Tennessee

A witness heard the airplane's engines make "sputtering" engine noises that cut out and quit prior to its impact with the wooded terrain. A sheriff's deputy who responded to the scene heard the pilot state the airplane ran out of fuel. The propeller blades of both engines displayed minor damage, and both propeller spinners were found intact and in place. No evidence of fuel odor nor fuel leakage was observed at the accident site. During examination, both engines were supplied fuel and a source of electricity. Each engine operated smoothly during testing at rpm ranging from idle to 1900.

About 20 minutes before the accident, the pilot reported to the air traffic controller that he had a problem with an engine and needed to shut the engine down. The flight had just leveled at 23,000 feet. The controller told the pilot that he was near Jackson, Tennessee, and that he could descend to 7,000 feet. About 10 minutes later, the pilot reported he was at 8,000 feet and requested radar vectors for the instrument landing system approach to runway 2 at the McKellar-Sipes Regional Airport, at Jackson. The pilot told the controller he had the left engine shut down. About 5 minutes later, the pilot reported he had a propeller runaway. About 1 minute later, the pilot reported he was in visual conditions and requested radar vectors direct to the airport. About 2 minutes later, the pilot reported he had a cloud layer under him and that he had the localizer frequency for runway 2 set. About 1 minute later, the pilot was told to contact the McKellar Airport control tower. The pilot acknowledged this instruction. No further transmissions were received from the flight. Examination of the left engine at the accident site showed the left propeller control was found disconnected at the point the propeller control extension bracket attaches to the propeller governor. The propeller control cable had also pulled loose from a swaged point at the control rod and was also separated further aft due to overstress. The housing for the propeller control rod was found securely attached to the engine and the control rod was securely attached to the extension bracket. The propeller governor control arm, which was disconnected from the propeller control cable and rod, was found spring loaded into the high RPM position. Examination of the fractured left propeller bracket assembly was performed by the NTSB Materials Laboratory, Washington, D.C. The bracket assembly was fractured in the area of the outermost eyehole, at the point a bolt passes through the bracket assembly and the propeller governor arm. The fracture surface contained small amounts of dirt, grease, and minor corrosion. The fracture surface features include flat areas that lie on multiple planes separated by ratchet marks, features typically left behind by the propagation of a fatigue crack. The fatigue crack emanated from multiple origins on opposite sides of the bracket. The total area of the fatigue crack occupied approximately 85 percent of the fracture surfaces. The fatigue fractures initiated on the outer edges of the surface and propagated inward toward the center. The remaining 15% of the fracture surface had features consistent with overstress separation. Near the middle of each fatigue region were microfissures suggesting that the crack propagated under high-stress conditions. The NTSB Materials Laboratory also examined the separation point between the left propeller control flexible cable and the rigid rod that connects to the bracket assembly. The cable and the swaged part of the rigid rod were in good condition with no fractures or damage. The Piper PA-31-T2 Pilot Operating Handbook, Section 3, Emergency Procedures, does not contain a procedure for loss of propeller control. Section 3 did contain a procedure for "Over speeding Propeller", which stated that if a propellers speed should exceed 1,976 rpm, to place the power lever of the engine with the over speeding propeller to idle, feather the propeller, place the engine condition lever in the stop position, and complete the engine shutdown procedures. Pilot logbook records show the pilot completed a simulator training course for the accident model airplane about 9 days before the accident and had about 13 flight hours in the Piper PA-31-T2.

Shortly after takeoff at less than 200 feet above ground level the pilot reported an engine failure, and requested to return to the airport. The controller saw the airplane in a right turn, descending, and observed the airplane level its wings just prior to impact with the tips of trees. The airplane collided with terrain approximately 2,000 feet east of the approach end of runway 20L. A post crash fire ensued and consumed a majority of the airplane. Examination of both engines displayed contact signatures to their internal components characteristic of the engines being powered, with the propellers out of feather at the time of impact, and a low power range. Examination of the propellers found the left propeller blades showed more damage then the blades from the right propeller. Both propellers were rotating with considerable rotational energy. However, examination showed that the left propeller had more power then the right.

The first officer failed to maintain the proper rate of descent (sink rate) resulting in a hard landing on touch down, and separation of the left main landing gear during landing rollout. The pilot-in-command stated he knew the first officer was not going to make a good landing. He did not take any corrective action other than informing the first officer initially to increase power. Examination of the left main landing gear assembly revealed a preexisting crack in the outer cylinder housing.

The flight was in radio contact with air traffic controllers at the FAA Memphis Air Route Traffic Control Center (ARTCC), and was level at 9,000 feet, when the pilot checked in on frequency. The last radio contact with the flight was at 0447. At 0520, radar contact was lost. The airplane impacted in rough terrain, at a steep angle of impact. Two other company pilots flying in trail of the accident aircraft said, they had radio contact with the pilot of N840FE about 5 minutes before the accident. They said he sounded fine and did not say anything about any problems. Radar data showing the flight's ground speed indicated that at 0512, the ground speed was 158 knots (182 mph). The ground speed then decreased to 153 (176 mph), 143 (165 mph), 138 (159 mph), and 132 (152 mph), until at 0519:40, when the ground speed of the flight was 125 knots (144 mph). After the radar read out at 0519:40, the next radar hit was coast (no information), and then the flight disappeared from the radar scope. The airplane's heading and altitude did not change during the decrease in ground speed. According to the NTSB Radar Data Study, calculated flight parameters indicated the airplane "...experienced a slow reduction of airspeed in the final 8 minutes of flight at altitude, and then abruptly exhibited a sharp nose down pitch attitude with a rapid increase in airspeed." About the time of the reduction in airspeed, pitch angle began to slowly increase also. When radar contact was lost, the calculated airspeed had reduced to less then 102 knots [118 mph], and calculated body angle of attack [AOA] had increased to 8.8 degrees. A large reduction in pitch angle, angle of attack, and flight path angle as the airspeed increases after peak AOA was reached. Examination of the engine Power Analyzer and Recorder (PAR) revealed that no exceedences were in progress at the time power was removed from the PAR. It was determined that no caution timing events were in progress. The PAR computer appeared to be operating correctly until power was removed at impact. Examination of the airplane's autopilot were not conclusive due to impact damage. Determination of whether the autopilot was engaged or not engaged at the time of the accident could not be determined. The NTSB Meteorological Factual Report revealed that at 0515, about 7 minutes before the flight was lost on radar, the radiative temperature in an area centered at Clarksville (4 kilometer resolution data), showed that the Mean Radiative Temperature was -6.26 degrees C (21F). The Minimum Radiative Temperature was -6.66 degrees C (19F). The Maximum Radiative Temperature was -6.06 degrees C (21F). According to the Archive Level II Doppler weather radar tape for a beginning sweep time of 0508:10, showed that N840FE had tracked into a weather echo from 0510:34, to 0516:28. The Doppler Weather Radar data, revealed that N840FE, had entered a weak weather echo about the same time that the airspeed of the airplane started to decrease, at an altitude of about 9,000 feet, and the airplane was in the weak weather echo for a few minutes. Based on the weather data, it was determined that in-flight airplane icing conditions were encountered by N840FE. Cessna Aircraft Company Airworthiness Directive (AD) 96-09-15; Amendment 39-9591; Docket No. 96-CE-05-AD, applicable to this airplane and complied with by the company, on December 12, 1996, stated: "...to minimize the potential hazards associated with operating the airplane in severe icing conditions by providing more clearly defined procedures and limitations associated with such conditions... operators must initiate action to notify and ensure that flight crewmembers are apprised of this change...revise the FAA-approved Airplane Flight Manual (AFM) by incorporating the following into the Limitation Section of the AFM. This may be accomplished by inserting a copy of this AD in the AFM...." The airplane was equipped with leading edge deicing boots on the wings, elevators, struts, and had a cargo pod deicing capability. Lights were installed to illuminate the leading edge of the wings, to aid the pilot in detecting ice on the leading edges of the wings during night operations. The airplane was not equipped with an ice detection device.

According to the pilot, upon landing in heavy rain, the airplane began hydroplaning. He said the airplane departed the left side of the runway, striking trees, which damaged both wings and collapsed the landing gear. A witness stated the airplane touched down in moderate rain.

The PIC stated he was established on an instrument approach when the left engine fuel pressure dropped to zero and the engine quit. He moved the fuel selector to the right rear fuel tank and the engine started. He continued the approach for about 2 miles when the right engine quit followed by the left engine. He made a forced landing to a sandbar. Examination of the airplane revealed the fuel tanks were not ruptured and the fuel tanks were empty.

The pilot reported that as he overflew the Tri-City airport, Tennessee at 9,000 feet, the right engine power slowly decreased. The airplane was in the clouds with an outside air temperature of about zero degrees Celsius. Right engine manifold heat was applied for a few seconds, with no noticeable difference in engine performance. The right magnetos were also checked with no obvious malfunctions noted. A descent for landing was initiated to the Tri-City airport and manifold heat was applied and removed several times, with no appreciable effect. On short final approach, when the landing gear was extended, the right main gear did not indicate down. There was insufficient time to perform the emergency gear extension procedure. A go-around was initiated, during which the pilot discovered that the left manifold heat control was now stuck in the 'ON' position. The pilot stated that with less than full power available on the left engine, and the right propeller unfeathered, the airplane could be climbed to about 200 feet. He flew the airplane until terrain clearance was no longer possible, then landed in a field, gear up. The airplane slid into trees and was substantially damaged.