American airmen recognized that the MiG-15 had a number of performance advantages over the F-86 due to its equivalent engine power and lighter weight. To increase the fighter’s flying performance, they pushed to reduce weight, increase engine power, and lessen drag. To improve fighting performance, the airmen attempted to increase the lethality of the Sabre’s armament.
Clearly the MiG was lighter and smaller than the F-86. Obviously, if the Sabre’s weight could be reduced, its performance would improve; the question was: what to remove? A complication was that even if weight reductions could be achieved by removing equipment or subsystems, the size and weight of the existing airframe could not be changed. North American proposed a scheme that would reduce the fighter’s weight by 2,800 pounds; however, as this called for a virtual redesign of the aircraft, this concept was dropped. Another factor that made the Sabre larger and heavier than the MiG was its designed range: the F-86 carried 17 percent more fuel internally than the MiG, which had been designed as a short-range interceptor. Certainly, the Air Force did not want to reduce range; if anything, it wanted to increase it. In the end, the North American fighter may have carried too much equipment, which made it heavier than the MiG; but the MiG probably carried too little equipment, which resulted in a better flying machine but not a better fighting machine.1
In October 1951 Col. Benjamin Preston, Commander of the 4th Fighter Group, filed an unsatisfactory report on the F-86E’s General Electric J47-13 engine. He wanted more power, recommending an engine with a minimum of 6,500 to 7,000 pounds of thrust. The USAF did have a better, but not much better, engine. The F-86F used the more powerful J47-GE-27 engine, which produced 770 more pounds of thrust (15 percent greater) than the J47-GE-13 used in the F-86A and F-86E. (Empty weight of the F-86F increased 4 percent, eating up some of that advantage.)2
The airmen wanted more power; however, the development of a bigger engine would take time. Therefore, on 11 January 1952, Gen. Earle Partridge, former commander of Fifth Air Force and now commander of Air Research and Development Command, gave priority to the task of improving Sabre performance.
The airmen were even willing to sacrifice engine life for this performance improvement unless it caused “an unsupportable burden on supply and maintenance.” During the war, Wright Air Development Center (WADC) initiated and tested five projects to increase engine power: over-temperature, water-alcohol injection, solid rocket booster, pre-turbine injection, and liquid rocket booster.3
The airmen realized that running the engine at higher temperatures would increase thrust but would also lessen engine life. It was relatively easy to affect the higher temperatures by inserting segments (called “rats” and “mice” in the field4) into the tailpipe, reducing the exhaust nozzle area. In addition, they placed a detent on the throttle control beyond which the engine operated at higher than normal rpm in the over-temperature range. The airmen attached a timer to record the duration of over-temperature operations, which they wanted to limit to one-minute cycles to get the maximum life out of the turbine blades. Tests of the concept demonstrated increased thrust, averaging 27 percent between fifteen thousand and forty-five thousand feet, with a maximum increase of almost 30 percent at thirty-five thousand feet. Consequently, performance markedly improved. Nevertheless, the system met considerable resistance at GE and WADC because of shortened engine life and supply and maintenance issues. The USAF canceled the project prior to September 1952.5
The water-alcohol injection program got its start in late 1951, when discussions began between General Electric and the Power Plant Laboratory at Wright Air Development Center. The idea was to increase engine power by increasing both the fluid weight flow passing through the engine as well as the exhaust gas temperature. Water would satisfy the first requirement but would sharply reduce the exhaust gas temperature, thereby nullifying any benefit. Adding alcohol to the water would accomplish both objectives.
The Air Force planned to test the concept in a B-45, but delays with the aircraft prompted the airmen to switch the tests directly to an F-86 in April 1952. The USAF ran flight tests with an F-86A between June and early October 1952 and with an F-86F between late October and late January 1953. Although the results were encouraging, two problems emerged. First, there were breakdowns of the injection pump. More serious were flameouts in climbs at high altitude. Although adjustment to the mixture solved the problem, it also reduced the augmentation gain.
The testers concluded that the water-alcohol injection could increase engine thrust by 20 percent at twenty thousand feet to 30 percent at forty thousand feet. If performance above forty thousand feet was desired, the mixture had to be adjusted to prevent the flameout problem, thereby reducing the gain to 15 percent at twenty thousand feet and 20 percent at forty thousand feet. The downside was that the water-alcohol injection reduced the life of normal combustion system components on the J47-27 engine to two hours. There was also the problem of where to store the mixture; trading fuel for mixture would reduce the fighter’s range, obviously not a viable solution when range was already a problem. Another difficulty was the realization that the device could not be tested for combat until late summer or fall 1953.6
The Air Force came up with a relatively lightweight and simple means to boost engine power called pre-turbine injection (PTI). PTI injected fuel into the engine forward of the turbine wheel and then burned it aft of the turbine wheel. The system also used a variable tailpipe nozzle that adjusted up to 15 percent of the area of the exhaust. This arrangement, in essence a very short afterburner, was necessary because of the space limitations on the F-86 and the requirement for minimal modifications.7 PTI flight tests began in 1953 and indicated an increase of 40 percent in thrust. Although this gave the F-86 much better performance, the price was considerable wear on the engine, as turbine blades burned up after one hour of operations, and turbine wheels after three.8 There were also problems with the nozzle binding. Nevertheless, one of the testers “concluded that the increase in performance of the F-86F due to PTI installation is of such a magnitude as to warrant further exploration of the system.”9 This was not to be, for despite considerable testing, PTI was not used; instead U.S. aircraft were (later) fitted with afterburners.
Another proposal to increase power was using solid fuel rocket boosters, with the misnamed JATO (jet assisted takeoff) bottles. In the summer of 1952, the Air Force flight-tested an arrangement that consisted of three rocket bottles mounted in an insulated compartment in the aircraft’s aft fuselage, directly below the speed brakes. Each unit produced one thousand pounds of thrust and could be fired individually in sequence (for thirty-eight seconds, increasing speed 15 kts) or simultaneously (for fourteen seconds, increasing speed 20 kts) at forty thousand feet.10
The USAF sent the rockets to the theater in August 1952 and installed them on six F-86Fs in the 334th Fighter Squadron, some with and some without the new leading edge. The rocket-equipped Sabres flew 136 combat missions and engaged in sixteen dogfights, during which they destroyed six MiGs. However, the rockets were only helpful in two of the credits, because one of these victories was accomplished without the rocket assist; and in three other kills the circumstances did not require the rocket boost. Aside from the fact that they were only minimally helpful, the rocket units added extra weight (600 pounds full and 450 pounds empty) and reduced maneuverability. On two occasions this loss of maneuverability prevented the F-86s from engaging the MiGs. In addition, the Air Force lost two of the modified F-86s. It should be noted that the test was not a fair comparison with standard F-86s: the tests employed highly experienced pilots who were given lead positions. Both of these factors enhanced the chances of success. In addition, the installation moved the center of gravity aft, creating a longitudinal instability problem that caused stick reversal (the aircraft moving in the opposite direction of that intended) under high “g” conditions and porpoising. Although maintenance problems were minimal, the airmen determined that the rocket-equipped aircraft required a turnaround time of three hours, compared with the normal one-half hour. The pilots emphasized the critical importance of timing in firing the rockets, the short duration of added power, and the betrayal of their position by the smoke trail. In brief, the pilots believed the rockets were of little value. Fifth Air Force received authorization to terminate the project in late November 1952.11
The Air Force also worked with a liquid fuel rocket booster that could be shut down and restarted. A study in early 1952 investigated an engine with thrust between one thousand and five thousand pounds, firing between 24 and 120 seconds. However, the installation would add weight and drag and thus reduce both speed and range when not operating. The study concluded that a small rocket motor of one thousand pounds thrust and 120 second duration was best suited for combat but went on to note that afterburning was a far superior approach.12
Nevertheless, the liquid rocket boost program continued, but slowly. It was not until January 1953 that the Air Force contracted with Aerojet to design, develop, test, and deliver a rocket of thirty-five hundred pound thrust (minimum), with a duration of 108 seconds, and capable of six starts. The Aerojet engine was not tested until fall 1956. Four years later, the Air Force tested a Rocketdyne rocket engine. None of these engines went into service on an F-86.13
The Sabre’s six .50-caliber machine guns, used so successfully during World War II, proved inadequate during the Korean War.14 Although some pilots liked this arrangement (the six guns provided redundancy, were a known quantity, were reliable, and had a high rate of fire that gave a shotgun effect), the Sabre’s guns did not get the same results over MiG Alley as AAF fighters had in World War II. There were twice as many multiple kills per sortie in World War II as in Korea. (A number of pilots in the earlier war scored three or more kills on a single sortie while only one Korean War pilot scored more than two credits.) Whereas in World War II fighter combat over Europe 71 percent of the total Eighth and Fifteenth Air Forces’ air-to-air claims were assessed as destroyed, that figure was 44 percent in Korea.15 Numerous American pilots “fired out” (fired their entire load of ammunition) without achieving a kill. Evgeny Pepelyaev, the top Soviet ace of the Korean War with twenty-three credits, dismissed the American .50s as being like mere peas and told of MiGs returning to base with forty or fifty hits.16
A number of factors account for this reduced effectiveness. Jet aircraft were less vulnerable than their prop predecessors: their engines were simpler; they lacked the large prop; and their airframes were more rugged to endure the higher speeds and altitudes of their expanded flight envelope. The MiG-15 was a small, simple, rugged aircraft that carried no fuel in its wings. As a result, it had few vulnerable areas.17 Jet fuel was less volatile than the high octane gas used in piston-powered aircraft, and the .50 API-ammunition (armor-piercing incendiary), which had proved to be so devastating in World War II, did not have oxygen to burn above thirty-five thousand feet, where most of jet combat took place. (Fighter combat in World War II took place at lower altitudes.) In short, armament had not kept pace with the expanded nature of jet fighter air combat.18
A May 1951 RAF report of their tests of the F-86A rated it as having superior all-around performance compared to any British jet, but the testers noted armament as the Sabre’s most serious limitation. An initial USAF report from Korea in January 1951 stated that the “fire power of the F-86 is not sufficiently destructive.”19 Almost all American leaders and fighter pilots who discussed the F-86’s armament wanted more firepower. As one report in August 1951 put it, “The F-86 pilots universally would like a heavier caliber gun so that when a hit is made, a sure kill results.”20
The call for better armament came quickly from the field, but the Air Force response was slow. Within weeks of the first engagement between MiG and Sabre, Far East Air Forces (FEAF) forwarded a requirement for heavier armament, followed in March with a request for USAF trials of a 20 mm cannon on the F-86. In mid-June 1951, Headquarters USAF pushed for 20 mm guns in fighters at the earliest possible time. By the end of October 1952, the USAF had fired seven hundred thousand rounds in various tests, one-tenth of these from F-86s in a project code-named GunVal.21
The USAF had developed the T-160 20 mm gun (later known as the M39) from a German design. During World War II, Mauser built a five-chamber revolver-type weapon with a very high rate of fire. After the war, the concept was taken up in Britain, Switzerland, and the United States. After work by the Illinois Institute of Technology and the Springfield (Mass.) Arsenal, Ford became involved in 1950 and completed the gun in June 1951.22
The 20 mm T-160 was a much more powerful gun than the F-86’s .50-caliber M-3. The 20 mm projectile was almost two and a half times heavier, and its muzzle velocity was 15 percent greater, giving it a shorter time of flight, flatter trajectory, longer range, and greater impact. It also had a 50 percent higher rate of fire. USAF tests showed that the .50 was more accurate at ranges below six hundred yards, while the 20 mm was more accurate beyond that range. Therefore, it had a greater kill potential than the .50.23
There were, of course, negatives. The 20 mm gun was heavier and larger than the .50, so the USAF replaced the six .50s with four 20 mm guns, which meant that the two packages fired the same number of projectiles (about one hundred per second). Because the 20 mm ammunition was also larger and heavier than the .50s, only 460 rounds were carried, compared with 1,600 rounds of .50s, giving a total firing time of 4.6 seconds versus 16 seconds. The 20 mm armament hurt aircraft performance because of extra weight and increased deceleration due to the guns’ greater recoil.24 Pilots in the theater were disturbed by this loss of flying performance and believed that it was “unwise to weigh a possible advantage with a known penalty.”25 Another factor was that the F-86’s fire control system limited the guns and made the two installations essentially the same.26
The USAF mounted four T-160s in four F-86Es and six F-86Fs. In March 1952, North American’s George Welch tested the guns between ten thousand to twenty-five thousand feet off of Catalina Island without a problem. After further tests at Edwards Air Force Base, the Air Force sent eight aircraft to the 335th Fighter Squadron. Along the way the airmen painted two additional gun ports on the GunVal aircraft so that they looked like the other Sabres with which they would fly.27
During combat tests between mid-January and the end of April 1953, the GunVal fighters flew 284 combat sorties, during which their pilots sighted 139 MiGs, fired at 41, and claimed hits on 21, 6 of which were claimed as destroyed. Proportionally, this record exceeded that of the standard F-86s; but this was not an equivalent situation because these pilots flew at times and positions that gave them a better chance of success. More significantly, they were very experienced; among them were seven Korean War aces who flew ninety-six GunVal missions. The pilots who flew the GunVal aircraft averaged almost 2,900 flying hours each—2,100 hours in fighters—and were credited with fifty-eight victories in the Korean War and twenty-nine in World War II.28
The GunVal aircraft encountered two problems. The first involved the reliability of the guns. The airmen fired ninety-eight thousand rounds of ammunition during the tests, with 210 jams, approximately twelve times the malfunction rate of the .50s. (In fairness, the 20 mm was a brand-new gun while the .50 had been in service for some years and had fired millions of rounds.) As might be expected, the jam rate declined through the course of the tests, from 3.4 per one thousand rounds for the first thirty-two thousand rounds fired to 1.2 for the last fourteen thousand.29
A more serious problem was engine stalls. Excessive gun gases caused twenty compressor stalls on 363 sorties, six of these in combat, causing the loss of two aircraft. What was not a problem below twenty-five thousand feet stateside turned out to be a severe problem above that altitude over Korea. Air Force engineers determined that the four 20 mm guns expelled four times the amount of gas as did the six .50s.30 The airmen took a number of steps to solve the problem. Engineers at Wright-Patterson AFB proposed blast deflectors, but these failed in Korea. The airmen welded shut doors that bled air from the gun compartments into the engine intake duct and drilled holes in the gun-bay doors to vent the gases. The airmen also installed a selector switch that permitted the pilot to either fire two or all four guns. Firing only two guns at a time reduced gun gasses and gave the pilot twice the duration of fire. But this did not completely solve the problem. North American engineer Paul Peterson designed a horseshoe-shaped clip mounted in the gun panel port that broke up the gasses.31
The commander of the Air Proving Ground Command (APGC), Maj. Gen. Patrick Timberlake, concluded in an August 1953 report on GunVal that the project was unsuitable for combat due to the limited quantity of ammunition carried and the engine compressor stall problems. Although the Air Force persisted with the project, an April 1954 test report noted that the stall problems continued. It concluded that the available fire control systems limited the effectiveness of the 20 mm installation and allowed the .50s to achieve essentially the same results. In brief, the 20 mm installation did not provide the “desired degree of improvement over the M-3 [.50-caliber].”32 But the Air Force had already made this decision. The T-160, now known as the M39, would arm the next air superiority version of the Sabre, the F-86H.33
It is true that many of the concepts the USAF tried did not work out, at least in time to see service in the Korean War. Besides technical barriers, there were problems with getting new ideas through the bureaucracy and, of course, time constraints. The Air Force did improve the F-86 over the years, sometimes in small ways, other times more dramatically. The USAF designated these major changes by adding a suffix to the F-86 designation; in this case, the daylight air superiority F-86A evolved into the F-86E and F-86F.34 (The Air Force also instituted changes within each suffix letter series known as block changes, usually in sequence of fives, e.g., F-86F-1, F-5.)
North American built 554 F-86As. It first flew in May 1948 and entered service in February 1949; the last was accepted in December 1950. The manufacturer mounted Mk 18 (optical, computing) gunsights into all but the last 24 F-86As. The builder put the A-1CM with a radar ranging device (AN/APG-5C or more commonly the AN/APG-30) from then on and retrofitted the rest of the series. The first 33 Sabres (the entire F-86A-1 block) had flush gun doors that opened when the guns fired and closed when they stopped, but these were removed for practical reasons, despite the loss of aesthetics and performance. The A-5 series also replaced the curved windshield with a “V” shaped one. North American moved the canopy jettison-release handle from the bottom forward cockpit panel to the right handgrip. The Air Force made other changes partway through the A-5 block, the most important of which was to reconfigure the wing slats to open and close at lower air speeds. This modification also allowed the builder to remove the slat locks (one less task for the pilot) and the stick shaker from the aircraft. Another improvement was to change the fuel flow meter reading from gallons to pounds and, later in the block, adding a totalizer that gave the pounds of fuel remaining.35 The accurate Sabre fuel gauge proved very important in combat over MiG Alley.
North American built 456 F-86Es. This variant first flew in September 1950 and went into service in May 1951. The F-86E-1 weighed 460 pounds more than the F-86A-5 but was powered by the same J47-GE-13. The main difference between the two aircraft was a new and much more effective flight control system. The “E” replaced the conventional, hydraulically boosted aileron and elevator control surfaces with a system of full-powered hydraulics. In addition, the entire horizontal stabilizer moved, not just the elevator. NACA had tested the all-flying tail starting in 1943 and incorporated it into the Bell X-1; North American began testing the concept on the F-86A in mid-1949. Test pilots reported that the flying tail system was far superior to conventional controls above Mach .85. The “E” was a good gun platform throughout its speed range (up to Mach 1) and had positive control above 500 kts where the “A” had only marginal control. The result of these changes to the control system was to make the already fine handling Sabre a much better flying machine at a small cost in performance due to the additional weight. Another change was to switch the flaps from a hydraulic system to an electrical one.36 The new flight control system was one of three major changes to the F-86 that greatly improved its fighting capability.
The USAF intended the “F” model to be a fighter-bomber, featuring three major changes to the F-86E: a more powerful engine, larger drop tanks, and improved armament (four 20 mm guns). The USAF planned to add the new guns and engine at the earliest possible point in production, but Headquarters Air Force revised this notion in late May 1951. The F-86F first flew in March 1952 and went into service a few months later. It weighed the same as the last blocks of the “E” series but was powered by the more powerful J47-GE-27 engine, which produced 12 percent greater thrust than did the J47-GE-13. The second major change was to use larger drop tanks. The first seventy-nine F-86Fs used 120-gallon drop tanks, as did the earlier F-86s; however succeeding F-86Fs could use either 120- or 200-gallon tanks. The USAF also added another pair of pylons to the existing two, enabling the F-86F-25 and F-86F-30 aircraft to carry four drops tanks or, more commonly when flying as a fighter-bomber, two drop tanks and two bombs. The 20 mm guns had some developmental problems and did not see service until the F-86H arrived after the Korean War.37 The increased power was a major factor in improving the performance of the F-86F. However, even more important was the installation of a new wing.
The third major improvement to F-86s was a wing modification. The existing wing with the automatic leading edge slats made the Sabre relatively docile at lower air speeds but penalized the fighter’s performance at higher altitudes and speeds. In 1951 North American experimented with an F-86E fitted with a larger wing without slats. It appears that North American originated the idea, most likely spurred on by their chief test pilot George Welch. In any event, at the end of the year, John Meyer (former commander of the 4th Fighter Group who had downed twenty-four German aircraft in World War II and two MiGs in Korea) flew the modified aircraft and was very much impressed; he therefore endorsed it with enthusiasm. Tests at Wright-Patterson and Edwards were also positive. Fifth Air Force learned of the work and in April requested combat testing.
The new wing was modified in two major ways. First it was slightly larger, six inches at the fuselage and three inches at the tip (hence the name “6-3” wing), increasing the wing area from 288 square feet to 302 square feet. Second, the slats were deleted. The problem of “tip stalling” that gave the aircraft a tendency to roll as it approached the stall was addressed by adding a small wing fence (five inches tall and extending sixteen inches from the leading edge) to each upper wing.38
The 51st Fighter Interceptor Wing received the first aircraft fitted with the 6-3 wing. In late August 1952 the unit reported on tests with the three modified Sabres (a fourth was wrecked in a crash landing) that flew for thirty-four hours, albeit without encountering MiGs. The pilots noted the improved performance: increased level speed (4 to 6 kts), dive speed (up to Mach 1.05), ceiling (four thousand feet), rate of climb (three hundred feet per minute), and range. The aircraft also turned tighter especially above Mach .85. There were no adverse handling problems.39 On the other hand, higher stall speeds forced an increase in approach, landing, and touchdown speeds (about 10 to 14 kts), and reduced drag necessitated a flatter approach. This called for more attention in the landing phase, but as one pilot put it so well, “Since the air war is not won in the traffic pattern these characteristics cause little concern.”40 Because the modification was cheap, simple, and effective, it was quickly adopted.41 In September, Fifth Air Force requested kits to retrofit all of its F-86E and F-86F aircraft. The change was also instituted on the production line in the midst of the F-86F-25 and -30 blocks.42 The F-86F, with its more powerful engine and fitted with the 6-3 wing, was equivalent—and in some respects superior to—its opposition. So, although much effort was expended for the few modifications enacted, the net result improved the F-86’s flying and fighting performance.43
One of the major advantages American pilots had over their Communist adversaries was better auxiliary equipment. This included flight helmets, “g” suits, and better cockpit defrosting, but probably the most significant of these advantages was a superior gunsight. (The Communists validated the importance of these devices by making special efforts to capture and analyze American “g” suits and gunsights.)44
Hitting a target that is maneuvering in three dimensions from an aircraft that is likewise moving in three dimensions is a difficult feat. A number of factors must be satisfied to score hits. One element is range. Two aspects are important. Since machine-gun bullets and cannon shells are limited to a finite distance, it is futile to fire them beyond a certain range. Moreover, since the projectiles follow a parabolic rather than straight-line trajectory, the gunner, except for short-range shots, has to compensate for bullet fall. A second element is deflection. Unless the attacker and target are flying in trail formation, the pursuer has to maneuver to fire his guns ahead of a turning target. The gunsight is the device to aid the fighter pilot in solving these problems.
Fighter gunsights used through the late 1930s were of the “ring and bead” type. This type consisted of a small bead at the muzzle and concentric rings near the breach to aid the pilot in aiming ahead of his maneuvering target. Because the system was very crude, only an experienced (or lucky) pilot could score hits that required deflection. As a result, the majority of kills were either stern attacks or those achieved at relatively short ranges.
The reflector sight was a major improvement over the ring and bead type. The pilot set the span of his target’s wings into the sight, which used lights and mirrors to project (or reflect, thus the term “reflector sight”) a center dot and circle on a transparent glass screen set on the top of the instrument panel. With a hand control on the throttle, the pilot adjusted the size of the circle so that it encompassed the target, thereby allowing the sight to calculate the range. This solved part of the gunnery problem.
The Germans first used the reflector sight during aerial combat in 1918. At the onset of World War II, all the major air forces either had reflector sights in their fighters or were installing them. But the breakthrough came when the reflector sight was mated with a gyroscope and a calculator that allowed the device to indicate where the pilot had to fire. The British began development of calculating sights in 1939, and they were the first, in 1944, to field them in both a turret and fighter version. The American copies were designated K-14 (USAAF) and Mark (Mk) 18 (U.S. Navy).45
The new gunsights met resistance from the fighter pilots. It was mounted directly in front of the pilot, creating a definite hazard to the pilot’s face on crash landings. Secondly, the pilot had to track the target for a second or so for the gyros to be effective. Perhaps most of all, it was new and different to pilots already familiar with existing equipment. However, the results could not to be denied. A Spitfire unit equipped with fixed sights scored kills in 26 percent of its combats, while another unit with the new sight claimed victory credits in 50 percent of its encounters. Another source states that pilots using the new sight scored hits five times greater, and kills three times greater, than pilots using existing sights. One reason for this success was that the sight achieved hits at greater ranges (as far as six hundred yards) and deflection angles (some over 50 degrees) than the older sight. The computing sight made average and inexperienced pilots good shots, especially when deflection was needed, whereas only a few pilots were able to master the older equipment.46
The next innovation linked radar to the gunsight. The British experimented with a radar-directed tail gun and installed it into a few Lancasters in 1944. In February 1943 the AAF had started a similar project for the B-24 but later fitted the equipment (APG-15) on the night-bombing variant of the Boeing Superfortress (B-29B). It performed well in tests but was a disappointment in combat.47
During the war, the AAF also developed a radar gunsight for fighters that was credited to Col. Leigh Davis and Dr. Stark Draper of MIT. The AAF conducted tests of the A-1 radar gunsight fitted in a P-38 against a banner towed by a B-26, and later with an F-84, and achieved excellent results.48 Nevertheless, North American fitted the initial Sabres (F-86As) with its modification of the tried and true Mk 18 gunsight. (There is no explanation why they picked the Navy gunsight over its AAF sibling, the K-14.) The Mk 18 had a number of limitations. It was designed to operate against slower moving, propeller-powered aircraft and had problems with excessive vibration (from the firing of guns). After a few months of combat, 4th Fighter Interceptor Wing personnel complained of the sight’s inaccessibility and the poor operation of the range control system, concluding that this “seriously hampered the effectiveness of the sight under combat conditions.”49 In short order, the USAF began to replace the Mk 18 with another gunsight.50
In March 1951, the USAF started to fit the F-86 with the A-1C sight, complete with automatic radar ranging. North American fitted the radar into the nose above the intake, giving the Sabre its “pouting upper lip” appearance. Gen. George Stratemeyer ordered the retrofit of the entire F-86 fleet with the new equipment, but the process was slow. By mid-June the USAF had modified seventy-six F-86As and estimated it would have four hundred completed by January 1952.51
The new sight—more correctly the additional radar rangefinder (APG-30)—improved accuracy at long ranges, allowing hits out to a range between twenty-five hundred to three thousand feet at the outer limits of the .50-caliber guns. The pilots needed the equipment because they consistently underestimated range and deflection—as have pilots throughout history. A 1952 USAF report made clear that, “despite popular opinion to the contrary, the bulk of the firing in combat is conducted at ranges and angles of in excess of 1,500 feet [range] and 10 degrees [deflection].”52
The new gunsight gave the American pilots a great advantage—that is, when it worked. Unfortunately, a combination of inadequate supply of parts, lack of test equipment, and—perhaps most of all—poor training of both pilots and maintenance crews led to malfunctions. A problem with the power supply caused inaccurate tracking. The sight proved fragile, especially in hard landings with the gunsight uncaged. The equipment also required extensive maintenance, prompting the Fifth Air Force director of operations to state that it was too difficult to maintain. In addition, the system had problems such as breaking radar lock at low angles off or extreme close range and having no way to check its accuracy while airborne. It performed erratically in clouds (due to moisture) and below six thousand feet (due to ground clutter). It was little wonder, then, that pilots came to distrust the system and some saw it as a useless 205 pounds of extra weight.53 In brief, the USAF had fielded the system prematurely.
Despite all these problems, the USAF did not take clear action until the spring of 1952. In March a team of Air Force and company technicians was sent to the theater to work out the supply and maintenance issues, as well as replace the radar rangefinder with a new unit. Project Jay Bird finished up in July after retrofitting 159 F-86s.54 Besides improving the gunsight’s reliability, they did make one significant modification.
Jay Bird added a feature that addressed the gunsight’s oversensitivity at long range. The original sight worked well at long ranges against non-maneuvering bombers but not against maneuvering fighters. The USAF’s APGC developed a device that became known as the Jenkins Limiter that reduced the sensitivity at long range, greatly increasing the tracking ability at these distances. The Jenkins Limiter also gave the pilot a visual indication that the target was within range and that the gunsight was working. As one official document put it, it was successful “from the very first.”55
At the same time, supply and maintenance issues caused major concerns. Fifth Air Force revised a number of supply and maintenance policies. For example, units were required to have a fifteen-day stock level at operating units. Critical items of radar and test equipment were to be hand carried to the theater in order to arrive in a timely manner. By April these issues had been resolved, and Headquarters Fifth Air Force described the in-commission rate of the A-1CM/APG-30 systems as excellent.56
In October 1952 the USAF began to replace the A-1CM with the A-4 sight. The A-4 was a redesigned and improved A-1 that eliminated or reduced the A-1’s problems and, from the outset, was judged to be more reliable than its predecessor. Nevertheless, there were residual problems with the system’s inverters and with spare parts.57
Unfortunately by this time the radar ranging gunsight had earned a poor reputation. Returning pilots complained about the gunsight’s unreliability, high maintenance requirements, and excess weight; and some suggested that it was too sophisticated for the theater and that the simpler K-14 gunsight would be an improvement. In response to a Headquarters USAF query of 11 July 1952 regarding the A-1CM gunsight, the Fifth Air Force Headquarters staff talked with five experienced pilots who had scored victories over Korea. They wanted the automatic features of the A-1CM system and saw this sight as superior to the Mk 18. They correctly noted that, while superior pilots could do well with manually ranging sights, these pilots were the exception, not the rule. (One of the shortcomings of many USAF studies was the use of a highly experienced test pilot, rather than the average combat pilot, as a tester.) The pilots also concluded that the low number of deflection shots had more to do with the deficiencies of the armament than of the sighting system. Therefore, they recommended that the A-1CM be retained, provided that immediate efforts to improve reliability and maintenance were implemented. They further noted that the radar equipment’s weight increase of 205 pounds was insignificant relative to its advantages.58
Despite this information from FEAF and firing statistics assembled by the APGC, the chief of staff of the USAF, Hoyt Vandenberg, was not satisfied. This came to a head in the summer of 1952 when fourteen Korean War aces meet with him and “vigorously” recommended that the radar gunsight be removed from the F-86. They assured their chief of the extremely high percentage of kills at short range during a stern chase and the need for 100 percent reliability. Further, they complained about the 205 pounds of extra weight in a complex and unreliable gunsight that was carried only for the “rather remote chance of a long range, high deflection kill.”59 Typical of this attitude was a comment attributed to World War II ace (28 aerial credits) Francis Gabreski (who also downed 6.5 in Korea) that: “I just stick a piece of chewing gun on my windscreen and use that as a sight.”60 The tendency of the older pilots to dislike and distrust the new equipment is understandable considering their lack of training with the radar ranging gunsight system, their success with the tried and true older gunsight, and—most of all—their experience of problems with the new one. On 8 September, Vandenberg directed that both the APGC and FEAF compare the radar ranging A-1C sight with the manual ranging Mk 18 series in fighter-versus-fighter combat conditions.61
The APGC used two methods to fulfill its assignment. First, it conducted flying tests with three F-86Es equipped with K-14 sights and three F-86Es equipped with the J-2 fire control system.62 The Air Force assigned five Korean War aces, one Korean War veteran who was also a World War II ace, and two proving ground pilots to the tests. They flew 307 gun camera passes on a towed banner, 191 of which were used in the study.63 This was the best the testers could do, but the results were questionable, since the target was non-maneuvering, slow, and had to be approached at angles greater than 15 degrees. The test report admitted that these conditions were “not realistic or comparable to fighter vs. fighter combat.”64
The testers noted that the K-14 weighed 205 pounds less than the radar ranging, fire control systems. However, because center of gravity considerations would require ballast to offset the weight of the K-14 in the nose, weight reduction could only be realized in future designs.65 The tests showed that the two sight systems achieved comparable results at short ranges, but the J-2 did much better at ranges in excess of twelve hundred feet. The pilots concluded that the A-4 was superior to the K-14 but that “the APG-30, despite the JAYBIRD fixes, is still not suitable for use in the F-86E due to its unreliability, limited performance at low altitudes, inadequacy against jet fighter targets, erratic performance in the presence of clouds, excessive maintenance and personnel requirements, etc.”66 They recommended that the APG-30 be deleted and the A-4 sight be used with manual ranging. Furthermore, they recommended an intensive effort be made to improve the APG-30 radar and that a new radar should be designed to improve maintenance and reliability.67
The APGC evaluators also reviewed Korean War gun camera film.68 They observed that World War II experience demonstrated that pilots grossly underestimated both range and angles off in engagements. They posited that the belief that fighter-to-fighter dogfights consisted of short range, minimum angle combat was reinforced “in the pilot’s mind by the experience in non-combat maneuvers and training, and in the non-flying public’s mind by films released for public consumption, showing close-in, highly effective attacks on enemy aircraft.”69 The selected film was of better pictures of targets (necessarily at shorter ranges) in which hits or damage could be seen. In fact, however, actual combat was at “considerably longer ranges” than indicated by either the APGC flying tests or selected Korean War gun camera film. Whereas 50 percent of the APGC test “firing” was at fifteen hundred feet and the selected film showed a range of sixteen hundred feet, the unselected (or unedited) film indicated an average range of three thousand feet. The evaluators divided the unselected film into three categories: pilots who scored two or more kills, those who scored one-half to two kills, and those who had no credits. They discovered that even the second group, which fired at shorter ranges than the other two, fired at longer ranges than that found in the APGC tests or unselected film. The tests indicated that the A-4 sight used in a manual mode was only 71 percent as effective as the complete J-2 system, and that the K-14 was 90 percent as effective as the A-4 in the manual mode and thus only 64 percent as effective as the J-2 system.70
Patrick Timberlake, the APGC commanding general, signed the overall report. He concurred with the comments on the APG-30’s reliability and maintenance problems but did not agree with the recommendation that it be removed from the F-86, for he noted the majority of the firing was done at long ranges with high angles off. Under these conditions, kill probability with the J-2 system using radar ranging was twice that of either the A-4 or K-14 using manual ranging. Therefore he recommended that the APG-30 be retained but efforts should be intensified to improve its reliability.71
As might be expected, FEAF’s response was less elaborate and elegant than that of the APGC. Three days after Vandenberg’s cable, Fifth Air Force held a seminar at its headquarters with nine officers, seven of whom had downed twenty-six aircraft on 707 missions by that time.72 Their conclusions did not vary from those put forth by FEAF in July. One view held in the theater, at least at the command level, was that the fourteen aces who influenced the chief of staff were probably not well-trained in the use of the radar ranging gunsight and had served in Korea when the gunsight reliability problem was at its worst. Those who had some training with the sight or familiarity with radar were less critical of it and saw its value.73 The FEAF seminar participants wanted all the automatic features of the A-1CM gunsight. Again, they mentioned that while superior pilots could do well with the Mk 18 manual ranging gunsight, such pilots were the exception and that the low deflection shots were a deficiency of armament not of the sights. The pilots reported that Project Jay Bird and the Jenkins Limiter had increased gunsight reliability and efficiency “immeasurably.” In brief, they wanted to retain the A-1CM sight, redesign it to ease maintenance, and reduce the numbers of technical personnel. They also believed that pilots should get comprehensive training in the use of the gunsight.74
In the end, the USAF retained both the radar rangefinder and its problems. For despite the Project Jay Bird, the gunsight’s reliability continued to dog the Sabre units. Between 1 January and 15 March 1953, the 4th Fighter Wing experienced malfunctions of the fire control system on 10 percent of its 2,860 sorties, despite the fact that the unit discovered and corrected 845 malfunctions on the ground. On the GunVal tests in 1953, in what we can assume were pampered aircraft compared to the other F-86s in Korea, gunsights failed on 8 percent of the missions. During the last months of the war (3 March through 27 July 1953) the ranging radar malfunctioned on 17 percent of the sorties. The older A-1CM had slightly fewer malfunctions than did the newer A-4. The aircraft equipped with the range limiter suffered malfunctions on 12.5 percent of the sorties while those without this equipment had a higher rate of 23.2 percent. (Perhaps the greater attention given to the new equipment accounts for the anomaly of additional and newer equipment combined with the same older equipment having better reliability than the same gunsight with less equipment.)75
Despite these difficulties, the radar ranging, computing gunsight gave American pilots an advantage over their foes, who used a sight equivalent to the World War II K-14. The poor marksmanship of the MiG pilots could be attributed to armament ill suited for fighter-to-fighter combat, inadequate training, and an outdated gunsight. John Meyer, the Fourth’s initial group commander in Korea, believed the gunsight gave good shooters a 25 percent advantage. Statistical analysis indicated that pilots using computing sights fired at longer ranges and fired 18 percent fewer shots per burst than those using fixed sights. Most importantly, gun camera film showed greater hit probability for the computing sights beyond three hundred feet. (A September 1952 study indicated that computing sights were 1.3 times as efficient as fixed sights.) The value of the range limiter was validated by figures that showed that while pilots scored hits in 42 percent of the bursts fired below the limiter setting, they scored only 18 percent in which the bursts were beyond the setting.76