Louisville

The flight crew was conducting a cargo flight in instrument meteorological conditions. Takeoff from the departure airport and the en route portion of the flight were normal, with no reported weather or operational issues. As the flight neared Charleston Yeager International Airport (CRW) at an altitude of 9,000 ft, the captain and first officer received the most recent automatic terminal information service (ATIS) report for the airport indicating wind from 080º at 11 knots, 10 miles visibility, scattered clouds at 700 ft above ground level (agl), and a broken ceiling at 1,300 ft agl. However, a special weather observation recorded about 7 minutes before the flight crew's initial contact with the CRW approach controller indicated that the wind conditions had changed to 170º at 4 knots and that cloud ceilings had dropped to 500 ft agl. The CRW approach controller did not provide the updated weather information to the flight crew and did not update the ATIS, as required by Federal Aviation Administration Order 7110.65X, paragraph 2-9-2. The CRW approach controller advised the flight crew to expect the localizer 5 approach, which would have provided a straight-in final approach course aligned with runway 5. The first officer acknowledged the instruction but requested the VOR-A circling instrument approach, presumably because the approach procedure happened to line up with the flight crew's inbound flightpath and flying the localizer 5 approach would result in a slightly longer flight to the airport. However, because the localizer 5 approach was available, the flight crew's decision to fly the VOR-A circling approach was contrary to the operator's standard operating procedures (SOP). The minimum descent altitude (MDA) for the localizer approach was 373 ft agl, and the MDA for the VOR-A approach was about 773 ft agl. With the special weather observation indicating cloud cover at 500 ft agl, it would be difficult for the pilots to see the airport while at the MDA for the VOR-A approach; yet, the flight crew did not have that information. The approach controller was required to provide the flight crew with the special weather report indicating that the ceiling at the arrival airport had dropped below the MDA, which could have prompted the pilots to use the localizer approach; however, the pilots would not have been required to because the minimum visibility for the VOR-A approach was within acceptable limits. The approach controller approved the first officer's request then cleared the flight direct to the first waypoint of the VOR-A approach and to descend to 4,000 ft. Radar data indicated that as the flight progressed along the VOR-A approach course, the airplane descended 120 feet below the prescribed minimum stepdown altitude of 1,720 ft two miles prior to FOGAG waypoint. The airplane remained level at or about 1,600 ft until about 0.5 mile from the displaced threshold of the landing runway. At this point, the airplane entered a 2,500 ft-per-minute, turning descent toward the runway in a steep left bank up to 42º in an apparent attempt to line up with the runway. Performance analysis indicates that, just before the airplane impacted the runway, the descent rate decreased to about 600 fpm and pitch began to move in a nose-up direction, suggesting that the captain was pulling up as the airplane neared the pavement; however, it was too late to save the approach. Postaccident examination of the airplane did not identify any airplane or engine malfunctions or failures that would have precluded normal operation. Video and witness information were not conclusive as to whether the captain descended below the MDA before exiting the cloud cover; however, the descent from the MDA was not in accordance with federal regulations, which required, in part, that pilots not leave the MDA until the "aircraft is continuously in a position from which a descent to a landing on the intended runway can be made at a normal descent rate using normal maneuvers." The accident airplane's descent rate was not in accordance with company guidance, which stated that "a constant rate of descent of about 500 ft./min. should be maintained." Rather than continue the VOR-A approach with an excessive descent rate and airplane maneuvering, the captain should have conducted a missed approach and executed the localizer 5 approach procedure. No evidence was found to indicate why the captain chose to continue the approach; however, the captain's recent performance history, including an unsatisfactory checkride due to poor instrument flying, indicated that his instrument flight skills were marginal. It is possible that the captain felt more confident in his ability to perform an unstable approach to the runway compared to conducting the circling approach to land. The first officer also could have called for a missed approach but, based on text messages she sent to friends and their interview statements, the first officer was not in the habit of speaking up. The difference in experience between the captain and first officer likely created a barrier to communication due to authority gradient. ATC data of three VOR-A approaches to CRW flown by the captain over a period of 3 months before the accident and airport security footage of previous landings by the flight crew 1 month before the accident suggest that the captain's early descent below specified altitudes and excessive maneuvering during landing were not isolated to the accident flight. The evidence suggests that the flight crew consistently turned to final later and at a lower altitude than recommended by the operator's SOPs. The flight crew's performance on the accident flight was consistent with procedural intentional noncompliance, which—as a longstanding concern of the NTSB—was highlighted on the NTSB's 2015 Most Wanted List. The operator stands as the first line of defense against procedural intentional noncompliance by setting a positive safety attitude for personnel to follow and establishing organizational protections. However, the operator had no formal safety and oversight program to assess compliance with SOPs or monitor pilots, such as the captain, with previous performance issues.

The instrument-rated pilot was issued a clearance to descend to 4,000 feet for radar vectors to a non precision instrument approach in instrument meteorological conditions (IMC). The last 1 minute 23 seconds of radar data indicated the airplane leveled at 4,000 feet for about 35 seconds and then varied between 3,800 feet and 3,900 feet for the remainder of the flight for which data was available. During this timeframe, the airspeed decreased from 131 knots to 57 knots. Witnesses observed the airplane descending in a spin, and one reported hearing the engine running. Recorded engine data showed an increase in engine power near stall speed, which was likely the pilot's response to the low airspeed. The airplane damage was consistent with a low-speed impact with some rotation about the airplane's vertical axis. The pilot did not make any transmissions to air traffic control indicating any abnormalities or emergency. Post accident examination of the airplane revealed no anomalies that would have precluded normal operation. During training on the accident airplane, the instructor recommended that the pilot get 25 to 50 hours of flight in visual meteorological conditions before flying in IMC in order to gain more familiarity with the radios, switches, and navigation equipment. The pilot only had 14 hours of flight time in the accident airplane before the accident flight, however it could not determined whether this played a role in the accident.

The pilot received a weather briefing before departure and when near the destination airport, cleared for the NDB approach. The pilot reported the procedure turn inbound; published MDA is 1,300 feet msl. Witnesses on the airport reported heavy low fog and heard the pilot announce over the UNICOM frequency, 'Oh there is fog rolling into Starkville too?' One of the witnesses advised the pilot they could go to another airport due to the fog; the pilot responded he would execute the approach. The witnesses heard the engines operating at full power then heard the impact and saw a fireball. The airplane impacted the runway inverted, slid across the runway, and came to rest in grass off the runway. A post crash fire destroyed the airplane. Tree contact approximately 972 feet northwest of the runway impact location separated approximately 51 inches of the left wing. Examination of the engines, propellers, and flight controls revealed no evidence of preimpact failure or malfunction. The pilot had twice failed his airline transport pilot checkride. The designated examiner of the second failed flight test indicated the pilot was marginal in all flight operations. The NDB was checked after the accident; no discrepancies were noted.

The pilot reported the airplane decelerated during the takeoff roll. He applied the brakes and as he advanced the throttles to full power the airplane accelerated. The airplane cleared the end of the runway then stalled into Lake Michigan, flipped inverted and sank. One passenger reported that it felt as if someone put on the brakes. One passenger drowned. The pilot used 32' of manifold pressure for takeoff versus 37.3' as placarded. The pilot operating handbook lists normal takeoff speed as 91 KIAS, however the airplane was equipped with vortex generators. The pilot reported looking for 105 to 110 KIAS for takeoff. No evidence was found of the pilot having a multi-engine rating. No evidence of a mechanical failure/malfunction was found.

The pilot reported that shortly after takeoff, the airplane's left engine started to surge. The airplane also began experiencing intermittent electrical surges which caused the instrument panel lights, cabin lights, and radios to go off and on. The pilot diverted to an alternate airport to land. He did not secure the left engine before landing because it was still developing some usable power. He placed the gear select handle in the down position and observed three green gear-down-and-locked lights. Prior to touchdown, both power levers were positioned to flight idle and no gear warning horn sounded. The airplane landed gear up. Postaccident examination revealed no abnormalities with the landing gear or electrical system. The landing gear emergency extension functioned properly. The landing gear indicating system showed a safe gear indication when the gear was extended during examination. Substantial damage to the gear doors was observed, but no damage to the landing gear was observed.

A witness stated he observed the airplane on climbout from runway 32. The airplane started a right turn estimated at about 30 to 45° angle of bank. The airplane stopped climbing and began descending. Subsequently, it collided with a tree line, while in a right bank, and then it impacted the ground. Weather conditions at the time of accident were described by the witness as very dark, with no ambient light or visible horizon. Examination of the airframe, flight control system, engine assembly, and propeller assembly revealed no evidence of a precrash failure or malfunction. The autopilot was found in the off position, and the autopilot circuit breakers were not tripped. The pilot and passenger were seriously injured and had no memory of the flight. A radio transcript revealed that after taking off, the flight had made one radio transmission to request an ifr clearance.

The pilot was conducting his initial revenue and solo flight for this company, in this type of airplane. The weather for takeoff included fog and low ceilings. The airplane was equipped with a primary (left) attitude indicator which was electrically operated via an independent switch. This aircraft was the only such airplane operated by this company, with an independent switch configuration for the primary attitude indicator. The airplane collided with the terrain on the airport, just after takeoff. Subsequent examination revealed no anomalies with the engines or airframe. The primary attitude indicator was located. On examination it was found to have a malfunctioning on/off flag which gave the indication of being operative regardless of power to the unit. No rotational damage was noted within the gyro housing. The pilot, sole on board, was killed.

At 0515 est, the flight was cleared for takeoff on a flight from Ypsilanti, MI to Louisville, KY. About 17 minutes later, the flight crew began to display indications of a deterioration of their ability to control the aircraft. At 1st, they deviated from instruments to hold west of the Findlay VOR at FL220. As the flight continued and was cleared to FL270, the crew displayed confusion about magnetic headings and basic instruments. At 1048 est, the aircraft deviated from the en route heading and the wrong heading was read back after a heading correction was given. Also, the aircraft continued climbing (to FL291), then radar and radio contact were lost at 0551 est. The controller noted the pilot's speech was slurred and some portions of the conversation were unintelligible. Subsequently, the aircraft crashed in a steep dive. No preimpact part failure was verified, though impact forces and post-crash fire resulted in extensive damage of the aircraft. The aircraft was equipped with oxygen and pressurization system. No audible warning was noted on ATC recordings to indicate the cabin altitude had exceeded 10,000 feet, though the aircraft was equipped with such a device. Both pilots were killed.

The aircraft was enroute to the Chicago Midway Airport. While letting down to 16,000 feet the pilot reported he was shutting down the right engine. ARTCC cleared the aircraft to 5,000 feet. Ten minutes later ARTCC lost radio transponder and mode C contact. 15 minutes later the aircraft descended out of the 1,800 feet overcast and struck a 170 feet utility tower at the 150 feet level. The aircraft then ground impacted and burned. Investigation revealed that the right engine had a bearing failure due to lack of lubrication. The pilot, sole on board, was killed.

The aircraft collided with the ground after stalling during takeoff. Witnesses reported that the aircraft pitched up in two increments, climbed sharply to about 200 feet agl, stalled dropped the right wing and descended, impacting the ground in a flat attitude.