EPA-600/2-76-247a
September 1976
Environmental Protection Technology Series
INFLUENCE OF AERODYNAMIC PHENOMENA ON
POLLUTANT FORMATION IN COMBUSTION
(Phase! Gaseous Fuels!
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/2-76-247a
September 1976
INFLUENCE OF AERODYNAMIC PHENOMENA
ON POLLUTANT FORMATION
IN COMBUSTION
(Phase I. Gaseous Fuels)
by
Louis J. Spadaccini, F. Kevin Owen, and Craig T. Bowman
United Technology Research Center
400 Main Street
East Hartford, Connecticut 06108
Contract No. 68-02-1873
ROAP No. 21BCC-014
Program Element No. 1AB014
EPA Project Officer: W. Steven Lanier
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
TABLE OF CONTENTS
Page
ABSTRACT v
LIST OF FIGURES vi
LIST OF TABLES xi
ACKNOWLEDGMENTS xi
SECTION I - INTRODUCTION 1
SECTION II - EXPERIMENTAL APPARATUS AND
INSTRUMENTATION 3
Combustor Facility .... 3
Probes 7
Sampling System 11
Laser Velocimeter 17
SECTION III - EXPERIMENTAL RESULTS 25
Description of the Experiments 25
Test Matrix 26
Input-Output Test Results 28
Flow Field Mapping Results 31
Fuel Injector Probing 73
SECTION IV - DISCUSSION OF RESULTS 75
SECTION V - RECOMMENDATIONS 79
APPENDIX A - LASER VELOCIMETER STATISTICAL ERRORS AND BIASING. 82
APPENDIX B - FUEL COMPOSITION 87
iii
-------
TABLE OF CONTENTS (Cont'd.)
APPENDIX C - COMBUSTOR HEAT BALANCE 88
APPENDIX D - TEMPERATURE DISTRIBUTIONS: TABULATED DATA ... 89
APPENDIX E - SPECIES CONCENTRATION DISTRIBUTIONS:
TABULATED DATA 95
APPENDIX F - MEANS AND RMS VELOCITY DISTRIBUTIONS:
TABULATED DATA 10?
PUBLICATIONS 133
REFERENCES 13l+
NOMENCLATURE 137
IV
-------
ABSTRACT
An experimental investigation of the effects of the interaction
between fluid dynamics and chemistry on pollutant formation and de-
struction in a natural gas fired turbulent diffusion flame burner has
been carried out. In this investigation, the effects of inlet air
swirl, combustor pressure and air/fuel velocity ratio on the time-mean
and fluctuating flow field have been determined using probing and op-
tical techniques, and the changes in flow field structure have been
correlated with changes in pollutant emissions from the burner. The
results of this investigation show that variation of these inlet param-
eters produces major changes in the time-mean flow field within the
burner which significantly influence pollutant formation. In addition,
it was found that there are substantial large-scale contributions to
the total rms turbulent velocity field. These large-scale fluctuations
results in significant departures from Gaussian turbulence and isotropy
in the initial mixing regions of the burner and have pronounced effects
on mixing, chemical reaction and pollutant formation.
This report was submitted in partial fulfillment of Contract 68-02-
1873 by United Technologies Research Center under the sponsorship of the
Environmental Protection Agency. Work was completed as of April 30, 1976.
v
-------
LIST OF FIGURES
Figure
M Page
No. a
1 Schematic diagram of axisymmetric combustion facility 4
2 Injector and swirl vane geometries 6
3 Photograph of combustion facility °
4 Exhaust sampling probe rake 9
5 Traversing gas sampling probe 10
6 Calibrated-conduction-loss thermocouple probe 12
7 Uncooled five-hole hemispherical pitot probe 13
8 Schematic diagram of on-line gas analysis system 14
9 Exhaust gas analytical system 16
10 Comparison of pitot probe and laser velocimeter 18
measurements of velocity in a swirling (S=0.3)
atmospheric-pressure natural gas-air flame
11 Schematic diagram of laser velocimeter 20
12 Schematic diagram of laser velocimeter data 21
processing equipment
13 Error due to directional ambiguity 24
14 High-speed motion picture (500 frames/sec) of flame near 32
injection plane--S=0, P=3-8 atm, Va/Vf=21
VI
-------
LIST OF FIGURES (Continued)
Figures
No.
15 Frequency spectra of transient pressure fluctuations 35
16 Time -averaged temperature distributions 37
17 Time-averaged species mole fraction distributions S=0, ^°
B=3.8 atm, Va/Vf=21
18 Time-averaged species mole fraction distributions- -3=0. 3 » ^
B=3.8 atm, Va/Vf=21
19 Time -averaged species mole fraction distributions S=0.3, ^
B=1.0 atm, Vg/V^ai
20 Time-averaged species mole fraction distributions S=0.6, ^
B=1.0 atm, Vg/Vf.^21
21 Time-averaged species mole fraction distributions--S=0.6, ^9
B=1.0 atm, Va/Vf=0.2
22 Time-averaged nitric oxide and nitrogen dioxide mole frac- 52
tion distributions--S=0.6, B=1.0 atm,
23 Mean axial velocity profiles--S=0.3> B=l atm, Va/Vf=21 53
2k Mean axial velocity distributions 5^
25 Mean tangential velocity profiles 57
26 Mean and rms tangential velocity distributions 58
27 Mean radial velocity profiles- -3=0. 3, E=l atm, Va/Vf=21 6l
28 Axial rms velocity distributions °2
29 Axial mean and rms velocity and directional intermittency 61+
profiles--S=0.3, B=3.8 atm,
VI1
-------
LIST OF FIGURES (Continued)
Figures
No.
30 Tangential mean and rms velocity and directional intermittency 66
profiles--S=0.3, B==3-8 atm, Va/Vf=21
31 Probability distribution functions of axial velocity 3=0.3, 67
P=3.8 atm, Va/Vf=21
32 Axial directional intermittency distributions 68
33 Probability distribution functions for shear stress measure- 72
ment -3=0.3, P=3-8 atm, Va/Vf=21
3^- Axial- tangential velocity cross correlations at X/D=0.05m-- 72
S=0.3, P=3-8 atm, Va/Vf=21
35 Mean axial velocity distribution within fuel injector 7)4.
number I
36 Comparison of mean axial velocities measured using 85
different seeding techniques--S=0.3, P=l atm, Va/Vf=21,
§=0.5
37 Comparison of probability distribution functions measured 86
using different seeding techniques- -S=0.3, B=l atm,
Va/Vf=21, 3=0.5
VI11
-------
LIST OF TABLES
Table
1 Test Matrix 27
2 Exhaust Species Concentrations 29
3 Summary of Transient Pressure Measurements 3)4
B-l Natural Gas Composition 87
D-l Temperature Distributions for Test No. 1 90
Di-2 Temperature Distributions for Test No. 3 9!
D-3 Temperature Distributions for Test No. k 92
T>-k Temperature Distributions for Test No. 6 93
D-5 Temperature Distributions for Test No. 7 9*1
'E-l Species Concentration Distributions for Test No. 1 96
E-2 Species Concentration Distributions for Test No. 3 98
E-3 Species Concentration Distributions for Test No. h
E-l). Species Concentration Distributions for Test No. 6
E_5 Species Concentration Distributions for Test No. 7 105
F-l Axial Mean and RMS Velocity Distributions for Test No. 1 108
F-2 Axial Mean and EMS Velocity Distributions for Test No. 3 111
F-3 Axial Mean and RMS Velocity Distributions for Test No. k 115
IX
-------
LIST OF TABLES (Cont'd)
Table
F-U Axial Mean and RMS Velocity Distributions for Test No. 6 118
F-5 Axial Mean and EMS Velocity Distributions for Test No. 7 121
F-6 Tangential Mean and EMS Velocity Distributions for Test No. 3 123
F-7 Tangential Mean and RMS Velocity Distributions for Test No. k 125
F-8 Tangential Mean and EMS Velocity Distributions for Test No. 6 127
F-9 Tangential Mean and EMS Velocity Distributions for Test No. 7 130
F-10 Radial Velocity Distribution for Test No. 3 132
-------
ACKNOWLEDGMENTS
A number of individuals at UTRC made significant contributions to
the experimental investigation. Dr. M. F. Zabielski and Mr. G. L. Dodge
designed the sampling system used in the investigation and developed
the calibration procedures employed in the gas sampling portion of the
experiments. Mr. T. A. Murrin assisted throughout the experimental pro-
gram and was responsible for operation of the combustor and for reduc-
tion of much of the experimental data. The high-speed motion pictures
of the reacting flows were made by Mr. R. J. Haas. Mrs. P. A. Rose and
Mrs. B. B. Johnson assisted in reduction and compilation of the experi-
mental data and in the preparation of the final report.
This research program was carried out under the sponsorship of the
Environmental Protection Agency, EPA Contract 68-02-1873, Research
Triangle Park, North Carolina, under the direction of Mr. W. S. Lanier,
Project Officer.
XI
-------
SECTION I
INTRODUCTION
Recent investigations of factors affecting pollutant emissions from
furnaces (Refs. 1-U) and gas turbines (Refs. 5,6) indicate that changes
in operating conditions, which alter mean flow patterns in the combustion
chamber, can have a substantial effect on pollutant formation and des-
truction. Experiments carried out on a laboratory-scale turbulent dif-
fusion flame burner (Ref. 7) under EPA Contract 68-02-1092 confirmed
these observations and demonstrated that the interaction between fluid
dynamics and chemistry is a major factor governing pollutant emissions.
At the present time, our understanding of the nature of this coupling
is insufficient to permit quantitative prediction of the effects of
changes in operating conditions on pollutant emissions. Analytical
studies of turbulent reacting flows (Refs. 8-12) have provided some
insight into the effects of mixing and turbulence on flow field struc-
ture and pollutant formation. However, it is uncertain whether
existing analytical models can provide accurate descriptions of turbu-
lent reacting flows of practical interest. Furthermore, it is difficult
to assess the limitations of these models because of inadequate local
flow field measurements in practical combustor geometries to serve as
test cases and because of a lack of information on the turbulent struc-
ture of reacting flows which can be used to assess the validity of pres-
ent turbulence models.
In a previous study (Ref. 7), fluid dynamic and chemical phenomena
in the regions near the injection plane were found to influence flame
stabilization, energy release and pollutant formation. If these signi-
ficant phenomena and their interaction with the pollutant formation
process are to be understood, then detailed information on the velocity,
temperature and species concentrations in the near-injector region is
necessary. Of particular importance are measurements of the location
and size of the flame-stabilizing recirculation zones and characteriza-
tion of the turbulent structure of the flow. To permit meaningful com-
parison of the experimental data with predictions of combustor flow
-------
model, careful determination of the combustor inlet conditions is
required.
The present report documents the results of an experimental inves-
tigation, sponsored by EPA Contract 68-02-l8?3, of the effects of
several operating parameters on the flow field structure near the injec-
tion plane in a turbulent diffusion flame burner and the subsequent
effects on pollutant formation and destruction. The investigation is
a logical extension of the previous contract effort (Ref. 7) in that it
addresses many of the questions outlined above and provides an expanded
data base on the effect of combustor inlet conditions on flow field
structure and pollutant formation. The principal objectives of the pro-
gram were (l) to obtain detailed maps of the combustor flow field,
including recirculation zones, as operating conditions were varied and
(2) to correlate changes in flow field structure with changes in pollu-
tant formation and energy release. The results will be used to evaluate
a combustor flow analysis being developed in the theoretical portion of
this program. Results from the analytical study will be documented in
a subsequent report (Ref. 12).
-------
SECTION II
EXPERIMENTAL APPARATUS AND INSTRUMENTATION
COMBUSTOR FACILITY
The experimental configuration and approach utilized in the present
investigation are similar to those employed in the previous effort
carried out under EPAContract 68-02-1092 (Ref. 7). Tests were conducted
in the instrumented, water-cooled combustion system shown schematically
in Fig. 1. The facility design was modified slightly from that employed
in the previous contract effort (Ref. 7). Particular emphasis was placed
on acquisition of species concentration, temperature and velocity data
throughout the initial regions of the reacting flow for comparison with
results obtained in the analytical study (Ref. 12).
Air from a 30-atm supply, at flow rates up to 0.65 kg/sec, may be
heated in an electrical heater section to provide inlet air temperature
up to 1000°K. Within the heater, the air flows through and around
four 6 m long stainless steel tubes which may be supplied with as much
as 720 kW of electrical power. The heated air enters the combustor
through a circular annulus formed by a replaceable axisymmetric fuel
injector and a 12.23 cm diameter entry section. Natural gas (~ 96 per-
cent CHi ) fuel, introduced through three (air foil shaped) struts into
the center deliver duct, is brought into contact with the annular air-
stream at the exit of the injector. Thereafter, mixing and chemical
reaction proceed at constant area in the remainder of the injection sec-
tion and into the instrumented combustor and extender sections. Flame
stabilization in the high velocity flows investigated was achieved by
producing a recirculation zone(s) in the initial region of the combustor
by imparting a swirl component to the air flow and/or by reducing the
fuel/air velocity ratio. For the present investigation, porous-metal
discs installed in the fuel injector and air entry sections serve to
provide uniform inlet flows. (The uniformity of the inlet flow was
verified by laser velocimeter and pitot probe measurements.) In order
to impart swirl to the airflow, straight swirl vanes are inserted into
the annular passage of the injector. In the previous contract effort,
-------
FIG. 1
SCHEMATIC DIAGRAM OF AXISYMMETRIC COMBUSTION FACILITY
MOVABLE INJECTOR
EXHAUST
SAMPLING
PROBE
REPLACEABLE
SWIRL VANES
ORIFICE
HEATER
76-05-198-1
-------
these swirl vanes were located at the exit of the injector section. In
the present study, the swirl vanes are located upstream of the injector
exit plane to permit measurement of the characteristics of the airflow
entering the combust or. These measurements together with measurements
made within the fuel injector should provide the inlet conditions needed
in the analytical modeling effort (Ref. 12). The two injectors utilized
in the study are described in Fig. 2 in terms of the ratio of the inner
and outer diameters of the air annulus, Z = dh/d, and the nominal air/
fuel velocity ratio, m = Vg/Vf, associated with air and natural gas
coaxial jets having a nominal overall fuel/air equivalence ratio, $, of
0.9. With the exception of the porous plug insert, these injectors
are identical to those used in the previous contract effort (Ref. 7).
Swirl vane designs are identical to those employed in the previous effort
and are shown in Fig. 2, where the swirl number, S, has been computed
from the injector geometry, Z, and the angle of the swirl vanes, T],
according to the following expression (Ref. 13):
The swirl number is simply the ratio of the angular momentum flux to the
axial momentum flux multipled by an effective nozzle diameter. A swirl
number of 0.3 connotes relatively low swirl, while S = 0.6 results in a
moderately high swirl situation. A practical upper limit of S = 0.8
exists for straight blades from the standpoint of packaging the vanes.
Observation of the combusting flow may be made through the 6.U-cm
diameter quartz window ports in the combustor section (Fig. 1). A pair
of window ports 180 deg apart are present at each location and permit the
use of optical measurement techniques (e.g., laser velocimetry and laser
holography). The location of a port directly downstream of the injector
exit plane allows an unhindered view of the flame in the vicinity of
the fuel delivery duct and permits acquistion of flow field data close
to the injector exit. The combustor probing devices used to make temper-
ature and species concentration measurements are compatible with all win-
dow ports and may replace a window or water-cooled plug in any given
port. In addition, the entry section was redesigned to permit axial
relocation of the fuel injector between tests, thereby greatly increasing
the number of axial locations at which radial traverses can be made.
The 12.23 cm diameter, 100 cm long instrumented combustor is
divided into five water-cooled zones of approximately equal length.
Water flow can be set independently in each zone, as needed, to keep
-------
INJECTOR AND SWIRL VANE GEOMETRIES
d0 = 12.23 CM
FLOW
11 I 111 I 11 11 11111
1.43 CM
INJECTOR STEP
VANE 0.163 CM
THICK 316 SS
INJECTOR
DESIGNATION
m, VELOCITY
RATIO*
df (CM)
Z = dh/d
L(CM)
S
7? (DEC)
NO. OF
VANES
X (CM)
21:1
6.314
0.677
1.499
0.3
0.6
28
47
18
12
7.37
I
01
00
0.2:1
0.757
0.084
4.290
0.6
60
BASED ON NATURAL GAS AND * = 0.9
1.59
Tl
-------
wall temperature (~ 500°K) roughly constant along the entire length of
the combustor. Wall temperatures are set and monitored using thermo-
couples installed on the outer surface and at various depths in the com-
bustor wall and cooling passages. Static pressure taps are also
installed at several locations along the combustor. Flow exhausts from
the combustor and extender sections to the facility exhaust stack. Com-
bustor extender pieces, 33.^ cm in length, are inserted when required
to fully contain the flame; the extender section consisted of two
extender pieces during all of the current experimental effort. Water-
cooled orifices can be installed downstream of the extender section to
raise the pressure in the combustor. A photograph of the combustion
facility is given in Fig. 3.
PROBES
Species concentration distributions within the combustor were
measured using a traversing gas sampling probe and an exhaust gas
sampling rake. Composition information is determined on-line by aspir-
ating flow through the cooled probes and analyzing the gas sample using
a Scott Model 119 Exhaust Gas Analyzer. Pressurized hot water at UOO°K
was used as the probe coolant to minimize wall-catalyzed reactions and
to prevent water condensation and loss of species within the sampling
lines.
The exhaust probe rake, located at the exit of the extender section,
consists of five identical probes centered on equal area annuli (Fig. k).
The individual probes are manifolded downstream and a single mixed sam-
ple is transferred to the gas analyzer. Radial traverses are made at
selected axial locations within the instrumented combustor section using
a single gas sampling probe of similar design (Fig. 5). The inlet flow
into both sampling probes was maintain choked, resulting in aerodynamic
cooling of the sample by means of a rapid internal expansion. This
expansion combined with the wall cooling effect served to quench chemical
reactions involving stable species. Errors associated with sampling
probe measurements in turbulent flames are discussed in Refs. Ik and 15.
Temperature profiles at the exhaust plane and within the combustor
were measured by traversing a calibrated-heat-loss thermocouple probe
across a combustor diameter. Although conventional thermocouple materials
limit application of these sensors to temperatures below about 2000°K,
cooling the exposed junction by conduction heat transfer extends the
range of thermocouple utilization above the melting point of the material
to the 2000-2500°K range. In order to obtain the local stream tempera-
ture, the measured stream thermocouple temperature must be corrected for
-------
COMBUSTOR FACILITY
OJ
t J .. ..-
. ~
-( .
Z TI
'" G)
I w
:::
I
'"
-------
'.0
EXHAUST SAMPLING PROBE RAKE
. PROBES CENTERED ON EQUAL AREAS
. FLOW FROM ALL PROBES MIXED BEFORE
ON-LINE ANALYSIS
/
{~ £~I
~~
""~
-..J
m
I
o
Ul
I
-..J
<1)
I
~
"It:
/
FLOW
FLOW -.
ALL DIMENSIONS IN CM
NOT TO SCALE
.::~~_~_'~88~'_'CC~D_~'~ o~
,N<- - - - - - Ci!?
6 ~~, t=:':HR2~;~R~=~.- ,~(~ 1 '5
0.~80 ~ - ~~~ 0.198 0.080 L 0.025
I 0.041, -. ' I CL
1-
TIP DET AI LS
.
'~
T1
G)
.,.
-------
TRAVERSING GAS SAMPLING PROBE
1.27 D-
0.20 D
x I ) t > t > t > in
tf / r i i / / i i7
0.31 D '
TIP DETAILS
0.95 D
45.72
(D
I
ALL DIMENSIONS IN CM
2.54
P
en
-------
conduction and radiation heat losses; therefore, calibration information
is acquired simultaneously with the required temperature measurement.
The probe consists of three thermocouples including, an iridium - 60
percent rhodium/iridium thermocouple which protrudes from a water-cooled
copper base into the reacting flow, and two platinum - 10 percent rhodium/
platinum thermocouples installed on the ends of the iridium wire to
record the base temperature and thereby permit calculation of the conduc-
tion heat- loss (Fig. 6). A thermocouple probe of this type was applied
without difficulty in the natural gas-air combustion environment. Con-
fidence in the accuracy of the temperature measurements was established
during the previous contract effort (Ref. 7) by measurements made at
identical test conditions using a conventional thermocouple probe and a
double-sonic-orifice probe. Potential errors in the use of thermocouple
probes to measure temperatures in turbulent flames are discussed in Refs.
Ik and 15.
An uncooled five-hole, hemispherical-nose pitot probe (Fig. 7) was
used to measure the radial distribution of the time-mean velocity of the
fuel jet by traversing within the fuel injector. A pitot tap is located
at the center of the probe and four static taps are symmetrically-located
on a centerline circle Uo deg from the tip. Flow velocity and direction
are determined from the differential pressures measured between various
static locations and the pitot pressure. The probe is calibrated in
pitch and yaw to measure flow angles of up to ± Uo deg.
SAMPLING SYSTEM
The gas samples withdrawn through the five-probe exhaust rake or
the traversing probe are analyzed on-line to determine the time-averaged
concentrations of carbon dioxide (C02), carbon monoxide (CO), oxygen (02),
nitrogen oxides (NO, N02) and unburned hydrocarbons (THC). The samples
are transferred to the analytical instruments through a teflon-coated,
flexible line which is heated (~ 400°K) electrically to prevent water
condensation. The sample is then directed through a condensate trap
(~277°K), where most of the water is removed, and it is pumped through
an unheated, teflon coated, aluminum line to a Scott Model 119 Exhaust
Analyzer. A schematic diagram of the sampling system is shown in Fig. 8.
Because natural gas was the only fuel considered in this phase of the
investigation, it was unnecessary to heat the sample lines for THC
measurements. A stainless steel bellows pump increased the sample pres-
sure from subatmospheric levels present downstream from the probes to
1 atm as required by the Exhaust Analyzer. The Analyzer, located in the
combustion facility control room, approximately 10 m from the combustor,
was used to measure the molar concentrations of CO, C02, 02, NO, N02 and
THC.
11
-------
f--'
f\)
-.J
m
I
o
U1
I
N
-.J
I
/
,. -IJ
'\
'-I
CALIBRATED-HEAT -LOSS THERMOCOUPLE PROBE
..
t
WATER OUT
t
\
WATER IN
. .t''' '"
~
,
h
f~"~~~ ~ 1.1,
,,-"f"" - -.
00
~w~
,~ -.~
i@
,,' iii \iI II! IL", if)" ,
'!Ii tI!.", ",'
~ "iI"""'" .,. ..' , .~
~II! -- S.>~.,~~/i .
rtffJ.iI!IIiII-
"
TRAVERSING MECHANISM
FLOW
\
iN
CM
- . -'~
!\ ~'" ',-~
--f.~-.
"~~ ~~-.. '.,
,,~~W
,.
'T1
o
m
-------
UNCOOLED FIVE HOLE HEMISPHERICAL PITOT PROBE
OO
0)
o
01
^1
ID
I
4 STATIC TAPS
AT 90 DEG
40°
TIP DETAILS
ALL DIMENSIONS IN CM
0.32 D
0.63 D'
-8.26-
1.59
45.72
-------
FIG. 8
SCHEMATIC DIAGRAM OF ON-LINE GAS ANALYSIS SYSTEM
ASPIRATED GAS SAMPLE
CALIBRATION AND
1
(
<
1
rf-
^*
^"
P^I
J' / / /
I
CHEATED LINE
CO
TF
_J
NDENSATE BELLOWS
3AP (277°K) PUMP
k - v i , i-J - - *""*
2I//x!2
^7 I//J
^f1^^
II
SCOTT MODEL 119
EXHAUST
ANALYZER
(CO,CO2,NOX,02,THC)
N12-161-1
lU
-------
The Scott Model 119 Exhaust Analyzer (Fig. 9), is an integrated
analytical system, with flow controls for sample, zero and calibration
gases conveniently located on the control panel. The incoming gas
sample passes through a refrigeration condenser (~275°K), to remove
residual water vapor. As the sample passes from the condenser, it is
filtered to remove particulate matter. The Exhaust Analyzer is com-
prised of five different pieces of analytical instrumentation. Beckman
Model 315B Nondispersive Infrared (WDIR) Analyzers were used to measure
the CO and C02 concentrations (mole fractions) in the gas sample. Con-
centration ranges available on the CO analyzer were from 0-200 ppm to
0-15 percent on several scales. Concentration ranges available on the
C02 analyzer were 0-h percent and 0-l6 percent. The accuracy of the
KDIR analyzers is nominally ± 1 percent of full scale. A Scott Model
125 Chemiluminescence Analyzer was used to measure the NO and N02 con-
centrations in the gas sample. Concentration ranges available with this
instrument were from 0-1 ppm to 0-10,000 ppm on several scales, with a
nominal ± 1 percent of full scale accuracy. The thermal converter used
in the chemiluminescent analyzer was stainless steel, and was operated
at a temperature of approximately 1030°K. The converter efficiency
(i.e., the percent NOg dissociated) was determined using the method out-
lined in Ref. 16. In this method, an NO/N2 span gas is diluted with 02
which flows through an ozonator. Measurements with the ozonator off and
on are made both going through and bypassing the converter to determine
converter efficiency. In the present study, a converter efficiency of
99 percent, was measured, with an uncertainty in the measurement of U
percent. During the course of the calibration tests, a loss of NO was
noted when calibration mixtures were flowed through the converter. The
observed loss never exceeded 2 percent of the NO in the stream entering
the converter. A Scott Model 150 Paramagnetic Analyzer was used to
measure the Cg concentration in the gas sample. Concentration ranges
available with this instrument were from 0-1 percent to 0-25 percent on
several scales, with a nominal accuracy of ± 1 percent of full scale.
A Scott Model 116 Total Hydrocarbon Analyzer was used to measure the
hydrocarbon concentration in the gas sample. This analyzer utilizes
an unheated flame ionization detection system to provide for measure-
ment of hydrocarbons (as carbon) in concentration ranges from 0-1 ppm
to 0-10 percent, with a nominal accuracy of ± 1 percent of full scale.
Output signals from the various analyzers are displayed on chart
recorders and on digital readouts. The Analyzer was calibrated prior
to each test by flowing zero gases and calibration gas mixtures having
compositions known to within two percent. Typically, at each test
point, sampling data were acquired for a period of 2-U min.
-------
f-'
0\
"
m
I
o
(J1
I
"
CD
I
w
"
~
C02 NDIR -
ANALYZER
EXHAUST GAS ANALYTICAL SYSTEM
NOX
CHEMI LUMINESCENT
ANALYZER
°2
.~~f.~;,~;ir~
W~
ho
, r>ri1Tt
7' .~.,
CO NDIR ~ r rr
ANALYZER ~I \ . M'
",JIll ._r, '"
. ""'" "'0......,..=", -.., ;..~
~
L
l_-
~.
ttrl~
".. I ?
_-I I I I Itf-
ij \If
,'. .'
t~!J ';J
-~
...
6i-1 I -I 1 1411-
.
"
~~
" "
[~ \II V ~
I ...LLJ .J
! ,
t'l'f'Tli
" " .'
- -.-
.~---'-""O'" _.-
tJ
"'1 Ii I I.
.-II'rHI
I
, l~ ('
JJ
FID
r- HYDROCARBON
~ ANALYZER
~ .t
I ",.. J' I'
.J. J- I
... . , :: .. j
,.0. .1.
!;:1;~ --
-~
~7'~
..
.' !-II '.
.. .. .-:k .;i
. . ". It
r;~
~
..
--
I
, 1
? 1 !
, I~
::!:I
GJ
<0
-------
LASER VELOC METER
The experimental examination of the interaction between fluid
dynamic and chemical processes inside combustors is complicated by the
fact that the mean flow fields and turbulence properties of combusting
flows with recirculation are difficult to determine with any degree of
reliability using conventional instrumentation. Flows with severe
adverse pressure gradients, which normally give rise to separation and
recirculation, are difficult to document as they are extremely sensitive
to local geometry and probe interference (Ref. Ik). In addition, stream-
line curvature and the associated static pressure variations make con-
ventional mean flow instrumentation techniques unreliable.
Although precalibrated pneumatic and microphone probes can be
used with acceptable accuracy in a wide variety of steady flow situa-
tions, significant errors can occur in highly turbulent or unsteady flows
since large (> 20 percent) velocity fluctuations affect the response and
subsequent interpretation of results (Ref. 15). Since most practical
combustor designs involve extensive regions of highly turbulent recircu-
lating flow these probes generally are inadequate for velocity measure-
ment. Figure 10 shows a comparison between mean axial velocity measure-
ments in a swirling natural gas-air flame with a cooled five-hole pitot
probe and a laser velocimeter. In regions with relatively high mean
flow velocities and relatively low turbulent intensities (as determined
by the laser velocimeter), there is relatively good agreement between
the two sets of data. However, in regions with low velocity and high
turbulent intensity, the probe data are scattered and significantly
different from the laser velocimeter results. Bennett (Ref. 1?) has
reviewed errors in pitot probe data resulting from turbulent fluctua-
tions. Bilger (Ref. 15) has discussed the averaging characteristics of
pitot probes in turbulent reacting flows and notes that proper interpre-
tation of probe measurements requiresknowledge of the turbulent structure
of the flow.
There are problems associated with turbulent structure measurements
because linearized hot wire data interpretations are not accurate in
highly turbulent flows (i.e., turbulent intensity > 20 percent) and
because these probe cannot withstand the high temperatures encountered
in combustors. With the advent of the laser velocimeter, linear non-
perturbing fluid mechanical measurements of complex three-dimensional
flow fields are possible provided light-scattering particles can be
relied upon to follow the local fluid velocity.
17
-------
FIG. 10
COMPARISON OF PITOT PROBE AND LASER VELOCIMETER MEASUREMENTS OF VELOCITY
O PROBE DATA
LV DATA
SWIRL = 0.3, 1 ATM, Va/Vf = 21
U(m/sec) 100T X/D = 0.15
-20-L
_L
_J
1.0
-1.0
0.5 0 0.5
RADIAL POSITION, R/RO
18
-------
Since the flows to be investigated involved regions of flow
reversal a laser velocimeter which could determine both the direction
and magnitude of the instantaneous velocity was required. Such a sys-
tem has been developed and used to obtain the detailed mean and turbu-
lence measurements which are presented in this report.
The mean velocity and turbulence measurements were made with a
dual-beam velocimeter utilizing a crystal Bragg cell which acted as a
beam splitter and frequency shifted the first deflected beam. A
schematic diagram of the optics and signal processing instrumentation
is shown in Figs. 11 and 12. The sensing volume determined by beam
crossover volume, off-axis collection and photomultiplier pin hole size
resulted in an elliptic sampling volume with principal axes of 0.2 mm and
2.0 mm, respectively. The velocity component sensed with this optical
arrangement lies in the plane of the two incident beams and is perpen-
dicular to their bisector. Single-particle, time-domain signal pro-
cessing was used to build-up the velocity probability density distri-
butions from which both the mean velocities and rms velocity fluctuations
were obtained using the following equations:
(a)
In the present experiments, a minimum of 1000 instantaneous velocity
determinations were used to build-up the probability distribution functions.
This number of determinations results in a maximum statistical error of
less than 5 percent in the computed values of both the mean and
variance with a confidence level of 95 percent (see Appendix A). Indeed,
a mass balance computed for the case of zero swirl agreed to within
10 percent of the metered fuel and air supply.
The instantaneous axial and tangential velocities were measured
by rotating the Bragg cell about an axis coincident with the laser
beam. With the beams in the axial plane U and-, u'2 are determined from
Eqs. (2) and (3). With the_bearns oriented 90° to the axial plane tan-
gential velocity provides (W and/'w''2 ) were obtained by traversing of
the optical system horizontally and radial velocity profiles (V and
Jv^T ) were obtained by traversing the optical system vertically.
With the beams orientated at ± k$ degrees to the axial plane the tur-
bulent shear stress component u'w1 was determined from the difference of
the two variances.
19
-------
SCHEMATIC DIAGRAM OF THE LASER VELOCIMETER
FIG. 11
COMBUSTOR
TEST SECTION
16 BIT WORDS TO
MINICOMPUTER
76-03-270-10
20
-------
FIG. 12
SCHEMATIC DIAGRAM OF THE LASER VELOCIMETER DATA PROCESSING EQUIPMENT
DATA SYSTEM
INTERFACE
MINI-COMPUTER
16 BIT WORD FROM LASER VELOCIMETER
SIGNAL PROCESSOR (COUNTER)
DATA RATE 200 TO 40,000 WORDS PER SEC
TELETYPE TERMINAL
CASSETTE TAPE
RECORDER
H
VISUAL DISPLAY OF DATA
76-03-270-9
21
-------
The optical sensitivity of the forward scatter system used in the
investigation was such that naturally occurring submicron particles could
be used for the velocity determinations. However, to increase the signal
to noise ratio and thus increase the data acquisition rate the air flow
was seeded with particles dispensed from a fluidized bed. A limited
number of measurements were made with the fuel stream seeded and with
both air and fuel streams seeded to evaluate biasing errors which can result
from seeding only the air flow (Appendix A).
A number of materials which had previously been used to seed small
open flames were tested but none proved suitable for the present experi-
mental arrangement. Both M^Q^ and TiOg deposited on the combustor win-
dows degrading the Doppler signals to an unacceptable extent, and
silicone oil droplets dispensed from a Laskin nozzle evaporated on or
before reaching the combustion zone. However, nominal 5 W& micro-
balloons (hollow spheres) of bakelite phenolic resin were used success-
fully. Due to their low initial density (< 0.1 gm/cc) and to the fact
that they charred to micron size in the combustion zone, these particles
gave adequate turbulence response and excellent signal/noise ratio
(> 10:1) without disturbing optical access.
For a spherical particle of diameter Dp suspended in a sinusoidally
vibrated column of air and acted on by Stokes drag, the ratio of
particle velocity to gas velocity can be expressed as (Ref. 18):
00
where
^P_/27Tf\
u - H j
a =
where u and u are the rms velocities of the particle and the gas p is
the particle density, f is the vibration frequency, I is the molecular
mean free path of the gas and K is the Cunningham constant (~ 1.8 for
air). Thus, a 5 nm phenolic resin microballoon in air at ambient con-
ditions will follow velocity fluctuations up to 10 kHz within 10percent.
Power spectral density measurements in the shear layer of nonreacting
jets (Ref. 18) indicate that for the reacting flows investigated in the
present study more than 95 percent of the turbulence energy will be
associated with Eulerian frequencies below 25 kHz. Hence, the scale of
22
-------
the smallest energy containing eddy will be on the order of
X = u/f « lOOm/sec/^SkHz = 4 xiO"3m
In the Lagrangian frame, this scale corresponds to the frequency on the
order of
f = (u - uc)/x . « 2om/sec/4 xio"3m = 5 kHz
so that errors due to particle response should be negligible.
Conventional laser velocimeters are subject to directional ambigu-
ity which can result in data interpretation errors in highly turbulent
and/or recirculating flows. This problem is illustrated in the insert
of Fig. 13 where Gaussian probability density distributions of the instan-
taneous velocities corresponding to local turbulent intensities of 20
and 70 percent are presented. It can be seen that, with directional
ambiguity, the negative velocities are assigned their equivalent posi-
tive values which leads to errors in the calculated mean value and
standard deviation. These errors rise sharply for turbulence intensities
above ko percent.
To circumvent problems associated with directional ambiguity, zero
velocity frequency offset was achieved by combining the primary and mod-
ulated beams at the detection volume where they generated moving fringes
so that a stationary particle produced a Doppler frequency, f0. Thus,
in the flow field, moving particles generated Doppler frequencies of fQ
± f~ depending on their velocities normal to the moving fringes. Hence,
the sign and the magnitude of the instantaneous velocities could be
determined as follows:
. 9
2sm-|
where \ is the wavelength of the laser light and 6 is the angle between
the incident laser beams.
23
-------
FIG. 13
ERROR DUE TO DIRECTIONAL AMBIGUITY
o
DC
LLJ
Q.
ID
ID
I
ID
4 -
LOCAL TURBULENCE LEVEL (a/U), PERCENT
76-05-198-2
-------
SECTION III
EXPERIMENTAL RESULTS
DESCRIPTION OF THE EXPERIMENTS
In a previous contract effort (Ref. 7) changes in combustor
operating conditions which altered the mean flow field structure in the
combustor were shown to influence pollutant emissions. The principal
changes in the mean flow were found to occur in the initial regions of
the combustor containing the recirculation zones. Furthermore, signifi-
cant fluctuations in the flame structure were observed, and the nature of
these fluctuations were found to depend on the operating conditions. The
present experimental program was carried out to further investigate the
effect of combustor operating conditions on flow field structure and
pollutant formation. The principal objectives of the program were --
(l) to obtain detailed maps of the mean and fluctuating flow field in the
vicinity of the injection plane, including recirculation zones, as
operating conditions were varied and (2) to correlate changes in flow
field structure with changes in pollutant formation and energy release.
The results will be used to evaluate the combustor flow analysis (CRISTY
code) being developed in the theoretical portion of this program (Ref.
12). Advanced optical and probing techniques were used to acquire
detailed data describing the mean flow field properties, including veloc-
ity, temperature and species concentration, and to obtain information on
the turbulence structure of the combustor flow field. The interaction
of fluid dynamic and chemical processes was investigated for a range of
test conditions using two different fuel injector geometries. Ulti-
mately, it is intended that the information obtained from the experimen-
tal and theoretical studies will be utilized for evaluating potential
emission control strategies.
The experimental program was comprised of two different types of
tests: (1) input-output tests and (2) flow-field mapping tests. The
input-ouput tests were conducted with the objective of determining the
relationship of exhaust species concentrations and temperature to
-------
selected combustor operating conditions. Measurements were made at the
exit of the extender section (Fig. l) using the exhaust probe rake and
the traversing thermocouple probe. Previous test results (Ref. 7)
demonstrated the importance of inlet air swirl, combustor pressure, and
air/fuel velocity ratio on governing pollutant emissions and therefore,
systematic variations of these parameters were performed in the present
investigation. Exhaust species concentrations were shown in Ref. 7 to
be less sensitive to variations in fuel-air equivalence ratio, inlet air
temperature and airflow rate, and therefore, these parameters were main-
tained constant. The ranges of variation inlet conditions in the present
input-output tests were selected generally to complement and expand the
data base obtained during the previous contract effort.
Combustor mapping tests were carried out for the purpose of
correlating changes in the fluid dynamic structure of the flow field,
resulting from variation of operating conditions, with the formation and
destruction of pollutant species. Detailed measurements were made
within the combustor at a minimum of four axial locations to determine
radial distributions of the time-mean and rms gas velocity, time-mean
temperature and time-mean species concentrations for five different
combustor operating conditions. The inlet conditions varied in these
tests were those which had the greatest effect on pollutant emissions
inlet air swirl, combustor pressure and air fuel velocity ratio. Mea-
surements were made using the laser velocimeter and temperature and gas
sampling probes described in Section II. Particular emphasis was placed
on augmenting and expanding the data base compiled in Ref. 7; therefore,
most measurements were made within the initial regions of the combustor
(i.e., at axial distances less than five injector diameters) and in the
recirculation zone(s). To assist in the interpretation of the test
results, high-speed color motion pictures (500 frames per second) of the
flame in the vicinity of the injector were obtained for each of the
mapping tests. In addition, the frequency spectra and relative ampli-
tudes of pressure fluctuations associated with the interaction of the
combustion process with the acoustic or mechanical properties of the
combustor were determined. Finally, mean velocity measurements were
made within the fuel port, using a five-hole pitot probe, and near the
exit of the fuel and air ports, using the laser velocimeter to determine
the combustor inlet conditions.
TEST MATRIX
A matrix of combustor operating conditions for tests conducted
using natural gas fuel is presented in Table 1. Eight input-output type
tests and five flow-field mapping experiments were performed to evaluate
the influence of air/fuel velocity ratio, pressure and inlet air swirl
26
-------
Table 1. NOMINAL TEST CONDITIONS (NATURAL GAS-AIR)
ro
Test
No.
1
2
3
h
5
6
7
8
Swirl
No.
0
0
0.3
0.3
0.3
0.6
0.6
0.6
Pressure
(atm)
3.8
7
1
3.8
7
1
1
3.8
Air/Fuel
Velocity Ratio
21
21
21
21
21
21
0.2
0.2
$
0.9
0.9
0.9
0.9
0.9
0.9
0.9
0.9
Mair
(kg/sec)
0.137
0.137
0.137
0.137
0.137
0.137
0.137
0.137
Tair
(°K)
750
750
750
750
750
750
750
750
Input -
Output
X
X
X
X
X
X
X
X
Mapping
X
-
X
X
-
X
X
_
-------
on the temperature, velocity and species concentrations within the
combustor. Selection of the parameters investigated was made with the
objective of supplementing the existing experimental data of Ref . 7,
with emphasis placed on studying those variables which had the greatest
influence on the structure of the flow field and rates of pollutant
formation. In each of the input-output tests the exhaust concentration
of NO, N02, CO, C02, 02 and THC were measured, while in the combustor
mapping experiments detailed radial distributions of temperature,
species concentration, and the mean and rms axial and tangential velo-
city were determined.
The natural gas fuel used for these tests was principally
(> 96 percent) with small amounts of other gaseous hydrocarbons, C02 and
N2 present. Fuel composition analyses are summarized in Appendix B.
Tests were conducted at nominal combustor pressures of 1, 3.8 and 7 atm,
and the overall fuel-air equivalence ratio was maintained constant at a
nominal value of 0.9. The level of swirl of the inlet air was varied by
changing the swirl vanes, and tests were conducted for zero, low (S = 0.3)
and moderate (S = 0.6) swirls. Air/fuel velocity ratios of 21 and 0.2
were achieved by interchanging fuel injectors, shown previously in Fig.
2, and the inlet air flow rate and temperature were held constant at
nominal values of 0.137 kg/sec and 750°K.
INPUT- OUTPUT TEST RESULTS
In the previous investigation, changes in inlet air swirl and pressure
were found to have a significant influence on WO and hydrocarbon emis-
sions. In the present study, the influence of inlet air swirl on exhaust
species concentration levels was evaluated at combustor pressures of 1
(Tests 3 and 6) and 3.8 atm (Tests 1 and U). The emissions data sum-
marized in Table 2, indicate that imparting swirl (3 = 0.3) to the airflow
resulted in a significant increase in nitric oxide emissions and in large
reductions in hydrocarbon emissions. However, with a further increase
in the swirl intensity to 3 = 0.6 there was only a modest increase in NO
and no significant change in THC emissions. The corresponding tempera-
ture distributions in the combustor exit plane, which are presented
together with the flow-field mapping data in a later section of this
report, indicated that flows with swirl resulted in higher peak tempera-
tures while nonswirling flows resulted in temperature profiles with
peaks of reduced magnitude.
Thermochemical considerations indicate that increased pressure will
result in increased temperatures, resulting in more rapid chemical
reaction. In addition, higher pressures result in reduced flow velo-
cities and longer combustor residence times. Each of these effects
28
-------
Table 2. EXHAUST SPECIES CONCENTRATIONS* (NATURAL GAS-AIR)
Test
No.
1
2
3
k
5
6
7
8
Test
No.
1
2
3
5
6
7
8
Swirl
No.
0
0
0.3
0.3
0.3
0.6
0.6
0.6
02
(Mole %)
3.25
1.1*9
3.17
1.67
1.60
2.07
5.60
U. 1*0
Eressure
(atm)
k.O
7.5
1.0
3.7
7.3
1.0
1.0
3.6
C02
(Mole %}
9-50
10.05
9.05
9.80
10.50
9-50
8.05
9.60
Air/Fuel Inlet Air
Velocity Ratio Temperature (°K) i
20.6
20.9
22.0
20.9
20.1+
21.3
0.17
0.17
CO
(Mole $)
0.96
1.52
0.89
1.00
0.73
1.21
0.68
0.35
)f")f
NO WOX
ppm ppm
366 koo
178
171 207
1+29 U76
320 337
175 207
117 15k
195 212
7^8
750
7^3
7^6
7k6
7^6
7kQ
7k8
THC
ppm, C
610
191k
kk
65
709
71
iQk
200
0.90
0.90
0.90
0.91
0.91
0.91
0.91
0.90
Carbon
Balance***
(percent)
-2A
+ 8.6
+ 5-5
+ O.k
- 5.7
+ 1.3
+18.6
+ 8.1*
* Expressed as measured on a dry basis.
** NOX = NO + N02
'.\ fi > n
" ' IN OUT
f \ n < C
k"; IN OUT
29
-------
favors increased NO formation, and th* effect of an increase in combustor
pressure from 1 to 3.8 atm for an inlet air swirl number of 0.3 (Tests 2
vs U) was a significant increase in exhaust gas temperature levels and in
NO emissions. However, a further increase in pressure to 7 atm (Tests 2
and 5) produced a decrease in NO emissions, substantially higher exhaust
concentrations of unburned hydrocarbons and a modest increase in CO
emissions (cf., Tests 1 and 2, and 3, 1+, and 5), suggesting a signifi-
cant change in flow field structure as the pressure is increased to
this level.
A final series of input-output tests was conducted to evaluate the
effect of air/fuel velocity ratio on pollutant emissions. Variation of
this parameter was effected by replacing the large-diameter, low-velocity
fuel injector with a smaller one (see Fig. 2), thereby increasing the
fuel injection velocity and simultaneously decreasing the inlet air
velocity. A change in air/fuel velocity from 21 to 0.2, for an inlet
air swirl number of 0.6 and atmospheric pressure (Tests 6 vs 7), resulted
in a reduction in TTO and CO emissions and in increase in hydrocarbon
emissions. Temperature measurements at the combustor exhaust indicated
that at air/fuel velocity 0.2 the profile was significantly less uniform
with large radial gradients.
A comparison of the emissions data obtained in the present
investigation with observations previously reported in Ref. 7 indicates
that there is general agreement with respect to trends resulting from
variation of inlet air swirl, combustor pressure and air/fuel velocity
ratio; however, exhaust concentrations of NO are lower, THC concentra-
tions are higher and CO concentrations are generally higher in the pres-
ent tests. This result suggests that the local temperature levels are
lower than those measured in the corresponding tests of Ref. 7; a pre-
sumption that was confirmed by the mapping data. These differences may
be due, in part, to modification of the previous combustion system which
included (l) elimination of the uncooled portion of the combustor in the
vicinity of the fuel injector and (2) relocation of the swirl vanes to
a station upstream of the injector exit. These changes should increase
the heat transfer to the combustor wall and reduce the temperature
levels in the combustor.
The repeatability of the exhaust species concentration measurements
was determined by obtaining several data points at each of the operating
conditions listed in Table 1, and the accuracy of the gas sampling and
analysis techniques was verified by performing a carbon balance between
the reactant and product species. The overall repeatability of the
measurements was approximately + 5 percent for the species 02, C02, CO,
NO and NOX, and approximately + 25 percent for THC. These variations
30
-------
are attributed primarily to small changes in input and combustor
operating conditions and, to a lesser extent, to errors in the sampling
and emissions measurement procedures. Normal acceptance criteria
specified in SAE ARP 1256 (Ref. 19) require that the carbon atom con-
centration determined from emission measurements agree to within 15 per-
cent of the concentration determined from fuel analysis. Carbon bal-
ances were calculated for all of the exhaust emissions data and the
results are tabulated in Table 2. Except for Test 7, in which large
radial concentration gradients were observed throughout the combustor
and at the exhaust station, the above acceptance criteria was satis-
fied and a balance to within nine percent was achieved.
FLOW FIELD MAPPING RESULTS
Detailed maps of the mean and fluctuating flow field were obtained
for the five test conditions listed in Table 1. These test conditions
were selected to encompass variations in combustor operating conditions
which have the greatest influence on pollutant emissions.
High-Speed Motion Pictures
As an initial qualitative indication of flow field structure, high-
speed (500 frames/sec) color motion pictures of the reacting flow in
the vicinity of the injector were obtained for each of the five test
conditions. These films showed that there were significant large-scale
fluctuations in the flame luminosity for all test conditions. The
nonswirling flow exhibited the largest unsteadiness, Fig. ih. The
axial motions of the flame in this flow were sufficiently large so that
the flame was observed to occasionally enter the fuel injector. Impart-
ing swirl to the air stream reduced the fluctuations but even in
swirling flows the flame is very unsteady in the vicinity of the injec-
tion plane. These visual observations of the flame structure support
the conclusions drawn later from laser velocimeter data regarding the
large-scale fluctuations of the flow in the Initial mixing regions.
Analysis of the transient pressure data, discussed below indicates that
for most of the conditions investigated there are no significant resonant
pressure fluctuations. Furthermore, similar large-scale fluctuations
have been observed in the initial mixing regions of nonreacting jets
(Ref. 20). Hence, the observed fluctuations are primarily fluid dynamic
in origin and are not the result of coupling of the combustion process
with the acoustic properties of the combustor or mechanical properties
of the injector.
In addition to providing information on the time-dependent structure
of the flame, the high-speed films also give qualitative information on
31
-------
FIG. 14
NO SWIRL
OVERALL EQUIVALENCE RATIO = 0.9
3.8 ATMOSPHERES
FRAMING RATE = 500/SEC
EXPOSURE TIME = 1 MSEC
UJ
~
I-
~
FLOW
-
76-06-168-5
32
-------
the spreading rate of the fuel jet. In all cases, the variations in
spreading rate with changes in inlet conditions observed on the films
are in agreement with the sampling data.
Transient Pressure Measurements
The high-speed motion pictures of the combustor flow field in Fig.
lU indicate a variation of the point of ignition from axial positions
downstream of the injector to positions within the fuel injection port.
Since the large-scale fluctuations will obviously complicate attempts
to analytically model the combustor flow, a series of tests were con-
ducted to evaluate the interaction of the combustion process with the
acoustic or mechanical properties of the combustor and to thereby ensure
that the configurations selected for detailed mapping were free of
large-scale combustion instabilities.
The frequency spectra and the relative amplitudes of the pressure
fluctuations occurring for each of the mapping experiments were deter-
mined by analyzing the output signal of a close-coupled high frequency
pressure transducer installed in the combustor window port nearest the
injector. A high-speed oscillograph was used to continuously record the
signal and permit measurement of the amplitude of pressure fluctuation,
and a spectrum analyzer was used to simultaneously determine the fre-
quency of the temporal component. The results, summarized in Table 35
indicate that for operation at 1 atm and zero swirl a periodic pressure
fluctuation occurs with a frequency of approximately 120 Hz and an
amplitude of + 6 percent. The measured frequency is in close agreement
with a calculation of the fundamental harmonic for the combustor.
Additional tests conducted at successively decreased fuel flow rates
revealed that the magnitude of the fluctuations decreased rapidly and
were less than +0.5 percent for 0 = 0.6.
When a swirl component was introduced on the inlet airflow at 1 atm,
the frequency of the oscillation remained unchanged; however, the ampli-
tude decreased steadily with increasing swirl and was less than + 0.5
percent for S = 0.6. Also, the use of an orifice plate to increase the
combustor pressure to 3.5 atm effectively changed the acoustic charac-
teristics of the combustor and significantly reduced the natural
frequency and the amplitude of the pressure fluctuations. The frequency
spectra obtained for typical test conditions are presented in Fig. 15.
Temperature Data
The combustor mapping data have been reduced to isopleth form to
permit visualization of the radial and axial variation of individual
33
-------
Table 3. SUMMARY OF TRANSIENT PRESSURE MEASUREMENTS
Test
No.
1
3
h
6
7
Swirl
No.
0
0.3
0.3
0.6
0.6
Pressure
(atm)
3.8
1.0
3.8
1.0
1.0
Air/Fuel
Velocity Ratio
20
20
20
20
0.2
$
0.9
0.9
0.9
0.9
0.9
Frequency
(Hz)
2
110
9
110
110
Amplitude Variation
«)
< ± 0.5
± 1.0
< ± 0.5
< ± 0.5
< ± 0.5
Ref. 7
UJ
0
1.0
20
0.9
120
± 6.0
-------
FIG. 15
FREQUENCY SPECTRA OF TRANSIENT
PRESSURE FLUCTUATIONS
lil
Q
I
111
H
<
UJ
21 : 1 FUEL INJECTOR
nli i >
(A) NO SWIRL, 3.8 atm
(B) SWIRL = 0.3,1 atm
A
(C) NO SWIRL, 1 atm
A, k
200 400 600
FREQUENCY-Hz
800
1000
76-06-282-21
-------
flow field parameters and to facilitate comparisons between these
parameters in each of the flow configurations investigated. However,
since the radial distributions of mean flow properties were determined
at four to six axial locations within the combustor, interpolation
between stations was necessary and, therefore, a certain amount of
"artistic license" has been assumed in constructing these plots. In
each of these isopleths the radial extent of the fuel delivery port and
the air annulus is indicated to aid in visualization of the flow field
structure. A complete tabulation of the experimental temperature data
is presented in Appendix D.
Contours of constant temperature which show the time-mean
temperature distributions obtained for all of the mapping test conditions
are presented in Fig. 16. An initial examination of the data reveals
the similarity of the flow field structure obtained for each of the test
configurations and the temperature distributions characteristic of
axisymmetric, turbulent diffusion flames, i.e., peak temperatures occur-
ring in an annular region off the centerline. Specific trends in the
temperature distributions resulting from variation of the inlet param-
eters are evident from a detailed examination of the isopleths. For
example, imparting swirl = 0.3 to the airflow at 3.8 atm (Figs l6a and
l6b) and increasing it from 0.3 to 0.6 at 1 atm (Figs. l6c and l6d)
resulted in increased locally high temperatures developed off the center-
line and increased radial temperature gradients. Furthermore, comparison
of the axial temperature gradients indicate that increasing swirl also
results in significantly higher temperatures in the vicinity of the fuel
injector.
Increasing the combustor pressure from 1 atm to 3.8 atm,with all
other inlet conditions remaining constant, resuits in longer combustor
residence times, higher temperatures and more rapid chemical reaction.
For the case of 0.3 swirl (Figs. l6b and l6c) the temperature distribu-
tions again exhibited the typical diffusion-flame-like structure; how-
ever, increasing pressure shifted the location of the peak temperature
closer to the centerline and resulted in higher radial gradients. As
expected, the temperature and heat release rate in the vicinity of the
injector were increased, primarily due to reduced flow velocity.
The influence of air/fuel velocity ratio on the temperature
distribution within the combustor is illustrated in Figs. l6d and l6e.
For a combustor pressure of 1 atm and an inlet air swirl of 0.6 decreas-
ing the air/fuel velocity (increased fuel velocity) from 21 to 0.2
resulted in a shift of the radial location of the peak temperatures away
from the combustor centerline of the duct. The conspicuous change in
the radial location of the maximum temperature suggests a significantly
36
-------
FIG. 16
1.0
0.5
§ 0
Q.
5-0.5
-1.0
1.0
0.5
55 0
O
Q.
<-0.5
O
-1.0
O
cc
O
1.0
0.5
o
a
< -0.5
5
-1.0
TIME-AVERAGED TEMPERATURE DISTRIBUTIONS
(A) ZERO SWIRL, 3.8 ATM
12 14 16
AXIAL DISTANCE, X/D
(B) SWIRL = 0.3, 3.8 ATM
4 6 8 10 12 14 16
AXIAL DISTANCE, X/D
(C) SWIRL = 0.3, 1 ATM
46 8 10 12
AXIAL DISTANCE, X/D
14 16
76-07-158-1
37
-------
TIME-AVERAGED TEMPERATURE DISTRIBUTIONS
FIG. 16
1.0
0.5
(D) SWIRL = 0.6,1 ATM
55 0
O
a.
<-0.5
a
-1.0
1.0
O
cc
0.5
V) 0
O
a.
1-0.5
-1.0
1200 j
4 6 8 10 12
AXIAL DISTANCE, X/D
14 16
(E) SWIRL =0.6,1 ATM, 0.2/1 INJECTOR
4 6 8 10 12 14 16
AXIAL DISTANCE, X/D
38
76-07-158-2
-------
more rapid rate of spreading of the fuel jet for the high fuel injection
velocity, an observation that was confirmed by measurements of the
radial concentration gradients of unburned hydrocarbons. In addition,
significantly lower temperatures were measured on the combustor center-
line and, therefore, the radial temperature gradients also were
increased.
Concentration Data
Because of the importance of the physical and chemical phenomena
occurring in the region near the injector (e.g., formation of the recir-
culation zone(s) and initiation of mixing and chemical reaction) measure-
ments of the species concentration profiles were made at four to six
locations within an axial distance equivalent to four combustor diameters
(X/D = If). Detailed maps of the species concentration distributions are
presented in Figs. 1? to 21. The concentration data are tabulated in
Appendix E.
Examination of the species contours reveals significant changes in
the mean species concentration distributions with variations in combustor
operating conditions. For example, comparison of Figs. 17 and 18 (Tests
1 and 4) reveals that introduction of swirl at a combustor pressure of
3.8 atm greatly increases the rate of oxidation of the fuel to CO, as
evidenced by the THC and CO concentration contours, accelerates the rate
of NO formation, with a significant fraction of the exhaust NO concen-
tration being formed within the first three combustor diameters, and
increases the rate of oxidation of CO to C02. All of these observations
point to a higher energy release rate in the swirling flow, which is
consistent with the measured mean temperature contours. Introduction
of swirl increases the spreading rate of the fuel jet, as evidenced by
the THC contours, and the principal region of fuel-air mixing and
reaction appears to be displaced radially outward. These observations
also are consistent with the temperature data, which show an outward
displacement of the temperature peaks upon imparting swirl.
Measurements of the species concentration distributions for the
S = 0.3 and the S = 0.6 flows at 1 atm combustor pressure are presented
in Figs. 19 and 20 (Tests 3 and 6). The THC concentration maps indicate
that as the swirl number is increased the spreading rate of the fuel
jet rapidly increases with a corresponding decrease in the concentration
of unburned hydrocarbons along the centerline. Unlike the other flows
investigated, moderately large fuel concentrations are measured near the
combustor wall in the S = 0.6 test. As noted earlier, there appears to
be significant large-scale fluctuations in all of the flows investigated.
Transport of fuel radially outward in the swirling flows may be the
39
-------
FIG. 17
TIME-AVERAGED SPECIES DISTRIBUTIONS
NO SWIRL, 3.8ATM,Va/Vf = 21
AXIAL DISTANCE, X/D
-1.0 L-
1.0
0.5
5> 0
O
a.
< -0.5
oc
-1.01-
234
AXIAL DISTANCE, X/D
EXHAUST THC = 610 PPM
1
1 1 1 1
234
AXIAL DISTANCE, X/D
76-07-158-19
-------
FIG. 17
TIME-AVERAGED SPECIES DISTRIBUTIONS
NO SWIRL, 3.8 ATM, Va/Vf = 21
1.0
O 0.5
O
£-0.5
-1.0
1234
AXIAL DISTANCE, X/D
1.0
O 0.5
5
2-0.5
-1.0 L
EXHAUST C02 = 9.5%
I
J_
1 234
AXIAL DISTANCE, X/D
76-07-158-18
-------
FIG. 18
TIME-AVERAGED SPECIES DISTRIBUTIONS
SW! R L = 0.3, 3.8 ATM, Va/Vf = 21
i.or
§ 0.5
§
O
a.
0
< -0.5
-1.0
1.0
I °'5
55 0
O
a
_i
5
tt -0.5
-1.0
1234
AXIAL DISTANCE, X/D
EXHAUST O2 = 1.67%
2 3
AXIAL DISTANCE, X/D
(B) SWIRL = 0.3, 3.8 ATM
1.0
O 0.5
cc
O
a.
-0.5
-1.0U
AXIAL DISTANCE, X/D
76-07-158-4
-------
TIME-AVERAGED SPECIES DISTRIBUTIONS
SWIRL = 0.3, 3.8 ATM, Va/Vf = 21
FIG. 18
1.0
0.5
« 0
o
a.
_i
a
a-0.5
-1.0
1 2 3
AXIAL DISTANCE, X/D
1.0
O 0.5
55 0
O
2-0.5
-1.0
1 2 3
AXIAL DISTANCE, X/D
76-07-158-5
-------
FIG.19
TIME-AVERAGED SPECIES DISTRIBUTIONS
SWIRL = 0.3,1 ATM, Va/Vf = 21
1.0r- r
0.5
V) 0
O
a.
_i
5
< -0.5
-1.0
1.0
P 0.5
O
o.
-0.5
-1.0
1234
AXIAL DISTANCE, X/D
1234
AXIAL DISTANCE, X/D
1.0
0.5
O
E
(f>
O
a
5 -0.5
E
-1.0
AXIAL DISTANCE , X/D
76-07-158-15
-------
TIME-AVERAGED SPECIES DISTRIBUTIONS
SWIRL = 0.3, 1 ATM, VaA/f = 21
FIG. 19
1.0
O 0.5
f
I
Q.
_l
5
2-0.5
-1.0
1.0
O 0.5
QC
(/> 0
o
Q.
-0.5
-1.0
I
I
I
1234
AXIAL DISTANCE, X/D
1234
AXIAL DISTANCE, X/D
76-07-158-14
-------
FIG. 20
TIME-AVERAGED SPECIES DISTRIBUTIONS
SWIRL = 0,6.1 ATM, Va/Vf = 21
1.0
O 0.5
55 0
O
o.
Q
tt 0.5
1.0
2 3
AXIAL DISTANCE,
1.0
O 0.5
|
g
M o
o
0.5 -
1.0
EXHAUST O2 = 2.07%
2 3
AXIAL DISTANCE, X/D
1.O.-
O 0.5
CO
O
Q.
Q
a 0.5
1.0
1334
AXIAL DISTANCE, X/D
76-07-158-8
-------
TIME-AVER AGED SPECIES DISTRIBUTIONS
SWIR L = 0.6,1 ATM, Va/Vf = 21
FIG. 20
1.0 r
§
0
o
Q.
a
a 0.5
1.0
2 3
AXIAL DISTANCE, X/D
O
1.0
0.5
55 0
_i
5
K 0.5
1.0
2 3 4
AXIAL DISTANCE, X/D
76-07-158-3
-------
result of entrainment of large fuel eddies by the swirling air stream,
with subsequent transport of the fuel toward the combustor wall due to
the radial spreading of the air stream. With increased swirl, oxidation
of the hydrocarbon fuel to GO is greatly accelerated with a lesser
increase observed in the rate of oxidation of GO to C02. These observa-
tions suggest an increase in energy release rate with an increase in
swirl from 0.3 to 0.6. This conclusion is supported by the measured
temperature distributions. The higher temperature levels for the S =
0.6 flow result in a more rapid NO formation rate.
One of the principal effects of elevated pressure is a decrease in
the flow velocity and increase in the time available for reaction.
Analysis of the NO concentration data presented in Figs. 19 and 18
(Tests h and 3) indicates that an appreciable increase in the axial rate
of NO formation occurred as a result of raising the combustor pressure
from 1 to 3.8 atm. However, comparisons made on the basis of equivalent
residence times (as opposed to equivalent axial distances) show a much
smaller increase in NO concentrations and, therefore, demonstrate the
residence time effect. At elevated pressure, radial spreading of the
fuel jet was noticeably diminished as was the penetration of oxygen to
the center of the flow. These trends are in agreement with the previ-
ously discussed shift of the temperature maxima toward the centerline.
Also, at higher pressures, the oxidation of hydrocarbons to CO is
accelerated, as is shown in Fig. 18, and significantly higher GO concen-
trations were measured along the centerline.
As indicated earlier, a significant alteration in the temperature
distribution within the combustor was observed as a result of inter-
changing fuel injectors and operating at an air-fuel velocity ratio of
0.2 (Test 7). Qualitative evaluations suggested that rapid spreading of
the fuel jet occurred. The 02 and THC concentration contours for air-
fuel velocity 0.2, Fig. 21, supports these observations and shows high
concentrations of fuel extending well beyond the lip of the injector and
low oxygen concentrations in the central portion of the flow. Low tem-
peratures and near zero concentrations of NO were measured along the
centerline. The appearance of the WO isopleth differs markedly from
those obtained at low injection velocity and shows that high NO concen-
trations exist in a torroidal-shaped region close to the injector,
approximately coincident with the recirculation zone. In this region,
the local temperature is very high and there is rapid oxidation of the
hydrocarbon fuel to form CO and ultimately C02. In contrast, low CO and
C02 concentrations were measured near the centerline due to the slower
rate of hydrocarbon reaction.
1*8
-------
TIME-AVERAGED SPECIES DISTRIBUTIONS
SWIRL = 0.6, 1 ATM, Va/Vf = 21
10r
FIG. 21
0.5
1
I-
55 o
o
a.
< -0.5
-1.0L
1.0
0.5
EXHAUST NO =117 PPM
1234
AXIAL DISTANCE, X/D
2 -0.5
-1.0
EXHAUST Oz = 5.60%
10
I
14
I
J
234
AXIAL DISTANCE, X/D
1.0
O 0.5
(0
O
a
_j
5
2 -0-5
-1.0"-
EXHAUST THC = 184 ppm
2 3
AXIAL DISTANCE, X/D
76-07-158-16
-------
TIME-AVERAGED SPECIES DISTRIBUTIONS
SWIRL = 0.6, 1 ATM, Va/Vf = 21
FIG. 21
i.o r
o o.s
o
Q.
-1.0L-
1234
AXIAL DISTANCE, X/D
1.0
O 0.5
55 0
O
a.
_i
S
E-0.5
-1.0
EXHAUST CO2 = 8.05%
1 2 3
AXIAL DISTANCE, X/D
76-07-158-17
-------
M02 Concentration Measurements
Evidence of significant nitrogen dioxide (N02) concentrations near
the primary reaction zone of turbulent diffusion flames has been reported
by several investigators (Refs. 21 and 22). In the present study N02
concentrations in excess of NO concentrations were measured along the
mean flame boundaries in the vicinity of the fuel injector. Because of
the difficulty in making quantitative measurements of N02 in strongly
reducing atmospheres using the stainless-steel thermal converter of the
chemiluminescence monitor (Ref. 16), W02 data could not be acquired as
the probe was traversed into the fuel-rich regions of the combustor.
Therefore, an NDUV analyzer was used to augment the chemiluminescence
detector by measuring W02 emissions in selected tests. A typical pro-
file, shown in Fig. 22, indicates that N02/N0 ratios greater than unity
were measured at the lean boundary of the flame zone, and N02 concentra-
tions subsequently decreased to a low level in the fuel-rich central
core region of the flow. As was stated in Ref. 7? significant N02 for-
mation was observed within an axial distance equivalent to three com-
bustor diameters downstream of the injector exit and thereafter the N02
concentration remained relatively constant with increasing axial dis-
tance. However, there remains considerable uncertainty in the observed
N02 levels because of potential sources and sinks for N02 during trans-
fer of the gas sample to the Exhaust Analyzer (Refs. 7 and 16).
Mean Velocity Measurements
The radial distributions of time-mean axial and tangential
velocities were measured at a minimum of four and a maximum of six
axial locations in the initial regions of the combustor (X/D £ 2). In
addition a limited number of measurements of the time-mean radial veloc-
ity were made. A typical set of mean axial velocity profiles is shown
in Fig. 23. From profiles such as these, mean axial velocity contours
were constructed, showing lines of constant velocity within the combus-
tor. The mean axial velocity contours obtained for the five test
conditions are presented in Fig. 2k. These profiles show the location
and shape of the time-averaged recirculation zones and indicate their
approximate longitudinal and lateral extent. As was the case for the
temperature and concentration isopleths, interpolation between data taken
at various axial stations was required to develop the velocity contour
plots. All the data show a consistent trend towards uniform velocity
profiles with increasing distance from the injector, which would be
expected for highly turbulent plug-like flows immediately downstream of
the initial mixing region.
-------
FIG, 22
NITROGEN OXIDE DISTRIBUTIONS
SWIRL - 0.6, 1 ATM,Va/Vf = 0.2
NO MEASURED BY CLA
A NOx = NO + NO2 MEASURED BY CLA
D NO2 MEASURED BY NDUV
A NOx = NO (CLA) + NO2 (NDUV)
100
EXHAUST NO = 117 ppm
EXHAUST NOx = 154 ppm
I X/D'1.6
-0.5 0 0.5
RADIAL POSITION, R/RO
76-06-282-27
-------
FIG. 23
MEAN AXIAL VELOCITY PROFILES
SWIRL = 0.3, 1 ATM, Va/Vf = 21
O X/D = 0.13
D X/D = 0.48
0 X/D = 0.05
J 20
J_
-1.0
-0.5 0 0.5
RADIAL POSITION, R/RO
LO
-------
MEAN AXIAL VELOCITY DISTRIBUTIONS
FIG. 24
(A) ZERO SWIRL, 3.8 ATM
1.0
Z
O
O
a.
1-0.5
1.0
1.0
0.5
55 0
O
a
<-0.5
-1.0
0.5 1.0 1.5
AXIAL DISTANCE, X/D
(B) SWIRL = 0.3, 3.8 ATM
2.0
0.5 1.0 1.5
AXIAL DISTANCE, X/D
2.0
1.0
0.5
55 0
O
a.
< -0.5
O
-1.0
(C) SWIRL = 0.3, 1 ATM
-+2C
0 +10 +20 +30 +40
0.5 1.0 1.5
AXIAL DISTANCE, X/D
2.0
76-07-158-20
-------
MEAN AXIAL VELOCITY DISTRIBUTIONS
FIG. 24
1.0
- 0.5
en 0
O
OL
_l
<-0.5
Q
-1.0
(D) SWIRL = 0.6,1 ATM
0.5 1.0 1.5
AXIAL DISTANCE, X/D
2.0
1.0
(E) SWIRL = 0.6,1 ATM, 0.2/1 INJECTOR
-------
In the case of zero swirl at 3-5 atm a large spheroidal time-
averaged recirculation zone is present immediately downstream from the
center (fuel) jet with associated mean reverse velocities significantly
larger than in any of the swirling flow cases. In addition to this
central zone, there is a second recirculation zone behind the backward
facing step at the nozzle exit plane.
The introduction of swirl brings about significant changes in the
time-averaged flow field: a much smaller toroidal recirculation zone
is present and the secondary racirculation zone is so reduced in size it
cannot be detected. The primary results of increasing swirl from 0.3 to
0.6 at 1 atm are to produce a more pronounced initial radial mean flow
and a somewhat larger, though still toroidal recirculation zone. How-
ever, in both these cases there is a much less rapid profile development
than at 3.5 atm which can be primarily attributed to the decrease in
combustor residence time.
For the small injector case there is extensive fuel jet coherence
even though the annular air swirl induces a significant initial spreading
rate.
The principal feature of each set of mean tangential velocity
profiles (Fig. 25) is the change from solid body rotation (i.e., forced
vortex flow) close to the injector towards a combined free/forced (i.e.,
Rankine) vortex flow downstream. Thus a region of irrotational flow
develops and progresses tcr-rards the center _" the duct as the flow pro-
ceeds downstream. As a result, the point of maximum tangential velocity
moves radially inward. This trend from forced to Rankine vortex flow
becomes more pronounced with both increased swirl and ambient combustor
pressure (increased residence time). Such a transition from forced to
Rankine vortex requires a sink, which in the present case is provided by
the inward radial flow downstream of the time-averaged recirculation
zones. This radial flow is evident from the increasing centerline axial
velocities in Fig. 2k, for example, and is more pronounced at 3.5 atm
thereby enhancing the rate of vortex transition. Flow angle calcula-
tions based on the measured tangential velocity at the location of the
peak axial velocity compare extremely well with the swirl vane blade
angles at S = 0.3. For the 0.3 swirl cases, blade angle 7] = 28°, the
flow angles at 1 and 3-5 atm are 26°. At 0.6 swirl, T] = 1+7°, the cal-
culated flow angle was 37°. However, in this latter case the number of
blades was reduced from 18 to 12. Such a large change in solidity
(chord/gap ratio), necessitated from a vane packaging standpoint, is the
likely explanation for this reduced swirl efficiency.
-------
MEAN TANGENTIAL VELOCITY PROFILES
SWIRU = 0.6,1 ATM, Va/Vf = 21
FIG, 25
w (m/sec) -p 100
X/D = 0.16
SS ? //V A
-1.0
"6.5 " "" 0 0.5
RADIAL POSITION, R/RO
1.0
57
76-06-282-24
-------
FIG. 26
MEAN AND RMS TANGENTIAL VELOCITY DISTRIBUTIONS
(A) SWIRL = 0.3, 3.8 ATM
O 1.0
Z 0.5
O
i o
Q.
< -0.5
Q
<
tt -1.0
0.5 1.0 1.5
AXIAL DISTANCE (X/D)
(B) SWIRL = 0.3, 1 ATM
1.0
0.5
-------
FIG. 26
MEAN AND RMS TANGENTIAL VELOCITY DISTRIBUTIONS
(C) SWIRL = 0.6,1 ATM
o -
gc
^ 0.5
O
P
to 0
O
a.
< -0.5
Q
cc -1.0
O
I
s
O
1.0
0.5
0.5 1.0 1.5
AXIAL DISTANCE, (X/D)
(D) SWIRL = 0.6, 1 ATM, 0.2/1 INJECTOR
8 '
Q.
< -0.5
Q
-1.0
0.5 1-0 1.5
AXIAL DISTANCE, (X/D)
2.0
10%
2.0
76-07-158-22
59
-------
The presence of forced vortex flow immediately downstream of the
injector has a pronounced effect on the flow field static pressure dis-
tribution since, neglecting viscous forces, there is a balance between
pressure and inertial forces given by dp/dr = -p₯2/r. Since War we
obtain the well-known result that the static pressure increase in the
core of a forced vortex is proportional to the square of the radius.
This result significantly affects the mean axial velocity flow field
because now the inviscid central fuel jet axial momentum is able to
overcome"the reduced static pressure gradient near the flow field cen-
terline. This results in the formation of a smaller toroidal recircu-
lation zone with reduced negative velocities made up primarily of lower
momentum boundary layer fluid as indicated by their locations in Fig.
2k.
To characterize the mean radial flow in the combustor, the mean
radial velocity was measured at one axial location just downstream from
the large fuel injector (X/D = 0.3*0 f°r a swirl number of 0.3 at atmo-
spheric pressure, Fig. 27. The flow is directed radially inward toward
the centerline in the central portion of the combustor, with peak mean
radial velocities of approximately 10 m/sec. These observations are
consistent with the description of the flow field determined from the
axial and tangential velocity contours.
RMS Velocity Measurements
Measurements of the associated root-mean-square axial velocity
fluctuations, presented in Fig. 28, indicate extremely high local fluc-
tuation levels in the initial mixing regions. For the large-diameter
injector, it was found that the normalized fluctuation levels decrease
with both increasing swirl and ambient pressure. These data, when
superimposed on the mean velocity contours, Fig. 29, show that the peak
fluctuation levels coincide primarily with the locations of high mean
shear, i.e., maximum local mean velocity gradient. However, measured
rms velocity fluctuation levels provide no information on the turbulent
scales involved. For instance, there are significant fluctuations
associated with the time-averaged recirculation zones with local peak
intensities occurring at points in the flow close to the time-averaged
axial stagnation points. Since the mean velocity gradients are rela-
tively low in these regions, the fluctuations must be due primarily to
large-scale motion associated with recirculation zone entrainment and/or
unsteadiness about its mean location. In the small diameter fuel injec-
tor case, the rms velocity contours are dramatically different and
appear to be a result of fuel jet "flapping" about its mean location
which was indicated by bi-modal probability density distributions
60
-------
FIG, 27
MEAN RADIAL VELOCITY PROFILE
SWIRL = 0.3. 1 ATM. Va/Vf = 21
10T
V (m/sec)
X/D = 0.34
-10-1-
1.0
-0.5 0 0.5
RADIAL POSITION, R/RO
1.0
6l
76-07-158-12
-------
FIG. 28
1.0
0.5
O
i
o
O
a
5 -0-5
-1.0
O
DC
a
z"
O
1.0
0.5
co o
O
u.
1-0.5
-1.0
pc
cc
O
CO
O
Q.
_l
<
Q
1.0
0.5
-1.0
AXIAL RMS VELOCITY DISTRIBUTIONS
(A) ZERO SWIRL, 3.8 ATM
0.5 1.0 1.5
AXIAL DISTANCE, X/D
(B) SWIRL = 0.3, 3.8 ATM
0.5 1.0 1.5
AXIAL DISTANCE, X/D
(C) SWIRL = 0.3, 1 ATM
30.
0.5 1.0 1.5
AXIAL DISTANCE, X/D
2.0
2.0
76-07-158-9
62
-------
FIG. 28
1.0
AXIAL RMS VELOCITY DISTRIBUTIONS
(D) SWIRL = 0.6,1 ATM
o
cc
40 130
0.5
55 0
O
Q.
<-0.5
-1.0
1.0
. 0.5
O
H
55 0
O
Q.
0.5 1.0 1.5 2.0
AXIAL DISTANCE, X/D
(E) SWIRL = 0.6, 1 ATM, 0,2/1 INJECTOR
-1.0
1
1 ' .
1 O.5
i -
1.0
i
1.5
J
2.0
AXIAL DISTANCE, X/D
76-07-158-10
63
-------
FIG. 29
AXIAL MEAN AND RMS VELOCITY AND
DIRECTIONAL INTERMITTENCY PROFILES
NO SWIRL, 3.8 ATM, Va/Vf ' 21
20
15
> 10
o
g
LLJ
_J
< 5
x
-5 L_
X/D = 0.15
'7777777]
80
60
c
s>
u
i
UJ
40
20
UJ
O
UJ
DC
5
FT^TVvv
_L
X
-1.0 -0.5 0 0.5
RADIAL POSITION, R/R
1.0
76_06-282-25
-------
obtained at the edge of the fuel injector close to the exit plane.
These instantaneous fuel jet direction changes give rise to the diverg-
ing rms velocity contours.
Tangential velocity fluctuation measurements at 3.8 atm are shown
in Fig. 30. These data obtained at X/D = l.i£ show a sharp peak near
the combustor centerline where the mean gradient is high. Contours of
constant tangential rms velocity (Fig. 26) show that this peak becomes
more pronounced as the flow proceeds downstream. This central rms
velocity distribution can be directly related to the increase in direc-
tional intermittency and mean tangential velocity gradient induced by
the transition from forced to Rankine vortex flow shown in Fig. 25.
The size, shape, recirculated mass flow and local turbulence levels
associated with recirculation zones are important to flame stability and
combustion intensity. High-speed motion pictures of the reacting flow
field in the vicinity of the injection plane show that there are signif-
icant fluctuations in the flame structure and large-scale motions asso-
ciated with- flow reversal. These large-scale motions are associated
with instantaneous movements of the recirculation zone location due to
local imbalances between fluid entrained from and fluid deflected into
the recirculation zone "which are in turn related to the local velocity
gradients, turbulence scales and recirculation zone size (Ref. 23).
Quantitative insight into this large-scale turbulent motion
associated with flow recirculation can be obtained from velocity prob-
ability density distributions such as those shown in Fig. 31. These
measurements, which can be obtained only with a velocimeter system with
zero velocity frequency offset, show the unsteadiness of the flow field
in the initial mixing region. For example, within the time-averaged
recirculation zone (R/RO = 0.35) there are a significant number of posi-
tive velocity occurrences (approximately 30 percent) which are the
result of either instantaneous recirculation zone breakdown and/or
extensive streamwise and lateral movement. These large-scale motions
result in significant deviations from Gaussian turbulence. In the
corner region (R/RQ = 0.9) approximately 25 percent of the instantaneous
velocity occurrences are negative and again the velocity probability
density distribution is skewed.
Defining directional intermittency (y) at a given point as the
fraction of the total observed velocity occurrences which are negative,
contours of constant directional intermittency can be constructed.
Such plots for the five test conditions are shown in Fig. 32. These
data show that there are a significant number of negative velocity
occurrences over most of the initial mixing region and that the
65
-------
FIG, 30
TANGENTIAL MEAN AND RMS VELOCITY AND
DIRECTIONAL INTERMITTENCY PROFILES
SWIRL = 0.3, 3.8 ATM, Vg/Vf = 21
20
15
10
l-
u
o
UJ
UJ
O
-5
-10
-15
X/D = 1.48
-1.0 -0.5 0 0.5
RADIAL POSITION, R/RQ
50
25 51 1
i^v
1.0
76-06-282-26
-------
FIG 31
PROBABILITY DISTRIBUTION FUNCTIONS
OF AXIAL VELOCITY
SWIRt - 0.3, 3.8 ATM,Va/Vf = 21
OCCURRENCE
PROBABILITY,
percent
X/D = 0.09
R/RO = 0.37
-20
- \e.
-iQ
1
R/RO
0.9
0.70
0.37
»/U (%!
195
32.5
200
->(%)
24.9
0
67.0
R/RO = 0.70
-10 0 10 20 30
INSTANTANEOUS AXIAL VELOCITY, m/sec
76-06-282-30
67
-------
AXIAL DIRECTIONAL INTERMITTENCY DISTRIBUTIONS
FIG, 32
(A) ZERO SWIRL, 3.8 ATM
O
O
Q.
O
<
OC
1.0
0.5
0.5
-1.0
0.5 1.0 1.5
AXIAL DISTANCE, X/D
(B) SWIRL = 0.3, 3.8 ATM
1.0
0.5
O
K
O
V)
O
a.
| -0.5
OC
-1.0 L
0.5 1.0 1.5
AXIAL DISTANCE, X/D
(C) SWIRL = 0.3,1 ATM
O
I
CO
O
Q.
Q
<
OC
0.5 1.0 1.5
AXIAL DISTANCE, X/D
I
2.0
2.0
2.0
76-07-158-6
-------
AXIAL DIRECTIONAL INTERMITTENCY DISTRIBUTIONS
FIG. 32
O
*
1.0
0.5
(D) SWIRL = 0.6, 1 ATM
55 0
O
a
_i
<-0.5
O
-1.0
0.5 1.0 1.5
AXIAL DISTANCE, X/D
I
2.0
1.0
O
CC" 0.5
O
o
a.
_i
<-0.5
O
(E) SWIRL = 0.6, 1 ATM, 0.2/1 INJECTOR
_ 60%
-1.0
c
Qx,io%
)
0
i
.5
1
I
.0
1
I
.5
2
I
.0
AXIAL DISTANCE, X/D
76-07-158-7
-------
probability of reverse flow into the fuel port and in the region behind
the backward facing step is high. Since the directional intermittency
never exceeded 90 percent, it can also be concluded that there are sig-
nificant spatial and temporal motions in the region of the time-averaged
recirculation zone.
To obtain additional insight into the relative magnitudes of the
small-scale and large-scale turbulent motions, consider the possible
sources of the total velocity fluctuations (UT) , namely, the small-scale
turbulent fluctuations associated with forward and reverse flow (u1 and
UI?EV' respectively) and the additional large-scale source due to sign
changes of mean velocity (U-Ujyjy) at the point in question.
Thus,
uVW = f[axu^EV,(l-ax)u',nx(u-DREV)] (7)
where (* is the percentage of time the mean flow is upstream and n^ is
the frequency of mean flow reversal. Assuming similarity of character
of the small-scale turbulence associated with forward and reverse flow,
the instantaneous velocity (U) may be expressed as
where u represents the large-scale fluctuations. With the assumption
that the small-scale and large-scale fluctuations are uncorrelated, i.e.,
u'u = 0,
crT2= U2-02= u'2+u
(9)
That is, the total mean square fluctuation level is the sum of the small
and large-scale contributions. An indication of the relative contribu-
tions of the large-scale and small-scale fluctuations to the total rms
velocity fluctuations may be obtained from Fig. 29. In this figure,
which shows the radial profiles of the mean and rms velocities and
directional intermittency, it is apparent that although the peak total
rms velocity fluctuations occur in the regions of maximum mean velocity
gradient, in the central and corner recirculation zones, where the mean
velocity gradients are small, significant fluctuations are present. It
is apparent that these fluctuations cannot be attributed solely to small-
scale gradient transport and that they must be associated with the large-
scale fluid motions. Two mechanisms for occurrence of the large fluctu-
ations can be envisioned -- (l) large-scale motions of fluid through the
70
-------
sample volume, as evidenced by high values of directional intermittency
and (2) large-scale convective transport of smaller-scale turbulence '
kinetic energy into the sample volume. It appears that these two
mechanisms can account for as much as 50 percent of the total axial rms
velocity fluctuations in the initial mixing regions.
Turbulent Shear Stress Measurements
At present time, efforts are being made at UTRC to develop numerical
methods for predicting turbulent flow behavior. The rate of development
of computational fluid dynamics, especially for combusting flow fields,
is hampered by the present levels of understanding of the physics of
turbulence and the structure of turbulent flows, which are required to
model the turbulent correlations which result from time-averaging the
Navier-Stokes equations.
Although significant progress has been made, the computation of
turbulent flows is still only a practical proposition when the turbulent
correlations (u'w1 for example) which arise from the process of time-
averaging the Navier-Stokes equations can be modeled by simple mixing
length or turbulent kinetic energy assumptions. It is difficult to
assess the potential of existing turbulence models due to the lack of
turbulent structure information which could be used to assess the
validity of present models or guide the formulation of improved models
to account for turbulent nonequilibrium effects present in these com-
busting flow fields.
The feasibility of measuring the turbulence shear stress has been
investigated in the present study. It is anticipated that more detailed
higher order correlation measurements will prove useful in future tur-
bulence model development. Measurements were made of the axial-
tangential turbulent velocity correlation (u'w1) at an axial distance
of 0.65 cm from the injection plane (X/D = 0.05) for the case of 0.3
swirl at 3.8 atm. Velocity probability density measurements obtained
at two different fringe orientations at R/RO = 0.8 are shown in Fig. 33.
The velocity components sensed in these two orientations are 1//2 (u +
w). Thus, the difference between the two measured variances yields
twice the turbulence shear stress, u'w'. Figure 3^ shows the measured
turbulent shear stress correlation coefficient (u'w'/auaw) obtained at
selected radial locations across the combustor. These data indicate^
that the maximum correlation coefficient is between O.k and 0.5, typical
of values observed in the wall region of boundary layer flows, although
there are substantial variations in the regions surveyed. These large
variations suggest that detailed measurements will have to be made
71
-------
PROBABILITY DISTRIBUTION FUNCTIONS FOR SHEAR
STRESS MEASUREMENTS
SWIRL = 0.3. 3.8 ATM, Va/Vf = 21
NUMBER OF OCCURRENCES 500
X/D - 0.05, R/Ro = 0.8
-40 -30 -20 -10 0 10 20
INSTANTANEOUS VELOCITY, (m/sec)
30
FIG. 33
AXIAL-TANGENTIAL VELOCITY
CROSS CORRELATIONS
SWIRL- 0.3, 3.8atm,Va/Vf = 21
76-06-282-22
FIG. 34
o
-1.0
X/D = 0.05
_L
_L
-0.5 0 0.5
RADIAL POSITION, R/RO
72
1.0
76-06-282-23
-------
throughout the flow field before the adequacy of existing two-equation
turbulence models can be assessed.
FUEL INJECTOR EROBING
The results of the present investigation and observations reported
previously in Ref. 7 indicate that for several of the combustor configu-
rations tested and, in particular, for flows without swirl, there may be
significant penetration of the recirculation zone into the fuel injector
duct. Therefore, in order to determine the initial conditions for the
combustor flow analysis, a radial traverse of the fuel jet was made
within the injector to obtain the time-mean fuel velocity distribution
upstream from the penetration region of the recirculation zone.
Measurements were made using a 0.95 cm diameter hemispherical-nose
pitot probe (Fig. 7) and differential pressure transducers having an
accuracy of 0.06 percent of the full scale range + 0.03 atm. The
results of a traverse made at a location approximately one injector
diameter upstream of the injection station are shown in Fig. 35 and
indicate a relatively flat velocity profile having a mean value of
approximately 5 m/sec, which is in approximate agreement with the cal-
culated value of k m/sec.
73
-------
MEAN AXIAL VELOCITY DISTRIBUTION WITHIN
FUEL INJECTOR I
FIG. 35
U [m/secl
X/D =-0.8
1.0
0.5 0 0.5
RADIAL POSITION, R/R,NJ
1.0
76-06-282-28
-------
SECTION IV
DISCUSSION OF RESULTS
Variation of inlet air swirl, combustor pressure and air/fuel velocity
ratio produces major changes in the time-mean flow field within the tur-
bulent flame burner which significantly influence energy release and
pollutant formation. Measurements in the initial mixing region in the
burner indicate that for the case of zero swirl at 3.8 atm there is a
large centrally-located time-average recirculation zone. With the
introduction of swirl (3=0.3), a much smaller toroidal-shaped time-
average recirculation zone is present. At 1 atm, increasing the swirl
number from 0.3 to 0.6 results in an increase in the volume of the toroi-
dal recirculation zone. Associated with these changes in the recircula-
tion zone, are changes in the temperature and species concentration dis-
tributions in the initial regions of the burner. The temperature peaks
in the nonswirling flow are broader and the maximum temperature is
lower than in the swirling flow, suggesting a more diffuse combustion
zone in the nonswirling flow. The radial species concentration gradients
support this observation. Energy release rates, as evidenced by
axial temperature gradients and by the hydrocarbon burn-out rate, are
larger in the swirling flow. The higher temperatures associated with
the swirling flow result in higher NO formation rates. Increasing the
swirl number from 0.3 to 0.6 at 1 atmosphere results in an increased
energy release rate, as evidenced by locally higher temperatures and
increased hydrocarbon burn-out rates. The higher temperatures associated
with the 3=0.6 flow result in an increased NO formation in the initial
regions of the flow.
A decrease in pressure from 3-8 to one atmosphere results in a
significant decrease in NO emissions and a modest decrease in THC emis-
sions for S=0.3. A major portion of this decrease may be attributed
to a decrease in residence time. However, measurements in the initial
regions of the flow show that the time-average flow field structure
changes as combustor pressure is varied. Decreasing the pressure results
in a substantial decrease in the recirculation zone volume and in the
75
-------
energy release rate, as evidenced by the mean temperature distribution
and lower hydrocarbon consumption rate. The combined effect of lower
temperature and reduced combustor residence times result in lower NO
formation at one atmosphere.
A decrease in air/fuel velocity ratio from 21 to 0.17, for an inlet
air swirl number of 0.6 and atmospheric pressure, results in a reduction
in NO emissions and an increase in THC emissions. This decrease produces
a significant change in recirculation zone geometry and location with
respect to the fuel injector, and a reduction in energy release rates
and peak temperatures, resulting in lower NO formation rates.
There also are significant changes in the turbulent structure of the
flow field with variations in inlet air swirl, combustor pressure and
air/fuel velocity. In all of the flows investigated there appear to be
substantial large-scale contributions to the total rms turbulent velocity
field. In the initial mixing regions of these flows, the total rms
velocity fluctuations can significantly exceed the local mean velocity.
In the cases of high air/fuel velocities, the large-scale fluctuations
are the result of low frequency movement of fluid in the central portion
of the flow. Reverse movements produce bulges upstream which extend into
the outer shear layer regions and result in rapid breakup and mixing as
the fluid is convected downstream. Downstream movements create voids
which draw outer shear layer fluid in towards the center of the combus-
tor. This "pumping" action, which is indicated by the directional inter-
mittency, is a function of the stability as well as the size of the
recirculation zone.
These large-scale motions have significant effects on the species
concentration, velocity fluctuations and chemical reaction rates. For
example, the initial mixing region of the nonswirling flow is charac-
terized by rapid apparent mixing of fuel and air and a relatively low
energy release rate. However, it must be borne in mind that conventional
time-average probe measurements provide no information on the scales at
which the fuel, air and combustion products are being mixed. If fuel and
air were well-mixed on a small scale, energy release rates, as evidenced
by hydrocarbon burn-out rates and temperature should be large. But, in
fact, the reverse is the case. This suggests that there are large-scale
inhomogeneities in the flow and that the mixing has been accomplished by
large eddies which have been generated by the pumping action of the
recirculation zone. Such eddies,when convected past the sampling probe,
would present the time-averaged appearance of a well-mixed flow although
the scales involved actually result in locally inhomogeneous mixtures
which tend to result in lower THC burn-out and overall energy release.
76
-------
This hypothesis is qualitatively supported by both directional inter-
mittency contours and the high-speed motion pictures which show substan-
tial large-scale fluctuations in the zero swirl case.
The introduction of inlet air swirl tends to stabilize the recircu-
lation zone, as evidenced by the directional intermittency data and the
high-speed films. One would expect, therefore, a reduction in large-
scale inhomogeneities in the initial mixing regions, thereby preserving
the separation of the reactants for a greater axial distance. Thus, mix-
ing should occur at a more clearly defined air/fuel interface. This
smaller scale (high shear layer) mixing results in localized chemical
reaction and produces locally high temperatures and NO formation rates
at the air/fuel interface. Increased combustor pressure, at constant
swirl number, also tends to increase recirculation zone stability, there-
by reducing large-scale mixing. One would therefore anticipate increased
segregation of the fuel and air streams and locally high energy release
rates in the initial mixing region and, indeed, this is the case.
Changing the air/fuel velocity ratio from 21 to 0.17 for an inlet
air swirl number of 0.6 at atmospheric pressure had a profound effect on
the combustor flow field. The high-speed motion pictures and the total
hydrocarbon contours indicate that the central fuel jet spreads very
rapidly. This enhanced spreading rate is primarily due to the radial
static pressure gradients induced by swirl and the large-scale inter-
action, i.e., entrainment of the fuel jet into the toroidal recircula-
tion induced by the inlet air swirl vanes. This later unsteady interac-
tion produces fuel jet motion about its time-averaged location, as evi-
denced by bi-modal fuel-velocity probability density distributions.
Thus, in this case three mechanisms are responsible for rapid time-
averaged fuel spreading rates.
Hence, it is felt that many of the changes in flow field structure
observed in the present study may be related to the interaction between
large-scale turbulent fluctuations, associated with the unsteadiness of
the recirculating flow or the fuel jet, and small-scale fluctuations
associated with shear-generated turbulence.
The large-scale fluctuations, discussed above, result in significant
departures from Gaussian turbulence and isotropy in the initial regions
of the burner. The intensity and inferred scale of the fluctuations in
these regions suggest that existing turbulence models, which utilize
local mean gradients, may not adequately represent turbulent transport
in the combusting flows studied. Comparison of the experimental results
obtained in the present investigation with predictions of the reacting
flow field using the CRISTY computer code will serve to evaluate the
77
-------
analytical procedures and turbulence models. Further evaluation of
existing turbulence models may be obtained by making detailed measurements
of the scale of turbulence and of turbulent shear stress and kinetic
energy throughout the reacting flow field with a laser velocimeter.
78
-------
SECTION V
RECOMMENDATIONS
The experimental investigations carried out under EPA Contracts
68-02-1092 and 68-02-1873 have shown that variations in inlet conditions,
e.g., inlet air swirl, inlet geometry and fuel injection geometry, pro-
duce major changes in the mean flow field, including recirculation zone
geometry and local fuel/air distributions, within a confined turbulent
flame burner which result in subsequent changes in pollutant formation
and destruction. In addition, it was found that turbulence significantly
influenced mean flow field structure and pollutant formation. In the
vicinity of the time-mean recirculation zones, the scale and intensity
of the turbulent fluctuations were sufficiently large so as to suggest
that turbulence models, which utilize local mean gradients, may not accu-
rately represent turbulent transport in these regions of the combustor.
The present data base is inadequate to permit definitive correlations
of mean flow field structure and pollutant emissions with burner inlet
conditions. Such correlations would provide useful "rules-of-thumb" to
the combustion designer in the selection of operating conditions to give
maximum efficiency while minimizing pollutant emissions. In addition to
obtaining the mean flow field structure, it is necessary to obtain infor-
mation on the turbulent structure of the flow. The combined time-mean
and turbulent flow field data will permit an assessment of the effects
of turbulence on mean flow field structure to be made and will be useful
in evaluation and development of turbulence models employed in analytical
procedures for predicting reacting flows.
A logical extension of the present experimental effort would be a
further investigation of the effects of burner inlet conditions and geo-
metry on the mean and fluctuating flow field structure in a gaseous
fueled axisymmetric turbulent diffusion flame burner and the subsequent
effects on pollutant formation and destruction. Two major modifications
in the combustor geometry employed in the present and previous studies
79
-------
should be made. The structure of the highly confined flames investigated
under Contracts 68-02-1092 and 68-02-1873 is significantly influenced by
the close proximity of the combustor walls to the initial mixing and
reaction zone, and the flames tend to be quite long. While the basic
fluid dynamic and chemical phenomena occuring in these confined flames
are the same as in practical combustion devices, the flame geometry dif-
fers considerably from that found in most industrial flames. To permit
investigation of more realistic flame geometries, the ratio of combustor
diameter to injector diameter should be increased to approximate ratios
found in practical combustion equipment. As the second modification,
the injection section would be altered to permit changes in fuel and air
injector geometry and to provide an expanded range of inlet air swirl.
Detailed measurements of the mean flow field, (velocity, temperature and
species concentration), the turbulent structure of the flow (turbulent
intensity and scale, shear stress and kinetic energy), pollutant emis-
sions and heat transfer to the combustor walls would be measured as
burner inlet conditions are varied. Changes in the mean flow field and
subsequent variations in pollutant emissions would be correlated with
changes in inlet conditions. Special emphasis should be placed on
measurement of the inlet conditions (including mean and rms axial, tan-
gential and radial velocities and mean temperature). State-of-the-art
instrumentation should be employed, especially if required information
on the turbulent structure of the flow is to be obtained. It is highly
desirable to use optical diagnostic techniques whenever possible to
avoid problems associated with probe techniques (flow disturbance, sample
perturbation, ill-defined averaging) in turbulent structure of the flow
can be measured using laser velocimetry.
Laser Raman scattering and laser fluorescence techniques for mea-
surement of temperature and species concentration in practical combustion
geometries are still in a developmental stage. However, as these techni-
ques are refined, they should be incorporated into the experimental pro-
gram. Until then, it still will be necessary to use probes to measure
temperature, species concentration and the chemical composition of parti-
culates.
Precise definition of the experimental program would be made follow-
ing a comparison of results of the on-going combustor modeling effort
(Ref. 12) with existing experimental data. A preliminary comparison of
the experimental and analytical results indicate that more detailed
information on the turbulent structure and inlet conditions are required
to assist in development of analytical combustor models. Hence, the
principal objectives of the proposed gaseous fuel tests would be:
80
-------
(1) To provide an expanded set of test cases to be used to evaluate
and develop both empirical and detailed analytical procedures
for predicting pollutant emissions.
(2) To provide experimental data which can be used for
assessment and development of turbulence models utilized in
the above analytical proceures.
(3) To provide additional data for development of phenomenological
correlations between pollutant emissions and burner inlet con-
dit ions.
A close coupling of the experimental program with the modeling effort
is required if maximum benefit from both programs is to be derived.
81
-------
APPENDIX A
LASER VELOCMETER STATISTICAL ERRORS AND BIASING
Statistical confidence levels within stated error limits in the
determination of both the mean and variance of any quantity with a
Gaussian probability variation may be defined according to Ref . 2h as
Sx
Error = P (lx-/8l)
-------
fD X fD
since
4
(15)
we see that
u
The confidence level for the standard deviation may be written
as
2|Sf-o-f|
(17)
Fow since sf_and fp are functions of the same random variable (fyj, the
error in au/U is the sum, not the square root of the sum of the squares
of each error, i.e., the confidence level in au/U = YM + YT-
For normal distribution functions, confidence levels may be
calculated using the following table.
Y
0.5
0.675
0.68
1.00
0.9
1.6
0.95
1.96
0.
2.33
0.99
2.57
For example: if at a particular location, the local turbulence level
was 1 percent and 100 instantaneous velocities were measured, 50 percent
of the mean measurements would be in error by less than 0.0675 percent of
the true value. Whereas only 1 point in a hundred would be in error by
more than 0.257 percent.
The velocity data presented in this report were obtained with seeded
air flow. To evaluate biasing errors which might result from seeding only
83
-------
the air stream, a limited number of axial velocity measurements were
made using three seeding techniques (1) seeded air stream, (2) seeded
fuel stream and (3) seeded fuel and air streams. These measurements
were made immediately downstream from the large diameter fuel injector
(x 10 = 0.16). It is in this initial mixing region, with high measured
velocity fluctuations where the greatest probability of biasing errors
exist. In these tests, the porous disk was removed from the fuel injector
to permit seeding of the fuel stream. The mean axial velocities measured
using the three different seeding techniques are shown in Fig. 36 for
the combustor operating at atmospheric pressure with an inlet air swirl
number of 0.3. These data show that there are no significant differences
between the velocities measured using the different seeding techniques
except in the outer regions of the flow where velocities measured with
fuel seeding were low. Comparison of the probability distribution
functions measured using the three seeding techniques, Fig. 375 shows
that near the combustor center line the measured pdf does not depend on
seeding technique. However, as expected, in the outer regions of the
flow the pdf measured with fuel seeding differs from that obtained when
both the fuel and air streams are seeded. These data indicate that for
the present experiments, valid velocity data were obtained using air
seeding.
-------
COMPARISON OF MEAN AXIAL VELOCITY PROFILES MEASURED
USING DIFFERENT SEED TECHNIQUES
SWIRL = 0.3, 1 ATM, Va/Vf = 21
FIG. 36
1.0
QFUELSEED
FUEL & AIR SEED
OAIRSEED
U(m/sec) T100 X/D=0.34
-0.5 0 0.5
RADIAL POSITION, R/Ro
-------
FIG. 37
EFFECT OF SEEDING TECHNIQUE ON PROBABILITY
DISTRIBUTION FUNCTIONS
j
O
o
O
UJ
CO
200
R/R0 = 0.013
O FUEL SEED
FUEL & AIR SEED
O AIRSHED
-20 0 20 40 60 80
INSTANTANEOUS VELOCITY,(m/sec)
500
UJ
400
oc
K
O
g 300
u.
O
UJ 200
o
D
100
R/R0 - -0.69
O FUEL SEED
FUEL & AIR SEED
-40 0 40 80 120 160
INSTANTANEOUS VELOCITY,(m/sec)
76-07-158-13
86
-------
APPENDIX B
FUEL COMPOSITION
Table B-l. NATURAL GAS COMPOSITION
Species
c%
C2H6
co2
C3H6, C3Hg
02 + Ar
N2
i - C^Q
n-°UHio
Mole Percent
Trailer No. 1
96.!^
2.16
0.65
0.21
0.06
0.69
0.05
0.0*4-
No. 2
97.1
1.80
0.56
0.15
0.06
0.26
0.03
0.03
Wo. 3
96.5
2.15
0.55
0.29
0.10
0.30
0.07
0.0*4-
Average
96.58
2.03
0.59
0.22
0.07
O.te
0.05
o.o^t-
87
-------
APPENDIX C
COMBUSTOR HEAT BALANCE
A thermal balance was performed on the combustion system for a
typical operating condition to evaluate the magnitude of the heat
transferred to the combustor walls and to provide a check on the accuracy
of the exhaust gas temperature measurements. The heat transferred from
the system was determined from measurements of the flowrate and the tem-
perature rise of the cooling water. The results of the heat balance are
summarized below:
Test No. 7
Swirl No. =0.6
P = 1 atm
T . =750 °K
air
Natural Gas - Air
Air/Fuel Velocity Ratio =0.18
Equivalence Ratio = 0.90
tn . = 0.137 kg/sec
air
Air Heater:
Fuel:
QOUT
Combustion Products:
Cooling Water:
Unreacted Fuel:
15.9 kg cal/sec
88.69 kg cal/sec
79.73 kg cal/sec
Ik.77 kg cal/sec
2.23 kg cal/sec
10U.59 kg cal/sec
96.73 kg cal/sec
"IN
= .075
*IN
-------
APPENDIX D
TEMPEMTUEE DISTRIBUTIONS: TABULATED DATA
-------
Table D-l. TEMPERATURE DISTRIBUTIONS FOR TEST NO. 1
CH^-Air
$ = 0.91- .01
Pressure = 3.9 + 0.1 atm
R/R0
-0-75
-0.62
-0.50
-0.37
-0.25
-0.12
0.00
0.13
0.25
0.38
0.50
0.63
0.75
0.88
1.00
Inlet Air Swirl No. = 0
Inlet Air Temperature = 751 + 8 °K
Air-Fuel Velocity Ratio = 20.7 + 0.2
Temperature, °K
X/D=0.3^
922
1011
1305
1371
1282
1237
1138
1060
993
109^
1205
1096
779
673
597
0.6o
1673
1753
1685
1659
1617
1518
1399
1328
1198
1255
1392
1733
Ihh6
1002
677
1.73
1889
1815
17^+
1722
1653
1631
1672
1811
1923
2028
2006
1879
1670
lUij-9
821
1.99
2079
1896
1868
1771
1736
1767
1861
1855
1952
2101
1971
1781
1528
12^-U
760
lk.36
1888
1900
1935
1985
1935
1980
1978
2095
2106
2065
1995
1955
1656
1565
1056
90
-------
Table D-2. TEMPEMTURE DISTRIBUTIONS FOR TEST NO. 3
CHv-Air Inlet Air Swirl No. = 0.3
$ = 0.91 - .01 Inlet Air Temperature = 7^3+ 9 °K
Pressure = 1.0 atm Air-Fuel Velocity Ratio = 21.7 + 0.5
Temperature, °K
.0.63
0.55
0.50
0.45
0.37
-0.25
-0.12
0.00
0.07
0.12
0.28
0.38
0.49
0.58
0.63
o 69
\mS m V^ J
0.75
X^ » } ^
0.88
1.00
x/D=0.34
1416
1545
1541
1481
1425
1416
1453
1479
i486
1473
1449
1359
1244
1144
1029
752
0.6o
1477
1502
1565
1634
1647
1626
1618
1560
1490
1431
1300
1160
1029
772
1.99
2034
1994
1897
1845
1830
1850
1911
1950
1955
1901
1777
_ _ _
1592
1446
1075
3.38
2061
2017
1927
1850
1777
1762
1798
1824
1850
1934
1993
1979
1912
1772
1625
1420
4.77
2102
2069
1958
1890
1828
1808
1850
1870
i860
1939
2026
2018
1925
i
1741
1625
1126
13.50
1953
2015
2013
1975
1960
1975
2030
2061
2032
1985
1925
1760
1620
1500
14.95
1734
1815
1837
1852
1840
1854
1884
1880
1846
1768
1679
---
- -
1600
1430
1220
91
-------
Table D-3. TEMPERATURE DISTRIBUTIONS FOR TEST NO.
CH^-Air Inlet Air Swirl No. = 0.3
$ = 0.91 ± 0.01 Inlet air Temperature = 7^9 + 7 °K
Pressure = 3.7 + 0.1 atm Air-Fuel Velocity Ratio = 20.5 +O
R/R Temperature, °K
-0.63
-0.50
-0.37
-0.25
-0.12
0.00
0.13
0.25
0.38
0.50
0.63
0.75
0.83
1.00
X/D=0.3^
1657
ll*8o
1220
1121*
1010
1015
1085
1235
1317
1230
10^0
825
695
626
0.60
1703
1559
11*51*
1373
1276
1361
ll*68
151*0
1582
1608
1538 .
1168
877
618
1.73
2081
2059
1975
1836
1776
1771*
1989
2105
2133
2116
1992
1769
1510
975
1.99
2133
2096
1950
1821
171*7
181*1
1978
2128
2189
2171
1993
1783
ll*79
935
ll*.36
1826
1863
1870
1832
1730
1761*
1836
1888
1875
1827
1757
1685
1626
ll*89
92
-------
Table D-U. TEMPERATUEE DISTRIBUTION FOR TEST NO. 6
CHjj-Air
i = 0.91 - .01
Pressure = 1.0 atm
R/RO
-0.63
-0.55
-0.50
-0.l»-5
-0.37
-0.25
-0.12
0.00
0.06
0.13
0.25
0.38
0.50
0.58
0.63
0.70
0.75
0.80
0.88
1.00
Inlet Air Swirl No. =0.6
Inlet Air Temperature = 750 + 10 °K
Air-Fuel Velocity Ratio =21.3 +0.5
Temperature, °K
1089
102U
1105
1537
0.60
1516
1591
1686
1700
1635
1597
1615
1622
1630
1600
1503
1352
1227
1079
2233
2181
20^6
1935
1933
2010
862
810
2235
22hO
2156
1928
1578
1231!
778
1.99
2237
2150
200^
1923
1925
2013
2136
2203
2221
2168
1920
198U
1980
1915
1835
1671
1336
1880
1925
1995
1997
1930
1856
1762
1691
1650
93
-------
Table D-5. TEMPERATURE DISTRIBUTION FOR TEST NO. 7
CH^-Air Inlet Air .Swirl Wo. = 0.6
Y= 0.91 - 0.01 Inlet Air Temperature = 7*1-6 + 7 °K
Pressure = 1.0 atm Air-Fuel Velocity Ratio = 0.17 + 0.01
Temperature, °K
-0.6*1
-0.50
-0.37
-0.25
-0.12
0.00
0.13
0.25
0.38
0.50
0.63
0.75
0.88
1.00
X/D=0.76
1610
1326
1193
1115
109*1
11*4-7
1*109
1667
1666
1632
1*129
1163
962
652
1.01
1570
1286
1177
11*19
11*4-9
11*4-7
1187
1*4-00
1565
1566
1399
1135
925
620
2.15
187*4-
1750
1616
1525
1*^88
1516
1591
1689
1797
18*4-3
1637
132*1
1086
662
2.14-1
1817
17*4-1
1669
160*4^
1588
1616
1718
1830
2005
2055
1792
1*4-51
1177
700
1*4-. 77
2050
2085
1850
1306
1230
1270
2135
22*4-5
2020
1800
1615
l*4-*4-8
-------
APPENDIX E
SPECIES CONCENTRATION DISTRIBUTIONS: TABULATED DATA
95
-------
Table E-l. SPECIES CONCENTRATION DISTRIBUTIONS FOR TEST NO. 1
CH^-Air Inlet Air Swirl No. = 0
$ = 0.91^0.01 Inlet Air Temperature = 751±8°K
Pressure = 3-9 ± 0.1 atm Air-Fuel Velocity Ratio = 751±8°K
R/R0 N°> Ppm
-0.85
-0.66
-0.1*1*
-0.25
-o.ok
0.16
0.38
0.58
0.80
R/RO
-0.85
-0.66
-0.25
-o.ok
0.16
0.38
0.58
0.80
R/RO
-0.85
-0.66
-0.1*5
-0.25
-o.oi*
0.16
0.38
0.58
0.80
X/D = 0.08
1
0
7
9
20
11
8
1
0
X/D =0.08
_
2
-
2
X/D = 0.08
0.00
0.00
1*.20
3-50
5.35
3.10
2.00
0.50
0.00
1.21
1*1
72
55
31
33
83
133
135
59
NOX, ppm
1.21
50
10U
168
88
CO, Mole f0
1.21
0.80
2.90
6.22
8.1*0
9.65
9.75
7.25
2.90
0.51
1.1*7
90
83
53
27
19
51*
108
120
69
l.H-7
122
11*2
200
90
1.U7
0.80
2.86
6.35
8.20
9.50
9.65
7.25
2.35
0.35
2.81*
198
185
115
81*
110
166
235
2Uo
158
2.8>4
2l*5
_
285
182
2.81*
1.82
6.15
9.80
11.3
Il.k
10.00
6.80
2.65
O.U8
-------
Table E-l. SPECIES CONCENTRATION DISTRIBUTIONS
(continued)
R/R0
-0.85
-0.66
-0.44
-0.25
-0.04
O.l6
0.38
0.58
0.80
R/R0
-0.85
-0.66
-0.44
-0.25
-0.04
0.16
0.38
0.58
0.80
R/RO
-0.85
-0.66
-0.44
-0.25
-0.04
0.16
0.38
0.58
0.80
C02, Mole %
X/D = 0.08
0.05
0.05
2.00
1.80
2.50
1.70
1.35
o.4o
0.05
X/D = 0.08
18.0
17,9
1.82
0.18
0.19
O.ll
0.13
14.4
20.2
X/D = 0.08
o.oia
0.022
>10.0
>10.0
>10.0
>10.0
>10.0
>10.0
0.011
1.21
3.95
5.30
4.10
3.48
3-50
4.32
5.60
5.95
3.51
02, Mole %
1.21
12.9
7.30
4.40
3.53
2.50
1.43
2.30
6.60
14.3
THC, Mole %
1.21
2.05
5.05
7.55
9.45
8.1+5
5.85
4.25
2.75
1.85
1.47
5.80
6.70
4.95
3.50
3-45
4.38
6.10
6.55
3.30
1.1*7
10.1
5.55
3.79
3.76
2.76
1.67
1.72
6.55
14. 7
1.47
0.50
2.31
7-45
9.69
9 = 30
6.18
U.oU
0.97
0.19
2.8U
7.15
6.75
5.05
if. 40
4.70
5.60
6.65
6.65
4.92
2.84
6.40
1.95
1.05
0.75
0.35
0.53
1.95
6.50
12.3
2.84
0.79
2.56
4.52
5.08
3.94
2.94
1.50
0.51
0.10
97
-------
Table E-2. SPECIES CONCENTRATION DISTRIBUTION FOR TEST NO. 3
Inlet Air Swirl No. = 0.3
§ = 0.9 - 0.01 Inlet Air Temperature = 7l*l*±9°K
Pressure = 1.0 atm Air Fuel Velocity Ratio = 21.7±0.5
R/R0 NO, ppm
-0.85
-0.66
-O.U6
-0.25
-O.Ol*
0.17
0.37
0.57
0.80
R/R0
-0.85
-0.66
-0.1*6
-0.25
-O.cU
0.17
0.37
0.57
0.80
R/R0
-0.85
-0.66
-0.1*6
-0.25
-o.oi*
0.17
0.37
0.57
0.80
X/D = 0.08
1
5
5
3
3
7
Ik
10
1
X/D = 0.08
18
29
35
13
X/D = 0.08
0.607
1.1*00
2.150
1.850
1.600
2.31+0
3.025
1.810
0.680
1.47
15
39
70
88
87
85
73
1*2
17
NOV ppm
1.1*7
^3
87
78
^7
CO, Mole <
1.1*7
0.9U
1.75
3.^0
5.30
6.UO
5.70
l*.20
1.82
0.92
2.86
27
66
102
100
87
93
103
79
27
,2.86
57
-
133
69
£
2.86
0.60
2.53
5.65
8.00
8.79
8.66
7.08
2.85
0.72
l*.25
U5
97
121
109
100
101
125
107
**7
l*.25
82
-
172
81*
i*.25
0.70
2.75
7.00
9.25
9.80
9.60
7.80
2.51
0.68
-------
Table E-2. SPECIES CONCENTRATION DISTRIBUTIONS
(continued)
R/RO
-0.85
-0.66
-0.46
-0.25
-0.04
0.17
0.37
0.57
0.80
-0.85
-0.66
-0.46
-0.25
-o.ok
0.17
0.37
0.57
0.80
-0.85
-0.66
-0.46
-0.25
-o.o4
0.17
0.37
0.57
0.80
X/D = 0.08
1.85
1.80
2.05
1.75
1.55
1.87
2.85
2.90
1.30
X/D = 0.08
15.5
lit.O
5.4o
3-05
1.80
2.03
2.65
11.6
16.5
X/D = 0.08
1.66
6.00
>10.0
XLO.O
>10.0
XLO.O
>10.0
8.31
2.23
C02, Mole <
1.47
4.62
6.3
7.15
6.80
6.17
6.58
7.04
6.44
4-79
02, Mole %
1.47
11.3
7.10
3.35
1.45
1.00
1.25
2.65
6.60
10.8
THC, Mole af
1.1+7
1.73
2.50
4.60
7.40
9-50
7.80
6.00
2.80
1.64
*
2.86
5.56
7-42
7.10
6.00
5.45
5.63
6.69
7.42
5.79
2.86
10.1
5.37
1.51
0.57
0.38
0.42
1.03
4.06
8.50
a
2.86
0.57
0.95
2.45
4.60
6.35
5.75
3-55
1.31
0.55
4.25
6.18
7.54
6.71
5.58
5.10
5.27
6.27
7.90
6.33
4.25
8.95
3.75
1.05
0.50
0.43
0.47
0.77
3.37
8.66
4.25
0.27
0.55
1.80
3.10
4.28
4.00
2.35
0.5^
0.18
99
-------
Table E-3. SPECIES CONCENTRATION DISTRIBUTIONS FOR TEST NO.
CH^-Air
$" = 0.91 i 0.01
Pressure = 3-7-0.01 atm
R/Ro
-0.85
-0.66
-O.U5
-0.26
-o.oi*
0.16
0.38
0.58
0.80
-0.85
-0.66
-0.1+5
-0.26
-0.01+
0.16
0.38
0.58
0.80
R/Ro
-0.85
-0.66
-O.U5
-0.25
-0.0*4-
0.16
0.38
0.58
0.80
Inlet Air Swirl No. = 0.3
Inlet Air Temperature = 7*4-9- 7°K
Air-Fuel Velocity Ratio = 20.5^0.4
NO,
X/D = 0.08
0
2
8
7
10
6
U
0
0
X/D = 0.08
1+
5
2
X/D = 0.08
0.00
0.28
3.58
1.70
1.86
1.10
1.23
0.19
0.00
1.21
28
86
123
116
71
88
ll+O
138
53
NOx, ppm
1.21
66
90
-
58
CO, mole %
1.21
0.9
3.3
8.6
11.1+
12.0
11.9
10.3
6.2
1.5
1.U7
53
110
119
130
69
93
152
131
55
1.1+7
89
133
-
-
1.1+7
1.17
1+.02
8.55
10.8
11.8
11.5
10.2
6.1+1
1.78
2.60
125
280
327
205
65
93
305
292
165
2.60
138
_
305
196
2.60
1.8
5.1
8.0
10.3
13.2
13.0
10.6
6.8
2.2
100
-------
Table E-3. SPECIES CONCENTRATION DISTRIBUTIONS
(continued)
R/R0
-0.85
-0.66
-0.45
-0.25
-0.04
0.17
0.38
0.58
0.80
R/Ro
-0.85
-0.66
-0.45
-0.26
-0.04
0.17
0.38
0.58
0.80
R/R0
-0.86
-0.66
-0.45
-0.26
-0.03
0.16
0.38
0.58
0.80
C0p? mole %
X/D = 0.08
o.oo
0.01
1.82
1.35
1.35
1.07
1.08
0.10
o.o4
X/D = 0.08
20.8
19. 4
1.3
0.4
0.1
0
o.4
8.3
20. 4
X/D = 0.08
o.ok
3.09
> 10.0
> 10.0
> 10.0
> 10.0
> 10.0
6.73
0.05
1.21
3.1
5.1
5.3
4.8
4.2
4.5
5.5
5-9
3.9
02, mole %
1.21
14.5
7.0
2.3
0.6
0.2
0.2
0.2
3.2
11.2
THC, mole %
1.21
2.74
4.58
4.55
4.55
7.51
6.00
3.43
3.84
4.44
J..47
4.11
5.48
5.16
4.90
4.09
4.46
5.39
5.90
4.26
1.47
11.7
6.0
1.0
0.3
0.1
0.1
0.2
2.8
10.3
1
1.47
2.05
3.92
3.99
3.46
6.76
4.95
2.78
3.61
3.67
2.60
5.6
7.2
7.1
5.6
4.1
4.4
6.4
7.0
6.1
2.60
8.6
2.5
0.6
0.1
0.1
0.1
0.3
1.5
7.6
2.60
0.86
0.62
0.68
1.69
4.32
3.24
0.71
0.44
0.30
101
-------
Table E-U. SPECIES CONCENTRATION DISTRIBUTIONS FOR TEST NO. 6
CH^-Air
i = 0.91 - o.oi
Pressure = 1.0 atm
R/RO
-0.85
-0.66
-0.1*5
-0.35
-0.25
-o.ok
0.17
0.27
0.38
O.k8
0.58
0.67
0.75
0.81
Inlet Air
Inlet Air
Swirl No. =
Temperature
Air-Velocity Ratio =
X/D =0.08
2
3
-
22
-
21
-
19
-
15
6
2
2
1
NO, ppm
1.21
3
17
62
-
89
86
88
-
80
-
31+
-
.
7
1.47
6
29
72
90
92
80
89
92
89
71
k7
20
-
5
0.6
= 750±10°K
21.3-0.5
2.86
30 .
8k
125
-
123
96
102
-
117
_
10
_
_
50
ppm
X/D = 0.08
9
15
1.21
17
k2
I.k7
28
2.86
72
123
R/RC
-0.85
-0.66
-0.45
-0.35
-0.25
-o.ok
0.17
0.27
0.38
0.58 - 70 87
0.67 - -5k
0.75 7 28 -
0.81 - - 29
102
-------
Table E-l*. SPECIES CONCENTRATION DISTRIBUTIONS
(continued)
R/Ro
0.85
0.66
OA5
0.35
0.25
O.Ol*
0.17
0.2?
0.38
0.1*8
0.58
0.67
0.81
R/Rf
-0.85
-0.66
-0.1*5
-0.35
-0.25
-O.Ql*
0.17
0.27
0.38
0.1*8
0.58
0.67
0.81
CO. Mole *
X/D =0.08
o.oi*
0.69
-
3.33
-
3.16
-
2.86
-
2.92
1.1*6
0.1*8
X/D = 0.08
0.06
1.27
_
2.35
_
2.16
_
2.10
MB
2.1*2
1.36
0.1*1*
0.05
1.21
1.29
2.91
5.52
-
8.31*
9.82
9.23
<>
6.58
-
l*.08
_
1.58
C09, Mole <-,
1.21
2.31
5.26
7.22
-
6.65
5.68
6.10
_
7.08
-
6.21*
-
2.80
1.1*7
1.1*1*
2.87
6.05
7-1*7
9.00
10.7
10.0
8.9^
7.57
5.92
l*-5l*
3.20
1.60
i
1.1*7
2.66
5.82
7- 12
6.75
6.26
5.50
5.90
6.2U
6.88
7.05
6.66
5.01
3.oo
2.86
2.25
3.59
6.2U
_
9.26
10.87
9.17
_
8.31*
-
i*.o6
_
2.26
2.86
5.3^
7.1*2
7.1*1*
-
6.21*
5.58
5-73
-
6.63
-
7.98
-
6.1*9
103
-------
Table E-U. SPECIES CONCENTRATION DISTRIBUTIONS
(continued)
R/R
-0.85
-0.66
-O.U5
-0.35
-0.25
-o.oU
0.17
0.2?
0.38
o.kQ
0.58
0.67
o.8l
R/RC
0.85
0.66
0.^5
0.35
0.25
0.0k
0.17
0.38
0.^8
0.58
0.69
0.81
X/D = 0.08
19.8
16. k
-
0.2
-
0.2
_
O.k
H
2.5
9.7
16.2
20.6
X/D = 0.08
_
7-k7
>10.0
>10.0
>10.0
>10.0
>10.0
>10.0
>10.0
>10.0
>10.0
1.0k
0?, Mole %
1.21
13.2
7-80
2.10
-
0.37
0.08
6.10
-
7.08
_
5.01
-
Ik. 00
THC, Mole
1.21
5.71
3.83
2.55
-
3.61*
3.1k
2.95
2.16
-
2.52
-
k.Qo
I.k7
13.2
6.80
1.65
0.75
0.30
0.09
0.10
0.32
0.70
1.85
k.10
7.60
12.6
i
I.k7
k.75
2.kO
1.99
-
2.62
2.86
2.71
1.81*
_
1.9k
2.87
1*.12
2.86
8.50
3.58
0.89
0.20
-
0.07
O.C
-
0.39
-
2.20
-
6.75
2.86
1.83
0.66
-
0.67
1.09
1.1*6
I.k2
1.18
_
0.1*6
-
1.00
10k
-------
Table E-5. SPECIES CONCENTRATION DISTRIBUTIONS FOR TEST NO. 7
CHl+~Air + Inlet Air Swirl No. = 0.6
$= 0.91 - 0.01 Iniet Air Temperature = ?!+(,-, ^
Pressure - 1.0 atm Air-Fuel Velocity Ratio = 0.17±0.01
iyjA0
-0.85
-0.66
-0.1+5
-0.25
-o.dk
0.17
0.38
0.58
0.8o
R/RO
-0.85
-0.66
-0.1+5
-0.25
-0.01+
0.17
0.38
0.58
0.80
-r^ /
R/Ro
-0.85
-0.66
-0.1+5
-0.25
-o.dk
0.17
0,38
0.58
0.80
X/D=0.2l+
2
31+
68
55
0
0
55
67
16
X/D=0.2l+
12
89
-
-
-
-
-
-
^0
X/D=0.2l+
0.1+5
2.60
7.08
6.72
0.1+6
1.1+5
6.81+
6.29
0.1+5
0.50
3
21
^8
16
0
1+
37
23
5
NOV, TDI
0.50
16
58
-
-
-
-
-
-
21
CO, mole
0.50
0.21+
0.70
6.05
If. 58
1.92
2.61
5.36
1+.63
0.27
1.63
1+
32
1+8
1+2
39
l+l
1+6
39
6
am
1.63
27
-
-
-
-
-
-
-
38
*
1.63
0.56
5.69
6.95
6.37
6.30
6.k6
6.97
6.3k
0.81+
1.8
>k
33
1+8
k5
k6
1+6
1+8
36
5
1.8;
26
-
-
-
-
-
-
-
31
1.89
0.53
6.10
7.1+0
7.20
7.22
7.22
i
7-3^
6.33
0.63
105
-------
Table E-5. SPECIES CONCENTRATION DISTRIBUTIONS
(continued)
-0.85
-0.66
-0.45
0.25
-o.o4
0.17
0.38
0.58
0.80
-0.85
-0.66
-0.45
-0.25
-O.o4
0.17
0.38
0.58
0.80
-0.85
-0.66
-0.45
-0.25
-o.o4
0.17
0.38
0.58
0.80
X/D=0.24
0.94
6.66
5.53
4.76
1.44
2.49
4.94
6.58
3-01
X/D=0.24
18.7
6.00
0.62
0.97
3.06
3.18
0.90
1.06
14.8
X/D=0.24
0.06
0.66
XLO.O
XLO.O
XLO.O
XLO.O
XLO.O
4.28
0.02
C02
0.50
16.1
4.48
5.66
3.95
2.47
2.83
4.45
5-97
1.59
02,
0.50
18.1
2-37
1-35
2.67
3-10
2-95
1.70
5.00
17.5
THC,
0.50
0.05
0.20
XLO.O
XLO.O
XLO.O
XLO.O
XLO.O
5.60
0.06
, mole%
1.63
2.30
5.59
4.71
4.31
4.20
4.24
4.54
5.53
3.03
mole'/o
1.63
15.8
2.84
1.23
1.48
1.53
1.44
1.20
2.04
14.3
mole'/o
1.63
0.64
0.82
XLO.O
XLO.O
XLO.O
XLO.O
XLO.O
9.32
8.00
1.89
2.11
5-67
4.81
4.53
4.63
4.64
4.92
5.81
2.63
1.89
16.3
2.39
1.01
1.14
1.13
1.14
1.03
2.60
15.2
1.89
4.40
7.65
XLO.O
XLO.O
XLO.O
XLO.O
XLO.O
8.15
4.35
106
-------
APPENDIX F
MEMS AND RMS VELOCITY DISTRIBUTIONS: TABULATED DATA
107
-------
Table F-l. AXIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO. 1
0%-Air Inlet Air Swirl = 0
§ = 0.91±0.01 Inlet Air Temperature = 751±8°K
Pressure = 3.9+0.1 atm Air/Fuel Velocity Ratio = 20.7±0.2
X/D = 0.052
R/RO u
u
,2
-0.917 ^.30 5.2
-0.788 18.6 6.0
-0.667 21.1 k.l
-0.5^2 15.0 6.1j-
-0.371 0.07 2.9
-0.20U 1.90 2.5
-0.038 2.86 2.8
0.1U6 3.1+3 2.5
0.35^ 3-30 2.8
0.563 20.7 k.k
0.708 18.7 5.9
X/D = 0.1U6
R/R0 U V u'2
-0.875 -l.U 3.3
-0.813
1.7 6.1
-°- 17.2 U.8
18.5 lt.6
13>9 5<3
1.9 U.6
-0.375 _0.7 3.3
-0.20U _1<2 3.0
-0.038 _1>9 3>i|
0.125 .2.1 3.3
0.350 _2.3 3.!
!. 9.0
0.1+58 0.6 5.3
13.8 1+.
108
-------
Table F-l. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D ~ Q.li+6
U /u.2
0.625 18.5 3.5
0.70^ 18.0 3.6
0.771 ll+-2 ^-8
0.833 M ^.8
0-908 -2.1 3.0
X/D = 0.187
R/RQ U
-0.917 lU.^ J+.5
-0.792 16.9 3.9
-0.667 12.6 U.7
-0.538 5.0 5.0
-O.U17 -l.k ^.6
-0.292 -k.k 3.8
-0.158 -5.2 3.7
-0.017 -6.0 3-8
0.125 -5.7 3.7
0.250 -5.6 3.5
0.375 -^.2 3.9
0 = 500 1.9 5.3
0.625 10-6 *-7
0.750 16.2 3.8
0.875 16.3 3.6
109
-------
Table F-l. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D = 1.98
R/RO u
-0.917 13.5 3.7
-0.788 lk.9 3.1
-0.667 l^.l 3.1
-0.5^2 13.1 3.6
-O.U17 12.3 3.5
-0.292 10.8 3.3
-0.163 10-2 3.0
-0.0*4-2 10.6 3.3
0.083 lOA 3.1
0.208 10.U 3.0
0.338 12.3 3.U
0.1463 12.0 3.5
0.588 13.0 3.2
0.708 13.0 2.6
0.838 12.9 2.7
0.9^2 12.3 2-7
X/D = 1.79
R/RO u
-0.913 lU.5 3.5
-0.771 15.9 2-8
-0.625 15.7 2.8
-0.^63 ill.3 3.1
-0.296 13.2 2.9
-0.125 12.14- 3.0
O.Ql+6 12.8 3.0
0.208 12.6 3.2
0.375 13.6 3.6
0.5^6 iU.6 3.3
0.713 llf.9 2.5
0.875 iU.3 2.U
110
-------
Table F-2. AXIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST WO. 3
C%-Air Inlet Air Swirl No. = 0.3
$ = 0.91±0.01 Inlet Air Temperature = ?Uli + 9°K
Pressure =1.0 atm Air/Fuel Velocity Ratio = 21.7±0.5
X/D = 0.052
R/RO u
-C-917 6-2 7.2
-0.792 36.9 19.7
-0.733 61.6 20.7
-0.679 70.2 17.2
-0.663 68.2 25.0
-0.600 59-3 22-°
-0.5^2 27.6 2U.5
-O.U13 -3-6 13'9
-0.267 1* n-9
-0.125 5-9 -"-O-2
o.ooo 10-3 2;-°
0.375 -j-° ;r:
0-500 ^-^ Ox'p
0 558 58.9 26'8
i 85 l 19-1
0.70^ °5"L ^ q
0.708 87-5 ^^
_ OQ A ^O.D
0.833 39'° 7<7
0.917 5'8
111
-------
Table F-2. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(Continued)
X/D = 0.128
R/RO
-0.917
-0.913
-0.763
-0.741
-0.671
-0.583
-0.579
-0.400
-0.229
0.042
0.167
0.254
0.421
0.588
0.588
0.663
0-758
0.833
x/D = 0.486
u
30.5
30.0
101.
44.3
84.4
15-3
46.0
5-3
11.9
18.7
13.9
4.1
107.
102.
13.0
u
33-9
1^.8
22.0
21.7
21.9
22.6
25-7
60.5
5^.8
32.8
22.5
-0.929
-0.875
-0.833
-0.833
-0.788
-0.767
-0.708
-0.667
-0.5S6
-0.563
-0.354
-0.155
0.042
62.4
80.7
87-5
85.4
87.5
86.4
66.1
48.6
49-7
36.2
42.2
31.6
41.6
49.6
55-7
55.3
55-7
69.0
62.3
52.7
51.1
40.9
39-3
36.4
112
-------
Table F-2.
X/D = 0.486 (Cont'd)
R/R0
0.254
0.458
0.567
0.667
0.729
0.846
0.850
X/D =0.519
AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(Continued)
U
27.8
45.9
45.6
91.0
83.1
61,3
54.2
R/R0
-0.871
-0.746
-0.688
-0.596
-0.446
-0.279
-0.113
0.054
0.142
0.146
0.263
0.638
0.808
0.858
U
59.0
61.8
6U
58
56
51.7
62.8
64.8
67.0
5^.6
63.1
59-8
59-8
63.7
26.2
25.3
25.8
2h.3
27.6
36.3
41.9
25-7
34.7
26.8
26.7
27-9
27.7
113
-------
Table F-2. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(Continued)
X/D = 1.88
R/R0
-0.871
-0.704
-0.546
-0.375
-0.208
-o.oia
o.oo4
o.oo4
o.i42
0.296
0.1*71
0.633
0.800
0.867
X/D = 3.00
R/R0
-0.875
-0.838
-0-746
-0.596
-0.^58
-0.267
-0.108
0.038
0.350
0.521
0.517
0.671
u
57-1
67.0
62.8
53.8
56.5
69.9
77.6
75.0
54.0
45.3
50.8
53.6
55.6
56.9
U
60.0
68.0
71.6
70.4
67.0
70.1
68.6
68.3
68.0
73.2
67.9
68.9
28.0
29.4
27.1
30.0
33.0
45.6
44.7
43.0
40.4
32.9
27.8
26.5
27-5
29.8
y
u
t2
24.4
30.5
30.9
29,
32,
36,
36.9
41.4
35-0
31-7
27.7
26.2
114
-------
Table F-3. AXIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO. 4
CH^-Air
§ = 0.91 ± o.oi
Pressure = 3 7*0.1 atm
X/D = 0.052
R/RO
-0.938
-0.813
-0.683
-0.542
-0.408
-0.271
o.o42
0.458
X/D = 0.093
R/RO
-0.892
-0.779
-0.683
-0.585
-0.1*71
-0.371
-0.167
0.038
0.250
0.346
0.458
0.458
0.563
0.667
0.771
0.833
- .
Inlet Air Swirl =0.3
Inlet Air Temperature =
Air/Fuel Velocity Ratio
U
0.7
5-2
14. 9
8.3
-1.1
0.2
3.6
1.0
U
2.1
16.6
19.8
13.3
7-8
-2.4
0.8
1.0
0.3
-1.9
5.0
4.8
18.1
21.6
14.5
7.1
- - .
749±7°K
= 20.5±0.4
/u^2
3.6
4.6
5.8
6.2
4.5
4.1
3.0
4.3
J^
4.1
6.8
6.4
8.1
10.7
4.8
3.9
3.8
4.3
4.6
.-
i
7-4
s*
5.6
5.2
60
.8
8_
.1
115
-------
Table F-3. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(Continued)
X/D = 0.291
-0.879 11.0 6.6
-0.708 19.2 6.2
-0.617 13.8 7.7
-0.538 3.9 7.1+
-0.458 -0.8 5.7
-0.375 -2.9 5,5
-0.167 0.9 5.5
0.038 2.1 i+.s
0.167 -2.6 5.3
0.458 o.4 5>2
0.562 9.8 7'o
0.667 17.0 s'.l
0-771 18.3 5.0
0.858 12.6 5 9
0.900 6.4 g"7
1.021 18.1 6>3
X/D = 0.550
R/RO u
-0.913 14.0 8.3
-0.879 17.2 5*6
-0.792 18.0 6*2
-0.750 15.6 6]5
-0.695 11.9 6.7
-0.600 10.2 8*7
-0.583 5.2 6;9
-O.U88 0.03 5 7
-OA13 5.9 9*2
-0.333 -i.u 8;0
-0.250 2.6 8 k
-0.083 4.6 6'9
0.083 4.3 5*3
0.188 2.0 6'7
0.346 -0.8 6'0
0.463 -2.0 *
lib
-------
Table F-3.
X/D = 0.550 (Cont'd)
AXIAL VELOCITY DISTRIBUTIONS (M/SECT)
(Continued)
R/RO
0.570
0.667
0.713
0.767
0.838
0.942
x/D = 1.48
R/RO
-0.888
-0.792
-0.688
-0.579
-0.458
-0.333
-0.208
-0.079
0.046
0.167
0.304
0.417
0.567
0.675
0.792
X/D = 3.00
R/RO
-0.913
-0.788
-0.621
-0.367
-0.163
o.o46
0.250
0.458
0.667
0.775
U
1.8
7.0
10.4
13-2
16.3
17.5
U
17-1
16.2
15.3
14.4
13.6
13.4
13.4
15.9
15-9
14.3
13.9
14. 3
16.1
17.9
18.8
U
15.1
17.2
17.8
18.7
18.7
20.3
18.9
18.4
17.7
17.3
f^
5.4
J T
5-5
f w f
5.8
^/ *-r
5.3
s '
5.1
^ w ^.
4.0
7u'2
3.8
3-9
4.1
3.9
4.3
4.4
4.7
6.2
6.0
4-7
4.0
4.0
4.3
4.0
3.7
/^
3.3
3.0
3.4
4.1
5.1
5.8
4.4
3.6
3.8
3.7
117
-------
Table F-k. AXIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST HO. 6
CH^-Air Inlet Air Swirl =0.6
§ = 0.91±0.01 Inlet Air Temperature = 750±10°K
Pressure = 1 atm Air/Fuel Velocity Ratio = 21.3±0.5
X/D « 0.160
R/RO u
-0.913 37.8 25.6
-0.867 67,9 29.0
-0.792 83.8 3^.5
-0.729 76.5 3k.k
-0.688 U5.7 ^7.0
-0.600 20.7 39.^
-O.k79 -5.^ 19.6
-0.371 0.6 19.3
-0.250 5.5 17.0
-0.075 lU.i 15.2
0.088 iu.2 15.8
0.250 6.0 18.1
0.^17 -3.1 18.8
0.588 0.06 31.5
0.671 15.0 36.3
0.750 57.2 1^.5
0.813 68.0 36.3
0.879 77.7 39.8
0.962 58.7 31.2
118
-------
Table F-4. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D = 0.405
R/R0 U
-0.892 46.5 18.3
-0.892 73.4 27.8
-0.833 63.4 29.4
-0.792 49.4 22.7
-0.754 49.0 28.3
-0.675 23.3 28.0
-0.575 8.6 25.1
-0.458 3.5 25.3
-0.313 3.3 26.8
-0.167 14.5 24.3
0.042 21.6 22.3
0.250 14.7 20.6
0.458 2.5 24.6
0.667 21.4 30.3
0.771 39.6 32.6
0.875 66.4 , 30.5
0.946 72.5 28.2
X/D - 1.55
R/RO u V
16.2
_n ft7 60.1
/ ^7 U ^'3
-0.746 57.4
-0.625 f0'1 %.l
-0.475 ^.0 29>Q
-0.367 3f'° 24.6
-0.242 ^-° 31.5
-0.100 59-2 31>0
0.033 j^'jj 30.3
0.150 ^'^ 23.5
0.275 37.6 18>8
0.395 36.1 16<6
0.525 ^'2 ^-2
0.667 ^'i 15-4
0.775 ?H 16.8
-. OO.^ -| q n
0.871 66<5 I8-0
0.904
119
-------
Table F-4. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D = 1.75
R/RQ U
-0.892 59-2 21.2
-0.792 62.1 15.1
-O.T5U 63.6 15.0
-0.688 57.U 13.6
-0.583 53.9 1^-^
-0.467 ^8-3 17.7
-0.375 ^5.3 19.8
-0.267 39.0 24.0
-0.167 43.6 28. 4
0.058 te.l 31.0
o.ote 29.7 3^.2
0.1U6 33.3 27.6
0.250 29.1 27.0
0.333 35.3 21^.1
0.467 ^5.6 20.5
0.558 50.2 16.8
0.667 58.2 15.9
0.771 62-3 15-0
0.879 67-1 17-^
120
-------
Table F-5. AXIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO. 7
= Air Inlet Air Swirl =0.6
$ = 0.91 ± 0.01 Inlet Air Temperature = 7^6 ± 7°K
Pressure = 1 atm Air/Fuel Velocity Ratio = 0.17 ± 0.01
X/D = 0.052
U
0.95*4 4.6 18.1
0.950 8.1 17.2
0.917 19.9 19.5
0.825 26.6 iu.o
0.813 49.3 24.2
0.750 85.0 3^.8
0.688 61.7 34.5
0.583 15.7 24.5
0.488 -0=9 22.2
0.329 10.0 17.2
0.204 24.4 17.8
0.075 38.4 25.1
0.025 98.1 73.0
0.017 87.5 86.5
0.046 56.2 67.5
0.175 34.8 12.9
0.292 19.8 12.6
0.417 2.5 12.3
0.479 -1-9 1-3-3
0.571 8-9 22.9
0.646 53.9 29-0
0.713 89-2 31.4
0.846 64.1 30.0
0.963 10.8 18-3
121
-------
X/D = 1.97
Tabe F-5. AXIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D = 0.615 _
R/R0 U
0.879 82-7 20.8
0.750 49-9 21.8
0.617 19-5 20.1
0.500 5.0 21.8
0.354 13.5 37.2
0.217 37=9 50.7
0.083 75-5 65.7
0.050 92.3 66.6
0.196 69.8 63.3
0.338 28.2 48.3
0.483 0.90 26.2
0.633 7.7 19-5
0.779 ^O-2 20.1
0.917 71.8 19.1
l.ooo 79-8 23.8
-0.908 61.3 11.7
-0.791 51.4 11.6
-0.671 1+3.6 11.2
-0.538 to. 5 15.6
-0.413 38.7 15.1
0.288 43.5 16.7
-0.167 50.9 17.1
0.042 58.9 17.2
0.083 59-4 17.1
0.217 53.6 15.7
0.338 46.9 15.6
0.458 42.2 16.1
0.592 44.6 14.0
0.667 48.2 12.3
0.708 55.3 13.6
0.829 60.2 14.8
0.925 65.7 10.1
122
-------
Table F-6.
TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO.
= Air
$ = 0.91 ± 0.01
Pressure = 1 atm
X/D = 0.128
R/R0
- 0.900
- 0.788
- 0.679
- 0.542
- 0.325
0.000
0.000
0.350
0.563
0.663
0.771
0.833
Inlet Air Swirl =0.3
Inlet Air Temperature = 7^4 ~ 9°K
Air/Fuel Velocity Ratio = 21.7 ± 0.5
W
52.1
52.8
47.0
26.0
12.7
1.9
3.9
-27.2
,54.6
-56.0
-44.6
-42.0
22.0
23.7
29.6
21.1
17.2
16.5
lit-.9
21.8
30.5
28.2
19.3
19.1
X/D =0.486
0.938
0.850
0.767
0.675
0.579
0.354
0.167
0.167
0.038
0.041
0.296
0.738
0.858
0.913
w
56.14
59.8
67.9
71.1
68.1
'58.9
44.5
37-7
-10.2
- 1.0
54.9
65.4
57.6
57-3
24.9
25.8
29-9
30.1
29.9'
26. h
28.4
28.8
-19-9
-27-7
28.0
27.5
26.
26.0
123
-------
Table F-6. TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC)
(Continued)
X/D = 1.60 _
W
0.86? 54.5 19.3
0.717 60.3 18.3
0.583 64.5 19-9
0.438 65.6 19-9
0.271 65.0 23.6
0.083 41.2 35.7
0.079 41.7 37.7
0.025 4.9 36.7
0.104 -45-4 37.0
0.229 -63.5 27.8
0.471 -66.4 17.9
0.683 -64.8 20.0
0.791 -61.5 18.3
X/D =1.88
W
0.867 96.9 UU.8
0.792 6^.^ 18.7
0.708 69.7 18.8;
0.5^2 68.h 20.1
O.U21 66.4 20.5
0.250 58.7 22.'9
0.088 39.8 32.3
0.000 28.h 33.8
0.017 '23,7 33.5
0.083 -63.9 22.2
0.146 -67.3 24.6
0.200 -73.2 24.2
0.354 -71.5 17.8.
0.533 -69.6 18.1
0.675 -66.5 16.7
0.804 -69.9 19.1
124
-------
Table F-7. TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO. k
= Air
= 0.91 ± 0.01
Pressure = 3-7 ±
atm
Inlet Air Swirl =0.3
Inlet Air Temperature = 7^9 ± 7°K
Air/Fuel Velocity Ratio = 20.5 ± O.U
X/D =0.291
0.885
0.783
0.688
0.579
O.U63
0.267
0.017
0.25^
0.663
0.817
0.90^
W
5.9
8.7
9-5
7-2
k.Q
h.k
0.8
-U.9
-6.6
-7.8
-9.6
-8.9
/(JU I
U.6
U.6
3.5
3.6
3.3
3.8
3.0
3.0
125
-------
Table F-7. TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC)
(Continued)
X/D = 0.550
- 0.896 8.9 U.I
- 0.792 9-0 U.I
- 0.667 8-0 U.7
. 0.583 8.U U.3
- O.U92 9-5 U.U
. 0.325 10-2 k'6
- 0.163 7.8 ^7
- 0.071 3-7 5-1
0.021 1.3 5.2
0.213 - U.6 U.7
0.383 - 6.2 U.2
0.508 - 5.6 U.I
0.575 - U-7 U.I
0.625 - U.I 3.9
0.688 - 3-3 3.8
0.750 - U.U U.I
0.833 - 5-6 3.5
0.875 - 6.6 3-5
0.900 - 6.2 3.1
0.979 - U.2 3.7
X/D = 1.U8 _
R/R0 W
- 0.892 8.0 3.U
- 0.875 6.6 3.2
- 0.750 8.7 3.2
- 0.7U5 7.3 3.U
- 0.613 8.3 3.U
- 0.538 10.8 3.6
- O.U75 10.2 3.8
- 0.329 lU.O U.U
- 0.267 12.7 U.3
- 0.121 lU.O 6.2
O.OU2 0.6 18.1
0.296 - 6.1 11.3
O.U75 -10.6 3.U
0.712 - 8.9 3.0
0.917 - 7.6 3.3
126
-------
Table F-8.
TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO. 6
CI% = Air
$ = 0.91 ± o.oi
Pressure = 1 atm
X/D = 0.162
0.917
0.808
0.708
0.583
0.1^58
0.329
0.171
0.000
0.000
0.021
0.171
0.333
0.500
0.633
0.750
0.875
0.9^2
Inlet Air Swirl = 0.6
Inlet Air Temperature = 750 ± 10°K
Air/Fuel Velocity Ratio = 21.3 ± 0.5
w
57.6
60.8
U8.5
38.6
3^.0
26.7
19.8
1U.9
9-0
-6.1
-27.3
43.9
-1*6.3
-V7.5
-5^.0
-65.5
-53.5
OD'
21.2
23-5
20.1
17.1
15.1*
15.8
17.3
22.0
25.2
27,9
22.5
21.0
17,9
18.2
20.k
zh.h
26.3
127
-------
Table F-8. TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D = 0.^05
0.891
0.791
0.68?
0.583
0.375
0.275
0.158
0.083
0.016
0.02Q.
0.25C
0.3^5
0.562
0.666
0.75C
0.875
0.916
w
66.0
63
60
57
63
65.0
50
35
8.2
-47.2
-60,
-66.
63
62
56
66.0
63.8
128
-------
Table F-8. TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D = 1.55
0.875
0.725
0.621
0.500
0.367
0.217
0.083
0.000
0.021
0.088
0.150
0.213
0.392
0.563
0.708
0.771
0.883
W
64.8
68.1
76.4
80.8
88.7
88.7
50.5
14.3
21.6
-20.2
-68.2
-81.8
-84.5
-76.0
-68.8
-66.6
-67,9
13.0
13.8
16.8
16.8
16.6
17.1
32.4
32.4
31.6
25.0
26.3
21.7
16.7
16.2
15.5
15.8
19.0
X/D = 1.75
R/R0
-0.900
-0.788
-0.688
-0.583
-0.371
-0.167
0.042
0.046
0.250
0.463
0.563
0.667
0.771
w
55.9
.1
.3
.9
.3
66.
65.
72.
82,
56.0
7.3
-38.7
-74.9
-79.8
-70.9
-65.8
-63.2
U)1'
15.1
14.7
16.3
17.2
17.2
23.3
32.5
23.3
19.1
17-4
16.8
13.8
13-3
129
-------
Table F-9. TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO. 7
= Air
$ = 0.91 ± 0.01
Pressure = 1 atm
X/D = 0.052
-0.958
-0.833
-0.708
-0.583
-0.1*58
-0.329
-0.20J4-
-0.079
0.171
0.292
0.1*17
0.667
0.708
0.708
0.833
0.833
0.958
0.958
X/D = 0.615
-0.895
-0.7^5
-0.620
-0.479
-0.333
-0.187
-0.037
0.087
0.125
0.254
0.375
Inlet Air Swirl =0.6
Inlet Air Temperature = 7*4-6 ±
Air/Fuel Velocity Ratio = 0.17 ± 0.01
7°K
w
63.5
63.8
65.6
53.9
64.8
71.5
73.7
79-7
-75.7
-67.4
-64.3
-89.9
-101.8
-103.8
-87.1
-100.5
-60.3
-60.1
w
56.9
63.4
66.3
69.1
63.4
42.5
12.6
27.0
-35.7
27.0
27^9
29.3
20.6
15.2
22.9
26.3
31.8
22.8
17.2
13.6
17. U
2U.O
21.7
22.9
23.7
Ik.7
lk.2
130
-------
Tabe F-9- TANGENTIAL VELOCITY DISTRIBUTIONS (M/SEC)
(continued)
X/D = 1.97
0.871 59.8 7<3
0.750 63.^ 8*.2
0.620 67.5 9'.8
0.500 70.9 11 '.3
0.375 66.3 15.1
0.250- 57.0 17.8
0.121 29.3 19.5
0.017 2.2 17,9
0.021 8.0 20.3
0.1^6 -17.3 17.2
0.288 -32.6 15.6
0.51(6 -62-° 1^.0
0.66? -6^.5 13.8
0.792 -62.1 8.6
0.875 -^9-6 9.6
0.938 -57.1 12.6
131
-------
Table F-10. RADIAL VELOCITY DISTRIBUTIONS (M/SEC) FOR TEST NO. 3
CH^ = Air Inlet Air Swirl =0,3
f =0.90 Inlet Air Temperature =
Pressure = 1 atm Air/Fuel Velocity Ratio = 21
R/RO v
-0.373 5.8 16.6
-0.373 3.8 1^.5
-0.3^2 2.U
-0.290 2.0 13.8
-0.207 0.9 11.3
-0.12U -0.8 1^.1
-1.5 10.9
-3.8 11.3
0.166 . ' -5.1 10.5
0.290 -6.9 11.2
0.415 -10.1 13.3
0.^98 -9.1+ iU.U
132
-------
PUBLICATIONS
The following publications have been produced as a result of
the research program described in this report:
Owen, F. K. Laser Velocimeter Measurements of a Confined
Turbulent Diffusion Flame Burner. AIAA Paper ?6-33 presented
at the AIAA l^th Aerospace Sciences Meeting, Washington, B.C.,
January 26-28, 19?6.
Owen, F. K., L. J. Spadaccini and C. T. Bowman. Pollutant
Formation and Energy Release in Confined Turbulent Diffusion
Flames, to be presented at the l6th Symposium (international)
on Combustion, Boston, August 15-20, 19?6.
133
-------
REFERENCES
1. Heap, M. P., T. M. Lowes and R. Walmsley. The Effect of Burner
Parameters on Nitric Oxide Formation in Natural Gas and Pulverized
Fuel Flames. Paper presented at First American Flame Days Meeting,
Chicago, Illinois, September 1972. p. 78.
2. Heap, M. P., T. M. Lowes and R. Walmsley. Emission of Nitric Oxide
from Large Turbulent Diffusion Flames. Fourteenth Symposium
(international) on Combustion. Pittsburgh, The Combustion Institute,
1973, PP. 883-895.
3. Shoffstall, D. R. and D. H. Larson. Aerodynamic Control of Nitrogen
Oxides and Other Pollutants from Fossil Fuel Combustion. Environ-
mental Protection Agency, Research Triangle Park, N. C., Publication
Number 650/2-73-033a.
k. Shoffstall, D. R. Burner Design Criteria for Control of Pollutant
Emissions from Natural Gas Flames. Paper presented at Symposium on
Stationary Source Combustion, Atlanta, GA., September 1975-
5. Mellor, A. M., R. D. Anderson^ R. A. Altenkirch and J. H. Tuttle.
Emissions From and Within an Allison J-33 Combustor. Comb. Sci.
Technol.6: 169-176, 1972.
6. Jones, R. E. and J. Grobman. Design and Evaluation of Combustors
for Reducing Aircraft Engine Pollution. Atmospheric Pollution by
Aircraft Engines, AGARD Document CP-125, April 1973.
7. Bowman, C. T. and L. S. Cohen. Influence of Aerodynamic Phenomena
on Pollutant Formation in Combustion. Environmental Protection
Agency, Research Triangle Park, N. C., Publication Number 650/
2-75-06la, July 1975. p. 159.
-------
REFERENCES (Cont'd)
8. Spalding, D. B. Mathematical Models of Continuous Combustion.
in: Emissions from Continuous Combustion Systems, Cornelius W-
and ₯. G. Agnew (eds.), New York, Plenum Press, p. 3-18, 1972.
9. Anasoulis, R. F. and H. McDonald. A Study of Combustor Flow
Computations and Comparison with Experiment. Environmental Pro-
tection Agency, Research Triangle Park, N.C., Publication Number
650/2-73-OU5, P. 9^, December 1973.
10. Bray, K. N. C. and J. B. Moss. A Unified Statistical Model of the
Premixed Turbulent Flame. University of Southampton, Southampton,
England. Report No. 335, p. 64, November 197!*.
11. Caretto, L. S. Mathematical Modeling of Pollutant Formation.
Prog. Energy Combust. Sci. 1: 47-71, 1976.
12. Buggeln, R. C. and H. McDonald. Work in Progress.
13. Kerr, N. M. and D. Fraser. Swirl. Part I: Effect on Axisymmetri-
cal Turbulent Jets. J. Inst. Fuel 38: 519-538, 1965.
14. Bowman, C. T. Probe Measurements in Combustion - A Synopsis.
Progress in Astronautics and Aeronautics (to be published).
15. Bilger, R. Probe Measurements in Turbulent Combustion. Progress
in Astronautics and Aeronautics (to be published).
16. Tuttle, J. H., R. A. Shisler and A. M. Mellor. Nitrogen Dioxide
Formation in Gas Turbine Engines: Measurements and Measurement
Methods. Combust. Sci. Technol. 9:261-271, 1975-
17. Bennett, J. C. Use of Five-Hole Pneumatic Probes in Unsteady
Flows. Progress in Astronautics and Aeronautics (to be published),
18. Becker, H. A., H. C. Hottel and G. C. Williams. On the Light-
Scatter Technique for the Study of Turbulence and Mmng. J.
Fluid Mech. 30:259-284,
135
-------
REFERENCES (Cont'd)
19. Anon. Procedure for the Continuous Sampling and Measurement of
Gaseous Emissions from Aircraft Turbine Engines. Aerospace
Recommended Practice 1256, SAE, p. l6, 1971.
20. Owen, F. K. Laser Velocimeter Measurements in Free and Confined
Coaxial Jets with Recirculation. United Technologies Research
Center (Presented at 13th AIAA Aerospace Sciences Meeting.
Pasadena, CA, January 20-22, 1975) p. 10.
21. Schefer, R. W., R. D. Matthews, N. P. Cernansky and R. F. Sawyer.
Measurement of NO and N02 in Combustion Systems. University of
California. (Presented at the Western States Section/Combustion
Institute Meeting, El Segundo, October 1973) p. 18.
22. Cernansky, N. P. and R. F. Sawyer. NO and N02 Formation in a
Turbulent Hydrocarbon/Air Diffusion Flame. Fifteenth Symposium
(international) on Combustion. Pittsburgh, The Combustion
Institute, pp. 1039-1050, 1975-
23. Owen, F. K. Laser Velocimeter Measurements of a Confined Turbulent
Diffusion Flame Burner. United Technologies Research Center.
(Presented at the lUth AIAA Aerospace Sciences Meeting. Washington,
D.C., January 26-28, 1976) p. 10.
2k. Lindgren, B. W. and G. W. McElrath. Introduction to Probability
and Statistics, New York, Macmillan, p. 165, 1959.
136
-------
NOMENCLATURE
a = defined by Eq. (5)
d = outer diameter of air annulus, cm
dn = inner diameter of air annulus, cm
D = Combustor diameter = 0.122 m
D = particle diameter,
f = frequency, Hz
f = Doppler frequency, Hz
f = offset frequency, Hz
K = Cunningham constant
-------
NOMENCLATURE (CONT'D)
R = radius, m
R = combustor radius = 0.06l2 m
S = swirl number as defined by Eq.. (l)
Sv = calculated variance in the variable x, Eq.. (11)
A
T = temperature, °K
Up = rms particle velocity, m/sec
U = mean axial velocity, m/sec
U = convective velocity, m/sec
c
IL = instantaneous axial velocity, m/sec
u = axial velocity fluctuation, m/sec
"u = large-scale axial velocity fluctuation, m/sec, Eq. (8)
Va = bulk mean air velocity, m/sec
Vf = bulk mean fuel velocity, m/sec
W = mean tangential gas velocity, m/sec
w = tangential velocity fluctuation, m/sec
x = random variable
X = axial distance, m
Z = dh/d
a = percentage of time mean flow is reversed, Eq.. (7)
P = true mean of the variable x, Eq. (10)
Y = directional intermittency
138
-------
NOMENCLATURE (CONT'D)
Ym = error in the mean, Eq. (10)
YX = error in the variance, Eq. (ll)
1\ - swirl vane angle, deg
9 = angle, deg
X = wavelength, m
p, = viscosity, gm/cm-sec
p = particle density, gm/cc
CT = rms velocity, m/sec
§ = overall fuel-air equivalence ratio = ("1fuel/"1airV(nifuel/iair)
S t-OlCil
139
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-247a
2.
3. RECIPIENT'S ACCESSION NO.
4.T.TLEANDSUBT.TLE JNFLUENCE QF AERODYNAMIC
PHENOMENA ON POLLUTANT FORMATION IN
COMBUSTION (Phase I. Gaseous Fuels)
5. REPORT DATE
September 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
L. J.Spadaccini, F.K.Owen, and C.T. Bowman
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING OROANIZATION NAME AND ADDRESS
United Technology Research Center
400 Main Street
East Hartford, Connecticut 06108
10. PROGRAM ELEMENT NO.
1AB014; ROAP 21BCC-014
11. CONTRACT/GRANT NO.
68-02-1873
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Phase Final; 4/75-5/76
14. SPONSORING AGENCY CODE
EPA-ORD
is. SUPPLEMENTARY NOTES jERL-RTP project officer for this report is W.S.Lanier, Mail
Drop 65, 919/549-8411 Ext 2432.
16. ABSTRACT.
The report gives results of an experimental investigation of the effects of
the interaction between fluid dynamics and chemistry on pollutant formation and
destruction in a natural-gas-fired, turbulent diffusion flame burner. The investiga-
tion determined the effects of inlet air swirl, combustor pressure, and air/fuel
velocity ratio on the time-mean and fluctuating flow field, using probing and optical
techniques. Changes in flow field structure were correlated with changes in pollu-
tant emissions from the furnace. The investigation also showed that varying these
parameters produces major changes in the time-mean flow field within the burner
which significantly influence pollutant formation. It was also discovered that there
are substantial large-scale contributions to the total rms turbulent velocity field.
These large scale fluctuations result in significant departures from Guassian
turbulence and isotopy in the initial mixing regions of the burner and have
pronounced effects on mixing, chemical reaction, and pollutant formation.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Natural Gas
Combustion
Furnaces
Nitrogen Oxides
Lasers
Aerodynamics
Turbulence
Swirling
Speed Indicators
Pollution Control
Stationary Sources
Gaseous Fuels
Laser Velocimetry
13B
21D
2 IP
13A
07B
20E
20D
07A,13H
14B
3. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
150
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
iko
------- |