Primary finding
Probable cause
The pilot's loss of control due to spatial disorientation during takeoff in instrument meteorological conditions.
Investigator assessment
Analysis narrative
The instrument-rated private pilot and four passengers boarded the multiengine airplane inside a hangar. The pilot then requested that the airplane be towed from the hangar to the ramp, since he did not want to hit anything on the ramp while taxiing in the dense fog. Witnesses heard the pre-takeoff engine run-up toward the end of the runway but could not see the airplane as it departed; the engines sounded normal during the run-up and takeoff. A witness video recorded the takeoff but the airplane was not visible due to the dense fog. During the takeoff roll the airplane's tires chirped, which is consistent with the wheels touching down on the runway with a side load. The video ended before the accident occurred. The witnesses stated that the takeoff continued and then they heard the airplane impact the ground and saw an explosion. The weather conditions at the time of the accident included visibility less than 1/4 mile in fog and an overcast ceiling at 300 ft above ground level. The airplane's weight at the time of the accident was about 105 lbs over the maximum takeoff weight, which exceeded the center of gravity moment envelope. The excess weight would have likely extended the takeoff roll, decreased the climb rate, and increased the amount of elevator pressure required to lift off of the runway. A majority of the airplane was consumed by postcrash fire. The ground impact marks and wreckage distribution were consistent with the airplane rolling left over the departure end of the runway and impacting the ground inverted in a nearly vertical, nose-low attitude. Examination of the engines revealed operating signatures consistent with takeoff power at the time of impact. The elevator trim tab and actuator were found beyond their full up travel limits and the trim cable exhibited tension overload separations near the actuator. It is likely that, when the cable separated in overload, the chain turned the sprocket and extended the actuator rod beyond full travel. No anomalies were observed with the airframe, engines, or cockpit instrumentation that would have precluded normal operation. The investigation was unable to determine the status of the autopilot during the accident takeoff. Based on the evidence it's likely that when the airplane entered instrument meteorological conditions the pilot experienced spatial disorientation, which resulted in a loss of control and descent into terrain.
Source record
Factual narrative
The Office of the District Medical Examiner, 10th Judicial Circuit of Florida, Winter Haven, Florida, completed an autopsy on the pilot, which attributed the cause of death to blunt impact. The FAA's Bioaeronautical Sciences Research Laboratory, Oklahoma City, Oklahoma, conducted toxicology testing, which revealed 20 mg/dL of ethanol in muscle, ibuprofen in the muscle, and no ethanol detected in the kidney. The toxicology samples exhibited putrefaction. Ibuprofen is in a class of medications called non-steroidal anti-inflammatory drug and are not considered impairing. After absorption, ethanol is uniformly distributed throughout all tissues and body fluids; therefore, the finding in one tissue but not another is most consistent with post-mortem production. The automated weather observation station at BOW reported consistent weather conditions from 0635 to 0715, which included visibility less than 1/4 mile, fog, an overcast cloud layer at 300 ft, temperature 56°F, and an altimeter setting of 30.18 inches of mercury. An area forecast discussion was issued for the region by the National Weather Service (NWS) that identified widespread shallow fog. An NWS dense fog advisory was in effect for Polk County, Florida. A center weather advisory was in effect for the accident area and advised of ceilings below 500 ft agl and visibilities below 1 mile in fog and mist. An AIRMET for IFR conditions was in effect for the accident area. There was no evidence that the pilot obtained a preflight weather briefing from a recorded source. The pilot's logbooks were not located, and the pilot's instrument currency or proficiency could not be determined. The mechanic who maintained the airplane stated that the pilot always flew with his feet flat on the floor and not on the rudder pedals. He also stated that the pilot never flew dangerously or recklessly. He added that the pilot's personal logbooks were always kept on the back shelf in the airplane. The pilot's personal assistant stated that he always flew the airplane a couple of days before a flight with passengers. She stated that everyone she talked to described him as a good pilot and diligent with his pilot duties. An acquaintance of the pilot, who also was the pilot's flight instructor in 2002, recounted flying the accident airplane with the pilot. He stated that the pilot mentioned an in-flight engine failure he experienced in the accident airplane. The pilot told him that he continued to his destination rather than making a precautionary single-engine landing because the logistics of diverting were too difficult. The acquaintance also stated that he and the pilot were supposed to fly the accident airplane together in early 2017. On the morning of the planned flight, he checked the weather conditions, which were about 1/4 mile visibility and 100 ft ceilings with dense fog. He told the pilot that they could not complete the flight because of the weather, and the pilot responded that, legally, they were allowed to fly under Part 91. The acquaintance had not talked to the accident pilot since that canceled flight. A local airplane mechanic, who was a business acquaintance of the pilot, stated that he flew with the pilot one time and then refused to fly with him again. The acquaintance stated that he was not a safe pilot and took unnecessary risks. The accident site was located on airport property about 190 yards east-northeast of the departure end of runway 9L and 10 ft south of taxiway delta (figure 1). Figure 1 – Aerial View of Accident Site The debris path was about 230 ft long and oriented northeast. The beginning of the debris path was defined by several ground impact craters in a line perpendicular to the debris path and the main wreckage. The middle impact crater contained pieces of the airplane's nose cowling and baggage door. Immediately to the right was the left engine's propeller. To the left of the middle crater was the right engine's propeller. A portion of the left wing tip fuel tank was found on the far right side of the initial impact area, and a portion of the right wing tip fuel tank was found on the far left side of the initial impact area. The main wreckage came to rest upright and oriented southeast about 30 ft from the initial impact craters. The fuselage was mostly consumed by fire and the empennage remained mostly intact with significant thermal damage. Both engines separated from the airplane and came to rest near the main wreckage. Airplane debris was found on the taxiway northeast of the main wreckage. The nose landing gear was separated from the airplane and found about 200 ft northeast of the main wreckage. The flight controls exhibited impact and thermal damage but did not reveal any preimpact anomalies. The wing flap position could not be determined due to the extensive damage. The elevator trim tab was found beyond its full up limit. The elevator trim cable exhibited tension overload separations near the actuator and the other side of the cable was intact in the aft fuselage area. The elevator trim actuator was found at 2 ¼ inches, which was beyond the full length of travel (1.9 inches). The round hole in the trim tab surrounding the actuator rod was damaged and pushed aft by the rod end and bearing. The damage on the rod end matched the damage to the hole. The trim cable and chain on the top side of the actuator sprocket was pulled during the examination to test its functionality. When the cable was pulled, the actuator retracted and the trim tab lowered from its full-up position. The landing gear actuator and the left and right main landing gear were all retracted. Two attitude indicator gyros and one directional gyro were disassembled and examined. All three gyros exhibited rotational scoring inside the housings and along the circumference of the gyros. An electric turn-and-bank indicator gyro was disassembled and examined; it also exhibited rotational scoring. The two vacuum pumps were separated from the engines and sustained significant impact damage. The right engine vacuum pump drive remained attached to the right engine and was melted by the postimpact fire. The left engine vacuum pump was separated from the engine, and the pump body and drive were intact. A portion of the vacuum manifold, which consisted of one of the end caps but no flapper valves, was found in the wreckage. The left engine was separated from the airframe and was identified based on the data plate and maintenance records. The crankshaft was completely fractured at the nose oil seal and the propeller flange remained attached to the propeller hub. The crankshaft fracture surface displayed tearing, shear lips, and discoloration consistent with an overstress separation on impact. The top sparkplugs displayed a normal worn appearance with no signs of lead or carbon fouling. The bottom sparkplugs were examined via a lighted borescope with no anomalies noted. The return line from the fuel pressure regulator to the fuel pump remained in place and, when removed, residual fuel poured from the pressure regulator. The pump was disassembled, and the pump vanes were intact with no anomalies noted with the internal components. The fuel manifold valve remained secured to the top of the engine and the fuel lines remained attached to the housing. The fuel lines remained secured to their nozzles and the upper deck reference line remained secured to the nozzles. The nozzles were free of obstructions. The fuel manifold screen was clear of contaminants and residual fuel was observed in the housing. The No. 2 intake inner valve spring was fractured. Residual oil was observed in each rocker cover. The cylinders were inspected and documented utilizing a lighted borescope. The No. 3 exhaust valve displayed two areas of green discoloration; otherwise, the pistons, valves, and valve seats were unremarkable. Manual rotation of the crankshaft was not possible due to impinge