N151TM – Papa 51 Ltd., Co. THUNDER MUSTANG – 1998-05-30 NTSB Accident Report

HISTORY OF FLIGHT

On May 30, 1998, approximately 1245 mountain daylight time, an experimental Papa 51 Thunder Mustang, N151TM, impacted the terrain about six miles south of Marsing, Idaho. The commercial pilot and his passenger, who held a private pilot certificate, received fatal injuries, and the aircraft, which was owned and operated by Papa 51 Ltd., of Nampa, Idaho, was destroyed. The 14 CFR Part 91 personal pleasure flight, which was being conducted to demonstrate the performance capabilities of the aircraft to the private pilot, departed Nampa Municipal Airport, Nampa, Idaho, about 20 minutes prior to the accident. No flight plan had been filed, and according to witnesses, the aircraft was being operated in visual meteorological conditions. There was no ELT transmission as the ELT had been removed for repair.

According to a company representative, the private pilot rated passenger had invested in Papa 51 Ltd., about four months prior to the flight, and this was the first opportunity to demonstrate the Thunder Mustang's performance and flight characteristics to this individual. About 1220, the company demonstration pilot and the investor departed Nampa Airport. Approximately the same time, a Papa 51 company official, piloting a Glassair III, took off from Nampa and joined up with the Mustang in formation flight west of Nampa Airport. After maneuvering together for about 5 minutes, the company official in the Glassair told the two pilots in the Mustang that he and his passenger were going to return to the airport. As the Glassair pilot turned east toward Nampa, the Mustang pilot turned to proceed to the west. Soon after the two aircraft separated, the demonstration pilot in the Mustang advised the company official flying the Glassair that the alternator light on the Mustang had just illuminated, but that they would complete the demonstration flight before returning to Nampa. According to the company official in the Glassair, since the Mustang had been flown before using battery power only after a loose wire had rendered the alternator inoperative, there was no further discussion of the illumination of the light. Except for the failed alternator indication, there had been no indication from the demonstration pilot that there was anything abnormal or unusual about the flight. In addition, neither of the individuals in the Glassair had noticed anything that would indicate that there was something wrong with either the Mustang or its pilot.

Soon after the two aircraft separated, a number of individuals noticed the Mustang performing a series of maneuvers in an area about five miles south of Marsing, Idaho. A number of these witnesses reported that when the aircraft was first seen, it was approaching the area from the direction of Marsing, and was traveling in relatively level flight at a high rate of speed. Once in the area near where the accident took place, the aircraft was seen climbing, descending, turning at varying degrees of bank angle, and maintaining level flight for brief periods of time. The witnesses reported that during this series of maneuvers, there were times when the engine sounded like it was running smoothly at a high power setting, times when the engine sounded as if it were "popping", "sputtering" or "backfiring" at a reduced power setting, and times when it could not be heard at all. During this period, which most witnesses indicated lasted between five and ten minutes, no smoke or fluids were seen trailing the aircraft, and no one reported seeing any structural components separate from the airframe.

Almost all of the witnesses reported that a short period of time before the impact, they heard "backfiring" and "popping" sounds, as they had during some of the previous maneuvers. This was followed by another period during which no engine sounds could be heard, except for an occasional backfiring sound heard by some of the witnesses. Individuals who heard this period of zero or reduced engine sound, gave estimates of a duration from 30 seconds to over a minute. According to the few witnesses who were still watching the Mustang during this last period of reduced engine sound, the aircraft appeared to climb for a short period of time after the reduction of engine noise, and then began descending while heading to the south toward slightly rising terrain. A number of witnesses lost sight of the aircraft as it proceeded to the south, but two witnesses reported watching the Mustang until it impacted the ground. According to one of these witnesses, during the final portion of its flight, the aircraft's southerly course paralleled Highway 95, about one mile west of the highway. This individual reported that he watched the aircraft as it passed about 1.5 miles east of his residence and proceeded to the south at high speed. He said that it flew over a ridge that runs between his home and Highway 95, and that the aircraft appeared to be between 1,000 and 2,000 feet above the ground. He also commented that when he initially heard the engine it was running fine, but as it continued to the south, it sounded to him like it "backfired" and then was silent. He reported that the aircraft began to slow as it proceeded south in level flight, and then its nose dropped sharply, and the aircraft appeared to lose a few hundred feet. The aircraft then returned to level flight and continued to the south. The witness said that during its return to level flight, the engine might have accelerated a little bit, but he was not certain. As it once again continued to the south, its speed over the ground appeared to slow significantly, and according to this witness became "very slow." When asked for an estimate of its speed, the witness stated that it looked to him as if the aircraft was going 35 to 40 miles per hour. As the aircraft approached the location where it ultimately struck the ground, the nose was once again seen to drop, followed by another loss of altitude. One of the two witnesses said that there was no engine sound after this second nose drop, and that during the altitude loss, the aircraft rolled left and descended into the ground. The other witness said that it sounded as if the pilot "got back on the gas" for a short period of time after the nose dropped, followed immediately by the initiation of a left turn. This witness further stated that while the aircraft was in the turn, it sounded to him like the engine "stopped altogether," and then the aircraft "...gently rolled into the ground." One of the witnesses reported that after the second nose drop, the aircraft was about 75 feet above the ground, and the other described it as about 2 to 2 1/2 times the height of the nearby power line towers. Both witnesses said that immediately upon contacting the ground, the aircraft was engulfed in a ball of fire.

METEOROLOGICAL INFORMATION

At 1215, about five minutes prior to the time the two aircraft departed Nampa Airport, the Aviation Routine Weather Report (METAR) taken at Caldwell, Idaho, which is located about 13 miles north-northeast of the accident site, reported calm winds, 10 statute miles visibility, scattered clouds at 3,100 feet and 3,700 feet, broken clouds at 5,000 feet, a temperature of 13 degrees Celsius, and a dewpoint of 6 degrees Celsius. At 1254, about 10 minutes after the accident, the METAR from the same location indicated calm winds, 10 miles visibility, scattered clouds at 3,800 feet and 4,700 feet, broken clouds at 6,000 feet, and a temperature/dewpoint of 13 and 7 degrees Celsius. The METAR taken about ten minutes after the accident at Boise Air Terminal, located about 30 miles east of the accident site, reported winds variable at 4 knots, 10 miles visibility, few clouds at 3,200 feet, scattered clouds at 4,000 feet, broken clouds at 6,500 feet, and a temperature/dew point of 13 and 6 degrees. According to the pilot of the Glassair, there was a broken to overcast ceiling of about 2,000 to 2,500 feet (AGL) in the area where the two aircraft separated, and it appeared to him that the ceiling became somewhat lower over the rising terrain southwest of Marsing. According to the witnesses in the area of the accident, there were "low" clouds in the area, with the height of the "overcast" varying throughout the general area, but generally becoming lower toward the hills. All agreed that the visibility underneath the clouds was very good.

WRECKAGE AND IMPACT INFORMATION

The initial impact was about one mile north of Wildcat Spring, and one mile west of Highway 95, at 43 degrees, 28.01 minutes North, 116 degrees, 53.10 minutes West. The terrain was covered with short dry grass and sparsely distributed 10 to 18 inch high brush. It was relatively flat, and sloped downhill to the north at less than five degrees. Aircraft system components and pieces of the structure were scattered along a wreckage distribution track, which ran on a heading of about 45 degrees magnetic, for a distance of approximately 225 feet (see Wreckage Distribution Diagram). There was a fan-shaped burned area, typical of that resulting from the ignition of fuel spray/mist dispersed by the force of the impact. This pattern extended approximately 150 feet down-track from the initial impact point, and much of the aircraft structure found within the pattern had been partially consumed by fire. Some sections of the aircraft were outside of this burn area, and were not damaged by the post-crash fire.

After the completion of the on-scene portion of the investigation, the wreckage was taken to a secure hangar at Nampa Airport, where further teardown inspections of the aircraft structure, system components, and engine were performed. All primary aircraft structure was present, and the elevator, rudder and aileron counterweights were accounted for. Both the landing gear and flap selectors were found in the up position, and the actuators for the landing gear and flap systems were consistent with those components being retracted. The cooling/radiator vent door was found in a partially open position, and there was no indication of any pre-impact anomalies in the flight control torque tube/hardware system. The throttle and propeller control handles were loose in their mounts, and the fuel selector valve position could not be determined because of impact damage. There was no indication of smoke or soot on any of the components which had come from the cockpit area but had been thrown clear of the post-crash fire. The inside surface of pieces of the canopy found outside of the post-crash fire also showed no sign of pre-impact fire damage. Both ailerons had been torn from the wings at their hinges, and the position of the aileron trim tab, which was still attached to and free to move in the right aileron, could not be determined. The rudder trim tab had been torn from the rudder at its hinges, but the elevator trim was still attached, and its trailing edge was approximately 1/4 inch higher than the trailing edge of the elevator. The elevator had been torn from the horizontal stabilizer at the hinges, and one of the elevator counterweights had been torn from the elevator, but was present at the scene. There were no indications of any flight control surfaces vibrating or fluttering past their respective limit stops, and no evidence was found of any pre-impact structural anomalies. Three of the four wood core, composite propeller blades had sheared where they protruded from the propeller hub, and the innermost portion of their span was still present within the hub. The fourth blade, which was severely burned, was still mounted in the hub. The propeller hub itself was still attached to the reduction gear box main drive gear. One of the separated blades had been thrown clear of the post-impact fire, and showed clear chord-wise scarring on the inboard 1/4 of its span. It had two easily seen partial fractures, one of which ran from its root to within about ten inches of its tip, and another that ran chord-wise about eight inches in from the tip. The severely burned remains of one of the other separated blades were found near the engine, and the fourth blade was not positively identified.

The Falconer V-12 fuel injected racing engine and the Papa 51 reduction gear box were subjected to full disassembly and inspection. A portion of the lower front of the gear box housing fractured during the impact, but the gear assembly remained intact within the housing. The quill shaft, spur gear, and main propeller drive gear were removed from the gear box and inspected, and showed no evidence of unusual wear, overheat, or any other anomaly. The left and right bank cylinder heads were removed from the engine block, and the valves, rocker arms, and push rods were removed from the heads and inspected. All twelve piston/rod/rod bearing assemblies where removed from the engine block and also subjected to inspection. In addition, the crankshaft, camshaft, timing gear and timing chain, and all six main bearings and the thrust bearing were removed and inspected. Except for a small area of copper smearing where the crankshaft had pushed the thrust bearing into its journal during the impact, none of these components showed any indication of abnormal wear, lack of lubrication, heat stress, misalignment, or other anomaly. All the spark plugs were removed and showed no signs of contamination, abnormal lead buildup, or unusual wear. The induction plenum, induction manifold, and injector orifices were found to be clear and clean, and the hydraulic and oil filters showed no sign of contamination or blockage. The engine driven fuel pump produced a suction when rotated, and the electric fuel pump, which had been distorted by impact damage, passed through a few degrees of rotation when electrical power was applied to its terminals. The propeller governor was taken to Precision Propeller, of Boise, Idaho, for a teardown inspection, and no evidence of pre-impact malfunction or anomaly was found.

TESTS AND RESEARCH

The two Motec electronic ignition engine control units (ECU's) were removed from the aircraft and the data stored within them was retrieved using a computer driven data recovery program provided by the United States distributor of the ECU's (J.G.M., Inc., of Huntington Beach, California). The data recorded by the ECU's had been gathered approximately once each second from engine start to the instant of impact (see ECU Data Chart), and included the following parameters:

a. Elapsed time. [Time] (in minutes and seconds). b. Engine speed. [RPM] (in revolutions per minute). c. Throttle position. [TP %] (in percent of full throttle). d. Manifold pressure. [MAP kPa] (in kilopascals). e. Engine coolant temperature. [ET C](in degrees Celsius). f. Ambient air temperature. [AT C] (in degrees Celsius). g. Main battery voltage. [BAT V] (in DC volts). h. Oil temperature. [AUX T] (in degrees Celsius). i. Engine load fuel metering. [EP %] (in percent of full). j. Fuel injector pulse duration. [APW %] (in milliseconds). k. Fuel injector duty. [DTY %] (in percent of capability).

Each of the data entry points from the time of power application at takeoff to the moment of data collection termination was reviewed by the investigative team, and selected data from the last five minutes of the flight was graphed on an elapsed time chart using Microsoft Exel (see Thunder Mustang Last Five Minutes graph). In addition, selected data from the last 117 seconds of the flight was graphed in an elapsed time sequence which allowed finer scrutiny of the relationship between data points (see Thunder Mustang Last Two Minutes graph).

According to the ECU's for both left and right bank cylinders, the throttle was moved from a taxi-level setting (around 20%), to the full forward (100%) position approximately 20.7 minutes prior to the time at which data collection terminated. As the throttle was moved to that position, where it remained for about 34 seconds, the engine speed increased from around 1,800 rpm to about 5,100 to 5,200 rpm. As both the throttle position and the engine rpm increased, the engine load (EP), injector pulse duration (APW), and injector duty (DTY) all followed the throttle/RPM increase in a fairly linear manner. According to both ECU's, the EP increased from around 20% to about 98 to 100%, the APW increased from about 3 milliseconds (mS) to about 7 to 8 mS, and the injector duty increased from around 10% to 60 to 70%. About one minute and twenty seconds after the full power application for takeoff, the left ECU indicated that the throttle was reduced to about 63% and the engine rpm dropped to around 4,250. The right ECU indicated approximately the same RPM, APW, and DTY, but its throttle position reading and EP reading, which takes into account throttle position, read approximately nine percent higher than that of the left ECU. This difference, according to both the engine and airframe manufacturers, is caused by a base-line measuring error which resulted in the throttle position reading on the right ECU being approximately nine percent greater than its actual position. This difference was constant throughout the flight, except when the throttle was in the full-power position, where the RPM and EP readings are limited by the reading limit of 100%.

The throttle position remained around 63% for about two minutes, and then there was a two and one-half minute period where the throttle was moved to various settings between 20% and 80% (although during most of this time the throttle settings were in the 40's). At the end of this period, the throttle was advanced to and remained in a range of 62% to 67%. It was operated in this range for the next twelve and one-half minutes, and then was reduced to the 0% (idle) position over a period of about eight seconds. The throttle remained at idle for the next minute and forty-five seconds, and over a period of about 35 seconds, the RPM gradually dropped to about 2,200 RPM. With the throttle still at idle, the RPM jumped back up to about 3,100 RPM for about five seconds, and then gradually dropped over a 20 second period to about 2,100 RPM. Then once again, the RPM rapidly increased, this time to a high of 3,900 RPM. Almost immediately the RPM once again began to drop, and about 23 seconds later had returned to the 2,100 RPM range. Then about five seconds before the end of data collection, the throttle was pushed forward to about 72% and the engine RPM climbed to just below 4,100 RPM. This throttle movement took place over a period of about four seconds. During the four second throttle push-up, the RPM, EP, APW, and DTY all followed in a linear manner as they had during the initial push-up at the beginning of the flight. Then about one second before the end of data collection, the throttle was retarded. The last throttle position recorded by the left bank ECU was 62%, and the last recorded by the right bank was 13% (idle). As the throttle was reduced during the last second of data collection, all four of the aforementioned parameters dropped in value. It was also noted that during the minute and forty-five seconds the throttle was at idle, the manifold pressure (MAP) slowly climbed, with minor fluctuations, from around 82 or 83 kilopascals (kPa) to around 89 or 90 kPa. According to the aircraft manufacturer, because of the location of the pressure sensor in the induction plenum, this increase of pressure is a direct reading of the change in the ambient pressure as the aircraft descended.

A review of the data also revealed that just over one minute after the throttle was advanced for takeoff, the battery voltage dropped below 12 volts, the level at which the alternator warning light comes on, and remained there (between 10.8 and 11.9 volts) for the remainder of the flight (except for one spike to 12.2 volts as the throttle was moved forward about four seconds before data termination). Additional data was collected by an auxiliary data terminal on the right ECU that had been set up to collect engine oil temperature information. It showed that throughout most of the flight, the oil temperature ranged between 110 and 114 degrees Celsius (230-240 Fahrenheit). Data also indicated that although the oil temperature had dropped to a range of 20 to 30 degrees Celsius (68-85 degrees Fahrenheit) during the minute and forty-five seconds the engine was at idle, at the termination of data collection, the temperature was recorded at 92.5 degrees Celsius (198.5 degrees Fahrenheit).

During the investigation, the Vision Micro Systems VM 1000 engine monitoring system was recovered from the wreckage and taken to Vision Micro Systems Inc., of Bellingham, Washington, for retrieval of the stored data. The RAM data taken from this unit represents only the information being processed at the moment power to the system was terminated. The recovered data indicated that 28 minutes and 15 seconds had elapsed from the time of engine start to the moment when power to the unit was terminated (see Vision Microsystems Data chart). The unit also indicated that the elapsed time from the application of takeoff power to the last moment of data collection was between .30 and .39 of an hour (18 to 23 minutes). The data further showed that although the maximum recorded battery voltage during the flight was 13.22 volts and the minimum was 10.84 volts, at the time of data termination the voltage was about 11 volts (10.96). Also, according to this system, although the last alternator output recorded was zero amps, at some time during the flight, the output had reached as high as 44.63 amps. The final data also revealed that the oil pressure was 55 pounds per square inch, the oil temperature was 200 degrees Fahrenheit, and the fuel flow was 30.5 gallons per hour. The final data also showed that the exhaust gas temperatures from nine of the twelve cylinders were between 430 and 630 degrees Fahrenheit during the last reading. According to the technician at Vision Micro Systems, the higher temperature readings from the exhaust of the other three cylinders (746, 835, and 1,198 degrees), were the result of the shorter elapsed time (in milliseconds) since the last time these cylinders had generated an exhaust pulse.

MEDICAL AND PAHTOLOGICAL INFORMATION

An autopsy was performed on each of the pilots by the Owyhee County Coroner, and the results of each indicated that the cause of death was accidental due to multiple blunt trauma from the aircraft impact. In addition, toxicology examinations were completed on both of the pilots by FAA's Toxicology and Accident Research Laboratory, and no alcohol was detected in the urine of the demonstration pilot, nor in the muscle or liver of the second pilot. Also, no drugs were detected in the urine of the demonstration pilot. An examination of the second pilot's liver and kidney confirmed 0.075 ug/ml of Diphenhydramine in his liver fluid and 0.08 ug/ml of Diphenhydramine in his Kidney fluid. According to the FAA laboratory, Diphenhydramine is commonly found in over-the-counter cold medications, and could cause some degree of drowsiness at the levels indicated.

ADDITIONAL DATA AND INFORMATION

During the investigation it was determined that twice before, after the illumination of the alternator light, the aircraft had flown using battery power only. During one of the flights, the aircraft ran on battery power for about 45 minutes while en route to Nampa, Idaho, from California. During the other flight, which was a return flight from Florida to Nampa, the aircraft flew on battery power for approximately three hours. According to the airframe manufacturer, the reason for both alternator failures was a wire that had broken due to vibration. Since that time, an expansion loop had been added to the wire to reduce the chance of future failure due to vibration. According to the engine manufacturer, the coil pack can produce a spark as long as the battery provides at least seven volts, and the ECU's can monitor/control engine performance down to a low of six battery volts. It was also noted that in addition to the 12 volt main battery, the aircraft also carries a 12 volt backup battery (although of lesser ampere-hour output).

During the investigation, it was also noted that, according to the airframe manufacturer, when the aircraft's throttle is reduced to idle in flight, the combination of short, direct exhaust stacks (no muffling system) and an intentionally enriched fuel mixture, produces a considerable amount of loud "backfiring."

On July 6, 1998, the aircraft wreckage was released to Dan Denny, CEO of Papa 51 Ltd., at Nampa Idaho.