In mid-1944 the Army Air Forces (AAF) developed a requirement for a single-place, medium-range, high-altitude day fighter that would serve as both an escort fighter and fighter-bomber. The AAF released the proposal to the aircraft industry in October that same year. Late the following month, North American Aviation started design studies, and they submitted four designs to the airmen in early 1945. After selecting a design, the AAF awarded the company a contract in May to build three copies of an aircraft designated XP-86. The next month the airmen approved a mock-up of the fighter.1 The AAF approved the design mock-up in June 1945.2
The design was of a straight-wing aircraft that borrowed from North American Aviation’s concurrent work on the FJ-1 Fury that it was developing for the Navy. The FJ-1 and XP-86 had the same general lines, bubble canopy, J35 engine, armament, and wing. The fighter’s straight wing was based on North American’s legendary P-51 wing and would mount four speed brakes that actuated above and below the wing, reminiscent of the original P-51 (A-36) design that had wing-mounted dive brakes. Deleting the Navy-specific equipment decreased the AAF aircraft’s design take-off gross weight by about half a ton to a design gross weight of 11,500 pounds and a maximum weight of 14,000 pounds. The designers also slimmed and lengthened the Fury’s bulky fuselage and shifted some of the fuel to the wing. A General Electric J35 engine that could develop 4,000 pounds of thrust powered the fighter. The FJ-1 could achieve a maximum speed of 547 mph; the XP-86 had an estimated top speed of 582 mph.3
North American understood that the project had problems. First, studies indicated the design would have essentially the same performance as its rivals, the XP-80 and XP-84. The company realized that, because these two competing fighters were further along in their development, the XP-86 design would probably be canceled. Perhaps more importantly, it could not meet the AAF’s top-speed requirement of 600 mph. North American met the challenge and came up with a dramatic change that would greatly improve the aircraft’s appearance and performance: the swept-back wing configuration.4
The swept-wing concept predated the North American design by a decade. For some time, the Germans had known that sweeping back the wing would facilitate higher speeds. At the 1935 Volta Scientific Conference on high-speed flight, Adolph Busemann delivered a paper discussing how swept-back wings would lessen drag at high speeds, enabling supersonic flight. The Germans quickly realized the importance of this concept, declaring it classified the next year and applying for a patent. A 1940 German report discussed the advantages of swept wings operating at high speeds (Mach 0.9); as early as 1941, the Germans conducted wind tunnel tests on the concept. Meanwhile the rest of the world virtually ignored this idea, and the Germans took the lead in the fields of jet propulsion and high-speed flight.5
The two principal German non-propeller fighters that entered combat during World War II featured swept-back wings, although neither adopted the swept-back wing for speed.6 The Germans had a number of other designs on the drawing board that also employed swept-back wings. In contrast, the jet-powered aircraft flown by the Anglo-Americans during this period had straight wings.
The Americans knew the advantages of swept-back wings.7 In mid-1944 the AAF conducted wind tunnel tests of swept-back wings, and by early the next year Bob Jones at National Advisory Committee for Aeronautics (NACA) had developed a theory about swept wings. At the same time, the American military recognized the German advantages in a number of aeronautic areas and therefore systematically gathered up German documents, aircraft, and personnel at war’s end.8 In 1945, the Technical Intelligence Survey team made German test data and interrogations of captured Germans available to North American. Larry Greene, a project aerodynamicist, who was competent in technical German, along with Dale Meyers and Harrison Storms, the chief technical engineer, worked late into the night for a week reviewing the material. They also studied copies of scientific papers at the Cal Tech library investigating the subject, and their assessment encouraged North American to set up an aerodynamics performance group, under the leadership of Jack Daniels, to calculate the performance of the swept-back wings. They were astonished at the improvement in performance.
By this time, however, North American was far along with the straight-wing design; in fact, they had begun construction. As a consequence, there was resistance to incorporating the new wing into the XP-86. Storms took the findings to Ed Horkey, chief aerodynamicist, who was thrilled by the results. But when he in turn took the news to his boss, Joe Greer, he met a roadblock. Greer thought it was horrible that the concept could not be used because the process was too far along. Storms, or Horkey, went over Greer’s head to Greer’s boss, Larry Waite, head of the technical section. He, too, thought it was too late to make a big change, remarking, “Get rid of it; bury it!” Storms was not dissuaded and instead went across the hall to Ed Schmued, head of advanced design (and designer of the P-51), who reacted “like a match dropped into a barrel of sky rockets.” He took the material to the heads of the company, Dutch Kindleberger and Lee Atwood. They in turn enthusiastically discussed the findings with the AAF engineering department in Dayton, Ohio, which gave verbal approval to go forward with the new wing. When the word got back to California, Ray Rice, vice president and chief engineer, was angry that he had not been consulted, an oversight that was quite embarrassing for Storms and Horkey. Nevertheless, Rice authorized wind tunnel testing of swept-wings in late August 1945.9 At this same time (the precise chronology is not clear), Ed Horkey went to the NACA to test the proposed North American airfoil.
North American built a .23-scale swept-wing model in two weeks and by mid-September had it in a wind tunnel. These tests indicated the benefits of the concept. The testers also applied a bent piece of metal around the leading edge to represent partial span slats, suggesting how performance could be improved at low speeds. North American used the extensive data the Germans had amassed for their proposed fully-swept Me 262 wing, coupled it with a North American laminar flow air foil, and extensively tested it in the wind tunnel. The company adopted the swept-back wing. The aircraft retained the straight empennage for a time, but by mid-October 1946, North American also swept back the empennage. This change slightly altered the aircraft’s dimensions.
Another consideration in the wing design was the aspect ratio (the ratio of the span [wing length] to chord [wing width]). North American began with the 5:1 aspect ratio wing, and tested both it and also one with 6:1 ratio. That the latter had obvious performance advantages over the former, especially greater range, encouraged North American to change the baseline wing design to a 6:1 aspect ratio in October 1945. However, the higher aspect-ratio wing had stability problems, vicious pitch up tendencies that the design team was unable to solve. Therefore, in March 1946, the design changed back to the 5:1 layout.10
Rice discussed the swept-wing with Gen. Bill Craigie, who was in charge of research and development at Wright Field. Although the change would add six months to the development time, the Air Force gave its formal approval to the swept-wing modification on 1 November 1945.11 The engineers calculated that the swept wing increased the critical Mach number of the wing (the speed at which the wing encountered supersonic flow) from Mach 0.8 with the straight wing (the same as the P-80) to Mach 0.9. This added 70 mph to the fighter’s top speed and thus enabled it (in theory) to exceed the AAF high-speed requirement.12 The decision to redo the aircraft’s wing also improved the fighter’s looks, giving it the elegant lines for which it is renowned. The F-86 was a good-looking airplane and would strongly support the old pilot saying, “if it don’t look good, it won’t fly good.”13 Better stated, “This bold decision to adopt the swept-wing changed the Sabre from a mediocre fighter into a great one.”14
Along with the wing change, North American also lengthened and strengthened the fuselage, moved the engine back about 2.3 feet to accommodate more fuel and to deal with center-of-gravity issues, enlarged the cockpit to make room for an ejection seat, and changed the engine from the J35 to the J47.15
The swept-wing brought with it problems: with this configuration the fighter was less stable at high angles of attack and had a higher stall speed than the straight-wing design. To deal with these problems the Germans had used automatically operating leading-edge slats on the Me 262. Although there was some resistance at North American to “Germanize our design,” the infamous “not-invented-here” mindset, the slats’ merits won out. North American tested full-span leading-edge slats on a wind tunnel model in September.16 Wind-tunnel tests with the 5:1 aspect-ratio wing yielded positive results, but testing 150 different slat configurations with the 6:1 aspect-ratio wing did not, pushing North American to adopt the lower aspect-ratio wing. Meanwhile, the AAF shipped to North American the Me 262 wing (with slats) that would provide the basis for the fighter’s slat system.17
In fact, the first seven aircraft built used the German design. American engineers went on to modify the system. At low air speeds, the slats automatically extended forward and downward on a curved track, increasing the wing’s area and camber, producing more lift, stability, and control at low air speeds. The slats automatically closed at higher air speeds as aerodynamic forces pushed the slats snug with the wing contour, decreasing drag. The pilot then locked the four moveable slats into place to preclude asymmetric deployment in maneuvering. When the pilot unlocked the slats, they automatically deployed as the aerodynamic forces decreased at lower air speeds. To ensure the slats were operable for takeoffs and landings, the designers interconnected the slat locks and landing gear so that the slats were unlocked when the gear was extended. North American used the slat locks on the first 159 fighters, and then removed them in early 1949 when it incorporated a modified slat system into the fighter.18
On 8 August 1947, North American rolled out the first XP-86. The fighter then went through taxi tests at the adjacent Mines Field (now Los Angeles International Airport) before the company disassembled the aircraft and trucked it to the Muroc test facility (now Edwards Air Force Base).19 Now theory and design would give way to practice.
North American chose George “Wheaties” Welch as the chief test pilot for the P-86 project.20 Unlike most AAF aviators, he was a son of privilege: his father was a senior research chemist with Dupont, and the family knew the Duponts personally. Welch graduated from a private secondary school and attended Purdue for several years before joining the Army aviation cadet program. He won his wings and commission in January 1941 and was posted to Hawaii. Francis Gabreski, of whom we will hear more later, was one of his fellow pilots there and described Welch as a rich kid, a real hell-raiser, but an excellent fighter pilot. Welch proved the latter when he gained everlasting glory as one of six AAF pilots at Pearl Harbor who got airborne and downed Japanese aircraft on 7 December 1941, getting the most kills: four Japanese aircraft on three sorties. Although nominated for the Medal of Honor by AAF chief Henry “Hap” Arnold, whose endorsement normally would have ensured that award, for unknown reasons he was instead awarded the Distinguished Service Cross.21 Welch went on to complete three combat tours, becoming a triple ace (sixteen credits) against the Japanese.
After 348 combat missions and a bout with malaria, Welch left the service in 1944. With Arnold’s recommendation, he joined North American in July of that year and became the chief test pilot of their Inglewood (Los Angeles) plant. Welch was to become the civilian pilot most closely associated with the F-86. Because he was independent and a loner capable of playing the prima donna role to the point of rudeness, he did not receive the publicity, acclaim, or credit that other test pilots received—and that he well deserved.22
On 1 October 1947, Welch took the XP-86 up for its initial flight and had to recycle the landing gear to get an “up and locked” indication on the nose gear. Having accomplished each of the test items as planned, he prepared to land thirty minutes later. Again, the nose wheel malfunctioned, failing to lock in the extended position. Welch recycled the gear a number of times, tried a 3 “g” maneuver, and used the emergency gear extension system, all without success. The nose wheel would not extend completely and was hanging at a 45-degree angle. The North American team on the ground radioed Welch that it was up to him to either bring the aircraft down without the nose wheel or to abandon the aircraft. Welch elected to land the fighter, but on the dry lakebed instead of the runway. He made a smooth approach with the nose high, touched down on the main gear, and held the nose wheel off the ground as long as possible. As the speed decreased and the nose began to drop, at the very last moment before touchdown, the nose gear snapped forward and locked. Welch radioed: “Lucky! Lucky!”23 Flight number one was complete, despite some anxious moments. Welch logged fifty minutes of flying time.24
The design problem was two-fold: the aircraft’s hydraulic cylinder was smaller than specified, and the nose gear retracted rearward and extended forward against the slipstream, an arrangement that was mandated by the aircraft’s air duct design. Thus, a relatively simple and well-tried system—the landing gear and particularly the nose gear—almost led to disaster on the first flight and continued to dog the fighter early in its lifetime.25
Welch was one of the first pilots to exceed the speed of sound. Clearly, this was a major feat, perhaps not so well appreciated today when military aircraft and the commercial Concorde have regularly flown at supersonic speeds. Five decades ago, however, the media wrote of a “sound barrier.” The death of Geoffrey DeHaviland, son of the British aviation pioneer and owner of the British aircraft firm, along with a popular film of 1946 led to the mystique and anticipated perils of supersonic flight.
The official story is that Capt. Charles “Chuck” Yeager was the first to break the sound barrier on 14 October 1947 in the rocket-powered Bell X-1. This has been highly publicized and burned into the public imagination by the photogenic and likeable Yeager, and most especially by the book and movie, The Right Stuff. The USAF acknowledges that the F-86 broke the sound barrier, albeit in a dive, in April 1948. (The Air Force made this disclosure after an excited British pilot, Roland Beaumont, gave details over an open microphone that appeared in the June 1948 Aviation Week.)
As in most cases, there is more to the story. There was a competition for bragging rights over who would be the first to exceed the speed of sound. The Secretary of the Air Force, Stuart Symington, and Bell (builder of the X-1) wanted the glory, as well as the accomplishment, to justify the expensive X-1 project. North American had the potential to ruin the show and was warned by Symington not to do so. Nevertheless, North American obtained the use of highly accurate NACA speed tracking equipment, which was in place for Yeager’s flight, to monitor some of Welch’s early flights at Muroc. The device showed that Welch flew faster than Mach 1 in October or November 1947, either two days or a month after Yeager’s feat.26
Some go further, claiming Welch was the first to go supersonic. Horkey implies Welch may have done so, since he was flying the same flight patterns prior to 14 October as he did when NACA registered his supersonic dive.27 Al Blackburn, a former North American test pilot, goes beyond this. He relates that, prior to the XP-86’s initial flight, Welch had asked one of the regulars (Millie Palmer) at Pancho Barnes’s famous “Fly In” bar just outside the field to be alert for “a sharp boom like a clap of thunder” and to write down the time and the reaction of others to the event. On the first flight, Welch put the XP-86 into a 40-degree dive from thirty-five thousand feet aimed at Poncho’s bar, and at twenty-nine thousand feet encountered slight wing roll and then the airspeed indicator jumped from 350 kts to 410 kts. As he pulled level at twenty-five thousand feet, he encountered the effect again. Palmer reported a “baboom.” The noise and the airspeed jump are consistent with supersonic speed. Welch repeated the feat on 14 October, only fifteen minutes before Yeager’s famous flight.28
So what? Clearly there is a difference between level supersonic flight, as the X-1 demonstrated, and supersonic flight in a dive. Certainly Yeager’s achievement remains historic. On the other hand, there is a major difference between accomplishing the feat with a rocket-powered test aircraft carried to altitude by another craft and capable of only a few minutes of powered flight, and doing so in sustained flight or in an operational aircraft.29 In any event, the Sabre was the first combat fighter to fly supersonic. The X-1 had its moment in history; the F-86’s was yet to come.
North American turned the first test plane over to the USAF in November 1948, after almost ninety-eight flying hours. It served at Wright-Patterson Air Force Base for two and a half years before returning to Edwards Air Force Base (formerly Muroc) in April 1951. In May 1952, it was transferred to Kirtland Air Force Base. According to most secondary sources, the aircraft was wrecked in September 1952. Air Force records, however, indicate that the aircraft was tested to destruction at Kirtland AFB in September 1952. North America flew, built, and tested two other XP-86s.30
The F-86 evolved as it was tested. One major change was to replace the J35 with the J47 engine. The first thirty-three planes (F-86A-1) were powered by the J47-GE-1, an engine that developed 4,850 pounds of thrust (21 percent more than the J35), which of course improved performance. Top speed increased from 618 mph at fourteen thousand feet with the J35 to 677 mph at sea level with the J47. By May 1950 the F-86 was powered by J47-GE-7 and GE-13, which had a rating of 5,200 pounds of thrust.31 North American also changed the windshield. The company installed a curved, bulletproof glass for the forward cockpit panel in the first F-86s. However, as this introduced an error into the gunsight, the company changed to a “V” shaped front glass in the F-86A-5.32
One of the unique features of the original F-86 design was gun doors that covered the muzzles of the six 0.50-caliber machine guns. These smoothed out the fuselage, cut drag, and created a more aesthetically pleasing design. When the pilot pulled the trigger the doors opened in 1/20 of a second and closed when it was released. While there was little gain in top speed, cruising speed and range improved. However, there were problems. The doors occasionally failed to operate and were shot off. They also cut the airflow to the guns, reducing cooling of the barrels, shortening barrel life, and causing “cook-offs,” ammunition that was inadvertently fired due to excessive heat. Additionally, there were also maintenance complications, since the doors required considerable adjustment and tightening of the fasteners. The USAF tested open gun ports in the summer of 1949 and concluded they were superior to the gun doors. This change was quickly made.33
Another much more important change was the placement of the speed brakes. Such devices were not necessary for propeller-powered aircraft, because changing the pitch of the propeller (blade angle) could rapidly slow down the aircraft. Only in specialized operations such as dive-bombing were other devices needed to slow an aircraft; hence, the origin of dive brakes. Jet-powered airframes were much more streamlined than their predecessors although extended landing gear and flaps slowed the aircraft. The original F-86 design called for wing-mounted speed brakes; however, these were deleted in September due to a change in the wing design. They were replaced by three hydraulically actuated devices mounted on the fuselage aft of the wing’s trailing edge, two on the side and one underneath. Wind-tunnel tests indicated problems, so they were rendered inoperative in the XP-86. (There is no evidence that they were ever employed during flight.) North American later changed the arrangement to two side speed brakes that increased the aircraft’s drag by a factor of 2.5.34
The new aircraft was a winner. Although the production model was slightly heavier (about 350 pounds) than test models, its new engine (J47-GE-1) was much more powerful than the J35 it replaced. The USAF was quick to show off its new aircraft. The world aircraft speed record, confirmed by the precise rules of the Federation Aeronautique Internationale (FAI), had first exceeded 600 mph in November 1945 when an RAF Meteor clocked 606 mph. An RAF Meteor raised it to 616 mph in September 1946. The AAF broke this record with a modified P-80 in June 1947 (623 mph); the navy in turn broke that record twice with a Douglas D-558 Skystreak in August 1947 (the higher at 651 mph). At the September 1948 Cleveland National Air Races, the USAF attempted to set a world speed record with an unmodified production model F-86. Maj. Robert Johnson of the flight test division at Wright-Patterson AFB flew a Sabre under the precise FAI rules: two runs in each direction (to negate any tail wind) at a precise altitude (below 165 feet, to preclude building up speed in a dive) over a 3 km course and recorded by camera. Before a crowd of eighty thousand, Johnson flew the course three times in each direction, at an average speed of 670 mph. This was not accepted as a new record, however, because the cameras recorded only three of the runs. Ten days later, Johnson fared better when he flew six courses in another F-86 at Muroc. The FAI certified the average speed of 671 mph as a world record, which stood for three years until it was broken by an F-86D at 699 mph in November 1952.35
In February 1949 the 94th Fighter Squadron, the renowned “hat-in-the-ring” squadron and the unit of Eddie Rickenbacker, America’s top ace in World War I, became the first USAF unit to receive the new fighter. It was part of the illustrious 1st Fighter Group, stationed at March Air Force Base. While the First would not see action in Korea, it did have an influence on the history of the F-86, for it named the fighter. In February, officers of the First examined seventy-eight names submitted by the group’s men in a naming contest, and selected “Sabre” (with the distinctive British spelling). Unfortunately, we do not know the author of what became the aircraft’s common name as of March 1949.36