NTSB Factual Report

HISTORY OF FLIGHT

On November 21, 1994, at 2000 eastern standard time, a Piper PA-28-181, N2949Q, co-owned and piloted by Harry E. Gee, was substantially damaged when it collided with trees during a night instrument approach to the Doylestown Airport, Doylestown, Pennsylvania. The pilot was fatally injured, and the passenger received serious injuries. Instrument meteorological conditions prevailed, and an instrument flight rules (IFR) flight plan had been filed for the flight conducted under 14 CFR Part 91.

The pilot, Mr. Harry Gee, was hired by the passenger, Mr. James C. Dalton, to fly him from Doylestown to Elmira, New York, and return, on November 21, 1994. Mr. Gee departed Robbinsville, New Jersey, where N2949Q was based, about 0830, destined for Doylestown to pick up Mr. Dalton.

Mr. Dalton stated that he met Mr. Gee at the Doylestown Airport (N88), and they departed about 0900 for the Elmira- Corning Regional Airport (ELM). Upon arrival at ELM, Mr. Dalton left the airport to conduct his business and returned, about 1750. Mr. Gee advised Mr. Dalton that, due to weather, the departure had to be delayed. Mr. Dalton believed that Mr. Gee then left to check the weather. When Mr. Gee returned, he stated that "they had a break in the weather," and were fueled and ready to go. They departed for N88 about 1830.

Mr. Dalton stated that the flight was conducted between cloud layers, and he described the flight as routine. Mr. Dalton could hear Mr. Gee talking on the radio, but could not hear clearly enough to understand what was said. Later in the flight, while the airplane was descending, Mr. Dalton was able to visually see the ground and recognize where they were. Mr. Dalton was able to maintain visual contact with the ground and recognize N88 when they flew over it. They circled the airport and flew over the runway twice, when he thinks they started a third circle. Mr. Dalton did not recall the last circle and the accident.

Mr. Dalton was not concerned when they circled the airport, and assumed that Mr. Gee was just getting his orientation to the airport and runway. The runway lights at N88 were visible during some of the circling. Mr. Dalton stated that Mr. Gee did not appear to be upset or worried, and he appeared to be concentrating on flying the airplane and the approach to the airport. Mr. Gee was talking on the radio while they were circling the airport.

Prior to the takeoff from ELM, Mr. Dalton observed Mr. Gee use a flashlight to look at charts on his lap because the airplane dome light did not work. He did not recall whether Mr. Gee used the flashlight at any other point during the flight. Mr. Dalton did not recall any problems with the airplane or the engine during the flight to ELM, or on the return flight.

According to the Elmira Flight Service Station (FSS) records and briefer statements, Mr. Gee received three in-person weather briefings at the Elmira FSS pilot briefing counter. These occurred at 1328, 1713, and 1805. Each briefing was a complete, standard weather briefing. At the completion of the 1713 briefing, Mr. Gee filed an IFR flight plan with the FSS. During the 1805 briefing, the FSS briefer recommended that the pilot take along all the weather information he had printed out, and for the pilot to consider delaying his flight until the next morning, when the weather was expected to improve. Mr. Gee took the weather information and indicated he would be leaving soon.

According to air traffic control records, N2949Q was issued an IFR clearance from ELM to N88 and was cleared to takeoff, at 1831, by ELM tower. The en route flight from ELM to N88 was unremarkable, and N2949Q arrived in the Philadelphia Approach Control area, at 1941. During the initial contact with the controller, N2949Q reported level at 5,000 feet. A clearance for the VOR RWY 23 instrument approach to N88 was issued to N2949Q, at 1949. This was followed by a frequency change to the advisory frequency for the airport, at 1950. N2949Q responded to the frequency change with, "Roger right now we are still in clouds." No further transmissions were received from N2949Q.

Several members of the local Civil Air Patrol (CAP) chapter were at N88 at the time of the accident. One witness was the Squadron Commander. She stated in a telephone interview that between 1950 and 2000, she and several of the CAP members observed an airplane make low passes to N88. She was standing on the west side of runway 23. The airplane was approaching N88 from northeast to southwest, flying low and slow. As the airplane flew overhead, the engine sounds increased, and the airplane performed a climbing left turn to the east. She stated that it did this twice.

She further stated she could see the airplane and the navigation lights located on the wings. She also thought the landing light was on during these low passes. Each time the airplane climbed and turned east away from N88, she was able to maintain visual contact with the airplane as it maneuvered for another approach. It appeared to her that the pilot was attempting to remain below the clouds. During the last go- around, she lost sight of the airplane in the clouds. A short time later she heard a "cracking" sound when the airplane crashed. She estimated that the airplane was flying between 300 and 500 feet above the ground. She heard no unusual noises from the engine.

The airplane struck 60 to 70 foot high trees and impacted the ground 1/4 mile northeast of N88's runway 23.

The accident occurred during the hours of darkness at approximately 40 degrees, 20 minutes north latitude, and 75 degrees, 8 minutes west longitude.

PILOT INFORMATION

The pilot, Mr. Harry E. Gee, held an Airline Transport Pilot Certificate with a rating for airplane multiengine land, and a Commercial Pilot Certificate with a rating for airplane single engine land.

His most recent Federal Aviation Administration (FAA) Second Class Medical Certificate was issued on April 12, 1994.

Mr. Gee's pilot log book indicated that his total flight time was approximately 2,500 hours, of which 1,400 hours were in single engine airplanes, and about 40 hours in this make and model. During the previous 12 months, Mr. Gee had logged 10 hours of actual weather flight and 1 instrument approach.

METEOROLOGICAL INFORMATION

The weather forecasts provided to Mr. Gee, by the ELM FSS consisted of:

North Philadelphia (PNE) Until 2100 EST - ceiling 1,500 overcast, light rain showers, winds from 180 degrees at 10 knots, gusting to 20 knots, low level wind shear; occasional 700 overcast, 3 miles visibility, rain showers and fog.

Philadelphia (PHL) Until 2100 EST - ceiling 900 broken, 4,000 overcast, light rain showers, winds from 180 degrees at 10 knots, gusting to 22 knots, low level wind shear; occasional 900 scattered, ceiling 3,000 broken 2 miles visibility, light rain showers and fog.

Atlantic City (ACY) Until 2200 EST - ceiling 500 overcast, winds from 160 degrees at 15 knots, gusting to 25 knots, low level wind shear; occasional 500 scattered, ceiling 2,000 overcast 5 miles visibility, light rain showers and fog.

Weather reported in the area at the time of the accident was:

Navy Willow Grove, (8 miles southwest): 1955 EST - Indefinite 100 foot sky obscured, 1/4 mile visibility, light drizzle and fog, winds calm.

Trenton-Mercer County, (14 miles southeast): 1950 EST - Measured 400 foot overcast, 2 miles visibility, rain and fog, winds from 320 degrees at 5 knots.

Allentown, (24 miles northwest): 1957 EST - 400 scattered, measured 1,500 foot overcast, 1 1/ 2 miles visibility, light rain and fog, winds from 250 degrees at 4 knots.

AERODROME INFORMATION

Runway 23 at Doylestown was equipped with a Simplified Abbreviated Visual Approach Slope Indicator (SAVASI), with a glide slope set for 5.0 degrees. The angle measured from the runway to the accident tree tops was 2.4 degrees.

The runway was also equipped with Medium Intensity Runway Lights (MIRL). The runway lights were controlled by an automatic timer and a light sensor. The timer prevented the lights from activating during the day when there was an overcast. Once the timer was activated, the sensor was then free to turn the lights on, which would remain on continuously during the hours of darkness.

WRECKAGE INFORMATION

The airplane wreckage was examined at the accident site on November 22, 1994. The examination revealed that all major components of the airplane were accounted for at the scene. The airplane came to rest inverted, in an open field, on a magnetic bearing of 245 degrees, at a ground elevation of about 420 feet above mean sea level (MSL).

Initial tree impact scars started approximately 315 feet from the wreckage. Numerous tree tops and subsequent lower branches were broken on a 200 foot path along a magnetic bearing of 230 degrees, in line with runway 23 of the Doylestown Airport.

Tree impact scars remained at the tops of the 60 to 70 foot high trees for about 50 feet. At the 50 foot mark, a tree branch several feet long, was observed on the ground with the left wing tip faring. From the 50 foot mark, broken branches became progressively lower on the trees in the direction of the wreckage. Numerous pieces of clean white paint chips were observed on the ground along the tree impact path.

The left main wheel was located in the initial ground scar, about 100 feet beyond the end of the tree line. The main wreckage came to rest inverted about 120 feet beyond the tree line. The right wing remained attached to the fuselage with minor damage to the leading edge. The left wing was rotated around with the leading edge facing aft, and held onto the fuselage by only a cable. The outer 1/3 of the left wing leading edge was dented and compressed inward up to 2/3 of the chord of the wing.

The right horizontal stabilator was attached to the fuselage and undamaged. The left side stabilator was twisted and dented with the outer third bent upward.

Control continuity was established from the stabilator and rudder into the fuselage where the floor was collapsed onto the cables. Continuity was established from the pilot controls to the right wing aileron and to the point where the cables exited the fuselage on the left side.

The right wing fuel tank was leaking fuel from the inverted fuel cap; however, approximately 5 gallons of fuel remained in the right tank. The left fuel tank was estimated to contain about 15 gallons of fuel. Fuel samples were taken from each wing tank. The samples were tested with water finding paste and found to be absent of water. The electric fuel boost pump on the engine firewall was damaged. Examination of the fuel filter revealed the screen was free of debris and obstructions.

The right side of the fuselage cabin was intact, while the left side was compressed inward and mud covered. The right side of the engine compartment was intact with minor damage to the cowling area. The left side engine cowling was separated from the engine compartment.

The engine was canted toward the left side of the airplane. The engine was removed from the wreckage for further examination.

The propeller blades and hub remained attached to the engine. The propeller spinner was compressed inward and displayed some rotational twisting of the metal. One propeller blade was relatively straight with a slight chord wise twist. A few faint chord wise scratches were observed on this blade. The other propeller blade was curved aft and displayed chord wise twisting. Several deep chord wise scratches and indentations were observed on the leading edge of the blade.

Examination of the cockpit area revealed the flap handle was extended to the first notch. Both VOR navigation receivers were set on 108.2, and the ADF receiver was set to 237.

MEDICAL AND PATHOLOGICAL INFORMATION

An autopsy was performed on Mr. Harry Gee, on November 23, 1994, by Dr. Halbert E. Fillinger, Jr. M.D., Forensic Pathologist, of the Bucks County Coroner's Office, Doylestown, Pennsylvania.

The toxicological testing report, from the FAA toxicology Accident Research Laboratory, Oklahoma City, Oklahoma, was negative for drugs and alcohol for Mr. Harry Gee.

TESTS AND RESEARCH

The engine was removed from the wreckage and shipped to the Textron-Lycoming Plant, Williamsport, Pennsylvania. On January 12, 1995, in the presence of the NTSB IIC and the parties to the investigation, the engine was removed from the shipping crate and prepared for an engine run.

The upper spark plugs were removed from each cylinder and the engine crankshaft flange was rotated. Valve train continuity was established and compression was confirmed by the thumb method. The vacuum pump was removed, bench checked, and found to operate normally. The oil pump screen was removed and found absent of debris. The left and right magneto timing was measured to be 27 degrees, plus or minus 2 degrees. A 1/4 inch hole in the oil sump was cleaned and plugged for the test. The lower number four spark plug lead was partially cut; therefore, the left hand spark plug harness was replaced for the test.

On January 13, 1995, the engine was mounted in a test cell. The starter motor would not rotate the engine more then two times. A replacement starter motor was installed and the engine started without hesitation. A complete engine run was accomplished over the next 30 minutes, and the engine was found to operate normally and produce rated power.

ADDITIONAL INFORMATION

Doylestown Instrument Approaches

At the time of the accident, the instrument approaches published for N88 were the VOR RWY 23 and the NDB RWY 23. Both approach plates were found in the cockpit of N2949Q.

The VOR RWY 23 approach to N88 was via a course defined by the Solberg VOR (SBJ) 240 degree radial (R-240). This required the SBJ VOR to be tuned on an aircraft navigation radio. While navigating to N88 via the R-240, position fixes could be determined by tuning in the Yardley VOR (ARD). There were two published cross fixes depicted on the VOR RWY 23 approach plate. The initial approach fix, Fling intersection, was defined via the ARD R-014. The final approach fix (FAF), Groom intersection, was defined via the ARD R-340.

The missed approach point on the approach plate was defined using the SBJ R-240 and the 22.8 mile DME (distance measuring equipment) fix, or measured time from Groom to the runway. The minimum descent altitude published for this approach was 1,020 feet above mean sea level (MSL). The published missed approach procedure for the VOR RWY 23 approach, was a climbing right turn (westerly) to 2,000 feet, and proceed to the Groom Intersection via the SBJ R-240.

The NDB RWY 23 approach to N88 was via a course defined by the Doylestown NDB (DYL) 237 magnetic bearing to DYL. The DYL beacon was located on the northwest side of runway 23, and was also the missed approach point. The NDB approach plate depicted a transition course from ARD to DYL, using the ARD R-307.

Airplane Instrumentation

The investigation revealed that both VOR receivers in N2949Q were set to the ARD VOR frequency of 108.2 MHZ. Each VOR receiver had a co-assigned course deviation indicator (CDI). The number one VOR receiver CDI was set to 350, the selection required to identify the FAF of Groom (ARD R-340). The number two VOR receiver was set to 320, a point about 2.4 miles, or half the distance from the FAF to the runway. The airplane's Automatic Direction Finder (ADF) receiver was set to the DYL NDB frequency of 237 KHZ.

Radar Data

The NTSB Office of Research and Engineering conducted a Recorded Radar Study. The report depicted N2949Q descending out of 1,000 feet MSL about 3/4 mile northeast of the runway. A gradual descent continued until N2949Q crossed the airport at 700 feet MSL, in the vicinity of the DYL NDB, at 1954. For approximately the next 5 minutes, N2949Q continued in oval shaped patterns around the Doylestown Airport, between 500 and 800 feet MSL.

NTSB Probable Cause

the pilot's descent below the published minimum circling altitude, and his failure to execute the published missed approach procedure during a night IFR instrument approach, which resulted in an in-flight collision with trees.