A new technique for rotating stall precursor identification in
high-speed compressors has been developed at the NASA Lewis Research
Center. This pseudo correlation integral method uses a mathematical
algorithm based on chaos theory to identify nonlinear dynamic
changes in the compressor. Through a study of four various configurations
of a high-speed compressor stage, a multistage compressor rig,
and an axi-centrifugal engine test, this algorithm, using only
a single pressure sensor, has consistently predicted the onset
of rotating stall.
Data for this algorithm have been collected from a single high-frequency-response
pressure transducer located one chord length upstream of the rotor
and flush mounted in the casing. The pressure signal data are
input to the correlation integral algorithm. The algorithm calculates
a phase plane portrait of the pressure data for one window in
time. As pressure conditions change in the compressor, specifically
through throttle area closing, the algorithm monitors dynamic
changes in the phase plane portrait. As the phase portrait shrinks
or enlarges, stretches or twists, the correlation integral calculation
monitors these changes. From calculations performed on four various
configurations of a high-speed compressor stage, the data confirm
a sharp decrease in the correlation integral prior to the onset
of rotating stall.
As an example, the phase plane portraits for a high-speed compressor
stage are presented at four operational conditions in the figure.
A phase plane portrait displays the pressure data on the
x
-axis
and a time-delayed version on the
y
-axis. Part (a) shows
the pressure signal for steady operation of the compressor stage
with the throttle fully open. Part (b) shows the phase plane portrait
of the compressor with the throttle partially closed, but still
far from the stall condition. Part (c) shows the pressure signal
very near the stall condition, and part (d) shows the compressor
in full rotating stall. The phase plane portraits in parts (a)
and (b) are very similar in physical nature, and their correlation
integrals are identical. In part (c), the physical nature of the
portrait has changed slightly and the correlation integral calculation
has decreased. In the rotating stall condition of part (d), the
portrait has changed significantly from the previous portraits
and the correlation integral has dropped significantly. This decrease
in correlation integral prior to rotating stall, which is a precursive
signal that rotating stall will occur, is present in all test
cases.
To provide a basis for comparison, we applied the traveling wave
energy technique (ref. 1), which has been used extensively to
study prestall data, to identical data sets (ref. 2). The correlation
method was shown to have a potential advantage over the traveling
wave energy method because it uses a single sensor for detection
and does not require the data to detect changes in the behavior
of the compressor. Both methods were used in this study to identify
stall precursive events in the pressure fluctuations measured
from circumferential pressure transducers located at the front
face of the compressor rig. The correlation method successfully
identified stall formation or changes in the compressor dynamics
from data captured from four different configurations of a NASA
Lewis single-stage high-speed compressor while it transitioned
from stable operation to stall. The traveling-wave energy technique,
which requires eight circumferentially located transducers, was
not able to predict the onset of rotating stall for one of the
test cases. The experimental results indicate that the correlation
method provides warning of the onset of rotating stall at high
speed. This method is being expanded into an online diagnostic
and prediction tool as part of an integrated active stall control
system.
Previous articleLast updated May 5, 1997
Responsible NASA Official:
Gynelle.C.Steele@nasa.gov
216-433-8258
Point of contact for NASA Glenn's Research & Technology reports:
Cynthia.L.Dreibelbis@nasa.gov
216-433-2912
SGT, Inc.
Web page curator:
Nancy.L.Obryan@nasa.gov
216-433-5793
Wyle Information Systems, LLC
