NTSB Factual Report

HISTORY OF FLIGHT

On March 6, 2009, about 1453 eastern standard time, an experimental amateur-built Stewart S51, N551X, was substantially damaged when it impacted trees and terrain in a recreational vehicle park in Ocala, Florida. The owner/pilot, the sole occupant, was fatally injured. The personal test flight was operated under the provisions of Title 14 Code of Federal Regulations Part 91. Visual meteorological conditions prevailed for the flight, and no flight plan was filed.

According to a Federal Aviation Administration (FAA) inspector and the pilot's relatives, the pilot was not the builder of the airplane, and he was its third owner. The pilot had just started the first phase of the flight test program, and the airplane's first flight by the accident pilot occurred on the day prior to the accident. The pilot told his wife that it was a "beautiful" flight, and his notes indicated that the duration was "0.3" hours.

A total of 17 witnesses provided information regarding the final portion of the accident flight. Almost all reported that the airplane was flying at a low altitude, and some reported it to be flying at a low speed. Nine witnesses reported that the engine was "sputtering" or making unusual sounds, and five witnesses, including a Florida Highway Patrol officer, observed that the engine was not operating. Most witnesses reported that the airplane was west- or north-bound and descending. Several reported that the airplane banked sharply in order to avoid a cell phone tower. They also reported that it struck trees in the final few seconds of flight. The airplane impacted the ground between two unoccupied recreational vehicles, and a fire ensued.

PERSONNEL INFORMATION

FAA records indicated that the 80 year old pilot held an airline transport certificate with an airplane multi engine land rating; a commercial pilot certificate with airplane single engine land, single engine sea, rotorcraft helicopter, and glider ratings; a flight instructor certificate with airplane single engine, multi engine, rotorcraft, and instrument airplane ratings; and a mechanic certificate with airframe and powerplant ratings. According to several persons who knew the pilot, he was a former test pilot on the North American P-51 airplane.

The pilot's most recent FAA third-class medical certificate was issued in March 2007, and he reported 13,917 total hours of flight experience on the application. A review of the pilot's logbook revealed that as of the day prior to the accident, he had a total of approximately 13,980 hours of flight experience. He had logged a total of 0.3 hours in the accident airplane; which occurred on the day prior to the accident flight. His next most recent flight occurred on February 12, about 3 weeks prior to the accident. In the 30, 60, and 90 days prior to the accident, the pilot had respectively logged 0.7, 1.3 and 3.5 hours, in four different airplanes. The airplanes were a Bellanca 7ECA ("Citabria"), an Aeronca 7AC ("Champ"), a Cessna 195, and the accident airplane. His most recent flight review was completed in September 2007, in a Cessna 195.

AIRPLANE INFORMATION

According to information on an S51 builder's website, the airplane was a two place, 70% scale kit-version replica of the North American P-51 fighter. Both the company that created the original plans, and the company that manufactured the kits, had ceased operations prior to the accident.

The airplane was of all-metal construction, and equipped with retractable landing gear and electrically-driven flaps. The standard fuel system configuration had a total capacity of 68 gallons, stored in integral wing tanks referred to as "wet wings." The primary fuel pump was the engine driven, rotary vane pump, which was supplemented by an electric auxiliary pump. The maximum gross weight was cited as 3,500 pounds. Nominal airplane performance values were cited as follows: Stall (at gross weight) 74 mph; Cruise (at 75 percent power) 289 mph at 28.5 gallons per hour (gph); Top speed (level flight) 334 mph at 38 gph; Maximum flap speed 150 mph; Maximum landing gear speed 130 mph. The data plate affixed to the airplane listed the maximum gross weight as 3,900 pounds.

The accident airplane was equipped with a liquid-cooled, port-injection Chevrolet 612 cubic inch displacement engine that produced approximately 400 to 600 horsepower, and a four-blade Hartzell constant speed propeller.

The airplane kit was purchased, but not completed, by its first owner. According to the second owner, he purchased the partially-assembled kit from the first owner, with the intention of racing the airplane. The second owner completed the construction in 2002, and he stated that the airplane had accumulated approximately 15 total hours of flight time while he owned it. Since the airplane was not "fast enough" to suit the second owner's needs, he partially disassembled it, and sold it to the accident pilot in March 2003. The disassembled airplane was transported by truck to the accident pilot's residence on his private grass airstrip, which was Idle Wild Airport (FL63), Ocala, Florida,

According to FAA records, a special airworthiness certificate was issued to the pilot for the airplane in July 2008. The operating limitations associated with the airworthiness certificate limited the airplane to flights of within a 40 mile radius of the Ocala very high frequency omnidirectional radio range (VOR). The airplane was based at FL63, which was located approximately 10 miles northwest of the Ocala VOR.

METEOROLOGICAL INFORMATION

The 1455 automated weather observation at Ocala International Airport-Jim Taylor Field (OCF), Ocala, Florida, located 5.6 miles southwest of the accident site, included winds from 170 degrees at 4 knots, clear skies, 10 miles visibility, temperature 25 degrees C, dew point 6 degrees C, and an altimeter setting of 30.34 inches of mercury.

COMMUNICATIONS

A review of FAA air traffic control voice recordings and ground based radar tracking data did not reveal any communications with the accident airplane, or any radar targets that could be associated with the accident airplane.

WRECKAGE AND IMPACT INFORMATION

According to information provided by the FAA inspector who responded to the accident, the accident location was 11.3 miles south-southeast of FL63. The airplane impacted the ground between two unoccupied recreational vehicles (RVs), which were damaged by airplane and tree debris. The airplane incurred significant fire damage.

A fence on the south side of the RV park separated a wooded area from the RV park. The ground impact point and main wreckage, which consisted of most of the airplane, was approximately 60 feet north of the fence, inside the RV park. A portion of the left wing was found approximately 35 feet south of the fence, and several trees in that area were damaged. The southern-most component was the tailwheel, which was found approximately 225 feet south of the fence.

A cell phone tower was located in the wooded area, approximately 475 feet south-southwest of the main wreckage. According to the FAA inspector, the cell phone tower was braced by multiple guy wires affixed at various heights and azimuth locations on the tower. The inspector stated that the "upper guy wires on the east side of the tower were near the suspected flight path," but his inspection of the tower and wires indicated that "no damage was evident." A review of the relevant aeronautical chart indicated that the tower height was approximately 500 feet above ground level.

The airplane, engine and propeller exhibited significant fire damage. The FAA inspector was able to confirm the propeller "blade connection to the pitch change mechanism," but he was unable to ascertain any engine control positions. The inspector observed that all engine "drive belts appeared to be on the pulleys" at the time of the post impact fire. He "suspected" that the tailwheel was in the extended position, but he was unable to determine whether the main landing gear was extended or retracted at the time of the accident.

MEDICAL AND PATHOLOGICAL INFORMATION

An autopsy conducted by the Florida District 5 Medical Examiner listed the cause of death as multiple blunt force injuries. Toxicological testing was conducted by the FAA Civil Aero Medical Institute. No carbon monoxide, cyanide, or ethanol was detected, but triamterene was detected in the blood and urine. Triamterene was a prescription diuretic medication typically used to control high blood pressure, was generally approved by the FAA for that purpose, and was not typically expected to result in impairment.

The pilot did not include any medications or medical conditions on his most recent airman medical certificate application, dated March 22, 2007.

ADDITIONAL INFORMATION

Other Witness Accounts

The pilot's widow stated that the airplane was reassembled by the pilot at his home airstrip. She said that he was not in a rush to complete and fly the airplane, that he had several other airplanes, and that he flew the other airplanes "regularly." She did not witness the pilot's first flight of the airplane, but she said that he told her that the airplane was "absolutely perfect." She watched the pilot take off on the accident flight; the takeoff time was about 1435 to 1440, and the pilot did not discuss his plans for the accident flight.

A long-time friend of the pilot stated that he had heard that the pilot was "struggling" with fuel leaks for some time, but that the pilot eventually solved the problems. The friend stated that he was not aware of exactly where the fuel leaks were located, or how the pilot stopped them, but that they were "frustrating" to the pilot. The friend said that the pilot did not conduct any taxi tests until the fuel leak problems were resolved, and that the engine "ran roughly" and/or "cut in and cut out," apparently from fuel-related problems, on more than one occasion during the taxi tests. The friend did not know the exact causes for, or resolutions to, the engine shutdowns during the taxi tests. He said that the pilot told him that the first flight was "very successful," and that the pilot was planning to move the airplane from his home airport to another airport very early in the flight test program, due to the other airport's multiple, paved runways. He also heard that the pilot planned to depart from his home field with a limited fuel supply, due to the field length and conditions. The pilot's friend did not witness either of the airplane's two flights.

A third individual, who was the builder of another S51, became acquainted with the pilot via an S51 builders' network. He stated that he was not very familiar with the pilot or the accident airplane, and that the last time he saw the airplane was 2 or 3 years prior to the accident. He said that the airplane design featured "wet wings," and that the pilot was "fighting a fuel leak" for about 6 months. He stated that one reason for the amount of time between the accident pilot's purchase of the airplane and its first flight was that there was a significant amount of reassembly required, and that the pilot was "meticulous" in his workmanship. The individual stated that he had heard that the pilot used a "sloshing sealant" to correct the fuel leaks, and that the particular sealant used by the pilot seemed to leave a "white oxide" residue. He also stated that subsequent to the use of the sloshing sealant, the pilot conducted at least two high speed taxi tests (on separate occasions), and both times the engine quit. The individual stated that he understood that on inspection following the taxi tests, the fuel "filters or screens" were found to be "clogged." He said that the pilot used the same sealant type or brand that was used until about 3 to 4 years prior to the accident by several builders of Vans airplanes, and that those builders switched the sealant they used due to its propensity to produce the white residue. The individual stated that his S51 fuel selector valve had three selectable positions (Right, Left and Off), and that it was equipped with two electric fuel pumps plumbed in parallel, and stationed forward of the firewall. The individual was unfamiliar with the accident airplane fuel system configuration.

Accident Airplane Construction Documents and Maintenance Records

Attempts to locate construction documentation and maintenance records specific to the accident airplane produced only two bound documents and a small notebook that belonged to the pilot. The two bound documents were entitled "S51D Builder's Guide" and "The Tractor Mechanic's Guide to Big Block Engines for the Stewart 51." These documents were copyrighted to an individual who had built an S51. Neither of these documents contained any information (handwritten or otherwise) that could be directly associated with the accident airplane. The small notebook was specific to the accident airplane. The only entries in the notebook were 6 1/2 pages of flight test notes entered by the second owner, and one brief hand-written paragraph by the accident pilot. The accident pilot's entry was "Mar 5, 2009 IDLW Local 0.30 RPM Hi, [a check mark symbol] Elev trim, Install Hobbs, Fit Gear Doors, Fuel Lk."

Fuel Tank Sealant

A partially-used can of PPG Aerospace brand "slosh sealant" was located in the pilot's hangar. The label listed the contents as "PR-1005-L One Part" and the label directions stated "PR-1005-L is a one-part synthetic rubber solution, ready to use as packaged." The label listed a shelf life of "1 year at temperature below 80 deg F when stored in original unopened container," listed a manufacture date of "06/08" and an expiration date of "06/09."

A technical data sheet for the sealant stated that "PR-1005-L is an aircraft integral fuel tank slosh coating…designed as topcoat or barrier coating. The cured coating is resistant to prolonged exposure to both jet fuel and aviation gas." The sheet also stated that the "uncured material is a thin syrup suitable for application by brush, fill-and-drain, dip or spray. It cures at room temperature by the evaporation of solvent to form a smooth, tough, flexible transparent film having excellent adhesion to common aircraft substrates and polysulfide sealants." The sealant conformed to military specification MIL-S-4383C.

The data sheet provided the following guidance on surface preparation: "Immediately before applying sealant to primed substrates, the surfaces should be cleaned with solvents. Contaminants such as dirt, grease, and/or processing lubricants must be removed prior to sealant application. A progressive cleaning procedure should be employed using the appropriate solvents and new lint free cloth (reclaimed solvents or tissue paper should not be used). Always pour solvent on the cloth to avoid contaminating the solvent supply. Wash one small area at a time. It is important that the surface is dried with a second clean cloth prior to the solvent evaporating to prevent the re-deposition of contaminants on the substrate. Substrate composition can vary greatly. This can affect sealant adhesion. It is recommended that adhesion characteristics to a specific substrate be determined prior to application on production parts or assemblies."

The S51D Builder's Guide specified the use of a commercial sealant called "Pro-Seal," which conformed to MIL-S-8802E or MIL-S-8784B, in the assembly of the wing bays designed to contain fuel. The builder's guide did not specify or prohibit the use of MIL-S-4383C sealant in the fuel bays. However, unlike PR-1005-L/MIL-S-4383C, which was to be applied after assembly, this material was to be used between the components during assembly. According to pilot/owners who built Van's kit airplanes, which also incorporated wet wings, they previously used PR-1005-L/MIL-S-4383C, but problems associated with the sealant flaking off the substrate caused them to discontinue its use, and use Pro-Seal instead.

No direct witnesses to when or how the pilot prepared the fuel bay interior surfaces, or applied the slosh sealant, were located. Examination of the builder's guide and the wreckage indicated that once wing assembly was completed, access to the fuel bays was extremely limite

NTSB Probable Cause

A loss of engine power due to contamination and clogging of the fuel system by a post-assembly fuel tank sealant.