When a rotor fails, some serious buck-passing begins (surprise!) between the various parties that might bear responsibility for the crash.
By Ken Russell, Contributing Writer
“A thingamajig that almost doesn’t work” is one expert’s description of a helicopter. Yet, the ability of these aircraft to take off vertically and hover in midair makes them highly valued in both civilian and military applications.
Helicopters are often referred to as “rotary winged” aircraft, as opposed to the more common fixed wing aircraft. The rotors must provide both propulsion and control. Therein lies a difficulty. Loss of all or part of a rotor causes a loss in both control and propulsion.
In this case, a private, European-built helicopter crashed. Injuries were fairly minor, but the helicopter was badly damaged. The helicopter cost about a million dollars new, so the loss was enough to cause serious buck-passing between the various parties that might bear responsibility for the crash, which was caused by the loss of part of the tail rotor.
My job was to deduce the chain of events leading to the failure. I received a part of the failed rotor including one side of the fracture surface. The fracture occurred straight across the rotor.
The rotor was made of flattish aluminum tubing that had been reinforced by a piece of stainless-steel sheet called a “doubler.” The innermost, highest-stressed part of the tubing was reinforced.
My study consisted of hardness tests, metallographic study of polished specimens and scanning electron microscopy (SEM) of the fracture surface. The hardness tests and the metallography showed nothing out of the ordinary. The SEM was equipped to do chemical analysis of the substance being viewed. It showed that the doubler was ordinary non-magnetic 18 percent chromium, 8 percent nickel stainless steel and the tubing to be aluminum 4 percent copper. Both are very common alloys suitable for use in the rotor.
Fatigue failure is always the prime suspect in dynamic failures, and is revealed by fine, more or less parallel striations on the fracture surface. These striations are formed as the fatigue crack moves during each loading cycle. Usually a fatigue crack starts at a single origin such as a crack or notch and gradually proceeds slowly and regularly outward at a gradually accelerating rate. Finally, the strength of the member is so reduced by the crack that the remaining cross section fails suddenly by ductile fracture.
The SEM study showed the fracture surface to be covered by striations. There had been a slew of fatigue cracks starting at various origins. I could not see any cracks or nicks that might have started the culprit cracks. There was no sign of the ductile fast fracture region. Fracture had occurred almost completely by propagation of the fatigue cracks. The fracture surface also showed a lot of rubbing, so the rotor had been operated a considerable time in the partially cracked condition.
The rotor resembled the celebrated “Wonderful One-Hoss Shay” in which all the parts were equally strong so that the entire conveyance failed at a single moment.
My observations led me to conclude that the failure was due to overload rather than to a defect in materials or workmanship. It could of course be that the rotor was simply under-designed. But, if such were the case, many rotors would have failed. It is more likely that the rotor was operated in an unbalanced condition which would give higher than expected stresses.
The rotor had a 1,200-hour estimated life and a 500-hour warranty. Failure occurred at 566 hours. My lawyer client wanted to know if the crack started before the 500-hour warranty expired. I assured him that the rotor was well on its way to failure before the warranty was void. This opinion seemed to satisfy him.
I never learned how the case ended, but suspect the manufacturer wrote the biggest check.
Contributing writer Ken Russell (firstname.lastname@example.org) is professor emeritus of Metallurgy and Nuclear Engineering at MIT. he speizlies in physical metallurgy, forensic metallurgy, and failure analysis. Cases presented here are drawn from his actual forensic files and court cases.