N52900North American SNJ-2 · 1943

On August 20, 2021, about 1236 eastern standard time, a North American SNJ-2, N52900, was destroyed when it was involved in an accident in Wilkes-Barre, Pennsylvania. The pilot was fatally injured. The airplane was operated as a Title 14 Code of Federal Regulations (CFR) Part 91 air show flight.

The accident airplane was operated as part of the Geico Skytypers Air Show Team, which was comprised of six North American SNJ-2 airplanes. The accident occurred during a practice flight for the Great Pocono Raceway Airshow.

The flight was initiated with a formation takeoff, which consisted of three, two-airplane elements (sections) with each element taking off at 15-second intervals. The accident pilot was leading the 2nd element, and the accident airplane was the No. 3 airplane in the formation. All communications were conducted for the flight by the flight leader in the No. 1 airplane, using the call sign “SKYTYPE 1.” At 1234:43, ATC cleared SKYTYPE 1 for takeoff from Runway 4. At 12:35:45, SKYTYPE 1 reported that they would be rolling momentarily. At 1237:44, SKYTYPE 1 advised the tower controller that they were coming around to land on Runway 4. ATC acknowledged SKYTYPE 1 and advised them they were cleared to land on any runway. At 1238:03, SKYTYPE1 advised the controller that they had an aircraft down at the departure end of runway 4.

According to the flight leader, after takeoff, about 100 feet above ground level, the No. 3 airplane made a left turn, the nose went down, and the airplane departed controlled flight. The smoke system then turned on, and the airplane impacted the ground.

According to the pilot in the No. 4 airplane, who was the accident pilot’s wingman in the 2nd section, they had just taken off, and he was sliding into position on the right side of the No. 3 airplane. Normally at that point, the accident pilot would look at him to check that his landing gear was up; but he did not. The pilot in the No. 4 airplane noted that the accident airplane’s landing gear was retracting and the airplane was starting to decelerate. He also heard, over the radio frequency that they used for communication between the airplanes, the accident pilot transmit, “Three’s got an emergency.” The No. 3 airplane turned to the left, and as it rolled through about 20° of bank, the airplane slowed further, and the bank continued to increase as the airplane descended. The smoke system then came on, the airplane impacted the ground, and a post-impact fire ensued.

A video provided by a witness on the ground captured popping noises similar to back-firing of the engine before the airplane turned left and pitched down to ground contact. No black smoke was observed emanating from the airplane before the accident.

At the time of the accident, the pilot was employed as a first officer for a 14 CFR Part 121 air carrier. The pilot was a graduate of the United States Air Force Academy and had served in the United States Air Force (USAF). During his time in the USAF, he flew the T-34C, T-6 (Texan II), C-5, C-21, and MC-12. He held type ratings for several transport category airplanes and also held a flight instructor certificate with ratings for airplane single and multiengine, instrument airplane, and glider. His most recent FAA first-class medical certificate was issued on January 4, 2021. On that date, he reported that he had accrued 7,006 total flight hours.

The accident airplane was a low-wing monoplane of conventional metal construction. It was equipped with retractable landing gear, wing flaps, and a two-blade, constant speed propeller. It was powered by a 600-horsepower, Pratt & Whitney R-1340-AN-1, 9-cylinder, air cooled, radial engine.

According to the Federal Aviation Administration (FAA) and airplane maintenance records, the airplane was manufactured in 1943, and had been modified from its original configuration by multiple alterations and additions, including conversion from a two-seat tandem configuration to a single-seat configuration, with a smoke oil tank taking the place of the rear seat.

The airplane’s most recent annual inspection was completed on February 4, 2021. At the time of the inspection, the airplane had accrued about 6,911 total hours of operation, and the engine had accrued about 439 hours of operation since its last overhaul.

The accident airplane was a low-wing monoplane of conventional metal construction. It was equipped with retractable landing gear, wing flaps, and a two-blade, constant speed propeller. It was powered by a 600-horsepower, Pratt & Whitney R-1340-AN-1, 9-cylinder, air cooled, radial engine.

According to the Federal Aviation Administration (FAA) and airplane maintenance records, the airplane was manufactured in 1943, and had been modified from its original configuration by multiple alterations and additions, including conversion from a two-seat tandem configuration to a single-seat configuration, with a smoke oil tank taking the place of the rear seat.

The airplane’s most recent annual inspection was completed on February 4, 2021. At the time of the inspection, the airplane had accrued about 6,911 total hours of operation, and the engine had accrued about 439 hours of operation since its last overhaul.

The airplane contacted the ground with its left wingtip first, then the nose, followed by the right wing. On impact, the airplane rotated to the left and came to rest on a 120° magnetic heading, just off the west side of taxiway B, about 185 ft south of intersection B5.

The airplane was severely fire damaged. During the impact sequence, the engine separated from its mounting location, was displaced to the right, and came to rest against the inboard right wing leading edge. The empennage had separated from the aft fuselage, the left outer wing panel had separated from its mounting location (the wing center section), and both the left and right ailerons were separated from their mounting locations. The landing gear and wing flaps were up, and the pitot tube was clear of obstruction.

Both fuel tanks were breached, and the top of the smoke oil tank was broken and melted away. The majority of the cockpit had been burned away. The hydraulic filter and hydraulic power control had separated from their mounting locations. The throttle and mixture controls were in the approximately mid-range position, and the propeller control was almost full forward. The engine control lock was burned away. The flight control lock was unlocked. The latching mechanism for the pilot’s five-point harness was in the closed and latched position.

Examination of the two-blade propeller revealed that one blade was bent back about 10° about 12 inches outboard of the blade root; the rest of the propeller blade was almost completely straight. The other propeller blade was partially curled aft starting about mid-span. Both blades displayed minimal leading-edge gouging and chordwise scratching.

Examination of the engine revealed that the starter motor was separated from its mounting location. The engine oil tank was also separated from its mounting location and was breached during the impact sequence. Trace amounts of oil remained within the tank. The oil cooler, though impact damaged, appeared functional, and the oil “Y” drain valve was separated from its mounting location. The spin-on oil filter was intact and contained engine oil; there was no metallic debris in the pleats of the internal filter material. The hand fuel pump had also been separated from its mounting location. Internal examination revealed that its screen was clear of contaminants. The air chamber mixing gate was open.

Some oil remained within the crankcase. Drivetrain continuity and thumb compression and suction were established on cylinder Nos. 2, 4, 5, and 9. Thumb compression and suction could not be established on cylinder Nos. 1, 6, 7, and 8 due to impact damage.

No oil was observed leaking from the blower, nor was any present in the intake or exhaust stacks. The carburetor was unremarkable; the carburetor floats were functional and were not leaking. The fuel pump was also functional.

Portable GPS

During the examination of the wreckage, a Garmin Aera 660 Portable GPS was recovered. An external examination revealed the device was missing its touch display and had sustained impact and moisture damage, rendering it inoperable. The nonvolatile memory chip was removed and read out was attempted; however, the chip was unresponsive using multiple read out and recovery methods and no data was recovered.

Fuel Facility Records

Review of fuel facility records indicated that the airplane had been fueled before the accident flight with 56.2 gallons of 100LL aviation gasoline. Further review of the fuel facility records indicated that the facility had completed all the required checks satisfactorily prior to dispensing fuel on the day of the accident.

Airplane Weight and Balance

Review of weight and balance information indicated that the airplane was within weight and balance limitations.

Break Out

According to the Formation Pilots Knowledge Guide, the purpose of a break out is to ensure immediate separation and to avoid a mid-air collision.

A wingman must break out of the formation if:

- He loses sight of his reference aircraft

- He is unable to rejoin or stay in formation without crossing directly under or in front of Lead

- He feels his presence in the formation constitutes a hazard.

- When directed to do so by Lead

The guidance continues that if a pilot has lost sight of the reference aircraft, clear, then break in the safest direction away from the last known position or flight path of Lead and other aircraft. One technique—look for blue sky and pull--is appropriate for many situations, but there may be conditions where a pilot would actually pull toward the ground, depending on the airplane’s attitude and relative location to the rest of the flight.

Aerodynamic Stalls

The Pilot’s Handbook of Aeronautical Knowledge (FAA-H-8083-25C), says in part, that an aerodynamic stall results from a rapid decrease in lift caused by the separation of airflow from the wing’s surface brought on by exceeding the critical angle of attack (AOA). A stall can occur at any pitch attitude or airspeed. The Handbook also states that:

Low speed is not necessary to produce a stall. The wing can be brought into an excessive AOA at any speed…

The stalling speed of an aircraft is also higher in a turn than in straight-and-level flight. Centrifugal force is added to the aircraft’s weight and the wing must produce sufficient additional lift to counterbalance the load imposed by the combination of centrifugal force and weight. In a turn, the necessary additional lift is acquired by applying back pressure to the elevator control. This increases the wing’s AOA and results in increased lift. The AOA must increase as the bank angle increases to counteract the increasing load caused by centrifugal force. If at any time during a turn the AOA becomes excessive, the aircraft stalls.

The accident airplane’s wings-level stall speed was around 74 mph with the wing flaps and landing gear retracted (referred to as a “clean” configuration). Review of a stall speed and load factor chart indicated that:

- At 45° of bank, the stall speed would have increased about 20% (About 89 mph).

- At 60° of bank, the stall speed would have increased about 40% (About 104 mph, which was above the wings-level clean stall speed).

- At 75° of bank, the stall speed would have increased about 95% (About 144 mph, which was above the 140 mph climb speed).

- At bank angles exceeding 75°, the stall speed would have increased to 100% or more (148 mph or more).

Magnetos

The airplane was equipped with two magnetos, which were driven by the engine and operated independently from the electrical system. In a magneto, a spinning magnet induces a large current and small voltage in the primary winding of a coil. A secondary winding then develops a small current and a large voltage which is then routed to the spark plugs. Even if the airplane electrical system fails, there should always be reliable ignition.

A way to shut off each of the magnetos is required. This is done through the P-leads on each magneto. The "P" in P-lead comes from the primary winding in the magneto's coil. To deactivate a magneto, the primary winding is grounded. An ignition (magneto) switch in the cockpit would open and close the P-lead circuits to a suitable ground.

The ignition switch had positions for “OFF” (off), “R” (right), “L” (left), and “BOTH” (both), to make the required connections:

When selected to “OFF”, both left and right P-leads were grounded and both magnetos were off.

When selected to the “L” position, the right P-lead was grounded.

When selected to the “R” position, the left P-lead was grounded. When selected to the “BOTH” position, both P-lead circuits were open, and both magnetos were on.

Preaccident Troubleshooting

According to the airplane operator’s Director of Maintenance (DOM), on August 19, 2021, the accident airplane arrived at AVP late due to weather. The pilot made a logbook entry after landing and performed a magneto check, stating that the engine was “banging” on the left magneto.

On the day of the accident, the DOM asked one of the other pilots to run the airplane for him so he could observe how the engine was running from outside the airplane. During the magneto check, the engine was “popping and banging” in the “L” position. The DOM then asked the pilot to shut down the engine. While at idle, the pilot conducted a P-lead check, as is the company’s standard operating procedure during engine shut-down. During the P-lead check, he observed that the engine seemed to turn off while passing though the “L” position on the way to the off position.

The DOM then asked the pilot to run the engine again so they could verify the left magneto position. The DOM climbed onto the airplane’s wing and observed the engine stop running while passing through the “L” magneto switch position. The DOM switched back and forth numerous times through the “OFF,” “L,” “R,” and “BOTH” positions, and verified that in the left magneto position the engine reacted identically to the OFF position. Then they shut down the engine.

Upon visual inspection of the back of the left magneto, the DOM discovered that the P-lead spring, located at the end of the P-lead insulator, was fully compressed, allowing the internal grounding spring, which grounds the magneto when the P-lead is removed, to make contact with the case of the magneto. He observed no other physical anomalies in the rear section of the magneto.

The DOM then unscrewed the P-lead and removed the insulator and P-lead spring. He checked for continuity of the wire to the switch with a multimeter, and for proper operation of the switch. Both functioned properly. He then replaced P-lead spring with a new one, re-installed the P-lead insulator back into the magneto, screwed down the cap, and connected the safety clip. The DOM verified that it applied ample pressure to move the internal grounding spring away from the case, then reattached the rear cover.

After finding the P-lead spring compressed on the left side, he decided to remove the cover for the right magneto to visually check the condition of the springs. The P-lead spring was extended and was pushing the internal grounding spring away from the case. The rear cover of the right magneto was then closed and secured. The DOM subsequently started the engine and performed an engine run-up and magneto check, which revealed no anomalies.

The DOM stated that, before the accident, the team’s engine start was uneventful, and engine run-ups and power checks were performed with no anomalies reported from any of the pilots before takeoff.

Pratt & Whitney Troubleshooting Guidance

According to the Pratt & Whitney Wasp (R-1340) Maintenance Manual (Chapter 5…Trouble Shooting), items listed in the published troubleshooting chart were presented with consideration given to frequency of occurrence.

Under the section on “Rough Running,” “Ignition” was the first item on the troubleshooting chart. The chart then listed the probable causes which could contribute to a rough running engine, such as:

Defective spark plugs

Dirty or glazed breaker points

Breaker out of adjustment

Fouled spark plugs due to prolonged idling

Moisture or oil in the magneto and/or distributor

Water in the ignition harness

Improper magneto timing

Faulty magneto internally

Defective ignition manifold

Defective sparkplug lead connectors

Magneto ground to cockpit switch connection partially grounded

The corrected action to be taken was listed in bold italics as “Apply continuity test.”

Under the section on “Engine Stops,” ignition was once again the first item on the troubleshooting chart and listed, “Short in the system” as a possible cause, in addition to the master switch or magneto switch being inadvertently shut off.

The corrected action to be taken was listed in bold italics as, “Check all wiring for security, breaks or chafing” and to, “Check all system components.”

Postaccident Testing of the Left Magneto

The input drive of left magneto would not rotate. The points looked serviceable with a gap of “8.” The coil was tested and found to be within operating limits of 5,100 ohms. The condenser was visibly damaged but tested serviceable.

Examination of the left magneto by the NTSB Materials Laboratory also revealed a crack in the threaded socket for the P-lead fitting, which was part of the aluminum upper case half casting. The crack ran from the top surface of the socket toward the interior and passed through a transverse hole near the top of the socket. The crack continued past the transverse hole further into the threaded socket. At the top surface of the threaded socket, just above the transverse hole, an indentation was present which was co-located with the crack. The indentation was consistent with impact-related damage. The top surface of the threaded socket exhibited an imprint consistent with the hexagonal shape of the P-lead fitting. This imprint was pressed into the top of the threaded socket. The interior surfaces of the threaded socket exhibited pulled and deformed thread crests, as well as areas where the threads were almost entirely missing. The fracture surfaces of the crack exhibited voids and an area with a cold shut casting defect in the aluminum. These features on the fracture surface were consistent with a low-quality casting, but were not indicative of a preimpact malfunction or failure.

Postaccident Testing of the Right Magneto

The right magneto rotated easily, indicative of weak magnetism. When the magneto was put on a test stand, the magneto would not perform at low rpm and would not fire on all points until it reached 180 rpm. Review of airplane maintenance records did not reveal when the magneto was installed on the engine.

An autopsy was performed on the pilot by Forensic Associates of Northeastern Pennsylvania. Cause of death was multiple traumatic injuries secondary to airplane crash.

Toxicological testing of the pilot was conducted at the FAA Bioaeronautical Sciences Research Laboratory, Oklahoma City, Oklahoma. The specimens from the pilot were negative for carbon monoxide, basic, acidic, and neutral drugs.