ACOUSTIC WAVE PROPAGATION IN PRESSURE SENSE LINES

Patrick Vitarius, Physics Dept, University of Alabama in Huntsville.

Don A. Gregory, PhD, Physics Dept, University of Alabama in Huntsville'
John Wiley, Transportation Directorate, NASA Marshall Space Flight Center
Valentin Korman, Madison Research Corporation

ABSTRACT per pound per square inch) and a maximum pressure

measurement of 2 psi.
Sense lines are used in pressure measurements to
passively transmit information from hostile The sensors themselves are quite expensive and cannot
environments to areas where transducers can be used. be sacrificed every test so a method must be used to
The transfer function of a sense line can be used to move the sensor itself away from the hot engine gasses.
obtain information about the measured environment A substantial part of the research performed was
from the protected sensor. Several properties of this directed toward determining the most efficient sense
transfer function are examined, including frequency line to use for the measure of pressure pulses from the
dependence, Helmholtz resonance, and time of flight engine. This is not a trivial task because the sense line
delay. itself has a unique transfer function that will affect
some frequencies more than others. The goal was to
determine if hollow tubes, liquid filled tubes, or gel
filled tubes would best transmit the actual acoustic
Data collection and analysis of rocket engine tests are spectra. A substantial amount of this work was done in
critical steps in understanding the physics of propulsion the laboratory before the engine tests were done.
devices that may be used for space exploration. Current Classical linear system concepts were utilized in
sensor technology must be exploited fully and understanding how the sense lines affected the signal
expanded to cover all the classical parameters such as detected by the pressure transducer.
temperature, pressure, and flow rate. Methods must be
established to infer such parameters as turbulence and A second task involved designing an experiment to
shock that cannot be measured directly. This report will detect off-axis thrust from an engine mounted in a test
describe several experiments that were performed at the stand. A test of the 24 inch SPTA engine provided a
Marshall Space Flight Center, East Test Area, on two platform for a simple optical technique that could at
different engines. One of the engines was a solid rocket least detect an off-axis (torque) component. The
motor and the other was a liquid fuel engine. Neither experiment was quite simple and involved mounting a
engine was dedicated to the experiments reported on mirror on the mounting ring that the engine was hard
here, but rather the experiments were added so as to not mounted to. A green solid state laser pointer was
interrupt scheduled tests. obtained and used as the source. The laser was mounted
a safe distance away from the motor and the beam
The research is centered on gathering and analyzing reflected from the mirror back to a screen. A digital
acoustic and vibration data. There were two separate video camera was then used to record the motion of the
but related reasons for collecting the data. The acoustic green laser spot on the screen. The eventual goal is to
experiments were done in preparation for testing a use several lasers and mirrors and screens so that a true
much larger solid rocket motor later in the year and an three-dimensional representation of the off-axis thrust
engine being developed for the NASA Orbital Space can be created. The goal of this first attempt was to
Plane (OSP). The testing done over the last few weeks discover the problems and solutions for the method so
was done on a small experimental engine called the that the technique would be ready for the OSP tests
LLT. It has a throat diameter of about 1.5 inches and is planned for later this year.
made substantially of a laminated material. It is a liquid
oxygen/ liquid hydrogen engine. The primary sensors The law of reflection can be used to create a device that
used are model 102A10 manufactured by PCB accurately measures small changes in angle*. A low-
Corporation, and have a sensitivity of 50 mV/psi and a power visible range laser is aimed at a small mirror
maximum pressure measurement of lo00 psi. Other mounted on the test article. The beam is reflected and
sensors used are manufactured by Enderco, model hits a large screen; a video camera is mounted facing
85108, having a sensitivity of 158.4 mV/psi (millivolts the screen in order to record the path taken by the laser
spot. Both the laser and the camera can be mounted transmitted through acoustic sense lines, Helmholtz
hundreds of feet from the test article; only a small resonance was clearly observed at frequencies
mirror need enter the test area. corresponding to the lengths of the sense lines.

For Phase I of this experiment, a green laser pointer The resonance condition for a tube opened at one end is
was used (5 milliwatts, peak wavelength 532 V
nanometers). The mirror was a first surface gimbaled (2n + 1)- =f ,
laboratory mirror mounted onto the thrust structure with 4L
epoxy. With an effective dispersion of 0.4 milliradians,
the laser spot fit nicely onto the two-inch wide mirror'. where n is an integer index ranging upwards from zero,
The mirror was aligned so as to reflect the spot towards fis the frequency of the sound, v is the speed of sound
a screen staked into the ground in the shade of some in the resonating tube, and L is the length of the tube".
pine trees 250 feet from the test engine. The recording
device was a digital video camera recording at 30 Time of flight analysis depends on the assumption that
frames per second, outfitted with an inexpensive green the transit time for each of the tubes is the same; the
filter. offset in signal onset from one sensor to the next is
solely a factor of the position of the open end of the
MOTIVATION sense line. If this is the case, then it is expected that
v=- Ar
At? (2)
Acoustic Sense Lines
where Ar is the difference in distance from the nozzle
The acoustic phenomena of interest in propulsion for any two sensors, and At is the time delay between
applications often occur in harsh environments where the two sensorstt. It is possible to calculate a mean
heat, high-velocity particulate matter, and volatile value for v by considering the linear fit for the various
chemicals preclude the use of sensitive and expensive data points plotted with time on the horizontal axis, and
sensors. To avoid the compromise between safe radial distance on the vertical axis.
distance and useful data range, a sense line may be used
to passively transmit information as a conduit. The
sense line should either be durable enough to withstand ODtical Off-Axis Thrust Measurements
such conditions, or else inexpensive so that it can be
replaced periodically. Axial thrust structures measure displacement along the
principle axis and infer thrust on this axis. Most
The variation in parameters at one end of the tube will configurations measure off-axis components of thrust
uniquely determine the condition at the closed end of poorly, or not at all. A non-invasive, inexpensive
the tube, where the sensor is to be located. The sense technique was desired for measuring these off-axis
line must have dimensions small enough that it does not components of thrust.
affect the parameters of the system it measures. It
should be emphasized that a sense line does not A system was developed in which a laser beam is
automatically transmit temperature, pressure, and other incident on a mirror mounted on the test stand. The
conditions from one end of the tube to the other. laser and mirror were positioned such that the reflected
Rather, every measurement taken through a sense line beam struck a Lambertian screen positioned several
must be analyzed in terms of temporal shift and transfer hundred feet away. Geometry insures that a small
function losses. In addition, any component of the change in orientation angle
system that alters data introduces potential error into the will lead to large displacements of the order
measurement, and sense lines are no exception.
Transfer functions for fluid-filled sense lines fall to zero Ax=DAB, (3)
for high frequencies as this vibrational energy is
absorbed by the fluid; any information about these high where Ax is the displacement on the screen, and A0 is
frequency components is lost. The time delay due to the change in orientation angle of the mirror. A more
the length of the sense line is a variable quantity that detailed analysis takes into account the incident angle
can depend on conditions (e.g., temperature) inside the of the laser beam, the orientation of the screen with
sense line. respect to the laser and other geometric factors to
produce a transcendental equation for Ax in terms of
Acoustic resonance is a phenomenon via which certain AO.
vibrational frequencies, or modes, are amplified by the
resonating structure. In the analysis of the pressure data
RESULTS None of the transfer functions are line r in the ense
that no sense line power spectrum differs from the
Acoustic Sense Line Transfer Functions freestanding transducer by a multiplicative constant.
All of the sense lines effectively damped out the 60
Various fluids were explored for use in sense lines. A Hertz (and integer multiple) noise indicated by the
fluid was desired that would transmit much of the sharp, low-frequency spikes in the freestanding
energy to the transducer with a known transfer function. transducer.
A laboratory experiment was designed to compare short
sense lines filled with water, agarose, and air to a
control transducer used to measure the unaltered shock. Acoustic Sense Line Time of Flight Results
Agarose is a polysaccharide obtained from agar. The
preparation used in this experiment is a suspension of The propagation rate of shock wave can be determined
agarose in a boric acid buffer with a gelatinous by comparing the onset times for each of the sensors.
consistency". The onset of the shock due to the combustion event
appears as a transient spike followed by a rapid rise to a
much higher plateau on each of the channels. As all
three sensors were separated from the event by identical
lengths of Tygon tubing, the time delay can be
attributed to the time of flight of the acoustic wave
travelling through the air. A linear fit to these three
data points indicated a propagation speed of 1100 feet
per second, comparing very favorably to the reference
value for acoustic waves, 1086 feet per second at STP.
A more extensive array of sensors could detect changes
in propagation speed, including regions where the
Figure 1: Power spectra of (clockwise from top left): propagation was supersonic, as well as directional
the freestanding transducer, and transducers in sense difference in propagation speeds.
lines filled with water, air, and an agarose suspension.
A qualitative comparison of the signals at each sensor
The general shape of the transfer functions of the sense indicates a rather rapid decrease in total energy as
lines can be obtained by taking the ratio of each of the distance increase, and a rapid decrease as the sensors
power spectra to the power spectrum of the freestanding are placed further from the axis. A non-directional
transducer. This ratio is not the true transfer function, ( l/r2) distribution would be independent of direction,
since the power spectra are the squares of Fourier whereas a collimated distribution would not show such
transforms and not the true Fourier transforms, but the a rapid falloff with distance". A more extensive array
general characteristics of the transfer functions should could determine the shape of the advancing profile by
be apparent from such an exercise. integrating power spectrum of the signal to determine
the energy received at each point.
Both water and agarose demonstrate a rough correlation
with the freestanding transducer at higher frequencies, Acoustic Sense Line Helmholtz Resonance
in that both mimic the falloff in power at loo0 Hz. But
these two sense lines also demonstrate high gain in the Helmholtz resonance was clearly observed in the
lower frequencies that correspond to Helmholtz acoustic wave tests. By plotting spectral intensity on a
resonance. This resonance can be seen much more time-frequency plane, broad bands could be observed at
clearly in the air-filled sense line, where the original regular intervals during the combustion event. Each of
profile is lost beneath a sharp resonance spike. This these events was indexed with an odd integer
resonance is analyzed in more detail in the field tests corresponding to a harmonic for a closed tube, and the
below. value for v, was used to solve the Helmholtz resonator
equation for the length of the tube. A value of 9.5 feet
was obtained, very close to the actual length of the tube.
 Thrust and acoustic measurement procedures were
proposed and given initial tests with two rocket motors,
one solid and one liquid. The liquid tests showed the
validity of using sense lines, and the solid test showed
the validity of optical techniques in measuring the off
axis torque of engines in test stands. Analysis of the
liquid tests confirmed the presence of Helmholtz
frequencies, and the time of flight analysis was
consistent with a shock wave propagating at sonic

- m u
1 ,
2-
speeds. Other measurements are planned for engines
later in the year.

*
t
'Hecht, E. Optics, Addison Wesley, Second Edition.
1987. (83)
* Ibid., (422)
$1
Halliday, D. et al. Fundamentals of Physics, John
Wiley and Sons, Fifth Edition. 1997. (436)
Ibid., (55)
" O'Neil, M. (editor) Merck Index: An Encyclopedia
of Chemicals, Drugs,& Biologicals, Merck Publishing
."., Group. 2001. (13, 184)
Figure 2: Time-domain data for time of flight '*Griffiths, David. Introduction to Electrodynamics,
measurement; time-frequency plots showing Helmholtz Prentice Hall, Third Edition. 1998. (31)
friquenc ies

ODtical Off-Axis Thrust Measurements

Phase I demonstrated that the recording device could

easily detect the full-intensity spot, and in separate
experiments it was shown that small changes in angle
resulted in measurable shifts in spot location. It was
also proven in the laboratory that the formulae
developed for calculating change in angle from these
shifts was accurate. However, the logistics of test firing
resulted in the laser not operating at full strength during
the firing of the motor, and as a result the spot was not
visible during that time. It will be necessary in future
tests to outfit the laser with a remote switch so that it
can be turned on immediately prior to the firing.

Figure 3: Laser spot on the target screen (left), and an

enhanced view of the same image.