F-86s in the Korean theater numbered less than a hundred until the second half of 1951 and then gradually increased to 352 in July 1953. By the end of the war, there were two wings of Sabres committed to air superiority (4th and 51st) and two wings to fighter-bomber duties (8th and 18th). Sheer numbers in the theater did not equate with the number of aircraft in combat; aircraft reliability, serviceability, supply, and ground support determined how many aircraft on the ramp made it into the air. One measure of aircraft reliability was the in-commission rate, or the percentage of aircraft on hand that were ready to fly. The F-86 averaged a 59 percent in-commission rate of fighters on hand during the war, almost the same as its two jet stable mates, the F-80 and F-84.1
Compared to the other two fighters, a greater percentage of Sabres were out of commission due to a shortage of parts. Some of these problems were caused by inefficient management and were easily solved. For example, in early 1952 the 4th Wing had a number of aircraft grounded because of a lack of replacement engines. The problem was that engines removed from the aircraft could not be shipped to Japan for overhaul because there was a lack of engine dollies, and the unit could only get replacement engines when a pulled engine was turned in, on a one-for-one replacement basis. When the 4th Group’s new commander, Walker Mahurin, took over in March 1952, there were eighteen engines waiting for dollies. He scouted around the base and found a dozen dollies in various places, as well as another half dozen still in packing crates.2
Were there any differences between the maintenance of the various F-86 units? For the ten months that we have comparable figures, the 4th had a higher in-commission rate in all except one month and a slightly higher average overall in-commission rate compared to the 51st.3 There was an intense rivalry between the two units, as might be expected from these young, aggressive fighter pilots who thirsted for promotion, bragging rights, and pride. One aspect of this was that the 4th posted a highly experienced supply sergeant to Japan who worked alongside the parts people there and personally tracked down parts requisitions from the 4th. He also put similar requisitions from the 51st at the bottom of the in basket. Another 4th tactic was more direct. Occasionally, fighters from the 51st short of fuel would land at the 4th base, because it was the closest U.S. airfield to MiG Alley. The 4th would use this opportunity to substitute their unusable parts for usable ones on these aircraft. This makes for a good story, but the fact is that the 51st had a lower percentage of aircraft out of commission for lack of parts than did the 4th in seven of ten months for which comparable figures exist. The 4th did beat down this rate from an average of 4.8 percent in the last half of 1952 to under 2 percent in May and June 1953.4
What of the Sabre’s various subsystems? While there were problems with the radar ranging gunsight, the F-86’s armament performed well. The Sabre’s tried and true .50-caliber machine guns ran up an impressive record and proved to be very reliable, reasonably accurate, and fairly lethal. During the course of the war, the 4th fired almost 25 million rounds of .50-caliber ammunition, averaging approximately 6,000 rounds per malfunction. Since the F-86 mounted six .50-caliber USAF machine guns and could carry a maximum ammunition load of 1,800 rounds, it is clear that gun stoppages were never a major problem. During the last half of 1952, the USAF attributed 60 percent of the gun failures to personnel.
Not all this ammunition was fired in combat. For example, only 37 percent of the ammunition expended by the 336th Fighter Squadron during the last half of 1952 and 79 percent in the first half of 1953 was in combat.5
There was one dramatic incident regarding the armament that ended in disaster for the Americans. On 20 May 1951, Capt. Max Weill (336FS) was leading a flight of four F-86s that engaged some MiG-15s. “At the beginning of the second burst,” Weill recalled, “I heard an explosion and saw several wild rounds of ammunition from the nose of my aircraft. Wild rounds were traveling 90 degrees to my line of flight.”6 The readings on his Mach meter, air speed indicator, and oxygen supply fell to zero, forcing Weill to break off the combat and return home. On his landing roll, he found he had no brakes because the hydraulics had been damaged, so he retracted the landing gear to stop the aircraft, slid off the runway, and totaled the fighter.
Three other fighters sustained similar damage. Capt. Milton Nelson (335FS) was leading a two-ship element when he engaged a MiG-15, which he destroyed. As he was returning to base, Nelson’s wingman noted a large hole on the Sabre’s left fuselage. Upon landing, his landing gear functioned properly although the indicator did not indicate that the nose wheel was down and locked. The left brake failed, however, due to damage to the hydraulic system; and Nelson attempted to ground loop the fighter to stop. Instead the fighter ran off the runway and was assessed to be beyond economic repair.7
Capt. James Roberts (335FS) had a similar, albeit less damaging, experience. As he fired three bursts at a MiG, he noted a large yellow flash just forward of the left side of the cockpit. He landed safely and found a large, jagged hole just forward of the lower left gun barrel.8 Inspection of the blast tubes of all aircraft found a fourth aircraft with internal blisters on two guns.
The accident board found that the Sabre’s ammunition (M-23 incendiary), which had a higher muzzle velocity than the standard M-8 ammunition, combined with inadequate cooling time between bursts, had exploded in the guns’ blast tubes (four guns in Nelson’s fighter, three guns in Weill’s, and one in Roberts’s). This was the first time the 4th had used the M-23 rounds and, understandably, the last. The incident had cost two F-86s.9
Other subsystems proved more difficult. The only evidence we have on the cause for aborts at the F-86 wing level covers the 4th Fighter Group over a seven month period, during which 24 percent of the aborts were caused by landing gear problems, 23 percent by engine problems, and 15 percent by radio problems. More extensive coverage from the 334th squadron supports these findings. The problems with the landing gear came as no surprise to those familiar with the F-86. One author explained that “the operation of the F-86 landing gear is dependent upon a multitude of micro switches, most of which are exposed to water and dirt splashed into the wheel wells . . . [that are] very difficult to check and can only be checked properly on a jacked-up aircraft.”10 The conditions on the Korean airfields was certainly much more primitive than those of the Sabres’ American bases.
This study found that a small number of pilots had a disproportionate number of aborts. Of the group’s seventy pilots who had aborts, a dozen (17 percent) had three or more aborts that accounted for 36 percent of the total aborts. One pilot had six, another five, and two had four aborts. But before blaming the pilots for the problem, note that a few aircraft had a disproportionate number of aborts. In fact, one third of the aborting aircraft had three or more aborts and accounted for two-thirds of the total aborts. One F-86 had five aborts and five others had four each. The pilot with six aborts (336FS) had three of these in the same aircraft on three different days, all caused by landing gear. All four aborts of another pilot (334FS) were caused by the landing gear in two different aircraft on four different days in June. Another pilot (336FS) aborted three times in two different aircraft, all due to inoperative radios. However, the pilot with five aborts (334FS) did so in five different aircraft, all to different causes. The study also noted that the F-86A had a much higher abort rate than did the F-86E, accounting for 45 percent of the aborts in the 336th although it comprised only 25 percent of the unit’s assigned aircraft. Because the “A” model aircraft were older and probably had accumulated more flying hours than the “E” model fighters, its greater maintenance is understandable.11
The F-86 Sabre experienced three major mechanical problems during the Korean War. These difficulties were quite different: one was very sophisticated; the second, routine; and the third, simple. The cutting-edge radar-ranging gunsight and the old and well-known landing gear have already been discussed. The third problem came from an even simpler device, the jettisonable, external fuel tanks known simply as “drop tanks.”
Range is one of the important elements of aircraft performance. While it is especially important for bombers and transports, fighter designers consider range after other performance aspects such as speed, ceiling, rate of climb, maneuverability, firepower, and acceleration. It is not that range is unimportant but that other aspects are more important in air-to-air combat. In short, to achieve greater range meant sacrificing other performance elements that are more important for fighter success.
Airmen understood the value of range early on and thought in terms of multi-place aircraft for escort duties. By the end of the 1920s, however, American airmen discussed the use of drop tanks to extend the range of fighters and tested a number of aircraft with droppable tanks.12 American airmen considered drop tanks only for ferry duties because of concerns of safety and reliability. In 1939 the Chief of the Air Corps banned tactical aircraft from using such tanks. In any event, top U.S. airmen questioned the effectiveness of fighters in war and believed instead that bombardment aviation “may be capable of effectively accomplishing its assigned mission without [fighter] support.”13 This led to the dogma taught at the top American air school of the day, the Air Corps Tactical School, that high-speed, heavily-armed bombers flying in tight formations, at high altitude, and without fighter escort could successfully defend themselves and attack critical enemy targets. The one American pre-war effort to develop an escort for American bombers was the abortive, twin-engine Bell FM-1.14 In fairness, it was not just Americans who went down this false path; aircraft designers worldwide believed that single-engine fighters would never be capable of long-range operations. The Germans went furthest, developing and mass-producing the twin-engine Me 110 for the escort role. While it performed well in a number of other ways, it failed as an escort because it could not stand up to single engine fighters.
Despite its major emphasis on bombers as its chief weapon and fighters designed as interceptors, in 1940 the Air Corps began to develop longer-range fighters. The air leaders realized that American fighters required longer ferry ranges if they were going to serve in European combat. In May 1941, Lockheed responded to an Air Force Materiel Command request with plans for 120-gallon tanks that would hang on bomb shackles. In September, the airmen modified the requirement by adding high-altitude escort duty as a fighter function. This change complicated matters, since it would require a boost pump to pressurize the tanks at operating altitudes. Nevertheless, by April 1942, Lockheed tested the tanks and used them to ferry P-38s across the Atlantic in mid-1942.15
World War II indicated that American fighters required greater range to close with and combat the enemy. Range can present major problems for the offensive fighter arm, as the Luftwaffe found out over Britain in 1940 and the AAF over Germany in 1943. The American airmen were slow in recognizing the need for escort fighters, and the ultimate solution (drop tanks) was hampered by managerial and technical difficulties: leaks, pressurization, shackles, jettison characteristics, and manufacturing. During the war, American fighters greatly expanded their escort radius of action: the P-47 increasing in a clean condition (without tanks) from 175 miles to 475 miles using two 150-gallon tanks; the P-38 to 585 miles with two 108-gallon tanks; and the P-51 from 475 miles in a clean configuration to 850 miles with two 108-gallon tanks.16
The AAF built more range into later versions of the P-47 and P-51. Efforts to develop other aircraft for this long-range escort role (the convoy defender YB-40, a heavily armed B-17) failed in action while the composite P-75 and twin-engine P-61 just could not perform as required. The twin-fuselage (from the P-51), twin-engine P-82 had the range and performance but arrived too late for combat.17
During the Korean War, American airmen had a number of problems with the F-86’s drop tanks. The war found the USAF unprepared for the large number of tanks required and their staggering cost. There were also technical problems with tanks either not releasing or damaging the fighter as they departed. This is surprising because the Air Force had considerable experience in World War II with drop tanks. In addition, the introduction of jets magnified the range problem because they were notorious fuel guzzlers. Therefore, the USAF neglect of the fighter range problem and drop tank solution is difficult to explain and even more difficult to excuse. In any event, drop tanks proved to be a considerable and continuing problem for the Sabre units.
The crunch came when the USAF decided that one F-86 Wing (the 4th) was inadequate to maintain air superiority and thus re-equipped the 51st Wing with the North American fighter. Drop tanks were already in short supply during the last half of 1951, so doubling the number of F-86 units pushed the situation to critical. Stateside production could not keep up with the new requirements. The Air Materiel Command history for the first half of 1952 noted that “The major item of tactical deficiency was the droppable fuel tank, for which requirements had jumped 500 percent in four months because of the increase in air opposition and the extraordinary fuel needs of the F-84 and F-86 jets.”18 In January 1952 the number of F-86 drop tanks in the field fell to a three-day supply (for all-out combat), triggering drastic measures.
One response to the shortage of tanks was to fly the F-86s with only one tank, beginning toward the end of January 1952. This reduced the Sabre’s range and endurance, cutting the coverage of MiG Alley and increasing the number of fighters returning home in a critical fuel state. Although such operations were feasible, they required changes for takeoffs. The pilots had to adjust the aileron and rudder trim and make additional use of brakes and nose wheel steering until they established sufficient rudder control. One exchange pilot noted that, while practical, “formation take-offs with single tanks hung on different sides tended . . . to be exciting.”19
Another effort to relieve this critical situation was to airlift drop tanks to the theater. Between 21 January and 4 February 1952, C-124s delivered twelve hundred tanks, a ten-day supply. In addition, by February 1952, every ship arriving at Yokohama carried some tanks. Nevertheless, the 4th was forced to limit the number of sorties it flew in February. It was not until May that the shortage of tanks was met. The problem returned in January 1953, however, when the 4th reported that its supply of drop tanks was “critically low.” Matters only grew worse when the USAF converted two existing fighter units to the F-86F fighter-bomber. In March the 4th used 3,531 tanks, the most it ever had used up to this point, and at the end of the month had 1,243 tanks on hand, about a ten-day supply.20
The original Sabre drop tanks had been designed by the North American engineers for ferry duty and were to be reused. Because of this and their high expense, the USAF authorized few spare tanks to the F-86 units. North American designed the shackle as part of the tank rather than part of the aircraft. This design was more streamlined than if the shackles were mounted on the airframe; however, it also meant that every time a tank was dropped, so too was a shackle. As one Fifth Air Force document explained, “these features caused expensive manufacturing processes and made the tank a luxury item.”21
Certainly the tanks were expensive, but they were necessary if UN aircraft were to operate over northern Korea. North American built both 120-and 200-gallon tanks. The first worked well but the second created a number of problems, such as damaging the aircraft after release and causing the aircraft to porpoise in flight. However, the greatest disadvantage of the North American tanks was their cost.22
Shipping the tanks was another problem because the bulky tanks took up much valuable shipping space and had to endure the rigors of the trans-Pacific journey. They were sent to the theater in three ways: completely assembled, partially assembled (fins, pylons, and fairings packed as loose items), and completely knocked down (“nested”). Tanks were sometimes damaged in shipping; one report stated that “a large percentage” was damaged even in shipping crates. One problem was that the crates were designed to be lifted by forklift trucks operating from a smooth concrete or asphalt surface. In Korea, however, the operating surface was likely to be muddy or dusty, depending upon the season, and the handlers used truck cranes that employed a steel cable wrapped around four to ten of the tank crates at a time. The crates were not designed for such treatment and suffered damage as a result.23
Readying the tanks for operations took some effort, not only because of damage to the tanks but oftentimes also because of measures taken to protect the unassembled tanks. Tanks shipped in knockdown condition were coated with oil (oil fogged) to inhibit rust. As these tanks were transported in the open, the oil collected sand, dirt, and grit, which had to be removed before assembly. A shortage of solvents in Korea forced the airmen to use labor-intensive steam cleaning, which required between six to nine man-hours to clean each tank.
It took additional time to assemble the tanks. To complicate matters, the tanks of the six different manufacturers were not interchangeable and required modification (termed “excessive” by Fifth Air Force) to fit the F-86. Perhaps the worst case of non-standardization was the Ingersoll tanks that used two sway braces in contrast to the other manufacturers that used one. The USAF experienced what it termed “considerable trouble” with the two-sway-brace installation in the second half of 1950, yet Ingersoll turned out at least fourteen thousand such tanks. It was not until March 1953 that the Air Force stopped delivery of the two-sway-brace Ingersoll tanks. The non-standardization and quality control problems can be attributed to inadequate USAF direction and supervision.
The demand for labor forced the Americans to rely on Korean help. This was a problem because “much of this Korean labor was totally inexperienced in the act of mechanics—many did not know the uses of pliers or screwdrivers.”24 This sounds like American impatience, arrogance, and hyperbole, but it nevertheless indicates there was a problem. Each wing set up a tank farm for the assembly of the tanks, manned by fifty or more men. This unanticipated labor demand forced the Air Force to use all available help, including cooks and other kitchen staff. As with other maintenance in Korea, most was done in the open under difficult conditions. Clearly the USAF needed standardized tanks requiring simple field assembly, but the effort to speed production pushed the Air Force in another direction.25
Unfortunately, these are not all of the difficulties the airmen encountered with drop tanks. Two operational problems emerged. First of all, some tanks did not properly release from the fighter. The F-86 pilots jettisoned the external tanks when they spotted enemy aircraft in preparation for combat, and if a tank or tanks did not release from the aircraft, Air Force policy dictated that fighter head for home. It was not so much that a stuck tank caused asymmetrical flight, something that could be overcome with some trim and some care, but that the hung tank would cut performance. Because the F-86 was at best equivalent to the MiG-15 in performance, any degradation was most unwanted. To make matters worse, USAF policy stated that an aborting fighter would be escorted home; therefore, one hung tank meant two less Sabres in action. The release problem diminished with time but was never completely resolved. In June 1951, the 4th reported that between 5 and 10 percent of the tanks experienced release problems. By the end of the year, that number was below 1 percent. For the remainder of the war, the failure rate hovered just over 1 percent.26
The airmen concluded that the principal cause of the release problem was the shackles. Because these were part of the drop tank and thus built by each manufacturer, it is little wonder that there were problems. The 4th dipped the shackles into a compound and incorporated a gun-heating element on each shackle to prevent icing from binding the mechanism. The next month the 4th reported “the problems of tip tanks failing to release has almost been eliminated by the adjustment of the tank shackles and the use of Dow Corning compound as packing for tank release shackles.” But this did not end the problem. In the first half of 1953, the 51st reported that shackles caused 90 percent of the failures of the Borg Warner tanks.27 A July 1953 report stated that “in almost every case reported [of the failure of a tank to release], the failure occurred in the tank shackle and not in the aircraft manual- or electrical-release system.”28
The second major operational problem was that all too frequently the released drop tanks flew back into, and damaged, the fighter. Some of the tanks slid forward and rode down the right wing’s leading edge, shearing off the pitot tube and knocking out the air speed indicator. (These damaged F-86s could land by flying on the wing of another fighter with a functioning air speed indicator.) Other errant tanks damaged the flaps or empennage. Even before the Sabre entered the war, the USAF experienced problems with drop tanks damaging the fighter, so this problem was no surprise to the airmen.29 Because this seemed to be caused by the fuel hose binding and the connections, the airmen lubricated the fittings, deleted hose clamps, and checked that the attaching bolts were properly torqued. The airmen instituted procedures that also seemed to help the situation. They found that if the air speed was above Mach 0.82 and there were some positive “g” forces when the tanks were released, separation went well.30 The 51st came up with another solution, adding a small (2.5 inch) airfoil (spoiler) on the tank to aerodynamically force the tank down and away from the fighter upon tank release. During the twelve-week period prior to the first use of the airfoil, the unit suffered damage on 0.86 percent of tank releases; in the twelve-week period afterward, the damage rate was cut in half. This worked well with the tanks fitted with one sway brace but in fact increased the problem with those using two sway braces.31 After a respite, Royal Heater tanks caused the problem to recur in the 4th in February 1953. In the first half of 1953, 63 percent of the 4th’s aborts were due to tank problems. In April 1953, the 51st recorded 0.76 percent damaged from tanks.32
A third problem with the drop tanks was cost. Astonishingly, drop tanks accounted for half of the operating costs of the Sabre units.33 Because the original North American tanks cost $850 each, the USAF sought cheaper alternatives. By July 1952 competition induced North American to slash its price to $386 per tank. Beechcraft tanks cost $587 each. By November 1952, tanks manufactured by Pastuchin and Ingersoll cost approximately $300 each. The least expensive were Japanese-built tanks: American-made tanks sold for an average cost of $457, including transportation; Japanese tanks could be delivered at half that, an average cost of $223. Although the percentage of operating costs attributed to drop tanks remained high, between 45 and 50 percent from November 1952 through May 1953, the average cost per drop tank gradually declined to around $300 for the period February through April 1953.34
Although the Japanese tanks proved to be much cheaper than American-built tanks, the airmen were slow in getting them into action. The 4th Fighter Interceptor Wing did not test them until February 1952, and the Fifth Air Force did not authorize their use until August. Compared with American-built tanks (specifically North American and Beech), Japanese tanks were cheaper but carried less fuel and had greater drag, requiring the F-86s to carry 5 percent more power to reach the same air speed. They were taken out of service for a time but then used again in 1953. While the 120-gallon Japanese tanks proved very satisfactory, the 110-gallon tank built by Kawasaki did not. In April 1953 one F-86 unit suffered 2.6 percent damage from the 1,014 Japanese-built 110-gallon tanks that were dropped, compared with 0.22 percent from the other 3,610 tanks dropped that were built by other manufacturers. The 4th discontinued use of these Japanese tanks in June, an action endorsed by an Air Force study group the next month.35
This same study also recommended discontinuing the use of tanks built by Royal Heater. These tanks caused the USAF considerable difficulties and more than their fair share of damage. Nevertheless, the airmen were forced to use them because of operational requirements. The airmen went so far as to name the damage from the Royal Heater tanks as one of two factors that were largely responsible for the declining efficiency of the 51st Fighter Wing in February 1953 because the unit’s victory-to-loss ratio fell from 26:1 in January to 6:1 in February. Finally, in the first half of 1953, the Air Force was able to discontinue or limit the use of the Royal Heater drop tanks.36
The USAF made a number of other efforts with drop tanks. The early Sabres (F-86A-1 through early production of the A-5) could carry two 206.5-gallon tanks. Then the manufacturer reduced this capability, apparently because these tanks caused buffeting at higher speeds that restricted operations to below Mach 0.8. North American replaced these tanks with two 120-gallon tanks. Later, the company increased the Sabre’s drop tank capability by fitting F-86Fs with either two 200-gallon tanks or two 120-gallon tanks. The company fitted the F-86F-25 (and subsequent aircraft) with four pylons that could carry four tanks (two 120-gallon and two 200-gallon) or a combination of tanks, bombs, or rockets. The manufacturer tested the larger tanks in the first half of 1952, using new pylons with leading edges that were cambered inward, allowing higher speeds. In December 1952, the 51st began testing the 200-gallon tank. These tanks increased the F-86’s radius of action from 330 to 465 miles but also exhibited “undesirable flight characteristics.”37
The USAF attempted to obtain a “standard contour fuel tank” for all of its aircraft. This certainly made sense in terms of logistics and lowering overall costs as opposed to having tanks that only fit one type of aircraft. After some resistance, four manufacturers agreed; as of mid-August 1952, however, the Air Force was convinced that North American was absolutely against such a program. This project had no impact on the war in Korea.38
There were other efforts to extend the F-86’s range. In early 1951, the USAF tested aerial refueling on a T-33, F-84, and F-86. The airmen flew four flights in an F-86A in March using a drogue system from a KB-29 tanker and a 3.5-foot probe on an F-86 120-gallon external fuel tank. This configuration encountered problems because it placed the fighter’s empennage in the tanker’s turbulence, making longitudinal control “very difficult” and causing the drogue to bob. As a result, contact in the F-86 was more of a problem than with the other two aircraft. Maintaining contact required care. The testers concluded that aerial refueling with this arrangement was feasible but difficult.39 Then, in the summer of 1953, Wright-Patterson AFB considered the use of a 200-gallon tank for in-flight refueling of the F-86F. The airmen concluded that although it was possible, requiring effort similar to formation flying, it “will require additional effort on the part of the pilot to maintain position due to the poor stability and control characteristics of the F-86F.”40 The Air Force later fitted an F-86E with tanks modified with refueling probes, but there is no record of any in-flight testing. The Air Force also conducted tests with a boom-equipped KB-29 and a Sabre fitted with a refueling receptacle just forward of the canopy. Photos document this effort, but, alas, there is no indication of when it took place. The USAF did not equip production Sabres with aerial refueling equipment, for by this time the F-100 with a fixed probe was about to enter service.41 Since then, all USAF fighters have air-to-air refueling capability.