SEA94LA235

The private pilot was forced to land the airplane on unsuitable terrain because of a loss of engine power. During the forced landing, the airplane impacted a tree stump, nosed over, and was substantially damaged. An examination of the engine revealed that the rocker arm for the no. 1 cylinder exhaust valve had failed. An examination of the rocker arm revealed that the arm had failed due to fatigue, and that the fatigue cracking was initiated at rough tool marks produced during the chamfering of the arm's oil hole. The chamfering was performed by the engine manufacturer during remanufacture of the engine.

On September 7, 1994, about 1300 hours Pacific daylight time, N61838, a Cessna 172M, operated by the owner/pilot, was substantially damaged during a forced landing in Coolin, Idaho. The forced landing was precipitated by a loss of engine power while climbing. The private pilot received minor injuries and her two passengers received no injuries. Visual meteorological conditions prevailed and no flight plan had been filed. The local flight departed from Cavanaugh Bay Airport, Idaho, and was conducted under 14 CFR 91.

According to the pilot, the flight began with a "good" pre- flight inspection and engine run-up. After a "normal" takeoff, the engine "started running rough" as the airplane climbed through 3,000 feet mean sea level above a lake. The pilot attempted to "smooth it out," but to no avail. The pilot then heard a "loud bang," and the engine subsequently had "no power."

The pilot turned back toward the shore and elected to perform a forced landing on a road. She noticed vehicular traffic on the road while on final approach, so she veered away from the road and performed a forced landing in a clear cut area. The airplane collided with a tree stump during the final approach, nosed over, and was substantially damaged.

A preliminary examination of the wreckage by an FAA Aviation Safety Inspector from Seattle, Washington, revealed that the rocker arm for the No. 1 cylinder exhaust valve was fractured. All 8 rocker arms are embossed with a part number "69443-D," located on the lefthand side of the rocker "nose," viewed in its normally installed orientation. Also, vibropeened on the lefthand side of the shaft boss is "LW-18790-85," and on the righthand side of that boss, also vibropeened, is "Rev. B."

The failed rocker arm was removed from the engine and sent to the Safety Board's Metallurgical Laboratory and examined. According to the Metallurgist's Factual Report (attached):

The rocker arm was fractured at two locations on opposite sides of the bearing bore.... [One of the fractures] intersected an oil drip hole drilled radially into the bearing bore.... The fracture regions adjacent to the oil hole on the fracture face... were smooth and silky and contained crack arrest positions indicative of fatigue crack propagation.... [An] examination with the aid of a bench binocular microscope revealed that a chamfering tool penetrated deep into the hole, enlarging the hole diameter through approximately 3/4 of its length. The examination further revealed that the reworked area of the hole was not concentric with the original hole, with much more material removed from the area....

The other seven rocker arms were subsequently removed from the engine and sent to the FAA Manufacturing Inspection District Office in New Cumberland, Pennsylvania, under the supervision of the Safety Board. All seven were intact on the engine prior to removal. They were then delivered to Lycoming and underwent an examination at the Lycoming Materials Laboratory under the supervision of the FAA. The oil hole area of each rocker arm was examined with a bench microscope for each rocker arms. Six of the seven rocker arms exhibited no scoring or light scoring. The No. 4 cylinder intake rocker arm exhibited a circular tool mark (groove) mid-way down the chamfer. Additional examination and measurements of the No. 4 rocker arm verified that Lycoming's material conformance specifications were met.

None of the seven rocker arms exhibited the same fracture patterns, cracks, or tool marks as the failed rocker arm.

An examination of the airplane's engine maintenance records (excerpts attached) indicated that the engine, a Lycoming O-320- E2D, underwent a complete overhaul by Textron Lycoming, Inc, Williamsport, Pennsylvania, and was "zero" timed on June 6, 1988. The engine had accumulated 292 hours of recorded tachometer time from the time it was overhauled to the time of the accident. The records did not indicate any maintenance which would have involved repair or replacement of the No. 1 rocker arm subsequent to the major overhaul.

Safety Board research revealed that the sharp outer edge in the oil drip hole of the Lycoming rocker arm part number 18790 was cited as a fatigue crack initiation area in several previous reports of rocker arm failures. As a result of these reports, on February 16, 1987, Lycoming published Service Bulletin No. 477A. The bulletin provides the instructions for oil drip hole inspection and states that if the outside edge of the hole is found to be sharp, "... it must deburred to the approximate .013 radius with a burr gun before reassembling of an acceptable rocker arm." Such radius, if correctly performed, would decrease the effect of a stress riser at the edge of the hole. The rocker arms with the reworked oil holes are identified by a letter "B." FAA Airworthiness Directive No. 87-10-06 made the requirements of Lycoming Service Bulletin 477A mandatory.

According to Lycoming, the accident engine was "built to remanufactured specifications and sold as overhaul" by Lycoming in June of 1988. During this process, rocker arms are normally inspected for verification of new engine specifications, and are to be reworked or replaced if necessary by Lycoming prior to the completion of the engine remanufacture. Lycoming Service Bulletin 477A was in effect at the time of the accident engine's remanufacture.