Share This Article

When the Soviets successfully tested their first atomic bomb in August 1949, the U.S. found itself dangerously unprepared to meet the new threat.

To say that the United States was unprepared for the dramatic expansion of Soviet military power during the immediate post–World War II period would be a gross understatement. No one could have realistically predicted that by 1949 the Soviets would acquire the technology to produce atomic bombs and aircraft capable of delivering them to North America (the Tu-4, a copy of the B-29). When the threat became apparent, U.S. Air Force Air Defense Command (ADC)—responsible for protecting America from air attack—was years away from being ready. A radar warning network across the northern U.S. and southern Canada was in the initial stages of development, but more acute, the only radar-equipped all-weather airplanes in ADC’s inventory were a handful of pro peller-driven night fighters, 41 Northrop F-61C Black Widows and 150 North American F-82E/F/G Twin Mustangs. Although the USAF had already made far-reaching plans to equip ADC units with missile- armed supersonic interceptors, these aircraft were not expected to become operational until the mid-1950s—and that was not nearly soon enough.

As new turbojet engines became available late in WWII, the U.S. Army Air Forces issued an Advanced Development Objective outlining the requirements for a jet aircraft to replace its existing fleet of prop-driven night fighters. The specification called for a two-seat, radar-equipped airplane with a top speed of 550 mph, an operational ceiling of 35,000 feet and a combat radius of 600 miles. In early 1946, after reviewing six design proposals, the USAAF narrowed the field to the Curtiss Model 29A and Northrop Model N-35, and Convair’s unorthodox delta-wing Model 7002 was spun off as a separate experimental project under the designation XF-92. Contracts to proceed with construction of night fighter prototypes were given to Curtiss as the XP-87 and to Northrop as the XP-89, with the expectation that prototypes would be flying within 14 months (mid-1947).

Although the two designs were similar in terms of their straight-wing layouts and takeoff weights, they differed substantially: The XP-87 was powered by four Westinghouse J34 engines paired in nacelles on the wings and featured a wide-section fuselage in which the pilot and radar operator were seated side-by-side, whereas the XP-89 incorporated a slender fuselage, with the crew seated in tandem under a long canopy and two Allison J35 engines housed in side-mounted nacelles below the wings. Both aircraft were designed to eventually be fitted with a nose turret housing four 20mm cannons. In the course of mockup inspections during 1946, Curtiss received the go-ahead to complete its flying prototype, but Northrop was directed by Air Material Command to make numerous design changes before proceeding further.

On March 5, 1948, nearly a year behind schedule, the XP-87 prototype was the first to fly. In June, most likely as a result of the USAF’s haste to obtain any type of jet interceptor, Curtiss received an order for 57 F-87A production aircraft, plus 30 RF-87A photoreconnaissance versions. Meanwhile, Northrop’s black-painted proto type, now the XF-89, made its first flight from the Muroc flight test base in California on August 16. In trials carried out between competing prototypes during the fall of 1948, the Air Force concluded that the Curtiss and Douglas entries were both seriously underpowered, while the somewhat faster XF-89 was deemed to have better potential for long-term development.


Canceling the Curtiss contract in late 1948, the USAF awarded Northrop a letter agreement in January 1949 to tool up for production of 48 F-89As, which was raised to 75 when the official production contract was granted in September. In the interval, the movable nose turret was discarded in favor of a fixed armament of six nose-mounted 20mm cannons, and the name Scorpion was officially adopted. The second prototype, delivered in a natural metal finish, began testing in November 1949, but the Scorpion program came to an abrupt halt in early 1950 when the first XF-89 was destroyed during a low-altitude speed run. The cause was found to be a failure of the horizontal tail surfaces induced by tail flutter, attributed to engine exhaust gas. The second prototype, redesignated YF-89, was subjected to extensive modifications before the flight program was allowed to resume. The changes included a longer, more pointed nose to house an AN/APG-33 radar and Hughes E-1 fire-control system, revised intakes and the addition of deflector plates behind the engine exhausts to minimize the tail flutter problem. Takeoff and climb performance was improved by fitting afterburning J35- A-21 engines that increased available power by 25 percent.

Eight F-89As incorporating the YF-89 improvements were delivered to the Air Force between September 1950 and March 1951, but were retained in the test inventory to conduct operational suitability trials. The first true operational Scorpions, 37 F-89Bs that were delivered between early 1951 and early 1952, differed from the A in having an autopilot and improved flight instrumentation. After initially entering service with the 84th Fighter Interceptor Squadron (FIS) at Hamilton AFB, Calif., the type equipped three more squadrons during 1951 and 1952.

Early operations with F-89Bs were marred by frequent engine failures and problems with the complex Hughes fire-control system. Starting in late 1951, the B models were followed by 164 F-89Cs, featuring improvements to the fuel system, new balance mechanisms on the horizontal stabilizer and a succession of engine up grades. F-89Cs became operational in early 1952, and went on to equip seven squadrons; however, the entire Scorpion fleet was grounded in September 1952 following the loss of six aircraft due to complete wing separations. An investigation revealed that the main attachment brackets had failed when twisting moments were imposed on the wings during high-G maneuvering. Over the next 15 months, all 194 F-89As, Bs and Cs were returned to Northrop in batches for the installation of stronger, machined attachment brackets plus fins added to the tip tanks that counteracted the twisting force. F-89s awaiting modifications were allowed to resume operations, but with restrictions on speeds and load limits.

The most numerous Scorpion production variant, the F-89D, debuted in October 1951 and became fully operational in early 1954. One of the early interceptor concerns facing ADC was adequate firepower plus the ability to fire weapons from greater ranges and at deflection angles that would permit better rates of closure. A new weapon designed specifically for that purpose, the 2.75-inch (70mm) Mighty Mouse folding-fin aircraft rocket (FFAR), was introduced in 1950. On the F-89D, the tip tanks were replaced by enlarged pods, each containing 308 gallons of fuel in the rear section and 52 FFARs carried in honeycomb tubes in the forward section. The rockets had an effective range of 2,000 yards, and when fired in salvo, could blanket an area the size of a football field. The fixed guns were removed to make room for an entirely new straight-tapered nose section that housed a new Hughes APG-40 radar and E-6 fire-control system, permitting beam attacks of up to 90 degrees of deflection. An upgrade to J35-A-35 engines boosted top speed to 635 mph (0.86 Mach) at 10,600 feet.

By the time the last of 682 F-89Ds were delivered in March 1956, they equipped 23 ADC squadrons located in the northern U.S., Canada and Alaska. Other proposed versions—the F-89E single-seat escort fighter, the F-89F powered by J47 engines and armed with guided GAR-2 Falcon missiles and the F-89G with a more advanced fire-control system—never progressed beyond the design stage.

The addition of a Hughes E-9 fire-control system and the ability to carry Falcon missiles resulted in the introduction of the F-89H in September 1955, with 156 examples delivered by August 1956, at which time Scorpion production ended. The wingtip pods of the H were designed to each accommodate three Falcons and 21 FFARs. Divided equally between semi- active radar and infrared homing types, the Mach 2.8 missiles could be fired at targets from four miles.

The final operational variant, the F-89J, came from 350 F-89Ds converted between November 1956 and February 1958, and was the first fighter of any type to be equipped with an air-to-air nuclear weapon. Douglas Aircraft had begun development of the unguided MB-1 Genie rocket in 1955. The 822-pound weapon could accelerate to Mach 3.3 and deliver its 1.5-kiloton nuclear warhead against targets within a radius of six miles, virtually annihilating everything within a 1,000-foot sphere. The Js retained the FFAR-armed tip pods of the D or mounted 600-gallon tip tanks. One or two Genies (and later up to four Falcons) could be carried on underwing racks. F-89Js first entered operational service in early 1957 and re-equipped Scorpion units through early 1958.

The process of phasing Scorpions out of active service began in 1954, when all remaining F-89Bs and Cs were transferred to Air National Guard (ANG) units. With the arrival of the long-awaited supersonic interceptors (F-102As in 1956, F-104As in 1958 and F-101Bs and F-106As in 1959), the USAF started turning over its F-89Ds and Hs to ANG units during 1957, and the last F-89J had been transferred from active service by the end of 1960. Scorpions continued to serve with ANG units through the 1960s, the last F-89Js being retired from the 132nd FIS of the Maine ANG in 1969.


In the fall of 1948, given the predictable delays associated with the testing and production of the XF-89, USAF officials looked at the alternative of adapting a jet interceptor from an already proven airframe. The most likely candidate was the Lockheed TF-80C (later T-33) trainer, a stretched, two-seat derivative of the F-80 that had flown in March 1948 and was already entering production. During October the Air Force authorized Lockheed to modify two TF-80Cs under the designation ETF-80C (later YF-94), and in January 1949 awarded a formal contract for procurement of 150 production aircraft as the F-94A. The conversion of the basic airframe to an interceptor configuration, however, was not as straightforward as had first been believed. In order to compensate for the loss of performance caused by the added weight of armament, radar and fire-control equipment, it was necessary to enlarge the aft fuselage so that the Allison J33-A-33 engine could be fitted with an afterburner. The installation of an AN/APG-33 radar set and Hughes E-1 fire-control system gave the nose its distinctive upturned profile, but left space for only four .50-caliber machine guns—exceptionally light armament for an interceptor.

Lacking guns and most operational equipment, the first ETF-80C flew on April 16, 1949, but almost immediately encountered unexpected flameouts during afterburner operations. While engineers from Allison and Lockheed were still working to resolve the problem, news that the Soviet Union had detonated an atomic bomb caused the F-94 contract to be increased twice, first to 288 aircraft, then to 368 before year end. Once the afterburner problem was fixed, the first F-94A was delivered to the USAF for testing in December 1949. With a top speed of 606 mph, it finally gave ADC an all-weather aircraft capable of intercepting a Soviet Tu-4 before it reached continental airspace. Thus despite many shortcomings such as unreliable electronic systems, frequent engine failures, inadequate cockpit space and unsafe ejection seats, F-94As were placed in operational service during the spring of 1950 with two ADC units based in Washington state, the 317th FIS at McChord and the 319th FIS at Moses Lake. Whatever may be said of the early F-94s, they were still the first frontline USAF aircraft equipped with afterburners and the first jets to equip ADC units.

The externally similar F-94B appeared in late 1950 with improvements to the canopy, cockpit arrangement, instrumentation and electronic systems, as well as Fletcher-type inline tip tanks in place of the earlier underslung versions. In April 1951, the 61st FIS at Selfridge AFB in Michigan became the first ADC unit to reequip with the B, and 356 examples had been delivered to the USAF by January 1952, bringing ADC’s force of F-94As and Bs to a total of 465 aircraft. While most were attached to units within Continental Air Command, with a smaller number serving with Alaska Air Command, three F-94A/B squadrons were deployed to Japan in 1951-52 for Korean War service. Their initial mission was to protect Japan from possible incursions by Soviet bombers, but they were later moved to forward bases in Korea for alert duty against North Korean night intruders and all-weather escort protection of B-29s. From early 1953 until the armistice, F-94Bs of the 319th FIS, operating out of Suwon, were credited with shooting down four enemy aircraft during night interceptions.

In mid-1948, even before the F-94A was ordered, Lockheed had tendered an interceptor proposal to the USAF for the considerably more advanced Model L-188. While utilizing a fuselage similar to the F-94’s, the L-188 was to be powered by a Pratt & Whitney J48 and feature an entirely new thin-section wing. Due to other interceptor projects in various stages of development, the Air Force expressed very little interest at the time. Lockheed, confident that the aircraft would eventually be procured, proceeded with construction of a company-funded demonstrator in 1949, which flew under civil registration N94C in early 1950. This resulted in a reappraisal of Lockheed’s project to the extent that the USAF purchased N94C and ordered a fully militarized prototype as the YF-97A. As a result of trials following delivery, the YF-97A underwent numerous refinements, acquiring a power-boosted swept horizontal stabilizer to reduce vibration at high Mach numbers, an enlarged vertical fin to improve directional stability and spoilers to enhance roll control. In lieu of gun armament, the nose section was modified to accommodate a battery of 24 FFARs surrounding a radome that housed an AN/APG-40 radar set. In September 1950, after an official decision had been made to procure more than 600 aircraft, the designation was changed to F-94C and the factory name Starfire was applied, apparently only to this version.

Although the first F-94C production model was delivered in July 1951, development problems with the fire-control system and cockpit seals, combined with engine flameouts caused by gas ingestion when the rockets were fired, delayed actual service entry until mid-1952, when the Starfire became operational with the 437th FIS at Otis AFB in Massachusetts. Starting with the 100th production model, streamlined fairings containing 12 FFARs were added to the leading edge of each wing, doubling firepower, and the feature was retrofitted to earlier Cs. Despite being 35 percent heavier than the F-94A/B, the F-94C’s added power and lower-drag wings yielded a top speed of 640 mph at sea level and allowed it to go supersonic in a dive. Progress in other interceptor programs caused procurement to be scaled back, so that a total of 387 F-94Cs had been delivered when production terminated in May 1954. At their peak during the mid-1950s, Starfires equipped 12 squadrons within ADC.

Due to advances in interceptor development, the active service life of all F-94 variants was comparatively brief. The process of phasing out F-94A/Bs commenced in mid-1953, and F-94Cs had been removed from ADC’s active inventory by early 1959, with the last operational Starfire retired by the Minnesota ANG that summer.


The most numerous of the ADC’s interim all-weather interceptors were North American F-86D/Ls, some 2,506 examples of which were accepted by the USAF from 1951 to 1955. In early 1949, soon after the F-94A had been ordered into quantity production, concerns over the viability of the F-89 program led Air Force officials to seek yet another alternative, this time based upon North American’s excellent F-86 Sabre, just then entering operational service. From the start, a decision was made to retain the single-seat configuration, but changes needed to incorporate the electronics and weapons for an interceptor led to departures from the basic F-86 design of such magnitude that it was redesignated the YF-95A. Instead of cannons or machine guns, it was to be armed with 24 FFARs carried in a tray that retracted into the belly, and in a parallel project, Hughes Aircraft was given the job of developing an all-new fire-control system (E-3 and later E-4) that would enable the new interceptor to fire its rockets from a head-on course as opposed to a traditional pursuit curve from behind.

When the final design of the YF-95A emerged, it bore only 25 percent commonality with the F-86A. Although the aircraft shared a similar J47 power plant, the rear of its fuselage was redesigned around an afterburner that boosted available thrust by 28 percent. To accommodate the AN/APG-36 radar and Hughes fire-control system, the forward fuselage was faired into a bullet-shaped radome over a reshaped, chin-type air intake. The cockpit layout was reorganized and enlarged for the additional electronic equipment, and access was improved with a redesigned canopy that hinged at the rear. Revisions to the empennage included increased fin area and an all-flying horizontal stabilizer in which dihedral was removed. Due to the urgency of the project, North American was authorized to start construction of two prototypes in July 1949 and tool up to build 122 production aircraft as the F-95A. Soon afterward, the official designation was changed to F-86D, and in September the contract was expanded to include another 31 aircraft.

Lacking armament and electronic systems, the first YF-86D flew from Edwards AFB in California on December 22, 1949. When the second, fully equipped prototype flew seven months later, trials indicated a top speed of 692 mph (Mach 0.91) at sea level. Although the aircraft was still in the preliminary development stages, mounting tensions between the U.S. and the Soviet Union— mainly due to the Korean War—generated significant increases in the F-86D contract, and the total order had risen to 979 aircraft by the middle of 1951. Deliveries of production models began in March 1951, but continuing problems with the fire-control and electronic fuel-control systems caused actual operational readiness to be delayed for two more years. The first production version to have true head-on intercept capability, Block 5s equipped with E-4 fire- control systems, did not begin operational evaluations until mid-1952, and further testing and development led to many other improvements incorporated into subsequent production blocks: power-boosted rudder, new radios and single-point refueling (Blocks 10-15); fuel filter de-icing and 120-gallon drop tanks (Blocks 20-25); automatic approach coupler control and omni- directional radarranging (Blocks 30-35); new glide-path indicator and J47-GE-17B engines (Block 40); and drag parachute and J45-GE-33 engines (Blocks 45-60). The final batch of Block 60 F-86D “Sabre Dogs” was delivered in September 1955.

The final interceptor variant, the F-86L, was actually an upgrade performed on 981 F-86Ds during 1956. Conversion entailed installation of the SAGE datalink system, which used a ground-based computer to transmit real-time radar information—target speed, altitude, bearing and range—to the aircraft’s E-4 fire-control system.

F-86Ds began entering frontline operational service in March 1953, and by mid- 1955 accounted for 73 percent (1,026 aircraft) of ADC’s overall aircraft strength. At their peak, a total of 1,405 F-86Ds and Ls equipped 20 ADC operational wings. As supersonic interceptors reached operational service, the Sabre Dogs were rapidly phased out of frontline service from August 1956 to April 1958. Like F-89s and F-94s, large numbers of F-86Ds and Ls went to ANG units, serving until the last examples were withdrawn in mid-1965.

As events actually transpired, the three stopgap interceptors ultimately bore most of the brunt of the Soviet bomber threat. Although history has revealed that the perceived threat was greatly overestimated, the huge ADC buildup—with more than 1,400 radar-equipped, rocket-armed interceptors in place by 1955—nevertheless represented a tangible deterrent against any Soviet inclination to make a first strike against the continental U.S.


 E.R. Johnson is a novelist, aviation author and practicing attorney who writes from Arkansas. He is a U.S. Navy veteran and a major in the Arkansas Wing of the Civil Air Patrol. Suggested reading: Encyclopedia of U.S. Air Force Aircraft and Missile Systems, Vol. 1: Post World War II Fighters 1945- 1973, by Marcelle S. Knaack.

Originally published in the May 2014 issue of Aviation History. To subscribe, click here.