KALAMAZOO, MI, USA
NX72WP
GRUMMAN F9F-2
According to the pilot, shortly after takeoff, about 200 feet above ground level, he felt a sudden deceleration and was thrown forward in his seat. He also noted that the airplane would not accelerate beyond 120 knots indicated airspeed, which he attributed to a problem with the pitot system, and that he decided to return to land. He said that while on the downwind leg, he retarded the throttle to idle, but the engine rpm reduced to 65 percent instead of the normal 28 percent. At this time, a red light illuminated in the cockpit, and the pilot stated that he believed this to be the warning light for the Emergency Fuel System. He said that as a precaution, he decided to shut the engine down and do a power-off landing. The pilot reported that he aimed for the first third to one-half-the-way down the runway, and he landed 'long and hot.' He was unable to stop on the remaining runway and steered the airplane on to a grassy area to the right side of the runway. He indicated that he had previously experienced anomalies with the fuel control unit and fuel pump; however, postaccident testing of these components did not reproduce the conditions reported by the pilot on the accident flight. Except for a slightly higher than normal idle (36 percent), the fuel controller performed satisfactorily. The airspeed indicator was damaged and could not be functionally tested.
HISTORY OF FLIGHT On September 19, 1996, at 1721 eastern daylight time, a Grumman F9F-2, NX72WP, overran runway 35 during an emergency landing at the Kalamazoo/Battle Creek International Airport, Kalamazoo, Michigan. The pilot was seriously injured. The airplane was substantially damaged during its collision with a chainlink fence, ditch, road shoulder, berm, and the postlanding fire. Visual meteorological conditions prevailed, and the runway was dry at the time of the accident. The flight departed Kalamazoo about 1715 on a 14 Code of Federal Regulations Part 91 personal flight operating on an instrument flight rules flight plan. During postaccident interviews, the pilot reported that the engine did not hesitate when he applied full power for the takeoff from runway 35. The pilot reported that the takeoff run and lift-off were normal for a hot day. He stated that about 200 feet above ground level (agl), he felt a sudden deceleration and was thrown forward in his seat. He stated that the airplane would not accelerate beyond 120 knots indicated airspeed (KIAS). Shortly after noting that the airspeed was not increasing, and thinking that an engine flameout was imminent, the pilot made a tight turn, continued to climb, and contacted the Federal Aviation Administration's (FAA) Air Traffic Control Tower (ATCT). He stated that he told the controller that he had a "problem" and wanted to return to Kalamazoo to land. The Kalamazoo ATCT transcript shows that the accident airplane was cleared for takeoff at 1715:58. At 1717:56, the pilot called the ATCT stating, "Whiskey Pops got a problem." The pilot told the ATCT that he was not certain what the problem was and that he intended to return to the airport. A few moments later the pilot said, "we'll stay in the pattern...normal pattern...seems as though we're not accelerating...it could be a problem with the pitot." The controller asked the pilot if he needed any assistance. The pilot replied, "Negative, we'll be all right, we're just...precautionary." The pilot stated that the airspeed remained at 120 KIAS, but he sensed that the airplane was actually moving faster than 120 KIAS. The pilot stated that he flew the airplane to 2,000 feet above mean sea level (msl) in a left climbing turn to the downwind leg for runway 35. He stated that this altitude was a little higher than the normal traffic pattern for the airplane. While on the downwind leg, the pilot retarded the throttle to the idle stop. He stated that the rpm rolled back to about 65 percent instead of the normal 28 percent. The pilot stated that during the downwind leg he recalled seeing a red light glowing on the left-hand console; however, he was unsure exactly when it illuminated. He stated that he believed that the red light was the Emergency Fuel System Warning Light. He stated that he did not manually switch on the emergency position fuel system. He stated that he knew that the emergency fuel system adds more fuel to the engine and that in the emergency mode rapid movement of the throttle could flame out the engine. Unable to obtain less than 65 percent rpm, and unsure of what the engine was going to do, the pilot stated that he shut down the engine to make a power-off landing. According to FAA Form 8020-6, "Report of Aircraft Accident," at 1723, the wind direction and speed were 030 degrees magnetic at 6 knots. The pilot stated that he aimed for a point about one-third to one-half the way down runway 35 and that the airplane landed "long and hot." He stated that he retracted the flaps and applied the brakes after touchdown. The pilot stated that he observed obstacles at the end of the runway and to avoid them he put all of the pressure on the right toe-brake and steered the airplane onto the grassy area to the right side of the runway. WITNESSES A witness at the fixed base operation on the airport's west side stated that he "first saw the aircraft west of the airport at...pattern altitude. The aircraft sounded and looked okay." He said that the airplane "banked from the downwind all the way around and came out on final just like all military landings. The airplane was going very fast and looked like it was going to be a high speed pass." A second witness, who was on Kilgore Road at the time of the accident-north of the airport-watched the airplane's landing roll. The police reported that the witness said he observed the accident airplane's wheels being cherry red during the landing roll as it neared the highway. According to the ATCT controller statement, the accident airplane's ground speed was "220 knots on final approach...less than a mile from the runway." The controller also stated that the airplane touched down between taxiways B-2 and B-3 and "was going at a high rate of speed." PERSONNEL INFORMATION The pilot held an airline transport pilot certificate. In addition, the pilot held a "Statement of Acrobatic Competency" (FAA Form 8710-7) dated June 6, 1996, with the following maneuver limitations: "All Aerobatic Maneuvers Within AC Operations Limitations / Formation Non Aerobatic." The altitude limitations were "Level 3, 500 FT" and the authorized aircraft was listed as "F9F-2 Panther." The pilot provided copies of his logbook that were directly related to the F9F-2 airplane to the investigator-in-charge (IIC). The logbook copies showed that he received 2 hours of dual instruction in a DeHavilland Vampire on September 25, 1985. He began flying the accident airplane on October 7, 1985. An entry in the pilot's logbook showed that he had received an "FAA Comp. Ck. 3 TO Ldgs" in the accident airplane on October 29, 1985. The pilot subsequently received Letters of Authorization (LOA) from the FAA to fly the F9F-2 airplane dated November 7, 1985; November 14, 1986; and October 20, 1987. A copy of the FAA Form 8710-7, a copy of the letter verifying the training in the DeHavilland Vampire, and a copy of the LOAs are appended to this report. The logbook showed a total of 226.85 hours in the accident airplane. No other Grumman F9F-2 airplane flight times were shown. National Transportation Safety Board Form 6120 1/2, "Pilot/Operator Aircraft Accident Report," listed "236+" hours. In subsequent correspondence, the pilot indicated that he had a total of 333.4 hours in the accident airplane. According to the logbook, the pilot had flown the airplane 5.5 hours during June 1996. The last entry in the pilot's logbook was on June 8, 1996, which showed that the pilot had flown the accident airplane 1.5 hours and made three landings that day. Excerpts from the logbook are appended to this report. AIRCRAFT INFORMATION The accident airplane was a Korean War vintage military jet airplane. The pilot provided copies of the airframe and powerplant logbooks to the IIC. According to the July 27, 1983, airframe logbook entry, NX72WP, serial number (S/N) 12078, was assembled "using new, spare, and surplus parts." The airplane received its special airworthiness certificate from the FAA on July 27, 1983. The logbook showed that the airplane had zero total time in service when it received the special airworthiness certificate. The airframe logbook showed that the accident airplane's last annual inspection was completed on May 21, 1996, and that it had 373.4 hours on its Hobbs meter. There were no other logbook entries describing any airframe maintenance after the May 21st entry. The accident airplane's total airframe time, which is based on a review of the accident airplane's airframe logbook and the pilot's logbook copies, was 378.9 hours at the beginning of the accident flight. According to a July 27, 1983, powerplant logbook entry, the accident airplane's Pratt and Whitney J42-P8 engine, S/N P400638, was "overhauled from new, spare, and surplus parts." This entry showed that the main fuel control unit (Bendix model TJ-Cl, S/N 17441) had been overhauled before installation on the accident airplane's J42-P8 engine. There were no logbook entries associated with the engine's main fuel control unit after the July 27, 1983, entry. Copies of the pilot-supplied airframe and powerplant logbooks from the accident airplane are appended to this report. The pilot supplied other documents to the Safety Board concerning the fuel control unit, including invoices for fuel control repairs, a certification of the history of fuel control repairs from a mechanic at Kal-Aero, a company that performed maintenance on the accident airplane, and a report of the metallurgical and physical examination of the fuel control. Copies of these documents are appended to this report. See the Powerplant Group Chairman's Factual Report for additional details about the powerplant and component maintenance history. AIRPORT INFORMATION The Kalamazoo/Battle Creek International Airport field elevation is 874 feet above msl. The touchdown zone elevation for runway 35 is 868 feet msl. Runway 35 is 6,499 feet long by 150 feet wide, and its surface has grooves perpendicular to its centerline. The grooves are about 1/4 inch deep and cover the runway surface, except for about a 3-foot border on each edge. Runway 35 is intersected by runway 09-27 about 3,400 feet from its approach end. WRECKAGE AND IMPACT INFORMATION During the on-scene investigation, runway 35 was examined for tire skid marks. Gray colored skid marks, spaced about 9 feet apart, were observed from the intersection of runways 35 and 27 to the blastpad at the departure end of runway 35, a distance of about 3,015 feet. As the skid marks neared the end of the runway, their individual width and color intensity gradually reduced. According to the airplane's maintenance manual (AN 01- 85FGA-2), the main landing gear tread is 8 feet, 3 inches wide. Excerpts from the manual showing this information are appended to this report. The skid marks continued off the north end of runway 35. A tire skid mark that correlates with the position of the airplane's left landing gear passed by one of runway 17's threshold lights, about 12 inches to its west side. This threshold light, and a second one about 10 feet west of it, were not damaged. These lights measured 29 inches high. The tire skid marks continued through the runway's clearway, across an east-west dirt access road, through a chainlink fence, up an embankment, and across an east-west four-lane highway. The airplane came to rest on the east side of an earthen berm on an approximate magnetic heading of 045 degrees. A diagram of the runway and off-runway airplane track is appended to this report. The accident airplane's nose section forward of the front cockpit pressurization bulkhead had separated from the fuselage. Equipment associated with this airframe section, such as the nose landing gear, was found scattered around the accident site. The windshield frame and glass had separated from their respective mounts. The bottom of the fuselage below the cockpit was crushed upward and aft to the rear pressure bulkhead. One dive brake was separated from its mount and was lying under the front fuselage. A second dive brake was hanging from its mounting. Portions of the fuselage bottom's exterior skin separated from the main structure. The cockpit floor was crushed upward. The pilot seat had been pushed upward with the floor. The top of the seat had penetrated the canopy just aft of the canopy's forward frame. The trailing and leading edge flaps were found retracted. The right main landing gear leg had separated from its mount below the leg's actuating rod attach point. The right wingtip fuel tank was split open. The forward half of the tank's bottom was crushed upward. The left wingtip fuel tank's bottom was crushed upward. This fuel tank was involved in the postimpact fire. The bottom of the left wing and the left side of the fuselage midsection also had fire damage. The airframe skins involved in the fire were not melted. The fire damage consisted of burnt paint and skin surface damage. Vegetation around and under the airplane's left side and aft fuselage had been burnt. The left wing's leading edge had two crush areas located about 18 inches inboard of the wingtip fuel tank and about 10 inches outboard of the air intake opening. The right wing had a crush area located about 6 inches inboard of the wingtip fuel tank. All three crush areas extended aft, into the wing, about 4 inches. The engine air ducts and the screened areas surrounding the engine air inlets did not have any debris blocking them. The fuel tank baffles were attached to their mounts. The plenum chamber hydraulics reservoir pressure gauge showed 835 pounds per square inch (psi). The hydraulic lines associated with this reservoir were not leaking. Hydraulic fluid was not found on the chamber walls. The engine-driven pump was connected to the system solenoid shutoff valve. The pitot tube under the left wing had a piece of wood pressed into its opening. The wood was about 3/4 inch long and teardrop shaped. The rounded end was about 1/2 inch in diameter. The piece of wood that had penetrated the pitot tube's opening matched a piece of weathered wood lying near the airplane's right wing. After the piece of wood was removed from the pitot tube end, air was blown into the open end of the pitot tube and was felt to exit the fitting end formerly attached to the airspeed indicator. The two fuselage-mounted static ports were found unblocked. The static lines were found crushed, bent, and fractured in the forward fuselage area. The left side of the instrument panel was bent about 70 degrees aft starting at the panel's center. The altimeter's case was cracked. The emergency brake "T" handle was found in the stowed position. All circuit breakers on the cockpit panels were broken, popped, or missing. The engine-driven fuel pump, landing gear, low-fuel boost pressure, and generator warning light filaments were stretched when examined under 10-power magnification. The emergency fuel system indicator light bulb filament was not stretched when examined under the same magnification. The emergency hydraulic pump switch, located on the right cockpit console, was destroyed, and its preimpact position was not identifiable. The engine fuel system selector switch was found in the "Takeoff and Flight" position. The engine's fuel control arm was found at the "Idle Stop" position. Alignment holes on the throttle bellcrank were in line with each other. Continuity was established between the fuel control and pilot's throttle. Control continuity was established for all three flight axes. The elevator push-pull tube had separated aft of the control stick attach point. The separation surfaces of the tube had shear lips on them. The separated ends of the tube were compressed. TESTS AND RESEARCH The main fuel control and both fuel pumps were individually bench checked and installed on the engine for a test cell run. The engine was accelerated in 5-percent increments until it attained 100 percent of its rpm. The engine's power lever was then moved to the idle position, which yielded 36-percent rpm, with a fuel flow of 1,900 pounds per hour (pph). The engine was accelerated and decelerated two additional times. In each case, the engine returned to 36 percent rpm, with a fuel flow of 1,900 pph. One of the two acceleration/deceleration modes was accomplished with rapid throttle lever movement. In both instances of acceleration/deceleration, the engine did not surge or hang up. The maintenance manual showed that the engine's idle rpm should be about 28 percent. According to the pilot, after the accident, and subsequent to the Safety Board's release of the wreckage, he sent the main fuel control amplifier from the accident engine to McSwain Engineering for disassembly and evaluation. The Safety Board was not notified and did not participate in the teardown and examination. The main fuel control amplifier was disassembled and the McSwain Engineering disassembly report states that the amplifier valve, also known as the half-ball, exhibited damage and wear. The report also states that the fractured surfaces of the two valve seat tangs that were found in the wreckage showed signs of overstress; the
The overrun of the runway as a result of the pilot's misjudgment of speed and distance because of an unreliable airspeed indication; reason for the unreliable airspeed indication was undetermined. Contributing to the accident was the pilot's perception of a power interruption and throttle/power control not being available for undetermined reasons.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
