NTSB Factual Report

HISTORY OF FLIGHT

On September 26, 1998, about 0810 hours Pacific daylight time, an amateur-built experimental O'Neil Van's RV-3, N244WW, experienced an in-flight breakup over Arbuckle, California. The airplane was destroyed and the pilot, the sole occupant, received fatal injuries. The airplane was being operated as a personal flight by the pilot/owner when the accident occurred. The flight originated from Redding, California, about 0730, and was en route to Atwater, California. According to witnesses, instrument meteorological conditions prevailed at the accident site and no flight plan was filed.

A witness, positioned about 1 mile from the scene of the accident, heard an aircraft fly overhead. He was unable to establish visual contact due to the presence of an overcast ceiling and misting precipitation, which also reduced visibility.

Another witness, positioned about 0.5 miles from the scene of the accident, reported that he saw the airplane break out of the bottom of an overcast cloud layer. He said that as he spotted the aircraft, it was accompanied by increased engine sounds and falling parts. During the descent, the right wing separated. The aircraft subsequently crashed in a grove of almond trees on the west side of Interstate 5. The witness estimated the base of the overcast to be about 400 feet agl.

The pilot had reportedly planned to attend an EAA Fly-In that was being held at Castle airport near Atwater.

PILOT INFORMATION

According to Federal Aviation Administration (FAA) Airman and Medical Records, the pilot held a commercial pilot certificate with an airplane single engine land rating. He was not instrument rated. A second-class medical certificate was issued on October 14, 1996.

The pilot's logbook was not found. He reported he had accumulated a total of 2,100 flight hours at the time of his last FAA medical examination.

AIRCRAFT INFORMATION

The aircraft's airworthiness certificate in the experimental amateur-built category was issued on March 3, 1981.

The aircraft was not equipped for flight in IFR conditions.

The aircraft's maintenance records were not recovered.

There was no data plate on the engine. A postaccident examination of the engine revealed that it was similar to a Lycoming O-290.

METEOROLOGICAL CONDITIONS

The 1,500 hour UTC, September 26, 1998, national mosaic reflectivity image for central California shows areas of precipitation in the general vicinity of the accident site.

COMMUNICATIONS

There were no reported communications from the pilot.

WRECKAGE AND IMPACT INFORMATION

Safety Board investigators did not observe the accident site. According to the aircraft retriever, the aircraft struck an almond tree in a grove located near the intersection of Marine and Wyre, southwest of Arbuckle. The right wing had separated from the aircraft and was located approximately 100 feet from the main wreckage.

The wreckage was examined by a Safety Board investigator on October 22, 1998, at the Plain Parts storage facility in Pleasant Grove, California.

The aircraft was laid out in a two-dimensional manner and an inventory of the aircraft structure was conducted. Investigators were able to account for all flight control surfaces; however, control continuity could not be determined due to the extent of the overall destruction of the flight control system.

The right wing was separated from the fuselage. The separation began about 10 inches outboard of the fuselage and extended to about 38 inches outboard of fuselage. The laminated spar caps showed evidence of buckling at either side of the inboard web. The web showed evidence of fore to aft diagonal distortion. There was no evidence of distortion on the right aileron's mechanical stops.

Both horizontal stabilizers and the vertical stabilizer were separated from the empennage. The empennage was separated from the fuselage at a bulkhead, about 24 inches forward of the tail cone.

The panel mounted G-meter was recovered and was found to read +7 G's and -4.5 G's.

The engine was separated from the airframe engine mounts. The two-bladed fixed pitch propeller was separated from the engine at the crankshaft flange. The crankshaft could not be rotated by hand. Mechanical continuity of the rotating group was established visually during the disassembly of the engine.

Visual inspection of the camshaft revealed no evidence of spalling or flattening on any of the lobes.

The combustion chambers of each cylinder showed no evidence of foreign object damage. The valve train continuity through each cylinder was established.

The number 1 top and bottom, the number 3 top and bottom, and the number 4 bottom spark plugs showed evidence of impact damage. The number 2 top and bottom, and the number 4 top spark plugs were undamaged. According to the Champion Spark Plug Check-a-Plug chart AV-27, the coloration and wear patterns were consistent with normal operation.

There was no evidence of lack of internal lubrication on any engine's internal rotating components. There was no evidence of contamination found in the oil filter, suction screen, or oil pump. When opened for inspection, the oil pump gears were intact and also showed evidence of lubrication.

The left magneto was detached from the engine at the mounting pad. The magneto data plate was missing from the housing and was not recovered. Magneto to engine timing could not be determined. The magneto could not be rotated by hand due to the distortion of the housing. The impulse coupling was securely installed. There was no evidence of an internal malfunction.

The right magneto was destroyed.

The carburetor was found attached to its mounting pad that was located on the oil sump. The throttle and mixture controls were separated from the carburetor control arms. There was no evidence of any obstructions in the carburetor throat. The carburetor bowl contained no fuel. No evidence was found of an internal mechanical malfunction. The metal float assembly showed evidence of hydraulic deformation on the outboard skin of the right float pontoon. The fuel screen at the carburetor inlet was clear and uncontaminated.

The engine driven fuel pump was detached from the engine. The pump was disassembled and was found to contain no fuel. There was no evidence of a mechanical malfunction.

MEDICAL AND PATHOLOGICAL INFORMATION

An autopsy was conducted on September 28, 1998, by the Calusa County Coroner's Office, with specimens retained for toxicological examination.

The toxicological results were positive for quinine. Volatiles were also detected in the toxicological specimens. According to the FAA Civil Aero-Medical Institute (CAMI) research laboratory manager, they were the result of postmortem ethanol production.

TEST AND RESEARCH

The kit manufacturer performed additional tests and inspections of the wreckage as the result of several previous in-flight breakups involving the main wing spar. The manufacturer determined that the optional manufacturing procedure of using epoxy adhesive to minimize drilling and clean up required during spar assembly had the unintended effect of increasing the load strength of the spar. According to the manufacturer, since the epoxy process was not intended to increase strength, it was not factored into the original strength calculations; however, the test article used for the wing proof load tests had been assembled using the epoxy process. In the original proof test, the wing strength was determined to be 9.25 G's. A second proof test performed on a test article without the epoxy process yielded strength of 7.3G's.

As the result of the tests, owners who had not used epoxy adhesive during spar construction were advised to operate their RV-3's within the utility category G-limits of +4.4 G's to -1.76 G's. This limitation was to remain in effect until they applied a kit manufacturer designed spar modification procedure. The accident aircraft had not had epoxy used in its original spar construction nor had it the spar modification applied. There was no placard for reduced G-limits.

The FAA Western Region Air Traffic Quality Assurance office (AWP-505) reported that there was no radar data available that could be identified with the time and location of the accident aircraft.

ADDITIONAL INFORMATION

The aircraft was recovered by Plain Parts and transported to their storage facility located in Pleasant Grove, California.

The aircraft was released to Dennis James, Plain Parts on March 8, 2000, as a representative of the registered owner.

No emergency locator transmitter (ELT) was found in the aircraft wreckage.

NTSB Probable Cause

The pilot's attempt to continue VFR flight in instrument conditions, which led to his spatial disorientation and in-flight loss of aircraft control, and, resulted in a pilot-induced overload failure of the airframe. The pilot's use of quinine may have been a factor in this accident.