Lockheed XF-90

Last revised September 15, 2003

The Lockheed XF-90 experimental long-range penetration fighter of the early 1950s always reminded me of the planes flown by the Blackhawks, who were comic book heroes when I was a wee lad. Hawkaaaaaa......!

The USAAF had found by painful experience in World War II that fighter escort was absolutely vital for the survival of bombers in enemy airspace. Unfortunately, the first jet aircraft were notorious fuel hogs and lacked the range and endurance of their piston-engined counterparts, and would be unable to escort long-range bombers such as the B-29, B-50, and B-36 all the way to their targets. In an attempt to solve this problem, the USAF considered all sorts of proposals for markedly increasing the range of jet fighter escorts, some of which bordered on the bizarre. One proposal for the solution to the escort fighter range problem was for the bombers to tow their escorting fighters into the combat zone and release them when their protection was needed. Several experiments were made with B-29s or B-36s towing P-80 or P-84 jet fighters, none of them being very successful and some being downright dangerous. Other proposals revived the parasite fighter concept of the 1930s, this time with jet fighters operating in parasite fashion from the bellies of large bombers. The best known example of this idea was the McDonnell XF-85 Goblin. Other ideas included the use of mixed power concepts such as that which produced the Convair XP-81. Others involved the construction of large, bulky fighters that were virtually flying fuel tanks, e.g., the Bell XP-83.

Initial attempts to produce jet-powered fighters with the endurance of piston-engined aircraft (e. g. the Bell XP-83 and the Convair XP-81) were disappointing, and in early 1946, the USAAF informally requested proposals for a "penetration fighter" with a combat radius of at least 900 miles and a performance capable of meeting all opposing fighters on more than equal terms. In addition, the USAAF wanted to keep the gross weight of the aircraft below 15,000 pounds. The requirements were somewhat ill-defined--combat range requirements changed from 900 miles to 1500 miles and then back to 900 miles, and climbing speed requirements were raised from an altitude of 35,000 feet reached in 10 minutes to 50,000 feet reached in less than 5 minutes. In addition, the aircraft was expected to be able to fly ground-attack missions if needed.

Spurred on by the USAAF request, the Lockheed Aircraft Corporation gathered a team under the direction of project engineers Don Palmer and Bill Ralston to work on a submission. In 1946, the Lockheed team submitted a design under the company designation of Model 090-32-01, which called for a delta-winged aircraft. The USAAF was sufficiently interested that on June 20, 1946, they issued Lockheed a contract for two prototypes under the designation XP-90. Serials were 46-687 and 46-688. A month earlier, the McDonnell Aircraft Corporation of St.Louis, Missouri had received a Letter of Intent from the USAAF for a competing design, the XP-88.

The XP-90 had originally been a delta-winged design, but wind tunnel tests carried out at the California Institute of Technology indicated that such a configuration would not work. Consequently, work on the construction of the delta-winged prototype was immediately halted, and those parts which had been already completed were scrapped. The Lockheed penetration fighter project was completely redesigned, and the design which finally emerged was the Model 90, which featured 35-degree sweptback wings, a sharply-pointed nose, and two Westinghouse J34 axial-flow turbojet engines mounted side-by-side in the rear fuselage and fed by side-mounted air intakes. The wings had leading-edge slats, Fowler flaps, and ailerons on the trailing edge. The pressurized cockpit was fitted with an ejector seat and a bubble canopy. Proposed armament was 6 20-mm cannon. The internal fuel could be supplemented by wingtip-mounted tanks, bringing total fuel capacity to 1665 US gallons.

The XP-90 was constructed of 75ST aluminum rather than the then-standard 24ST aluminum alloy. The aircraft was built with heavy forgings and machined parts. The result was an extremely well-constructed and robust aircraft. However, these innovations also resulted in an aircraft which had an empty weight more than 50 percent greater than that of its primary rival, the McDonnell XP-88. Since both aircraft were powered by the same engines, the performance of the XP-90 was certain to be much poorer than its rival.

In late 1947, the XP-90 was to acquire yet another competitor in the form of an enlarged version of the North American P-86 Sabre. In December of 1947, the Air Force ordered two examples of this North American design under the designation P-86C. Unlike the XP-88 and the XF-90, the North American entry was powered by a single engine, an afterburning Pratt & Whitney J48 turbojet. Since the P-86C design was so vastly different from the P-86A day fighter, the aircraft was later redesignated YF-93A.

On June 11, 1948, the designation XP-90 was changed to XF-90 when the P designation was replaced by F.

In the penetration fighter competition, the USAF initially favored the North American design because of its commonality with other Sabre variants, and in June of 1948 they supplemented the contract for the two F-86Cs with a contract for 118 production aircraft. It would seem therefore likely that the F-93 would be assured of a long and fruitful career with the USAF, and that the XF-88 and XF-90 would be consigned to oblivion. However, the F-93A production contract was suddenly cancelled in February of 1949. Several reasons were given. Perhaps the most important reason was a severe reduction in the military budget for that year. With limited funds available, it was decided to give priority to interceptors and to strategic bombers. In addition, a Senior Officers' Board felt that no production order for any penetrator fighters should be awarded until a competitive flyoff between the three contenders could be carried out.

The XF-90 was ready for flight testing in the spring of 1949. The engines were a pair of 3000 lb.st. Westinghouse XJ34-WE-11 engines. The XF-90 was trucked out to Edwards AFB, where it underwent its maiden flight on June 3, 1949 with the well-known test pilot Tony LeVier at the controls. Although test flying initially proceeded without many problems, the performance of the XF-90 was (as expected) rather sluggish because of its excess weight and the low power of its non-afterburning J-34 engines.

In search of more power, Lockheed decided to fit the second prototype (46-688) with afterburning Westinghouse XJ34-WE-15 turbojets, each offering 3000 lb.st. dry and 4200 lb.st. with afterburning. These engines were also retrofitted to the first prototype. When powered with the afterburning J34s, the aircraft were redesignated XF-90A. The XF-90A achieved a maximum speed of 668 mph in level flight at 100 feet, and reached a maximum speed of Mach 1.12 in a dive. It was on one of these maximum speed dives that one of the XF-90As was almost lost when Tony LeVier experienced great difficulty in pulling out of a dive.

Even though the afterburning engines gave the XF-90A a better performance, it was still not good enough to be considered acceptable to the USAF. The performance of the XF-90A was actually poorer than that of the F-86A Sabre, which was already in service. In pursuit of better performance, Lockheed explored three other developments of the XF-90: the Model 190-33-02 powered by a single Allison J33-A-29, the Model 290-34-03 powered by a pair of Westinghouse J46-WE-2s, and the Model 390-35-02 powered by a single General Electric J47-GE-21 engine. None of these projects ever saw the light of day, as they would have all required a major redesign of the intakes and fuselage to accommodate the larger air-flow requirements and the larger diameters of the engines.

The flyoff between the Lockheed XF-90, the McDonnell XF-88, and the North American YF-93A took place between June 30 and July 8 of 1950. On August 15, 1950, the Evaluation Board declared the McDonnell XF-88A to be the winner of the contest.

Work on the XF-90 was formally terminated in September of 1950. In 1952, the second XF-90A (46-688) was deliberately destroyed on the ground during a nuclear test at Frenchman's Flat in Nevada. The first XF-90A (46-687) was shipped to the NACA laboratory in Cleveland, Ohio in 1953. By this time it was no longer flyable, and was used for structural testing, exploring the limits of the extremely robust structure of the design. Presumably it was tested to destruction. In 2003, the hulk of the second XF-90 (46-688) was recovered by the USAF Museum from the Nevada nuclear test site. I assume that it will eventually be restored and put on display.

McDonnell's victory in the penetration fighter contest turned out to be rather hollow, since no penetration fighters were ever actually manufactured or placed in service by the USAF. At that time, wartime pressures mandated that priority be given to the procurement of existing types for use in Korea. In addition, the development of long-range, high-speed jet bombers such as the B-47 and the B-52 eliminated any real need for penetration fighters.

Specification of the XF-90A:

Engines: Two Westinghouse XJ34-WE-15 axial-flow turbojets, each rated at3000 lb.st. dry and 4200 lb.st. with afterburning. Dimensions: wingspan 40 feet 0 inches, length 56 feet 2 inches, height 15 feet 9 inches, wing area 345 square feet. Weights: 18,050 pounds empty, 27,200 pounds loaded, 31,060 pounds maximum. Performance: Maximum speed: 668 mph at 1000 feet. Climb to 25,000 feet in 4.5 minutes, service ceiling 39,000 feet, normal range 1050 miles, maximum range 2300 miles. The planned armament of six 20-mm cannon was not installed.


  1. Lockheed Aircraft Since 1913, Rene J. Francillon, Naval Institute Press, 1987.

  2. Fighters of the United States Air Force, Robert F. Dorr and David Donald, Temple Press Aerospace, 1990.

  3. The American Fighter, Enzo Angelucci and Peter Bowers, Orion, 1987.

  4. E-mail from Mike Hoffman on survival of hulk of 46-688.