N78GP

On May 9, 2024, about 1052 mountain standard time, an experimental amateur-built Robert Sterling 4P, N78GP was substantially damaged when it was involved in an accident near Lake Havasu City, Arizona. The pilot sustained minor injuries. The airplane was operated as a Title 14 Code of Federal Regulations Part 91 personal flight.

The float-equipped airplane had just undergone a condition inspection by the pilot, which was completed on the morning of the accident. The airplane had not flown in a long period of time, and as part of the inspection the pilot changed the engine oil and replaced the fuel. The pilot stated that the engine performed normally during multiple postmaintenance run-ups. With no anomalies observed, he planned to depart from runway 32 and remain in the traffic pattern so he could assess the airplane’s operation and performance.

The pilot stated that the takeoff was uneventful and the climb out was normal. After reaching about 500 ft above ground level, he began the crosswind turn and the engine lost power. He started to maneuver the airplane for a landing on runway 14 while verifying the fuel boost pump was on, checking the throttle and fuel mixture control positions, and switching from the right to left fuel tank. With no change in engine power, he recognized that a return to the airport was not possible. With a set of power lines and a busy highway ahead, he decided to maneuver the airplane to a flat open area in front of the power lines. The airplane landed, bounced on its floats with the gear still extended, and then tumbled after it struck the ground again. The pilot’s door opened during the impact, and he was able to egress after unbuckling his seatbelt.

The airplane sustained substantial damage to both wings and the entire empennage.

The airplane was originally certificated as a 1956 Piper PA-22 Tri-Pacer, but had been extensively modified in 1997 to the extent that it was reclassified to the experimental amateur-built aircraft category. Some of the modifications included an increase in fuselage and wing length along with replacement of the original four-cylinder engine with a six-cylinder Lycoming O-540 engine that had been upgraded with the addition of an Airflow Performance fuel injection system. In 2012, amphibious floats were added.

The airplane was equipped with a JPI 900 engine data monitor that was configured to record engine parameters at 6-second intervals. Maintenance records, along with the engine monitor data, appeared to indicate that although the engine was run after a condition inspection in October 2022, the airplane was last flown in 2018.

According to the engine monitor data, seven engine runs were performed in the hours leading up to the accident, with the last completed about 15 minutes before the accident flight.

During the accident flight, the monitor started recording after the engine had been started. The engine speed was 1,345 rpm, manifold pressure 15.7 inHg, and fuel flow 4.6 gph. The cylinder head temperatures (CHT) all averaged about 290°F, while the exhaust gas temperatures (EGT) were about 1,175 °F. For the next three minutes, a series of changes in manifold pressure, fuel flow, and engine speeds were consistent with the pilot performing an engine runup. One minute later, the engine speed climbed to 2,775 rpm, manifold pressure reached 28.2 inHg, and the fuel flow reached 25.3 gph; EGT and CHT climbed to 1,300°F and 360°F, respectively. For the next 1 minute and 40 seconds the values remained stable, consistent with takeoff.

For the following 12 seconds, with engine speed and manifold pressure unchanged, fuel flow dropped to 13.6 gph while EGT increased to 1,450°F. Six seconds later, the engine speed had dropped to 1,600 rpm, and fuel flow was now at 6.1 gph, while manifold pressure remained unchanged. The next recorded data point showed that the EGT had now dropped to 890°F, but manifold pressure still remained unchanged. Engine speed then began to rise to 2,100 rpm, with an accompanying rise in fuel flow and EGT. For the last three data points, the manifold pressure began to drop, (consistent with the pilot pulling back the throttle control), reaching 11.5 inHg, accompanied by reductions in engine speed, fuel flow, and both EGT and CHT.

Postaccident examination did not reveal any evidence of catastrophic engine failure. All four of the airplane’s fuel tanks contained fuel, which was light blue in color and free of water when tested with water-detecting paste. No failures were observed to the fuel delivery system that would have precluded normal operation, and residual quantities of fuel were present in all the lines up to the inlet of the fuel injection servo.

The servo and fuel flow divider were both examined and tested at the Airflow Performance (AP) facilities, under the oversight of the NTSB. AP records indicated that the servo was manufactured in 1997 and the flow divider in 2008. The original factory lock wire seals were still in place on the servo, and there was nothing in the airplane maintenance records to indicate either unit had ever been serviced or overhauled since installation.

The AP Installation and Overhaul Manual indicated that the servo and flow divider should be overhauled every 12 and 8 years, respectively. Nevertheless, both units were tested and met the manufacturer’s performance specifications, and disassembly did not reveal any anomalies that would have precluded normal operation.

The installation manual stated, “Gascolators [fuel strainer] are not required for proper operation of the fuel injection system but if used, should not be installed on the engine side of the firewall.” The manual also stated that, in order to avoid fuel vapor or hot operation issues, if a strainer is installed it should be enclosed within a closed box that is configured to receive blast air.

Examination of the airframe revealed that a strainer was installed but was unboxed and mounted on the engine side of the firewall in the lower left corner, about 8 inches aft of the shared exhaust pipe stack for the Nos. 2, 4, and 6 cylinders.

The manual also stated that all installations using an engine-driven fuel pump should incorporate a purge valve as part of the flow divider assembly. The purge valve is used during all start operations, but according to AP, it is particularly effective during hot start operations, when it is used to flush hot fuel and vapor from the system and cool the fuel injection components.

AP supplied a purge valve assembly, and recommended that for it to be effective, it should be installed either directly onto the flow divider, or as close as possible to it. Examination of the airframe revealed that the airplane had been equipped with a purge valve that was comprised of a brass household-grade gas valve mounted under the instrument panel, where it was routed through the engine compartment and to the right fuel tank. An undated entry in the maintenance logbooks gave instructions for use of the valve, stating that it was to be used to purge the fuel system of air during hot start operations, and when used, it bypasses fuel into the right fuel tank.