HISTORY OF FLIGHT Use your browsers 'back' function to return to synopsisReturn to Query Page
On June 18, 1998, at 1116 Eastern Daylight Time, a rebuilt Hawker Aircraft Limited Hunter T Mk 8, N745WT, was destroyed during an approach to the Manchester Airport (MHT), Manchester, New Hampshire. The certificated airline transport pilot was fatally injured. Visual meteorological conditions prevailed at the time of the accident. The aircraft airworthiness test flight was conducted under 14 CFR Part 91.
The builder stated that on the day before the test flight, he and the pilot met for about 2 hours for cockpit familiarization and discussion of the flight test scenario. They briefed again for about 30 minutes on the morning of the test flight. The pilot wanted to use a modified Hunter test schedule that was agreed upon by the builder, and that schedule was later found with writing on it, near the wreckage.
The builder wrote that the pilot wanted to use the lower altitude structure for the test flight, based upon the prevailing weather. The builder then "insisted that the flight test be a maximum of 30 minutes plus taxi time because of limited fuel duration at the lower altitude. We both agreed on this plan." After the preflight, the builder again reminded the pilot of the 30 minute flight duration.
Several witnesses stated that the Hunter's internal fuel tanks and two pylon fuel tanks were full prior to the flight.
Witnesses also said that the Hunter started up about 1030. According to tower audio tapes, the jet took off at 1041. The builder wrote that it then turned to "a right downwind, building speed for a continued right onto a final approach with gear retracted for an unanticipated low pass at an estimated 400+ knots" before departing the airport area.
At 1047, the pilot radioed to approach control that he had a hydraulic problem, and that he was declaring an emergency. At 1048, he cancelled the emergency, and told approach control he would "stay out here and work for about 30 minutes."
After another 13 minutes of maneuvering, the pilot received vectors for a long final approach to Runway 17. At 1113:50, he checked in with Manchester Tower and requested "a touch and go, and some multiple visual patterns if you can arrange it." At 1116:03, he radioed: "Hunter 1 will be a full stop," and at 1116:10, the pilot reported: "Hunter 1 is gonna have to eject."
A witness heard and saw an airplane that was silver-gray in color, and looked and sounded like a fighter. "It was coming in pretty fast, and then it got silent" as it passed the witness, about 1/2 mile prior to the initial impact point. According to other witnesses, the jet continued to fly south over a populated area, crossed Interstate 293, then flew low over a rental truck parking lot. An employee of the rental company was standing near an open garage door when he "heard a noise like a plane - like a swoosh. I'm sure it wasn't under power. I looked out the door and saw it gliding past - the tail was down somewhat."
Another witness said he saw the canopy depart the airplane just as it was clearing the parking lot, where a number of people had been standing. The accident occurred during the hours of daylight, about 42 degrees, 57.1 minutes north, 71 degrees, 26.7 minutes west.
The pilot was experienced in the Hawker Hunter, and had previously flown with the builder. He was an active duty military pilot, but flying the test flight under civil authority only. A month earlier, the pilot had flown a Hunter in an air show, and he possessed a current Federal Aviation Administration (FAA) Statement of Acrobatic Competency for the Hawker Hunter.
According to the pilot's civilian logbook, he had flown over 230 sorties, and 240 flight hours in the Hunter. On his February 1997 application for a Second Class Medical Certificate, he stated that he had 4,500 hours of civilian flight time. His military flight time was not provided by the U.S. Air Force. However, his biography stated that he had over 8,500 hours of total military and civilian flight time, with over 3,500 hours in military fighters. It also stated that he had flown for Northern Lights Aircraft, Incorporated, and in that role, flew the Hunter in "stunts for cinematography, air displays, test work, and instructing test pilots in out-of-control/spin recovery and prevention procedures at the Navy Test Pilot School..." It went on to state that he was also "fully mission capable and qualified in the F-16."
The truck parking lot was bordered to the south by a chain link fence. Running along the outside the fence, was a narrow stand of trees, which rose 35 to 40 feet. To the south of the trees was a large sand pit, approximately 250 feet in length, and to the south of that, was a stand of dense woods.
The first ground impact area was observed in the sand pit, about 140 feet south of the fence. The second impact area, where the jet came to rest, was about 275 feet south of the fence, and slightly less than a mile from the airport runway. The pilot was found in his seat, in the woods, about 380 feet south of the fence. The drogue parachute was found deployed, while the main parachute was not.
Tree tops outside the fence were found sheared off, corresponding to a left wing being down at a 40-degree angle. The descent angle between the top of the sheared trees to the initial impact point was 11 degrees.
The jet's cockpit area was destroyed. There was substantial leading edge damage to the left wing, and minor damage to the right wing. There was no evidence of fire.
An inspection of the fuel system revealed that the fuselage center tanks and front tanks were empty. The two pylon fuel tanks were destroyed, and no evidence of fuel was found near the remnants. The right inboard wing tank had fuel to a maximum depth of about 1 1/2 inches, and the left inboard wing tank had fuel to a maximum depth of about 2 inches.
The amount of fuel remaining in the wing tanks when the accident occurred could not be determined. Firefighters on-scene reported seeing fuel leaking from the right wing area at a rate of 1/2 to 1 gallon per minute for about 20 minutes. However, it was not known what quantity of fuel had leaked before the fire fighters' arrival.
MEDICAL AND PATHOLOGICAL INFORMATION
On June 19, 1998, an autopsy was conducted on the pilot by the State of New Hampshire, Office of the Chief Medical Examiner, Concord, New Hampshire. No toxicological testing was performed.
Subsequent to the accident, its ownership became a matter of dispute, and was being determined through litigation. Several documents listed the builder as the owner. However, the builder also provided a copy of a bill of sale which indicated that the airplane had been sold to Grace Aire, Incorporated, of Corpus Christi, Texas.
The Hawker Hunter was a 1950's-vintage jet trainer/fighter. According to the builder, the accident aircraft was rebuilt from two others over a 12-year period. The cockpit-forward section came from a two-seat T Mk 8, while the remainder came from a one-seat F Mk 50. The jet was issued a "Special Airworthiness Certificate for the purpose of exhibition" on June 15, 1998.
Descriptions of the Hawker Hunter's fuel system, from Hunter T Mark 7 Air Publications A.P. 101B-1302-1, Chapter 2, Fuel System, include:
1. The fuel system installation consists of four flexible bag-type tanks installed within the center fuselage, four in each wing, and a provision for drop tanks. Delivery to the engine is from the front pair of fuselage tanks, the supply being supplemented by means of electrically-driven booster pumps....Transfer of fuel from the remaining tanks to the fuselage front tanks is accomplished by means of air pressure obtained from a restricted tapping on the engine compressor.
2. The four fuselage tanks are...mounted in front of the engine in the center fuselage, two on each side between and around the engine air-intake ducts.
3. ...All four tanks are each provided with a refueling valve, these being in connection with a pressure refueling coupling situated in the port wheel bay.
4. The two front tanks are each provided with a combined pressure and suction relief valve. Vapor release valves, one to each front tank, are also incorporated. The two rear tanks...(referred to as center tanks), are each provided with a lightly-loaded suction relief valve. These valves, like their counterparts in the combined valves of the front tanks, operate in an inward direction opposite to that of a normal relief valve, thus permitting air to enter the tanks should the pressurization fail and, consequently, prevent the setting up of a negative pressure and the possible collapse of the tanks. Non-return valves are fitted in the pipelines between the front and center tanks to prevent back-flow during the refueling of the tanks.
5. Each wing accommodates four tanks....The four tanks are interconnected and, for practical purposes, may be considered as one tank. The fuel from these tanks is transferred to the center fuselage tanks by air pressure...each inboard wing tank to its respective center tank in the fuselage.
6. The aircraft can be fitted with four under-wing pylons on which drop tanks or alternative stores may be carried....[There are] air pressure pipes which feed engine air into the drop tanks for fuel transfer purposes....
7. For effective transfer of fuel to the fuselage front tanks at all altitudes and rates of flow, a pressure of about 6 lb/in squared is required. The air supply for tank pressurization is taken from a restricted tapping on the engine compressor and then through a non-return valve and a filter to pipes which feed the port and starboard tank installations. Each of these pipes contains a reducing valve.... From these valves, the pipes convey the air to the drop tanks through non-return valves. The air pressure causes the necessary displacement of fuel from the drop tanks to the wing tanks, and thence to the center tanks in the fuselage, and finally to the front tanks.
Except in the case of transfer failure, the front tanks remain full as long as there is fuel in the other tanks. Should the transfer system fail, however, no fuel will be transferred from the wing and drop tanks, and only part of the fuel from the center tanks.
Under these circumstances, the fuel transfer pressure switches will operate, causing the transfer magnetic indicators to show. The warning switches are also in connection with the fuel contents gauges, giving an 'empty' indication for the center tanks. Thus, when the warning indicators show, the contents of the front tanks only are indicated, this being the only amount of fuel available to the engine.
8. A fuel flow proportioner is fitted in the main delivery line from the booster pumps. It consists of two cells of equal capacity, each equipped with a vane-type rotor, a non-return valve and a bypass valve. The non-return valves are open under normal conditions but if the rotors become stiff or inoperative, the bypass valves open to allow the fuel to flow unmetered. The inlet of each cell is connected to its associated tank system, and the outlet from each delivers into a common manifold. Providing that the inlet pressures do not differ by more than 2 lb/inch squared, the flow proportioner ensures that equal amounts of fuel are taken from both sides of the tank system.
9. The fuel is pumped to engine-driven pumps from the front tanks by means of tank booster pumps with inverted flight valves.... The two booster pumps are matched units which should empty the front tanks simultaneously, but provision is made to enable the pilot to balance the fuel levels manually should such action become necessary. The manual balancing controls consist of two switches mounted on the cabin starboard shelf. The pumps,...under normal conditions are running continuously throughout the flight.... Indicators, situated adjacent to the pump switches, provide warning in the event of pump failure.
From the pumps, the fuel passes to a fuel proportioner, which ensures a balanced flow from the two sides of the system under normal conditions, and thence to the low-pressure cock. From the low-pressure cock, the fuel passes to the engine-driven pumps, connection being made by means of flexible couplings which allow for engine expansion.
10. The dual engine-driven pump is of the positive-displacement, multi-plunger type, the delivery being controlled by variation of the pump plungers.
11. The low-pressure fuel cock control is mounted on the cabin port shelf and moves forward from OFF to ON. The cock is fitted in the pipe lines [between] the tank booster pumps [and] the engine inlet. The high-pressure fuel cock control is also mounted on the port shelf and moves forward from OFF to ON. It is used to stop the engine by shutting off the fuel supply to the burners....
12. The throttle control is mounted...on the cabin port shelf.... The throttle consists of a profiled plunger, movement of which varies the flow of fuel to the burners.
13. The fuel contents gauges are grouped together at the top of the center instrument panel. The tank units...consist of small condensers on flexible straps. They are fitted into pockets in the tanks and are secured by press fasteners. The condensers operate the gauges through amplifier units.
14. A switch mounted on the cabin port shelf is provided to energize a solenoid-operated valve in the engine fuel pump circuit to isolate one pump from the other in the event of a defect in the engine fuel pump system.
16. The suction and pressure relief valves (one of which is fitted to each front tank), consists of a body which contains a spring-loaded relief valve. The valve is adjusted to open at a pressure of 11lb/inch squared by means of an adjuster which screws into a cap ring. After adjustment, the cap ring is sealed by a screwed cap. The valve allows excess tank pressure to escape via the orifice which is in connection with the vent pipe. This pressure relief valve incorporates a small spring-loaded suction valve, the action of which is opposite to that of the pressure relief valve in that it admits air into the tanks should the fuel tank pressurization system fail, thus preventing the formation of vacuum and possible collapse of the tanks.
17. One [suction relief valve] is fitted in each of the fuselage center tanks, and one in each port and starboard pipe to the drop tank bypass valves.
18. Under certain conditions of flight, particularly during inverted flying, air may pass into the front tanks. Air also comes out of solution from the fuel at altitude, or the fuel may boil. The expansion of this air or vapor while climbing may prevent fuel transfer and allow the front tanks to empty while fuel remains in the other tanks To prevent this, a vapor release valve is fitted to each front tank.
From the Pilot's Flight Manual, section 1-1, page 9, paragraph 18, Fuel Gauge Errors: "The fuel contents gauges have been found to give erroneous indications due to temperature effects on the electrical gauging system. The magnitude of the error depends on both temperature and flight conditions. Low temperatures at high altitude give gauge underreading; high temperatures at high speeds at low altitude give gauge overreading. During a descent from altitude, if the inaccuracy is a gauge under reading, the gauges progressively become more accurate and may eventually tend to overread."
The builder stated that "new light bulbs [were] used throughout [the] aircraft."
TESTS AND RESEARCH
Six light bulbs were submitted for laboratory examination. The bulbs from "RH TANK PUMP FAILURE," "FUEL LOW PRESSURE," and "LH TRANSFER FAILURE" warnings had intact filaments that exhibited no stretching. The bulb from the "LH TANK PUMP FAILURE" warning had a broken filament that exhibited no stretching. The filaments from the "RH TRANSFER FAILURE" and the "HYDRAULIC PRESSURE" warnings were intact, and exhibited "slight stretching." Further examination of these filaments under a scanning electron microscope showed both to be "severely notched, consistent with a well-aged filament."
Fuel gauges, which were found near the cockpit, were also analyzed. Both gauges were missing their glass covers. One gauge's needle was frozen about the 575-pound indication. No marks were found on the gauge face consistent with needle impingement. The needle of the other gauge moved freely, and the end was curled under, approximately 360 degrees. Two marks were found on the gauge face. One was consistent in length and angle to the position of the needle at 740 pounds of fuel. The second mark coincided with the curled end of the needle, around 660 pounds of fuel.
In the cockpit, the left tank pump switch was found in the off position, and the right tank pump switch was found in the on position.
Areas related to the engine flame-out included:
-- Fuel burn rates at low altitude. Radar data was requested from the Federal Aviation Administration (FAA) to make fuel-burn calculations, but was not provided, and was subsequently lost. Air traffic control voice transcripts were provided which revealed some assigned block altitudes, a few specific altitudes, and a time line, but did not provide groundspeeds or airspeeds.
Scenarios were created, based on the information available and the data contained in the flight manual. One scenario showed that the airplane could have flamed out at the time of the accident, provided it was at full throttle almost the entire flight. Another showed that the airplane could have had about 1,700 pounds of fuel remaining, if the pilot had throttled back to best range speed, 310 knots, during the 17 minutes the airplane was maneuvering at altitude.
-- Fuel quantities taken onboard the airplane prior to the flight. The builder stated that he had run the airplane out of fuel about a week before the test flight, in order to recalibrate the fuel indication system. Fuel receipts from 2 days before the test flight indicated that the airplane had received 214 gallons of fuel, and the owner stated that he had then performed four to six high speed acceleration/parachute deployment ground runs. A fuel receipt from the day of the accident revealed that the airplane had received an additional 380 gallons, for a total of 594 gallons received since the airplane was run dry. The fuel capacity of the Hunter was about 737 gallons. The builder stated that it would have been possible for the airplane to have received additional fuel, and it not been logged, since at the time, he was also the owner of the company that supplied the fuel.
The Vice President of Airline Services, the company that performed the refueling on the Hunter, stated that he rechecked all refueling documents, and that he could not find any that were previously unaccounted for. He said that he had to perform a fuel reconciliation for each day, and that the maximum it might have been off each day was 1 gallon. He said he would have caught any discrepancy over that, especially one as large as what would have been missing from the Hunter.
-- Refueling procedures. According to the maintenance manual, the Hunter had a single-point refueling system which filled the fuselage front tanks first, then the fuselage center tanks, then the wing tanks, and finally, the drop tanks. If the drop tanks were filled, then the other tanks should have been filled, as well. According to the mechanic who fueled the airplane on the day of the accident, he removed the 2-inch filler caps from both drop tanks to monitor the fuel levels. The port drop tank filled up first, then the starboard drop tank. As the fuel levels reached the tops, he replaced the filler caps, then opened a screw on the tops of the tanks to bleed off any remaining air and allow them to fill completely. When fuel began flowing from the screws, he tightened them, then had the fuel truck stop pumping. After the fuel truck was disconnected, he did a walk-around inspection and saw some fuel "weeping" from the port wing negative pressure relief valve. After engine start-up, the weeping ceased. The mechanic noted that the fuel gauges were reading full.
The mechanic's refueling actions on the day of the accident were confirmed by the fuel truck driver.
The areas where the screws were located on the drop tanks were not found at the accident site, but were found months later, in the pile of wreckage, at the storage facility. The screws were checked by the FAA Inspector, and found to be securely in place.
A review of the video of the Hunter's taxi and takeoff operations revealed that there were no fuel leaks during those events, and that upon takeoff, neither wing dipped.
-- Collapsed fuel tanks. The tops of the Hunter's fuselage fuel tanks were secured in place by strings. If one or more strings broke, a fuel tank could have folded over on itself, which then would have allowed less fuel to be carried onboard. However, when the fuel tanks were checked, the strings were found to be securely in place, with none of the tanks folded over.
-- Extended engine storage. The builder stated that he had discussions with other people familiar with the Hunter. In those discussions, he was told that when an engine sat in storage for long periods of time, the barometric pressure compensator would pump unusually high rates of fuel into the engine until a break-in period had passed. However, according to personnel at the Empire Test Pilots' School, Boscombe Down, Salisbury, United Kingdom, the additional fuel flow, if it had occurred, would have resulted in a higher engine rpm, which in turn, would have been monitored by the pilot. In addition, the engine would have been both temperature and rpm limited, so as not to allow excess fuel into it, unbeknownst to the pilot.
-- Fuel tank transfer pipe sizes. Empire Test Pilots' School personnel noted that, if the "short" internal fuel tank transfer pipes were inadvertently substituted in positions requiring the "long" pipes, it would have allowed excess fuel to remain at the bottom of the wing tanks. When the FAA Inspector reinspected the aircraft, he found the proper-sized pipes to be in their correct positions.
On June 20, 1998, the airplane was released to the builder.