NTSB Factual Report

HISTORY OF FLIGHT

On June 17, 2000, approximately 1200 Pacific daylight time, an experimentally certificated homebuilt Haessler Cosmo Phase II "Trike," N9022F, registered to and being operated by a private pilot, was destroyed during an in-flight collision with terrain following a loss of control while maneuvering, about two miles southeast of Banks, Oregon. The pilot was fatally injured. Visual meteorological conditions existed and no flight plan had been filed. The flight, which was personal, was operated under 14CFR91, and originated from a field near the pilot's residence earlier on the morning of the accident.

A witness reported that about noon he heard the noise of an airplane, looked up, and saw the aircraft fly overhead about 150 to 200 feet above ground heading east. The witness stopped following the aircraft but reported that shortly thereafter "...something made me look back up, maybe a change of noise, when I saw the plane it had made a 180 degree change in direction and was already out of control..." He further described the aircraft as being in a nose down attitude during which it made two or three turns, and that he did not hear any noise "...except when the plane went into the dive, the engine revved to high R.P.M.'s..." (refer to STATEMENT CH-I).

PERSONNEL INFORMATION

According to records maintained by the Federal Aviation Administration (FAA), the pilot (owner/builder) was issued a private pilot certificate with airplane single engine land rating on April 9, 1999. His personal logbook showed his first instructional flight commencing December 15, 1998, and the last flight was entered June 15, 2000. His total logged flight experience was approximately 133 hours of which the majority was in Cessna model 150 and 172 type aircraft.

The first flight in the accident aircraft make/model was logged by the pilot on June 9, 2000 (N9022F) and was for a duration of 0.2 hours. The second flight in the accident aircraft make/model was logged June 13, 2000 (N9022F) and was for a duration of 1.0 hours. The third flight in the accident aircraft make/model was logged June 15, 2000 (N9022F) and was for a duration of 1.4 hours. This flight was the last flight entered in the pilot's logbook.

AIRCRAFT INFORMATION

The aircraft, which was manufactured by Cosmos, a French company, was assembled and modified by the owner/builder in Oregon. The model was a Phase II generically known as a "Trike." The owner/builder obtained a special airworthiness certificate for the aircraft on June 6, 2000, with an experimental certification. The aircraft was powered by a 65 horsepower Rotax engine equipped with two three-bladed propellers longitudinally stacked and mounted at the rear of the engine and aft of the pilot/passenger seats. The Cosmos Phase II was a tandem seated, two placed aircraft, which, by virtue of its 14 gallon gas tank, weight, and dual seating was not qualified for ultralight status in the United States (refer to ATTACHMENT GI-I). The Phase II was controlled by means of a control bar which allowed the pilot to shift the center of gravity in the same fashion a hang-glider is controlled. This control system did not include a vertical stabilizer or rudder system (refer to ATTACHMENT GI-II).

The owner/builder modified the aircraft with the addition of a vertical stabilizer and controllable rudder at the end of an aluminum tube attached to the structural tube at the aft trailing edge of the wing (longitudinal centerline). According to entries in the owner/builder's flight log and corroborated with his personal flight log, the aircraft had a total 2.6 hours of flight time prior to the accident flight (Refer to ATTACHMENTS AL-I and EL-I).

The aircraft could be broken down into two basic parts by detaching the "trike" portion (i.e., engine, pilot/passenger gondola) from the wing and added rudder assembly. This would necessitate disconnection of the rudder cables either at the rudder pedals or their opposing ends (the attach points at the rudder bellcrank - refer to photograph 5).

METEOROLOGICAL INFORMATION

The aviation surface weather observation for the Hillsboro airport, located seven nautical miles east of the accident site, for 1153 hours Pacific daylight time on the date of the accident reported the following:

Sky condition clear, visibility 10 statute miles, winds from 350 degrees magnetic at four knots, temperature 26 degrees C., dew point 12 degrees C., altimeter setting 29.92 inches of mercury.

WRECKAGE AND IMPACT INFORMATION

The aircraft crashed at the edge of a blackberry field on a narrow dirt/mud lane. The accident site coordinates were estimated to be 45 degrees 35.5 minutes north latitude and 123 degrees 6.0 minutes west longitude (refer to CHART I). The wreckage was confined to a small area and all major components from the aircraft were located at the ground impact site.

The left wing displayed extensive leading edge deformation while the leading edge of the right wing was relatively undamaged (refer to photographs 1 and 2). The gondola or "trike," including the two seats and engine, was observed lying on its left side directly aft of the wing assembly. The two tandem mounted three-blade propellers were observed to have remained attached to the Rotax engine crankshaft and displayed successive blade damage decreasing circumferentially (refer to photograph 3). An aluminum rudder/stabilizer tube, approximately three inches in diameter was observed separated from the wing at its forward end and separated from the rudder panel. This assembly was observed lying on the top of the wing (refer to photograph 3).

The control cables from the rudder bellcrank at the end of the attach tube forward to the rudder pedals were found to be continuous. The rudder pedals were observed to be jammed in the full right rudder position (right pedal forward).

MEDICAL AND PATHOLOGICAL INFORMATION

Post-mortem examination of the pilot was conducted by Karen Gunson, M.D., at the facilities of the Oregon State medical Examiner's Office, 301 NE Knott Street, Portland, Oregon, on June 18, 2000.

Toxicological evaluation of samples from the pilot was conducted by the FAA's Toxicology Accident and Research Laboratory, Oklahoma City, Oklahoma. All tests were negative (refer to attached report).

TESTS AND RESEARCH

The wreckage was re-examined in detail at a storage facility in Hillsboro, Oregon, where it had been transported to. The examining FAA inspector reported that "...the rudder cables were connected opposite to conventional rigging practices (refer to photographs 4 and 5). When the right rudder was applied at the foot pedals, a left deflection was observed..." (refer to ATTACHMENT FAA-I).

Additionally, the re-examination revealed that the throttle had been relocated from the control bar held by the operator to the foot pedals. The throttle control had been modified so that the top half of the foot pedals were hinged. Foot pressure on the upper hinged portion would then result in an increase of power (refer to ATTACHMENT FAA-I).

ADDITIONAL INFORMATION

An acquaintance of the owner/builder provided a statement regarding his earlier conversations relative to the aircraft. He reported in part that he and the owner/builder had a conversation on June 16, 2000, and that he (owner/builder) "...had flown it on Thursday, 15 June 2000, and had experienced problems with controllability when using the rudder. Although the original design of the aircraft did not include a rudder, he had added one at the end of the central boom..."

Additionally, he reported that the owner/builder "...stated that, during his flight on Thursday, 15 June 2000, he experimented with sideslipping while at an altitude of approximately 2000 feet AGL [above ground level] and experienced a problem with rudder control. With 'small' rudder inputs..." "...the aircraft would yaw in the appropriate direction (i.e., left yaw with left rudder and vice versa). He stated that, with small rudder inputs, the aircraft responded similarly to light aircraft he had flown, such as the Cessna 152. With larger rudder inputs..." "...the aircraft would initially yaw in the appropriate direction. Then, after a delay of less than a second, would begin a rapid uncommanded roll in the opposite direction (i.e., right roll with left rudder). The roll would continue as long as he held pressure..." The owner/builder further remarked that on at least one occasion the aircraft reached a roll angle between 45 and 60 degrees angle of bank and that at higher airspeeds, the rate of roll was greater while the delay before the initiation of the roll was shorter (refer to STATEMENT JL-I).

Initial on-site examination of the wreckage was conducted by an inspector from the FAA's Hillsboro Flight Standards District office on June 17, 2000. The wreckage was released for removal and subsequently re-examined by an FAA inspector on June 21, 2000. The wreckage was verbally released following this latter examination to a representative for the owner.

NTSB Probable Cause

The pilot's improper installation, i.e., reversed connection, of the rudder control cables resulting in reverse rudder operation and a subsequent stall/spin. A contributing factor was the pilot's intentional design change/addition of a rudder to the aircraft.