Sabre Jet
XP-86 SWEPT WING DEVELOPMENT

by Larry Davis

The decision to radically re-design the XP-86 (NA-140) was both easy and difficult for North American officials. Costs incurred on the straight wing jet fighter were absorbed by the company. Some of this would be recovered through the Navy decision to produce the FJ-l Fury. However, it was easy for North American to make such a drastic move because the NA-140 would never meet the AAF General Operational Requirement (GOR) - a top speed in excess of 600 mph. The thin straight wing simply wouldn't allow that type of speed.

Wing sweep had long been known as one of the answers to lower drag. Drag Coefficient was significantly lower when the wing was swept; i.e. at 0.9 Mach, drag coefficient for a straight wing was 0.05, while a swept wing was 0.01. But the problems associated with wing sweep were as great as the end results. Sweeping the wing did lower the thickness ratio, thus reducing the drag coefficient, resulting in higher speeds. But sweeping the wing just 12 inches created wing tip stall and low speed stability problems that no one had been able to overcome.

In August 1944, Ed Horkey, North American Chief Aerodynamicist, went to NACA at Langley Field to study the effects of a very thin wing operating at high Mach numbers. He was informed that no data existed for such a design. In 1945, Allied forces overran German test facilities, including one conducting research into the effects of wing sweep. The Me-262 jet fighter had a mild (150) sweep to the leading edge of its wing. It was also discovered that Messerschmitthad been working on a radically swept (350) version of the Me-262, called the Pfeilflugelor 'arrow wing'.

George Schairer of the Boeing Company, went to Germany after the end of the war, with von Karman and Robert Jones (an early NACA proponent of swept wings), to investigate German data on swept wing technology. Schairer was very enthusiastic about wing sweep, proposing that Boeing incorporate it into the new XB-47, and noting that this information should be made available to the US aeronautical industry.

Larry Green, head of Design Aerodynamics at North American, came up with an answer to the swept wing instability problems. Green had been conducting wind tunnel tests on the Curtis XP-55 Ascender, which indicated severe non-linear instabilities present at high angles of attack over a swept wing, causing an extreme pitch up' attitude. Green, Walter Koch (who spoke fluent German), Dale Meyers, and Harrison Storms set up four Recordaks, and began translating the material being funneled to North American by Wright Field. Within the captured material were considerable data concerning the use of wing leading edge movable surfaces, commonly called 'slats' as a possible solution to the instability problems.

The North American Technical Section headed by Ed Horkey, included Harrison Storms as Chief Aerodynamicist, Walt Fellers, Larry Green, Meyers, Bill Wahi, and a host of others. This group finally convinced the powers-that-be that wing sweep would put the XP-86 over the top speed requirement of the G.O.R.. On 18 August 1945, North American received a research and development grant to develop a swept wing XP-86 (RD 1369). Two weeks later a .23 scale model of the swept wing XP-86 was ready for wind tunnel testing. The results were what North American had been looking for, and clearly indicated the drag rise and compressibility had been lowered enough to bring the XP-86 into the 600+ mph range.

In September 1945, the straight wing XP-86 fuselage was mated to a 350 swept wing and tested in North American's low speed wind tunnel. The results were satisfactory and seemed to indicate that the slats were probably the answer to the instability problems. On November 1st 1945, General Bill Craigie, head of R&D at Wright Field, gave North American the go ahead for the swept wing XP-86.

Slowly but surely, North American's engineers brought the design to its final shape. But the slat design remained a problem. Finally, an entire Me-262 wing was flown in from Wright Field. North American's engineers disassembled the slats and modified the slat track mechanism to fit the XP-86 wing, using the Me-262 slat lock and control switch. Although not perfect, it was a start and the slat worked. In fact, the first seven aircraft used Me-262 slat locks and tracks.

Both 5 and 6 aspect ratio wings were tested before finally settling on a 5 ratio wing with a 350 sweep The trailing edge of the wing was extended 4" at the root, thus reducing trailing edge separation which had caused some loss of aileron control. Modifying the trailing edge angle increased the amount of overall sweep to 35.20.

Originally, the swept wing proposal retained the tail assembly from the straight wing design. However, by the time the XP-86 mockup was built, both the vertical and horizontal tail surfaces were also swept at 350 Additionally, the horizontal stabilizer was fully trimmable to achieve a better balance between low speed control and high speed requirements. Sweeping the tail surfaces caused the overall fuselage length to increase from 35.5' to 37.54'.

The cockpit sat high on the forward fuselage with a PlexiGlas canopy offering excellent all-round vision (3600) for the pilot, something not matched in a fighter aircraft until the McDonnell/Douglas F-15 Eagle was unveiled some 25 years later. The XP-86 had three speed brakes - a pair of small doors on the rear fuselage sides that opened at the rear; and a larger slab door under the rear fuselage.

The XP-86 had six .50 caliber M3 machine guns in banks of three on either side of the cockpit. Ammunition bays in the bottom of the fuselage held a maximum of 300 rounds per gun, although 267 rounds was the normal load. The gun muzzles were recessed behind individual frangible 'doors' that opened in 1/20th of a second when the trigger was pulled. All radio and radar antennas were enclosed in fibreglass fairings within the design of the aircraft. The engine powering the prototype was the Chevrolet-built, General Electric J35-C-3 rated at 4,000 lbs static thrust. However, production aircraft would be powered by the GE TG-190 (J47) engine offering 5,000 lbs. thrust.

The design of the wing wasn't the only new innovation found on the XP-86. The wing was of a totally new construction and manufacturing process. The conventional 'rib and stringer' wing construction was replaced by a double skin structure with 'hat sections' between the layers that provided room for installation of self-sealing fuel tanks. The wing skins were tapered throughout their length and width, being .250" thick at the wing root, tapering down to .064" at the joint with the outer wing skin, and .032" at the wing tip. So complex were these tapered skins that they required special milling machines, taking 45 minutes to complete a single skin. The fuselage was unique in that it was divided into two sections, joined near the wing trailing edge, providing easy access to the engine and accessories.

On February 28th, 1946, the mockup received AAF approval. In August 1946, basic engineering drawings were finished and metal was cut. So excited was Army Air Force over the prospect of the new swept wing XP-86, and its projected much higher performance, that a contract was awarded on December 20th, 1946, to build 33 production P-86As. No YP-86 service test aircraft were built on August 8th, 1947, the wait was over. The doors of North American's Inglewood factory opened to reveal the first swept wing production aircraft in the world -XP.-86 #1, serial 45-59597.

During the next month, taxi and brake tests were conducted at Mines Field, adjoining the North American factory. On September 11th 1947, the XP-86 was disassembled and trucked to Muroc Dry Lake Army Air Base, now known as Edwards AFB. The XP-86 was re-assembled, and all systems retested and adjusted. On the morning of October 1st 1947, George Welch, Engineering Test Pilot for North American, taxied the XP-86 to the edge of the runway at Muroc, released the brakes and pushed the throttle forward. Three thousand feet down the runway, the XP-86 lifted smoothly off the dry lake bed for the first time. Everything went smoothly, and 30 minutes later Welch made a final circuit, pulled the lever to lower the landing gear, and began his landing approach.

Ed Horkey recalls; "Ed Virgin, Head of Engineering Flight Test, Jim Sullivan, and I were clustered around the radio listening to George's excited chatter about the first flight. The first flight went very smoothly, with each item on the First Flight Card being checked off. George could tell from the speeds he was obtaining, versus the power settings, that he was riding something pretty fast After about a 30 minute flight, it was time to land and George lowered the flaps and gear."

"The main gear lights were in the green, but the nose wheel light didn't show the gear down and locked. We had him make a pass. We could clearly see the nose gear was down about 45 degrees George tried everything he and we, could think of. Nothing worked. Decision time was rapidly coming upon us. Ed Virgin and I were immediately unanimous in letting George make the decision on whether to jump or bring the airplane in. The company would not influence a decision to save the airplane if it meant jeopardizing George's life."

"George radioed that he would stay with the airplane and try to bring her home. But he was going to land on the dry lake bed, not on one of the runways. He made a smooth, very nose high approach. Touching down on the lake bed, George let the airplane just coast along with no brake application. As the airplane slowed and the nose started irretrievably over, the nose gear swung forward and locked into place, with George quietly exclaiming "Lucky! Lucky!""

"We found out later that someone in landing gear hydraulics wasn't impressed with the nose gear load data furnished by the wind tunnel crew. They installed a cylinder/piston size on the nose gear retraction system that was too small - not the one called for. Normally, nose wheels rotated down to the rear, so that air loads would force the gear down even if the hydraulics failed. However, the XP-86 nose gear rotated down to the front, against the air stream! The immediate fix on the XP-86 was to use two of the original cylinder/pistons, then replace them later with the correct one."

Other than the nose gear problem, Welch had only one complaint. The J35 didn't have enough power! With only 4,000 lbs. of thrust, the XP-86 had a rate of climb of only 4,000 ft/mm. But since the J35 would not power the production airplanes, no one got too excited. Production P-86As would have J47 power with 5,000 lbs thrust available. Especially in light of what happened next while still using J35 power.

On October 14th 1947, Chuck Yeager took the rocket-powered Bell XS-1 beyond the magic number of Mach 1.0 - the first piloted vehicle to do so. But George Welch may have beaten that date in history during some of the routine tests with the XP-86.

Ed Horkey: "Had George gone Mach 1.0 before October 14th 1947? It's an intriguing question. Recently discovered data in the North American archives indicates that possibility. During the Phase One Flight Tests, George had been telling me that he was getting oscillation of the air speed and altimeter readings, indicating Mach 1.0 on the pitot head. (During flight close to Mach 1.0, shock waves will affect both the airspeed and altimeter readings.) But at that time, North American had no way of calibrating airspeed indicators at that speed."

"NACA had a flight test operation at Muroc. They put a tracking theodolite (like a surveyers transit) together with a large radar receiver which measured speeds very accurately at any altitude. It was the same system that had tracked Yeager in the XS-1. We heard about it and talked Walt Williams, NACA Director, into tracking George in the XP-86. They asked Welch to dive the XP-86 in a certain pattern. Lo and behold, George hit a reading of Mach 1.02! The date was October 19th 1947 - five days after Yeager's flight in the XS-1. The tests were flown again on the 21St with the same results. But George had been performing some of those very same flight patterns before October 14th 1947!"

"Just like the people involved in the XS-1 program, everyone involved in the XP-86 flight test program was immediately sworn to secrecy. In fact, Stuart Symington, the first Secretary of the new independent United States Air Force, called Dutch Kindelberger, and told him not to let anything out concerning the XP-86 going over Mach 0.935. It would remain a secret for many years!"

Phase II Flight tests flown by Air Force pilots, began in December 1947. Maj. Ken Chustrom was the Phase II pilot - "In late November 1947, North American called Col. Al Boyd, Chief of Air Force Flight Test, recommending delay of the start of Phase II flight tests because of heavy rains at Muroc, that flooded the dry lake bed. Colonel Boyd suggested I visit Muroc and inspect the conditions to determine if I could operate from the runway at North Base. After a few days at North Base evaluating the lake bed, the runway, and the XP-86, I called Colonel Boyd and recommended we proceed."

"Colonel Boyd notified North American that the Air Force would begin flying the Phase II tests of the XP-86 immediately. North American expressed concern because of the extremely short runway conditions. However, the Air Force prevailed and I made my first take off on December 2nd 1947. This flight was a get acquainted flight, and since there were 110 squawks, I asked that the XP-86 be refueled for a second flight that day! This allowed for performance checks and speed points at intermediate altitudes."

"1 was very impressed with the XP-86's speed improvement over the Republic X-84 with the same engine, but having a straight wing. The maximum speed for the X-84 was 615 mph, while the XP-86 maximum speed was in excess of 650 mph! Our remaining Phase II flights were accomplished in 11 flights totaling 10 hours and 17 minutes - all in six days! The XP-86s performance envelope was investigated up to Mach 0.9, and altitudes near 45,000 feet. My conclusion to the Phase II tests, and supported by our data, was that the Air Force now had the very best jet fighter developed to this date, anywhere in the world."

In early 1948, XP-86 prototypes #2 (45-59598) and #3 (45-59599) were finished. They were different from the #1 airplane, as well as from each other. Both #1 and #2 had different fuel gauges, a stall warning system built into the control stick, a by-pass for emergency operation of the hydraulic boost system, and hydraulically- actuated slat locks. The #1 bird was the only one with an on-board fire extinguisher. The #3 was the only one with automatic slats that opened at 135 mph, and full armament.

The #1 airplane was the only one with the rear-opening fuselage speed brakes, and a ventral brake under the fuselage. A new aft section had been constructed for the #3 prototype, with production-style speed brakes installed. These brakes had hinges at the front, and opening out and down. The ventral brake was eliminated. Air Force had approved the new brake design, but the new aft section would not be finished until late 1947. It was installed on the #1 prototype in late January following test flight #77. The other two prototypes were completed with production-style speed brakes.

The spring of 1948 saw many significant events take place in the early P-86 development. In March, the first P-86A, 47-605, come off the assembly line. In May the rest of the world was informed that George Welch had exceeded Mach 1.0 in the XP-86, the first airplane to do so (airplane being defined as a vehicle that takes off from the ground, flys, then returns and lands - under its own power). But the date was April 26th 1948, and George Welch wasn't the pilot.

Ed Horkey - "A visiting British pilot came over and checked out in the XP-86. He was told about the phenomenon he might encounter (i.e. breaking the sonic barrier), and the secrecy restrictions. Unfortunately, he had an open radio channel and all the nearby towers got an earful when he went through Mach 1.0. The facts soon became common knowledge throughout the aviation industry. I suppose the media never brought it to the public's attention, as it would tarnish the otherwise exotic story of Yeager and the XS-l. The June 14th 1948 issue of Aviation Week announced to the world that the XP-86 had gone supersonic."

It has been said that the April 26th flight took place after the XP-86 had been re-engined with the General Electric J47 engine. But that didn't take place until later in the XP-86 test program. On May 20th, the first flight of a production P-86A took place. On May 29th, Air Force officials placed a verbal order for an additional 333 P-86As, bringing the total production of the P-86As to 554 airplanes.

On September 16th 1947, Congress made the Air Force a separate branch of service from the Army. In June 1948, the new US Air Force re-designated all Pursuit aircraft to Fighter aircraft, changing the prefix from P to F. Thus all XP and P-86A aircraft became XF and F-86A. The three prototypes continued to serve in various test and evaluation programs into the 1950s, unofficially re-designated YP-86 following J47 installation. The #1 airplane crashed and was destroyed in September 1952, while #2 and #3 were retired from service in April 1953 and finally scrapped.


No portion of this article may be used or reprinted without permission from the President of the F-86 Sabre Pilots Association or the editor of Sabre Jet Classics magazine.


Return to Classics Page