NTSB Factual Report

HISTORY OF FLIGHT

On September 8, 2002, at 1821 eastern daylight time, a Piper PA 32R-301T, Canadian registration C-GKLY, was destroyed during a forced landing to wooded terrain in Byram Township, New Jersey. The certificated private pilot and one passenger were fatally injured, and two passengers were seriously injured. Visual meteorological conditions prevailed for the personal flight that originated at Teterboro Airport (TEB), Teterboro, New Jersey, at 1809. A visual flight rules flight plan was filed for the flight destined for London, Ontario, Canada, and conducted under 14 CFR Part 91.

Federal Aviation Administration (FAA) transcripts of air traffic control communications revealed that the pilot declared an emergency at 1817. The airplane was in cruise flight at 3,500 feet mean sea level (msl) when the pilot reported a loss of engine oil pressure. The pilot stated that he would attempt to reach Aeroflex-Andover Airport (12N), Andover, New Jersey. The controller advised the pilot that he was 6 miles from the airport and provided a heading and runway information.

Two witnesses near the accident site reported that the airplane maneuvered across their neighborhood, just above the trees, as it descended in a level attitude. The airplane then banked slightly to its left and descended into the trees.

One witness stated that the engine was sputtering. The second witness stated that he didn't recall hearing the engine, nor did he recall hearing anything unusual.

The accident occurred during daylight hours, in the vicinity of 40 degrees, 59.8 minutes north latitude, 74 degrees, 40.5 minutes west longitude.

PILOT INFORMATION

The pilot held a Transport Canada private pilot license, issued August 21, 1997, and valid for aeroplanes, non-high performance, single engine land. A night endorsement was added to his license January 12, 1998. The pilot's logbook was recovered, but his total flight experience could not be determined. The logbook was designated Logbook #1 on July 1, 1996. The logbook entries annotated his initial flight training, and a total of 165.6 hours of flight experience up to November 28, 1997, the date of the last entry. In November 2001, the pilot reported 900 hours of flight experience to his insurance company.

The pilot's medical certificate was issued on October 30, 1998, with a requirement that he wear glasses. The pilot declared 375 hours of flight experience on that date. According to an air safety investigator with the Transportation Safety Board of Canada, the pilot's medical certificate was valid for 60 months from the date of issue in Canada, and valid for 24 months outside Canada.

AIRPLANE INFORMATION

The airplane was a 1997 Piper PA 32R-301T. A review of the airplane's "journey log" revealed the most recent annual inspection was completed February 21, 2002, at 415.6 aircraft hours. The airplane's technical logs were not recovered.

According to Lycoming records, the airplane's engine, a TIO-540-AH1A, was built with a zinc-plated crankshaft gear bolt, part number STD-2209. The engine, which was manufactured on September 27, 1997, had accrued 488.8 hours since new.

METEOROLOGICAL INFORMATION

The weather, reported at an airport 16 miles southeast of Byram Township, included clear skies, with winds from 200 degrees at 5 knots.

MEDICAL AND PATHOLOGICAL INFORMATION

An autopsy was performed on the pilot by the Sussex County Medical Examiner, Newton, New Jersey. Toxicological testing was performed at the FAA Toxicology and Accident Research Laboratory, Oklahoma City, Oklahoma.

WRECKAGE AND IMPACT INFORMATION

The wreckage was located in a residential neighborhood, about 3 miles east of Aeroflex Airport. It was examined on September 9, 2002, and all major components were accounted for at the scene. The airplane came to rest inverted, and wedged between two trees and a storage shed. Rescue workers had used logs and lumber to prop the fuselage onto its left side.

The wreckage path headed about 245 degrees magnetic, and was 110 feet in length. The initial impact point was in a tree about 80 feet above the ground. The left wing tip was at the base of the tree below the initial impact point. Trees along the wreckage path were broken off progressively closer to the ground in the direction of the main wreckage.

An examination of the wreckage revealed that the roof was crushed by impact, and that rescue personnel had cut the windshield posts. The empennage and tail section were separated aft of the cabin area, but remained attached by wires and cables. The vertical stabilizer, horizontal stabilator, and rudder were damaged, and wrapped around a tree where the airplane had come to rest.

Both wings were separated at the fuselage attach point, and the left wing had been consumed by fire. The inboard section of the right wing, with the right main landing gear attached, was connected to the right side of the fuselage by an aileron cable and a hydraulic line. The outboard section of the right wing was entangled with the left wing, and had also been consumed by fire.

Aileron cable continuity could not be established due to breaks in both aileron cables and the aileron balance cable. All cable breaks were 'broomstrawed'. Each aileron bellcrank had its respective aileron and balance cable ends still attached, and both aileron cable ends were still attached at the sprocket and chain assembly.

Rudder cable continuity was established from the cockpit to the rudder. Continuity could not be established to the stabilator due to breaks in the cables. However, the cables were traced from the flight controls, to the individual breaks, then back to the stabilator. All of those cable breaks were also 'broomstrawed'.

The interior of the airplane was damaged by fire. The throttle, propeller, and mixture levers were all in the full-forward position. All engine instruments indicated 'zero'.

Of the three propeller blades, one had no damage, one had little leading edge damage but was bent back about 45 degrees, and one had little leading edge damage but was curled back about 90 degrees. The propeller and crankshaft were subsequently rotated by hand.

The engine's top spark plugs were removed, and piston movement was noted in all cylinders as the crankshaft was rotated. However, no movement of engine accessories was noted. The rocker arm covers were removed, and no rocker arm movement occurred during crankshaft rotation.

The engine-driven fuel pump drive cover was removed, and the accessory section was rotated through the drive. The vacuum pump rotated, the magneto impulse couplings "snapped," and rocker arm movement was noted on all cylinders. However, during rotation of the accessory section and the valve train, the crankshaft did not turn.

The examination of the engine was suspended, and the engine was shipped to Textron Lycoming, Williamsport, Pennsylvania, for disassembly and further examination.

TESTS AND RESEARCH

- Engine Examination -

On September 17, 2002, the engine examination was resumed at the factory under Safety Board supervision.

The examination revealed that the zinc-plated crankshaft gear attachment bolt was fractured, and that the crankshaft no longer drove the gear. An initial review of the manufacturer's specifications revealed that a cadmium-plated bolt was indicated for the crankshaft gear application.

A Scanning Electronic Microscope (SEM) examination of the fracture surfaces on the zinc-plated bolt revealed "intergranular separation due to hydrogen-assisted cracking."

- Additional Research/Documentation -

The following is a compilation of Lycoming letters, additional reference materials, FAA briefing materials from a meeting held on June 18, 2003, at Safety Board headquarters, and a combined FAA/Lycoming factual review letter, dated February 15, 2005. In order to provide historical continuity, the source documents listed below are not in chronological order.

According to a Lycoming letter, dated April 29, 2003, IO-540 engines had previously been built with cadmium-plated bolts. However, the bolt manufacturer, Champion Bolt Corporation, requested that zinc plating be used, because it was accepted throughout the industry, it was more readily available, and cadmium plating was being phased out due to environmental concerns.

Per Textron Lycoming Supplier Discrepant Material Report (SDMR) 78347, dated July 17, 1995: "Lycoming drawings will be revised to add AMS 2402 as an alternate...per DMR 78347."

According to Champion Bolt Corporation memorandum, to Lycoming, dated August 7, 1995: "Request the substitution of AISI C4037 material in lieu of AISI C1038 material. The reason for this is material availability. C4037 is readily available and also of aircraft quality. Use of the C4037 has a positive impact on present and future production schedules and deliveries."

Per FAA/Lycoming joint letter of February 15, 2005, Lycoming response: "The requirement for the Lycoming crankshaft gear bolt material is SAE J429 as called out on the print. AMS (Aerospace Material Specification) and AISI (American Iron and Steel Institute) are references that industry suppliers would use for reference to their products. Both AMS and AISI are industry material specifications. AISI 1038 and AISI 4037 are essentially the same. Both are medium carbon steels that may be heat treated to meet the requirements of SAE J429 for a grade 8 bolt. The 4037 contains Molybdenum (Mo) and is considered a slightly better material.

According to the Lycoming letter, dated April 29, 2003: "AMS 2400 and 2402 are essentially identical electrodeposition plating specifications, excepting 2400 is for cadmium plating and 2402 is for zinc plating. Both plating specifications are primarily used to provide corrosion resistance to metal parts (including threaded fasteners). These two specifications detail very similar plating thicknesses, levels of corrosion protection, and processes for hydrogen embrittlement prevention."

Also attached to the letter, was a materials list which rated "AISI-4037" as "Grade 8" Alloy Steel, and "AISI-1038 as "Grade 5" medium carbon steel.

According to the Lycoming letter, dated April 29, 2003: Although SDMR 78347 and 78348 did not request a change or use of either cadmium- or zinc-plated bolts, "either...bolt would have been acceptable." In addition, "the metallurgical properties of the zinc-plated bolts are not considered to have an adverse effect on the crankshaft gear bolt application."

In addition: "It is general practice to consult the materials laboratory when an SDMR involves a metallurgical question." Also, "the alternate plating material for the bolts was approved pursuant to review by the Lycoming Materials Review Board (MRB) and its issuance of DMR 78347."

Per FAA/Lycoming joint letter of February 15, 2005, Lycoming response: "No hard evidence was found regarding the extent of the consultation with the materials lab. The consultations are generally verbal. The determination may also be based on previous experiences."

According to a Lycoming letter, September 23, 2002: Lycoming received 8,500 zinc-plated bolts from Pentacon (formally known as Champion Bolt Corporation and AXS Company), between November 11, 1996, and July 17, 1998. A total of 4,250 bolts were subsequently purged from Lycoming stock and scrapped.

From the FAA Briefing:

During the timeframe the zinc-plated bolts were in use, Lycoming manufactured or re-built/overhauled 1,088 engines. In addition, Lycoming reported that it shipped 2,667 zinc-plated bolts as spares.

Two crankshaft gear bolt failures had been documented in 1998, due to hydrogen embrittlement. One of the bolts had fractured at 30 hours of operation, and the other at 19 hours of operation. Both bolts had been used in O-540-F1B5 engines that had been installed on Robinson R-44 helicopters. The shipping date of the first engine was November 14, 1997, and the second engine was January 30, 1998.

Bolt recalls were issued by both Robinson and Lycoming. Robinson Service Bulletin SB-29, dated July 9, 1998, had to be complied with by August 7, 1998.

Lycoming issued Special Service Advisory Number 48-798, dated July 21, 1998, for compliance within 10 hours, or before August 7, 1998, whichever occurred first. In the background section, the advisory stated: "Lycoming has determined that hydrogen embrittlement has caused two crankshaft gear retaining bolts to fail in service...Failure of the retaining bolts results in a total loss of power without prior warning."

An SAIB was drafted to advise owners and operators of Robinson R-44 helicopters of the bolt recall by Robinson and Lycoming.

According to the Lycoming Letter April 29, 2003: "In addressing the Robinson Helicopter issue, Lycoming conducted an extensive investigation and maintained continual communications with the FAA. The localized failures in the Robinson Helicopters indicated that the root causation was due to either operational problems or the higher loads generated by helicopters. Therefore, when Lycoming issued Special Service Advisory 48-798, it was limited to the Robinson Helicopter fleet. The FAA agreed with Lycoming when it issued [AD] 99-03-05, which was also limited to the Robinson Helicopter fleet." In addition, "Lycoming and the FAA believed the failures to be limited to helicopter applications. Additionally, there had been no reported fixed wing aircraft engine failures due to the zinc-plated crankshaft gear bolts.

According to the Lycoming letter of April 29, 2003: "Lycoming and the FAA did discuss the status of the remainder of the STD 2209 zinc-plated bolts and Lycoming removed those bolts from its stock on or about July 17, 1998." The bolt removal was a "purely precautionary measure. While the environmental concerns with cadmium-plated bolts remained an issue, Lycoming had utilized the bolts for many years without experiencing any failure."

From the FAA Briefing: When the recall failed to capture all affected engines, Airworthiness Directive 99-03-05 was issued on February 3, 1999, to replace the crankshaft bolts in order to ensure capturing all affected bolts in helicopter engines worldwide.

From the FAA Briefing: No quantitative risk analysis was done. The decision to issue the AD was based on the above facts. Subsequently, there were five related engine failures on airplanes.

From the FAA Briefing: On July 29, 1999, on a PA-32R-301, a bolt on the IO-540-K1G5 engine failed at 327 hours, due to hydrogen embrittlement. The engine ship date was March 18, 1997. The FAA was notified of the event on September 8, 1999. Lycoming laboratory report 11763 was received by the FAA on October 27, 1999.

From the FAA Briefing: On March 6, 2000, on a Royal Jordanian Falcon, a bolt on the AEIO-540-L1B5 failed at 292 hours, due to hydrogen embrittlement. The engine ship date was December 26, 1996. The FAA was notified of the event on February 22, 2001. Lycoming laboratory report 12158 was received by the FAA on September 25, 2002.

From the FAA Briefing: On January 21, 2001, on a Royal Jordanian Falcon, a bolt on the AEIO-540-L1B5 engine failed at 179 hours due to hydrogen embrittlement. The engine ship date was January 30, 1998. The FAA was notified of the event on February 22, 2001. Lycoming laboratory report 12689 was received by the FAA on September 25, 2002.

According to the Lycoming letter of April 29, 2003: "The Falcon trainers were all aerobatic aircraft, which, like the Robinson Helicopters, generate a consistently high load. As a precautionary measure, Lycoming decided to replace the crankshaft gear bolts in the entire fleet, which consisted of five affected aircraft."

In addition: "Lycoming and the FAA discussed safety issues relative to the ongoing use of zinc-plated crankshaft gear bolts following the Robinson Helicopter and Falcon trainer incidents. Based on the information available at the time, the in-flight incidents involving STD-2209 zinc-plated bolt failures were not thought to extend beyond helicopter and aerobatic applications, such as those used by the Falcon trainers."

Further

NTSB Probable Cause

the improper manufacturing of the crankshaft gear bolt, which resulted in the failure of the bolt due to hydrogen embrittlement, and the subsequent loss of engine power. Also causal was the engine manufacturer's failure to remove the affected bolts from fixed wing engines during the 4-year period that transpired between the initial failures and removal of the bolts from stock for installation, until the accident flight. A factor in the accident was inadequate oversight (lack of action) by the FAA during the 4-year period in which the bolts remained in serviceable engines.