylons create a great deal of drag, but they still do less to impact a fighter’s size than internal weapons bays because the entire fighter must be scaled up to accommodate the weapons bays. Thus, the fighter has to carry around that extra weight and volume with it all the time. Another reason is that it’s difficult to add space for fuel in a stealthy aircraft because you can’t just change the shape of the plane to make more room for fuel without compromising its stealth qualities. The issue of fuel is pertinent because an engine running on an afterburner guzzles gas like there’s no tomorrow. Since changing the shape of stealth aircraft is very difficult, conformal fuel tanks are out and drop tanks are draggy. Although it does have the capability to carry drop tanks (Sweetman, 2004, p. 113).
When compared to the F-22, the 35’s weapon bays are deeper in order to accommodate a 2,000-pound GBU-31, whereas the 22 can only fit a 1,000-pound GBU-32 (Sweetman, 2004, p. 107 – 108).
A word on stealth. Bear in mind, I’m no expert and the true details are very hush, hush, super, ultra, top secret. If I told you, I’d have to kill you…sorta deal. Basically, there’s more to stealth than making an airplane have Low Observability (LO). You can deck a plane out with funny angles and RAM to give it a low RCS, but that’s just one part of the equation and not the end-all-be-all of stealth. A great deal of planning goes into the route that the aircraft will fly to avoid or minimize exposure to radars. Other considerations are the systems that detect, locate, and ID radar emitters and estimate their ability to detect the aircraft. The idea behind a stealth aircraft is similar to that of a submarine. It tends to operate alone, with a bare minimum of emissions and no non-stealthy aircraft in the area. That being said, it is still heavily supported by other assets. The USAF considers stealth aircraft survivable on the assumption of appropriate mission planning, force packaging, and tactics. USAF F-117 and B-2 flights over Yugoslavia during the Kosovo War were supported by EA-6B Prowlers providing jamming escort. One famous incident was in April of 1999 when an F-117 was shot down by an SA-3 over Serbia. A number of factors were probably to blame for this. Serbians were known to move their missile sites around to avoid detection, thus the SA-3 may not have been properly located. The Serbians may have also modified their targeting radars to improve their detection capabilities and were on the lookout for F-117s. Furthermore, the F-117 had flown repeated missions over the same course and the aircraft had just opened its bomb bay doors to launch a weapon, possibly creating a strong radar return. Additionally, support assets for this plane were not all in the right place at the right time. F-117 strikes during Desert Storm were also supported by jamming and simultaneous attacks by Tomahawk cruise missiles. All in all, even stealth aircraft rely on the element of surprise and the increasingly complex nature of air defense networks and capabilities of SAMs, LO aircraft are seen as essential for prosecuting some missions (Sweetman, 2004, p. 109 – 110).
One key feature contributing to the stealth of the F-35 is the design of the air intakes. They’re S-curved so as to conceal the fan blades which are a big contributor to RCS. That being said, radar waves entering the inlet can still bounce off the duct walls, hit the fan, and then bounce out. Therefore, the walls of the air intakes are coated in RAM which gradually attenuates the radar energy with each bounce until the escaping signal is so small it can’t be detected. Comparatively, the F-35’s inlet is simpler than the F-22’s (Sweetman, 2004, p. 115 – 116).
Much of the shape of the F-35, while having some rounded edges, is still composed of flat, canted sides and angles. This is a further development of the Have Blue project that developed into the F-117. Essentially, a flat surface illuminated by radar waves at right angles has a huge RCS. If it’s tilted at 30 degrees in one dimension, the reflectivity is reduced by a factor of 1,000. Additionally, if it’s rotated away from the beam on a diagonal axis, then the RCS is further reduced. Therefore, the same radar reduction can be achieved with an 8-degree angle if properly canted and swept back. Anyway, there’s a lot more to the concepts of stealth than just making an airplane out of funny angles, such as how the plane presents itself to an emitter. Even things like the edges of doors, apertures, and access panels can cause spikes in the radar signature. This is why many of them have that sawtooth-looking edge. The composition of the Radar Absorbent material itself is also a factor, but I won’t go into it further. As with many things, there are tradeoffs. One thing is that the F-35’s stealth characteristics were designed to last longer than previous aircraft such as the B-2 or F-22. The B-2 used dozens of different materials which had their own supply chain and maintenance schedules, some of which required specialized training just to handle, and could only be restored or replaced very slowly and laboriously. For example, the caulking around the edges of doors had to be applied in an air-conditioned hangar and only a few feet at a time, then left to cure. The F-22 uses about 1/3rd as many different types of LO materials. The F-35 was designed around the idea that maintenance could be performed through actuated doors and hinged doors rather than removable panels as on the F-22. The skin of the 35 is expected to last 8,000 hours under tactical conditions before needing to be repaired or replaced (Sweetman, 2004, p. 116 – 121). Anyway, as you can probably tell by now, there’s a ton of work that goes into making and keeping an aircraft stealthy.
Well, that does it for now. The next post will discuss the envisioned role of the F-35.
Barrett, P. (2017, April 4). Is the F-35 a Trillion-Dollar Mistake? Retrieved from https://www.bloomberg.com/news/features/2017-04-04/is-the-f-35-a-trillion-dollar-mistake.
Newdick, T. (2015). The World’s Greatest Military Aircraft: An Illustrated History. London, UK. Amber Books Ltd.
Sweetman, B. (2004). Ultimate Fighter: Lockheed Martin F-35 Joint Strike Fighter. St. Paul, MN: MBI Publishing.