N268AFBeech A45 · 1961

On June 7, 2024, about 1140 Pacific daylight time, a Beech A45 airplane, N268AF, was destroyed when it was involved in an accident near Deer Harbor, Washington. The pilot was fatally injured. The airplane was operated as a Title 14 Code of Federal Regulations part 91 personal flight.

The airplane was based at Skagit Regional Airport (BVS), Burlington, Washington, and stored in a museum facility founded by the pilot. The pilot’s son, who was working at the museum on the day of the accident, reported that the pilot arrived that morning and appeared to be in good spirits. He stated that he was going to perform an “Orcas sortie,” a phrase the pilot used to describe a flight around the San Juan Island archipelago, with a touch-and-go landing at Orcas Island Airport (ORS), Eastsound, Washington, and a pass over his previous home on Orcas Island.

The pilot boarded the airplane about 1050, and at 1100, a family friend (who ultimately witnessed the accident) received a text message from him indicating that he would be flying past her house, which was located close to his old home on the western shore of Orcas Island, about 1140. The friend reported that such flights were not unusual, and this was the first time he had performed one this year. She stated that he typically performed two flybys, and although sometimes he rocked the airplane’s wings as he passed by, he had never performed an aerobatic maneuver.

Limited ADS-B data was available for the flight; however, data from nearby radar facilities showed a target that departed BVS about 1113 transmitting a 1200 beacon code with no altitude information. The target flew a westerly track toward ORS that generally followed the shorelines. At 1125:32, the target was aligned with a final approach for runway 34 at ORS, and about 4 minutes later, the target had departed the airport area and was tracking southwest (see figure 1).

Figure 1 – Radar data flight path

Over the next 9 minutes, the target flew a series of passes back and forth along the coastline of the San Juan Channel, and by 1137:30, it was heading north along the west coast of Orcas Island (see figure 2).

Figure 2 – Final segment of radar data flight path (data is approximate, with variations potentially exceeding 1,000 ft)

The friend reported that, about this time, she could hear the familiar sound of the pilot’s airplane, and a short time later she observed the airplane overhead traveling north along the shoreline in front of her house.

The airplane then flew behind trees and out of view, and she could hear what sounded like the airplane making a turn to the south. The airplane then came into view as it rolled out on a southerly heading, flying over the water at a higher altitude than the previous occasions. After it passed by, she could see the left wing drop, and she thought this was part of his usual routine; however, the wing continued to drop as the airplane began to rapidly descend toward the water.

Another witness, who was standing in the same location, commented that the airplane was diving toward the water in a sweeping arc. She recalled that the transition from level flight was so graceful and smooth, and the maneuver looked so perfectly executed, that it was reminiscent of something she might see at an airshow.

A witness who was located about 1/3 mile north of the two other witnesses along the same shoreline heard what he thought was a vintage airplane. He stated that this was not unusual for the area, and he went out onto his deck to look for it. He was facing west as it came into view, and he began to record it with his phone. At the beginning of the recording, the airplane was inverted with a slight nose-down attitude and heading generally to the south. Over the next 3 seconds, the airplane had transitioned to an almost vertical dive. As the airplane approached the water, it began to pull out of the dive, now facing the opposite direction. By the time it had recovered to an almost wings-level, upright attitude, the belly impacted the water and the airplane bounced across the water on a northerly trajectory.

The pilot, who was aged 90 at the time of the accident, had an extensive and distinguished career in aviation, including service as an astronaut and Air Force pilot, followed by roles as a United States Ambassador, test pilot, and senior positions within the public and private sectors.

By 1996, he had retired and founded the Heritage Flight Museum. According to family members, his plan was to create a legacy of historic airplanes, with a P-51 he owned being the first in the collection. The collection grew to 18 aircraft, 15 of which were in flying condition. The pilot flew all the airworthy airplanes, including the P-51, which he flew regularly at airshows until 2006. The airshow work generally comprised heritage flights, though the pilot held a “surface level” aerobatic competency card with approval for “rolls only.” He did not maintain that card’s currency after retiring from airshow flying.

According to the pilot’s son, who was a flight instructor and conducted his flight reviews, the pilot was “fastidious” in his preparation and always wore a parachute when performing aerobatics. Throughout his career, the pilot always flew from the pilot/primary seat, and this remained the case as he aged.

By 2015, the pilot voluntarily stopped flying higher-performance airplanes such as the P-51 and T-6 out of concern for damaging them. He shifted primarily to simpler, seasonal flying in the T-34 (the military designation for the accident airplane make/model), the airplane in which he originally learned to fly. He purchased three T-34s, one of which he kept near his home in the San Diego area. However, he did not like to fly it in Southern California, because he did not like the airplane’s communication system and the congested airspace. The decision was eventually made that he would not fly solo in that area.

The pilot’s son stated that he still had impeccable flying skills, that he was excellent in formation flight, and was particularly comfortable in the T-34. His only concern was that, due to the pilot’s background, he tended to fall back into military flying practices and could be impatient with other aircraft in the traffic pattern. He no longer performed aerobatics; however, he did perform a barrel roll in the T-34 the week before the accident during a flight review conducted by his son. They were not wearing parachutes because they had not planned to include aerobatic maneuvers in the flight; however, he was flying so well that they decided to, and his son reported that it was performed flawlessly.

His son stated that the pilot did not perform low-level maneuvers and would have been well aware of the minimum altitudes required to perform aerobatic maneuvers. He stated that the pilot had recently performed a low pass over a friend’s ranch in the mountains near Ramona, California, but that was the extent of his low-level flying activities.

Family members indicated that the pilot was in good health for his age, but his son and a family friend reported that in recent years, he had displayed slowed and reduced mobility. The pilot always insisted on entering the airplane unaided, but his son always monitored and followed him closely as he climbed the ladder into the airplane. The pilot used to wear a flight suit when flying, but had not for a while and had become more relaxed and generally wore sweatpants and a sweatshirt. For the accident flight he also appeared to be wearing casual sheepskin-style moccasin shoes. In the past, he had insisted on wearing a personal flotation device when flying over water, but had become more lenient and was not wearing it for the accident flight.

Review of the pilot’s logbook, the last entry of which was dated June 1, 2024, indicated that he performed 23 flights in the year leading up to the accident, but took a break from flying between November 17, 2023, and May 31, 2024. Except for one flight, which was in the T-34B in San Diego, all other flights were in the accident airplane, and flown from either BVS, ORS, or Anacortes Airport (74S). Although not noted in his logbook, ADS-B data indicated that the airplane also flew locally on June 5 and 6, and the pilot’s son confirmed that he had flown those flights as part of the D-Day Commemoration.

The pilot’s most recent FAA medical certificate expired in 2014, and he subsequently flew under the provisions of BasicMed.

The airplane was manufactured in 1961 and purchased by the pilot in 2018.

It was approved for aerobatic maneuvers and had a design limit structural load factor of +6G and -3G. The flight manual provided a series of recommended entry speeds for the aerobatic maneuvers.

The airplane was equipped with two Garmin G5 electronic flight instruments, one configured as an attitude indicator, and the other as a horizontal situational indicator. Although these instruments could store flight parameters to a panel-mounted microSD card, cards were not installed in either unit. The airplane was also equipped with a tail-mounted mounted ADS-B out beacon. According to the pilot’s son, the beacon was operating intermittently, and review of the last two flights revealed that the transmitted ADS-B flight data was sporadic.

Although the pilot had a ForeFlight account, there was no evidence that he had accessed weather imagery on the day of, or the day before, the accident.

Visual meteorological conditions prevailed throughout the area of the accident flight, with clear skies and 9 miles visibility at ORS. The weather observation station at ORS reported an atmospheric pressure of 30.05 inHg 5 minutes before the accident.

The airplane was manufactured in 1961 and purchased by the pilot in 2018.

It was approved for aerobatic maneuvers and had a design limit structural load factor of +6G and -3G. The flight manual provided a series of recommended entry speeds for the aerobatic maneuvers.

The airplane was equipped with two Garmin G5 electronic flight instruments, one configured as an attitude indicator, and the other as a horizontal situational indicator. Although these instruments could store flight parameters to a panel-mounted microSD card, cards were not installed in either unit. The airplane was also equipped with a tail-mounted mounted ADS-B out beacon. According to the pilot’s son, the beacon was operating intermittently, and review of the last two flights revealed that the transmitted ADS-B flight data was sporadic.

The fuselage was located about 1,200 ft west of the last radar target, in about 30 ft of water.

The debris field was about 400 ft in length and oriented on a magnetic heading of about 320°. The located debris, which comprised the fuselage, empennage, both wings, the engine, and the propeller assembly, were found between depths of 30 to 45 ft. Underwater terrain depths increased to about 150 ft on the north side of the debris field. The canopy, nose wheel assembly, engine cowling, and right outboard wing skins and tip were not recovered.

Underwater dive video recorded by first response personnel confirmed that the pilot was in the front seat and secured with the lap and shoulder harness. He was not wearing a parachute or personal flotation device.

The forward portion of the fuselage sustained significant damage to the aft cabin; the aft fuselage and empennage were mostly intact. The left wing remained attached to the fuselage, and the recovered inboard section of the right wing remained partially attached by its aft spar.

The forward instrument panel was crushed and deformed and most of the flight instruments were impact-separated, leaving remnants of their cases and fitting screws attached to the panel. The altimeter was still partially attached to the panel, and its indicator needles had become dislodged internally. The Kollsman window showed a value of 30.05 inHg. The associated flight instrument air tubing behind the panel was significantly damaged, with multiple breaches and fitting separations, and the transponder was damaged and dislodged from the panel.

The fuel selector valve and both cockpit fuel selector handles were found in the “left” tank position. Both landing gear handles were found in the “UP” position and the landing gear actuator was found in the retracted position. The left flap and both flap actuators were found in the flaps retracted position.

Examination of the flight control system did not reveal any evidence of pre-accident malfunction or failure.

The engine was intact and internal drivetrain continuity was confirmed, with no evidence of catastrophic failure. Both propeller blades and the hub assembly remained attached to the engine crankshaft. The blades exhibited leading edge nicks, dents, and paint abrasion. One blade had bent about 90° aft while twisting about 180° along its rotational plane. The other blade sustained similar twist damage, and a 4-inch section of its leading edge had separated.

The Office of the Snohomish County Medical Examiner performed the pilot’s autopsy. According to the autopsy report, the pilot’s cause of death was multiple blunt force injuries. As a result of the extent of the pilot’s injuries, examination of the brain was not possible. Examination of the heart identified 60% narrowing of the left anterior descending coronary artery by plaque, with mild disease in the remaining coronary arteries. The remainder of the autopsy examination, including of the heart, did not identify other significant natural disease.

The FAA Forensic Sciences Laboratory performed toxicological testing of postmortem specimens from the pilot. Ethanol was detected in liver tissue at 0.01 g/hg and was not detected in cavity blood or muscle tissue. Loperamide was detected in cavity blood at 0.9 ng/mL and was detected in liver tissue. Rivaroxaban and acetaminophen were detected in cavity blood and liver tissue.

Ethanol is a type of alcohol. Alcohol consumption is not the only possible source of ethanol in postmortem specimens. Ethanol can sometimes be produced by microbes in a person’s body tissues and fluids after death. Postmortem ethanol production is made more likely by extensive traumatic injury and can cause an affected toxicological specimen to test positive for ethanol while another specimen from the same person tests negative.

Loperamide is an over-the-counter medication commonly used to treat diarrhea by slowing intestinal motility. Use of loperamide may cause drowsiness, and caution is advised when using this medication if driving or operating heavy machinery. The FAA over-the-counter reference guide for pilots states that loperamide is a “no go” medication, noting also that gastrointestinal illnesses may cause dehydration, cramps, and abdominal pain with increases in altitude. Loperamide levels measured in postmortem cavity blood cannot be reliably compared to reference ranges in living people.

Rivaroxaban is a prescription blood thinner medication. One of its common uses is to help prevent stroke in persons with atrial fibrillation. Rivaroxaban is not typically impairing.

Acetaminophen is an over-the-counter medication commonly used to treat minor aches and pains and reduce fever. Acetaminophen is not generally considered impairing.

Video Study

A video study of the phone footage recorded by one of the witnesses was performed by a specialist from the NTSB Office of Research and Engineering.

Analysis indicated that, as the recording began, the airplane was about 1,500 ft above the water and diving at a speed of about 339 ft/second (201 knots). The airplane’s speed remained relatively constant during the maneuver, and the airplane achieved an almost constant arc radius of 930 ft throughout. Extrapolation of the arc revealed that its highest point was about 1,800 ft agl, and at its lowest point passed about 30 ft below the water level (see figure 3).

The centripetal acceleration experienced by the pilot was estimated at 3.8g during the last 4.5 seconds before water impact, and factoring the vertical component of gravity, the value experienced by the pilot just before impact would have been about 4.8g.

A sound spectrum analysis indicated an engine speed of 2,120 rpm when the recording started, increasing to 2,261 rpm when the airplane impacted the water.

The airplane was intact throughout the video and was not trailing smoke or vapors.

Figure 3 - Estimated airplane trajectory