NTSB Factual Report

HISTORY OF FLIGHT

On March 12, 2013, about 1330 mountain standard time, a Franklin Glasair II, N61RA, experienced a total loss of engine power and nosed over during a forced landing to a field in Spanish Fork, Utah. The airplane was registered to the commercial pilot and operated under the provisions of 14 Code of Federal Regulations (CFR) Part 91. The pilot, the sole occupant, sustained minor injuries; the airplane sustained substantial damage. The local personal flight departed from McCarran International Airport, Las Vegas, Nevada, about 1300, with a planned destination of Spanish Fork-Springville Airport, Spanish Fork. Visual meteorological conditions prevailed, and no flight plan had been filed.

The pilot stated that while en route, the airplane was experiencing fuel flow problems, which were alleviated by him keeping the fuel system boost pump on. As he entered the vicinity of Spanish Fork, he reduced the engine power and in response the engine experienced a total loss of power. The pilot performed a forced landing in a field about 1,500 feet from the runway. During the landing roll, the airplane nosed over inverted and the wings separated from the fuselage.

AIRPLANE INFORMATION

The amateur-built experimental Galsair II, serial number 761, was completed in 1994 at which point it received its airworthiness certificate. The powerplant, a Continental Motors, Inc. (CMI) TSIO-360-GB1B, serial number (s/n) 309281, was the originally installed engine and, at the time of installation, had a total time in service (TTIS) of 1,118 hours. The engine records indicate that following manufacture in 1979, the engine was originally installed on a Mooney M20K. The engine's last major overhaul occurred in April 2008. The engine logbooks show that from March 1994 to August 2011 there was no maintenance done on the engine with a note indicating between entries that the engine was run "periodically."

According to the logbooks, in September 2012, at a Hobbs time of 85.6 hours and an engine TTIS of 1,203.6 hours, the aircraft had been inspected in accordance with a condition inspection, but was endorsed as unairworthy due to a number of discrepancies (a list of discrepancies was provided to the owner). An entry dated several weeks later, at a Hobbs time of 87.6 hours, indicated "remediation of all discrepancies." The Hobbs meter at the time of the examination read 109.3 hours, equating to an engine total time of 1,227.3 hours.

The airplane was equipped with a 24-volt Dukes Aerospace electric fuel pump, part number 4140-00-15 (s/n 8553). The system was designed with 1 electric boost-pump which was operated by the pilot's selection of either a low-boost or a high-boost switch, the latter of which was guarded. The system was designed where activation of the high-boost switch would enable electrical current to travel through the electrical system directly to the electric pump, which would operate at its maximum capacity (35 gallons per hour). If the low boost switch was selected, electric current would travel through a resistor before reaching the electric pump, which would operate the pump at a reduced capacity.

METEOROLOGICAL INFORMATION

A routine aviation weather report (METAR) generated by an Automated Surface Observation System (ASOS) at the Provo Municipal Airport, Provo, Utah (located about 5 nautical miles northwest from the accident site), indicated that about 25 minutes after the accident the wind was variable at 4 knots; temperature 55 degrees Fahrenheit; dew point 30 degrees Fahrenheit; and altimeter 30.28 inHg.

TEST AND RESEARCH

Examination

Despite National Transportation Safety Board (NTSB) investigators requests for the pilot to not attempt to troubleshoot the engine, he had done some work to the engine prior to the examination. The pilot stated that following the accident he plumed a fuel source to the inlet port of the fuel strainer and started the engine without a propeller installed. The engine operated with the electric boost pump engaged on the high setting, but when he manipulated the pump to operate on the low-setting, it would lose power. He noted that he only ran the engine for a few seconds because there was no propeller installed. The safety wire securing the fuel manifold valve's cap was cut. The owner said that he removed the manifold cap after the accident and found the screen to be clear.

An external visual examination of the engine revealed that it had remained attached to its respective mounts. Manual manipulation of the engine controls in the cockpit resulted in coinciding movement of the throttle, mixture, and propeller control levers with no slipping noted between the control levers and control shafts.

The magnetos remained secured to the engine and the ignition leads remained secured to the spark plugs. The ignition leads were removed from the upper spark plugs, which were then removed and photographed. The spark plug faces and fine-wire electrodes were all dark grey/black and were similar in this soot-covered, carbon-fouled appearance.

Investigators removed all cylinders' rocker box covers and noted a light oil film on the rocker arms and valve assemblies. The cylinders' combustion chambers were examined through the upper spark plug holes utilizing a lighted borescope. The combustion chambers remained mechanically undamaged, and there was no evidence of foreign object ingestion or detonation. There was no evidence of valve to piston face contact observed. The gas path and combustion signatures observed at the spark plugs, combustion chambers, and exhaust system components displayed coloration that the CMI representative said was consistent with a rich operation. There was no oil residue observed in the exhaust system gas path.

Investigators achieved manual rotation of the crankshaft by rotation of the crankshaft propeller flange. Thumb compression was established in all cylinders. Valve train continuity was observed, with approximately equal lift action at each rocker assembly. The accessory gears were undamaged and had a light residue of oil. The top sparkplugs were reattached to their respective ignition lead and the engine's crankshaft was again rotated and spark was observed on all the upper spark plugs.

The inlet filter remained intact and in place in its bracket. There was some organic debris adhering to the outside of the filter, but it was considered accident-related and not of an amount that would disrupt airflow into the induction system. The over boost relief valve, throttle body, intake tubes, and upper deck reference lines all remained intact and attached to their respective components. There was no evidence of an induction leak.

The pilot's lap belt fractured in flight. The complete materials report is appended to this accident in the public docket. In pertinent part, the fracture surface was consistent with that of an overload failure due to impact forces.

Fuel System

There was no evidence of a fuel leak on any of the engine's fuel system components.

Investigators removed the fuel manifold and upon disassembly observed that internal screen was clear, and the diaphragm, plunger, and spring were intact and in place. Several remnants of unidentified debris were found in the fuel manifold valve housing under the screen. Removal of the fuel injector lines and nozzles from their respective cylinders revealed that they were clear of contaminants.

In an attempt to functionally test the electric fuel pump, investigators supplied power to the electrical system. The electric boost pump would operate when the high-boost switch was selected, although would fail to operate when the low-boost switch was selected. Examination of the system by the owner revealed that the resistor had an open circuit, which prevented electrical current from passing through to the pump. When the resistor was bypassed, the electric boost pump ran with the low-boost switch selected.

The engine-driven fuel pump was removed and the drive coupling remained intact. Manual manipulation of the drive coupling while installed in the fuel pump drive shaft revealed that the drive shaft would rotate with a slight binding felt about 15 degrees of its complete 360 degrees rotation. The complete examination of the fuel pump and throttle body were conducted at the CMI facilities, which produced a report that is appended to this accident in the public docket. In pertinent part, the fuel pump did not meet the specified production pressures for a given fuel flow; however, according to CMI representatives that could be attributed to adjustments that were made to the pump's adjustable orifice and low-pressure relief valve during fuel system setup on the accident engine. The pump was capable of producing operational pressures at the parameters tested. The fuel injector nozzles, lines, and manifold valve assembly were flow tested on the production test stand and no leakage was noted. The fuel manifold valve failed to meet the new production test limits as specified in the engineering documents; however, the fuel manifold valve was capable of metering fuel at operational pressures. The fuel nozzles were unrestricted and exhibited normal operating signatures. The fuel lines were intact and undamaged.

ADDITIONAL INFORMATION

Following the accident, the pilot had the following electric fuel pumps bench tested: the accident pump and another Duke, Inc. pump that he had installed on a Mooney 231, which was equipped with a similar engine. Both of them passed the bench test and no anomalies were noted. After installing both pumps back on the same engine (one after the other), it was noted that fuel was still inhibited from reaching the engine-driven fuel pump or developing full pressure.

There were no blockages in lines, filters, gascolator, and fuel selector. Disconnecting the output hose from the pump and attaching a clear vinyl hose revealed that a steady stream of air bubbles was produced from the electric pump. Upon further testing, he found that the pump introduced air into the fuel line when not pressurizing the output, or when the pump was turned off but under suction, as would happen with the engine-driven pump trying to draw fuel without the use of the electric boost pump. After overhauling the pumps with a new shaft bearing and seal, he has had no problems with the fuel system or engine in indication the shaft and seals were worn. He opined that the air leak originated from the shaft on both pumps and that the engine-driven pump could not produce enough suction with introduction of air in the system.

Both Mooney and Duke Aerospace reported that there have been no user complaints or customer service calls about such a problem occurring.

NTSB Probable Cause

A loss of engine power due to fuel starvation, which likely resulted from air being introduced into the fuel system through the worn shaft of an electric fuel boost pump.