MIA05LA093 · Piper PA-28-161

The pilot/mechanic arrived at the Seminole Lake Gliderport to repair the engine following an uneventful precautionary landing the day before due to a bent exhaust pushrod of the No. 2 cylinder. He replaced the damaged No. 2 cylinder exhaust pushrod, pushrod housing, pushrod seals and lock tab. He reported visually inspecting the valve springs which did not appear to be damaged, but did not inspect the valve, nor did he remove the cylinder. Following the repairs, the engine was started and a full static run-up was performed with no discrepancies noted. The flight departed for Kissimmee Gateway Airport and shortly after takeoff, the engine started running rough though it was still producing power. He elected to return to the Seminole Lake Gliderport and after recognizing he was going to overshoot the runway with the engine not producing full power, elected to perform a 360-degree turn back onto final approach. During the turn, the aircraft impacted trees and came to rest in a swamp south of the south edge of runway 18. A pilot-rated witness observed the airplane on final approach and noted that as it descended to treetop level, or the height of the buildings, the airplane began to abruptly pitch up and down and was abruptly yawing. It appeared to the pilot-rated witness that some of the motions of the airplane did not coincide with the wind conditions that he observed. Postaccident examination of the engine by an FAA inspector revealed the replaced pushrod was damaged on both ends, with no obvious indication of a stuck valve. Additionally, the rotator type exhaust valve stem cap was missing from under the exhaust rocker arm of the No.2 cylinder. The exhaust valve and guide were within new limits in compliance with the Lycoming overhaul manual. There was no evidence of carbon inside the guide; however, there was swirl marks from reaming of the guide. Review of Lycoming Service Bulletin (SB) No. 388C revealed inspection procedures to be carried out at 400 hours or earlier, when valve sticking is suspected. The SB indicates that failure to comply with this provision could result in sticking valves, or broken exhaust valves which could result in engine failure. A section of the SB outlines the inspection procedure that uses a "GO/NO-GO" gage, and recommends moving the exhaust valve completely out of the guide to avoid interference when using the GO/NO-GO gage to check the guide for wear or carbon build up. The mechanic did not report using the GO/NO-GO gage. Additionally, the mechanic did not report using a local manufactured tool.

Verbatim NTSB analysis and probable cause from the NTSB dataset