It’s difficult to believe that an airplane flying in the tropics in the summer could have an engine fill up with ice, freeze, and shut down. But the phenomenon, known as engine core ice accretion, has happened more than 150 times since 1988 – frequently enough to attract the attention of NASA aviation safety experts, who are preparing a flight campaign in northern Australia to learn more about this occasional hazard and what can be done to prevent it.
“It’s not happening in one particular type of engine and it’s not happening on one particular type of airframe,” said Tom Ratvasky, an icing flight research engineer at NASA’s Glenn Research Center in Cleveland. “The problem can be found on aircraft as big as large commercial airliners, all the way down to business-sized jet aircraft.” And it has happened at altitudes up to 41,000 feet.
No accident has been attributed to the phenomenon in the 23 years since it was identified, but there have been some harrowing moments in the air. In most of the known cases, pilots have managed to restore engine power and reach their destinations without further problems. According to the Federal Aviation Administration, there have been two forced landings. For example, in 2005, both engines of a Beechcraft business jet failed at 38,000 feet above Jacksonville, Fla. The pilot glided the aircraft to an airport, dodging thunderstorms and ominous clouds on the way down. Engine core ice accretion was to blame.
Little is understood about ice crystal properties at high altitude and how ice accumulates inside engines. The engines may be toasty warm inside at such heights, but the air outside is frosty cold. The prevailing theory holds the trouble occurs around tropical storms in which strong convection currents move moist air from low altitudes to high altitudes where the local temperatures are very cold, creating high concentrations of ice crystals. But the properties of the ice crystals, such as their size and how many of them are in a given volume of air, are a mystery – one that an international research team led by NASA aims to solve.
The FAA has proposed new certification standards for engines that will be operated in atmospheric conditions that generate ice crystals. The rules will take effect next year, just as the NASA team heads to Darwin, Australia, aboard an aircraft specially equipped with instruments to study cloud physics during the Southern Hemisphere summer. Analyses of the Darwin flight tests and additional tests in ground-based facilities in the United States and Canada will provide the FAA the means for ensuring compliance with the new standards.
“We need to understand what that environment is out there and, even though it may be a rare case, be able to fly through those icing conditions unscathed. Or if we can find ways of detecting this condition and keep aircraft out of it, that’s something we’re interested in doing,” Ratvasky said.
Researchers explain the phenomenon this way: Small ice crystals found in storm clouds get sucked into the core of an aircraft engine, where the pressure is high and the temperature is warm. Some of the ice melts and covers the warm engine parts with a thin film of water that traps additional ice crystals. The super-cooled water chills the engine components enough that ice can accumulate on them. If the built-up ice breaks away in chunks it can damage compressor blades, reduce the power level, or snuff out the engine altogether.
For the flight research, NASA is outfitting a Gulfstream 2 business jet with more than 20 meteorological sensors that will be used to probe cloud properties, such as water content and the size and concentration of ice particles, which can lead to engine and air data sensor failures that threaten aviation safety. The data gathered will aid scientists’ understanding of cloud growth processes, help them create reliable detection methods and realistic ground-based simulations, and provide a foundation for possible new aircraft design and certification standards. FAA can use what the team learns over the course of its research project to verify the range of atmospheric conditions addressed in the new standards.
The flight campaign has three primary goals:
- Characterize the range of environmental conditions in which internal engine icing can take place, with an emphasis on how much water or ice is present in a given volume of air.
- Determine how to identify geographic regions where such weather threatens and ways to detect the conditions in real time in order to develop guidance that pilots can use to avoid the hazard.
- Collect enough data to enable researchers to simulate the weather conditions for aircraft engine tests in ground facilities such as Glenn’s Propulsion Systems Laboratory.
“Our plan is to study the weather patterns that lead to these conditions, not to test a particular engine configuration. We do not plan to intentionally cause our engines to have an icing event,” Ratvasky said.
The Propulsion Systems Laboratory recently underwent upgrades to equip it for ground-based simulations of high-altitude icing conditions. Work to transform the Gulfstream 2 into a working airborne science laboratory is under way at a NASA contractor site, Flight Test Associates in Mojave, Calif., and will be completed early in 2012. Engineers will mount six instruments on each wing and additional instruments on the fuselage to measure cloud particle size and shape and water content, whether the particles are liquid or crystal, and the speed of the updraft as cloud particles form.
The research team – with representatives from FAA, The Boeing Company, the U.S. National Center for Atmospheric Research, Environment Canada, the National Research Council of Canada, Transport Canada, Airbus and the Australian Bureau of Meteorology – will conduct trial runs during the monsoon season in February and March 2012, develop findings and address lessons learned, and then return in January through March 2013 for the primary flight campaign.
The team chose Darwin for several reasons: its ground-based weather observing systems are the best in the tropics, there will be plenty of storms to sample, there is plenty of data from previous atmospheric characterization efforts with which to compare, and the Southeast Asia region has seen a large number of engine power-loss events.