NTSB Factual Report

On June 3, 2004, at 1215 central daylight time, a Beech M35, N693V, collided with a fence and a guy wire during a forced landing after the number 2 cylinder departed the engine in Strafford, Missouri. The pilot and passenger were not injured. The airplane was substantially damaged. The 14 CFR Part 91 flight was operating in visual meteorological conditions and a visual flight rules flight plan was filed. The flight originated in Tell City, Indiana, at 1005, with a planned destination of Tulsa, Oklahoma.

The pilot reported he was flying at 8,500 feet mean sea level when the number 2 cylinder departed the engine resulting a total loss of engine power. He reported that with no airports within gliding distance, he elected to land in a farmer's field. The airplane was damaged when it contacted a fence and a guy wire during the landing.

The number 2 cylinder and the left magneto were not located after the accident.

The Continental IO-470-C, s/n 242060-R, engine was torn down under the supervision of an inspector from the Federal Aviation Administration Kansas City, Missouri, Flight Standards District Office. Half of the crankcase, two fractured cylinder through-bolts, two cylinder through-bolt nuts with pieces of bolts inside, four cylinder hold down nuts with pieces of studs inside, a connecting rod cap with the connecting rod cap bolt, and a connecting rod nut were submitted to the NTSB for a metallurgical examination.

The following are excerpts from the metallurgical examination report.

"For reference in this report, the studs and through-bolts were numbered clockwise starting with the upper forward stud... . Stud number 1 was contained in the fractured piece of the crankcase.... . The crankcase fracture intersected the hole for stud 2, and the inboard piece of stud 2 was missing. The outboard piece of stud 3 and the nut for stud 3 also were missing. Three of the four stud pieces and nuts ... mated to studs 1, 4, and 6. One stud and nut did not mate to any of the stud fractures and was therefore identified as the mating side of stud 2. The two through-bolt pieces with nuts mated to the other two submitted through-bolt pieces at positions 7 and 8."

"Each stud and through-bolt fracture surface had a relatively smooth flat region oriented approximately perpendicular to the longitudinal axis, features consistent with fatigue. ... Crack arrest lines were observed emanating from multiple origins at one side of the fracture surfaces for studs 2 to 5. The fracture surfaces for studs 1 and 6 were slightly rougher and did not have distinct crack arrest lines, but ratchet marks were visible indicating fatigue from multiple origins. Although the fracture surface for stud 1 ... was nearly obliterated by post-fracture damage, the mating fracture surface showed the described fatigue features. In each case where orientation could be determined, the fatigue emanated from origins at the cylinder side of the stud or through-bolt."

"Of all the stud and through-bolt fractures, stud 3 had the smallest overstress region. The fatigue region covered more than 90 percent of the fracture surface."

"A lip was observed on the crankcase surface at a location corresponding to the forward edge of the cylinder 2 barrel flange, adjacent to the through-bolt positions. At the aft side of the cylinder 2 area of the crankcase, linear marks aligned with the axis of the cylinder were observed at the aft side of the hole for cylinder 2 corresponding to sliding contact with the cylinder barrel skirt. Also, fretting was observed in the barrel flange contact area around through-bolt number 7."

Studs 3 and 4 were sectioned longitudinally. "Cracks and/or rolling laps were observed at the thread roots of both sections. On stud 3, the cracks generally followed the material flow lines and were limited to within 0.001 inch of the surface."

"On stud 4, cracks, many of them branching cracks, ... extended radially for some distance then turned, following the flow lines in the material. Many of the thread roots had sharp radii... ."

The hardness of studs 3 and 4 was measured. The hardness of stud 3 was 33 HRC, and the hardness of stud 4 was 31 HRC. The manufacturer specified harness range for the studs was 32 - 38 HRC.

Two studs from cylinder 4 were also removed and sectioned. One of these studs had thread roots with sharp radii and many cracks at the thread roots similar to those on one of the studs from cylinder 2. The other stud from cylinder 4 had a small rolling lap in one thread. No sharp thread root radii were noted on this stud.

"The connecting rod cap bolt from cylinder 2 ... was slightly bent. ... The thread peaks on the connecting rod bolt ... were deformed and worn, visually appearing flattened and shiny. The threads closest to the end of the bolt, the thread peaks were mostly deformed toward the end of the bolt. At the other end of the threaded region closest to the head, the thread peaks were deformed both toward the end and toward the head."

The engine was overhauled on July 6, 2000, at a tachometer time of 2440.33 hours. The engine accumulated 194.03 hours since the overhaul.

Parties to the investigaiton were the Federal Aviation Administration and Teledyne Continental Motors.

NTSB Probable Cause

The improper installation of the cylinder during the engine overhaul which resulted in the fatigue failure of the cylinder studs and through-bolts and the subsequent separation of the cylinder. Factors associated with the accident were the fence and the guy wire which the airplane contacted during the forced landing.