NTSB Factual Report

HISTORY OF FLIGHT

On August 31, 2014, about 1834 central daylight time, a Schafer Foxtrot 4, experimental, amateur-built, single-engine airplane, N729AR, was destroyed after impacting terrain during initial climb near Abilene Regional Airport (ABI), Abilene, Texas. The pilot and one passenger were fatally injured, a second passenger was seriously injured, and a 2-year old child sustained minor injuries. The airplane was registered to and operated by the pilot. Day visual meteorological conditions (VMC) prevailed and a flight plan had not been filed for the 14 Code of Federal Regulations (C.F.R.) Part 91 personal flight. At the time of the accident the airplane was departing ABI for a local flight.

The airplane had departed ABI runway 17R to the south and was about 400 feet above ground level (agl) when the pilot declared an emergency and reported to the tower controller that he needed to return. Witnesses saw the southbound airplane in a steep left bank and several witnesses heard the sounds of impact. The airplane was generally northbound when it impacted mesquite trees and terrain about one mile south of the airport. The impact resulted in the complete separation from the fuselage of the forward fuselage, engine and propeller. The wing was also completely separated from the fuselage and both wing fuel tanks were compromised. All three adult occupants were ejected or partially ejected. The child in the rear seat remained restrained by a 4-point safety harness and was removed from the wreckage by emergency responders who arrived quickly. There was a fuel spill at the scene, but no postimpact fire.

PERSONNEL INFORMATION

The pilot, age 35, was a career active duty officer in the United States Air Force (USAF). He was rated as a Senior Pilot and was then assigned on active flight status as a military instructor pilot. He held a Federal Aviation Administration (FAA) commercial pilot certificate with a rating in airplane multiengine land limited to center thrust, a rating in instrument airplane, and a rating in airplane single engine land limited to private pilot privileges. The pilot also held a current FAA instructor pilot certificate with a rating in only instrument airplane.

Complete copies of the pilot's personal flight records were not available for examination by the NTSB during the course of the investigation. The investigator-in-charge was able to review six selected pages of the pilot's personal logbook, the aircraft logbook, the engine logbook, FAA and USAF documents and other records. The pilot's flight experience on August 31, 2014, was estimated as a total of 2,594 hours of flight experience in all airplanes, with an assumed total of more than 2,000 hours in military airplanes. The pilot's flight experience in a Foxtrot 4 airplane was estimated as a total of 23 hours, with all of that experience being within the previous three months.

AIRCRAFT INFORMATION

The low wing, fixed tricycle landing gear, single-engine airplane, serial number (s/n) F-008, was powered by a 300-horsepower Lycoming IO-540-K1H3 fuel injected engine, s/n L-22038-48, which drove a McCauley model B3D36C431-C, 3-blade aluminum alloy controllable pitch propeller, s/n 040071.

The airplane had a cantilever low-wing with conventional tail surfaces, and a four-seat enclosed cockpit which was accessed through "gull-wing" doors on both the left and right sides of the front cockpit seats. The aircraft structure was made mostly from composites with the fuselage halves molded as single pieces. The kit manufacturer's website described the airplane as having an expected empty weight of about 1,750 pounds, a maximum gross weight of 3,200 pounds, and an expected stall speed in the landing configuration of about 58 knots.

FAA records show the airplane was completed by the pilot, with assistance from the kit-builder, as a kit-built airplane. On June 10, 2014, it was issued an FAA special airworthiness certificate in the experimental category as an amateur built airplane with a requirement for the next annual condition inspection due on June 30, 2015.

Entries in the aircraft logbook showed that the first recorded flight in N729AR was by a Team Tango employee on June 11, 2014, at a beginning aircraft total time of 0.0 hours. Based on a review of entries in the aircraft logbook, the engine logbook, pilot logbook entries and other records, the IIC estimated the total flight times on August 31, 2014, as: aircraft total time 53.0 hours; engine total time 1,498.5 hours; and engine total time since major overhaul 362.5 hours.

Records show that N729AR flew from Golden Triangle Regional Airport (GTR), Columbus, Mississippi to San Marcos Municipal Airport (HYI), San Marcos, Texas, on August 29, 2014. Refueling records at HYI showed the airplane was refueled on that day with 65.80 gallons of 100LL aviation gasoline.

METEOROLOGICAL INFORMATION

At 1752 the Automated Surface Observation System at ABI reported wind from 150 degrees at 12 knots, visibility 10 miles, few clouds at 7,500 feet, temperature 36 degrees Celsius (C), dew point 17 degrees C, and an altimeter setting of 29.87 inches of mercury.

Data from the National Oceanic and Atmospheric Administration showed that, at the accident location, at 1834, the altitude of the sun was about 18 degrees above the horizon and the azimuth of the sun was about 269 degrees. Apparent sunset occurred at 2006.

COMMUNICATIONS AND RADAR

Following is a timeline of selected communications between the pilot of N729AR and FAA Air Traffic Control (ATC), including a summary of selected FAA ATC radar contacts.

1755 N729AR landed at ABI after a flight from HYI. All ATC services with the inbound flight appeared to be normal and N729AR taxied to Abilene Aero, the general aviation fixed base operator, located east of taxiway C2 and west of the control tower.

1827:33 N729AR requested taxi instructions from Abilene Aero, to runway 17R at taxiway C3 for an intersection takeoff. The controller instructed the pilot to taxi via taxiway C. The pilot correctly read back the instructions and advised he would be operating west of the airfield in the local area for about 15 minutes before returning to ABI.

1832:01 N729AR was at taxiway C3 and runway 17R and advised he was ready for departure.

1832:19 The controller cleared N729AR for takeoff; the pilot read back the takeoff clearance and repeated that he would be staying in the local before returning to ABI.

1833:03 Radar showed a primary target over the runway abeam taxiway C2.

1833:22 Radar showed the primary target was southbound.

1833:27 Radar showed the primary target was southbound.

1833:32 Radar began receiving the Mode 3 code of 0322 and showed N729AR was southbound at a transponder reported altitude of 2,100 feet mean sea level (msl).

1833:37 Radar showed N729AR was southbound at a transponder reported altitude of 2,100 feet msl.

1833:41 Radar showed N729AR was beginning a left turn at a transponder reported altitude of 2,200 feet msl.

1833:42 N729AR transmitted: "tower experimental seven two nine alpha romeo ah emergency landing one seven right opposite direction".

There were no further communications from N729AR.

1833:46 Radar showed N729AR was in a left turn at a transponder reported altitude of 2,100 feet msl.

1833:51 Radar showed N729AR was in a left turn at a transponder reported altitude of 2,000 feet msl.

Radar contact was then lost.

AIRPORT INFORMATION

The FAA Airport/ Facility Directory, South Central U. S., indicated that ABI was a towered airport with a field elevation of 1,790 feet msl. The longest runway was 17R/35L, which was a grooved asphalt runway 7,203 feet long by 150 feet wide. Runway 17R was oriented to 180 degrees true and 172 degrees magnetic. Records show that runway 17R had a 0.4 percent upslope gradient to the south. Other runways at ABI were also listed.

WRECKAGE AND IMPACT INFORMATION

The airplane impacted 15 to 20-foot-tall mesquite trees and terrain on flat ranch lands at an estimated terrain elevation of about 1,755 feet msl. The impact location was about 2,500 feet south from the departure end of runway 17R.

Evidence at the scene showed that the airplane was moving to the north-northwest when it impacted terrain in a slightly nose down attitude with the right wing low. The first impact was to an upper branch of a mesquite tree at about 11 feet agl. The next impact was a ground scar about 16 feet away at an estimated angle of about 35 degrees from the horizontal. An uprooted mesquite tree about 28 feet from the initial tree impact showed evidence of a nearly vertical slice in its trunk at an angle to the first tree impact estimated about 12 degrees from the horizontal. The separated propeller spinner was found nearby. Other ground scars, wreckage debris and numerous broken branches littered the area on a wreckage distribution path of about 330 degrees. Several trees were completely uprooted by the impact. The main wreckage came to rest about 140 feet from the initial tree impact. The completely separated forward part of the fuselage, including the attached engine and propeller, came to rest about 195 feet from the initial tree impact.

The instrument panel and forward cockpit, including both front cockpit seats, were completely separated from the main fuselage and fragmented parts littered the wreckage distribution path leading to the main wreckage. The entire wing was completely separated from the fuselage and came to rest inverted with the left wing tip oriented to the south. Both integral wing fuel tanks were completely breached. The main part of the fuselage came to rest upright on top of the separated wing. The tail surfaces and aft empennage remained attached to each other, but were completely separated from the fuselage behind the rear cabin bulkhead and came to rest about 45 feet southeast from the main wreckage. Parts of the engine cowling, other fragmented parts, and a noticeable amount of engine oil littered the paved road in the path leading to the separated engine.

The upper skin on both wings was completely separated and both fuel tank caps remained securely fastened. Each fuel port was marked "AV GAS 60 GAL 100LL".

Impact damage prevented an assessment of the position of the flaps.

All flight control surfaces, and all major components of the airplane were accounted for.

The USAF Air Force Rescue Coordination Center reported that they had not monitored any emergency locator transmitter (ELT) beacon signals in the area on the day of the accident. An ACK E-04 ELT installed in the airplane's tail was examined. The 406 MHz ELT was marked with a 15-digit hex number: 2DC86 A42C2 FFBFF and the manufacture date code was: "23 AUG 2013". Investigators opened the case to the ELT and observed that no batteries were installed.

Parts of the four foot tall barbed wire fence in the debris path on the south side of the east-west paved road were destroyed. The electric power lines, about 35 feet above the fence, were undamaged.

Emergency responders reported they did not see evidence of liquid fuel and did not smell fuel when they first arrived at the accident scene. The investigative team arrived on the following day and did not see evidence of liquid fuel and no evidence of a vegetation kill was evident. Other investigative team members returned to the scene on the third day following the accident and they then observed clear evidence of a significant fuel blight of vegetation which showed a fuel spray pattern on the leaves of the mesquite trees in the accident debris field.

While on-scene, the investigative team observed that many of the large areas of flat cultivated agricultural fields to the south appeared to be a more adequate emergency forced landing area than the wooded area of the accident location.

MEDICAL AND PATHOLOGICAL INFORMATION

An autopsy was performed on the pilot by the Tarrant County Office of the Chief Medical Examiner in Fort Worth, Texas.

Forensic toxicology was performed on specimens from the pilot by the FAA Civil Aerospace Medical Institute (CAMI), Oklahoma City, Oklahoma. The toxicology report stated that no carbon monoxide was detected in blood. Tests for cyanide were not performed. No ethanol was detected in vitreous. Diphenhydramine was detected in urine and femoral blood. No other listed drugs were detected. The amount of diphenhydramine detected was too low to quantify and thus is considered to have not been impairing.

The pilot's most recent FAA medical certificate was an unrestricted third class certificate issued on January 17, 2001. USAF investigators reported that the pilot's most recent military flight-check as an instructor pilot was completed on October 5, 2013. A USAF "Medical Recommendation for Flying or Special Operational Duty" (form AF IMT 1041) showed that the pilot completed a USAF flight physical exam on January 24, 2014, and was then medically cleared for military instructor pilot duties. Under the provisions of 49 C.F.R. 61.23 (b) (9), the pilot's current military medical clearance met the FAA's equivalent requirements for operations requiring an FAA third class medical certificate.

TESTS AND RESEARCH

The wreckage was moved to another location and examined.

Examination of the separated wing showed the right wing was almost unrecognizable and highly fragmented. The fuel tank in the right wing was completely impact breached. The only section of the right wing remaining was a largely broomstrawed spar cap and impact damaged and fragmented spar web. The right hand wing ribs, wing skins and flight controls were all of fiberglass construction and were broken, most of them being completely separated from the adjacent structural members. The right landing gear fitting was completely separated from the wing. The left side of the wing still retained the shape of a wing, with the exception that the leading edge was fragmented, the entire upper skin of the left wing was separated, and the integral fuel tank was breached. The bond quality of the left wing skin bond line exhibited process issues that resulted in decent adherence to spars and ribs, but poor adhesion resulted in a clean separation of the bond to the attaching wing skin surface. The adhesive was not consistently applied on the whole bond surface and in some areas the adhesive had a very thick bond line.

The flight control linkage and engine control linkage were examined. Severe impact damage prevented a confirmation of preimpact control continuity, however all of the control linkage separations examined appeared to be a result of impact damage.

The engine mount remained attached to the firewall and the separated section of forward fuselage and was partially collapsed just to the engine mount's point of contact with the firewall. The impact vector was primarily rearward with an upward component. The main cabin area of the fuselage was largely intact and consisted of the barrel section from a plane just behind the front spar and instrument panel back to the empennage break. This remaining fuselage cabin area was laterally flexible due to compromise of the aft cabin bulkhead. The forward fuselage from the firewall and below the windscreen to just aft of the front spar was fragmented and separated.

The two front seat lap belt anchorages were adhesively bonded to the interior structure. The front seats and the inboard lap belt attachments and shoulder harness reel attachments were impact separated. Most of the fuselage shell around the front seat occupants had disintegrated into small pieces. The forward lap belt anchor locations consisted of large hex head bolts placed adjacent to the airplane structure with the fastener head covered with a generous mound of structural filler. On top of this filler the assembly was then covered with what appeared to be two four inch diameter circular wet-layup plies of glass cloth. The lap belt attachments were not otherwise mechanically attached. The shoulder harness inertia reels for the front seat occupants were anchored to the floor of the airplane and were not attached to the ceiling or the upper side of the door jamb. Both inboard lap belt attachments and both shoulder harness inertial reel attachments for the front

NTSB Probable Cause

The pilot's decision to attempt a steep turn to return to the airport after reporting an emergency. The reason for the emergency could not be determined because postaccident examination of the airframe and engine did not reveal any mechanical anomalies that would have precluded normal operation.