Statistically, an aircraft flying into a cloud of ash spewing from a volcano was a rarity—until April 2010, when Iceland’s Eyjafjallajokull erupted. The ash cloud closed airspace over Europe for days, canceling more than 100,000 flights and incurring airline losses approaching $2 billion. The disruption exposed how little the industry comprehended the effects of volcanic ash on aircraft and engines.
It was known that ingesting volcanic ash could damage engines, but the concentrations at which safety and performance effects could become critical were not well understood. This led to conservative assumptions for closing airspace, which were eased slightly to limit flight disruptions as the crisis persisted. The result was an upsurge in research into the effects of ash ingestion.
The biggest of these efforts is the-led Vehicle Integrated Propulsion Research (VIPR) program, which culminated in July with an ingestion test involving a Pratt & Whitney engine on a airlifter—on the ground but on-wing, to simulate prolonged flight through a volcanic ash cloud. “The engine surprised us,” says Paul Krasa, VIPR program manager at Research Center.
The VIPR program is a multiagency-industry partnership created in 2010. At that time,was looking at developing engine health management systems and sensors for next-generation aircraft engines. Then Eyjafjallajokull erupted and both the and U.S. Air Force “became very interested in the impact of volcanic ash on high-bypass turbofans,” says Krasa.
Previous encounters with ash had been inadvertent. In 1982, all four engines onFlight 9, en route from London to Auckland, flamed out when the -200 flew through ash from Indonesia’s Mount Galunggung. In 1989, all four engines failed on a 747-400, ’s Flight 897 from Amsterdam to Tokyo, when it flew through ash from Alaska’s Mount Redoubt. Both aircraft landed safely.
In 2000, en route from Edwards AFB, California, to Kiruna, Sweden, NASA’s own McDonnell Douglas DC-8-72 airborne laboratory inadvertently flew at high altitude through a diffuse ash cloud from Iceland’s Mount Hekla volcano. All fourengines had to be replaced. Coming after these events, the 2010 airspace closures increased interest in understanding the effects of ash.
“The Iceland eruption caused three weeks of disruption,” says Krasa. In addition to the impact on commercial airline flights over Europe, there were wars underway. “There was a huge impact on Air Force logistics to Iraq and Afghanistan. They had to send flights westward, the long way round.”
The result was a project that has involved not only all four of NASA’s aeronautics research centers at Armstrong, Langley, Ames and Glenn but also the FAA and Air Force Research Laboratory, while the Big Three engine manufacturers—Pratt & Whitney,and —and have contributed resources and research.
“When you do a test that has never been done before—introducing volcanic ash directly into an engine on the wing of an aircraft, you really need to understand the full system effects,” Krasa says. “We could not do it on a test stand. We needed to do it on an aircraft, as an integrated system, to understand the effects on the engine and how the flight crew perceive them.”
The Air Force loaned NASA two(PW2000) engines, flyable spares taken off a C-17 prototype in the Air Force Museum and overhauled by Pratt to restore their operating limits. The team was careful not to stress the engines too soon, he says, and VIPR testing was conducted in three phases.
In the first test, VIPR 1, “we ran a lot of peripheral sensors attached to the outside of the engine. We ran simulated faults, but they were not detrimental to the engine,” he says. VIPR 2 moved into modifying the engine to integrate sensors into the core. “We loaded it up with science, but did not harm the engine.” Tests included inducing faults to see how the sensors reacted, and injecting powered chalk to simulate ash and understand how to do the ultimate test.
“VIPR 3 was always the vision, to inject ash into the engine,” says Krasa. “And we learned a tremendous amount from the test.” VIPR 3 was the first controlled exposure of an engine to ash. Ground tests were performed by Calspan in the 1980s on a Pratt & Whitney F100 fighter engine to assess the performance deterioration from exposure to dust from nuclear explosions, “but we did it in a more controlled way,” he says.
Before VIPR 3 what was known about volcanic-ash ingestion is that the rapid impact can include erosion of the compressor and melting of the ash in the hot section, blocking the fuel system, clogging the combustor and coating the turbine, and blocking the cooling holes. Longer-term effects include loss of compressor efficiency, lubrication system contamination and reduced turbine component life.
For the tests, ash was injected at two flow rates—1 mg and 10 mg per cu. meter. Over Europe in 2010, “no-go zones” were established where ash concentrations exceeded 2 mg/cu. meter, later raised to 4 mg/cu. meter. Today engine manufacturers do not recommend operations in concentrations above 2 mg/cu. meter. The ash encountered by KLM Flight 897 was estimated at about 2,000 mg/cu. meter.
Conducting the ingestion tests was not as simple as shoveling ash into the engine. The material had to be selected carefully. Ash from Mount Mazama in Oregon was chosen with help from the U.S. Geological Survey, in part because the material occurs naturally on the dry lake bed at Edwards, where the test was conducted. “The majority of the ash went into the core, but some went through the fan and out the back, and this made the environmental release much easier,” says Krasa.
GE developed the volcanic ash distribution rig. “We did a lot of CFD [computational fluid dynamics] simulations and predicted over 90% of the ash would go into the core. The actual test was about 99%—that’s how tightly it was designed,” he says. “We never could see the ash going in, but we could see erosion on the fan—a cleaning of the blades over the first couple of inches.”
But GE had to redesign the rig. “They took the sand ingestion standard used in engine certification and put factors on top, because ash is more erosive and corrosive. They thought they were conservative, but the ash rig ate itself,” says Krasa. “Ash cut through fittings in a short period. So we recharacterized the ash. It is much nastier than we thought. Ash is fine like talcum, but under a microscope you can see it is so angular that it has cutting ability.”
The VIPR 3 ingestion tests totaled 14 hr. of engine runs over multiple days at the low flow rate, then two days at the high rate. For each run, the engine was operated at a nominally constant pressure ratio, and ash was fed into the core only when it was at the correct power.
Preliminary data has surprised the team. “At 1 mg/cu. meter we predicted we would see some [performance] degradation at 1 hr. Then at 10 mg/cu. meter we expected a redline breach [the engine no longer airworthy] after 3 hr.,” he says. “We ran at low flow rate for a week, followed by two days at a combined low and high rate and at no time was there a redline breach.”
There was degradation, but after a longer time than expected. “We saw the engine performance shift at the 10 hr. point. When we went from low to high flow rates we hit a knee in the degradation curve, but when we stopped the test at 14 hr., the engine was still running and putting out power,” Krasa says.
Borescoping the engine after the first day of high-flow tests, the team thought it would never start again because there was so much glassification in the high-pressure turbine section where the ash had melted and coated the rotors and stators. “We had the combined expertise of GE, Pratt & Whitney and Rolls-Royce going over the borescope results to determine if we could go ahead safely,” he says. “Rolls has one of the world’s definitive experts, Rory Clarkson, and he said not to worry.”
The glass-laden engine started and tests continued beyond the 10-hr. mark. “For the first 10 min. the engine would chug and cough, and a big brown cloud would come out the back. The glass is very brittle. It would accumulate and accumulate then, when thick enough, break off,” Krasa says.
“We expected in the first day of high-flow tests to have all the research done in 1.5-3 hr. By the end of that day, we saw degradation as the engine loaded up with glass. The second day we decided it was safe to start the engine, but thought it would not last the day. We started out predicting a redline breach in 3 hr.; at 14 hr. it had not hit the stop point,” he says. “Could we have gone on? The major reason we stopped is we had met our success criteria.”
Now that the tests are done, the head-scratching over the results has begun, he says. “At low flow, we saw an increase in performance in the beginning, which we think was a cleaning effect. They used to use walnut shells to clean engines early on, and our hypothesis is that at low flow we saw a slight cleaning [of the compressor] at first. As we went on we saw a knee in the curve and erosion in the compressor.”
In the high-pressure turbine, the team saw glassification effects that looked like ice shapes—“the shapes you get when supercooled water droplets hit the leading edge of an airfoil,” Krasa says. “The same thing may be happening with ash. As it comes through the hot section, it turns into very small volcanic glass droplets and you may get the same supercooling effect as with water droplets. Scientists look at the borescope images, see familiar ice-type shapes and wonder if there is a correlation. Can we take our analytical capability for ice shapes and apply it to ash?”
The results from VIPR are to be published early in 2016. The FAA and International Civil Aviation Organization (ICAO) are waiting for the data to better understand the effect ash has on engines. “ICAO will look at the data and understand more about to how to fly safely and whether the current guidance is too restrictive or not,” says Krasa.