Immediately after WW 11, when Grumman and the Navy first discussed the Panther, they were privy to the Navy Technical Team's knowledge of Germany's sweptwing research, and they also knew that many top German fighter designers had migrated to Russia-our new enemy. The research information firmly emphasized the greater critical Mach numbers possible with sweptwings compared with conventional straight-wing fighters for the U.S. and the Russia. The F9F-2 Panther straight-wing jet fighter con-tract awarded to Grumman in October 1946 therefore contained a provision for the development of a sweptwing version.
Because we were concerned about sweptwing jet aircrafts' higher stall speeds and about their compatibility with WW II Essex-class carriers' catapults and arresting-gear, we naturally gave top priority to the straight-wing version, which was under a firm production contract. Grumman was also under pressure from competitor McDonnell, whose twin-jet, straight-wing FH-1 Phantom had been carrier-qualified on July 26, 1946.
Meanwhile, the Navy Bureau of Aeronautics decided that proof of the concept would be obtained by flying a sweptwing experimental research aircraft equipped with slats to see whether its handling characteristics and stall speeds could be made suitable for carrier operations. Grumman and Bell were invited to put in their bids to get one
flying as quickly as possible. Bell judiciously proposed sweeping the wings of two P-63 King Cobras-one with leading-edge slats and one without. Grumman proposed Design 77-an entirely new sweptwing version of the Wildcat. Grumman's proposals were far more expensive, so Bell was awarded the contract.
On June 19, 1946, 1 flew a Tigercat to the Bell plant in Niagara Falls, New York, and evaluated the Bell prototypes. My flight in the first L-39-a sweptwing P-63 with no leading-edge device-was short. It cavorted like a cat on catnip during the stalls and required excessive altitude for recovery. The second L-39 prototype with leading-edge slats was docile during stalls and accelerated stalls. These stalls could he done with little wing dropping and the usual loss of altitude. The two prototypes made it clear that slatted, sweptwings would provide carrier-suitable flight characteristics and stall-speed performance for fighters. I was soon to discover that my L-39 flights were merely Sweptwing Course 101.
Progress on Grumman's sweptwing Panther (Design 83) was slow because the Navy put a much higher priority on straight-wing Panther production. Design 83 therefore slowly evolved into a completely new design with a variable-sweep wing that moved from 13 degrees sweep for landing to 45 degrees for combat. Grumman's preliminary design engineering led us to think that this was the only design that would guarantee sweptwing carrier suitability; it later evolved into the XF1OF-1 Jaguar.
Unknown to us at Grumman, the Navy Bureau of Aeronautics discovered too late that Rolls-Royce had sold to Russia the same Nene engine rights as Pratt & Whitney had bought for Panther manufacture. When, in late 1949, Naval Intelligence determined that the Nene engine was being installed in a sweptwing MiG-IS fighter, the Grumman
Panther and the McDonnell F2H-I Banshee straight-wing fighters then entering the Navy's inventory became instantly obsolete. Work on the sweptwing Panther (Design 93) was now a top Grumman priority, and XFIOF-l Design 83 was temporarily shelved.
To make the XF9F-6 Cougar design easier to produce, the Panther's fuselage, engine installation, wing center section, fin, rudder and landing gear were retained. The forward fuselage was lengthened by two feet to accommodate more fuel. The upper rudder was converted into a yaw damper; this was required to cope with the increase in yawing and rolling inherent to sweptwing aerodynamics in rough air. The wings were swept to 35 degrees, and hydraulic-powered leading-edge slats similar to the Bell L-39's fixed slats were installed. The stabilizer was swept 35 degrees and was made fully trimmable for high-Mach-number flight. Normal aerodynamically powered elevators were installed.
The first aircraft flew with short hydraulic-power-operated spoiler flaperons and aerodynamic ailerons (in case the flaperons' power failed) as a combination lateral control. A contract was eventually signed March 2, 1951. Only six and a half months later, on September 20, test pilot Fred Rowley flew the XF9F-6 Cougar on its first flight. The first production Cougars were delivered to VF-32 in November 1951 just one year after the MiG-15 debuted in Korea. It was interesting for me to watch the huge increase of interest in and effort given to the Cougar when it was accorded top-priority status. Combat necessity was eventually the speedy mother of Cougar invention.
As the senior engineering test pilot, I relinquished the program of demonstrating additional external stores for the F9F-5 Panther and was brought into the Cougar program shortly after its first flight.
COUGAR GROWING PAINS
Checkout in the Cougar could be rapid because the only difference between its cockpit and the Panther's was the flaperon/aileron lateral control system: it switched automatically to the regular ailerons if the flaperons' hydraulic power failed. A temporary paper placard specifying preliminary air-speed limits had been pasted on the instrument panel: speed was not to exceed 575mph. This was slightly less than the Panther's maximum speed at sea level and it was the Cougar's estimated best climb speed. I stupidly assumed that the Cougar had already been tested to its "placard" limits. Wrong!
As I accelerated through 525 mph on my first flight, I felt a strong vibration in the stick grip, and this was quite visible when I removed my hand. I immediately closed the throttle and extended the speed brakes; while decelerating through 500mph, the vibration stopped. I landed immediately for a complete aircraft inspection-no damage. I was then told that no other Cougar pilot had exceeded 475mph! On my next flight, at 525mph, my chase pilot noted that the spring tab on the elevator (which was there to reduce stick forces) was a one-inch blur when the vibration was happening. After I had landed, engineers soon determined that non-balanced tabs, which had been used on all previous Grumman fighters, had to be 100 percent static balanced when on a 35-degree-sweep control surface. New, balanced tabs solved the problem by increasing the tabs' flutter speed to well above the Cougar's designed maximum limit speed of 650mph indicated airspeed. We had just matriculated into "35-degree-sweep university."
LANDING-CONDITION STALL PROBLEMS
For its first few flights, the Cougar seemed to have satisfactory landing characteristics, and other, more pressing, problems were being investigated and corrected. On an early flight, I decided to look into landing-condition stalls under accelerated-stall flight conditions, which would occur when lining up for a carrier landing and flying into the turbulent wake of a carrier island structure. Carrier aircraft are required to have thoroughly tested landing-condition accelerated-stall characteristics to avoid having any spin tendencies.
For my first check of this condition, I was at 9,000 feet in a level-flight landing condition and 20 knots above the carrier-approach stall speed. I carefully pulled the stick aft to force an accelerated stall, which I estimated would occur at 1.5G. As the aircraft pitched upward, all of my pull-stick forces disappeared and were reversed! Even after I had instantly applied full for-ward stick, the nose continued to rise as the stall progressed. Without any application of rudder from neutral, the airplane yawed rapidly to the left and entered a spin. I immediately applied full-right anti-spin rudder. The Cougar was now in a fully developed left spin even though I was using full anti-spin control to stop it!
I was no longer the pilot. I was now an aghast passenger and wondering what had happened as a result of such a small aft stick motion. As the aircraft entered the second turn, I could see the village of Oyster Bay on the north shore of Long Island rotating me slowly below. During the third turn, the rotation slowed slightly and after the fourth, the aircraft's nose slowly dropped to the vertical and the rotation stopped. I applied full power to pick up airspeed and get out of the stall. I was staring straight down into oyster Bay's main street, which was only 2,000 feet below, and I was now too low to eject safely. I made as delicate a pullout as I could, but I was between a rock and a hard place: I didn't want to precipitate another accelerated stall, but I also wanted to avoid colliding with the homes of Oyster Bay's good townsfolk. My seemingly endless pullout bottomed at 300 feet above the ground!
I felt washed out, and during the recovery climb to 2,000 feet, I noted with great surprise that my landing gear was retracted. I must have instinctively retracted it, subconsciously cognizant of the fact that the Panther's and Tigercat's tricycle gear had much less spin stability when the nose gear was extended. I do not remember taking that action. This beast was totally unfit for any further test flights; something would have to be changed radically before I'd fly it again. I landed back at Grumman in a state of disbelief and shock after surviving the Cougar's first accelerated stall.
In previous-non-accelerated-stalls, I noted that the short yarn tufts on the wings (installed to indicate airflow) showed that there was a considerable spanwise airflow that promoted the stalling of the wingtips. Because we were all neophytes in sweptwing stalling aerodynamics, I suggested a solution that the Russian MiG-IS used: it had fences on the wing's upper surface. I was overruled. Also, just out-board of the engine air ducts, the Cougar had very sharp leading-edge contours that were supposedly designed to give proper stall airflow characteristics to all straight-wing aircraft. At my now highly motivated insistence, fences were installed, and the sharp leading-edge contour was given a more usual, rounded shape.
These two changes solved the landing-condition accelerated stall/spin-control-reversal problem. The yarn tufts showed that the fences stopped almost all of the low-G spanwise airflow-inboard of the fences-and restored chordwise airflow beyond the fences and to the wingtips. This was a satisfactory fix for stall speeds, but we soon discovered that we were still in sweptwing kindergarten.
THE BUFFET BOUNDARY
All fighter aircraft of the early '50s could maneuver at their designed 7.5G structural limit only at altitudes of less than 12,000 to 15,000 feet because at higher altitudes, transonic flow over the swept wings caused strong Mach-number buffeting during pullouts. This strong buffeting occurred at lower and lower G numbers during pullouts as altitude increased, and it was referred to as the "buffet boundary." Pull-ups into the buffet boundary also showed a stong, unacceptable reversal of stick forces and pitch-up. Yarn-tuft motions during these maneuvers clearly indicated that the inboard airflow was jumping the fences and causing the problem. During my first buffet-boundary pull-up attempt, the aircraft went into a spin that resembled my previous debacle over Oyster Bay, but the controls now reacted somewhat faster, and recovery was fairly prompt with only a few thousand feet of altitude loss because of the new wing-fence installation. Unwanted spins that start at 40,000 feet, however, are much easier on the nerves than those starting at 9,000 feet or lower!
Increasing the height and length of the new wing fences partially improved the buffet-boundary pitch-up problem. We also had to re-camber the long flat area of the engine air duct's upper surface to give it a four-inch higher airfoil shape to cure its tendency to stall prematurely and to ensure acceptable high-altitude buffet-boundary pullout characteristics without stick-force reversals. With these changes, the wing tufts now showed that almost all of the spanwise airflow that jumped the fence had been eliminated and was now a normal, chord-wise airflow. The complete tip stall-which caused the pitch-up-was a thing of the past. The pilot could now release the stick at peak G and be confident that the airplane would immediately reduce its angle of attack and return to 1G level flight. We had now graduated from high school transonics, but college finals loomed ahead
THE COUGAR GETS A FLYING TAIL
The conservative nature of Grumman's engineering had been demonstrated by its original aileron/powered-flaperon lateral-control system; this was proven to be unsatisfactory in transonic flight and was replaced exclusively by the powered flaperon system. With the same conservatism, the Cougar also had an industry-standard electric trim adjustable stabilizer and a normal elevator longitudinal-control system for simplicity and reliability. This was a similar configuration to the original North American F-86A Sabre, which flew four years before the Cougar. USAF test pilots soon determined that for acceptable transonic and supersonic maneuvering flight, the F-86A model Sabre required the newer, dual, hydraulic-powered, all-moving stabilizer that was dubbed "the flying tail." The trimmable stabilizer was much too slow to counteract the effects of fast trim changes and too slow for the sensitive maneuvering required for transonic gunnery and evasive maneuvers.
To follow up on the Navy's post-evaluation request for better Cougar transonic control, on April 23, 1952, Grumman sent me to Edwards Air Force Base to evaluate North American's Sabre F-86E number 91849. This was the first U.S. fighter to have the highly touted dual hydraulic-powered flying-tail system. This instantly user-friendly longitudinal-control system demonstrated that subsonic, transonic and supersonic flight could now be smoothly integrated. It also demonstrated much greater combat maneuvering capability in transonic and supersonic flight. On my return, I heartily recommended it for the Cougar.
Grumman's conservatism again prevailed, and engineers designed a single hydraulic system flying tail that reverted to its original trimmable stabilizer and mechanical elevator control when the hydraulics failed. It had some complexities that took time to resolve, but the Navy soon accepted it for the Cougar's operational future.
The F9F-6 Cougar was now over its major teething troubles and ready for squadron operations. It was too late for the Korean War, but it did provide the Navy with a carrier-based transonic fighter that could compete against the M1G-15 during the protracted Cold War.
COUGAR SERVICE HISTORY
The Cougar was first introduced to squadron operations by VF-32 in November 1951-one year after the MiG-IS appeared in Korea. Because of Grumman's production capability and the F9F-6 Cougar's pleasant flight qualities in carrier operations, 20 Grumman Panther fighter squadrons were re-equipped in the following two years. Seven more fighter squadrons were also assigned the F9F-6 Cougar, and four Navy attack squadrons flew them before converting to the Douglas A4D Skyhawk. Five reconnaissance squadrons were assigned the F9F-6P-the "photo-graphic" version of the Cougar. By the mid-1950s, Cougars out-numbered all other fighter types in the Navy's inventory.
By Corwin H. Meyer, Flight Journal
