NTSB Factual Report

On February 19, 2015, about 1440 eastern standard time, a Cirrus SR22 airplane, N358CD, was substantially damaged during a runway overrun while attempting to depart from Bluegrass Airport (LEX), Lexington, Kentucky. The private pilot and both passengers were not injured. Visual meteorological conditions prevailed and an instrument flight rules (IFR) flight plan was filed for the cross-country flight that was destined for Oakland County International Airport (PTK), Pontiac, Michigan. The personal flight was conducted under the provisions of Title 14 Code of Federal Regulations Part 91 and was originating at the time of the accident.

According to the pilot, he did not observe any abnormalities with the engine during any of the five individual flights that he completed in the accident airplane about one week prior to the accident. The pilot reported that both the preflight inspection and subsequent engine run-up did not present any anomalies on the day of the accident. The pilot then taxied to runway 22 and began a takeoff roll. The airplane's initial climb appeared normal until it reached approximately 200 feet above ground level (agl). The engine "backfired" several times, which was immediately followed by a partial loss of power. The pilot elected to discontinue the flight, retarded the throttle to the idle position, and initiated a descent to land on the remaining runway. The pilot stated that he had "too much energy" to stop the airplane before it overran the end of the runway and collided with the precision approach path indicator lights and a snowbank. According to a police report, the pilot stated that the airplane touched down near the approach end of runway 04.

A Federal Aviation Administration (FAA) inspector interviewed a witness who was in his office, which was located about midfield on runway 22/4, at the time of the accident. According to the witness's recount, he did not observe any anomalies as the airplane began its climbout. However, once the airplane was "abeam his office window" and approximately 200 feet agl, the witness heard the engine surge, which was followed by a reduction in power and multiple loud "pop" sounds. The airplane then entered a nose-low attitude and began to descend. The witness observed the airplane begin a landing flare from approximately 30 feet agl. During the airplane's subsequent touchdown attempt, it bounced three times and then overran the runway.

The 1454 recorded weather observation at LEX included wind from 280 degrees at 12 knots, gusting to 15 knots, 10 statute miles visibility, overcast clouds at 3,400 feet, temperature -14 degrees C, dew point -23 degrees C; barometric altimeter 30.31 inches of mercury.

The four-seat, low wing, fixed-gear airplane was manufactured in 2004 and powered by a Continental Motors IO-550-N27, 310-horsepower reciprocating engine. According to the maintenance records, the airplane's most recent annual inspection was performed on December 5, 2014, at a total airframe time of 3,700 flight hours, 14 flight hours before the accident. At the time of the inspection, the engine had accumulated 1,598 total flight hours since its last overhaul, which took place on December 18, 2009 at 2,116 hours, total time in service. A 500-hour magneto inspection was completed at the time of the annual inspection.

According to the engine logbook, the ignition harness was replaced with a factory new unit on August 24, 2009, approximately 100 hours before the engine was overhauled. The logbook entry that pertained to the airplane's most recent inspection stated that the spark plugs were "cleaned, gapped, and inspected" and the engine was inspected in accordance with the manufacturer's maintenance manual. The ignition harness inspection and cleaning requirements were included in a service bulletin, but not in the manufacturer's maintenance manual. The most recent inspection logbook entry did not reference the service bulletin nor did it indicate that the ignition harness spark plugs terminals had been cleaned.

The airplane was equipped with an Avidyne multi-function display (MFD) that was capable of recording airplane and engine performance data to a compact flash card. The compact flash card was removed and successfully downloaded. The data contained recorded engine parameter data and GPS coordinates for the accident flight. The data were recorded at a rate of once every 6 seconds, and did not include altitude or airspeed; however, the airspeed was computed using time and the airplane's GPS-derived location. According to the data, the airplane began a takeoff roll at 1437:12 at which point the engine rpm increased from 1,470 rpm to 2,460 rpm, on its rise to takeoff power. In the 18 seconds that followed, the engine maintained 2,400 – 2,700 rpm, which corresponded to a fuel flow of about 30 gallons per hour (gph).

After the airplane passed the first third of the runway, the engine rpm, fuel flow, and cylinder exhaust gas temperatures (EGT) began to decline simultaneously; however, a precise rate of decline could not be captured due to the rate at which the data was recorded. The fuel flow decreased to 3 gallons per hour in the 12 seconds that followed the power reduction. The engine rpm and cylinder EGTs continued to decline as the airplane reached the departure end of the runway. At 1438:12 the engine rpm leveled out at approximately 450 rpm for about 12 seconds, when the airplane came to rest. The rpm then decreased to 0 rpm and the fuel flow was reduced to 0 gph almost simultaneously.

Postaccident examination of the airplane revealed that the spark plugs and ignition harness functioned normally when field tested, and electrical continuity was established through the magneto switch and primary leads. The magnetos had been timed to approximately 22 degrees below top dead center (BTDC), consistent with the manufacturer's specification. Both magnetos were subsequently field tested, but only the right magneto produced a spark at the ignition leads.

A set of new magnetos, furnished by the manufacturer, were installed and timed to 22 degrees BTDC and a set of test leads were attached to the disconnected primary leads to bypass the magneto switch. The ignition harness was not replaced. A subsequent engine test run revealed that the engine ran smoothly on both magnetos. When the right magneto was selected the engine lost approximately 20 rpms, but continued to run smoothly. Once the left magneto test lead was selected, the engine lost power and began to backfire.

All 6 fuel injectors were cleaned after an inspection showed that some of the injectors were contaminated and restricted. The injectors were reinstalled and another engine run was attempted; however, the engine still lost power and backfired when the left magneto was selected. The airplane was secured until the engine could be re-run with new spark plugs and a new ignition harness.

A follow-up engine run was completed with a new set of spark plugs installed and a subsequent engine-run revealed that the engine lost approximately 200 rpm when the magneto switch was moved from BOTH to LEFT, but the engine did not backfire as it did during previous tests. After the ignition harness was replaced, the engine dropped only 20 rpm when the left magneto was selected and did not backfire.

The ignition harness and magnetos were submitted to the NTSB Materials Laboratory for further examination. An examination of the ignition harness revealed the presence of radial carbon tracks on the sleeves of 8 out of 12 terminals on the harness. The sleeves exhibited pitting, discoloration and flat spots consistent with wear contact. Black deposits were observed on the sleeve surfaces, including the areas that sealed against the spark plug insulator. Each terminal spring was covered in black deposits and several of the springs and sleeves were bent. Multiple leads displayed wear damage, and in one case the damage extended to the underlying metal braid.

Examination of the right magneto revealed that 11 teeth were fractured and two teeth were cracked. The left magneto exhibited 9 fractured teeth and one partially fractured tooth. A set of teeth from the right distributor gear were deliberately fractured under impact loading conditions and the resulting impact signatures were consistent with those observed in the teeth that had been previously fractured in both magnetos. Laboratory testing showed that each distributor gear had a Fourier-transform infrared spectrum consistent with the specific material prescribed by the magneto manufacturer.

Champion Aerospace Aviation Service Manual, AV6-R, dated August 2014, stated that if the terminal well in the spark plug became dirty with moisture or other foreign material, current could track through the dirty terminal well to ground on the shell, which could result in an erratic misfire of the spark plug. This condition was known as connector well flashover. The service manual further stated that spark plugs with dirty terminal wells should be replaced with serviceable units.

The Australian Civil Aviation Safety Authority Airworthiness Bulletin (AWB) 17-005, Issue 3, dated October 2014, listed a number of potential causes for nylon distributor gear failures, including propeller strikes, kick back during start-up events, and any other event that can cause shock on the gear train driving the distributor gear.

According to Service Bulletin (SB) SB-643B, published by Continental Motors, Inc. on April 6, 2005, all ignition harness outlet plates, covers, or cap assemblies should be cleaned and inspected in concurrence with the 500 hour magneto inspection. Any damaged parts, including those that were broken, brittle, cracked or burned, must be replaced. The SB required that all ignition harness spark plug terminals be removed, cleaned, and inspected during each 100 hour, annual inspection, or progressive maintenance inspection.

NTSB Probable Cause

Maintenance personnel's failure to properly inspect and clean the engine ignition harness spark plug terminals, which resulted in a partial loss of engine power during an attempted takeoff.