0
Select Articles

Out Through the Intake PUBLIC ACCESS

Stall and Surge are Endemic Scourges of Jet Engine and Gas Turbine Operation

[+] Author Notes

Lee S. Langston is professor emeritus of mechanical engineering at the University of Connecticut in Storrs and a frequent contributor to Mechanical Engineering magazine.

Mechanical Engineering 139(04), 36-41 (Apr 01, 2017) (6 pages) Paper No: ME-17-APR2; doi: 10.1115/1.2017-Apr-2

This article elaborates various challenges presented by stall and surge to the gas turbine research community and jet engine designers. The article also presents several examples of stall and surge faced by the pilots. Stall and surge emerge from basic physics: the behavior of the boundary layer on the compressor blades and stators; however, current technology has no means to completely eliminate it. Engine control systems, such as the Full Authority Digital Electric Control (FADEC), are programmed to keep the operating point of the compressor well away from the surge line. Researchers have been studying stall and surge for decades, looking for ways to predict and combat the phenomena. Meanwhile, there has been some success in using FADEC to either prevent a stall and surge or to limit the number of repetitions. FADEC can also step in during flights in heavy rain or hailstorms. In those conditions, extra fuel is needed to process and evaporate the water being swallowed by the engine. The electronic control system can sense the mismatch between the power setting and the fuel flow and take action to prevent possible engine instability.

Modern jet engines are so reliable that airline pilots can fly an entire career without experiencing an engine failure or having to shut down an engine for a minor problem. Most passengers never have to experience mechanical failure except as the reason for an otherwise unexplained grounding of a flight.

One engine of this C-17A suffers a compressor surge while taxiing on an airfield.

Photo: Will Mallinson

Grahic Jump LocationOne engine of this C-17A suffers a compressor surge while taxiing on an airfield.Photo: Will Mallinson

Engine failure—especially one caused by engine stall and surge—may be rare, but it is unforgettable.

Recently, I started discretely asking people in the gas turbine community about stall and surge incidents they had encountered. A retired pilot I know described to me one event, on a flight from Portland bound for Tokyo. He said that during the takeoff run, a single Herring gull was “ingested” into one of the engines. In a blink, the cowling blew right off the front of the engine and destroyed the tires of the landing gear. The takeoff was aborted.

The problem wasn’t the bird itself. Jet engines are designed to withstand strikes from small birds like gulls (though the birds are not designed so resiliently). Instead, the bird disturbed the flow of air entering the engine and the compressor experienced a stall. In an instant, superheated gases from the combustion chamber that normally power the turbine in the rear of the engine surged forward.

Stall and surge are two words that catch a gas turbine engineer's immediate attention. Those two types of air flow disturbances in the axial compressor of a jet engine or gas turbine represent the breakdown of orderly flow through them. Modern design and fuel control systems try to keep a jet engine or an electrical power gas turbine away from operating conditions that bring about stall and surge. But they don’t always succeed.

I have been thinking about stall and surge ever since the most recent International Gas Turbine Institute Scholar Lecture, when Ivor Day of the Whittle Laboratory at Cambridge University summarized the progress that has been made in understanding the phenomenon and highlighted some steps that are being taken to control it. In spite of those efforts, stall and surge present a challenge to the gas turbine research community and a worry to jet engine designers.

Gas turbines can produce power over many orders of magnitude, from kilowatts to hundreds of megawatts. The useful output is, however, only a fraction of the power produced by the turbine component of the machine. Some 50 to 70 percent of the turbine output is diverted to drive the axial flow compressor upstream. Compressing a gas is not easy.

Image credit: Robert Mazzawy, Trebor Systems.

Grahic Jump LocationImage credit: Robert Mazzawy, Trebor Systems.

Axial compressors get their name because the path of flowing gas runs more or less in a straight line parallel to the gas turbine's axis of rotation. The compressor is assembled from stages, each comprised of a ring of moving rotor blades mounted on a rotating disc or drum and a downstream ring of case-mounted stationary stator blades simply known as stators. The rotor blades act on the gas path air flow, increasing its static pressure, total pressure, and kinetic energy. The stators remove blade-induced swirl velocity, and thus decrease the kinetic energy, while serving to increase static pressure and align airflow for blades in the next stage.

Each additional stage increases both static and total pressure of the gas path. Typically, for industrial gas turbines, each stage operates in a pressure ratio range up to about 1.4:1. Air entering a six-stage compressor with a pressure ratio of, say, 1.2:1 would see an overall compression of 1.2 raised to the power of 6, or 2.99 times the initial pressure.

The thermal efficiency of a gas turbine is directly tied to the overall pressure ratio, which creates an incentive to maximize compression. Some 70 years ago, an axial compressor with 15 stages might have an overall pressure ratio of 4:1. Today, a 231 MW GE 7F.05 gas turbine has a 14 stage compressor with an overall pressure ratio of 18.4:1 and for jet engines the ratio can be as high as 58:1. As a consequence, gas turbine efficiencies have more than doubled.

Together, the compressor blades and stators operate on the gas path flow to produce what aerodynamists call an adverse pressure gradient in the flow direction—they turn low static pressure air into high-pressure air. It's analogous to pushing water up an inclined channel, with many small, rapid brush strokes. If the incline is too steep, the water runs backward, down the slope. (By contrast, the turbine operates in a decreasing, favorable static pressure field—as the expanding combustion gases act to turn those blades, their static pressure decreases.)

Over the course of the testing, engineers will determine for each pressure ratio the air flow rate below which the compressor fails to operate properly.

No matter how advanced axial compressors become, they must be carefully controlled in their operation to avoid the power-robbing effects of stall and the convulsive effects of complete flow reversal, brought about by surge.

What does this convulsion look like? Another correspondent described a videotape of the first few minutes of an airline flight shot through a cabin window by a passenger. First, the runway streamed by during takeoff and the jet began to climb. “Suddenly the inboard right engine gives out a boom,” he wrote, “and sheets of flame start pulsating out of the engine as the countryside and fields below fall away.” Alarmingly, as the video continued, the ground (and farm animals in nearby fields) started to get closer even as the nose of the plane continued to point up. Eventually, the plane recovered enough altitude, though not before the tower crew, which saw the plane dip below the tree line, alerted emergency responders of a probable crash.

To avert disaster (and near disasters like the one above) any gas turbine company that designs a new compressor will test it thoroughly, varying air flow rates and rotor speed to measure pressure ratio and efficiency, creating what they call a compressor map. Over the course of the testing, engineers will determine for each pressure ratio the air flow rate below which the compressor fails to operate properly.

In industry parlance, that's called the compressor surge line, and that line is labelled as such as it runs across the compressor map. But the actual phenomenon it measures is aerodynamic stall.

To understand why, consider the streamlines of the air flowing across the compressor blades. Going from low pressure at the blade's leading edge to a higher pressure at its trailing edge, the streamlines closely follow the blade's suction and pressure surfaces.

The flow around the blade is controlled by its boundary layer: the very thin, almost immeasurable layer of air on the blade surface, where viscous frictional effects are concentrated. Within the boundary layer, the air velocity relative to the blade decreases from that of the external streamlines, to zero at the blade surface. The concept of the boundary layer was discovered by the engineer Ludwig Prandtl in 1904, at an appropriate time to profoundly influence the design and operation of aircraft—and turbomachinery.

In aircraft, when the angle of attack exceeds a certain level, the boundary layer (which is exceedingly sensitive to an adverse pressure gradient) separates from the airfoil and the wing loses lift; the plane stalls. The same effect plagues axial compressors. Generally, what changes the angle of attack within a compressor is a reduction in air flow due to a blockage downstream in the combustor or turbine or an interference to the upstream air inlet. Roughness on the compressor blades or too large a gap between the rotating blades and the engine casing can also disrupt the boundary layer. Regardless, when the boundary layer separates from the surface such that the streamlines no longer follow along the suction side of the blade, the compressor blade is stalled.

When an airplane stalls, the aircraft loses altitude. If the pilot cannot restore lift across the wing, the plane will crash.

A stalled blade in an axial compressor loses its ability to increase a pressure gradient. What then?

With the axial flow so disrupted, flames from the combustor can shoot out the back of the turbine outlet, or even forward through the compressor inlet.

As Ivor Day explained in his IGTI Turbo Expo talk, a stall disturbs the compressor flow in the tangential direction, while the average axial airflow through the compressor remains steady. But a stalled compressor blade can itself create a blockage which diverts the approaching stage flow. This can trigger separation in adjacent blades opposite to the direction of rotation on the same rotor, creating something called a stall cell.

In certain circumstances, that stall cell will start to move, rotating in the opposite direction of the compressor and at half its rotational speed. Needless to say, those rotating stalls can lead to greatly shortened blade life, through increased stress and vibration.

In extreme cases, however, rotating stall can cause the compressor flow to fail altogether, disturbing the airflow in the axial direction. The airflow rate will pulse on millisecond timescales— sometimes so violently that the flow through the compressor is reversed. (This reverse flow is often accompanied by a loud bang, like a backfire.) With the axial flow so disrupted, flames from the combustor can shoot out the back of the turbine outlet, or even forward through the compressor inlet.

That degree of compressor flow failure is called a surge. It is to be avoided.

Another surge incident I learned about involved an airliner with an engine that had undetected erosion in the lining of its compressor casing. As the plane took off on a transcontinental flight, the engine let out a loud bang—then 57 more in the 70 seconds it took the pilot to shut down the engine. The crippled plane had to circle and dump fuel for an hour before it was able to make a single-engine landing.

A Sukhoi T-50 PAK FA suffers a compressor surge.

Credit: Wikimedia

Grahic Jump LocationA Sukhoi T-50 PAK FA suffers a compressor surge.Credit: Wikimedia

How can we avert these sorts of engine failures?

Stall and surge emerge from basic physics: the behavior of the boundary layer on the compressor blades and stators. Current technology has no means to completely eliminate it.

Right now, gas turbine manufacturers resort to trial and error, plumbing the contour of the surge line of a compressor through extensive testing to determine what conditions to avoid. Engine control systems, such as the Full Authority Digital Electric Control, or FADEC, are programmed to keep the operating point of the compressor well away from the surge line. Mechanical systems within the engine, such as compressor bleeds, casing treatments, tip clearance controls, and variable pitch stators that control the direction of airflow to rotors are all used to avoid stall conditions.

Those mechanical systems work within the blink of an eye. The time needed to adjust a variable stator or a bleed is on the order of 200 milliseconds.

That is not fast enough.

Researchers have been studying stall and surge for decades, looking for ways to predict and combat the phenomena. According to Robert Mazzaway of Trebor Systems, who in 1980 was one of the first to report on the engine stress created by surges, researchers have found that subtle modal waves lead to the rotating stall that precipitates surge. Detect those waves, the thinking goes, and FADEC could act to prevent the stall. Unfortunately, a rotating stall develops over the course of just a handful of rotor revolutions—which take only 20 milliseconds for an industrial gas turbine and only 5 millisecond for a jet engine. There's not enough time, then, for the mechanical systems to counteract the rotating stall and ensuing surge.

Even so, there has been some success in using FADEC to either prevent a stall and surge or to limit the number of repetitions. In twin-spool engines, for instance, stall can be caused by a mismatch between the rotation rates of the low-spool and high-spool rotors, usually due to a disruption in the fan stream. In such cases, however, the low-spool rotor increases it rotational speed slowly enough for the FADEC to sense the mismatch in RPMs between the spools and take measures to counteract it.

FADEC can also step in during flights in heavy rain or hailstorms. In those conditions, extra fuel is needed to process and evaporate the water being swallowed by the engine. The electronic control system can sense the mismatch between the power setting and the fuel flow and take action to prevent possible engine instability.

Fortunately for me, my interest in stall and surge is one of curiosity, spurred by Ivor Day's comprehensive IGTI lecture. For others, however, it is a vivid experience that still raises the hairs on the back of their neck.

“It felt just like we had hit a telephone pole with the right wing,” one acquaintance told me about a surge event on a flight he was on. “It was short and abrupt, but scary as hell.”

I. J. Day. “Stall, Surge and 75 years of Research.”ASME J. Turbomach. 138 (Oct. 2015): 011001, 16 pages. [CrossRef]
R.S. Mazzaway. “Surge-Induced Structural Loads in Gas Turbines.”ASME J. Engr. Power 102 (Jan. 1980): 162–168. [CrossRef]
E.S. Taylor. “Evolution of the Jet Engine.”Astronautics & Aeronautics 8 (1980) 64–72.
Copyright © 2017 by ASME
View article in PDF format.

References

I. J. Day. “Stall, Surge and 75 years of Research.”ASME J. Turbomach. 138 (Oct. 2015): 011001, 16 pages. [CrossRef]
R.S. Mazzaway. “Surge-Induced Structural Loads in Gas Turbines.”ASME J. Engr. Power 102 (Jan. 1980): 162–168. [CrossRef]
E.S. Taylor. “Evolution of the Jet Engine.”Astronautics & Aeronautics 8 (1980) 64–72.

Figures

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In