F-104 Starfighters for Belgium

Much has been written about the operational use of the Lockheed F-104 Starfighter jet fighter with West Germany, Italy, and Japan, bearing in mind the fact that the F-104s operated by the Luftwaffe (the air force of West Germany and ultimately the reunified German nation) were given nicknames like "Flying Coffin" and "Widowmaker" because they suffered a high accident rate. However, almost lost in talk regarding the F-104's operational career with Western and Central European air forces is the service career of the F-104 with the Belgian Air Force.

A Belgian Air Force F-104G Starfighter (serial number FX-82) on outdoor static display at the Planes of Fame Museum, photographed by me on January 21, 2024.

In the late 1950s, the Belgian Air Force was shopping for a new combat jet to replace the subsonic CF-100 Canuck interceptors and F-84F Thunderstreak fighter-bombers in its inventory, especially given that both the Canuck and Thunderstreak were becoming technologically obsolete. With the Luftwaffe as well as the post-World War II Italian air force (Aeronautica Militare) and Royal Netherlands Air Force purchasing the F-104, the Belgian Air Force decided to order the F-104G, and contracts were signed for the purchase of 112 F-104Gs, with twelve TF-104Gs ordered from Lockheed and 100 F-104Gs to be built under license by the homegrown Belgian aircraft manufacturer SABCA. The TF-104Gs were given the serial numbers FC-01/FC-12 while the SABCA-built F-104Gs for bore the serial numbers FX-1 to FX-100; twenty-eight of the F-104Gs ordered by Belgium (three TF-104Gs, 25 F-104Gs) were funded under the Mutual Assistance Program (MAP). Deliveries of the F-104G to the Belgian Air Force began in February 1963, with the F-104Gs and TF-104Gs replacing the F-84F and CF-100 in service with the 23 and 31 squadrons (both of 10 Wing) at Kleine Brogel and the 349 and 350 squadrons (both of 1 Wing) at Beauvechain (Bevekom) respectively. One F-104G ordered by the Belgian Air Force (serial number FX-27) crashed during a training flight at Sart-Dames-Avelines on November 21, 1963 prior to delivery due to a flameout of the turbojet engine, and a new F-104G (c/n 9082) also bearing the serial number FX-27 was built and delivered to replace the crashed aircraft, so a total of 111 F-104Gs and 12 TF-104Gs were built for the Belgian Air Force.

Three Belgian Air Force F-104Gs (serial numbers FX-12, FX-30, FX-82) in flight, 1971. 

Operational use of the F-104G by the Belgian Air Force in its role as an interceptor began in August 1964. Although the 10 Wing and 1 Wing were exclusively tailored to the air defense and tactical nuclear strike roles respectively, the F-104Gs were swapped between the two operational wings to balance airframe fatigue between medium/high-altitude fighter missions and nuclear-armed fighter-bomber missions. In its role as a fighter-bomber for the Belgian Air Force, the F-104G was armed with B61 tactical nuclear free-fall bombs stored at facilities owned by the 52nd Special Ammunition Group at Meeuwen. By 1968, the F-104Gs of the 10 Wing switched to dual air defense/fighter-bomber missions and began training with conventional weapons such as the 20-mm Vulcan rotary cannon, three napalm bombs or two Snakeye bombs or two LAU rocket launcher pods each armed with nineteen 2.75-inch FFAR rockets. Beginning in late 1979, the Belgian Air Force began replacing the F-104G with F-16 Fighting Falcons (the TF-104Gs were phased out in 1980), and the last Belgian Air Force F-104Gs were retired from service on September 26, 1983. Besides the F-104G with serial number FX-27 that crashed in November 1963, a total of 38 F-104Gs and three TF-104Gs were lost in accidents during operational service. In the meantime, 18 Belgian Air Force F-104Gs were transferred to the Turkish Air Force and 23 were given to Taiwan. 

Nuclear-armed stinger from Hawthorne: the Northrop F-89J

In 1955, the Douglas Aircraft Company began full-scale development of a short-range unguided air-to-air rocket to carry a 1.5 kiloton W25 nuclear warhead, the MB-1 Genie, after it became clear that traditional World War II-era US fighter armament would be inadequate to repel a bombing attack by squadrons of the Soviet Union's new gas turbine powered strategic bombers, the Myasishchev M-4, Tupolev Tu-16, and Tupolev Tu-95. The MB-1 Genie obviated the need for precise accuracy when targeting enemy bombers because it was designed with a large nuclear blast radius. Beginning in March 1956, Northrop modified numerous F-89D Scorpion all-weather interceptors from the F-89D-35 to -75 production blocks to carry the MB-1 Genie under Project Bellboy, and the company designation N-160 was allocated to this scheme. The resulting Genie-armed Scorpion, designated F-89J, carried two MB-1 Genies below launching rails that were mounted on the underwing pylons and had the standard wingtip missile pod/tanks replaced with 600 gallon (2,271 liter) fuel tanks, although a few F-89Js retained the wingtip tanks of the F-89D. Later, the F-89J received an extra modification by adding two more underwing pylons inboard of the launching rails for the Genie to carry four Falcon air-to-air missiles tipped with non-nuclear warheads. The F-89J was equipped with the Hughes MG-12 fire-control system (a upgraded and more advanced development of the E-5 fire control system installed on the F-89D), which could allow it to attack enemy bombers at much higher altitudes by making it easier for the crew to launch the Genie rockets while in a nose-up, climbing altitude. During interception of an enemy bomber formation, the MG-12 fire-control radar tracked a target and assigned a Genie to its target, after which the pilot armed the nuclear warhead and fired the Genie at the bomber pack before pulling the interceptor into a tight turn to escape the nuclear detonation and then proceeding to use remote control to allow the Genie's nuclear warhead to explode and destroy enemy bombers.

Left: Two F-89Js (serial numbers 52-1848 and 52-1862) in flight, 1958
Right: An F-89J (serial number 52-1949) at the March Field Air Museum, photographed by me on December 17, 2022.

In November 1956, the US Air Force began taking deliveries of the F-89J, the 84th Fighter Interceptor Squadron based at Hamilton AFB in Novato, California being the first unit to receive the F-89J, and standing active alerts of the F-89J with the Genie started on January 1, 1957. A total of 350 F-89Ds were converted to F-89J standard, with modifications completed by February 1958, and the US Air Force assigned the system code WS-205G (Weapons System 205G) to the F-89J. On July 19, 1957, as part of Operation Plumbbob, the F-89J carried out the first and only live firing of a Genie (codenamed John) when an F-89J with serial number 53-2547 fired an MB-1 Genie over the Yucca Flats Nuclear Test Site in southern Nevada, with the rocket's warhead detonating at an altitude of 15,000 feet (4,500 meters). To prove that the Genie was safe for use over populated areas in the event that Soviet bombers would penetrate US airspace, a group of five Air Force officers volunteered to stand uncovered in their light summer uniforms underneath the blast, and they were apparently spared from the effects of the blast after the live-firing test of the Genie over the Yucca Flats.

Despite proving to the US Air Force that a live-firing of a nuclear-armed unguided air-to-air rocket was feasible, the F-89J was destined to have a brief operational career with the Air Defense Command, and beginning in July 1959 it was replaced in ADC units by the supersonic F-101B Voodoo and F-106 Delta Dart, which also carried the Genie air-to-air rocket. The F-89Js were then transferred to the Air National Guard, operating with ANG until late 1968, when they were retired. In an interesting footnote, in 1963 ten F-89Js were stripped of their nose radars, fitted with additional underwing fuel tanks, and eventually used for testing Nike missile defenses in Japan, being redesignated DF-89J.  

References:

Balzer, G., and Dario, M., 1993. Northrop F-89 Scorpion. Leicester, UK: Aerofax.

Chong, T., 2016. Flying Wings & Radical Things: Northrop's Secret Aerospace Projects & Concepts 1939-1994. Forest Lake, MN: Specialty Press.

Davis, L., and Menard, D., 1990. F-89 Scorpion in Action (Aircraft Number 104). Carrollton, TX: Squadron/Signal Publications.

The miniature bee from San Diego: the Beecraft Wee Bee

Everyone, myself included, has been fascinated with gigantic aircraft from the annals of heavier-than-air powered flight, like the Hughes H-4 Hercules, Antonov An-124 and An-225, Convair B-36 Peacemaker and XC-99, Sikorsky Il'ya Muromets, Tupolev ANT-20, and the Riesenflugzeugen-type bombers built in Germany in World War I. However, very small airplanes have tended to fly under the radar, although many miniature heavier-than-air flying machines existed in the earliest days of homebuilt airplanes, and most people don't know that once upon a time in the late 1940s, the aircraft industry in San Diego built the smallest ultralight plane anywhere in southern California, the Wee Bee. Hence, I am dedicating this post to discussing the oft-neglected story of the smallest aircraft manufactured in San Diego.

 

Convair engineer William "Bill" Chana (1921-2012), who worked on design of the Wee Bee and became one of aircraft's test pilots.

During the late 1940s, the US aviation industry was intoxicated by the nearly-monthly trend of new and improved airplanes appearing in the United States, so in 1947 Convair engineer Kenneth Coward toyed with the notion of a balsa wood airplane having an empty weight of 70 pounds. Despite being the chief designer for his proposed miniature airplane, Coward also enlisted five engineers from Convair, including William "Bill" Chana, Karl Montijo, James Wilder, Tom Bossart, and A.B. Mandeville, to explore the possibility of building an extremely small airplane. Chana himself took a leading role in working out the layout of Coward's proposed aircraft, having built award-winning model planes since he took an interest in aviation at age 6 and studied aeronautical engineering at Purdue University before joining Consolidated Aircraft in June 1941 to become a flight test engineer for a variety of aircraft built by the San Diego division of Consolidated (which became Convair after 1943). Coward, Chana, and other members of the group decided that balsa probably would not be an appropriate building material for use in construction of their proposed aircraft because war-surplus aluminum was available in sufficient quantity and the tooling equipment for fashioning could be readily obtainable. 

The aircraft design by Coward and Chana that soon emerged, dubbed the Wee Bee, was an all-metal monoplane whereby the pilot lay prone atop a girder-like fuselage to allow for a reduction in weight and drag and which had the elevator control beneath the pilot despite being in the same spatial position as in a conventional airplane as well as rudder pedals housed in slots on top of the fuselage. The Wee Bee had a length of 14 feet 2 inches (4.32 meters), a height of 5 feet (1.52 meters), an empty weight of 210 lb (95 kg), and a maximum takeoff weight of 410 lb (186 kg). The Wee Bee's designers formed a new company to undertake manufacture of the aircraft, which was initially called Ken S. Coward & Associates and then was known for a while as Beecraft Aviation Associates before finally changing its name to Bee Aviation Associates, and Bill Chana became president of this new firm.

Left: The Beecraft Wee Bee taking off on its second flight in November 20, 1948.
Right: The Beecraft Wee Bee parked next to the Convair XC-99 prototype heavy-lift transport near the Convair factory, late 1948. The huge size disparity between these two piston-powered planes is evident.

The Wee Bee aircraft was completed in early 1948, eventually receiving the civil registration NX90840. After receiving a few modifications, including replacing the original 15 foot (4.57 meter) wing with a slightly bigger 18 foot (5.49 meter) wing that had a wing area of 44 square feet (4.1 square meters), it began taxi tests at El Cajon's Gillespie Field in August, but the tail skid originally incorporated onto this aircraft carved up with asphalt runways of the airfield, and even though a tail wheel was substituted, it created a ground-looping tendency and the 20 hp (15 kW) Kiekhaefer O-45-1 flat-twin piston engine installed on the Wee Bee did not generate enough power. Therefore, the Wee Bee was fitted with a tricycle landing gear, which not only cured the ground-looping tendency but also allowed the nose to be lifted off the ground at 30 mph (48 km/h). On September 26, 1948, the Wee Bee carried out its first flight with Bill Dana himself at the controls, flying at an altitude of one foot at a distance of less than 100 feet. After a few additional flights by Karl Montijo and William Bouck, which were also hops in ground effect, the Wee Bee was fitted with a more powerful 30 hp (22 kW) O-45-35 engine by a US Navy officer at a hangar in Ream Field at Chana's behest, and on March 12, 1949, Montijo carried out the first "true" flight of the Wee Bee, reaching an altitude of 40 feet (12 meters) and banking to the right in level flight. He  further reduced the weight and drag of the Wee Bee by changing the aircraft's wing incidence to 19 degrees and installed fairings on the wing junctures to the fuselage and engine to fuel tanks while fitting the Wee Bee with a shorter nose gear strut and a new Sensenich propeller. On April 20, 1949, the Wee Bee reached an altitude of 400 feet and flew for 10 minutes on its next flight, and after being fitted with wingtips, it was ferried to the UK via the Queen Mary for a demonstration flight at Gatwick Airport in London. A planned demonstration flight at Belfast, Northern Ireland, was shelved due to bad weather, but the Wee Bee caught the attention of the print media as the smallest plane ever flown when it impressed crowds at the 1949 National Air Races in Cleveland, Ohio, with a flight that lasted seven minutes. Kenneth Coward bolstered his messaging that “the Wee Bee was big enough to lift a man and small enough to be lifted by a man” by wrapping his arm around the airplane and lifting it off the ground. Bill Chana again changed the aircraft's wing incidence again to 12 degrees in order to further enhance performance, and with this modification, the Wee Bee attained a top speed of 82 mph (132 km/h) during a series of flights by Chana himself on March 11, 1950. The thirteenth and last full-fledged flight of the Wee Bee occurred on  April 20, 1950, by which time newsreel companies were utterly impressed by the performance of the Wee Bee.

A replica of the Beecraft Wee Bee on display at the San Diego Air and Space Museum in Balboa Park, San Diego, photographed by me on August 24, 2019.

Even before the last flight of the Wee Bee, Beecraft Aviation Associates proposed a military version of the Wee Bee for the US Air Force, the Military Mite, which would have been armed with six underwing rockets and featured provisions for folding wings to allow for easy ground transportation, while being capable of takeoff and landing from any road. However, the USAF had no interest whatsoever in this proposal. After being retired from flying, the Wee Bee became a static exhibit at US airshows before becoming an exhibit at the San Diego Air and Space Museum in San Diego in 1963. On February 22, 1978, the Wee Bee was destroyed when the original museum building was set on fire by an arsonist, even while plans were underway to move the SDASM and its aircraft collection to the Ford Building in Balboa Park. Despite the loss of the original aircraft, a replica of the Wee Bee was constructed and is now on display at the current home of the San Diego Air and Space Museum in Balboa Park.

Competing designs to the T-45 Goshawk from southern California

The T-45 Goshawk is the most modern advanced jet trainer aircraft in service with the US Navy, jointly built by McDonnell Douglas (which was acquired by Boeing in 1997) and British Aerospace (renamed BAE Systems in 1999) as a derivative of the UK's British Aerospace (Hawker Siddeley) Hawk land-based jet trainer. However, I should emphasize that the T-45 Goshawk did not exist in its own right despite the fact that it owes its heritage to the Hawk jet trainer. During my first visit to the Western Museum of Flight at its current location in Torrance, I happened to notice a wind tunnel of an aircraft on a display stand across from an F-86 Sabre jet fighter, and when I saw a label on the stand identifying the wind tunnel as being of an early 1980s Northrop jet trainer design for the US Navy, I realized than more than one company undertook design studies for a carrier-based advanced jet trainer to replace the Navy's fleet of T-2 Buckeyes. Given that I have in my personal possession a copy of Tony Chong's book Flying Wings & Radical Things: Northrop's Secret Aerospace Projects & Concepts 1939-1994 and thanks to the Secret Projects Forum, it is now possible for me to discuss advanced jet trainer designs from southern California's aircraft industry that competed with T-45 Goshawk.

Top left: Lockheed derivative of the Alpha Jet for the VTX-TS competition
Top right: Artist's impression of the Rockwell International T-2X project
Bottom: Company artwork of the proposed Rockwell International NA-424

In 1978, the US Navy launched VTX-TS requirement for an advanced jet trainer to replace the North American T-2 Buckeye and Douglas TA-4 Skyhawk. In addition to the joint McDonnell Douglas/British Aerospace team, Rockwell International submitted bids for the VTX-TS competition, as did a joint team between Northrop and Vought as well as Lockheed. The Lockheed submission was a navalized derivative of the Dassault/Dornier Alpha Jet jet trainer and light attack aircraft, and it featured nose landing gear modified for nose-tow catapults, a slightly longer nose, reinforced legs of the landing gear, and a stronger arrestor hook. The Lockheed trainer derivative of the Alpha Jet was to be powered by a pair of Teledyne CAE 490 turbofans (a proposed version of the SNECMA Turbomeca Larzac turbofan to be built under license by Teledyne Turbine Engines) on the sides of the fuselage below the wings. In September 1980, a French Air Force Alpha Jet serialled A58) was flown to the US and made a total of 88 demonstration flights at US Navy air bases in Florida, Maryland, Mississippi, and Texas, and 67 Navy pilots carried out those flights. Two Rockwell International designs for the VTX-TS requirement were proposed in 1980, the NA-424 and T-2X. The T-2X was an evolutionary derivative of the T-2 Buckeye which retained the straight wings of the T-2 but differed in having two turbofans with square-shaped air intakes situated in the wing roots, while the NA-424 was to have a new airframe featuring backswept wings, a slightly pointed nose, a dorsal fin protruding from the base of the vertical stabilizer and reaching an area behind the cockpit canopy, and two turbofans on the sides of the fuselage above the wing roots. Two variations of the NA-424 were studied by Rockwell International, one with a low-mounted horizontal stabilizer (in contrast to the T-2 Buckeye's mid-mounted horizontal stabilizer that gave the tail empennage of the Buckeye a cruciform appearance in front view) and another having a T-tail configuration.

Top: A wind tunnel model of the N-351 design at the Western Museum of Flight, photographed by me in June 2016
Bottom left: A three-view drawing of the N-350 from Northrop project documents
Bottom right: An artist's rendering of the N-351 jet trainer preparing for a landing on an aircraft carrier

Now all of this brings my attention to Northrop's bidding for the VTX-TS competition. When the US Navy in 1975 began discussions with aircraft manufacturers about possible replacements for the T-2 Buckeye and TA-4J, Northrop envisaged the N-328 subsonic advanced jet trainer with tandem seating, straight wings, and two turbofan engines (possibly Garrett TFE731s) on the sides of the fuselage and mounted above the wing roots. The N-328 would have been the Navy equivalent of Northrop's N-325 advanced jet trainer proposal for the US Air Force designed in 1974, and Northrop also conceived the N-329 for use by both the Air Force and Navy. The N-328 and N-329 remained design studies only, but in late 1977 Northrop undertook design work on the N-334 project, which was powered by two TFE731 turbofans on the sides of the fuselage above the wing roots and drew its design heritage from the CASA C-101 and AIDC AT-3 jet trainers (single-engine and twin-podded engine configurations were also studied by Northrop for the N-334). The N-334 was submitted to the Navy Air Development Center in early 1978, and following further study under contract from the NADC, Northrop in 1980 submitted the final N-334 design for the VTX-TS requirement. After being selected along with five other companies by the Naval Air Systems Command (NAVAIR) on August 19, 1980, for the next phase of the VTX-TS competition, it teamed up with Chance Vought (which had worked on the V-532B advanced jet trainer project for the US Navy in the late 1970s) to conceive the N-350 advanced jet trainer, for which initial drawings had been devised a few months earlier. The N-350 was 38 feet 8.4 in (11.8 meters) long with a wingspan of 32 feet 4.8 in (9.88 meters), and power was provided by either two 3,500 lb (15.57 kN) thrust Garrett TFE731 turbofans or two 3,300 lb (14.68 kN) thrust Pratt & Whitney JT15D turbofans. By December 1980, the Northrop/Vought team envisaged its final design for the VTX-TS contest under the company designation N-351, which differed from the N-350 in having wings with greater leading edge sweep and a more pointed nose, and which measured 40 feet (12.2 meters) long with a wingspan of 29 feet 7.3 in (9.02 meters).   

The McDonnell Douglas/British Aerospace T-45 Goshawk, which was selected over the NA-424 and N-351 as the winning design for the VTX-TS competition.

On November 19, 1981, the US Navy announced that the proposal by the McDonnell Douglas/British Aerospace for a derivative of the British Aerospace Hawk had been declared the winner of the VTX-TS competition. The winning McDonnell Douglas/British Aerospace design became the T-45 Goshawk and made its first flight on April 16, 1988.

References:

Chong, T., 2016. Flying Wings & Radical Things: Northrop's Secret Aerospace Projects & Concepts 1939-1994. Forest Lake, MN: Specialty Press.

https://www.secretprojects.co.uk/threads/u-s-navy-vtxts-advanced-jet-trainer-proposals.17230/

https://www.secretprojects.co.uk/threads/northrop-vought-vtxts-jet-trainer-projects-and-derivatives.15677/

https://www.secretprojects.co.uk/threads/lockheed-dornier-alpha-jet-vtxts-advanced-trainer-proposal.12926/

https://www.secretprojects.co.uk/threads/north-american-rockwell-vtxts-proposal-t-2x.17233/

Jayhawk's progeny from San Diego: the AQM-81 Firebolt

In the mid-1960s, the US Air Force launched the Sandpiper program to modify some Beechcraft AQM-37 supersonic target drones with hybrid fuel rocket motor system using a solid grain fuel and liquid oxygen because the liquid-fuel rocket motor used to power the AQM-37 used dangerous hypergolic propellants. In 1968-1969, flight tests of the Sandpiper were conducted and they were so successful that the USAF in the early 1970s initiated the HAST (High Altitude Supersonic Target) program, later renamed HAHST (High Altitude High Speed Target), for a production target drone based on the Sandpiper configuration, and the designation XAQM-81A was assigned to the HAST/HAHST program. However, technical difficulties meant that a definite configuration for the HAST/HAHST program was not settled upon until the late 1970s. By this team, the USAF deemed Beech's offer for a full-scale development contract too expensive, and thus called for competitive bids for HAHST development from the aerospace industry. Among the submissions for the HAST/HAHST competition was a proposal from Teledyne Ryan, the Model 305 Firebolt, and in December 1979, the the XAQM-81A development contract was awarded to Teledyne Ryan, with an order placed for nine XAQM-81As.

An AQM-81A Firebolt on display at the March Field Air Museum in Riverside, California, photographed by me on December 17, 2022

The XAQM-81A Firebolt had the same layout and airframe as the AQM-37, particularly the slender delta wings and triangular vertical fins at the wingtips, but it mainly differed in having a single 1,200 lb (5.3 kN) hybrid-fuel rocket motor built by Chemical Systems Division (CSD) of United Technology which used IRFNA (Inhibited Red Fuming Nitric Acid) as the oxidizer for the solid grain fuel. The thrust of the rocket motor could be adjusted in flight to alter speed and altitude to give interceptor pilots a more flexible and realistic target. The air inlet below the fuselage of the Firebolt utilized a ram air turbine that pressurized the IRFNA oxidizer before delivering it to the thrust chamber of the rocket engine, and it also provided electrical power for the drone. The XAQM-81A was 17 feet (5.18 meters) long with a wingspan of 3 feet 4 inches (1.02 meters), a fuselage diameter of 13 inches (33 cm), and a weight of 1,230 lb (560 kg), and it had a top speed of Mach 4.3 and an endurance of five minutes. The F-4 Phantom II served as the launch platform for the AQM-81A, and after being launched from an F-4 at a speed of Mach 1.5, the Firebolt would ignite its rocket rocket to reach an altitude of 103,000 feet (31,400 meters) at speeds exceeding Mach 4. The XAQM-81A Firebolt could fly a pre-programmed course and/or respond to guidance commands from the ground, and it was equipped with a parachute recovery system to enable either a soft landing or a mid-air retrieval by a helicopter, the latter via the Mid-Air Retrieval System (MARS). The US Navy version of the Firebolt, the AQM-81B (erroneously listed in a few sources as "AQM-81N"), had the same mode of launch and performance characteristics as the planned production AQM-81A but differed in incorporating the Navy's AN/USW-3(V) ITCS (Integrated Tracking and Control System), radar augmentation for ground tracking requirements, and  flotation gear for recovery over water.

The XAQM-81A began flight tests on June 13, 1983, at Eglin AFB in Florida. Even before the Firebolt began its flight test program, the number of Firebolt test vehicles had been increased to 21 drones, including six each of the AQM-81A and AQM-81B. In July 1984, test flights of the AQM-81B began at Point Mugu NAS in southern California, by which time the Air Force had began service tests of the Firebolt. Flight tests of the AQM-81 continued until the fall of 1984, by which time more than twenty flights had been conducted. Although the AQM-81 exhibited outstanding performance during its flight test program, it was not ordered into production because the Air Force and Navy realized that the unit cost of the AQM-81 was far more expensive than the simpler AQM-37. 

References:

Munson, K., 1988. Jane's World Unmanned Aircraft. Coulsdon, UK: Jane's Information Group.

Wagner, W., and Sloan, W.P., 1992. Fireflies and other UAVs (Unmanned Aerial Vehicles). Arlington, TX: Aerofax.

Sabres for the US Navy: the FJ-2, FJ-3, and FJ-4

When I made my first visit to the Yanks Air Museum back in July 2016, it was a blessing for me that the museum has on display the only surviving example of the first jet fighter for the US Navy built anywhere in southern California, the North American FJ-1 Fury, whose straight wings happened to be the original wing design for the F-86 Sabre before captured wartime German aeronautical research prompted North American to redesign the Sabre with backswept wings. On my visits to the Planes of Fame Museum, I got to see in person for the first time one of a handful of variants of the F-86 Sabre for the US Navy, the FJ-3, and as has been noted before, the FJ-2, FJ-3, and FJ-4 were swept wing aircraft unlike the FJ-1, which is quite odd because the FJ-2/3/4 should have been designated F2J. Even though the FJ-2, FJ-3, and FJ-4 were manufactured by North American in Ohio rather than California, I am including the navalized F-86 Sabre variants for convenience on this blog as they were derived from the Sabre. 

Two FJ-2s of Marine Corps squadron VMF-235 in flight in 1954

In late 1950, the US Navy's first generation of jet fighters were recognized as inferior in performance to the Mikoyan-Gurevich MiG-15 six months into the Korean War, and although the Vought F7U Cutlass was the first swept wing jet fighter for the Navy designed from scratch, it was not yet operational. Thus, on January 30, 1951, North American Aviation envisaged a navalized variant of the F-86E Sabre, the NA-181, which had the slatted wing of initial production F-86Fs but differed in having folding wings, a modified cockpit canopy, catapult attachment points and arrester gear, a General Electric J47-GE-2 (a navalized version of the F-86F's J47-GE-27), the wheel track widened by 8 inches, and strengthened landing gear, and armament comprised four 20 mm cannons with 600 rounds. The NA-181 proposal was submitted to the US Navy on February 6, and the Navy expressed considerable interest, signing a contract on February 10, for 300 production NA-181s (BuNos 131927/132126) to be built at a newly opened North American factory in Columbus, Ohio. Although the NA-181 was a navalized F-86E and thus different from the FJ-1, the Navy instead designated it FJ-2, perhaps hoping that Congress would approve of an aircraft which was a "logical extension" of an existing type. On March 8, the US Navy ordered three XFJ-2 prototypes (BuNos 133754/133756), which were to be built in Inglewood due to the Columbus plant not yet being ready for manufacturing operations. The first two XFJ-2s (company designation NA-179) differed from the F-86E-10 in having a V-frame arrester hook, catapult points, and a longer nosewheel to raise the angle of attack during takeoff and landing. The third XFJ-2 (BuNo 133756), designated NA-185 by North American, retained the F-86E's nosewheel and lacked naval equipment, and it was armed with four Colt Mk 12 20 mm cannons with 150 rounds and designated XFJ-2B (B stood for special armament). The XFJ-2B was actually the first of the three XFJ-2s to fly, making its first flight on December 27, 1951, before being flown to Inyokern, California, for armament tests, and the first XFJ-2 prototype took to the skies on February 14, 1952. The three prototypes were accepted by the Navy from June to December 1952, and the XFJ-2s underwent carrier qualification tests aboard the USS Midway and USS Coral Sea in the second of half of 1952, but problems surfaced during those tests, such as the new landing gear and arresting hook bumper being too weak for carrier landings, and poor handling of the aircraft during carrier approaches and landings. Nevertheless, the Navy had chosen to go ahead with the start of FJ-2 production, and the first production FJ-2 was flown on November 22, 1952 and delivered to the Navy on December 12. Whereas the horizontal stabilizer of the FJ-2 prototypes had dihedral, the production FJ-2 had a flat horizontal tail. After the Korean War, production orders for the FJ-2 were reduced to 200 aircraft, and deliveries of the FJ-2 to the Navy and US Marine Corps were completed by September 1954. By this time, the Navy deemed the Grumman F9F-6 Cougar to be better at operations from carrier flight decks despite its slightly slower cruising speed given the FJ-2's increased weight compared to the F-86F, and thus only a few FJ-2s saw Navy service, even though the FJ-2 entered service with US Marine Corps units in January 1954. In USMC service, FJ-2s operated with squadrons VMF-122, VMF-232, and VMF-312 of the Atlantic Fleet Marines, and squadrons VMF-235, VMF-224, and VMF-451 of the Pacific Fleet Marines, and even though most of their operations were carried out from land bases, the FJ-2 Furies went to sea aboard the USS Coral Sea and a few other carriers. Even so, the FJ-2's carrier handling characteristics weren't really satisfactory, and it was retired from frontline service in 1956 and phased out by reserve units in 1957. 

Left: An FJ-3 Fury (BuNo 135867) at the Planes of Fame Museum, photographed by me on April 13, 2019. 
Right: Four FJ-3s of US Navy squadron VF-21 in flight in the late 1950s.

Shortly after flight tests of the XFJ-2 and XFJ-2B began, in March 1952 North American undertook design of a new Fury variant, the NA-194, to be powered by one Wright J65 turbojet (the American license-built version of the Armstrong Siddeley Sapphire), which had greater thrust that the J47-GE-2 that powered the FJ-2. Viewing the NA-194 as having potentially enhanced performance, on April 18 the Navy placed an order for 289 aircraft (BuNos 135774/136162) to be built in Columbus and the designation FJ-3 was assigned to the NA-194. The fifth production FJ-2 (BuNo 131931) was chosen to serve as a testbed for the FJ-3, being fitted with a J65 and assigned the designation NA-196 by North American, and it first flew in this iteration on July 3, 1953. Despite retaining the FJ-2's slatted wings and hydraulic power-operated horizontal tail and ailerons, the FJ-3 featured a larger nose intake to encapsulate the J65's increased thrust, and cockpit armor including a 52-pound back plate and an 88-pound plate in front of the instrument panel. The FJ-3 made its first flight on December 11, 1953, and it reached operational deployment with US Navy units beginning in September 1954, with the first carrier landings aboard the USS Bennington on May 8, 1955. Even before deliveries of the FJ-3 to operational units had begun, the Navy felt so impressed with the performance of the FJ-3 during flight testing that on March 11, 1954 it ordered 169 more FJ-3s (BuNos 139210/139278, 139324/139423), designated NA-215 by North American; the second FJ-3 batch ordered on March 11 with BuNos 139324/139423 were cancelled. In addition, North American developed a missile-armed FJ-3 variant, the FJ-3M, which was armed with four underwing pylons for external stores, the inboard weapons pylons being capable of carrying 500-pound bombs or rocket packs, and the outboard stations fitted with either 1,000-pound bombs or launching rails for AAM-N-7 (later AIM-9) Sidewinder air-to-air missiles, On November 2, 1954, eighty FJ-3Ms (BuNos 141364/141443) were ordered, and 105 FJ-3s were modified to FJ-3M configuration as well, with operational deployment of the FJ-3M commencing in 1956. Deliveries of the FJ-3 and FL-3M continued until 1956, by which time a total of 538 FJ-3s had been completed. A few FJ-3s were converted in 1957-1960 to drone control aircraft, with FJ-3s modified to carry the Vought KDU-1 (a target drone conversion of the Regulus submarine-launched cruise missile) being designated FJ-3D and those used for handling control of F9F-6Ks and Ryan KDA-1 Firebee drones receiving the designation FJ-3D2. The Wright J65 had some severe lubrication problems that could cause it to seize up and lose all power during a catapult launch, forcing the aircraft to drop into the ocean, and it also suffered from occasional catastrophic turbine blade failures, which would cause the engine to shed its turbine blades and send them flying out the sides of the fuselage. Despite these mishaps, the FJ-3 was fairly popular with its pilots, with Captain James Powell of the Navy fighter squadron VF-142 describing the FJ-3 as taking off faster than the F9F-6 Cougar. The FJ-3 Fury never saw combat, although it carried out combat air patrols during the 1958 Lebanon crisis.

Left: The first FJ-4 prototype (BuNo 139279) in front of the North American plant in Columbus, Ohio before its first flight, October 1954.
Right: Four FJ-4B Furies of US Navy attack squadron VA-63 in flight, late 1958.

The last Fury variant to be manufactured, the FJ-4, was envisaged by North American in June 1953 as an all-weather interceptor with 50 percent more fuel capacity than the FJ-3, necessitating an extra fuel tank below the engine and a dorsal spine stretching from the rear of the cockpit all the way to the tail empennage. The FJ-4 also differed from the FJ-2 and FJ-3 in having a new, thin wing of slightly greater span and wing area with a six percent thickness-to-chord ratio and skin panels milled from solid alloy plates, tapering more sharply towards the tips, but also thinner horizontal stabilizers that lacked dihedral and had a slightly shorter span, and a taller vertical stabilizer. The new "thin" wing of the FJ-4 required the main landing gear design to be considerably modified to fold the wheel and strut within the wing's contours, and the track of the main wheels was increased and because they were closer to the center of gravity, there was less weight on the nosewheel, while wing folding was limited to the outer wing panels. On October 16, 1953, the US Navy ordered two FJ-4 prototypes (BuNos 139279/139280) and 107 production aircraft (BuNos 139281/139323, 139424/139530), known internally by North American as NA-208 and NA-209 respectively. The first FJ-4 prototype was flown on October 29, 1954, and deliveries of the FJ-4 to US Navy and US Marine Corps units began in February 1955 and continued until March 1956. Armament of the FJ-4 consisted of four Colt Mk 12 20 mm cannons with reduced ammunition to create space for extra armor on the nose as well as four AAM-N-7 Sidewinder air-to-air missiles, and in its secondary role as a fighter-bomber the FJ-4 could carry 3,000 lb (1,360 kg) of bombs and six LAU-3/A pods containing 70 mm rockets while using its cannons for ground strafing. North American also developed a dedicated fighter-bomber version of the FJ-4, the FJ-4B (company designations NA-220, NA-229, and NA-244), which had a stronger wing with six underwing stations for carrying 6,000 lb (2,721 kg) of missiles, bombs, and rockets, reinforced landing gear, and extra aerodynamic brakes under the aft fuselage which made landing safer by allowing pilots to use higher thrust settings and were also useful for dive attacks. The FJ-4B was capable of carrying the ASM-N-7 (later AGM-12) Bullpup air-to-surface missile, which was usually fired from a shallow dive with the pilot putting his gunsight pipper on the intended target and guided to the enemy target by the pilot operating a miniature control stick and sending radio control signals to the missile to move the fins on the rocket, but it could also carry a nuclear weapon on the inboard port station, and it was equipped with the LABS (Low-Altitude Bombing System) for delivery of nuclear weapons. An order was placed for 25 FJ-4Bs (BuNos 139531/139555) on July 26, 1954, followed by a contract signed on November 2 for 46 more FJ-4Bs (BuNos 141444/141489), and the first FJ-4B made its first flight on December 4, 1956, with operational deployment in 1957. An additional 184 FJ-4Bs (BuNos 143493/143676) were ordered on April 5, 1956, but this was cut back to 151 aircraft, and a total of 222 FJ-4Bs had been built before deliveries were completed in May 1958. In all, 374 FJ-4s were built, but like the FJ-2 and FJ-3, the FJ-4 would never see combat. When the Defense Department introduced the Tri-Service aircraft designation system on September 18, 1962, the FJ-3, FJ-3M, FJ-3D, FJ-3D2, FJ-4, and FJ-4B were redesignated F-1C, DF-1C, DF-1D, F-1E, and AF-1E respectively. By this time, the FJ-3 and FJ-4 had been phased out of service with Navy and Marine Corps combat units, and they were transferred to Navy reserve units where they served until the mid-1960s, when they were retired. One retired FJ-4B with BuNo 143575 was later acquired by Flight Systems International in 1971 with the civil registration N400FS and operated on various military contract duties until 1982, when it was withdrawn from use, and it was eventually purchased by T-Bird Aviation in 1991 and given the new civil registration N9255. After years of restoration, in 2002, this aircraft reverted to its initial civil registration and once again became airworthy for a second time, and it remains the only FJ Fury still flying.

Left: One of two FJ-4Fs (BuNo 139284) in flight, early 1958
Right: A desktop model of the NA-295 ("AF-1E") light attack aircraft that lost the VAL competition to the A-7 Corsair II.

Before concluding this post, mention must be made of FJ-4 developments that either made it to the experimental stage or remained unbuilt. In November 1954, shortly after the FJ-4's first flight, North American Aviation explored the possibility of a rocket-augmented FJ-4 in response to US alarm at the USSR's deployment of the Myasishchev M-4/3M and Tupolev Tu-16 jet bombers and Tupolev Tu-95 turboprop bomber. The proposed rocket-augmented FJ-4 would have a search radar housed in a fairing above the nose and provisions for two Sidewinder air-to-air missiles or fifty 2 inch Gimlet unguided rockets (or alternately 168 1.5 inch NAKA unguided rockets) housed in internal weapons bay and four Sidewinder or Sparrow II missiles mounted below the underwing pylons. This initial proposal evolved into the NA-234 design study of August 1955 with a monopropellant rocket motor above the tailpipe of the FJ-4's J65-W-16A turbojet, and by late 1956 the NA-234 was superseded by the NA-248 which had an auxiliary 5,400 lb (24.02 kN) thrust Rocketdyne XLR36-NA-2 liquid-fuel rocket motor fueled by a mixture of hydrogen peroxide and JP-4. The US Navy showed immediate interest in the NA-248 and on July 3, 1957, it awarded North American a contract to convert the second and fourth production FJ-4s (BuNos 139282 and 139284) to NA-248 configuration, allocating the designation FJ-4F to the NA-248. (The NA-251 was an unbuilt FJ-4F proposal envisaged in 1957 with a variable-thrust rocket motor.) The FJ-4Fs not only utilized the auxiliary XLR36-NA-2 rocket motor but also had the enlarged fairing above the nose that would have house the search radar. The FJ-4Fs were delivered to the Flight Test Division of the Naval Air Test Center at NAS Patuxent River in Maryland on January 29, 1958, and the following day the FJ-4F made its first flight. A total of 22 flights were made with the rocket motor switched on, and the FJ-4F attained a top speed of Mach 1.3 during flight tests, its flight behavior similar to that of the FJ-4 although a deterioration of directional stability was noted above Mach 1.2 under positive load factor. After the conclusion of flight testing on May 10, 1958, the Navy issued published recommendations for further development of the FJ-4F to include a throttleable rocket motor, but the development of more powerful afterburning jet engines meant that the FJ-4F did not enter production. North American Aviation also worked out a handful of derivatives of the FJ-4B for the close air support role in the early 1960s. The FJ-4BF design proposed in late 1960 retained the FJ-4 wing but had a nose radome to house a terrain-clearance-and-search radar, a slightly bigger wingspan, seven store stations (six under the wings and one blow the centerline), and two General Electric TF37 turbofans on the sides of the fuselage. In response to the short-lived VAX requirement issued by the Navy in 1961, an FJ-4 derivative was conceived by North American as an interim design powered by one Pratt & Whitney TF30 turbofan (which necessitated a deeper fuselage) and a radome on the upper lip of the air intake for the APG-53A radar given that the VAX program was deemed too ambitious and therefore canceled in 1962. When the Navy announced the VAL requirement for an A-4 replacement with subsonic capability in early 1965, North American submitted a derivative of FJ-4 for the VAL competition, designated NA-295 by the company and unofficially called "AF-1F", which resembled the FJ-4 but had a radome protruding from the lower lip of the air intake that housed the APN-149 terrain-clearance radar, a TF30 turbofan, five stores stations with an option for two more, a length of 40 feet (12.04 meters), and a height of 16 feet 4.8 inches (5 meters). On February 11, 1964, the rival Vought V-463 design (based on the F-8 Crusader) was declared winner of the VAL contest and subsequently became the A-7 Corsair II, and thus the NA-295 was never built.

References:  

Buttler, T., 2021. American Secret Projects 4: Bombers, Attack, and Anti-Submarine Aircraft 1945 to 1974. Manchester, UK: Crécy Publishing.

Friedmann, N., 2022. U.S. Navy Attack Aircraft 1920-2020. Annapolis, MD: Naval Institute Press.

Kinzey, B., 2003. FJ Fury (Detail & Scale Volume 68). Carollton, TX: Squadron Signal Books.

Kinzey, B., 2021. FJ Fury in Detail & Scale, Part 2: FJ-4 and FJ-4B Variants. Detail and Scale Publications.

T-28 Trojan: the tricycle propeller-engined flying classroom from Inglewood

The North American Texan that I discussed in a March 2016 post is arguably the most prolific trainer aircraft ever manufactured by North American Aviation, having trained thousands of crewmen to fly technologically complex front-line combat aircraft that served with the militaries of the US and British Commonwealth in World War II. However, North American's production of piston-engine trainers for the US military did not stop with the end of Texan production despite the fact that the Jet Age had come into full swing. During my visits to the Palm Springs Air Museum, I happened to see for the first time in person another piston-powered trainer aircraft built by North America, the T-28 Trojan, and while North American surely made its mark on US combat aviation during the 1950s by manufacturing jet fighters while undertaking development of the XB-70 Valkyrie prototype supersonic bomber, what is sometimes overlooked is the fact that the T-28 not just replaced North American's own Texan as the mainstay of the US Air Force and US Navy's primary training squadrons in the 1950s but also had a secondary role as a light attack aircraft for use in the Vietnam War and by some US allies around the world.

Left: The first XT-28 prototype (serial number 48-1371)
Right: A trio of US Air Force T-28As in flight, 1960

The T-28 Trojan traces its roots back to the North American XSN2J prototype primary trainer (company designation NA-142) design, which was designed in response to a 1945 US Navy requirement for a new trainer to replace the N3N and N2S biplane primary as well as the SNJ Texans and SNV Valiant. The XSN2J first flew on February 10, 1947 and exhibited good performance, but did not enter production due to a tight Navy. Nevertheless, in response to a US Air Force requirement issued in late 1947 for a training aircraft to replace the Texan, a new trainer utilizing some design attributes of the XSN2J was conceived by North American under the company designation NA-159 along with a competing design by Douglas, which were designated XT-28 and XT-30 respectively. Like the XSN2J, the XT-28 had a squared-off vertical stabilizer with a spine at its base but utilized a tricycle landing gear and one Wright R-1300 Cyclone 7 radial engine. After inspection of mockups of the XT-28 and XT-30, in early 1948 the XT-28 was declared the winner of the USAF competition for a Texan replacement, and two XT-28 prototypes (serial numbers 48-1371/1372) were ordered. The first XT-28 prototype flew on September 24, 1949, and the Air Force ordered 266 production examples (serial numbers 49-1491/1756) of the baseline T-28 variant, the T-28A (company designations NA-171, NA-174, and NA-189), which was officially named Trojan in the early 1950s. The T-28A was 32 feet (9.75 meters) long with a wingspan of 40 feet 1 in (12.22 meters), a wing area of 268 square meters (24.90 m²), an empty weight of 5,111 lb (2,318 kg), and a gross weight of 6,759 lb (3,065 kg). Following suitability tests of the T-28A for the advanced training role at the Air Proving Ground at Eglin Air Force Base in Florida by the 3200th Fighter Test Squadron in June 1950 were deemed satisfactory by the US Air Force, production contracts were awarded for an additional 928 T-28As (serial numbers 50-195/319, 51-3463/3796, 51-7482/7891, 52-1186/1242, and 52-3497/3498) in 1950-1952. Deliveries of the T-28A to the Air Training Command began in 1950 and continued until 1953, by which time 1,194 T-28As had been built. The T-28A served a number of USAF training squadrons with the ATC in the mid- to late 1950s, replacing T-6Gs in service, but in 1957 the Air Force began phasing out the T-28A as the Cessna T-37 Tweet jet trainer began entering operational deployment, with retirement of the T-28A completed in the early 1960s, after which some T-28As were transferred to the Air National Guard to assist in proficiency training for squadrons transitioning to jet aircraft, many others were placed in storage or transferred to the air forces of Argentina, Cuba, Dominican Republic, Ecuador, Ethiopia, Mexico, Philippines, Saudi Arabia, South Korea, and Taiwan. In addition, 94 ex-USAF T-28As were given to the US Navy in the early 1960s and reserialled with BuNos 150356/150405, 150692/150716, and 153643/153659.

An early-production T-28B in flight, 1954

Despite having not ordered the XSN2J into production due to a tight fiscal climate in the late 1940s, the US Navy continued shopping for a successor to the SNJ Texan. In 1952, two T-28As (serial numbers 51-7527/7528) were acquired by the Navy for evaluation testing, receiving the BuNos 137636/137637 while being given the company designation NA-199 by North American. Evaluation tests of the T-28A showed that the Trojan required substantial modifications in order to meet Navy training requirements, including airframe strengthening to withstand the “drop-in” landing technique, a more powerful piston engine with a three-bladed propeller for better approach and wave-off characteristics, a hydraulically-operated speed brake in the belly, a fully-castering nosewheel, and a lower profile canopy. In response, North American developed a carrier-based version of the Trojan with a belly-mounted speed brake and one Wright R-1820-9 radial engine driving a three-bladed propeller, the T-28B (company designations NA-200 and NA-219). On September 2, 1952, a Navy contract was signed for 489 T-28Bs (BuNos 137638/137810, 138103/138367, and 140002/140052), and deliveries of the T-28B to the Naval Air Training Command began in 1953, continuing until 1955. The T-28C (company designations NA-226, NA-252, and NA-307) was a T-28B with arrestor hooks, strengthened landing gear, and shorter propellers with wider blades, designed to meet a new US Navy requirement for a carrier-qualification trainer, and in June 1954 production contracts for 372 T-28Cs (BuNos 140053/140077, 140449/140666, 146238/146293, and 154658/154729) were signed. The first flight of the T-28C occurred on September 19, 1955, and deliveries were made to US Navy units from 1956 to 1957; the T-28C batch with BuNos 154658/154729 was canceled before any could be built. The T-28B and T-28C were used for training student aviators at NAS Whiting Field and NAS Saufley Field in Florida as well as NAS Corpus Christi in Texas, and they served NATS units until the early 1980s, when they were retired from service as Beechcraft T-34Cs began reaching Navy training squadrons; before long, in late 1959, T-28Cs used for carrier-qualification training were supplanted in that role by the Lockheed T2V/T-1 SeaStar jet trainer. The Japanese Air Self-Defense Force was the only foreign operator of the T-28B, with one T-28B delivered to the JASDF in 1955 and given the serial number 63-0581; the company designation NA-218 was assigned by North American Aviation to the T-28B for the JASDF.

Top: Two T-28D Nomads (serial numbers 53-8367 and 53-8368, originally BuNos 138112 and 138127 respectively) in South Vietnamese markings during Operation Farm Gate, early 1960s.
Bottom left: Ex-USAF T-28A (serial number 51-3663), one of 148 T-28As converted to T-28S Fennec standard by France, photographed by me at the Palm Springs Air Museum on March 11, 2023.
Bottom right: A T-28S Fennec (ex-USAF T-28A serial number 51-7799) of the French Air Force in flight, 1960

In the late 1950s, with the Algerian National Liberation Front engaging in guerrilla warfare against French colonial rule over Algeria, 148 retired T-28As were sold by Pacific Airmotive (PacAero) in 1959 to France to be modified by Sud-Aviation into the T-28S Fennec counter-insurgency (COIN) aircraft, which had four underwing hardpoints for unguided air-to-surface rockets or 0.50-caliber machine gun pods, and one Wright R-1820-97 supercharged radial engine (thus the letter S in T-28S standing for "supercharged"). The T-28S Fennec was used by the French Air Force for airstrikes  against against Algerian independence fighters from 1959 to 1962, and after Algeria won its independence in 1962, the French Air Force sold most of the Fennec aircraft to other countries, with 65 sold to the Argentine Navy to be converted to the T-28P carrier-based attack aircraft with shortened propeller blades and arrester gear, and 25 given to the Royal Moroccan Air Force; several the Fennecs in service with the Argentine Navy and Royal Moroccan Air Force were transferred to Honduras, Tunisia, and Uruguay. For its part, in the early 1960s, PacAero modified 313 retired T-28s for the counter-insurgency, reconnaissance, search and rescue, and forward air control roles, and when these aircraft were delivered to the US Air Force, they were designated T-28D and christened Nomad. The T-28D had two underwing hardpoints for unguided air-to-surface rockets and other armament; the T-28D-5 sub-variant had ammo pans inside the wings that could be hooked up to hardpoint-mounted gun pods for greater center of gravity and aerodynamics. Three sub-variants of the Nomad differed in the variants of the R-1820 Cyclone engine, the Nomad Mark I with a supercharged R-1820-56S, the Nomad Mark II with one R-1820-76A, and the Nomad Mark III with one R-1820-80. Fairchild Hiller, in the meantime converted 72 T-28Ds to light attack aircraft under the designation AT-28D, fitting them with six underwing hardpoints and the rocket-powered Stanley Yankee ejection seat. The USAF began using its Nomads in combat in Vietnam in the Air Commando role in December 1961 as part of Operation Farm Gate, with the 4400th Combat Crew Training Squadron carrying out airstrikes against Viet Cong targets using trained South Vietnamese pilots. The USAF operated the Nomad until 1964-1965, when it was withdrawn from service and replaced by the A-1E Skyraider due to losses from ground fire and structural failures, and several ex-USAF Nomads were given to the Republic of Vietnam Air Force; various T-28Ds became inducted into the Vietnamese People's Air Force after the fall of Saigon in April 1975 and Vietnam's reunification under Hanoi's control in July 1976. The USAF and the Royal Lao Air Force, Royal Thai Air Force, and Khmer Air Force, in the meantime, used the Nomad for clandestine airstrikes against the Khmer Rouge and Pathet Lao over Cambodia and Laos, and Royal Laotian Air Force T-28Ds fitted with a camera pack behind the wing below the fuselage when used for photo-reconnaissance were called RT-28D. The CIA used the T-28D in secret operations against the Marxist Simba rebellion in the eastern provinces of the Democratic Republic of Congo in the early 1960s, and T-28D and AT-28D also served the air forces of Bolivia, Ethiopia, and Nicaragua in the late 1960s and 1970s, with several Nomads from USAF stocks delivered to those countries. The Hamilton Aircraft Company of Tucson, Arizona, undertook development of a civilian version of the Mark III subvariant of the Nomad, the T-28R Nomair, which had the R-1820-80 engine of the Mark III but had the wingspan increased by 7 feet (2.13 meters) to reduce stalling speed. The T-28R prototype, converted from an ex-USAF T-28A (serial number 50-201), first flew in September 1960 and received an FAA Type Certificate on February 15, 1962; six T-28As (serial numbers 49-1605, 49-1665, 49-1720, 50-202, 50-270, and 50-299) were modified by Hamilton Aircraft as the T-28R-1 Nomair I military trainer for use by Brazil and ten more T-28As were converted to the T-28R-2 Nomair II utility aircraft with provisions for one pilot and two rows of two passengers housed in a cramped cabin that opened from the port side (one T-28R-2 was later sold to a high-altitude photographic company).

The first YAT-28E prototype (serial number 0-21242) during a test flight in 1963. 

Not too long after the Nomad began operations in Vietnam, North American envisaged a derivative of the T-28D with one 2,445 shp (1,823 kW) Lycoming YT55-L-9 turboprop, the NA-284, which was armed with two 0.50 in machine guns and up to 6,000 lb (2,730 kg) of weapons on twelve underwing hardpoints. On September 17, 1962, the US Air Force awarded North American a contract to convert one ex-USAF T-28A (serial number 52-1242) to NA-284 configuration, and the YAT-28E designation was assigned to the NA-284. The first flight of the YAT-28E took place on February 15, 1963, and the first YAT-28E prototype made twelve successful flights before it was lost in accident on March 27 after it went into a flat spin during high G maneuvers, killing North American test pilot George Hoskins. Despite the crash of the first YAT-28E, the US Air Force was so impressed with performance of the aircraft that it authorized the conversion of two more T-28As (serial numbers 51-3786 and 51-3788) to YAT-28Es. The second and third YAT-28E prototypes differed from the first aircraft in having an LW-2 ejection seat, a taller vertical stabilizer, and a reinforced fuselage. The second YAT-28E prototype had a 9 inch (22.86 cm) tall vertical stabilizer, and the third YAT-28E was representative of the design of the proposed production YAT-28E, featuring a redesigned flat-topped cockpit canopy that allowed clearance for the ejection seats and improved rudder control and a slightly taller stabilizer measuring 12 inches (30 cm) high. The second YAT-28E prototype flew on November 15, 1963, and the third aircraft took to the skies in July 1964, but after a joint evaluation with North American of these two aircraft over a timespan of nine months, the US Air Force concluded that further development of the YAT-28E required extensive and costly redesign, and with the advent of the North American OV-10 Bronco and Cessna A-37 Dragonfly, it chose not to order the YAT-28E into production. The second and third YAT-28Es were returned to North American in January 1965 and placed in long-term storage, but when the US Navy began shopping for a new trainer to replace the T-28B and T-28C, North American proposed a navalized version of the YAT-28E, and the third YAT-28E prototype was loaned to the Navy for evaluation tests under a joint Air Force-Navy contract. Evaluation testing of the third YAT-28E by the US Navy began on January 24, 1966 and 27 flights were conducted, but the US Navy in May 1966 pronounced the YAT-28E unsuitable for the training role due to both technical and operational drawbacks, and the proposed navalized YAT-28E never materialized. The second and third YAT-28Es were again put in long-term storage, and they were initially donated to private owners before being given to C&J Aviation in 1999 a dismantled state in Camarillo, southern California; the third YAT-28E currently awaits restoration in Camarillo, while the fuselage and tail empennage of the second YAT-28E have been in storage in a private collection in Port Hueneme since 2014.

As a side note, the T-28 design formed the basis of the AIDC T-CH-1 Chung-Hsing, the first indigenous Chinese trainer to serve with the post-World War II Republic of China Air Force (which relocated to Taiwan after the communist takeover of mainland China in late 1949), which was powered by one Lycoming T53 turboprop and had tricycle landing gear like the Trojan. Fifty-two T-CH-1 trainers were built, and the first prototype flew on November 23, 1973 followed by series production starting in 1976 and continuing until 1981.     

References:

Adcock, A., 1989. T-28 Trojan in Action. Carrollton, TX:: Squadron/Signal Publications Inc. 

Avery, N., 1998. North American Aircraft: 1934–1998, Volume 1. Santa Ana, CA: Narkiewicz-Thompson.

Darke, S.M., 2013. "The North American T-28D". Air-Britain Aeromilitaria 39 (156): 147–155.. 

Hellström, L., 2014. "T-28s in the Congo – Part 1: Stemming The Rebellion". Air-Britain Aeromilitaria  (159): 117–128.

Hellström, L., 2014. "T-28s in the Congo – Part 2: Heyday of the Trojan". Air-Britain Aeromilitaria  40 (160): 147–157.

Hellström, L., 2015. "T-28s in the Congo – Part 3: The Twilight Years". Air-Britain Aeromilitaria 41 (161): 4–17.

Johnson, E.R., 2015. American Military Training Aircraft: Fixed and Rotary-Wing Trainers Since 1916. Jefferson, NC: McFarland & Company.

Thompson, K., 1999. North American Aircraft: 1934–1998 Volume 2. Santa Ana, CA: Narkiewicz-Thompson.