The XF-86C / XF-93A Penetration Fighter

While the Sabre was engaged in heavy combat across the Pacific, the design was being updated back in the States. Design work on the "NA-157" or "XP-86C" was begun in late 1947, in response to a USAF requirement for a "deep penetration" fighter. North American's XF-86C, as it was redesignated in 1948, won the competition against the Lockheed "XF-90" and the McDonnell "XF-88" (which would eventually evolve into the F-101 Voodoo), with the Air Force ordering two prototypes.

The XF-86C had a bigger and longer fuselage than the F-86A. The increased size was to accommodate 5,909 liters (1,561 US gallons) of internal fuel to meet the range requirements, while the the increased length was to allow fit of an afterburning Pratt & Whitney J48-P-6 engine. The J48 was an improved, afterburning version of the Rolls-Royce Nene centrifugal-flow turbojet, manufactured in Britain as the Tay, providing 27.8 kN (2,835 kgp / 6,250 lbf) dry thrust and 35.6 kN (3,625 kgp / 8,000 lbf) afterburning thrust. The J48's exhaust had a two-piece clamshell variable-size outlet.

The fuselage was designed using the new "area ruling" concept developed by NACA engineers, which specified that changes in an aircraft's cross-sectional area should be minimized to ensure smooth airflow at high speeds; this meant that the fuselage was "pinched" slightly along the wing roots. The result was a somewhat inelegant machine compared to the F-86A, with a porpoise-like body and fat appearance. In fact, the aircraft was so clearly different that the USAF redesignated the type the "XF-93A" in 1948.

The nose intake was replaced with air intakes at the sides, leaving the nose available for SCR-720 radar, and the new aircraft was armed with six 20 millimeter cannon instead of six 12.7 millimeter Brownings, with 225 rounds per cannon. The XF-86C's greater weight required reinforced landing gear, with dual wheels on the main gear. The twin air brakes of the Sabre were replaced with a single large air brake under the fuselage.

The XF-93A had excellent performance and range, and it could be fitted with stores pylons for external tanks to give even greater range, or to carry up to 900 kilograms (2,000 pounds) of bombs, rockets, or other stores. The Air Force ordered 118 production F-93As in 1948.

The first XF-93A prototype performed its initial flight in January 1950, again with George Welch at the controls. Only two were built, neither being armed; the contract had been cancelled a year earlier, since the USAF's new Boeing B-47 bomber was so fast that it didn't really need fighter escort, and money was tight anyway. Mid-air refueling would soon kill the "penetration fighter" concept completely. The two XF-93A prototypes were finished as test articles, and flown in this role by NAA, the USAF, and NACA, eventually ending up as NACA property. They were used in various experiments into the late 1950s, in one case fitted with scoop-type air intakes instead of the original flush intakes, and then scrapped.

The F-86D (F-95) Sabre Interceptor

Even while the Air Force had building up fighter groups with the F-86A, work was proceeding on an interceptor variant of the F-86. Development of the big Northrop "F-89 Scorpion" interceptor on which the USAF was resting their hopes had been delayed, and the USAF was desperate to get an aircraft to defend North America from Soviet bombers. The service decided to obtain a derivative of the Lockheed T-33, to become the "F-94 Starfire", and a derivative of the Sabre as interim interceptor solutions.

Development of the first Sabre interceptor variant, originally designated the "F-95A", was initiated in early 1949, with rollout of the first of two prototypes, lacking armament and combat avionics, in November 1949, and first flight on 22 December 1949. Once again, George Welch was at the controls, and once again the nosewheel wouldn't go down. This time it didn't pop open at the last minute, and Welch had to pancake the prototype onto the runway, damaging it.

The second prototype first flew in September 1950, and was fitted with a Hughes E-3 fire-control system (FCS). By this time, the aircraft had a new name. The F-95A was an almost complete redesign of the original F-86, with only about 25% commonality. However, the US Congress would not approve new aircraft at the time; a modification of an existing aircraft wasn't such a problem, and so the aircraft was redesignated the "F-86D" in the summer of 1950 to ensure funding.

The prototypes were redesignated "YF-86D". The first production "F-86D-1" was rolled off the Inglewood assembly line in March 1951, with first flight in June. 37 were built. This was followed by the "F-86D-5", featuring the definitive Hughes E-4 FCS, with much longer range radar than the E-3. The first was flown in July 1952, though the new FCS proved unreliable and operational delivery of the subtype was protracted. NAA built a total of 26 F-86D-5s.

The F-86D had only a general resemblance to an F-86A, the most noticeable difference being the F-86D's big nose radome, and modified air intake to accommodate the radome. Of course, this arrangement required a complete redesign of the Sabre's forward airframe. The big nose gave the F-86D a faintly cartoonish appearance. Although the F-86D prototype had a standard F-86A sliding canopy, production F-86Ds had a clamshell canopy and a flat armor-glass windscreen. The F-86D also had a "slab" all-moving horizontal tailplane with no elevators.

The F-86D was fitted with the J47-GE-17 afterburning turbojet, with 22.3 kN (2,270 kgp / 5,000 lbf) dry thrust and 29.6 kN (3,015 kgp / 6,650 lbf) afterburning thrust, plus an electronic fuel-control system. The prototypes had been fitted with development J47-GE-17 engines, with about ten percent less thrust. The rear fuselage was widened and lengthened to accommodate the new engine, with some changes in the design between the prototype and production aircraft.

The J47-GE-17 gave the F-86D impressive performance for the time. In late 1952, an F-86D would set a world speed record of 1,124.6 KPH (698.505 MPH), and in July 1953, another F-86D would establish a new record of 1,152.3 KPH (715.7 MPH).

The F-86D had no guns. Its armament consisted of 24 70 millimeter (2.75 inch) unguided "folding-fin air rockets (FFAR)", stored in a belly tray that could be lowered in half a second. The rockets could be fired in salvos of 6, 12, or 24 at intruding bomber formations. Each rocket had a range of over 4 kilometers (2.5 miles) and a warhead weighing 3.4 kilograms (7.5 pounds).

By the standards of the time, the F-86D was an electronic marvel. Most radar-equipped aircraft were two-seaters, carrying a pilot and a radar operator. The F-86D was a single-seat aircraft. Although early concepts for the aircraft had envisioned a two-seat machine, this led to reductions in performance and range, and new developments in electronics encouraged NAA and the Air Force to think that advanced technology could replace the "back-seater" in guiding the pilot to the target.

In operation, the pilot would be directed to the vicinity of a target by a "Ground-Control Intercept (GCI)" operator. At a range of about 50 kilometers (30 miles), the pilot would acquire the target with the aircraft's AN/APG-7 radar. The E-4 fire-control system, which integrated the radar with an AN/APA-84 electronic analog computer system, would provide a course to intercept the target. At the appropriate time, the E-4 would indicate that rockets should be fired as the two paths intersected, the F-86D streaking past the target from the side or front. The E-4 would also provide a warning if there were danger of a collision. The F-86D had a manual lead-computing gunsight as a backup.

In reality, dumping the workload normally handled by a pilot and a radar operator in a two-seat interceptor on a single man proved to be challenging. Despite the fact that the electronics were intended to take up much of the load, the F-86D was said to require the most training of any contemporary USAF plane, even more than the Boeing B-47 bomber.

The small batch of F-86D-5s was followed by another small batch of 36 "F-86D-10s", featuring a power-actuated rudder with no trim tab. Earlier Sabre variants had used a manually-actuated rudder with a trim tab. This led in turn to a long sequence of other subvariants with generally minor changes:

• F-86D-15 (32 built): Featured a "single-point" refueling system that allowed quick turnarounds on the ground, as well as a new radio.
• F-86D-20 (188 built): Featured a fuel filter de-icing system.
• F-86D-25 (88 built): Featured twin 454 liter (120 US gallon) drop tanks.
• F-86D-30 (200 built): Reverted to the manual rudder with trim tab.
• F-86D-40 (300 built): Fitted with improved J47-GE-17B engine, with 24.1 kN (2,460 kgp / 5,425 lbf) dry thrust and 33.4 kN (3,400 kgp / 7,500 lbf) afterburning thrust. Also provided with a few additional cockpit controls.
• F-86D-45 (300 built): Fitted with a compartment for a ring-slot-type drag chute with a diameter of 4.88 meters (16 feet) at the base of the vertical tailplane. The drag chute was required because the Air Force had begun deploying F-86Ds to Japan, Okinawa, and Formosa, where the runways were shorter. Interestingly, NAA obtained the expertise on installing the drag chute from Lockheed, in trade for NAA information on FFAR armament. The last 62 of this batch had the J47-GE-33 engine, with 24.7 kN (2,520 kgp / 5,550 lbf) dry thrust and 34.0 kN (3,470 kgp / 7,650 lbf) afterburning thrust.
• F-86D-50 (301 built): Minor electronic and instrument update.
• F-86D-55 (225 built): Another minor electronic and instrument update.
• F-86D-60 (399 built): Yet another minor electronic and instrument update, and the last production batch of the F-86D. Including the two prototypes, a total of 2,506 F-86Ds were built.

The F-86D was known as the "Sabre Dog" or "Dogship". It seems that these two names were not entirely affectionate, and the second undoubtedly led to a cruder nickname when things weren't going well. The advanced technology used in the interceptor led to painful teething troubles.

The leading-edge afterburning J47 engine proved to be unreliable. Delivery of the E-4 fire-control system was delayed, and when it did arrive, it was plagued by manufacturing defects. At one point, the delays in delivering acceptable fire-control and other electronic subsystems were so bad that there were 320 F-86Ds lined up on the field outside the North American manufacturing plant, awaiting arrival of the boxes needed to complete them.

Problems continued to plague the F-86D after it was delivered. A rash of 13 fires and explosions grounded them all in late 1953, until changes in the electronic fuel-control system were implemented. Then, in early 1954, another streak of 19 accidents occurred, and the Sabre Dogs were grounded again. The answer was "Project Pullout", in which the Air Force, working with North American, implemented a comprehensive set of fixes to the 1,000-plus F-86Ds in service, bringing them up to F-86D-45 standard. The refit took 18 months and cost $100 million USD. After the update, the Sabre Dogs proved much more satisfactory in service. They were deployed in large numbers in the US, Europe, and Far East.

F-86E Through F-86L

The US continued to refine the Sabre design to obtain absolute air superiority over Korea, and refinements continued after the cease-fire. The F-86 would be a major weapon in the early years of the Cold War, assigned to roles including reconnaissance, air defense, and nuclear strike.

Sabre F-86E, F-86F / Conclusion in Korea

The XF-86C / XF-93 had been a major departure from the F-86A. However, NAA also developed a more evolutionary enhancement of the F-86A in the form of the "F-86E".

The F-86E was difficult to distinguish from the F-86A, but featured improved flight controls, an uprated J47-GE-13 engine with 24.2 kN (2,470 kgp / 5,450 lbf) thrust, and an "all-moving tail" or "all-flying tail" with hydraulic actuators. In the F-86A, the tailplane was fixed, but with the all-flying tail the entire tailplane assembly could be rotated. The moving elevator was retained. The all-flying tail was introduced because the original tailplane became ineffective at transonic speeds, with pilots having difficulty pulling out of high-speed dives. The all-flying tail provided much better control authority at transonic speeds. However, the hydraulic actuation system meant that the elevators no longer provided any force feedback to the pilot. As a result, NAA engineers devised an "artificial feel" mechanism, using bungee connections and counterweights.

The first F-86E flew on 23 September 1950, once again with George Welch at the controls. The first production batch of 60 "F-86E-1s" began to flow to fighter groups in early 1951, replacing war-weary F-86As in combat beginning in the spring. The next batch was of 51 "F-86E-5s", which were identical except for some tweaky changes in switches and the like.

The third batch of 60 "F-86E-6s" was actually built by Canadair of Canada. Canadair had signed a license agreement with NAA to build the Sabre, and the USAF took advantage of the arrangement to use Canadair as a second source. The F-86E-6, or "F-86E-6-CAN" as it was sometimes called, was essentially a Canadair Sabre Mark 2, which is discussed in the next chapter.

There wasn't supposed to be a fourth batch of F-86Es, since production was planned to move on to the "F-86F" with the uprated J47-GE-27 engine with kN 27.1 kN (2,760 kgp / 6,090 lbf) thrust. However, GE couldn't deliver the new engine in time, so NAA went ahead and fitted the F-86F airframe with the current J47-GE-13 engine. The result was the "F-86E-10", and 132 were built beginning in September 1951. Since they had been designed to accommodate the J47-GE-27 engine, some were retrofitted with the more powerful engine in the field. There would actually be another unplanned E-series variant, to be discussed later.

The first proper "F-86F-1" with the J47-GE-27 engine flew in March 1952. The new engine fit provided a significant improvement in performance, particularly in terms of rate of climb. 78 were built, with operational deliveries beginning in June 1952.

The F-86F-1 was followed by 16 "F-86F-5s", which featured a stronger pylon attachment point on each wing to allow the fighter to carry 758 liter (200 US gallon) drop tanks, providing extended range and combat endurance. The next production batch consisted of 34 "F-86F-10s" with the new A-4 gunsight, replacing the less reliable A-1CM gunsight.

The "F-86F-15" featured a redesign to the hydraulics system, which had proven vulnerable to battle damage, with a single hit capable of disabling the aircraft. However, only seven F-86F-15s were built because of problems with delivery of the J47-GE-27 engine. As a result, NAA then completed 100 of the variant with the J47-GE-13 engine, redesignating it the "F-86E-15". This was the last of 404 F-86Es, including the prototype.

The F-86F would prove to be the definitive production variant of the Sabre. By this time, the refinements to the Sabre had produced a fighter that was termed "viceless" by the men who flew it. For example, the Sabre had to be forced into a spin and could be returned to normal flight simply by neutralizing the controls.

The demand for F-86Fs was so great that NAA couldn't keep up with production at the Inglewood, California, plant, and so the company leased a factory from Curtiss in Columbus, Ohio, that had been used to build Helldiver dive-bombers but had been shuttered since the end of World War II. The NAA Columbus plant began Sabre production with 100 "F-86F-20s", the first flying in March 1952. The F-86F-20 was almost identical to the F-86F-15s from Inglewood, differing only in minor changes to radio fit. The USAF gave Inglewood Sabres an "NA" suffix and Columbus Sabres an "NH" suffix, for example "F-86F-15-NA" versus "F-86F-20-NH".

The next improvement in the F-86F was to make it a better fighter-bomber by adding an extra set of stores pylons, one at the midpoint of each wing, for a total of four. This "dual store" wing allowed the Sabre to carry two drop tanks and two 450 kilogram (1,000 pound) bombs. 859 of this subvariant were built by Inglewood as the "F-86F-30-NA", while 600 were built by Columbus as the "F-86F-25-NH". The first was rolled out in October 1952, and they were rushed to Korea to replace outmatched North American F-51 Mustangs and Lockheed F-80 Shooting Stars in close-support units.

North American had only begun tinkering with the wing, however. Aerodynamic studies had shown that eliminating the leading-edge slat and providing a sharper leading edge on the wing provided substantial benefits in high-speed flight, though this led to "hotter" landing characteristics. To compensate for the hot landings, the wing was extended 15 centimeters (6 inches) at the root and 7.6 centimeters (3 inches) at the wingtip. The new wing, which also featured fuel tanks in the leading edge that increased capacity by 264 liters (70 US gallons), was known as the "solid 6-3" wing. The wing was incorporated into F-86F-25/30 production in midstream, and was also sent to Korea in kit form to upgrade most F-86Fs and a few F-86Es in service there.

Once the USAF got the new and better F-86F into the skies over Korea, the service decided to demonstrate the superiority of the Sabre once and for all. In the summer of 1953 the USAF had a "turkey shoot" against Communist pilots that proved once and for all who was boss. The final tally for the entire war was estimated by the USAF as 800 MiGs lost to Sabres, with 80 Sabres lost to MiGs, a 10:1 kill ratio. 30 other Sabres were lost in combat, and 114 more were lost to accidents. Even assuming that the MiG losses were exaggerated by a factor of two, that still made the Sabre the clear winner of the contest. If the war on the ground ended in a stalemate, the USAF had reason to feel pleased with themselves in the air.

Although refinements of the aircraft helped, it was still pilot training that gave the Americans the edge. Of the 40 American aces in Korea, 39 flew the F-86. Many of them had extensive World War II experience, and these 39 men tallied over 300 of the 800 MiG-15s claimed destroyed. The top-ranking aces in the war were Captain Joseph D. McConnell JR, with 16 kills; Major James Jabara, with 15 kills; and Captain Manuel "Pete" Fernandez with 14 kills. Jabara had actually been the first jet ace and was in competition with McConnell, but Jabara's eagerness got the better of him and his 16th kill turned out to be a Sabre. The pilot escaped unharmed, though a 12.7-millimeter bullet grazed him close enough to tear open his life jacket.

McConnell was killed in an F-86 crash in 1954. He had a "Hollywood" look to him, and would have a movie made about him with Alan Ladd in the starring role. Colonel Francis "Gabby" Gabreski, well-known P-47 Thunderbolt ace of World War II, obtained 6.5 kills in a Sabre over Korea, adding to his 31 kills against the Luftwaffe from the previous conflict.

Sabres operating in the ground-support role dropped roughly 8,000 tonnes (8,800 tons) of bombs and napalm and fired about 300 rockets. This was a modest quantity compared to the other aircraft assigned to the "mudfighter" role in the war.

Even as the war in Korea bogged down to its inconclusive end, North American was working on new subvariants of the F-86F. The "F-86F-35-NA" was similar to the F-86F-25, but incorporated a "Low Altitude Bombing System (LABS)" for delivery of a single 545 kilogram (1,200 pound) Mark 12 atomic bomb. LABS was used to perform targeting for a "toss bombing" attack, in which the Sabre went into a climb from low altitude, released the bomb to fly upward and then fall back down on the target, while the aircraft departed the area as fast as possible to be clear before detonation. 264 F-86F-35s were built, becoming nuclear operational in January 1954. They were deployed in NATO bases to help meet the Soviet threat to Europe.

The final production subvariant of the F-86F was the "F-86F-40", which was intended for Japanese use. Confronted by the Soviet Union, China, and North Korea in the Far East, the US finally decided to permit Japan to re-arm, though the new Japanese military services did not trace their ancestry back to Imperial Japanese units and were carefully described as "Self-Defense Forces". In fact, the Japanese adopted a clearly defensive military posture. After the disaster of the Pacific War, even they had misgivings about rebuilding their forces.

The Japanese Air Self-Defense Force (JASDF) was given the highest priority for bringing up to strength, and decided to adopt the F-86F as their first-line fighter. In July 1954, North American came to an agreement with Mitsubishi Heavy Industries of Japan, who had built the famed Zero fighter in the war, to co-produce a new version of the F-86F, the F-86F-40, to be designed by NAA and supplied to Mitsubishi in kit form for assembly.

The F-86F-40 was similar to an F-86F-25/30, but incorporated a modified wing, featuring a 30 centimeter (1 foot) wingtip extension on the "6-3" wing, plus restored leading-edge slats. This wing was logically known as the "F-40" wing. The first F-86F-40 was rolled out at the Inglewood plant in October 1955. Performance matched that of the F-86-25/30, while the F-86-40 had tamer landing characteristics. The initial batch of kits had already been sent to Japan in August, with the first JASDF F-86F-40 rolled out a year later. Mitsubishi would assemble a total of 300 F-86F-40s, with the last produced in 1961.

Late Japanese production would be equipped to carry the new "Sidewinder" heat-seeking air-to-air missile (AAM), which was first test-fired in 1952. This weapon was then designated "GAR-8" by the USAF and "AAM-N-7" by the US Navy, though the type is now designated "AIM-9". These first-generation Sidewinders needed to be accurately boresighted on the target's exhaust to be effective. They provided an audio signal to the pilot through his earphones, in the form of a growling tone that grew louder as the missile achieved lock for firing. Sidewinders would become a common store on many late-production Sabre fighters.

In the meantime, the USAF was in a desperate need for new fighters and also decided to buy F-86F-40s, with a total of 280 built by NAA, and in addition ordered a large number of "F-40" wing kits to retrofit to older F-86Fs. These wing kits were also supplied to many US allies who used the Sabre. The F-86F-40 was the final production variant of a total of 2,538 F-86Fs built, not counting some of the unusual variants described below.

Gunval Sabres / TF-86F Trainer

The lack of killing power of the Sabre's six 12.7 millimeter Brownings was apparent both to the Soviets and the USAF. As a result, the Air Force initiated a program named "Gunval", in which an evaluation batch of Sabres were fitted with four T-160 20 millimeter revolver-type cannon. The T-160 was a derivative of a wartime German Mauser design, developed by the US Springfield Armory and NAA, to be eventually produced by Pontiac as the M-39 and fitted to the F-101 VooDoo and F-100 Super Sabre.

A Sabre nose section was pulled out of production in early 1952 and used to test the gun fit. The nose assembly required considerable redesign to accommodate the four cannon, not just because of their different configuration and size, but because the recoil shock was more substantial than for the Brownings, straining the gun mounts. One change that would prove significant is that to vent gun blast gases, small ports were installed in the inside of the air intake.

Ten Gunval Sabres were built and given the designation "F-86F-2". They were fitted with the four T-160 cannon, with 100 rounds per gun, with all four guns firing simultaneously. George Welch performed the test firings and all seemed to be well. Eight Gunval Sabres were shipped to Japan in late 1952 and went into combat evaluation in early 1953.

Problems quickly cropped up. One Gunval Sabre suffered a flameout while firing on a MiG-15, though the American pilot was able to drop to lower altitude and relight. However, a few days later another Gunval Sabre pilot had the same problem and had to eject. The Gunval Sabres were grounded. It turned out that Welch had performed the test firings at medium altitude, about 6 kilometers (20,000 feet), while air combat in Korea generally took place at much higher altitudes. In the thin air, blast gases vented into the air intake choked off air to the engine, causing a flameout. The gun ports inside the air intake were welded shut, and a selector switch was installed to allow the pilot to fire two or four cannon as desired.

This helped considerably, but it still wasn't perfect. Another Gunval Sabre was lost to an engine flameout, the pilot again ejecting safely. It turned out that the blast gases were accumulating in front of the intake, choking off airflow. An NAA field service representative devised a horseshoe-shaped clip that was installed in the muzzle troughs of the cannon to divert the blast gases away from the nose of the aircraft, finally fixing the problem.

The combat evaluation was finished in May 1953, with generally satisfactory results once the fix was in place, and the aircraft were returned to the US. Two other Sabres were fitted with four Oerlikon 206RK 20 millimeter cannon, redesignated "F-86F-3", and put through evaluation in 1954, but the Oerlikon cannon were heavier than the T-160s, making the aircraft nose-heavy, and were also not as reliable.

Following a USAF request for a high-performance trainer to replace the Lockheed T-33 Shooting Star, in 1953 an F-86F was pulled from the production line for conversion to a two-seat "TF-86F" operational trainer prototypes. Initial flight was on 14 December 1953, with NAA test pilot Ray Morris at the controls. The TF-86F featured a 1.6-meter (5 feet 3 inch) fuselage stretch and tandem seating, with a single piece cockpit canopy that hinged up from the rear in a "clamshell" fashion, in contrast to the sliding canopy of most of the fighter F-86 variants. The stretch unsurprisingly shifted the aircraft's center of gravity, and so the wing was moved forward 20 centimeters (8 inches).

The TF-86F had short slatted wings and additional fuel tankage. No armament or gunsight were fitted. The result was that the TF-86F only weighed about 45 kilograms (100 pounds) more than a stock F-86F and had very similar performance. Unfortunately, the TF-86F crashed on 17 March 1954, killing test pilot Joe Lynch. The Air Force ordered a second TF-86F conversion a few days later, with the aircraft making its first flight on 5 August 1954. The second TF-86F differed from the first in that it had the "F-40" long slatted wing, two 12.7-millimeter machine guns with 100 rounds per gun, and a small ventral fin under the tail. The ventral fin was added as a fix to the flight instability problem that was believed to have destroyed the first aircraft.

The second TF-86F was delivered to the Fighter School at Nellis AFB on 31 January 1955, but the program was cancelled a week later, with the Air Force settling on the TF-100 Super Sabre instead. However, the second TF-86F had a long and productive life, operating at Nellis for several years as an operational trainer for Sabre pilots, and then as a chase plane at Edwards Air Force Base well into the 1960s.

Author: Greg Goebel