“The prop was rotating at impact.”
An hour later, the class has it down: They can spot bowed flanges, shorn bolts, torsion, and ground scars immediately. Learning how to accumulate such details and objectively assess their significance, Wall tells them, is the first step toward narrowing the possibilities for the cause of the accident. The real answers, after all, are often very far from what first impressions may suggest and usually include multiple failures, so investigators must have a comprehensive understanding of what the aircraft—and, ultimately, its crew—has gone through. Wall describes some of the analytical tools available. He tells them to document the site with videos, photographs, and grid sketches showing debris distribution lines. Another trick is vector analysis, drawing arrows on the wreckage that show the direction that forces are being applied. “You’ll get the big picture very fast,” Wall says.
Emphasis on the big picture is clear in the range of the course curriculum. It’s a general introduction, and it’s just one of many that investigators will take throughout their careers. The class, which is taught by TSI, NTSB, and FAA staff as well as aviation industry experts, covers all types of aircraft and gives equal weight to both clinical discussions of aircraft component failure modes and the human side, describing how to work with witnesses, survivors, and family members and what investigators might experience at the site. “When you arrive at a bad accident site, I guarantee you you will not sleep that night,” warns Frank Del Gandio, an FAA investigator who lectures on crash-scene biohazards, including blood-borne pathogens such as the hepatitis virus and HIV. “In fact, you might find you’re not sleeping for days. That’s normal. Don’t worry. But you need to focus on that investigation. Focus on the people you might be able to help in the future.”
That’s a virtual mantra for accident investigators, who must often work at remote and inhospitable sites and with a mind-numbing collection of variables. Their cause, though, is safety—to figure out precisely what happened at each accident so that the problem can be prevented from happening again. According to the NTSB, there are roughly 2,000 aviation accidents per year, most involving small general aviation aircraft and about 700 involving fatalities. All are thoroughly investigated with a variety of team configurations that can include representatives from the airline, if applicable, the aircraft manufacturer, and any part suppliers or airport personnel who might contribute useful records or data. If there are fatalities, the NTSB will lead the on-site investigation. If not, the board might delegate the investigation to the local FAA office. (The FAA, which is responsible for regulating the aviation industry and operations, always participates in investigations in support of the NTSB.) In all cases, the facts are reported to the five-member board, which will review the accident, develop a probable cause, and possibly issue recommendations for the FAA to enact.
The Transportation Safety Institute, a Department of Transportation division charged with training investigators of aviation, highway, and marine accidents, trains some 600 FAA investigators a year. It also opens its doors to tuition-paying military and commercial airline safety specialists. In the course I’m participating in, most of my classmates are these specialists, who oversee flight operations with an eye toward safety procedures, training, and maintenance and who might eventually participate in investigations that their companies or military units conduct in support of or in addition to an NTSB inquiry.
All training stresses that accident investigation is a team effort. After a crash is reported, the investigators, led by the Investigator In Charge, break into groups, each of which gathers evidence about the pilot, the powerplant, the aircraft structure, air traffic control, the weather, and other factors. They interview witnesses and begin sorting through debris. The students are taught that they must account for every component of an aircraft, if for no other reason than to rule it out as a contributing factor.
One of the biggest challenges in that respect is determining whether components were damaged before the accident or during it. Among the most thoroughly addressed subjects is fire. Wall passes around a small charred canister. It’s an oxygen container identical to the one that started the fire that brought down ValuJet Flight 592 in the Florida Everglades in May 1996, killing 110 people. This one was used to test whether such a fire could have occurred, and its presence silences the room. “Always suspect the possibility of in-flight fire,” Wall says. “When a fire is in the slipstream and fueled by lots of oxygen, such as in an engine, temperatures can exceed 3,000 degrees, whereas on the ground they will usually stay below 2,000 degrees.”
Materials leave distinctive signatures when subjected to certain temperatures, he continues, thus possibly revealing whether the fire occurred in the air or after the crash and where it originated. Rubber hoses, for example, will melt at 400 degrees, aluminum alloys at 1,000 degrees, and stainless steel at 3,100 degrees. Furthermore, when aluminum melts on the ground, it will puddle due to gravity, but when it comes apart in an in-flight fire, it produces the “broomstraw effect,” in which the points where the metal has come apart will look stringy.
Other damage might require closer inspection. Andy McMinn, a TSI staff instructor, briefs us on metallurgy and how to use it to determine what caused a part to fail. Throughout his talk, he refers to crystal structure—how atoms are arranged—and discusses the various ways parts can change, depending on whether they have been subjected to fatigue, violence, or some other stress. Parts usually fail by overload, the material can be ductile or brittle, and the failure will leave important clues. “Fatigue manifests itself in ‘beach marks,’ like tide marks in the sand,” McMinn says. “If a rectangular plate is pulled and compressed thousands of times, beach marks will indicate the direction of crack propagation. When the cracks reach a critical size, you’ll see progressive failure and then instantaneous failure.”