Scientists in the propulsion system laboratory (PSL) at NASA’s Glenn research center in Cleveland, Ohio, have developed a test facility that can recreate high-altitude engine icing, a long-awaited capability that should equip the aviation industry to tackle a poorly understood hazard.
The researchers have adapted a variable-pressure wind tunnel that was capable of simulating altitude for engine tests and modified it to generate ice crystals at simulated altitudes up to 40,000 feet and airspeeds up to Mach 0.8. The modified test chamber can lower temperatures to -60 degrees F (-51 degrees C). As part of the process, water is introduced via spray bars that allow the researchers to control the size and amount of ice in a “cloud” it creates. The ice particles generated are spherical frozen droplets, Dr. Judith Foss Van Zante, icing engineering technical lead, told AIN.
Explaining the principle, Van Zante said the particles can be as small as 15 microns (a micron measures one-millionth of a meter). Actually, 15 microns is a median volumetric diameter (MVD), a term that accounts for the distribution of particle sizes, she said. Larger microns also can be created, “but if we want them to be entirely glaciated, we have to keep the MVD below 100 microns,” she continued.
This is not an issue because it’s the smaller particles that the researchers want. Smaller particles are entrained into the air more readily and hence flow into the engine core (see box), according to Dr. Michael Oliver, aerospace engineer and research lead for PSL engine icing tests. Larger particles tend to hit the fan and either break up into smaller particles or be centrifuged out to the fan bypass flow path.
To date, the PSL has performed only one series of tests, which lasted 80 hours logged in February last year. They were run on an obsolete 8,000-pound-thrust Honeywell ALF502-R5 turbofan, which the OEM had been flight-testing after it experienced icing problems (since resolved) on the BAe 146 regional jet.
During the tests, the PSL researchers created conditions that caused the engine to lose power completely. The study involved more than 130 test points. To protect the test engine, only three test points were allowed to extend to a full rollback. For the remaining test points, the researchers developed a so-called rollback procedure that tracked key test parameters as ice built to operational limits. This allowed them to halt the test when the rollback was imminent.
They found that, for a given ice content, variations in altitude and temperature effected a rollback. “We identified a boundary,” Oliver said, and within the boundary there was sufficient ice build-up to cause rollback. As the boundary was approached, ice build-up occurred less rapidly. Beyond the boundary it ceased to build.
The icing facility’s main limitation is its maximum airflow–330 pounds per second. This limits the thrust of the engines at test for higher-altitude points to 30,000 to 40,000 pounds. However, the PSL’s flow constraint meets the core airflow rate for all current commercial engines in revenue service, Oliver pointed out, suggesting further that an engine core or driven compressor rig could be used to research icing in large commercial engines.
Oliver made it clear that the PSL is strictly a research-and-development tool and cannot be used for certification purposes. However, manufacturers may use the icing wind tunnel in Cleveland. The group hopes its research data will assist the industry with its certification efforts.
“With the controlled environment, we can study where and how the ice is accumulating,” Van Zante said. It should also help fine-tune simulation software programs. “Our codes have been able to predict test results,” she added.
How Ice Forms in Engines
The principles governing how ice crystals can create internal engine icing in flight are now understood. Engine icing is caused by ice crystals, as opposed to the supercooled liquid droplets that can build ice on external surfaces. First, ice crystals bounce off freezing surfaces near the front of the engine and enter the core. Then, in the compressor, they melt on vane surfaces, creating a film of water. Particles keep impinging on the wet vane, cooling it enough for ice to form, which starts an ice-accretion process. Eventually, the ice breaks off and can cause engine malfunctions.
Some 250-plus events related to operations in ice-crystal conditions have been recorded in 30 years. The consequences for the engines range from partial loss of power to surge and complete flameout. Sometimes the pilots don’t even notice the effect, but sometimes the result is more serious. In 2005, the crew of a Beechcraft 400A had to perform a dead-stick landing at Jacksonville (Fla.) Airport as a result of engine icing.