WPR14FA165

The partial loss of engine power due to an improperly maintained carburetor and the pilot's subsequent failure to maintain aircraft control.

The plans-built single-seat airplane had been constructed about 8 years before the accident by another individual, who had flown it about 30 hours before it was purchased by the current owner/pilot. In the year since the purchase, due to the low build-quality of the airplane, the pilot had made several modifications and repairs to the airplane. The accident flight was the pilot's fifth flight in the airplane. Review of the pilot's flight logbook indicated that his most recent flight review occurred about 7 years before the accident and that he had flown only 15 hours in the 2 years before the accident. A witness reported that shortly after takeoff, when the airplane was about 3 miles from the airport, the engine began making a sound as if power was intermittently being interrupted. The nose of the airplane began to pitch up aggressively as it flew out of view. The wreckage location, wreckage distribution, and impact signatures indicated that the airplane struck the ground in a steep nose-low attitude, consistent with an aerodynamic stall event. Postaccident examination of the carburetor revealed multiple maintenance-related discrepancies, any one of which could have resulted in the loss of engine power. Additionally, before the accident, the pilot reported to a friend that the airspeed indicator was not reliable and that the airplane exhibited roll control anomalies. Both of these conditions would have hindered the pilot's ability to safely operate the airplane.

Carburetor The Stromberg NA-S3B carburetor was examined at the facilities of Uni-Tech Air Management Systems, Kankakee, Illinois in the presence of the NTSB investigator-in-charge. The carburetor serial, "Continental", and model numbers correlated to the gravity fuel-feed application for use with a Continental A-50 or A-65 engine. The carburetor was of the "low-altitude" fixed mixture control type, and was therefore not configured with a cockpit adjustable mixture control arm. The carburetor sustained minimal damage and was externally examined. No obvious fuel leaks were observed, and according to the Uni-Tech representative, the fuel inlet hose was of the automotive type. No safety wire was present on the venturi retainer or the throttle valve lock adjustment screws. The idle screw mixture appeared to be set at 3/4 of a turn back from fully closed, rather than the typical 3 turns. The Uni-Tech representative stated that it was not common to see an idle mixture screw set so low. The throttle control arm moved smoothly when manipulated by hand. The arm was moved to the fully closed position, and the throttle valve completely obscured the venturi orifice. No gap was observed between the valve and the throat in this position, indicative of an incorrectly adjusted valve stop screw. According to the Uni-Tech representative, a valve without a gap at the venturi intersection results in an almost completely closed air inlet, and would inhibit or limit the engine's ability to operate at idle speed. The control lever was then moved to the full-open position, and the valve appeared to open beyond its center position by about 5 degrees. The fuel bowl was separated from the upper casing, and internal components were examined. An undamaged Delrin float needle had been installed, with the appropriate rounded valve seat; however, no accompanying brass float balance weight had been installed as required by Bendix (Stromberg) Service Bulletin Number 84. Examination of the engine maintenance logbooks revealed an unattributed entry dated March 10, 2006 (before the first accident), stating, "Installed delron needle in carburetor". The entry did not indicate the installation of the accompanying float weight. The main metering jet body appeared to be touching the float base, preventing full travel of the float. The float drop was then measured at the valve needle and exhibited 0.019 inches of travel as opposed to the minimum specified in Stromberg documentation of 0.048 inches. The bowl assembly was mounted and leveled on a flow gauge test assembly, and fuel was applied to the inlet at a pressure of 1 psi. The float moved upwards, and fuel immediately overflowed out of the bowl consistent with binding of either the float or float assembly. The operation was tried again, and this time the fuel flow stopped once it had reached within 1/16th of an inch from the bowl seam. According to Stromberg specifications, the fluid level (float level) should be 13/32nds of an inch from the seam. The float and float pin assembly was removed and examined. An indentation was observed on the lower portion of the float adjacent to the chafe strip, in an area corresponding to the main metering jet. The valve was reinserted within the seat, and would not seal when low air pressure was applied. The needle seat assembly was removed, and two level-adjustment gaskets were present. The gaskets were 0.03 and 0.062 inches thick, respectively. According to the Uni-Tech representative, the maximum allowable combined gasket thickness was 0.05 inches. The fuel inlet screen plug appeared to have been tightened with excessive force, and required considerable force to remove. Furthermore, the fuel inlet screen had been installed upside down, such that fuel would have flowed around the screen seat thereby limiting the screen's ability to capture debris. Fueling records from Mariposa Airport indicated that the pilot purchased 5.12 gallons of 100 octane low-lead aviation gasoline about 1 hour prior to the accident. Daily records provided by Shell Aviation revealed that the fuel in the airport's tanks was clean and clear during the week leading up to the accident. Three other aircraft were serviced with fuel from the same tank on that day, and none reported problems. The airplane met the FAA criteria for the light sport aircraft category, and the pilot's stated medical status along with his commercial pilot rating allowed him to operate with sport pilot privileges. As such, he was not required to hold a current FAA medical certificate. The carburetor icing probability chart from FAA Special Airworthiness Information Bulletin CE-09-35, Carburetor Icing Prevention, dated June 30, 2009, shows a probability of "icing at glide power" at the temperature and dew point reported at the time of the accident. An autopsy was conducted by the Stanislaus County Sheriff's Department, Coroners Division, on behalf of the Mariposa County Coroner's Office. The cause of death was reported as the effect of blunt injuries, with no other contributing conditions. Toxicological tests on specimens recovered from the pilot were performed by the FAA Civil Aerospace Medical Institute (CAMI). Analysis revealed negative findings for carbon monoxide and ethanol with the following positive drug findings: >> Warfarin detected in Urine >> Warfarin detected in Blood (Cavity) Refer to the toxicology report included in the public docket for specific test parameters and results. According to CAMI, Warfarin is an anticoagulant medication, with no specific warnings pertinent to flight. The airplane came to rest at the base of an oak tree, within densely wooded terrain at an elevation of 2,250 feet mean sea level (msl), about 3 miles northwest of Mariposa Airport. The terrain surrounding the accident site was comprised of grass and poison oak, interspersed with rocky outcroppings and various oak trees ranging in height from saplings to 20 feet tall. The airplane came to rest on a magnetic heading of about 60 degrees, facing uphill on a 20-degree slope. A freshly cut swath through the tree branches was located directly above the airplane; the swath was nearly vertical. Although the airplane was surrounded by trees, no other damage to limbs or branches was noted. The wings came to rest inverted, with the forward fuselage and engine located underneath the wing root. The tailcone and empennage structure had separated aft of the seat, and was resting undamaged on its right side. Both wings sustained aft crush damage to their leading edges. The entire cabin structure forward of the tailcone was fragmented, and the firewall was compressed against the rear of the engine. All cockpit flight controls exhibited varying degrees of bending damage, but remained functionally intact. The fuel tank sustained multiple breaches, and was detached but still located within the center section of the wreckage. The airframe and engine did not display any indications of bird strike or fire. The engine remained attached to its mount, which remained attached to the firewall. Both magnetos (Eisemann, Model AM-4) remained firmly attached to the engine, however, their plug caps had both fragmented, crushing and exposing the timing gears, points, and coils, as well as detaching all ignition wires and both P-leads. The carburetor had broken away from the inlet manifold, and the inlet air filter assembly exhibited crush damage. The throttle cable was attached and continuous from the cockpit to the butterfly valve; the cable was in the full-forward position at the cockpit control. The carburetor heat control cable was continuous from the cockpit control to the heat box. The control was in the aft (carburetor heat on) position. The fuel primer was in the forward and locked position. The top spark plugs were removed and examined. They were of the three-prong type, with the electrodes covered in grey deposits and exhibiting m