History of the Flight
The aircraft, a Piper PA-28-161 Cherokee Warrior II registered G-MAND, was on its third flight of the day, an instructional trip for the revalidation of the second pilot's Private Pilot's Licence. The technical log showed fuel contents around 100 litres; fuel water drain checks had been completed before the first flight. The instructor reported that the second pilot's pre-flight inspection and engine power checks revealed no anomalies. The second pilot executed a normal takeoff on Runway 28 at Wolverhampton Airfield, with wind from 280° at 10 kt gusting to 20 kt. Runway 10/28 is 1090 metres long and 23 metres wide with an asphalt surface.
Shortly after takeoff, both pilots checked the engine instruments and found them normal. At approximately 700 feet agl, the engine began to run roughly, and the instructor took control. Almost immediately, engine speed reduced to around 1800 rpm from full power. The instructor levelled off and decided to fly a low level circuit and land back, but the engine then appeared to lose all power. Unable to locate a suitable forced landing area ahead, at around 600 feet agl he turned back towards the airfield to land on Runway 10, informing the Tower of the situation and his intentions.
On approach, the instructor sideslipped to lose height but, with the strong tailwind, could not touchdown until about halfway along the runway. He could not stop in the remaining runway length, and the aircraft ran off the end, through the airfield boundary hedge, across a lane, and came to rest in an adjacent field. All three landing gear legs failed during the overrun, and the left wing detached, coming to rest inverted on top of the right wing. There was no fire. The pilots, each wearing a harness with lap strap and inertia reel diagonal upper torso restraint, were uninjured except for minor whiplash effects and evacuated through the right hand main door.
Engine Examination
A strip examination of the engine and testing of magnetos, carburettor, and engine-driven fuel pump were carried out by the AAIB at an engine overhaul facility. Accessory tests were normal, but two anomalies were found:
Valve Operating Mechanism: Severe distress to the operating mechanism for Nos 3 and 4 inlet valves was evident. The camshaft lobe was severely worn, with height reduced to 0.309 inch (compared to 0.352 inch for others), and both follower bodies had extensive surface pitting. Metallurgical examination found that the Nos 3/4 inlet valve lobe had suffered overheating and consequent softening, leading to rolling contact fatigue damage in the followers. The most likely cause of degradation was insufficient lubrication.
No 1 Cylinder: Extensive cracking of the No 1 cylinder assembly (Part No 65099-REV-1) was evident, comprising two circumferential cracks in the aluminium cylinder head wall that linked the upper and lower spark plug holes via forward and aft sides. The cracks extended through the whole wall thickness, almost completely separating the head from the cylinder, leaving a 0.8 x 0.4 inch unfailed portion. Damage to the exhaust system from ground contact indicated abnormal loads had been applied during the overrun. However, heavy sooting on a portion of the fracture between the lower spark plug hole and inlet valve port showed that part of the crack had been present while the engine was operating. This crack (approximately 30% of total length, 3 inches long) and a second crack between the lower spark plug hole and exhaust valve port also had sooted areas indicating pre-existing cracks. Metallurgical examination found no fatigue progression except in two small regions near the unfailed area; other portions had a rough, dimpled morphology from tensile overload. Material hardness and microstructure were normal.
Aircraft History
At the time of the accident, the aircraft had 6,237 operating hours since manufacture. A factory newly rebuilt engine had been installed in 1997; since then the aircraft had been used for flying training with the same operator and maintainer. In late 1999, four new cylinder assemblies were fitted, 693 hours before the accident. The engine had accumulated 1,865 hours since the 1997 rebuild. Documentation showed regular operation and servicing per the maintenance schedule. A compression check was performed every 50 hours (more frequent than the 150-hour schedule), and internal visual checks for cracking, especially near the lower spark plug bore, were done at 50-hour checks. The last 50-hour check was 45 hours before the accident; no abnormalities were recorded.
Background on Valve Wear
The overhaul agency noted that camshaft and follower distress to this extent was common on engines received for overhaul, especially those near overhaul time, estimating about 70% exhibited similar damage, but associated significant power loss had not been reported. The engine manufacturer estimated a lower rate (5% to 10%) and noted that appreciable wear could go unnoticed due to hydraulic tappet compensation. Severe wear could be detected by metallic debris on the oil filter. Progression was rapid, becoming severe within one or two 50-hour oil change periods. The manufacturer had not found material deviations; the problem was believed due to inadequate lubrication from infrequent operation, insufficient oil changes, or repetitive accelerated starts. A Mandatory Service Bulletin (Textron Lycoming No 446D) specified the addition of oil additive LW-16702 at each oil change for certain engine types, unless the oil already contained an equivalent additive.
Although the valve operating mechanism wear was not considered a factor in this accident, a safety recommendation was made concerning reduction of such wear through mandatory use of oil additives.