One at a time

By David Jack Kenny

When a loss of engine power precipitates an aircraft accident, it can be difficult for investigators to determine exactly what went wrong. In fact, nearly 30 percent of the time the cause of the engine failure is never satisfactorily explained. Sometimes the resulting damage is so extensive that the pre-impact condition of critical components can’t be determined. In other cases, a largely undamaged engine runs just fine when tested after the fact. Fuel starvation and carburetor ice no doubt account for some of these—but don’t explain the stoppages of fuel-injected engines with ample quantities in all tanks.

There’s also a third category: engines that suffered from enough different discrepancies to make it impossible to reconstruct which of those faults or interactions between them actually made the engines quit.

On the morning of July 23, 2014, the 67-year-old owner of an amateur-built Great Lakes 2T-1A biplane took a close friend’s 12-year-old son for a sightseeing flight. They departed from Ferguson Airport just north of Pensacola Naval Air Station in Florida before 10 a.m. and flew for about 15 minutes. The weather was pleasant for the Gulf Coast in July, with light southeast winds, a temperature of about 80 degrees Fahrenheit, and scattered clouds at 2,000 feet. The pilot did not attempt any aerobatics, but according to the passenger, the pilot did “set the engine to idle or off so that they could glide for a while,” an experience the boy enjoyed.

At 10:10 a.m. the pilot took off with the boy’s 15-year-old sister. This flight also was intended to last about 15 minutes, but the airplane did not return. After tracking the pilot’s cell phone, an Alabama State Police helicopter found the wreckage about 4 p.m. in a swampy area just across Perdido Bay. Although both seats were fitted with five-point harnesses, neither occupant survived the crash.

First responders noted a strong smell of fuel on the scene; the fuel tank in the upper wing had been breached. Damage to the trees suggested the biplane had struck them in a level attitude on a westbound heading. Flight-control systems were intact aside from impact damage and cables cut during recovery of the wreckage. The throttle and mixture controls were full forward and the carburetor heat was on; the primer was locked, and the magneto switch was in the “both” position. The carburetor was intact and functional.

However, two spark plugs—the upper from cylinder No. 1 and the lower from No. 4—were fouled, drawing attention to the magnetos. Sure enough, the right—originally manufactured by Bendix in 1973—produced sparks on all leads, while the left, which dated to 1968, did not. Although the wreckage had come to rest in about two feet of water, investigators concluded, “There was no evidence that the magnetos were submerged.”

The logbooks indicated that the engine had been installed in the airframe on Feb. 11, 2013, after major service that included the installation of four new Superior Air Parts cylinder assemblies, an oil filter adapter, an overhauled carburetor, and “two [overhauled] yellow-tag Bendix mags” together with new spark plugs and wiring harnesses. The magnetos’ part numbers and serial numbers and the name(s) of the facilit(ies) that had performed the overhauls were not specified.

Disassembly and testing by the NTSB’s Materials Laboratory found that the breaker points opened and closed, but even while closed would not conduct electricity. Microscopic examination found them to be coated with deposits of some foreign substance. The gap between the armature and coil inside the case was “fouled with thick, greasy deposits,” and “the carbon electrodes and gear electrode exhibited heavy deposits and erosion, consistent with extended age and operation. Additionally, the points, distributor and brushes appeared to be at the end of their useful life and not consistent with a magneto that was recently overhauled.”

Of course, one of the principal reasons reciprocating aircraft engines use dual magnetos is to provide redundancy in case one fails. Finding a bad mag during runup is reason to ground the aircraft. We don’t know whether the pilot checked his mags before each flight, whether the right magneto actually stopped working during the accident flight, or whether the left might have misfired afterward. (Its condition is not discussed in the NTSB’s report.) However, the worn-out magneto wasn’t the airplane’s only weak point. Repairs to the left exhaust manifold in January 2014 included the installation of ball joints to prevent further cracking. Unfortunately, to accommodate this modification the size of the carburetor heat shroud had to be cut back by about half. The humid Gulf air created conditions conducive to serious carburetor icing at glide power, making the extended glide at idle power during the first flight potentially relevant to the accident on the second.

It seems like a sure bet that the owner didn’t know his “overhauled” magnetos hadn’t been. When carrying out scheduled overhauls, though—or even replacing old magnetos with factory-new units—many owners prefer to change them one at a time, flying the engine for 25 or 50 hours with one tried and tested component still in place. The reason is simple: to assure that one is known to be good. It just makes sense to avoid introducing multiple points of uncertainty.

One at a time

Related links