Piper PA-61 Aerostar (Ted Smith 601)

During takeoff, the twin-engine airplane was observed to roll left, pitch nose down, and impact terrain shortly after the pilot reported to ATC that he had a problem. Witnesses reported that the left engine was producing black smoke during the takeoff roll. One witness stated that the airplane had slowed to approximately 60-70 mph prior to rolling to the left. A mechanic, who worked on the airplane prior to the accident, stated that the pilot reported being unable to maintain manifold pressure (MP) with the left engine. The mechanic found that the left engine's rubber interconnect boot, which routes induction air between the turbocharger controller elbow and the fuel servo, was 'gaping open.' The mechanic reseated the boot and tightened the clamp. The pilot flew the airplane and reported no problems. During a second flight, the pilot reported that the left engine was again unable to maintain MP. Prior to the accident flight, the pilot informed the mechanic that the 'hose had slid off again' and that it had been reinstalled and he 'felt sure it was o.k.' A witness stated that he saw the pilot working on the left engine the morning of the accident. At the accident site, the left engine's interconnect boot was found disconnected. The clamp securing the boot was not located. No other preimpact anomalies were found with the engines, propellers, turbochargers, or fuel servos.

The pilot departed on runway 06 with zero degrees of flaps. A witness said that she noticed that the airplane appeared to wobble and shudder, and immediately went into a steep right bank turn right after takeoff. The airplane then went into the clouds which were 200 to 400 feet agl. Radar data indicated that the airplane made several 90 degree turns prior to impacting the mountainous terrain 2.55 nm from the departure end of the runway. The pilot normally used 20 degrees of flaps for takeoff. A test pilot said that the airplane handles significantly different during takeoff if zero degrees of flaps are used verses 20 degrees of flaps. The upper cabin's entry door was found, with the locking handle and locking pins, in the closed position. No preimpact engine or airframe anomalies, which might have affected the airplane's performance, were identified.

The airplane took off from Port Huron, Michigan, on April 1, 1999, at 1130 est. The airplane was scheduled to arrive in Freemont, Ohio. An employee of the pilot's company said that the pilot was going to meet a customer there. At 1230 est, the customer called the company inquiring about the pilot. The employee said that the pilot 'would have taken the shortest route, over [Lake] St. Clair, Ontario [Province], and [Lake] Erie,' to get to Freemont, Ohio. An ALNOT was issued at 1803 est. Search and rescue operations were conducted by the U. S. Coast Guard, Civil Air Patrol, and the Canadian Search and Rescue Center. The search was suspended on April 10, 1999, at 2125 est. The passenger's body was discovered on May 1, 1999, in the Lake St. Clair shipping channel, approximately 6.9 miles east of St. Clair Shores, Michigan. On July 2, 1999, the pilot's body was found in Lake St. Clair. Parts of the airplane identified from the make and model of aircraft were recovered with the bodies.

Shortly after the airplane took off, a witness about 1 mile from the airport observed the airplane about 150 feet above the ground in a left turn, before it disappeared into the clouds. A witness across from where the airplane crashed stated he was in his shed when he heard the sound of an airplane overhead. When the sound faded and returned, like the airplane had circled above the shed, he stepped outside and looked for the airplane. He saw the airplane exit the clouds in a near vertical position and impact the ground. He described the engine noise as loud and smooth. The airplane impacted in a field about 3/4 miles from the departure airport and was consumed by a post crash fire. Streaks of oil were observed on the leading edge of the right horizontal stabilizer extending to its upper and lower surfaces. Disassembly of both engines did not reveal any pre-impact mechanical malfunctions. A weather observation taken after the accident reported included a visibility of 2 miles with light drizzle and mist, and the ceiling was 400 foot overcast. Witnesses described the weather at the accident site as '...pretty foggy,' and worse than the conditions reported at the airport.

During arrival at night, the flight was being controlled by a developmental controller (DC), who was being supervised by an instructor (IC). The pilot (plt) was instructed to descend & cross the STILL Intersection (Int) at 3,000 ft. STILL Int was aligned with the localizer (loc) approach (apch) course, 10.1 mi from the apch end of runway 17 (rwy 17); the final apch fix (FAF) was 3.9 mi from the rwy. About 5 mi before reaching STILL Int, while on course & level at 3,000 ft, the plt was cleared for a Loc Rwy 17 Apch. Radar data showed the aircraft (acft) continued to STILL Int, then it turned onto the loc course toward the FAF. Shortly after departing STILL Int, while inbound on the loc course, the acft began a descent. Before the acft reached the FAF, the DC issued a frequency change to go to UNICOM. During this transmission, the IC noticed a low altitude alert on the radar display, then issued a verbal low altitude alert, saying, 'check altitude, you should be at 1,500 ft (should have said '1,800 ft' as that was the minimum crossing altitude at the FAF), altitude's indicating 1,200, low altitude alert.' There was no response from the plt. This occurred about 2 mi before the FAF. Minimum descent altitude (MDA) for the apch was 720 ft. The acft struck tree tops at 750 ft, about 1/2 mi before the FAF. The IC's remark 'you should be at 1,500 ft' was based on an expired apch plate with a lower FAF minimum crossing altitude; the current minimum crossing altitude at the FAF was 1,800 ft. Apch control management had not made the current plate available to the controllers. Investigation could not determine whether a current apch plate would have prompted an earlier warning by the controllers.

About three hours and twenty minutes after departing Bellingham, Washington, for Midland, Texas, the pilot contacted Klamath Falls (Oregon) Tower and told the controller of his intention to land. About 10 minutes later, while about 30 miles north of Klamath Falls, the pilot reported he was low on fuel and was not able to find the city. The tower responded with instructions that would take the pilot south to the airport. But because the pilot seemed not to be following the instructions, but was instead continuing to the west, he was switched to Seattle Center. Center provided the pilot with a southeasterly heading direct to Klamath Falls, but less than a minute later radar and radio contact with the aircraft was lost. Other pilots overheard the pilot transmit that he had lost power in one engine, and later state that he had lost power in both. Soon thereafter the aircraft was seen to descend to about 200 to 300 feet above the surface of Lake of the Woods. The aircraft then began to slow and its nose began to rise. As it was slowing, one of the engines surged back to a high power setting, and the aircraft almost immediately rolled quickly to the side and dove nearly straight down into the lake. During the post-accident inspection of the airframe, the throttle for the right engine was found retarded to idle, but the throttle for the left engine was found in the full-forward (maximum power) position. A review of the Aerostar owner's manual revealed that the Engine Failure/Restart checklist called for the throttle for a failed engine (both engines in this case) to be retarded to the 'Cracked 1/2 inch open' position. Toxicological results indicate the presence in the pilot's blood of chlordiazepoxide and three of its active metabolites, norchlordiazepoxide, nordiazepam, and oxazepam. Chlordiazepoxide (Librium) is a tranquilizer often used to treat anxiety and tension. At sufficient levels it can have significant adverse effects on judgement, alertness, and performance. It is known to cause drowsiness, mental dullness, and euphoria. The results also indicate the presence of diphenhydramine in the pilot's blood. Diphenhydramine is a sedating antihistamine, and in sufficient quantities is known to produce drowsiness, impaired coordination, blurred vision, and reduced mental alertness.

The airplane departed the airport and crashed shortly thereafter. Before departure, the airplane was fueled with 120 gallons of 100LL aviation fuel. According to the refueler, the airplane had full fuel tanks. The refueler also indicated the pilot had stated he wanted to be airborne prior to the arrival of bad weather. After the accident, the engines and propellers were disassembled and examined. No engine or propeller discrepancies were noted, except (post impact) heat damage.

The non instrument-rated pilot filed an IFR flight plan, but did not request nor was given a weather briefing. Shortly after taking off into low instrument meteorological conditions, he reported he was returning to the airport, but did not give a reason why. He never declared an emergency. The last transmission was when the pilot said he had 'the problem resolved,' and was continuing on to his destination. Various witnesses said the engines were 'revvying' and 'unsynchronized,' and that the propellers were being 'cycled.' One witness said brownish-black smoke trailed from the right engine. The airplane struck one ridge, then catapulted approximately 1,000 feet before striking another ridge. There was post impact fire. Both propellers bore high rotational damage. Disassembly of the engines, propellers, turbochargers, and various components failed to disclose what may have prompted the pilot to want to return to the airport. Internal components of the right engine, however, were black and, according to a Textron Lycoming representative, were indicative of 'an excessively rich mixture.' A psychiatrist had recently treated the pilot for depression, attention deficit and bipolar disorders. The pilot also had a history of alcohol and drug abuse. Postmortem toxicology protocol disclose the presence of Fluoxetine (an antidepressant), Norfluoxetine (its metabolite), and Hydrocodone (the most commonly prescribed opiate).

The pilot had departed Swainsboro, Georgia at 1930 EST on an IFR flight. About 12 minutes later, he informed Jacksonville Center that he was very dizzy and could not see. There were no other recorded transmissions from the pilot. The airplane was tracked on radar until radar contact was lost at 2130. The airplane was at a heading of 110 degrees and an altitude of 9,000 feet the entire time. Attempts to locate the airplane by aerial intercept were uneventful. All shipping vessels along the airplane's expected course, were notified of the airplane's estimated fuel exhaustion point. No contact was reported and the search was suspended. Prior to departing Swainsboro, the pilot had mentioned to his wife that he had a headache. A review of the pilot's medical records revealed that he had twice indicated on his application for a medical certificate that he had a medical history of unconsciousness. In addition, he was being treated for hypertension with Norvasc and chlorthalidone prescription drugs.

The pilot stated that the airplane was in cruise flight at 8,000 feet MSL, when the right engine lost power. He advised ATC of the loss of power and received radar vectors toward an airport. The pilot said he maintained the best single-engine rate-of-climb speed, but the airplane's altitude 'drifted down.' When the airplane entered clouds, it began to accumulate structural icing and would not maintain sufficient altitude. The airplane impacted mountainous terrain about 16 miles northwest of the airport. The pilot stated that he had departed on the cargo flight with 5 hours of fuel on board for what he estimated to be a 2 1/2 hour flight. Also, he reported that conditions were dark and foggy, when the accident occurred. Postaccident examination of the engines and their systems revealed no evidence of preimpact mechanical malfunction. Examination of the airplane wreckage revealed no evidence of preimpact failure of the airframe or its systems. During a postaccident engine test run, the right engine started normally and operated satisfactorily.