NTSB Factual Report

HISTORY OF FLIGHT

On June 29, 2014, at 0917 eastern daylight time, a Czech Sport Aircraft Piper Sport, N957MD, was substantially damaged during a forced landing in Peachtree City, Georgia. The private pilot was seriously injured. Visual meteorological conditions prevailed, and no flight plan had been filed for the flight from Atlanta Regional Airport – Falcon Field (FFC), Peachtree City, Georgia, to Newnan Coweta County Airport (CCO), Atlanta, Georgia. The personal flight was conducted under the provisions of 14 Code of Federal Regulations Part 91.

According to the pilot's wife, and due to the extent of his injuries, the pilot was not able to recall "anything from that day," except that there was a loss of engine power for unknown reasons.

According to a witness, she was watching the airplane take off [from runway 31] and it became clear that the airplane was not going to clear the trees beyond the runway. The airplane veered to the right, like it was trying to turn around, "but fell like a rock nose first into the ground."

In a recorded television interview, a golfer who had been with a group on a nearby golf course stated that he saw the airplane taking off from the airport, and that the engine was "sputtering just like it was out of gas or not hitting on all cylinders." He thought the airplane was returning to the runway and was headed for some trees when it suddenly turned and dove nose-first into the ground.

The golfers raced to the airplane, and when they arrived, they saw fuel on the ground. They tried to extricate the pilot who could not move and remained secured via a seat belt. One of the golfers turned the ignition off and removed the keys. The pilot was subsequently removed and taken to the hospital.

AIRCRAFT INFORMATION

The airplane was powered by a Rotax 912ULS carbureted engine, driving a three-blade composite propeller.

According to the pilot's wife, the pilot's father originally owned the airplane. When he passed away in 2013, the pilot's mother gave it to the pilot.

Maintenance records indicated that, on October 20, 2011, at 36.3 hours of operation, a condition inspection was performed on the airplane. At 67 hours of operation (no date noted), another condition inspection was performed. At the time, the spark plugs were replaced. On November 20, 2013, at 93.2 hours of operation, another condition inspection was performed. At the time, the spark plugs were "cleaned, gapped and tested."

On December 2, 2013, the accident pilot's brother and a flight instructor flew the airplane. According to the flight instructor, the accident pilot's brother made the takeoff, and shortly thereafter, a "couple" hundred feet in the air, the engine rpm started decreasing to about 3,000 rpm. The engine deceleration was rapid and the engine never quit, but the flight instructor thought it was beginning to seize, and it "sounded horrible." He thought from what he saw on the engine analyzer, that there was a partial loss of power from the right side of the engine. The flight instructor took control of the airplane and subsequently flew a teardrop pattern back to the runway without further incident. The flight instructor was not sure if the airplane would have maintained altitude since the engine rpm was continuing to decrease and he was more concerned with making it back to the runway safely.

After landing, a witness advised the flight instructor that the engine had been blowing black smoke out of the exhaust. Other individuals stated that the engine likely had a stuck or sunk carburetor float. However, subsequent examination by a mechanic revealed no anomalies except that the dark appearance of the spark plug electrodes indicated that the fuel mixture had been burning rich. The flight instructor had a mechanic clean the spark plugs, and they drained fuel from the sumps, after which, there were no further difficulties with engine power that he knew of until hearing about the accident. There was no airplane logbook entry regarding the cleaning of the spark plugs.

After that incident, the flight instructor flew with the accident pilot's brother a couple more times in a local traffic pattern. There were some engine cooling issues, with cooling fluid coming out of an overflow valve, but after the cooling system was "burped," there were no further issues.

The flight instructor also noted that several months prior to the accident flight, on March 2, 2014, he and the accident pilot flew the airplane about 3 hours from Spruce Creek, Florida, to FFC. During that flight, the airplane performed "flawlessly."

The flight instructor further stated that the airplane had originally utilized automotive gasoline (MOGAS). With no MOGAS available, they utilized 100LL aviation gasoline to fly to Peachtree. Upon arrival, someone came up to them and said he had one or two of the same type of airplanes and knew of a mechanic to use.

The flight instructor suggested to the accident pilot that he get a Rotax engine specialist to go through the engine at the next inspection. The accident pilot said he was going to get a flight instructor and mechanic, but the flight instructor did not know if he ever did. The flight instructor also noted that the accident pilot was an excellent pilot who utilized "correct" procedures.

According to the responding Federal Aviation Administration inspector, a local flight instructor who flew with the pilot twice in the airplane stated that both times they flew together, the fuel pressure gauge read "high." After the second flight, the flight instructor told the pilot that he needed to "get that fixed before we fly again." The flight instructor spoke with the pilot at a later date, and the pilot stated that he was working on getting the fuel pressure gauge repaired.

According to the pilot's wife, the airplane was evaluated for safe flight to CCO by a mechanic in order to conduct a thorough maintenance review in the mechanic's hangar.

The mechanic stated that a couple of weeks before the accident, the pilot had called him, noting that the mechanic was certified on the Rotax engine, and that he would like to bring his airplane over to the mechanic's airport (CCO) so the mechanic could look at the engine. The mechanic advised to bring it over when he could.

Within the next couple of weeks, or several days before the accident, the mechanic and about five other mechanics were working on a DC-3 on the upper ramp at FFC. He heard the pilot's airplane take off, and at that time, the engine "didn't sound right." He added, a Rotax 912 engine normally had a hum to it, but that engine sounded abnormal.

The next day, the mechanic was back working on the DC-3 and went over to meet the pilot for the first and only time. They had a discussion about the airplane and the engine. The mechanic told the pilot that the engine did not sound correct on the previous flight. During the discussion about the engine and in just looking it over, the oil dip stick was pulled and no oil registered on the dip stick. The propeller was turned through about 7-8 times at least twice, with no oil ever showing on the dip stick. The pilot said, "I can assure you there is oil in the engine." The mechanic left at that point, never looked into the oil tank to see if it had oil in it, and did not evaluate the airplane as safe for flight.

A representative of the engine manufacturer noted that it was possible for an engine to be turned over 10 to 15 times before an indication of oil would show on a dipstick. The key was to listen for a "gurgling" sound before checking the dipstick.

PILOT INFORMATION

The pilot, age 55, held a private pilot certificate. He reported 111 hours of total flight time, with 18 hours in airplane make and model.

AIRPORT INFORMATION

Runway 31 was 5,219 feet long and 100 feet wide, at an elevation of 807 feet. Published airport information indicated 60-foot trees, 500 feet from the end of the runway and 525 feet right of centerline. Internet satellite views of the airport revealed a road that was perpendicular to the runway, about 575 feet from runway end, with small trees on either side in the vicinity of the extended runway centerline. Beyond the road was a golf course, and beyond that, a lake.

METEOROLOGICAL INFORMATION

Weather, recorded at the airport at 0853, included clear skies, calm winds, temperature 25° C, dew point 22° C, and an altimeter setting of 30.24 inches Hg.

WRECKAGE AND IMPACT INFORMATION

Photographs of the airplane revealed that it came to rest on the shoulder of the road that ran perpendicular to the runway. The airplane exhibited severe front end crushing.

The airplane was later moved to a recovery facility where, on August 29, 2014, an additional examination was conducted under NTSB oversight. Upon initial examination, the Rotax 912US engine, serial number 6776264, exhibited impact damage, including damage to the oil system pump and oil tank. Both carburetors were displaced, and the coolant radiator was crushed and breached.

Two of the three composite propeller blades were found broken off from the hub.

The engine was subsequently removed from the airframe and further examined on a workbench. Crankshaft continuity, piston movement, and cylinder compressions were confirmed.

When the NGK DCPR8E spark plugs were removed, the electrodes were found to be black and some were encrusted with debris.

The lubrication system was breached and most of the oil had drained. However, there was no evidence of inadequate lubrication within the engine.

The propeller gearbox was inspected for condition, modifications, proper components, smooth operation and for any unusual sounds or other discrepancies, with no anomalies noted.

Fuel was supplied to the engine via an engine-driven pump and an electric-driven supplemental pump. The electric fuel pump was an automotive design that functioned when tested without any anomalies noted. The mechanical fuel pump was the original fuel pump that came from Rotax. It was recommended to be replaced with a new-style pump per Rotax Service Bulletin 912-063UL due to possible malfunction.

The engine was subsequently prepared for an engine run. Due to the extent of impact damage, the carburetor for cylinders Nos. 2 and 4 was replaced with a carburetor from another engine, and the exhaust muffler, oil tank, carburetor air box, and coolant radiator were removed. The ignition wires were modified to operate the engine safely, and connected to the airplane's rotary switch for starting and stopping the engine. The original spark plugs were used. The fuel system was simplified with the use of a suction hose from a fuel canister. The electric fuel pump and airframe gascolator were not utilized, and the remaining propeller blade was cut off.

A fork lift was used as a makeshift engine test stand and the engine was secured to the forks. A battery was connected to the starter of the engine, and the rotary switch was turned to start the engine. The throttles were manipulated by hand and the engine ran for several minutes at varying rpm settings until it was manually shut off by turning the rotary switch to the off position. No anomalies were noted during the engine test run. The maximum rpm obtained could not be quantified.

ADDITIONAL INFORMATION

Electronic Devices

The airplane was equipped with several electronic devices with downloadable data. Data from those devices revealed that the airplane impacted the ground about 1 minute, 10 seconds after the application of takeoff power.

About 10 seconds after the application of takeoff power, engine rpm stabilized in excess of 3,800 rpm, and during that time, the nose of the airplane assumed a takeoff attitude of about 9-10 degrees. Groundspeed rose to about 60 knots, while indicated airspeed was 57 knots.

Forty seconds after the application of power, the engine rpm dropped about 400 rpm, then recovered 300 rpm, then gradually diminished over the next 30 seconds to a final recorded rpm of 3,237, when the data stopped recording. Concurrent with the final rpm recorded, data indicated that the airplane's nose was raised to 19 degrees in a right turn of 28 degrees. Ground speed dropped to 48 knots while the indicated airspeed dropped to 42 knots. Latitude-longitude positions of the airplane then became constant, and indicated the airspeed dropped to 24 knots, consistent with the airplane having had entered an aerodynamic stall.

Additional data parameters revealed: 1) the airplane climbed to a maximum of about 50 feet above runway end elevation, 2) both fuel flow and fuel pressure remained relatively constant throughout the flight, 3) the No. 1 cylinder head temperature was relatively constant through the first half of the flight, but had gained an additional 5° C by the end of the flight (the other cylinders were not instrumented), 4) the No. 1 exhaust gas temperature climbed about 70°C as the takeoff progressed, then dropped 60° C after the engine rpm began to drop, 5) the No. 2 cylinder head temperature climbed about 50° C as the takeoff progressed, then dropped 20° C after the engine rpm began to drop (Nos. 3 and 4 cylinders were not instrumented for exhaust gas temperature), and 6) the electronic tachometer time was 114.65 hours.

Service Bulletin 912-065 R1

On November 13, 2014, Rotax issued mandatory service bulletin (SB) 912-065 R1 to check carburetor float buoyancy on affected part numbers. The reason for the SB was stated as, "Due to a deviation in the manufacturing process some floats could absorb more fuel thus having more weight. This leads to a loss of float buoyancy and wrong regulation of the fuel in the float chamber. Possible effects may be a rough engine during running, especially at low speeds and under circumstances loss of performance and/or fuel leakage in the area of the carburetor."

The accident engine had been sold and was not available for additional examination subsequent to the release of SB 912-065 R1; however, a representative of Rotech confirmed that the engine, as shipped from the factory, did not contain floats with a part number that was identified in the SB.

According to the spark plug manufacturer's website, "a spark plug is considered fouled when the insulator tip becomes coated with a foreign substance such fuel, oil or carbon. This makes it easier for the voltage to follow along the insulator nose, leach back into the metal shell and ground out rather than bridging the gap normally." In addition, "once a spark plug is fouled, it will not provide adequate voltage to the firing tip and that cylinder will not fire properly."

NTSB Probable Cause

The pilot’s use of excessive pitch and roll following a partial loss of engine power on takeoff, which resulted in an exceedance of the airplane’s critical angle of attack and a subsequent aerodynamic stall. Contributing to the accident was an excessively rich fuel/air mixture, which resulted in spark plug fouling and the partial loss of engine power, and the pilot’s decision to take off with known engine problems.