FTW95LA340

During takeoff on a mountain flying instructional flight, the aircraft (with a flight instructor & 2 pilots aboard) lifted off about halfway down the 4903' runway. However, after lift-off, it failed to climb sufficiently to clear the perimeter fence, which was on higher terrain about 1560' from the runway. The fence was at an elevation of about 60' above the end of the runway. The airplane came to rest in a field about 140' from the fence. No aircraft or system failure or malfunction was found. There was evidence that the fuel boost pump was on during takeoff (although the engine driven pump was operating normally). During a test run, the engine operated normally with the boost pump off. With the boost pump at low boost, there was a 500 to 700 RPM loss of RPM; at high boost, the engine quit running. The location and design of the boost pump switch was non-standard having been moved during an STC aircraft alteration. The switch was located next to the throttle, where it could be inadvertently operated, and it had no safeguard to prevent such operation. Investigation also revealed that there were no leaning procedures incorporated in the pilot manual. The instructor (CFI) had briefed the pilots to lean the engine by using the RPM. The manufacturer reported that the engine could not be properly leaned by using the RPM gauge, since it was equipped with a constant speed propeller.

On August 14, 1995, at 0842 mountain daylight time, a Beech J35, N8344D, sustained substantial damage during during takeoff at Centennial Airport, Englewood, Colorado. The three pilots aboard were not injured. Visual meteorological conditions (VMC) prevailed for this Title 14 CFR Part 91 local area mountain flying instructional flight and a visual flight rules flight plan was filed.

According to information provided by the pilots under instruction, (hereafter referred to as students), they came to the Denver area on August 11, 1995, in their aircraft, from Saint Louis, Missouri, for the specific purpose of receiving mountain flying training. They selected the instructor based on information they had received from a friend and on information provided to them by the flight instructor.

The students stated that they met the instructor shortly after 0700 the day of the accident. Prior to the instructor's arrival, the students said they had conducted a preflight of the aircraft and had received a weather brief from Denver Automated Flight Service Station. Weather was VMC and expected to remain so with the possibility of rain showers in the afternoon.

At the time they conducted the preflight, the main fuel tanks were serviced to full and the auxiliary fuel tanks were empty. They said they had the aircraft fueled in this manner to conserve weight as they had been told that fuel was available at the mountain airports they were going to.

Following a discussion of the flight and mountain flying with the instructor, the students were told by the instructor to have the auxiliary tanks filled and that the additional weight would not be a problem. The students were also told that they would depart using a short field takeoff using zero flaps.

They then filed a VFR flight plan and figured takeoff ground roll to be approximately 2,200 feet on runway 10 which, although it was the shortest runway, the instructor told them they would use due to the winds being 100 degrees at 10 knots and decreased taxi time.

According to the students, during the preflight discussion, they were not briefed on emergencies or given any instructions on who would fly the aircraft in the event an emergency should occur.

In the post-accident interview and written statement, the flight instructor related that she did not consider herself to be pilot in command.

The students said that they decided the least experienced would fly first. Following taxi to the hold short area and run up, the instructor (according to written statements and interview notes) told the student flying to hold the brakes, set the propeller to full increase, apply maximum throttle and lean the engine by sound until it ran rough and then advance the mixture two or three turns. The aircraft was equipped with both fuel flow and exhaust gas temperature (EGT) gauges and they were functional.

According to the attached instructor's statement, she showed the student flying how to lean the mixture by reducing the mixture until a slight rise in RPM was noted followed by a drop, then to add two to three good turns to adjust the mixture for altitude and then to watch the EGT gauge during takeoff roll to be sure the mixture was not too lean.

According to both the engine manufacturer's representative and Beech engineering, proper leaning using the rpm method is not possible on a constant speed propeller equipped aircraft.

According to the student flying, (see attached statement) she rotated the aircraft at 75 knots and though the aircraft felt sluggish, it lifted off and the gauges appeared normal. The student said she drifted left of centerline and the instructor added right rudder to bring the aircraft back. The student said she thought she felt a main landing gear touch down but then became airborne again climbing slowly. The flight instructor then suggested the student add some flaps and about 10 degrees were added. The instructor then reportedly tried to add some more flaps by reaching across the cockpit but she could not operate the flap lever. The student said she put down more flaps.

The instructor provided information in her statement which indicated the power appeared to be normal following liftoff and that the EGT gauge white needle was straight up with the red needle being to its right. The instructor also stated that they became airborne about halfway down the runway which is 4,093 feet in length.

According to the pilots, by this time the aircraft was passing over the overrun area approaching the airport perimeter fence which was about eight feet in height, wire, with large wooden posts. The student flying said she saw the fence and must have pulled up because the stall warning horn sounded as the aircraft impacted the fence. The student said the aircraft then impacted in a plowed field beyond the fence and slid to a stop. She said she secured the aircraft and during that evolution was surprised to find the fuel boost pump on low. All occupants then exited the aircraft.

An on scene investigation was conducted by this investigator. Witness marks and measurements provided evidence that the aircraft came to rest approximately 1,700 feet beyond the runway threshold and had struck the perimeter fence breaking off two posts. Terrain slope from the runway end to the aircraft was an upslope and the change in elevation from the end of the runway was estimated to be about 60 feet up.

The aircraft was sitting upright in a plowed field 140 feet beyond the fence. The landing gear was separated and the flaps were down 20 degrees as indicated by the position marks on the left flap. The propeller blades were bent in an 'S' curve and exhibited chord wise scratches and leading edge gouges. Both blades were loose in the hub. The leading edges of both wings were damaged and pieces of fence post wood were found imbedded in the wing leading edge impact crush area.

Examination of the engine/aircraft revealed that an engine test run could be conducted. This test run was conducted on August 22, 1995.

The engine started on the first attempt. Following leaning based on pressure altitude, using fuel flow/EGT, the engine produced 25 inches of manifold pressure and redline RPM with the propeller set at full increase. The magneto check was within limits as were all other engine parameters. When the boost pump was activated to the low position, the engine lost between 500 and 700 rpm and went into a bog down condition, becoming unresponsive to additions of power. When the boost pump was activated to the high position, the engine quit.

An examination of the propeller assembly by a propeller overhaul facility provided evidence that the pitch change links failed in overload due to impact forces.

Examination of the aircraft logs revealed that a supplemental type certificate (STC) had been issued covering a redesign of the center forward portion of the instrument panel. The STC relocated some gauges and the throttle, mixture, and propeller controls. It did not mention relocation of the fuel boost pump switch from the lower left panel to the center panel or the type of switch to be used.

Examination of the aircraft revealed that the boost pump switch was a three position switch: Up-high pressure, center-off, and down-low pressure. Switch design allowed the switch to be inadvertently activated and no safeguards to prevent inadvertent activation were incorporated. In addition, the switch was located beside the throttle in such a position that inadvertent activation was plausible. Tests conducted during the engine run supported this possibility. (See attached photographs).

A review of the aircraft operating handbook and airplane flight manual was conducted as regards leaning procedures. On page 15 Section IV of the Beechcraft Bonanza J35 handbook item 15, the procedure for leaning reads as follows: "Mixture - FULL RICH (or as required by field elevatio