In 1984, NASA and the FAA remotely crashed an empty Boeing 720 to test whether anti-misting fuel could reduce post-crash fires

In 1984, NASA's Dryden Flight Research Center teamed up with the Federal Aviation Administration for an experiment known as the Controlled Impact Demonstration. The test was designed to answer a specific question: could a special fuel additive reduce the risk of fire after an aircraft crash? The full-scale test did not produce the result researchers had hoped for. It was a planned crash carried out remotely, with no one aboard the aircraft.

Preparing the Boeing 720 for its final flight

Post-crash fires were a major danger in aeroplane accidents. NASA and the FAA tested a fuel additive called FM-9 to see whether it could reduce that risk. FM-9 was a high-molecular-weight, long-chain polymer that, when blended with Jet-A fuel, had shown an ability to inhibit ignition and flame propagation in simulated impact tests.


To see how the additive performed in a full-scale crash, the teams used an obsolete Boeing 720, a four-engine airliner obtained from the FAA. The aircraft was to be deliberately crashed into steel structures installed on Rogers Dry Lake at Edwards Air Force Base. Known as wing cutters, the structures were designed to tear open the wing fuel tanks during the impact. The test was also designed to generate data on aircraft structure and crash survivability.

The modified fuel created a technical problem of its own. Because the anti-misting fuel could cause issues such as clogged filters, it could not be fed directly into the engines. Instead, each of the Boeing 720's four engines was fitted with a degrader that restored the fuel to nearly Jet-A standard before combustion.

The flight also carried a separate set of crashworthiness experiments. NASA's Langley Research Center installed instrumented dummies in the passenger cabin and collected data on seat designs, flight data recorders, cabin fireproof materials, burn-resistant windows and other components. The project involved more than four years of preparation before the final demonstration, including work to establish impact conditions that the FAA considered survivable.

A remotely controlled crash

The Boeing 720 was modified so it could be flown from the ground without a flight crew aboard. Before the final crash, engineers conducted 14 preliminary flights with safety pilots on board while gradually introducing the modified fuel and monitoring engine performance. During those flights, the remotely controlled aircraft accumulated 16 hours and 22 minutes of flight time. The test programme included 10 takeoffs, 69 controlled landing approaches and 13 landings on the abort runway.

The final flight took place on December 1, 1984, with no one aboard. The remotely controlled aircraft completed a nine-minute flight before reaching the prepared crash site. The aircraft was supposed to approach the impact area wings level and along the planned flight path so that the wing cutters would open the fuel tanks. As it approached the crash site, however, it deviated from the intended flight path.

The left wing struck the ground before the aircraft reached the wing cutters, sending the aircraft into a yaw as it skidded across Rogers Dry Lake. One of the cutters subsequently struck an engine on the right wing rather than the fuel tank. The engine was destroyed, a large fuel leak developed and the fuel caught fire. The aircraft continued to yaw as the damaged right wing broke away and the fuselage became engulfed in a fuel-fed fireball. The fire took more than an hour to extinguish.

The fire did not make the entire experiment unsuccessful. Cameras inside the aircraft captured the behaviour of the instrumented dummies and cabin components during the impact, while the crash generated data on structural loads, seats and other survivability measures. The structural and crashworthiness side of the experiment produced significant full-scale data.

FM-9, however, did not perform as hoped in the demonstration. Earlier small-scale tests had suggested that the anti-misting additive could reduce post-crash fires, but the Boeing 720 test showed that the modified fuel was ineffective at reducing the propagation or intensity of the fire. The result also ended the FAA's effort to require the additive in airliner fuel. The agency's requirements for the additive were subsequently cancelled.

The Controlled Impact Demonstration therefore produced two very different outcomes. FM-9 did not deliver the fire-reduction effect seen in earlier tests, while the crash generated full-scale data on aircraft structure and passenger survivability. The experiment also demonstrated the difficulty of translating promising results from smaller tests into the conditions of a full-scale aircraft crash.