N96- 15595

QUANTITATIVE MEASUREMENT OF OXYGEN IN MICROGRAVITY

COMBUSTION

Joel A. Silver
Southwest Sciences, Inc.
Santa Fe, New Mexico

Introduction

A low-gravity environment, in space or in ground-based facilities such as drop towers, provides

a unique setting for studying combustion mechanisms. Understanding the physical phenomena
controlling the ignition and spread of flames in microgravity has importance for space safety as
well as for better characterization of dynamical and chemical combustion processes which are
normally masked by buoyancy and other gravity-related effects. Due to restrictions associated
with performing measurements in reduced gravity, diagnostic methods which have been applied
to microgravity combustion studies have generally been limited to capture of flame emissions on
film or video, laser Schlieren imaging and (intrusive) temperature measurements using
thermocouples. Given the development of detailed theoretical models, more sophisticated
diagnostic methods are needed to provide the kind of quantitative data necessary to characterize
the properties of microgravity combustion processes as well as provide accurate feedback to
improve the predictive capabilities of the models. When the demands of space flight are
considered, the need for improved diagnostic systems which are rugged, compact, reliable, and
operate at low power becomes apparent.

This research builds on our earlier work I and combines two innovations in an experimental
system which should result in a new capability for quantitative, nonintrusive measurement of
major combustion species. Using a newly available vertical cavity surface-emitting diode laser
(VCSEL) and an improved spatial manning method, we plan to measure the temporal and spatial
profiles of the concentrations and temperatures of molecular oxygen in a candle flame and in a
solid fuel (cellulose sheet) system. The required sensitivity for detecting oxygen is achieved by
the use of high frequency wavelength modulation spectroscopy (WMS). 2"3 Measurements will
be performed in the NASA Lewis 2.2-second Drop Tower Facility.

The objective of this research is twofold. First, we want to develop a better understanding of the
relative roles of diffusion and reaction of oxygen in microgravity combustion. As the primary
oxidizer species, oxygen plays a major role in controlling the observed properties of flames,

Program Starting Date: 28 June 1994.

279
including flame front speed (in solid or liquid flames), extinguishment characteristics, flame size
and flame temperature. The second objective is to develop better diagnostics based on diode laser
absorption which can be of real value in microgravity combustion research. We will also
demonstrate diode lasers' potential usefulness for compact, intrinsically-safe monitoring sensors
aboard spacecraft. Such sensors could be used to monitor any of the major cabin gases as well
as important pollutants.

Experimental Approach

A schematic of the experimental setup is shown in Fig. 1. The output of the diode laser is
focussed onto an optical scanning mirror positioned at the focus of an off-axis paraboloidal
(OAP) reflector. Since all light rays emanating from this point are reflected along parallel paths
by the OAP, this results in a collimated laser beam (-1 mm dia.) which linearly traverses the
flame region as the mirror is scanned. A second OAP refocuses the beam onto a single detector.
The mirror pair is configured so that the total optical path at any transverse position is nearly
constant.

This arrangement permits a temporal map of concentration and temperature of 0 2 to be obtained

during the microgravity experiment. In a typical run, the mirror is dithered back and forth at a
rate of 20 Hz, so that the laser samples a 4 cm wide region encompassing the flame. As the
mirror is moved, an absorption spectrum is recorded at 1 mm spatial resolution. The laser
wavelength sweeps across two adjacent absorption lines. From these data, the temperature and
concentration of 0 2 can be recovered for each spatial element as a function of drop time. For
the candle flame experiments, where the profiles of interest are radially distributed from the flame
center, Abel inversion techniques are used to convert the measured projection spectra to radial
profiles for selected heights in the flame. 1,4

In our prior work, 1 an eight line-of-sight fiber optic system measured water vapor mole fractions
in the NASA Lewis 2.2-see Drop Tower. The electronic modulation and detection circuitry
developed in that program will be used here. The spatial scanning mechanism now used has the
advantage that much higher spatial resolution is achieved with a concurrent simplification in
complexity and size of the detection electronics. A single detection channel sequentially records
all spatial information, which avoids problems associated with multiple detection channels
(relative calibrations, drift, cost, etc.) A commercial state-of-the-art digital signal processing
(DSP) data acquisition board is used to generate both the mirror dithering and laser ramp
waveforms, as well as acquire, pre-process, and store all the data. This board will be controlled
by a stand-alone 486 computer board. All of the electronics (including the computer) will be
housed in a small enclosure which will be mounted on the drop rig.

This program utilizes a new type of diode laser. Vertical cavity surface-emitting lasers, which
are now available at selected wavelengths below 1 #m, are true single-mode devices. Unlike the
Fabry-Perot GaA1As lasers typically used in this spectral region, VCSEL lasers can be scanned
without mode-hopping over a relatively large wavelength range (3 nm). We have obtained
custom VCSEL lasers at 760.3 nm, chosen to access a pair of 0 2 rotational lines in the blE-x3E
electronic transition.

280
Wavelength Modulation Spectroscopy

In traditional absorption measurements, the spectral lineshape is obtained by subtracting the signal
transmitted through the absorbing medium from a reference (no absorber) spectrum. The factor
limiting detection sensitivity is source noise, and limits of 10 .3 fractional absorption are common.
High frequency wavelength modulation absorption spectroscopy allows the measurement of much
weaker optical absorbances by shifting the detection band to high frequencies where excess laser
source noise becomes unimportant. For WMS, a small sinusoidal modulation is superimposed
on the diode laser injection current. This current modulation produces a modulation of the laser
wavelength. Typically, the amplitude of the current modulation is chosen so that the induced
wavelength modulation is comparable to the width of the spectral feature under study. Phase-
sensitive electronics are then used to detect the signal at the second harmonic of this frequency.

Diode laser intensity noise power per unit frequency has been found to decrease rapidly at higher
frequencies. Therefore, improved sensitivity is attained by modulating the laser wavelength at
megahertz frequencies. In this way we have achieved fractional absorption sensitivities at the
near shot-noise limit (-10 "7) for both near- and mid-IR diode lasers. 3 For the greatest sensitivity,
WMS (or lower frequency harmonic detection) has usually been implemented using a "sweep and
fit" approach. The laser is modulated and scanned in wavelength across the absorption feature;
the entire 2f WMS spectrum is recorded. By numerical fitting techniques (multilinear regression,
matched filtering, etc.) the measured 2f lineshape is compared to a calibrated value so as to
determine the absolute sample concentration.

Given the flexibility of the data acquisition system, both in-phase and quadrature components of
the phase-sensitive detector output will be recorded. Thus the detection phase does not have to
be preset and can be optimized during data analysis. Also, the de photocurrent will be acquired
as well. This signal is proportional to the laser intensity (I0) and is used to account for all
variations in alignment, beam attenuation, etc.

Measurement of Oxygen Mole Fraction and Temperature

The absolute absorber concentration N is related to the WMS signal by

Signal = C'ot'fwMs, where a = oN/, (1)

and C is a system electronic calibration constant, fWMS is a factor relating the WMS signal to the
direct absorption peak value (approximately 0.6), ct is the absorbance, o the absorption cross
section and l is the absorption path length. Given the system temperature and pressure, the
absolute concentration is linearly proportional to the measured signal and is readily obtained.

The optimum spectral lines for use in simultaneously measuring temperature and concentration
(or mole fraction) of 0 2 have been determined. Our goal is to use two closely spaced, but not
overlapping, lines where the composite numerical factor which is used to convert signal to mole
fraction should be very sensitive to temperature for one of the lines, and insensitive for the other,
over the flame temperature range of interest. In this manner we can use a single line to get mole

281
fraction without accurately knowing the temperature profile, or we can use the signal peak ratios
to obtain temperature. 5 We have identified four pairs of lines which exhibit these characteristics.
These pairs lie in the 759 to 761 nm wavelength range. The best pair shows a temperature
sensitivity of 47% ratio change per 100 K in the range 1300-1800 K. The best temperature
independent line in this region has a constant conversion factor (+5%) between 1000 and 1650
K. Since we should be able to measure signal height ratios to better than 5%, the expected
temperature accuracy (assuming we have an accurate knowledge of all spectral parameters) is
±25 K.

Results

At this point in the program, we have tested a benchtop layout of the optics. It appears that the
dither scanner can readily move a collimated beam of diameter -1 mm back and forth across the
required 40 mm range at 20 Hz. The laser power on the detector was fiat to better than 2%
during these scans (although any variations will be accounted by measurement of Io). The DSP
board has been programmed to generate both laser and scanning waveforms and can acquire the
2f spectra. We have received and are now characterizing the VCSEL lasers. We plan o_gm
to '---:-
preliminary benchtop candle flame measurement shortly.

References

1. J.A. Silver, D.J. Kane and P.S. Greenberg, "Quantitative Species Measurements in
Microgravity Flames With Near-IR Diode Lasers, Appl. Opt. (in press, 1995).
2. J. Silver, "Frequency Modulation Spectroscopy for Trace Species Detection: Theory and
Comparison Among Experimental Methods," Appl. Opt. 31, 707-717 (1992).
3. D.S. Bomse, A.C. Stanton and J.A. Silver, "Frequency Modulation and Wavelength
Modulation Spectroscopies: Comparison of Experimental Methods Using a Lead-Salt Diode
Laser," Appl. Opt. 31, 718-731 (1992).
4. C.J. Dasch, "One-Dimensional Tomography: A Comparison of Abel, Onion-Peeling, and
Filtered Backprojection Methods," Appl. Opt. 31, 1146-1152 (1992).
5. M.P. Arroyo, S. Langlois, and R.K. Hanson, "Diode-Laser Absorption Technique for
Simultaneous Measurements of Multiple Gasdynamic Parameters in High-Speed Flows
Containing Water Vapor," Appl. Opt. 33, 3296-3307 (1994).

282
 Detector

Laser _,.

Scanning Mirror

Figure 1 - Schematic of Optical Setup.

283