APTD • 1441
Low NOx Emission Combustor
For Automobile Gas Turbine Engines
Prepared By
David J.White, P.B.Roberts and W.A.Compton
SOLAR DIVISION INTERNATIONAL HARVESTER COMPANY
2200 Pacific Highway
San Diego, California 92138
CONTRACT NUMBER:68-04-0016
EPA Project Officers
H. F. Butze and Robert B. Schulz
Prepared For
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Mobile Source Pollution Control Program
Advanced Automotive Power Systems Development Division
February 1973
-------
The APTD (Air Pollution Technical Data) series of reports is issued
by the Office of Air and Water Programs, U. S. Environmental Protection
Agency, to report technical data of interest to a limited number of
readers. Copies of APTD reports are available free of charge to
Federal employees, current contractors and grantees, and non-profit
organizations - as supplies permit - from the Air Pollution Technical
Information Center, U. S. Environment Protection Agency, Research
Triangle Park, North Carolina 27711, or may be obtained for a nominal
cost from the National Technical Information Service, U. S. Department
of Commerce, 5285 Port Royal Road, Springfield, Virginia 22151.
This report was furnished to the U. S. Environmental Protection Agency
by Solar Division of International Harvester Company, San Diego,
California, in fulfillment of Contract Number 68-04-0016. The contents
of this report are reproduced herein as received from Solar Division of
International Harvester Company. The opinions, findings, and conclusions
expressed are those of the author and not necessarily those of the
Environmental Protection Agency.
Office of Air and Water Programs Publication Number APTD - 1441
-------
APTD . 1441
Low NOx Emission Combustor
For Automobile Gas Turbine Engines
Prepared By
David J.White, P.B.Roberts and W.A.Compton
SOLAR DIVISION INTERNATIONAl HARVESTER COMPANY
2200 Pacific Highway
San Diego, California 92138
CONTRACT NUMBER: 68-04-0016
EPA Project Officers
H. F. Butze and Robert B. Schulz
Prepared For
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Mobile Source Pollution Control Program
Advanced Automotive Power Systems Development Division
February 1973
ROR \ 105.5
-------
FOREWORD
This report entitled "Low NOx Emission Combustor for Automobile Gas
Turbine Engines", describes the work performed pursuant to Contract No. 68-04-0016,
for the Advanced Automotive Power Systems Development (AAPSD), Office of Air Pro-
grams (OAP), Environmental Protection Agency (EPA). The work was conducted at the
Solar Division, International Harvester Company in San Diego, California, under the
technical direction of the Research Laboratories. Mr. W. A. Compton, Assistant
Director - Research, was the technical director and program manager with Mr. David
J. White, Research Staff Engineer, as the principal investigator.
Two individual Project Officers have been associated with this program.
Initially Mr. Robert Schulz of the Advanced Automotive Power Systems Development
Division of the Environmental Protection Agency, held the post, which he relinquished
to Mr. H. Fritz Butze, approximately a quarter of the way through the program.
Mr. Butze is employed by the National Aeronautics and Space Administration, Lewis
Research Center in Cleveland, Ohio.
Solar's internal report number is RDR 1705-5.
iii
-------
ACKNOWLEDGEMENTS
. In addition to the authors, Solar personnel and consultants who made an
important contribution to the program include L. F. Blinman, Product Engineer"- .
initial principal investigator; J. R. Shekleton, Engineering Specialist - combustor
.. design cpnsultation; R. T. LeCren, Group Engineer - experimental investigation of
conventional combustors; T. E. Duffy, Research Staff Engineer and W. E~ Reed, .
Group Engineer":, controls analysis; Dr. B. P. Breen and C. Bodeen of KVB
Engineering - combustion analysis and computer programs; and Professor R. . F.
Sawyer, University of California at Berkeley - consultation.
Spec~al credit is given to Mr. John V. Long, Director-Research, for
supporting the program fr:om concept to completion.
. .'
iv
-------
iU u- ---~-
Section
1
2
3
4
, .
, , CONTENTS
INTRODUCTION
1.1
1.2
Background
Program Phases
SUMMARY
2.1
2.2
2.3
Background
Objective
Results
3.1
OVERALL PROGRAM DESCRIPTION
3.2
3.3
3.4
3.5
4.1
Analytical Investigations
3.1.1 Computer Analyses
3.1.2 Flow Visualization
Combustor Modifications - Rich Primary Zone
3.2.1 Conventional Combustors
3.2.2 Rapid Mixing Transition Zone Combustor
LEAN PRIMARY ZONE COMBUSTORS
3.3.1
3. 3. 2
Lean Primary Zone Pilot System
Lean Primary Zone Vortex Induced Circulation
System
Lean Primary Zone Jet Induced Circulation
System
3.3.3
Parametric Combustor Designs
Parametric Combustor Results
3.5.1 Class A-Mod (JIC-3)
3. 5. 2 Class B-Mod (JIC-4)
PROGRAM EXTENSION OUTLINE
Addendum I - Optimized A 1 Federal Driving Cycle
Results (JIC-3)
v
Page
1
1
2
3
3
3
3
7
7
7
8
9
9
17
24
26
26
33
37
43
43
55
65
66
-------
Section
5
6
4.2
4.3
CONTENTS (Contd)
, ,
Addendum II - Primary Zone'Model and Variable Area
Port Tests
4. 2. 1 Primary Zone Model Tests
4.2.2 Variable Geometry Port Tests.
Addendum ill - Combustor Control System Analysis
and NOx Correlations
4. 3. 1 NOx Correlation Parameters
4.3.2 Control System Analysis
CONCLUSIONS
RECOMMENDA TIONS
vi
~
66
, 66
74
84
84
85
89
91
-------
Figure
1
2
3
4
5
6
7
8
9
10
11'
12
13
14
15
16
17
18
19
20 "
21
22
ILL USTRA TIONS
Program Organization Chart
Schematic Representation of Flow Visualization Rig
Emission Characteristics of Original (Reference) Combustor
Emission Characteristics of Original (Reference) Combustor
Emission Characteristics of Mod. 9 Combustor
Emission Characteristics of Mod. 9 Combustor
Emission Characteristics of Mod. 9 Combustor
Emission Characteristics of Mod. 9 Combustor
Emission Characteristics of Mod. 9 Combustor
Emission Characteristics of Mod. 9 Combustor
Emission Characteristics of Mod. 9 Combustor
Emissions Characteristics of Mod. 9 Combustor
Rapid Mixing Transition Zone Combustor
Results of Rich Primary Zone Combusto r
SChematic of Lean Primary Zone Pilot Flame Combustor
Results of Lean Primary Zone Pilot Flame Combustor
Results of Lean Primary Zone Pilot Flame Combustor
Schematic of Vortex Induced Circulation (VIC) Combustor With
Air-Assist Atomizer
Page
8
10 "
13
14 .
15
16
18
19
20
21
22
23
24
25
27
27
28
29
Vortex Stabilized Lean Primary Zone Combustor (NOx Emissions) 30
Vortex Stabilized Lean Primary Zone Combustor (CO Emissions)
Resuits of Vortex Induced Circulation (VIC) Lean Primary Zone
Combustor
High Recirculation Stabilized Lean Primary Zone Combustor
(N0x Emissions)
vii
31
'.
32 .
33
-------
Figure
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
ILL USTRA TIONS (Contd)
High Recirculation Stabilized Lean Primary Zone Combustion
(CO Emissions)
Schematic of Jet Induced Circulation Combustor (JIC)
Results of Jet Induced Circulation (JIC) Lean Primary Combustor
(Initial Fuel Injector Position) .
Results of Jet Induced Circulation (JlC) Lean Primary Zone
Combustor (Modified Fuel Injector Position)
Results of Jet Induced Circulation (JlC) Lean Primary Zone
Combustor (Modified Fuel Injector Position)
Assembly Drawing of Parametric Combustor Class A
Class A-Mod - Low Pressure Combustor Tests
Class A-Mod Combustor - JP-4 - FDC #1
Class A-Mod Combustor - JP-4 - FDC #2
Class A-Mod Combustor - JP-4 - FDC #3
Class A-Mod Combustor - JP-4 - FDC #4
Class A-Mod Combustor - JP-4 - FDC #5
Class A-Mod Combustor - JP-4 - FDC #5 Rev. 1
Class A-Mod Combustor - JP-4 - FDC #6
Class B-Mod Combustor, FDC #1 Rev. 1 (Repeat) (Fuel: JP~4)
Class B-Mod Combustor, FDC #2 Rev. 2 (Fuel: JP-4)
Class B-Mod Combustor, FDC #3 Rev. 2 (Fuel: JP-4)
Class B- Mod Combustor, FDC #4 (Fuel: JP-4)
NOx Variation With Inlet Pressure, Class B-Mod Combustor
(JIC-4)
Schematic of Primary Zone Modular Construction
Model Primary Zone Combustion System With Attached Tailpipe
Brayton Cycle Combustor Program Primary Zone Model Testing
Configuration Summary
viii
~
34
3q
36
38
39
41
44
45
46
47
48
49
50
51
57
58
59
60
64
68
69
70
-------
ILL USTRA TIONS (Contd)
Fil?;Ure Page
45 Lean Blow Out Correlation for a Jet Induced Circulation (JIC)
Combustor Primary Zone 73
46 Primary Zone Model Tests, JP-4, Reference Combustor 75
47 Primary Zone Model Tests, JP-4, Mod. 19 76
48 Variable Port Model 77
49 Variable Geometry Port Tests Translating Valve 79
50 Variable Port Model - Slide Valve 79
51 Controlled Variable Area Combined Primary and Dilution Port
System 80
52 Details of Slide for Slide Valve Port 80
53 Mass :flow Characterization of the Slide Valve Variable Area
Port 81
54 Variable Port Model - :fluidic Operation 81
55 :flow Characteristics of the :fluidic Variable Area Device 82
56 Control System Block Diagram S6
57 Error Analysis - Full Speed Digital 87
58 Error Analysis - Idle Speed Digital 88
ix
-------
TABLES
Table ~
I 1976 Federal Standards (Emission Level Goals) 4
II Analytical Computer Programs Available 9
III Summary of Combustor Test Configuration 11
IV Brayton Cycle Combustor Test Program Configuration Summary 12
V Brayton Cycle Combustor Class A-Mod (Preliminary Results) 52
VI Brayton Combustor Program FDC and Steady State Results, 53
Class A-Mod Combustor
vn Aldehyde Analysis 55
vm A3 Range Mode Results - High Baseline 56
IX A3 Range Mode Results - Low Baseline 56
X Brayton Cycle Combustor Class B-Mod 61
XI Brayton Combustor Program - Bl and B2 Test Results 63
Class B-Mod Combustor
XII Class A-Mod Combustor Optimized Federal Driving Cycle 67
XIII Primary Zone Model Tests Summary 71
XIV Comparison of the Reference Primary Zone with the Best of the 77
Various Modifications
XV Summary of NOx Correlation Parameters 86
xi
-------
1
INTRODUCTION
1.1
BACKGROUND
Contract #68-04-0016, entitled "Low NOx Combustors for Automobile Gas
Turbine Engines", has advanced a combustor concept which shows excellent promise
for low emissions on an automotive gas turbine. The objective of the program was to
develop the necessary design criteria for such a combustor and demonstrate the
emission characteristics, on the test stand, of a model combustor employing these
criteria.
A number of combustor concepts were evaluated for the low pressure regenera-
tive (A-Mod) cycle and the high pressure recuperative (B-Mod) cycle against the require-
ments of the EPA prototype vehicle performance specification.
The combustor concept finally selected provides the ability to stabilize flames
at lean equivalence ratios and reduce the primary zone oxygen content (through dilution
with recirculated combustion products) by a method termed "Jet Induced Circulation"
(JIC). In order to achieve this type of recirculation, jets are arranged to flow forward
toward the dome in a conventionally shaped can combustor so that they impinge on the
centerline, producing a derived jet that in turn impinges on the combustor dome. The
resulting primary zone flow field is of the nature of a toroidal vortex. It is possible
by this method to ensure that the ratio of recirculation mass flow rate to the entering
air and fuel mass flow rate be greater than one: a necessary requirement for satis-
factory lean extinction limits with a lean reaction zone system.
It has been shown, using interchangeable "fixed" orifices at the primary and
dilution ports, that the JIC combustor can operate at set test points representing the
Federal Driving Cycle to yield integrated emission levels within the EPA prototype
vehicle specifications.
A Data Book documenting this program in considerable detail, is available
on request from AAPSD/EPA. This particular report provides the results and dis-
cussion of each investigation and test performed during the program.
1
-------
1.2
PROGRAM PHASES
The program approach dictated that the work be split into five phases:
. Phase I - Development of combustor design criteria through combustor
test and analysis
. Phase n - Establishment of parametric combustor designs
. Phase ill - Parametric combustor fabrication; test rig fabrication
. Phase IV - Parametnc combustor test and final evaluation
. Phase V -Program extension work
2
-------
2
SUMMARY
2.1
BACKGROUND
Because of the increasing concern for a cleaner environment, automotive
engines utilizing unconventional thermodynamic cycles have been given serious con-
sideration. In particular, the Brayton cycle engine is considered a prime candidate
for reducing the undesirable pollutant emissions produced by present automobiles.
The Environmental Protection Agency, Division of Advanced Automotive Power
Systems, has recognized that the combustors of present day production automotive
gas turbine engines have not been optimized for low emissions of NOx' and is there-
fore supporting programs intended to provide both basic design data and demonstra-
tions of combustors that will meet or better the 1976 standards.
2.2 OBJECTIVE
The prime objective of this particular program was to develop, through both
analytical and experimental studies, the basic design criteria and data necessary to
produce a low emission combustor. In addition, this information was to be utilized in
the production of two combustor designs, one for a typical low-pressure regenerative
type of engine and the other for a high-pressure engine with partial recuperation.
Initially these two designs were referred to as the Class A and Class B combustors,
respectively. Later, because of the uprating of the combustor inlet conditions
(essentially both classes had their inlet temperatures increased), these classifications
were changed to Class A-Mod and Class B-Mod. The 1976 Federal Standards for
vehicle emissions together with the derived emission index goals for the two classes
are given in Table I.
2.3
RESULTS
Several model combustors were produced employing various concepts to
obtain low emissions, and these were evaluated as to their suitability for incorporation
into a practical engine system. The criteria for evaluation were in order of importance:
. Low emissions of oxides of nitrogen (NOx)
3
-------
TABLE I
1976 FEDERAL STANDARDS
(EMISSION L~VEL GOALS)
Equivalent Equivalent
Emission Index Emission Index
at at
10 mpg 12.7 mpg
Emission Level (s.g. "'0. 763) (s. g. = 0.763)
Constituent (gm/mile) (Class B Mod.). (Class A Mod.)
gm/Kg gm/Kg
Oxides of Nitrogen asN02 (NOx) 0.4 1.38 '1. 78
Carbon Monoxide (CO) 3.4 11.8 14.9
"
Unburned Hydrocarbons as CHI. 85 0.41 1.42 1.81
(UHC)
Note: At the start of the progr~m the goals were to have emission indices lower than
those in the table for: each combustor class. Later in the program Solar set
itself the goals of meeting one-half or less of each of the emission indices.
. Low emissions of carbon monoxide (CO) and unbur.ned
hydrocarbons (UHC), (high efficiency)
.Acceptable pressure loss (ilP/P) $ 5%
. Stability equivalence ratio range sufficient for entire engine
operating range "
. Simplicity of variable area control (where necessary)
.No smoke or odor
. Reasonable size and weight
Of the various models tested, when operated' with liquid fuels, the concept
that appeared to show the most promise, particularly in terms of low NOx emissions,
was that designated the Jet Induced Circulation (JIC) combustor system. This system
initially had high pressure losses, but a logical approach for reducing these losses
without impairing emission performance was apparent; thus this system became the
prime choice.
4
-------
Two combustors utilizing this basic JIC co~cept were designed and built, one
for.a low-pressure, regenerative engine cycle and the other for a high-pressure
cycle system with partial recuperation. These two designs were designated Class A~
Mod and Class B-Mod or JIC-3 and jIC-4, respectively. The resuIt"s of these two
combustors, when tested over a simulated Federal Driving Cycle (conditions specified
by EPA), indicated (particularly the Class A-Mod system) that the 1975-76 emission
requirements can be easily met and bettered, provided some satisfactory variable area
system and control could be devised.
A program extension was granted, which had as one of its aims the genera-
tion of additional design data to enable the JIC concept to be scaled to meet any given
set of inlet conditions. In addition, an attempt was also to be made to both devise a
control system for the combustor and to evaluate various variable area port concepts.
The basic design information, intended for the purpose of combustor sizing,
was obtained from a series of uncooled model combustor primary zones and model
varia.ble area primary ports. The results obtained have indicated several approaches
that could lead to increased operating range for the low emissions section, which in
turn would minimize the need for variable geometry.
The evaluation of the various variable area port models led to the conclusion
that all of the various concepts tested could, under certain circumstances, be suitable
for incorporation into a practical combustor. In general, the problems associated
with variable area ports are mainly connected with the necessary auxiliary systems
such as the linkage and actuator mechanisms in those cases where mechanical move-
ment is required, and in the provision of the control air supply in the case of fluidic
ports. The fluidic ports do have some extra disadvantages when compared with purely
mechanical systems, in that under certain flow conditions instability in the form of
bistable operation can occur, although this can be allowed for. .
In investigating a control system suitable for a typical combustor operating
with variable area ports to control local equivalence ratios and thus NOx' a series of
NOx correlations as a function of the main combustor inlet conditions was initially
evolved. Based on the accuracy of these correlations and the errors involVed in
measuring the various independent parameters, an estimate of the total possible error
associated with the control system was obtained. This estimate was then compared with
the acceptable operating range and it was found that the two were very close. Thus, the
conclusion drawn from this initial phase is that although such control is possible, it
would be difficult. In consequence, an approach was taken to minimize the number of
independent variables. This was achieved by correlating the NOx emissions as a
function of the inlet pressure and a synthesized primary zone reaction temperature.
The ratio of the equivalence ratio entering into the primary zone reaction temperature
calculation to the overall equivalence ratio obtained from the fuel control system is a
5
-------
function of the primary port to dilution port area ratio. Thus a schedu!e of this latter
area ratio for different overall equivalence ratios, inlet pressures, and. temperatures
can be correlated diJ:'ectly as a function of NOx. This approach does lead to a viable
solution to the contrc:>l problem, assuming that the present operating bandwidth can be
increased slightly.
6
-------
3
OVERALL PROGRAM DESCRIPTION
The overall program approach adopted can be represented by the block dia-
gram of Figure 1.
This diagram shows that Solar's considerable background experience in
combustor technology was util~zed, together with flow visualization techniques and
computer simulations, to provide some basic combustor model designs. These initial
models were rig tested, and they provided a basic understanding of the type of param-
eter important from a point of view of NOx production and the design of a low NOx
combustor system. From these results, various combustor concepts were designed
and tested, including examples of both rich and lean reaction zone systems.
From this second series of rig tests the most promising combustor concept
was selected to be the basis of the parametric Class A-Mod and Class B-Mod designs
which, after AAPS approval, were fabricated and put through the final test evaluation.
3.1 ANALYTICAL INVESTIGATIONS
The analytical studies were used as a qualitative guide towards the formation
of various conceptual designs of low emission combustors during the initial stages of
the program. A dual-level approach involving the application of analytical studies in
conjunction with an experimental test program was thought to be the optimum ~ethod
that would minimize the effects of the separate limitations of each approach.
. .
3. 1. 1
Computer Analyses
. Early in the program it was realized that to obtain a detailed understanding
of the mechanisms involved in the formation of nitric oxide (NO) within a practical
combustion system, a number of analyses involving aerodynamics, heat transfer and
chemical kinetics would have to be made. These analyses, although usually applied
only to an idealized section of a practical system, are complex, and efficient usage
dictated that they would have to be incorporated as computer programs. .
7
-------
CONTRACT NO.68-01-0016
FLOW
VISUALIZATION
COMPUTOR
SIMULATION
(KVBJ
CONVENTIONAL
DESIGNS
i .
APPS APPROVAL
FIGURE 1. PROGRAM ORGANIZATION CHART
The various programs, although specifically applicable to com-
bustor analyses are capable of extensive use in other areas of interest in gas turbines.
A listing of the various computer programs available to Solar is given in Table II.
3. 1.2
Flow Visualization
As an aid in visualizing the effect on the flow patterns of changes in combustor
geometry, and also to verify the results of certain analytical computer programs, a
series of Plexiglas models of the preliminary combustor were built. These models,
when used with smoke injection, provided qualitative definition of the flow patterns
within the combustor.
The flow visualization rig consists basically of the Plexiglas model mounted
to an exhauster fan unit, with a light source above the combustor producing an axial,
narrow slit of light. This slit of light is directed along a diametrical plane as shown in
8
-------
TABLE II
ANAL ITICAL COMPUTER PROGRAMS AVAILABLE
1. Droplet Evaporation and Combustion
2. One Dimensional Chemical Equilibrium (ODE)
3. Two Dimensional Chemical Equilibrium (TDE)
4. One Dimensional Chemical Kinetics (ODK)
5. Two Dimensional Chemical Kinetics (TDK)
6. Gosman-Spalding Model
7. One Dimensional Combustor Aerodynamic Analysis
8. Two Dimensio"nal Combustor Aerodynamic Analysis
9. One Dimensional Combustor Heat Transfer Analysis
110 Two Dimensional Combustor Heat Transfer Analysis
11: Two Dimensional Mixing Program
12. Jet Penetration Analysis
13. Jet Impingement and flow Split Analysis
14. Atomizer Design Program
15. Swirler Design Program
Not a Production Program
but Available
Production Program
Production Program
Production Program
Production Program
Experimental Program
Production Program
Experimental Program
Production Program
Experimental Program
Experimental Program
Production Program
Production Program
Experimental Program
Experimental Program
the schematic of Figure 2. When operated in a darkened room, with smoke injected
discretely in turn through a hole in each row that is in the plane of light, a clear
tracing of the flow pattern in that particular plane results.
3.2
COMBUSTOR MODIFICATIONS - RICH PRIMARY ZONE
3. 2. 1
Conventional Combustors
A key element in this program was the decision to conduct an extensive series
of low cost model tests with simplified combustors and use the resultant data, in con-
junction with analytical studies, to design a final combustor suitable for engine tests.
The lack of detailed analytical data, especially as regards prediction of mixing, was the
justification for this broad cover, low cost approach.
The initial rig testing was confined to the regenerative Class A combustor
configurations, generally at atmospheric pressure conditions. The FID hydrocarbon
9
-------
TRANSITION DUCT
AXIAL SLIT LIGHT SOURCE
I-
EXHAUST DUCT
\
VANE-AXIAL FAN
o
o
o
o
o
o
o
o
o
o
o
o
o
'~o \
TEST COMBUSTOR
o
o
o
J
o
o
o
o
o
FLOW DIRECTION
I
.... ....
~
CLAMP AND
SEALING
COLLAR
°\ZKE GENERA TOR
CIGAR LOCATING PIN
CONTAINS LIGHTED
OIL SOAKED CIGAR
FIGURE 2. SCHEMATIC REPRESENTATION OF FLOW VISUALIZATION RIG
analyzer was not operable, hence results include nitric oxide and carbon monoxide
production only.
The various combustor configurations shown in Tables III and IV were tested.
The configurations included variations in the fuel atomization system and the air-fuel
distribution throughout the length of the combustor. Configurations were tested at
varying air mass flow rates, inlet temperatures, inlet pressures and temperature rises.
Typical results are presented in Figures 3 and 4. In these graphs the NOx
(calculated as NOz) and CO emission indices in grams per kilogram of fuel are .. .
plotted as a function of combustor temperature rise for various values of combustor
loading, expressed as combustor pressure loss, and combustor inlet temperature.
Original Combustor
This combustor was the initial Class A combustor configuration consisting.
of a rich primary zone for moderate reaction temperatures (and hence .low NO product-
ion rates) followed by a secondary reaction zone, where enough air was suddenly added
to lean out the mixture to a point where the NO production rate was maintained at a
low level but the local temperatures were still high enough to permit the continued
reaction of CO to C02. The remaining air was introduced at the beginning of the
10
-------
TABLE III
SUMMARY OF COMBUSTOR TEST CONFIGURATIONS
. MNAL
6 6 22 66*
LIQUID FUEL ~
MOO 1 ,J
3 317 77
LIQUID FUEL ~
MOO 2 ~
6 3 16 75
AIR ASSIST OR NG**
LIQUID FUEL ~I MOO 5 ~
6 2531 38
LlQUIDA;:;~~:: 6 ~
9 2529 37
LIQUID FUEL
N G ---f
~ MOO 7 ~
2732
41
AIR ASSIST ---"1
LIQUID FUEL---L~I M003 ~
6 316 75
NG-t
~ MOO 8 t J
3718 45
AIR
ASSIST OR NG
LIQUID FUEL ~ MOO 4 ~
6. 330 61
* AIR FLOW PERCENT
** NATURAL GAS
NG--t
~
55
MOO 9 ~
. 45
tertiary zone, where final mixing was carried out. A prevaporized fuel injection
system was utilized, with the liquid fuel introduced through a plain liS-inch diameter
tube coaxially with the vaporizing airflow. The results of the concentration of CO and
N02 are given in Figures 3 and 4 .
At a given combustor inlet temperature the N02 concentrations did not
appear to be a function of combustor loading, in contrast to the CO concentrations,
where for a given inlet temperature there was an increase of CO level with increased
combustor loading, presumably due to the reduction in overall residence time. The
NO concentrations, however, are mainly a function of a chemical reaction which is
dependent upon the actual reaction zone residence time, which in turn is apparently
mixing controlled.
Combustor Mod. 9
This configuration represents the final one in this series and simplifies the
combustor to a set of primary and dilution holes only. Figures 5 and 6 give the results
11
-------
TABLE IV
BRA YTON CYCLE COMBUSTOR TEST PROGRAM
CONFIGURATION SUMMARY
11.75
AIR SWIRLER.
SIX TANGENTIAL SLOTS.
LENGTH - D
WIDTH - E
I! I
.25
1--.75
I~F
-1
3.0CA.
FUEL DELIVERY TUBE.
.....
t\j
MOD. "I< DSK 13410 MOD. 1 MOD.2 MOD.3 MOD.4 MOD. 5 MOD. 6 MOD. 7 MOD. 8 MOD.9
FUEL SYS. ST, TUBE ANGLE TUBE ANGLE TUBE AIR ASSIST AIR ASSIST AIR ASSIST AIR ASSIST HI-VELO. HI-VELO. HI, LO VELO.
0= .75 0= .375 D= ,75 D= .75 0= .75 D= ,75 0 = 1,0
AIR SWIRL. E= .125 E= .125 E= .125 E= .125 E= .125 E= .125 E= ,187 NONE NONE NONE
F = 3.0 F = 6.0 F = 6.0 F = 6.0 F = 6.0 F = 6.0 F = 6.0 i
PRIMARY A = 1.75 A=1.75 A = 1.75 A = 1.75 A = 1.75 A = 1.75 A = 1.75 A = 1.75 A=1.75 A = 1. 75
1x30x.94* 1x30x.94 1x30x.094 1x30x.094 1x30x.094 1x30x.157 1x30x.157 1x30x,157 2x30x.157 2x30x .19 2
SECONDARY B = 3.25 B = 3.5 B = 3.5 B = 3.5 B = 3.25 B = 3.25 B = 3.25 B = 3.25 B = 3.25 NONE
2x15x.188 1x15x.218 lx15x.218 1x15x.218 2xl5x.218 2x15x.218 2x15x.218 2x15x.218 1x15x.218
DILUTION C=7 C = 10 C = 10 C = 10 C = 10 C = 10 C = 10 C = 10 C = 10 C = 10
2xllx.375 2xllx.404 2xllx.404 2xllx.404 !4c*2xllx. 404 lxllx.404 1xllx.404 1xllx,404 1x11x.404 1xllx. 404
FUEL JP4 JP4 JP4 JP4 JP4 JP4, NG. JP4,NG. N .G. N.G. N.G.
. I
* 1 x 30 x . 94 Eq~ivalent to one Row of
30 Holes of . 94-Inch Diameter .
** ONE ROW PARTIALL Y BLANKED TO MAINTAIN SAME TOTAL HOLE AREA.
-------
-
...J
I.JJ
=>
lJ..
~
~ 4.0
:2'
<.:)
V>
Z
o
V>
V> 3.0
I-'
'"
:2'
I.JJ
N
o
Z
6.0
RECUPERATIVE COMBUSTOR
AVERAGING PROBE - HOLES DOWNSTEAM (JP-4)
5.0
850 of
2.0
CONTRACTUAL LIMIT
1.0
200
6':'0 6P /P
4°(0 6P /P
COMBUSTOR
PRESSURE
DROP
1000
400 600 800
COMBUSTOR 6T (oF)
FIGURE 3. EMISSION CHARACTERISTICS OF ORIGINAL (REFERENCE) COMBUSTOR
-------
~
,.j::.
...J
W
=>
LL 30
<.:)
::£
"-
:2
<.:)
V1 25
2:
o
V1
V1
~ 20
o
u
50
RECUPERATIVE COMBUSTOR
45
40
35
15
10
80/0 6P /P
5 .
1050 of
AVERAGING PROBE - HOLES DOWNSTREAM
(JP-4)
6"/0 6P /P
400 600
COMBUSTOR 6T ( OF)
800
200
1000
400
COMBUSTOR
INLET
TEMPERATURE
750 of
4"/0 6P /P
COMBUSTOR
PRESSURE
600
FIGURE 4. EMISSION CHARACTERISTICS OF ORIGINAL (REFERENCE) COMBUSTOR
800
1000
-------
-.J
W
:;)
I.J..
~
~ 3.0
:2
~
VI
Z
o
I-'
c:.n
VI
VI
:2
w
N
o
Z
5.0
4.0
2.0
1.0
o
o
CONTRACTUAL LIMIT
8"10 LlP /P
AVERAGING PROBE - HOLES DOWNSTREAM - EXTENDED TAILPIPE
NA TURAL GAS - HIGH VELOCITY INJECTOR
6't~Ll PIP
I
I
I
I
COMBUSTOR INLETI
TEMPERATURE I
750°FI
800
I
i
J
1000
41oLlP/P COMBUSTOR
PRESSURE DROP
FIGURE 5.
400 600 800
COMBUSTOR TEMPERATURE RISE (OF)
EMISSION CHARACTERISTICS OF MOD. 9 COMBUSTOR
-------
1--
......
0':>
:J
w
:::>
LL
~ 25
"-
:2
8
z
a
~ 20
:2
w
a
u
45
40
35
30
15
10
8')',L'.P
P
5
CONTRACTUAL LIMIT
o
AVERAGING PROBE - HOLES DOWNSTREAM - EXTENDED TAILPIPE
NATURAL GAS - HIGH VELOCITY INJECTOR
1000
200 400 600 800
COMBUSTOR TEMPERATURE RISE (oF)
200
1000
400
FIGURE 6.
EMISSION CHARACTERISTICS OF MOD. 9 COMBUSTOR
600
4 ')',L'.P
P
1000
-------
of running on natural gas fuel at atmospheric pressure, and Figures 7 through 12
depict the results obtained from runs at pressure. The NO results were the lowest
obtained during the series of combustor modifications and are below the allowable
limit at the lowest inlet temperatures and at atmospheric pressures.
The final Mod. 9 combustor configuration was tested in some detail. In
addition to the usual atmospheric runs, the combustor was tested at various pressure
levels with two inlet temperatures, 750 and 1000°F. The effect of gas nozzle injection
velocity is also shown with two different pressure runs at 750°F inlet temperature.
The pressure dependency of the NOx formation rates can best be seen at the
higher combustor temperature rises. At the lower temperature rises the dependency
is masked by the mixing effects, as mentioned before, although this effect is reduced at
the higher inlet temperature (1000° F). Although the basic NO reaction rate equation
contains the pressure to only the one half power, there appears to be only a slight
tendency towards a reduction of pressure effects at higher pressure levels. In all three
sets of pressure tests, however, there are reversals where the NOx emissions at a
high pressure are lower than those at some intermediate pressur~. The effects of a
change in gas injection velocity can be seen from Figures 7 and lL. The repeatability
is considered good for those pressure levels that are common to both figures, although
some mixing effects are apparent at the low temperature rises.
3.2.2
Rapid Mixing Transition Zone Combustor
The final example of a rich primary zone combustor which was tested incorpor-
ated a specially designed transition zone between the primary and dilution sections.
The main purpose of this zone was to facilitate the extremely rapid mixing between
the fuel rich primary zone product gases and the secondary air, necessary to prevent
the formation of long-lived pockets of stoichiometric mixture. A schematic of the com-
bustor showing the main features of the system is provided in Figure 13. Premixing
of the air and fuel for the primary zone was accomplished with the swirling air atomizer
system shown. This swirler also provided part of the stability by creating a flow
reversal within the primary zone. An increase in the amount of products recirculated,
thus adding to the stability, was obtained because of the blockage created by the plenum
chamber supplying air for the transition mixing zone.
ConceptUally, it was felt that a system with a primary zone sufficiently fuel-
rich to prevent high temperatures and consequent NOx formation, together with a
secondary and dilution zone so lean to similarly obviate the production of NOx, could
be successful if a transition mixing system could be designed to provide sufficiently
rapid mixing rates. Provided these mixing rates are sufficiently high, dwell times
at high equivalence ratios or temperature levels could be minimized, which would in
turn minimize the NOx emissions from this zone. The mixing rate in the transition
zone would have to be considerably faster than the fuel oxidation rate at any instant
17
-------
:::;
w
:::;)
I..L.
<.:) 3.9
~
......
:2;
<.:)
VI
Z
0,
VI
VI
:2;
W
N
o
Z.
. ,!" ~
4.0
AVERAGING'PROBE - HOLES UPSTREAM - EXTENDED TAILPIPE
COMBUSTOR PRESSURE DROP - 3.5'/0 - 4.5'/0 .'"
COMBUSTOR INLH. TEMPERATURE - 750?F ; .
NA TURAL GAS - LOW VELOCITY GAS INJECTOR
," ,"
2.0
. .'
"
...
I' ,>
'"
o COMBUSTOR INLET PRESSURE
15 PSIG '.
.0
'.
..'
1.0
100
'1000
FIGURE 7. EMISSION CHARAQTERISTICS OF MOD. 9 COMBUSTOR
18
-------
140
130
120
110
100
15 PSIG
COMBUSTOR
PRESSURE
90
-'
w
::>
IJ.
~ 80
""
"-
~
8
V1
:z
2
V1 70
V1
~ '
-------
6.0
-
...I A
w
;:)
LL
<.:I 5.0
:..:
.......
:2
8
(J')
:2
a
(J')
(J') GI
:2 4.0 [:J
W
N
a
:2
3.0
9.0
8.0
7.0
2.0
1.0
100
CONTRACTUAL LIMIT
AVERAGING PROBE - HOLES UPSTREAM - EXTENDED TAILPIPE
LOW VELOCITY GAS INJECTOR - NATURAL GAS
COMBUSTOR INLET TEMPERATURE - 1000°F
COMBUSTOR PRESSURE DROP - 3.5';', - 4.5';',
400 500 600 700
COMBUSTOR TEMPERATURE RISE (0 F)
1000
FIGURE 9. EMISSION CHARACTERISTICS OF MOD. 9 COMBUSTOR
20
-------
60
40
:J
UJ
=>
L!..
<..:)
::<:
"'-
:2
<..:)
;;:; 30
:2
o
VI
VI
:2
UJ
o
u
AVERAGING PROBE - HOLES UPSTREAM - EXTENDED TAILPIPE
NATURAL GAS - LOW VELOCITY GAS INJECTOR
COMBUSTOR INLET TEMPERATURE - 1000°F
COMBUSTOR PRESSURE DROP - 3.5'10 - 4.5"/0
20
700
CONTRACTUAL LIMIT
10
00
800
900
FIGURE 10.
EMISSION CHARACTERISTICS OF MOD. 9 COMBUSTOR
21
1000
-------
:J
UJ
::>
IJ..
<.;)
::.::
~ 3.0
<.;)
V'I
Z
o
V'I
V'I
:2;
UJ
N
o
Z
5.0
4.1
2.0
1.0
AVERAGING PROBE - HOLES UPSTREAM - EXTENDED TAILPIPE
NATURAL GAS - HIGH VELOCITY GAS INJECTOR
COMBUSTOR INLET TEMPERATURE -750°F
COMBUSTOR PRESSURE DROP - 3.5'Yo - 4.5'Yo
CONTRACTUAL LIMIT
15 PSIG
o
COMBUSTOR PRESSURE
100
1000
FIGURE 11. EMISSION CHARACTERISTICS OF MOD. 9 COMBUSTOR
22
-------
160
150
140
130
15 PS IG
120
110
100
a 90
:::>
...
....
'"
;;
....
:;; BO
z 0 PSIG
o
Vi
In
~
~ 70
u
60
50
40
CONTRACTUAL LIMIT
I
AVERAGING PROBE - HOLES UPSTREAM - EXTENOEO TAILPIPE
NA TURAL GAS - HIGH VELOCITY GAS INJECTOR
COMBUSTOR INLET TEMPERATURE - 750.F
COMBUSTOR PRESSURE DROP - 3,5Y. - 4.5Y.
200
300
400 500 600
COMBUSTOR TEMPERATURE RISE C. FJ
700
1000
30
20
10
o
o
100
FIG T)RE 12.
EMISSION CHARACTERISTICS OF MOD. 9 COMBUSTOR
23
-------
SECONDARY AIR
SECONDARY AIR
DILUTION AIR
FLAME ZONE
=
AIR SWIRLER
.0
o
o
o
FIGURE 13.
= \INTENSE MIXING ZONE
RAPID MIXING.TRANSITION ZONE COMBUSTOR
during the mixing, if the producti01?- of NOx is to be minimized. In an attempt to
provide such mixing rates, a transition zone was constructed to provide :opposed
air jets, each offset from the other. i The resulting turbulent zone along the edge of
each jet was intended to provide rapid mixing of the fuel rich product gases from the
primary zone as they exit with the incoming secondary air.
The results obtained on the NOx emissions indicate that at atmospheric
pressure the contractual limits can almost be met, but the emissions of carbon
monoxide are outside their limit'~ as can be seen in Figure 14. At higher pressures
the results show that although the emission of carbon monoxide is much reduced, the
NOx levels are substantially increased. The reduced NOx levels at the atmospheric
condition could be due to the loW combustion efficiency, as'indicated by the high emission
level of carbon monoxide at the, same conditions. Low combustion efficiencies in
general would reduce local high temperature zones and thus reduce the NOx produced.
At higher pressures since the combustion efficiency is approximately proportional to
the pressure squared, the increased combustion efficiency ensures low carbon monoxide
levels but high emission rates of NOx.
3.3
LEAN PRIMARY ZONE COMBUSTORS
One of the. most promising alternate approaches to the de~ign .of a low emission
(particularly low NOx) combustor determined during the conceptual model phase of the
program was the lean primary zone system. In this approach the problem of low NOx
emission is' 'essentially traded for a problem in' flame stability. Provided some means
24
-------
25
::J
~ 15
lJ..
t.:)
~
~
8
, 10
V1>
z
9
V1
V1
~
uJ 5
o
u
CONTRACTUAL LIMIT: 11.8 GM/KG FUEL
20
o
12
10
::J
uJ
::)
lJ.. 8
t.:)
~
~
8
~
~
0-\,\0
~'
q,c.;
~c.;
~ 6
Q
V1
V1
~
uJ
~ 4
z
2
CONTRACTUAL LIMIT: 1.38 GM/KG FUEL
\~
":?~
q;'o
~~
~X;
c.,~
~r:::,
~
o
100 200 300 400 500 600 700 800 900 1000
COMBUSTOR TEMPERATURE RISE - ( T> "F
-2
0.022 0.045 0.068 0.090 0.114 0.138 0.163 0.187 0.213 0.238
EQUIVALANCE RATIO
FIGURE 14. RESULTS OF RICH PRIMARY ZONE COMBUSTOR
25
-------
of stabilizing a flame at primary zone equivalence ratioslow, enough to produce local
zone temperatures less than 3200°F can be devised, then low NOx levels can be ob-
tained. In addition, sufficient residence time has to be provided in the primary zone
to allow the complete reaction of the fuel, thus ensuring concomitant low emissions of
CO and UHC. Generally, in lean primary zone combustor designs the problem of
mixing the fuel and air is more difficult than that in the rich primary zone des i~s,
mainly because of the increased volume of air to be mixed with the fuel. The percentage
of fuel in the primary zone mixture drops from 2 percent for the fuel rich system to
0.55 percent for the fuel lean approach. ' ",'
Several different methods of obtaining stability at the low equivalence ratios
necessary were evolved and these are discussed in the following section.
3. 3. 1
Lean Primary Zone Pilot System
One of the major problems, as mentioned above, of the homogeneous lean
stabilized system is the reduced stability range at low temperature rise levels or low
overall equivalence ratios. This in turn indicates a definite need for equivalence ratio
control through variable area or geometry control. Since the mechanical control of the
flow area could become cumbersome and possibly difficult to operate, an approach was
sought which might eliminate the need for such control. The concept used initially to
provide the combustion stability needed, over the entire operating range, was that of
a" "pilot" combustion system. In operation, the main fuel and air were premixed and
introduced upstream of a conical stabilizer, which is shown in Figure 15. This
stabi~~,z,er also provided the housing for the pilot flame. '
,
It can be seen in Figures 16 and 17 that although the contractual limits for NOx
are met for a wide range of conditions, the emissions of carbon monoxide and hydro-
~arbons are generally above their limits. This poor efficiency was caused basically by
using a single flame-stabilizer device with medium blockage (approximately 32%~. If
multiple, annular ring'stabilizers had been used, the efficiency could have been much
highe'r and consequently the emissions of carbon monoxide and unburned hydrocarbons much
lower. As aresult of the low combustion efficiency, it could be argued that the local
temperatureE! were much reduced, and that this in turn led to the low levels of NOx. If
this were true, - t~en increasing the efficiency by designing a better s tabilizing d~vice
could lead to increased NOx emission levels.
3.3.2
Lean Primary Zone Vortex Induced Circul:ition System
. " .one proven.method for achieving rapid mixing of the fuel and air is a vortex
or swirl generator. These devices, by vir~ue of the high shear gradients created, tend
to have high mixing rates. In this concept the method of recirculating a large quantity
of hot products provides the stability required;; :by both preheating the incOniing air and
fuel and then igniting the mixture. The preheating effect is important since, by lowering
26"
-------
MAIN FUEL
INLET
AIR
INLET ---
FIG URE 15.
..J 3.0
UJ
::>
"-
'"
~
~ 2.5
II
'"
Z
o
~
:i 2.0
UJ
N
o
"
PILOT FUEL
INLET
DILUTION
AIR INLET
MIXING
ZONE
FLAME
ZONE
C)
o EXHAUS~
LL F~ME
STABILIZER
SCHEMATIC OF LEAN PRIMARY ZONE PILOT FLAME
COMBUSTOR
MIXER ~
C)
4.5
PILOT FLAME ONL Y
NA TURA L GAS
T =lOOO'F
INLET
P ;ATMOSPHER~
INLET
4.0
3.5
1.5
CONTRACTUAL LIMIT: 1.38 CM/KG
.. --
1.0
.5
o
0 100 200 300 400 500 600 700 800 900 1000
COMBUSTOR TEMPERATURE RISE, (t>T! OF
I I I I I I I ,
0.022 0.045 0.068 0.090 0.114 0.138 0.163 0,187 0.213 0,238
EQUIVALENCE RA no
FIGURE 16. RESULTS OF LEAN PRIMARY ZONE PILOT FLAME COMBUSTOR
27
-------
..J
W
:;;)
...
<.:>
""
~ 150
<.:>
V1
z
a
V1
~ 100
:;;.
w
. ""
CX)
....
:I:
U
300
..J
w 250
:;;)
...
<.:>
'"
"'-
:;;
<.:>
. 200
V1
z
a
V1
V1
:;;
~ 150
u
. 100
FIGURE 17.
,~ "
'. .
. : . , \. '\,
200
~ ~ ~ .
T = 1000'F
INLET
P 0: ATMOSPHERIC'
INLET
. ,
50
o
CONTRACTUAL LIMIT 0: .1,42 GM/KG FUEL
350
.",
50
PILOT FLAME,
ONLY
o
300 400 500. 600 700
COMBUSTOR TEMPERATURE RISE, (L~T! - of
I
0.022
I
0,068
I I I
0.090' 0,1140,138
EQUIVALENCE RATIO
I
0,045
I
0,163
I
0,187
I
0,213
I
0.238
RESULTS OF LEAN PRIMAItY ZONE PILOT FLA,ME COMBUSTOR
-' '. ,,','. ..' , . '. . ': . .
28
-------
the ignition temperature, it allows stabilization of the combustor at primary zone
equivalence ratios lower than the room temperature flammability limits.
The results of running the swirl or vortex stabilized lean primary zone
combustor, as shown in Figure 18, with natural gas as fuel, are shown in Figures 19
and 20. These show that the NOx level increases rapidly as the equivalence ratio in-
creases toward the stoichiometric value. Essentially, this effect appears to be a
chemical kinetic controlled mechanism, in which increased combustion temperature
increases the reaction rate. As the equivalence ratio decreases toward zero, the
NOx level again increases rapidly, as does the carbon monoxide level, which is due to
a rapid deterioration in mixing between the air and fuel. It is important to Dote that
due to rig limitations the combustor mass flow was such that the pressure drop across
the combustor during this test was slightly in excess of 30 percent.
A second version of this type of combustor was tested with the gaseous fuel
injector replaced by an air-assist liquid fuel atomizer. Detailed tests with this arrange-
ment of atomizer revealed poor fuel atomization and mixing characteristics, resulting
in large areas at near stoichiometric conditions and, consequently, high NOx emission
levels. In order to evaluate the effects of fuel atomization and mixing, and also to
provide lower NOx emissions, a high pressure, simplex, swirl atomizer providing a
spray with a SMD* of less than 40 microns was incorporated, replacing the earlier air-
APPROXIMATELY SCALE
LIQUID
FUEL
}=.
o
SWIR LER
t
SWIRL TUBE
o
SWIR L OR
RECIRCULATION ~
CHAMBER~ f--
,.,,".J
I.
FIGURE 18. SCHEMATIC OF VORTEX INDUCED CIRCULATION (VIC)
COMBUSTOR WITH AIR-ASSIST ATOMIZER
*SMD is Sauter Mean Diameter
29
-------
:J
~ 3..0
I.L.
'"
~
E
'"
2:
o
;::2.0
«
0:::.
f-
2:
.w.
u
~,
U
N 1.0
o
2:
4.0
@
INLET TEMPERATURE 1000"1=: .'
INLET PRESSURE 21.2 PSIA
FUEL - NATURAL GAS
CONTRACTUAL LIMIT
-----------------
@
'0.02 0.04 0.06 0.08' 0.1 0.12 0.14 0.16 0.18 0.2
. EQUIVALENCE RATIO (OVERALU
FIGURE 19. VORTEX STABILIZED LEAN PRIMARY ZONE
COMBUSTOR (N0x EMISSION)
assist system. In addition to the simplex atomizer, a further change was incorporated
into the design. This change consisted of a cooling strip on the back face of the
recirculation zone chamber, as shown in Figure 18. Incorporation of this cooling
device was to correct problems encountered during the air-assist atomizer testing
when corner weld failures held up the progress of testing. As was expected, the
trend of the NOx emissions is similar to those obtained with the air-assist atomizer but
the emissions were lowered and approached the results provided by natural gas, obtained
earlier in the program.
Because of the success of the liquid fuel high pressur6, simplex atomizer
combustor, in meeting the contractual limits in the _exploratory tests, it, was decided that
a map of the emission characteristics at various pressure levels was necessary for a
full evaluation. A map of the NOx emission level at a 1000°F inlet temperature, and
with a constant pressure drop, was recorded as a function of combustor temperature
rise at various inlet pressures. The results of these tests are shown in Figure 21.
As can be seen, the atmospheric pressure results do not extend down to the low tempera-
ture rise levels, and thus do not go below the contractual limits. The reason for the
omission of these points is that failure of the gas sampling system occurred during the
test. .It can be. surmised from earlier test results with similar trends, that the NOx
emissions would fall below thecont~actual goals ~t low temperature rises.
30
-------
1000
NATURAL GAS
o
o
~ 100
~
lL.
l.:)
:.::
"-
2
l.:)
V1
Z
o
V1
V1
10
o
2
LJJ
LJJ
o
X
o
z
o
2
z
o
CD
a::
«
u
CONTRACTUAL LIMIT
o
1
o 0.02 0,04 0.06 0,08 0.10 0,12 0,14 0,16 0,18 0,20 0,22 0.24
EQUIVALENCE RA TIO
FIGURE 20, VORTEX STABILIZED LEAN PRIMARY ZONE COMBUSTOR
(CO EMISSIONS)
31
-------
---.-.--
27.3
5.0
CONTRACTUAL LIMIT - 11.8 GMlKG FUEL
- - --.-- _.-
- ---- -'-
:J
w
:::J
lJ..
~ 4,0
::;;
::!
-------
As the pressure level increases, the trend is for the curves of N02 versus
temperature rise to increase their overall N02 levels, although this dependency is not
as marked as that exhibited by the initial rich primary zone combustors.
3.3.3
Lean Primary Zone Jet Induced Circulation System
A third lean primary zone approach was also investigated where a large scale
recirculation with recirculating mass flows greater than the inlet flow rate by a factor
of 1. 4 was accomplished by an ejector method. In this system, the injected air and
fuel are premixed and act as the motive jet in a jet educer in which the fluid entrained
is the hot products of combustion. Using this method of stabilization it is possible to
provide stability at an equivalence ratio of about 0.28 in the primary zone with a pre-
mixed air and fuel mixture. This model was not produced as a complete combustor
but as a primary zone alone without a dilution section.
In Figure 22 the results of the ejector recirculation system show a fairly con-
stant trend for the NOx level versus equivalence ratio. Little variation of the carbon
monoxide (Fig. 23) emissions with equivalence ratios greater than 0.40 is also evident.
The increase in carbon monoxide at lower equivalence ratios was considered a chemical
kinetic phenomenon. Apparently at these low equivalence ratios the hydrocarbon
reaction rates decrease rapidly, creating quantities of carbon monoxide. This parti-
cular configuration was constructed without any form of cooling, which prevented
14.
. ------------------------
CONTRACT GOAL=1.38 GM/KG FUEL@ 10 MPG JP4
'lleolll
1,2
uJ 1,0
~
u.
""
~
"-
~ 0,8
ATMOSPHERIC PRESSURE
MAIN FUEL
INLET
!
EXHAUST
COMBUSTOR INLET
TEMPERATURE 10000F
G REClRCUlAT"GFl.. f
l-
FLAME
ZONE
VI
Z
o
~ 0,6
:2:
UJ
N
~ 0,4
FUEL-NATURAL GAS
J FLAM: OUT
0-
0,2
o
0,25
0.30
EQUIVALENCE RATIO
0,35
FIGURE 22. HIGH RECmCULATION STABILIZED LEAN PRIMARY
ZONE COMBUSTOR (N0x EMISSIONS)
33
-------
----------------------------------
Contractual Limit
10
- 9
..J
t.J
;:;
t.. 8
"
:.:
"-
:; 7
2-
'"
z
~ 6
'"
51'-
t.J 5
t.J
o
~ 4
z
o
::g 3
z
o
C!:>
~ 2
U
o
0.1
0.2
0.3
0.4
o. :)
0.6
0.7
E'\.UlVALENCE RATIO
FIGURE 23. ,HIGH RECmCULATION STABILIZED LEAN PRIMARY
ZONE COMBUSTION (CO EMISSION~)
extended testing at equivalence ratios greater than 0.4. In addition, the injected air
and fuel was fully premixed, which in general reduced the stability limits. The lowest
equivalence ratio at which stable combustion occurred was approximately 0.28. This,
if considered on an overall combustor basis, remembering that the model was com-
posed only of a primary zone, would represent an equivalence ratio of about 0.07.
Developments of this initial recirculation combustor were also tested. In
these combustors a high level of recirculation was obtained by injecting the primary
air an~ partially premixed liquid fuel through a series of ports angled toward the
dome at a 45 degree angle to the combustor axis. These angled jets impinged at the
combustor axis, forming two axially derived jets, one moving towards the dome and the
other downstream toward the exhaust. The major portion of the flow moved toward the
dome, providing a jet that behaved in a similar fashion to the motive jet of an ejector.
A schematic of this jet induced circulation (JIC) system is shown in Figure 24.
A measure of premixing between the liquid fuel and the air is obtained by using multiple
air atomizers as the primary zone feed ports. This premixing prior to reaction mini-
mizes the production of stoichiometric "islands" or areas, where NOx can be formed,
that are normally associated with co-current mixing and reaction. In order to achieve
stabilization at these extremely lean equivalence ratios a high degree of recirculation
34
-------
//PRIMARY FUEL (LIQUID)
/
<
o
o
"-
~
"
o
Q
,
,
,
,
ENTERING MEAN EQUIVALENCE
RATIO 0.36 TO 0.45
'\
DILUTION AIR
FIGURE 24. SCHEMATIC OF JET INDUCED CIRCULATION COMBUSTOR (JIC)
is required, and this has been provided as mentioned above, by the movement toward
the combustor dome of the motive jet produced from the interaction of the impinging
jets from the primary zone air atomizers. Recirculation ratios obtained in this
particular design are approximately 60 to 70 percent of the primary zone inlet mass
flow. A reduced performance map of this combustor was produced and is shown in
Figure 25. These results indicate that with an inlet temperature of 1000°F at both
atmospheric conditions and at 40 psig the contractual requirements are met over the
equivalence ratio (overall) range of approximately 0.1 to 0.126.
The results of the tests at an inlet temperature of 750°F surprisingly show
higher NOx levels than those recorded with a 1000° F inlet. This indicates that the NOx
produced is mainly mixing controlled under these conditions since the major phys ical
change between the two tests is that of the rate of vaporization of the fuel. At the higher
inlet temperatures increased vaporization rates result in overall increased air-to-fuel
mixing rates and consequently fewer deviations from the mean primary zone equivalence
ratio. With the combustor primary zone operating at an essentially uniform (premixed)
equivalence ratio, no deviations toward the stoichiometric region occur and hence no
"islands" of high temperature occur, which at these lean conditions would produce NOx.
With the success of the Jet Induced Circulation (JIC) combustor in meeting the
contractual requirements at ambient pressure, it was decided to pursue the investigation
further and obtain the emissions index at a variety of pressure levels~ The results of
35
-------
24
22
20
2 18
::;)
u..
~ 16
'-
:;;
<..:>
--; 14
tf)
z
~ 12
tf)
tf)
i:j 10
o
u
24
22
20
18
-'
w
~ 16
<..:>
'"
'-
:;; .14
<..:>
~ 12
c
tf)
~ 10
:;;
w
. ON 8
Z
FIGURE 25.
JET INDUCED CIRCULA TION COMBUSTOR (JIC II) 6-17
8
CONTRACTUAL LIMIT: 11,8 GM/KG FUEL
6
4
2
o
6
0--0 750' F INLET TEMPERATURE
COMBUSTOR INLET PRESSURE; ATMOSPHEmC:
:lti; PHESSUlh: DUO})
o---a 1000' F INLET TEMPERATURE
COMBUSTOR INLET PRESSURE; ATMOSPHEmC,
11. 5fh PRESSURg DROP
lr---l> 1000°F INLET TEMPERATURE
COMBUSTOR INLET PRESSURE = 40 PSIG
11;5'10 PRESSURE DROP'
FUEL: JP-4
4
200
300. 400 500. 600 700
COMBUSTOR TEMPERATURE RISE - (6T! OF
1000.
I
0,022
I
0,068
I
0,213
I
0,238
I I I
0.090 0,114 0,138
EQUIVALENCE RA no
I
0,187
I
0,045
I
0,163
RESULTS OF JET INDUCED CIRCULATION (JIC) LEAN PRIMARY
COMBUSTOR (INITIAL FUEL INJECTOR POSITION)
36
-------
these investigations are shown in Figures 26 and 27. A slight modification to the pri-
mary zone ports/air blast atomizers was made to the model before running the
above tests, which consisted of moving the fuel injector tip further out toward the
cooling annulus. This modification in essence provided a longer air-fuel mixing
length before reaction thus allowing the production of a.nearly homogeneous equivalence
ratio in the inlet reactant stream.
As the results indicate, this particular combustor concept is capable of meeting
the contractual requirements of both the emissions of NOx and CO over a discrete range
of (overall) equivalence ratio, approximately 0.057 to 0.1. The corresponding primary
zone equivalence ratio range, which is the controlling factor for the emissions at a
fixed set of inlet conditions, varies from 0.175 to 0.3 approximately. It may be noted
from the results that the stability range is less at low pressures, near ambient, than
at higher values. This phenomenon may be related to reduced fuel droplet penetration
at the higher pressures resulting in a higher peak equivalence ratio in each of the
primary zone jets. If this is the correct explanation to the results then a similar effect
could be achieved at low pressures by reducing the fuel velocity at the fuel injector tip.
3.4
PARAMETRIC COMBUSTOR DESIGNS
The conceptual designs of both the Class A and B combustors were based on
the jet induced circulation (JIC) model combustor configuration (see Fig. 24) and have
been designated JIC-3 and JIC-4, respectively. The key feature of the design is the
combined primary zone, ports and air blast atomizers, which can be seen in the assembly
drawing, Figure 28. By incorporating an air-blast atomizer into each of the primary
zone inlet ports, which number eight in total, a very even fuel to primary zone air ratio
distribution is achieved. Although this multiple atomizer arrangement poses some
mechanical difficulties, it has shown itself to be considerably superior in performance
to a single fuel injector. In general, it is extremely difficult to provide the same
degree of mixing with a single atomizer that can be obtained with multiple fuel injectors,
even when high pressure air-assist atomizer types are considered. In addition to the
equivalence ratio uniformity obtained by having multiple fuel injectors, by arranging the
fuet'to be injected into the primary zone air jets, a substantial mixing length prior to
ignition and reaction is obtained. This mixing length is increased further by angling
the inlet ports to the combustor axis, as shown in Figure 28. Thus by combining the
effects of multiple fuel injectors and a premixing zone, a nearly homogeneous fuel and
primary zone air mixture can be obtained. The primary air and mixed fuel jets impinge
on the combustor centerline and form two axially derived jets, one moving toward the
dome and the other rearward toward the exhaust. The effect of angling the ports
toward the dome is to ensure that the bulk of the fuel-air mixture moves into the
primary zone section. This main derived jet impinges on the combustor dome, which,
in conjunction with the combustor walls, causes the mixture (substantially reacted by
this time) to recirculate toward the incoming primary air jets. .
37
-------
24,0
22,0
20,0
18,0
16,0
:J
lJ.J 14,0
:::>
LL
<.:>
::<:
"-
:2
<.:>
;:;; 12,0
z
o
-------
2 4.0
:::>
...
'"
><
"-
:;
~ 3.5
I
VI
Z
o
VI
VI
~ 3,0
w
o
u
FIGURE 27.
-. ---. - ----- - - -.-.- - ---
-----
CONTRACTUAL LIMIT = 11,8 GM/KG FUEL
FUEL: JP.4
COMBUSTOR INLET TEMPERATURE 1000 of
6.0
5.5
5.0
JET INDUCED CmCULATION
JIC-2 COMBU1n"OR (RETRACTED INJECTOHS)
COM BUSTOH ~p/p 8.8';
COMBUSTOR INLET
PRESSURE - 16 PSIA
o
4.5
2.5
8.
2.0
1.5
1,0
o
.5
I I I I I I I I
100 200 300 400 500 600 700 800 900 1000
COMBUSTOR TEMPERATURE RISE - T ( . F)
I I I I I I I I I I
0,022 0,045 0.068 0,090 0.114 0,138 0.163 0.187 0.213 0.238
EQUIVALENCE RATIO
RESULTS OF JET INDUCED CIRCULATION (JIC) LEAN PRIMARY ZONE
COMBUSTOR (MODIFIED FUEL INJECTOR POSITION)
39
-------
..-
--
!> t 410 AO:J. + bJlr SEE SH f.
"-t 409 ';) T.OP
I '+-1~~ ~IN~ - SH ~€
I ~ 4OE. . ADJ. 8Q..T
~l4OS PM:K..N\}!.-.-
lit ~. ~04 PR!. 8050:. - - --
\t -~~-- ::~ Tlft: I
.4~L_-- ~O~ .. -~
- -~--""!
~5~ A:~1- DI~ .!~-~ ~ _5U: S~-~ -- ------+
~--~ --- - I
PORT I=I1II(, 5..£E 5H 5 I
-2-)0
, ,
LJ
tr=~-1
~.~-o
(OUTlR (AS[ ':'5..1)
SEE s..."
. "0
"--
AEAR
(/("", ~H 5'"' 5
'-~._~
22~
FRONT ( -i SlE 5,", '!o
~ ~ '-"~+f-'- ~~L(,A~E p- S~~-~ ~
~10
DOMe:
20.
2e,
2 "'
20'
'::(12
20'
OCV",LER ~E '5...)
LC'e. foci"''' SloE. S,", )
SPI.A'SHPlATt ..
(roo.'I<.. :TRI
c.f.<:\.,N(. :TR
Tu",
----1-----
I II' HC.'IOR.....
I --I"i'S-- 1"1 r 1Iv..
I 114 t>ASlo:l T
1'~.-5lff;';l
112 ":ZI.[ A""">
II t HOvS'N(.
'02
'0'
-
-
5'A....~t'H 5~t 5- <-
pc~r AJ,,(, 'in Sw 2
U21
SO. "'A
2.0Dt....O~w X2."
-
~Ob,,20C I
--;oTftciEil.1
~~!)'ZCI
"O~5t 1...1
<;C51",(<'1
.,OSSS~C I
--- U8I' 01 --- ~ I.8fI
=Y£EF'.=,!5.'ir..~.r::=~-Zr~
-~
---..
..w ." "I.' ,.
DO ;:'L,(. ~ 12, i, ,
.. '
. TI\...", I.. - 7 -~I
HI:.9 ':'I,r~-
.
~ ~t.
..~
H
G
F
E
!-
D
'e
I!
B
':f\
00:;" 13' .
FIGURE 28.
ASSEMBLY DRAWING OF PARAMETRIC COMBUSTOR CLASS A
41
-------
By splitting the primary zone air into a number of jets, the rate of entrainment
and thus mixing is much increased over a single jet produced by the same combust.or
pressure drop. This is to be expected since the entrainment rate increases as the
length to diameter ratio of the jet is increased. Consequently, by arranging the ai r to
enter as a number of discrete jets, and by adjusting the aerodynamic blockage of these
jets, the quantity of hot products entrained can be maintained at any desired value. Thjs
quantity entrained fixes the recirculation ratio of the paired vortex system in the
primary zone, and determines to a large extent the stability range of the combustor. It
has been shown earlier, with the first of the jet induced circulation combustors, that
with high recirculation rates and long reaction lengths or times, the stability of the
primary zone can be extended to primary zone equivalen~e ratios as low as 0.28.
At equivalence ratios as low as this, provided the system is well mixed, virtually
no NOx is produced. With this arrangement of primary air injection and fuel atomi-
zation it is apparent that good air and fuel mixing is achieved, together with a high
level of hot pr ,ducts recirculation, which can be adjusted to meet the low equivalence
ratio stability requirements.
In order to minimize the NOx emissions over a broad range of fuel to air ratio
conditions, it becomes necessary to introduce some variability into the combustor to .
provide a constant or fixed primary zone equivalence ratio schedule. The primary t
zone equivalence ratio is critical in obtaining low NOx emissions, and it has been shown
that a value of the order of 0.3 is necessary to provide levels below the contractual
requirements when the inlet air temperatures are of the order of IOOO°F and pressu~es
in excess of 70 psia. There are several ways of achieving a constant primary zone
- .
equivalence ratio. The simplest mechanical system would be one in which the secondary
or dilution port area was reduced to provide the desired area and thus mass flow
proportionality. This simple approach suffers from the rather severe disadvantage Of
high pressure drops at high fuel flow conditions, since the total flow area of the com-
bustor would be decreased. Consequently, to obviate the effect of pressure drop during
the parametric combustor investigation, both the primary and dilution areas were.
varied simultaneously to maintain a constant pressure drop and the correct primary zone
equivalence ratio. .
3.5
PARAMETRIC COMBUSTOR RESULTS
3. 5. 1
Class A-Mod (JIC-3)
Al (Federal Driving Cycle) Test Mode
The first section of the test program involving emission testing was the
section Al of the simulated Federal Driving Cycle test mode. This FDC test m~de con-
sisted of a series of steady state combustor inlet conditions specified by EPA which were
designed to simulate the emissions from a vehicle driven over the Federal Driving.
Cycle. The various steady state point conditions are given in Figure 29. A s'eries of
43
-------
A.1. Simulated Federal Driving Cycle Mode
Wf PI Tl T2 Wa* Time
(pph) (psig) (oF) (oF) (Ib/sec) ('/0> (Sec)
6 18 1380 1560 0.59 3 41
8 13 980 - 1300 0.44 34 466
10 13 1000 1400 0.44 22 302
11 13 1100 1550 0.'44 22 302
12 18 1380 1740 0..59, 18 247 "
20 13 1000 1900 0.44 1 14
100')', 1372
Compute the projected Federal Driving Cycle Emission Levels'
(HC, CO, NOx) by avera~ing, weighing time. ' ,
Total Distance: 7.50 mi es. '
(See Nomenclature)
A.2. Steady Speed Mode
Wf PI Tl T2 Wa* Veh.Speed,
(pph) (psig) (oF) (oF) (Ib/sec) (mph)
8.5 13 1060 1410 0.44 30
9 13 1330 '1700 0.44 40
12 ,18 1380 1740 ,0.59 50 (repeat
13 20 1320 1680 0.63 60 point>
20 25 1120 1600 0.74 70
32 33 950 1600 0.90 8U
40 40 960 1670 ,1.06 90
60 51 1070 1850 1.31 100
67 59 1100 '1900 1.50 108
*Reference only
FIGURE 29. CLASS (A-MOD) - LOW PRESSURE COMBUSTOR TE~TS
, '
~atched inse~ts for the primary and dilution ports had already been designed to pro-
vide what was throught to be the most effective primary zone equivalence ratio.,
, '
, ,
(\s a preliminary to the A1 testing, the combustor was painted with tempera,..
ture sensitive paints and test run at the A1#5 test ,conditions to evaluate the combustor
wall temperatu~es at the extreme case of 1380° F combustor inlet temperature. lr}/:Ipection
:qwealed that the maximum liner temperature was of the order of 1650° Fin the inte~7
mediate combustor zone between .the primary and dilution ports. No hot spots were, '
apparent and the circumferential temperature distributions were quite even. '
"
, Each test point in the Federal Driving Cycle was run as part of a band or :r~e
'of temperature rises at the correct inlet pressure, temperature and ,mass flow conditions.
,In general, each point or rang'e was repeated at least once at a slightly different,
pressure drop or mass flow to ensure that if the first mass flow was not quite corr,ect
an accurate estimate of the emission of the correct value could then be made. The
results obtained at each of the simulated Federal Driving Cycle test points are shown
in Figures 30 through 36. These basic results were then integrated over the driving
cycle according to the test procedure described in the contract and are depicted in ,
Table V. As can be seen, the emissions of nitric oxide, carbon monoxide (CO) and
unburned hydrocarbons (UHC) are in excess of the levels desired. On inspection of , ,
the test results, this failure to meet the CO and UHC emissions is due to high emissions
at two points of the Federal.Driving Cycle, #2 and #3. In fact, these high emissions
~ere due to incipient instab~lity at the desired test points, since the operating conditions
were such that the combustor was very close to the lean blow-out limit at the destgn ,
point temperature rise. It was felt the emissions of points #2 and #3 could be reduced,
by resizing the secondary ports, these modifications were later carded out. as part of a
program addendum, and did indeed reduce the integrated emissions below the 1976
Federal Standards (see Table XII).
44
-------
~ 20
::>
u.
. 18
'-'
'"
~ 16
:;;
'-'
, 14
x
o
z 12
...J 14
w
::>
u. 12
'-'
~ 10
:;;
'-' 8
,
:3 6
1.2
~ 1.0
::>
u.
.; .8
'"
"-
,:;;' .6
'-'
, 4
IJ"I'
co
...i .2
:I:
u 0
34
---.-
------ --
600
600
500
600
32
FIGURE 30. CLASS A-MOD COMBUSTOR - JP-4 - FDC #1
30
T1 = 13800F
T2 = 15600F
p)= 18.0 PSIG.
wa = 0.585 Ib./sec.
waV~ = .767
P
wf = 6.0 pph.
28
26 .
24
22 -
* OESIGN POINT
o .\P = 36.0" WATER
O.\P = 50.0" WATER
10
8
6
4
2
o
20
18
16
4
2
o
100
200 300 400
COMBUSTOR C>T - of
o tJ
300
C>T - of
400
45
-------
10 --
R 9ROoF T of
1 * DESIGN POINT
7 13 PSIG PI
...J 13000F T of 0 ,\P --, 28.0" WATER
w 6 2
::J 0 ,\P = 54.0" WATER
LJ.. wa -= 0,463 Ib.lsec,
..; 5 wa'/'r = 0 634
""
"; 4 -P-"
:2 wI = 8 pph,
<.:>
I 3
x
0
:2 2
1
0
300 400 500 600 700 800
COMBUSTOR ~T of
100
90
80
...J 170
w
::J
LJ.. 60
..;
~ 50
:2
<.:> 40
I
0 30
u
20
10
0
100 200 300 400 500 600 700 800
COMBUSTOR ~ToF
22 0
20
18
16
.1
w 14
::J
LJ..
,; 12
""
";
:2 10
<.:>
I 8
U'\
0:
..... 6
:r
u
4
2 0
0
600 700 800
FIG URE 31.
CLASS A-MOD COMBUSTOR - JP-4 - FDC #2
46
-------
24
22 T 1 = 10000F
°
20 T 2 = 1400 F
P - 13 PSIG
13 1 -
wa = 0.471lb.lsec.
...J 16 wa \fT = ,650
UJ ----p
::>
11.. 14
~ W f = 10 pph.
:>.: 12
';
2
~ 10
'
x *DESIGN POINT
a 8
z
O,\P = 44,0" WATER
6 O.\P = 52.0" WATER
4 O:\P = 28.0" WATER
2 *
FLAME
0 OUT
400 500 600 700 800 900
COMBUSTOR 6T - of
30
...J
UJ
::>
11.. 20
~
:>.:
'~
2
~ 10
0
u
0 600 7QO 800 900 1000
400 500
COMBUSTOR 6T - of
5
4
...J
UJ
::>
11..
~ 3
:>.:
'~
2
~
, 2
'"
co,
....
;x:
u
1
0 500 600 700 800 900 1000
400
COMBUSTOR Ln - of
FIGURE 32. CLASS A-MOD COMBUSTOR. - JP-4 - FDC #3
47
-------
'.
2,4
2,,2
~ ' .
, 2,0
1,8
C;j 1,6
:;:>
u.
, 1,4
'-'
::<:
'-; 1,2
:2
'-' 1,0
I X
o
z 0,8
0,6
0,4
0,2
o
..J
,UJ
. ~ 30
'-'
::<:
'-;
~. 20' ,. . ..'f
o
u
T 1 = 1l00oF
T 2 = 15500F
'u.
~ 20
::<:
i'.
::2
'-'
~ 10
co,
,...,
.:t:
.U
O. . :." ~,
100 200 . 300 400 , 500 600 700
COMBUSTOR. sr - of
:EK~.URE 3,3,... CLAf?SA-MOD COMBUSTqR -JP-4- FDC"#~
48
-------
')()
34 T 1 = 13800F
32 T 2 = 17400F
P = 18 PSIG
30 - 1
wil = .585 Ib.lscc.
28 wilVT =.767
P
26 wf = 12 Ib.lhr.
24
-' 22
I.J.J
::J
lJ.. 20
'-" * DESIGN POINT
~ 18
~ 0 \ P 40.0" WATER
::;;
'-" 16 0 \ P 48.0" WATER
'" t::. \ P 56.0" WATER
a 14
z
12
10
8
6
4
2
0
200 300 400 500 600 700
COMBUSTOR 6T - of
30
-'
I.J.J
::J
lJ.. 20
~
~
~
::;;
'-" 10
0
u
0
100 200 300 400 500 600 700
COMBUSTOR t. T - of
-'
I.J.J 2
::J
lJ..
~
~
~
::;; 1
'-"
I 0
Lf'I 6
00. 0
...... 0
:r: 100 200 300 400 500 60U 700
u
COMBUSTOR 6T - of
"
FIG URE 34.
CLASS A-MOD COMBUSTOR - JP-4 - FDC #5
49
-------
-' 4
L.LJ
::I
lJ..
'N
c.:IO
~23
. V1 '
::
-------
10
9 T 1 = 10000F
8 T2=1900oF
-l 7 P = 13 PSIG 0
11.J 1
=> wa = .393 Ib./sec.
Ll..
. 6 t.JaVT =
<.:> .542
~
" 5 P FLAME OUT
.
~ w f = 20 pph.
<.:>
I 4
x
0
:2 3
* DESI GN POI NT
2 0 L\P = 17.0" WATER
1
*
0 600 700 800 900 1000 1100 1200
500
COMBUSTOR 6T - of
10
9
8
-l 7
11.J
:;)
U.
. 6
<.:>
~
" 5
.
~
<.:>
I 4
0
u 3
2
1
0 700 800 900 1000 1100 1200
500 600
COMBUSTOR 6T - of
FIGURE 36. CLASS A-MOD COMBUSTOR -,JP-4 - FDC #6
51
-------
TABLE V
BRAYTON CYCLE COMBUSTOR CLASS A-MOD
JP-4, MODE Al
SIMULA TED FEDERAL DRIVING CYCLE
(Preliminary Results - See Also Table XII)
NOx NOx NOx I
lb/1000 lb. Fuel Flow lb/hr Time lb/test point
FDC # Fuel lb/hr x 103 sees x 106
:.Total Cycle = 1,878.5 x 10-6 lb NOx
1 4.5 6.0 27.0 41.0 307.50 per 7.5 miles
I 2 0.2 8.0 1.6 466.0 20~.10 :.gm/mile = 1,878.5 x 10-6 x 453.6
7.5
3 0.2 10.0 2.0 302.0 167.78
4 0.22 11.0 2.42 302.0 203.01 C' O.H36
5 Rev. 1 0.8 12.0 9.6 247.0 658.67 (Program goal = 0.4 gm/mile NO",
6 4.3 20.0 86.0 14.0 334.44
1; 1,878.5 Note: Using FDC # 5 instead of FDC
#5 Rev. 1 gives 0.6514 gm/mile
5 11. 6 12.0 139.2 247.0 9,550.67
CO CO CO
lb/1000 lb. Fuel Flow lb/hr Time lb/test point
FDC # Fuel lb/hr x 103 sees x 106 :.Total Cycle: 268,549.05 x 10-6 lb CO
1 .. 4.1 6.0 24.6 41.0 280.17 per 7. 5 miles
2 *200.0 8.0 1600.0 466.0 207,111.11 :.gm/mile = 268,549.05 x 10-6 x 453.6
7.5
3 *60.0 10.0 600.0 302.0 50,333.33
4 8.0 11. 0 88.0 302.0 7,382.22 C' 1623
5 Rev. 1 3.0 12.0 36.0 247.0 2,470.0 (Program goal = 3.4),
6. 12.5 20.0 250.0 14.0 972.22 Note: Using FDC #5 inst~ad of FDC
268,549.05 #5 Rev. 1 gives 16.14 gm/mile
5 1.0 12.0 12.0 247.0 823.33
* Estimates only. '
CH CH CH
1.85 1. 85 1. 85
lb/1000 lb. Fuel Flow lb/hr Time lb/test
FDC # Fuel lb/hr x 103 sees x 106 ,
1 0.9 6.0 5.4 41.0 61. 5
2 *40.0 8.0 320.0 466.0 41,422.2 :.Total Cycle = 44,405.37 x 106 lb CH
1. 85
3 2.5 10.0 25.0 302.0 2,097.2 per 7.5 miles
4 0.5 11. 0 5.5 302.0 461. 39 / 1 44,405.37 x 10-6 x 453.6
:.gm mi e =
5 Rev. 1 0.27 12.0 3.24 247.0 222.3 7.5
6 **1.81 20.0 36.2 14.0 140.78 C - 2.6"
1: 44,405.37 (Program goal = 0.41)
5 0.4 12.0 4.8 247.0 329.3
* Estimated only. Note: UsingFDC # 5 instead of FDC
** Not available - assume 1. 81 gm/kg fuel. #5 Rev. 1 gives 2.692 gm/mile
52
-------
A2 (Steady Speed) Test Mode
Tests at the Steady Speed Mode for the Class A-Mod combustor were run
over a band or range of temperature rises, and the trends exhibited by these results
follow closely those of the Federal Driving Cycle.
A summary of the results (including A 1) is shown in Table VI. In general,
these results are similar to those obtained in the Simulated Federal Driving Cycle.
As can be seen, the stability limit for most of these various test points does not
extend to sufficiently lean equivalence ratios to suppress the emission of unburned
hydrocarbons (UHC) and carbon monoxide (CO).
TABLE VI
BRA YTON COMBUSTOR PROGRAM FDC AND STEADY STATE
RESULTS, CLASS A-MOD COMBUSTOR
Data Tin L1T L1P% Calc. gm/kg Fuel P
Point of of P (j prim. NO ** CO HC** psig
x
I 1380 180 4.26 0.3233 4.5 4.1 0.9 18
2 980 320 4.81 0.3302 0.2 200.0* 40.0 13
3 1000 400 4.02 0.3799 0.2 60.0* 2.5 13
4 1100 450 4.15 0.3821 0.22 8.0 0.5 13
5 1380 360 4.99 0.3509 11.6 1.0 0.4 18
5Rl 1380 360 5.19 0.2682 0.8 3.0 0.27 18
6 1000 900 2.49 0.3821 4.3 12.5 - 13
7 1060 350 4.46 0.367 0.2 16.0 1.3 13
8Rl 1330 370 3.67. 0.275 0.3 10.0 - 13
9R1 1320 360 5.2 0.266 0.3 22.5 0.7 20
9R2 13~0 360 6.38 0.25 0.4 25.0 - 20
10 1120 480 4.59 0.376 0.3 4.0 0.55 25
14Rl 1100 800 4.52 0.4036 1.1 2.8 - 59
11Rl 950 650 3.13 0.411 2.0 110.0 16.0 33
13R1 1070 780 4.22 0.509 2.6 0.55 - 51
12R2 960 710 3.54 0.453 3.5 8.0 4.3 40
* Estimates only.
** HC expressed as CHI. 85' NOx expressed as NO 2.
- Not available.
53
-------
Several of the Steady Speed A2 test points were "rematched" in primary zone
equivalence ratio together with point #5 in the A 1 Simulated Federal Driving Cycle.
This was done where, as previously mentioned, the stability limit of the tested config-
uration was insufficient and flame-out was occurring at temperature rises higher than'
the design point temperature rise.
, Rematching was generally performed by increasing the size of the dilution
ports., This effectively increased the reaction zone equivalence ratio at the expense,
of a slight decrease in overall pressure drop. By reducing only the size of the dilution
ports, the recirculation and entrainment rates of the primary jets were unchanged.
Rematching in this manner gave generally predictable results, whereas when primary
port sizes were changed during rematching, the emission characteristics were more
difficult to predict.
A second method of improving the lean stability of a given combustor con-
figuration was to reintroduce the torch igniter fuel flow. Introducing a minimum,
amount of fuel flow to the torch certainly stabilized the combustion process to ~e~ow
the normal flame-out but resulted in some compromise to the NOx emissions. The
attempts were rather cursory in nature and merely served to demonstrate the. efficacy
of the technique, as optimization of torch ignited fuel flow or method of introduction
was not carried out.
Emissions of aldehydes and smoke were also monitored for selected points
of the A2 test mode.
No visible smoke was observed from the combustor at arw of the test points
comprising the Simulated Federal Driving Cycle (A 1) or Steady Speed (A2) test modes.
Nevertheless, exhaust smoke samples were taken from the rig tailpipe during test
pOInts #8R1 and #l1R1 of the Steady Speed (A2) test mode using a Von Brand Smoke-
meter operating at 0.108 cubic feet of sample per square inch of filter paper. The,
samples were drawn from a single-point probe mounted at the center point of the tail-
pipe and were taken over a wide range of temperature rises, from a value in excess of
the test point rise down to the lean limit point. Analysis of the resulting filter paper
tapes with a reflectometer revealed essentially zero smoke emissions at both of the
test points, the reflectometer readings being identical to the "clean paper" calibration
point.
It is concluded, therefore, that the Class A-Mod combustor operating on JP-4
not only produces no visible smoke at any operating condition, but that the sub-visible
smoke emissions are negligible.
54
-------
Grab samples for aldehyde analysis were taken during test points #7, #10,
and #llR1 of the Steady Speed (A2) test mode. The samples were drawn from the
tailpipe gas sampling probe into evacuated glass flasks and analyzed using the standard
3-MBTH method. The results are given in Table VII. The samples at all three test
points were taken at temperature rise conditions close to the flame-out point; hence,
it would be expected that the percentage of aldehydes in the total hydrocarbons would
be much greater than at higher temperature rise conditions. An improvement in flame
stability margin would result in a much lower percentage of aldehydes.
TABLE VII
ALDEHYDE ANALYSIS
Total Aldehydes
A2 Test Mode (H/C) From FID as Formaldehyde Aldehyde % of
Test Point # (ppm) (ppm) Total (H/C)
7 10.0 7.0 41. 0
10 10.0-12.0 12.0 52.0
llR1 . 75.0-150.0 29.0 20.0
'--- --
A3 (Rarw;e) Test Mode - Low Baseline
The low baseline test point is a repeat of the Al FDC #5 Rev. 1 test point ,and
the combustor configurations utilized in terms of primary and dilution port sizes
were identical.
...
Tables VIII and IX show the emission results of the high and low baseline
configurations, respectively, together with the approximate ranges of the combustor
variables within which the emissions are maintained below the contractual goals.
3.5.2
Class B-Mod (JIC-4)
. By arrangement with the Program Officer only the B1 Simulated Federal
Driving Cycle and B2 Steady Speed Modes were tested.
55
-------
TABLE VIII
A3 RANGE MODE RESULTS - HIGH BASELINE
CLASS A-MOD, JP-4
Combustor Parameters
- Available Ranges -
T _oF PI - psig T - of
1 2
1100 50.0 1880
t t t
1100 59.0 1900
~ ~ ~
1100 >64.0 1960
High Bas eline
Parameters
High Baseline
Emissions - .
gm/kg Fuel
NOx CO UHC
1. 17 9.0 -
TABLE IX
A3 RANGE MODE RESULTS - LOW BASELINE
CLASS A-MOD, JP-4
Combustor Parameters
. - Available Ranges -
T -OF PI - psig T _oF
1 2
1100 <5.0 1715
t t t
1380 18.0 1740
~ ~ t
-- >25.0 1920
Low Baseline
Parameters
Low Baseline
Emissions -
gm/kg Fuel
NOx CO UHC
0.3 6.8 0.3
,
B1 (Federal Driving Cycle) Test Mode
The emission levels monitored during this test section are shown in Figures
37 through 40 and are summarized in Table X. As can be seen, the NOx and CO
emission levels meet the 1975-76 requirements, but the hydrocarbon emissions are
slightly in excess. These results are encouraging in that they indicate that this. type
of combustor is capable of meeting the 1975-76 emission levels with only a little more
development.
56
-------
FIGURE 37.
60
50
40
....
w
::>
LL
.; 30
""
':
::<
<.:>
'x 20
o
z
uj 18
::>
LL
. 16
<.:>
~ 14
::<
'7 12
o
u 10
..J
W
::>
LL 2
<.:>
""
':
::<
<.:> 1
I
'"
co
....
B 0
T 1 = 840°F
T 2 = 10400F
PI = 66 PSIG.
"'f = 6 L8.1HR.
"'a = 0.575 LB.lSEC.
wi/.fr
P = ,257
10
o
FLAME OUT
~
06P = 2.0"HG.
DESIGN POINT: (*)
!!'..{r = 1203
AP .
- 6P = 1.8" HG.""'"
600
700
800
*
100
200
300 400 .500
COMBUSTOR 6T - of
CONTRACTUAL GOAL-ll.8
57
600
700
800
28
26
o
600
700
800
24
22
20
8
6
4
2
o
100
200
300 400 500
COMBUSTOR 6T - of
CLASS B-MOD COMBUSTOR, FDC #1 REV. 1 (FUEL: JP-4)
3
CONTRACTUAL GOAL-1.42
100
200
300 400 500
COMBUSTOR 6T - of
-------
24
22 T 1 = 860°F
20 T 2 = 13000F
18 PI = 57 PSIG
W a = .533 LB./SEC
16 wf = 12.01b LB ./HR
..J
W
::> 14 !niT
u. ....L- = 270
'-' 12 P .
:.::
'0
::;; 10 Ol:>P = 4.0" HG.
'-'
, ol:>P = 3.0" HG.
x 8
0 DESIGN POINT: (*)
z
6 m-vT
AP = .1808
4
l:>P = 4.1" HG.
2 CONTRACTUAL GOAL
-.. ....
1.38 I
0
FLAME OUT
*
200 300 900
50
40
..J 0
W
::>
U. 30
.;
:.::
'0
::;;
'-' 20
,
0
u
CONTRACTUAL GOAL
10 11.8
0 200 900
300 700 800
;r...-i 0
u
CONTRACTUAL GOAL
1.42
..J
~ 20
u.
'-'
:.::
"-
i 10
'-'
I
U"\
o:J
900
FIGURE 38.
CLASS B-MOD COMBUSTOR, FDC #2 REV. 2 (FUEL: JP-4)
58
-------
24
22 850°F
Tl =
20 T2 = 13400F
18 PI = 66 PSIG.
W = .575 LB.lSEC
16 a
wf = 15.0 LB.lHR.
...J .14 ofT
UJ wa - = .258.
::>
lJ.. 12 P
.;
~ 110 Ol:.P=4,O"Hg
::;: Ol:.P=5,O"Hg
l:) 8 DESIGN POINT: (*)
, mf} =,173
x
0 6
z l:.P = 4,2" Hg
4 CONTR"CTUAL. GOAL ~~..
2
0
*
200 300 400 500 600 700 800
COMBUSTORt:T - of,
50
40
...J
UJ
~ 30
.;
'"
':
~ 020
,.
0
u
10
0
200 300 400 500 600 700 800
COMBUSTOR-l:.T - °F~
...J 10
UJ
::>
lJ.. 8
.;
'" 6
':
::;:
l:) 4
1/'1
~ 2 CONT.~ACTU".L GOA~.I,42 ..
.....
:I:
U 0
200 300 400 500 600 700 800
COMBUSTOR l:.T - of
FIGURE 39.
CLASS B- MOD COMBUSTOR, FDC #3 REV. 2 (FUEL: JP-4)
59
-------
11
10 T 1 :::: 8600F 0 0
°
9 T 2 :::: 1800 F 0
8 PI :::: 57 PSIG.
W :::: .477 LBjSEC
a
..J 7 wf :::: 25 LB/HR.
I..LI
:::>
Li... 6 'VI ~4. ~
. wa , ~~ ",-,,0
<.::) -:::: 242
~ 5 p. . ~"'-
-------
L--- -
I
TABLE X
BRAYTON CYCLE COMBUSTOR CLASS B-MOD
JP-4, MODE Bl: FEDERAL DRIVING CYCLE
NOx Fuel NOx NOx
Ib/1000 lb. Flow lb/hr Time lb/test point
Test Point Fuel lb/hr x 103 sees x 106 ".Total Cycle = 5,402.28 x 10-61b (NOX>
1 Rev. 1 0.5 6.0 3.0 41.0 34.167
-6
.'. gm/Mile = 5402.28x10 x453.6
2 Rev. 2 0.7 12.0 8.4 1043.0 2433.67 7.5
3 Rev. 2 2.4 15.0 36.0 274.0 2740.00 = 0.327
4 2.0* 25.0 50.0 14.0 194.44 (Program Goal = 0.4 gm/mile)
= 5402.28
* Estimate only.
CO Fuel CO CO
Ib/1000 lb. Flow Ib/hr Time Ib/test point
Test Point Fuel Ib/hr x 103 sees x 106
.'. Total Cycle" 43,054.997 lb (CO) x 10-6
1 Rev. 1 30.0* 6.0 180.0 41. 0 2,050.00
-6
2 Rev. 2 8.0 12.0 96.0 1043.0 27,813.33 . /MU - 43,054.997 x 10 x 453.6
..gm e- 7.5
3 Rev. 2 9.0 15.0 135.0 274.0 10,275.00
= 2.604
4 30.0* 25.0 750.0 14.0 2,916.67 (Program Goal = 3.4 gm/mile)
k f 43,054.997
* Estimate only.
Test Point
CHI. 85
Ib/1000 lb.
Fuel
Fuel
Flow
lb/hr
CHI. 85
Ib/hr
x 103
Time
CHI. 85
Ib/test point
x 106
sees
1 Rev. 1
3.0*
6.0
18.0
41.0
205.00
'-6
".Total Cycle = 8,238.84 x 10' Ib (CHI. 85)
-6
. m/ '1 - 8,238.84 x 10 x 453.6'
. .g Ml e - 7.5
2 Rev. 2
1.8
12.0
21. 6
1043.0
6,258.00
3 Rev. 2
1.3
15.0
19.5
274.0
1,484.17
= 0.498
4
3.0*
25.0
75.0
14.0
291.17
(Program Goal = 0.41 gm/mile)
~ = 8,238.84
*Estimate only.
61
-------
r
I
The results of FDC #4 test point of the Class B-Mod combustor are given in
Figure 40. As can be seen, lean flame-out was occurring at a temperature rise much
higher than the design point. This phenomenon was subsequently demonstrated to be
unconnected with the combustor configuration but a result of fuel system problems.
A retest of this point was not considered at the time, due to the pressure of the
program schedule and the fact that FDC #4 test point represents only a 1 percent
contribution to the integrated emissions.
The results for FDC #4, shown in Table X, are' not intended to be extrapolations
of the test figures but are estimates, considered to be on the pessimistic side.
The final inspection of the Class B-Mod (JIC-4) combustor at the conclusion
of the testing, after an estimated total of 30 hours test time at various conditions,
revealed some minor erosion/corrosion of the dome cooling strips at the inside
diameter of the trailing edges. This was not considered serious and could probably
be avoided by a slight redesign of the air admission hole positions on the dome cooling
s~rips .
B2 (Steady Speed) Test Mode
Because of limitations imposed by time and budget, some of the test points
of the B2 section were tested with common rather than individual port inserts. In
general, the results of these tests follow the saine trends as'those of the simulated
Federal Driving Cycle; a summary of the results is shown in Table XI.
During the testing points 10, 11 and 12 of the Steady Speed Mode (B2) with
common port inserts, a reversed NOx pressure dependency was noticed, complementing
the observations of the Class A-Mod results. Essentially, the indications are that at
these higher pressure levels, the NOx emissions decrease with increasing pressure
ratios rather than increase. This particular trend is shown in Figure 41 for two
different primary zone equivalence ratios. During the tests each of the points 10, 11
and 12 had essentially the same percentage pressure drop and inlet temperature, but
the mass flow through the combustor varied as the inverse of the inlet pressure -
maintaining a constant ratio. Thus since not all tne inlet conditions were held constant,
it may be incorrect to infer a true pressure dependency, although it is likely.
Wall Temperatures and Outlet Temperature Profiles
Maximum wall temperatures were found to be of the same levels as found
with the Class A-Mod combustor, i. e., around 1650° F at any of the test point
conditions.
62
-------
TABLE XI
BRA YTON COMBUSTOR PROGRAM - B1 AND B2 TEST RESULTS
CLASS B-MOD COMBUSTOR
Data Tin Pin 6T 6P% gm/kg Fuel
2
Point of psig of P NOx CO UHC 1
1R1 840 66.0 200 1.10 0.5 30.0 3.0
2R2 860 57.0 440 2.81 0.7 8.0 1.8
3R3 850 66.0 490 2.56 2.4 9.0 1.3
4 860 57.0 940 2.24 2.0 30.0 3.0
5 860 60.0 460 2.5 1.4 5.5 -
6 850 62.0 490 2.44 2.5 2.8 -
7 850 70.0 520 2.77 2.2 50.0* 15.0*
8 830 81. 0 600 2.72 6.2 27.0 25.0
9 820 97.0 700 3.28 2.0 100.0* 10.0*
10 810 112.0 800.0 3.38 2.0 70.0 1.0
11 800 138.0 970.0 3.11 6.5 4.0 0.3
12 800 161. 0 1100 3.17 11.3 2.0 0.32
* Estimates only
1. Expressed as CH1. 85
2. Expressed as N02
Outlet temperature pattern factors for both the JIC-3 and JIC-4 combustors
were generally less than 0.1, with a maximum of 0.15 at high temperature rise levels,
(above design point).
63
-------
Pressure . NOx at NOx at
Test Point at Ms. Abs. T1 jP/P% Up =0. 4 Up=0.5
10 8.60 8100 F 3.38 . 5.0 14.4
11 10.40 8000 F 3.11 3.1 10.4
12 11.96 8000 F 3.17 2.8 9.,95
15
14
13
12
11
-
-------
4
PROGRAM EXTENSION OUTLINE
In April 1972 certain extensions to the original scope of work of Contract
68-04-0016 were agreed to between Solar and the Environmental Protection Ag~ncy,
Advanced Automotive Power Systems Development.
The extensions are conveniently grouped into Addendums I, II and III and are
outlined below.
. Addendum I - Optimized Federal Driving Cycle
This task consisted of optimizing the configurations of the integrated
Federal Driving Cycle test points for the Class A-Mod (JIC-3) combustor
in order to ensure that the integrated emissions were each lower than
the contractual goals.
As noted earlier, the results of the initial Al test mode were encouraging
inasmuch as the NOx level was within requirements but CO and UHC
were much higher than requirements.
. Addendum II - Primary Zone Model and Variable Area Port Tests
The objectives here were twofold: (i) to obtain information on possible
approaches to increasing the operating range of the Class A-Mod (JIC-
3) type of combustor by rig testing various primary zone mopel config-
urations; (ii) to test various cold flow models of variable geometry port
configurations in order to investigate such effects as discharge coefficient
variation, leakage and repeatability that would contribute to the operation
of a successful control system.
. Addendum III - Combustor Control System Analysis and NOy Correlations
The two interrelated tasks were as follows: (i) using data obtained during
the parametric combustor testing, to develop correlation functions for
the NOx production in terms of the combustor operational parameters;
(ii) to perform an analysis of a typical variable JIC type combustor
65
-------
control system in order to assess the complexity of such a system and
to estimate overall error levels of such a control.
- The detailed results of the above addenda are presented in the following
sections.
4.1 ADDENDUM I - OPTIMIZED A1 FEDERAL DRIVING CYCLE RESULTS (JIC-3)
The original results of the A1 test mode on the Class A-Mod combustor
showed the NOx emissions to be well within requirements but the CO and UHC to be
approximately a multiple of 5 in excess of the goal.
It was recognized at the time that the undesirably high CO and UHC were the
result of the contribution to the integrated cycle of only two test points, #2 and #3,
where the combustor had been matched with the test design point too close to the lean
extinction limit. The program schedule at that time prevented any further retesting
of these points.
Consequently, the optimization of the A1 Federal Driving Cycle was concen-
trated exclusively on test points #2 and #3 and consisted of rematching the primary
. zone equivalence ratio to lower the unburned constituents without excessive com- '
promise to the NOx.
Table XII gives the results of the optimization and, as can be seen, the
integrated emissions are all below 50 percent of the contractual goals. The data for
test points #1, #4, #5 Rev. 1 and #6 are as previously reported, with #2 Rev. 5 and
#3 Rev. 3 representing the optimized results for test points #2 and #3.
4.2 ADDENDUM II - PRIMARY ZONE MODEL AND VARIABLE AREA PORT TESTS
4.2.1
Primary Zone Model Tests
The series of primary zone models produced were intended to provide basic
information into the lean extinction and operating bandwidth characteristics of the JIC
type combustor reaction zone.
Essentially each model was constructed from two modular sections: a primary
zone section in the form of a simple, closed-end cylinder; and a primary port holder
ring which was attached to the open end of the primary zone section. To complete the
combustor, a plain dilution section tube was attached; no dilution ports were Incorpora-
ted, however, and thus the models were truly primary zones only. No skin cooling
provisions were made for the models.
66
-------
TABLE XII
CLASS A-MOD COMBUSTOR
OPTIMIZED FEDERAL DRIVING CYCLE
NO NOx NOx
x
LIJ/I000 LB FUEL FLOW LB/H~ TIME LB/TESl PT :. NOx (AS N02) PER TEST CYCLE = 2796.06 x 10-6 LI;
FDC NO. FUEL LB/HR x 10 - SEC S x 10
1 4.5 6.0 27.0 41.0 307.50 :.GM/MILE = 2796.06 x 10-6 _4~\6
2R5 0.6 8.0 4.8 466.0 621.33
3R3 0.8 10.0 8.0 301.0 671.11 = 0.169
4 0.22 11.0 2.42 302.0 203.01
5Rl 0.3 12.0 9.6 247.0 658.67 CONTRACTUAL GOAL = 0.4 GM/MILE
6 4.3 20.0 36.0 14.0 334.44
.[= 2,796.06
CO CO LB/TESl PT
LB/I000 LB FUEL FLOW LB/~ TIME x 10 :.CO PER TEST CYCLE = 14,853.50 x 10-6 LB
FOC NO. FUEL LB/HR x 10 - SEC S
280.17
1 4.1 6.0 24.6 41.0 2,071.11 GM/MILE = 14,853.50xlO-6x 4V~
2R5 2.0 8.0 16.0 466.0 1,677 .78
3R3 2.0 10.0 20.0 302.0 7,382.22 = 0.898 GM/MILE
4 8.0 11.0 88.0 302.0 2,470.00
5Rl 3.0 12.0 36.0 247.0 972.22 CONTRACTUAL GOAL = 3.4 GM/MILE
{, 12.5 20.0 250.0 14.0
.[= 14,853.50
CH1.85 CII1.85 CH1.85
LB/I000 LB FUEL FLOW LB/~ TIME LB/TES6 PT :.CHl.85 PER TEST CYCLE = 1739.31 x 10-6 LB
FOC NO. FUEL LB/HR x 10 - SEC S x 10
1 0.9 6.0 5.4 41.0 61.50 GM/MILE = 1739.31 x 10-6 x 4~~~6
2R5 0.5** 8.0 4.0 466.0 517.78
3R3 0.4 10.0 4.0 302.0 335.56 = 0.105 GM/MILE
4 0.5 11.0 5.5 302.0 461.39
SRI 0.27 12.0 3.24 247.0 222.30 CONTRACTUAL GOAL = 0.41 GM/MILE
6 1.81* * 20.0 36.2 14.0 140.78
**ESTIMATES ONLY
.[= 1,739.31
There were three different diameter primary zone sections manufactured and
also three additional sections with varying lengths at constant diameter. Three primary
port holders at constant diameter were manufactured with various angles. Each port
holder had 16 mounts, hence up to 16 primary ports/fuel injectors could be
accommodated.
The modular sections are depicted schematically in Figure 42. A photograph
of a built-up model with the three sections tack-welded together is given in Figure 43.
The chart shown on Figure 44 depicts all the various configurational arrange-
ments considered together with the various primary zone lengths and diameters. Not
all of the configurations shown have been tested by any means; in fact, much of the
hardware described was not produced.
67
-------
TWO OTHER DIAMETER PRIMARY ZONES mOTH
5 INCHES LONG) ARE AVAILABLE - &INCH
DIAMETER AND 3-INCH DIAMETER
I
r,----------------i
I ~ - - - - - - - - - - - - - - -; I
II . II
II ,I
II I:
I II
PORT MOUNTING
HOLDER (16)
\
%
FIXED GEOMETRY PORT
INSERTS ARE POSITIONED
IN HOLDER (COMBINED
WIT H INJECTOR S)
TORCH
IGNITER
ENTRANCE
- RING HAS FIXED
DIAMETER OF .
5 INCHES
'I ,I
I I
II II
II II
II . I'
1_-_-==-_-- ~ :::_-==- :_-- -=- = =.=.=J
BASELINE PR IMARY ZONE
SECTION 5 INCHES LONG
5-INCH DIAMETER (TWO
OTHER LENGTHS AVAILABLE)
THREE DlFFERENX
ANGLED RINGS ARE -
AVAILABLE (30°,
45°,60°)
FIGURE 42. SCHEMA TIC OF PRIMARY ZONE MODULAR CONSTRUCTION
The chart outlines the reference configuration utilized for the initial test.
This version essentially reflects the configuration of the Class A-Mod (JIC-3) combustor
inasmuch as the length and diameter of the primary zone are the same. The primary
port angle and the number of primary ports are also common to both systems. For the
reference configuration, representative primary port diameter and length- to diameter
ratios were chosen, and a test point combustor pressure drop of 4 percent was chosen
- for all configurations tested.
Trble XIII lists the geometries of the various configurations tested. Also
shown are the lean "blowout" equivalence ratios, and the values of the equivalence
ratio where NO emissions exceeded the contractual goals. The difference between
x
these two values represents the useful operating range from both a blow-out and NOx
emission standpoint. From these data are multiple regression analysis a correlation
between various combustor parameters and lean blow-out was developed. A plot of
predicted value against actual value is shown in Figure 45.
68
-------
;'\,
II
II
'IIi
II
}
~
w
~
FIGURE 43. MODEL PRIMARY ZONE COMBUSTION SYSTEM
WITH ATTACHED TAILPIPE
69
-------
PRIMARY PORT ANGLE
30°
PRIMARY PORT NUMBER
-.::J
o
PRIMARY ZONE
LENGTH
PRIMARY ZONE
DIAMETER
60°
FIGURE 44. BRAYTON CYCLE COMBUSTOR PROGRAM PRIMARY ZONE MODEL TESTING.
CONFIGURA TION SUMMARY
-------
TABLE XIII
PRIMARY ZON E MODEL TESTS SUMMARY
CONFIGURATION Equiv. Equiv. Equiv. wAIR
Tl PI Ratio Ralio
D L Ratio at at '"Ox ° peraling L* L** D* LB.!
a Blow-out
N of. Limit Range INS. INS. INS. SEC. NOTES
TEST DATE DEG. INS. INS. PSIG. ~BO NOx ~
800 20 .304 .308 .004 .509
7-29 45 8 5 5 800 40 .301 .400 .099 3.655 2.75 .683 .803
BL 1000 20 .259 .289 .030 .473
1000 40 .250 .316 .066 .746
800 20 -- -- -- .509 UNSTABLE BEFORE FLAME
Ml 8-2 45 16 5 5 BOO 40 -- -- -- 3.655 2.75 .500 .803 OUT - EXCESSIVE CO LEVELS.
1000 20 .217 .292 .075 .473
1000 40 .208 .263 .055 .746
800 20 -- -- -- .509 ONLY 8 FUEL INJECTORS.
M2 8-3 45 16 5 800 40 -- -- -- 3.655 2.75 .500 .803 *TOTAL LINE OVER NOx LIMIT.
5 1000 20 .238 .246 .008 .473
1000 40 .244 --* -- .746
800 20 -- -- -- .509
M3 8-4 45 4 5 5 800 40 -- -- -- 3.655 3.5 1.000 .803
1000 20 .240 .283 .043 .473
1000 40 .243 .308 .065 .746
800 20 -- -- -- .509 *ESTIMATE ONLY.
M4 8-5 45 4 8 5 800 40 -- -- -- 3.655 3.5 1.000 .803 UNSTABLE BEFORE FLAME
1000 20 .159 .242 .083 .473 OUT - EXCESSIVE CO LEVELS.
1000 40 .160 .282* .122 .746
800 20 .233 .320 .087 .509
M5 8-10 45 8 8 5 800 40 .240 .331 .091 3.655 2.75 .683 .803
1000 20 .187 .270 .083 .473
1000 40 .179 .304 .125 .746
800 20 .258 .270 .012 .509
M6 8-11 45 8 3 5 800 40 .289 .332 .043 3.655 2.75 .683 .803
1000 20 .261 .270 .009 .473
1000 40 .265 .297 .032 .746
800. 20 .276 .330 .054 .509 *TOTAL LINE OVER NOx LIMIT.
8-12 45 4 3 5 800 40 .298 .347 .049 3.655 3.5 1.000 .803 AT T 1 = 1000oF, NOX INCREASES
1>17 1000 20 .250 --* -- .473
1000 40 .246 --* -- .746 TOWARDS FLAME OUT.
800 20 .243 .327 .084 .509
M8 8-14 45 4 5 8 800 40 .240 .315 .075 5.875 3.5 1.00 .803
1000. 20 .198 .210 .012 .473
1000 40 .187 .288 .101 .746
800 20 .286 .300 .014 .509
M9 8-15 45 8 5 8 800 40 .274 .356 .082 5.875 2.75 .683 .803
1000 20 .264 .282 .018 .473
1000 40 -- -- -- .746
800 20 .238 .259 .021 .509
MI0 8-16 30 8 5 8 800 40 .247 .330 .083 4.125 4.125 .683 .803
1000 20 .238 .259 :021 .473
1000 40 .223 .285 .062 .746
800 20 .282 .285 .003 .237 .683 INSERTS RETAINED IN
M11 8-18 30 4 5 8 800 40 .238 .341 .103 4.125 4.125 1.000 .375 ERROR.
1000 20 .233 .255 .022 .221 FUEL SYSTEM PROBLEMS
1000 40 .233 .273 .040 .348 DURING THIS TEST.
800 20 .279 .322 .043 .509 RETEST OF M.11.
M12 8-19 30 4 5 8 800 40 .296 .351 .055 4.125 4.875 1.000 .803 WITHOUT INSERTS.
1000 20 .213 .248 .035 .473
1000 40 .230 .309 .079 .746
800 20 .227 .350* .123 .509 *ESTIMATE ONLY.
M13 8-24 30 4 5 11 800 40 -- -- -- 6.625 4.875 1.000 .803
1000 20 .176 .230 .054 .473
1000 40 .178 .278* .100 .746
800 20 .216 .216 .000 .509 *ESTIMA TE ONL Y.
M14 9-6 30 8 5 11 800 40 .230* .285 .055 6.625 4.125 .683 .803 .EXCESSIVE CO LEVELS.
1000 20 .168 .233 .065 .473
1000 40 .170* .238* .068 .746
800 20 .210 .370* .160 .509 INJECTOR SPACING 4,4,8.
M15 9-8 45 3 8 5 800 40 - -- -- 3.655 3.5 1.000 .803 *ESTIMATE ONLY.
1000 20 .173 .336* .163 .473
1000 40 .166 .370* .204 .746
800 20 .205 .312 .107 .509 REPEAT OF M.4.
M16 9-11 45 4 8 5 800 40 -- -- -- 3.655 3.5 1.000 .803 EXCESSIVE CO LEVELS.
1000 20 .139 .253 .114 .473 *ESTIMATES ONLY.
1000 40 .100 .270* .170 .746 UNSTABLE BEFORE FLAME OUT.
CONFIGURA TlON Equiv, E1Uiv. Eqv;v. "'AIR
Tl PI Ratio Ratio
D L Ratio at at '"0. Operating L* L** D* LB.!
a N Blow-out Limit
TEST DATE DEG. INS. INS. of. PSIG. 9BO Range INS. INS. INS. SEC. NOTES
9,"°x "'P
800 20 .237 .313 .076 .509 INJECTOR SPACING 4,4.8.
M17 9-13 45 3 5 5 800 40 -- -- -- 3.655 3.5 1.000 .803
1000 20 .210 .260 .050 .473
1000 40 .198 .258 .060 .746
800 20 - -- -- .509 INJECTOR SPACING 5,5,6.
M18 9-18 45 3 8 5 800 40 .242 .278 .036 3.655 3.5 1.000 .803
1000 20 .178 .240 .062 .473 UNSTABLE BEFORE FLAME
1000 40 .182 .238 .056 .746 OUT.
800 20 .195 .275 .080 .509
M19 9-21 30 4 8 8 800 40 -- -- -- 4.125 4.875 1.000 .803
1000 20 .190 .270 .080 .473
1000 40 .160 .285 .125 .746
800 20 -- -- -- .509 UNSTABLE BEF"ORE FLAME
M20 9-25 30 8 8 8 800 40 -- -- -- 4.125 4.125 .683 .803 OUT.
1000 20 .145 .253 .108 .473
1000 40 .160 .253 .093 .746
D DIA.
D*
PRIMARY ZONE
DIMENSIONS
NOMENCLATURE:
T 1 - COMBUSTOR INLET TEMP.
PI - COMBUSTOR INLET PRESS.
-------
.31
.30
.29
w.28
=>
..J
~ .27
..J
~ .26
I-
u
< 25
I .
o
~ .24
<
a:::
w .23
u
z
~ .22
<
~ 21
=>.
C1
w 20
w.
z
~ .19
>-
~ .18
::2:
g: .17
.16
cpB.O = [WAiR 8 2 J 0.1.6881 NL~0.024gexp [1.779 - 0.0009643 T ~
~1' D L~ [D* J ~
wAIR - (LB/SEC) AIR MASS FLOW. cpBO - REACTION ZONE EXTINCTION EQUIVALENCE RATIO
PI - (PSIA) COMBUSTOR INLET PRESSURE.
D - (INS) COMBUSTOR REACTION ZONE DIAMETER
T 1 - (oR) COMBUSTOR INLET TEMPERATURE
N - NUMBER OF PRIMARY PORTS
D* - (INS) PRIMARY PORT DIAMETER. L** - (INS) PRIMARY JET FREE LENGTH
L* - (INS) REDUCED REACTION ZONE LENGTH
BRA YTON CYCLE COMBUSTOR PROGRAM
PRIMARY ZONE MODEL TESTS
LEAN EXTINCTION DATA CORRELATION
o
o
o
o
o
o
o
o
o
o
o
o
o
o O~
00
o
o
o
o
o
o
o
.15
.14
.13.14 .15.16:.17 .18 .19 .20 .21 .22 .23 .24 .25 .26 .27 .28 .29
PRIMARY ZONE EQUIVALENCE RATIO - PREDICTED VALUE
.30 .31 .32 1.33
FIGURE 45. LEAN BLOW OUT CORRELATION FOR A JET INDUCED
CIRCULATION (JIC) COMBUSTOR PRIMARY ZONE
73
-------
I
Characteristic emissions of NOx' CO and UHC were plotted against overall
(reaction zone) equivalence ratio at constant values of combustor pressure drop,
combustor inlet pressure and combustor inlet temperature.
The results of the reference combustor are shown in Figure 46, arid display
the now familiary characteristics of the JIC combustor system. Increasing the com-
bustor inlet temperature from 800 to 1000° F results in increas~d NOx levels but
improves the lean stability of the combustor. Increasing the combustor pressure from
20 to 40 psig produces an attendant reduction in NOx with a slight improvement in
lean extinction. '
, The Mod. 19 combustor test results are shown in Figure 47. As can be seen
a considerable increase in operating range over the baseline configuration has been
obtained, together with a concomittant decrease in the lean blow-out.equivalence
Compared with the baseline combustor configuration the operatingr~ge has been
increased approximately threefold, a fact which increases the chances of having a
successful variable area control system.
4.2.2
Variable Geometry Port Investigation
Variable Geometry Port Goals
The goals of the variable geometry port investigation were, in general, to
provide a design such that its effective discharge coefficient was independent of the;
annulus crossflow conditions. In the case of the mechanical variable area ports,' it
was also desired to have a design that produced a perfectly stable exit jet (a jet with
non-varying exit angle) with any slide or plug position. Additionally for the mechanical
variable area primary ports, a secondary requirement of ensuring that the fuel atomi-
zation characteristics were invariant with slide or plug position was considered neces-
sary.
Fluidic variable area ports were required to meet the same general goal as
the mechanical devices, but, in addition, a restraint preventing any swirl in the exit
~as also imposed. It was recognized that fuel atomization could be 'a problem with
fluidic primary ports, and the design had to be such that the fuel spray quality did not
vary with the valve internal flow conditions.
Translation Port
A photograph of this arrangement is given in Figure 48. In this system the
rectangular sectioned primary and dilution ports are formed between stationary and
translatory sections of the ,combustor. A typical angled primary port and radial
dilution port can be seen in the figure.
74
-------
6
FUEL: JP-4
7-29-72
5 8 PORTS x ,703 DIA. T 1 = 10000F
8 INJECTORS
BASELINE COMBUSTOR P - 20 PSIG
COMBUSTOR PRESSURE DROP = 4.01'"r~:] 1 -
4 0
:J
w 111 r-
:::> -1!.VT 1 = 0,521
.... PI
~ 3
'"
"
::;; 20 PSIG
8
I 2 CLASS A M.q,D. CONTRACTUAL
0 ~
2: GOAL -1.75
o
.20 ,22 .24 .26 .28 .30 .32 .34 .36 .38 .40
IDEAL PRIMARY ZONE EQUIVALENCE RATIO
11
10 CLASS A MOD. CONT~_ACTUAL GOAL -14.9
9
8
:J 7
w
:::> 6
....
e,:; 0
'" 5
"
::;;
8 4
I
0 3
u
2
0 .28 .30 .32 .34 .36 .38 .40
,20 .22 .24 .26
IDEAL PRIMARY ZONE EQUIVALENCE RA TlO
8
7
:J 6
w
:::> 5
....
~
'" 4
"
::;; 3
8
'" ~?A MOD. CONTRA_C.TUAL GOAL -1.81
"". 2
.....
:I:
U 0 0
0
,20 .22 .24 .26 .28 .30 .32 .34 .40
IDEAL PRIMARY ZONE EQUIVALENCE RATIO
FIGURE 46.
PRIMARY ZONE MODEL TESTS, JP-4, REFERENCE COMBUSTOR
75
-------
::J
w
::>
W-
e,;
~ 2.5
':Ii
<.:>
x
o
z 2.0
::J
w
::>
W- 3.
e,;
~
':Ii
~
, 2.
o
u
FIGURE 47.
4.0
JP-4
, SEPT. 21 72
MOD. II 19
PRIMARY ZONE MODEL TESTS
4 PORTS/INJECTORS
8"DIA. x 8" PRIMARY
30°
3.5
3.0
CLASS A MOD. CONTRACTUAL GOAL 1. 75
1.5
1.0
T = 1000°F
1
P = 40 PSIG.
1
o
0.5
o
0.1
.15 .20 .25 .30
IDEAL PRIMARY ZONE ,EQUIVALENCE RATIO
.35
.40
5.
~- CONTRACTUAL GOAL-14.9
4.
1.
o
.15 .2 .25 .3
IDEAL PRIMARY ZONE EQUIVALENCE RATIO
.1
.35
.4
PRIMARY ZONE MODEL TEsrS, JP-4, MOD 19
76
-------
TABLE XIV
COMPARISON OF THE REFERENCE PRIMARY ZONE WITH
THE BEST OF THE VARIOUS MODIFICA TIONS
-
Reference Mod. 19
8 Ports @ 45', 5" dia. x 5" Primary 4 Ports @ 30', 8" dia. x 8" Primary
Bandwidth on Bandwidth on
Combustor Inlet Conditions ~BO NOx Average ~ ~NO
Bo x Average (~~)
Tl = 800'F, PI = 20 psig 0.304 0.308 0.004 O. 195 0.275 0.080 oJ: 17.02%
Tl = 800' F , PI = 40 psig 0.301 0.400 0.099 -- -- -- --
Tl = 1000'F, PI = 20 ps ig 0.258 0.289 0.030 O. 190 0.270 0.080 oJ: 17. 40%
Tl = 1000'F. PI = 40 psig 0.250 0.316 0.066 O. 160 0.285 O. 125 oJ: 28.05%
,a~_-
''I' "
~......... - ...... ----- "
,u ~
I I I I
I INCH
FIGURE 48. VARIABLE PORT MODEL
77
-------
The system promotes relative simplicity in manufacture and provides an
aerodynamically "clean" combustor annulus but has a variable leakage rate with
position and a possible problem with fuel injector positioning. The actuation forces
required might also be difficult to predict, depending on the state of the sliding surfaces.
The model was tested on the airflow rig previously described, and the results
are shown in Figure 49. The results apply only to the 45 degree angled primary port.
Slide Valve Port
The second mechanical system is shown in Figure 50. As depicted, the model
consists of combined primary and secondary ports, both rectangularly sectioned,
where the area control is performed by the central slide. The primary slide is
independent of the secondary slide but could be coupled together for certain ranges of
combustor operation. The fuel delivery tube is shown positioned at the primary port
throat. A splitter is provided at the dilution port outlet to divide the jet penetration
between the primary jets. The dilution port outlet is shown angled to direct the flow
downstream to avoid premature quenching. Photographs of the system and the twin
central slide are given in Figures 51 and 52, respectively.
The slide valve port was tested only at three discrete positions: the first'
with the primary side fully open and the dilution closed, the second with the primary
and dilution each half open, and the last with the dilution open and the primary closed.
, A typical characteristic is shown in Figure 53. It can be seen that cross flow over the
port entrance has a major effect on the discharge coefficient. In addition to this, it
was found that the primary and dilution port jet angles varied with the slide position.
With the primary open full, and the dilution side closed, the primary jet angle was
found to be 65 degrees. When the slide was adjusted, however, so that both the
primary and dilution ports were each half open, the dilution jet angle was 58 degrees
while the primary jet angle had increased to 72 degrees. At the other extreme when
the dilution port was fully open and the primary closed to its minimum position, the
primary jet angle was 90 degrees while the dilution was 65 degrees.
Fluidic Valve Port
The third system is fluidic in nature with no internal moving parts. A section
through this design is shown in Figure 54. The static nature of the device is an
obvious attraction, although an external compressor would be required to supply the
control air. The operation is as follows:
The main supply airflow passes into the swirl chamber through the outer
perforated plate and then t~rough the fixed orifice at the swirl chamber outlet into an
78
-------
1.0
0.8
I-
Z
UJ
!:2
L\.. 0.6
L\..
UJ
o
u
UJ
C)
0:
~ 0.4
u
V1
Ci
0.2
10
o VALVE OPENING=0.47"
o VALVE OPENING=O.21"
o
o
00
o
o
~P-PRESSURE DROP ACROSS PORT
q -DYNAMIC HEAD OF CROSS FLOW
100
1000
10,000
~.,~)
FIGURE 49. VARIABLE GEOMETRY PORT TESTS TRANSLATING VALVE
PRIMARY SLIDE
SLIDE GUIDE
DILUTION SLIDE
PR,MARY PORT~
FUEL
DELIVERY TUBE
COMBUSTOR
WALL
FIGURE 50. VARIABLE PORT MODEL - SLIDE VALVE
79
100,000
-------
00
o
""
..
... .
'J- '''' ,;
I!:MEe~ .'
_.sIDt~~BIf<;:.'.
"i~ ~~gJ(ri!(
;GI!i.WAUi'3,
.~;: .~
. ~",,,,.
~. '..)
',"~~":::+'...
.... """"i,
!III
.
FIGURE 51.
CONTROLLED VARIABLE AREA COM-
BINED PRIMARY AND DIL urION PORT
SYSTEM
FIGURE 52. DETAILS OF SLIDE FOR
SLIDE VALVE PORT
-------
u
w
~
'"
en
-'
, 0,20
~
-'
...
'"
'"
~ 0,15
,
;:-
0,35
0,30
0,25
0,10
PRIMARY VALVE FULL OPEN AND DILUTION
FULL CLOSED POSITIONS WITH AND
WITHOUT THE EFFECTS OF CROSS FLOW,
0,05
o
2,0
(6 p)1/2
7,5
3,0
3,5
4,0
0,5
1,0
1,5
FIGURE 53. MASS FLOW CHARACTERIZATION OF THE SLIDE VALVE
VARIABLE AREA PORT
CONTROL FLOW
TANGENT tAL SWIR L
OR IFICES
SUPPLY AIR FLOW
:::ij:::j:~::;~:j: ;::ij:i::
EXPANSION
CHAMBER
ANTI-SWIRL
VANES
FIGURE 54. VARIABLE PORT MODEL - FLUIDIC OPERATION
81
-------
.30
~
o
~ .15
V1
V1
«
2
-I
«
I-
o
I- .10
PLENUM PRESSU~E,
Pp = 30 PSIG
.25
." .
.05
, .
. ,
" .
.01
.02 .', . '.03
CONTROL MASS F'LOW - (LB/SEC',
.04
.05
.' . . . . .
FIGURE 55., FLow CHARACTERISTics QF THE FLUIDIC
, V 4\RIABLE AREA DEVIC'E .
expansion chamberl;>ef9re leaVing the port exit tube. The port exit tube is equipped
with de-swirl vaIl~s to produce essentially axial flow in the exit tube.
. ,
The discharge coefficient of the fixed orifice is varied by imposing a degree
of swirl to the airflow. '
The swirl component is provided by the control airflow, delivered from an
external source, which enters the swirl chamber through tangential orifices.
A generalized characteristic of this device is shown in Figure 55 where total
mass flow through the port' is plotted against cOD;trol mass, flow for va~iou~ levels of
. : '. .': ' . .. . \ .-' ~ . ' .:.' .
82
-------
plenum pressure. The overall pressure drop across the port is the plenum pressure
less the atmospheric pressure.
It can be seen that control begins to occur when the control mass flow is
approximately 10 percent of the total mass flow and the valve characteristic becomes
fairly linear thereafter. Such a characteristic would be amenable to control scheduling
although it is considered that the required control flows are excessive. A redesign of
the port to increase the swirl chamber diameter, for example, would reduce the
required control mass flows.
None of the results from any of the m~odel designs dictate that a particular
idea should be ruled out completely, although some redesign would be required in
each case before a satisfactory system would result.
It should be possible to design a combustion system around either of the
mechanical variable geometry port designs and obtain operation of the port with
minimum discharge coefficient variation, although such a variation could, after
calibration; be built into a control schedule.
Some redesign would be required to the fluidic port system to reduce the
external pumping horsepower requirements.
4.3 ADDENDUM III - COMBUSTOR CONTROL SYSTEM ANALYSIS AND NOx
CORRELATIONS
4.3.1
NOx Correlation Parameters
The formulation of NOx correlation parameters is important from two aspects.
First, such parameters can provide useful information as to the controlling mechanisms
and hence indications as to possible methods of reduction of NOx. Second, in the
absence of a suitable NOx sensor, the control system of a variable geometry combustor
must synthesize the NOx level from the other more readily obtainable parameters such
as pressure, temperature and fuel flow in order to maintain the correct combustor port
area relationships. Thus the control must have the correlation parameter 'built inn.
The Addendum lIT work included all the test points of the Class A-Mod and.
Class B-Mod combustors.
The emissions data obtained on the Class A-Mod and Class B-Mod combustors,
when tested over the various specified ranges, have been divided into three categories.
The first are those results obtained with the Class A-Mod combustor equipped with a
co-axial port fuel injector. Comprising the second set of results are those obtained
with the Class A-Mod combustor fitted with a slotted face injector, with the slot
83
-------
oriented toward the primary zone dome. The final set of results were those of the
Class B-Mod combustor, which exclusively utilized the coaxial fuel injection system.
The analysis required that each NOx characteristic be reduced to discrete
operating points described by NOx level and overall equivalence ratio. Approximately
four points were selected from each characteristic. On the Class A -Mod results the
points were selected below a maximum NOx level of 3.0 gm/Kg fuel and below a .
mmmum Of 5.0 gm/Kg fuel for the Class B-Mod combustor.
Each of the data sets corresponding to the above categories was in turn.
analyzed to provide a NOx correlation parameter, using a linear multiple regression
computer program. The results for the two fuel injector configurations of the Class
A-Mod combustor are shown in Table XV together with a Class B-Mod result. An
extremely large Variety of correlation parameters can be obtained by simply rearrang-
ing the various independent variables, although it is likely that none of these would
provide any clues to the controlling physical processes involved in NOx production. In
an effort to provide a correlation parameter that would have some physical significance,
the results were constrained to follow, instead of a straight line, a curve of thEb form
NOx = exp(1/I) - 1, where 1/1 is a parameter composed of the independent variabl~s;and a
constant. Such a curve typically represents the behavior of a generalized kinetically
controlled chemical process, and it was hoped that the independent variables might be
grouped into a form that had some chemical or physical significance. Unfortunately.,
as can be seen, the results of this latter approach did not produce any correlation
functions that could be easily interpreted to provide insight into what might control t~e
production of NOx.
A further correlation was produced that included the ratio of primary to overall
equivalence ratio (see Table XV). This latter term is a function of the primary zone
port area to the total area ratio, and thus can be utilized as part of a control function
providing a positional reference for the proposed variable area actuating mechani~m.
Presently the approach to the problem of providing a control function for the actuating
mechanism is to maintain the combustor pressure drop fraction (~P Ip) constant, which
provides an easy solution to the total combustor area as a function of the combustor
. inlet conditions, expressed as (w.[T/p). Then, from the NOx correlation described
. above, the ratio of the primary zone area (plus a constant primary zone cooling area
portion) to the total combustor flow area can be obtained, and this, combined with the
total value obtained from the pressure drop relationship, provides the actual primary
zone area required. This latter value can then in turn be translated into a spatial
coordinate reference for the actuator mechanism. This part of the present investigation
will be discussed in more detail in the next section.
4.3.2
Control System Analysis
An examination has been made of the feasibility of successfully operating a
variable geometry control system.
84
-------
TABLE XV
SUMMARY OF NOx CORRELATION PARAMETERS
LINEAR MULTIPLE REGRESSION ANALYSIS
1. Results of the Class A-Mod Comwstor (Side Slotted Fuel Injectors) Provide:
NOx =
.10.54 2.96 4.79
W £\T D exp(34.7 + 6.67 (1 + 0.00487 T )
P P 1
P 15. 1 (~p )1. 43
1 PI
2. Results of the Class A-Mod Combustor (Coaxial Fuel Injectors) Provide:
NO
x
,;}.57 £\T2.69 Dp 1. 48 exp(4.76 + 3.98 (1 P + ~~~"~38~T 1)
P 16. 32 (~P )2. 87
1
3. Results of the Class B-Mod Combustor (Coaxial Fuel Injectors) Provide:
6.767
D exp(15.6 + 12.8 (1 - 0.0000123 T )
P P 1
P 17. 158 (~P) O. 336
I
CORRELATION WHEN RESULTS ARE CONSTRAINED TO FIT A CURVE OF
NOx = exp(I/J) - 1 WHERE I/J IS THE CORRELATION PARAMETER
cj6.56 £\T2. 59
NO
x
1. Results of the Class A-Mod Combustor (Side Slotted Fuel Injectors) Provide:
0/1=
'" 4.14 A 2.84
(WI P ) uT exp(0.6 + 0.99 (1 - 0.0032 T )
1 P 1
Pl1.26 (l1P/PI)2.6~.6
CORRELATION USING LINEAR MULTIPLE REGRESSION BUT INCORPORATING
THE VARIABLE AREA CONTROL FUNCTION «(1T/(1p)
1. Results of the Class A-Mod Combustor (Side Slotted Fuel Injectors) Provide:
.10.08 A 4.32 / 3.9
W uT «(1T u) exp(18.37+0.00537T)
NO= p "1
x P 11.86 D 9.33 (£\p/p 5.63
1 P )
NOMENCLA TURE
Dp
PI
(AP/P)
Tl
hT
~
Up
UT
primary port diameter (incb)
combustor inlet pressure" (psia)
combustor pressure drop (%)
combustor inlet temperature rR)
combustor temperature rise r F)
combustor air mass flow (lb/sec)
primary zone equivalence ratio
overall comwstor equivalence ratio
NOTE: NOx is the emissions index in gm/Kg of fuel.
85
-------
(0
T3 -1
Ln
(+( T
e 0 1 3
N T l 1- T / T ( N><
>< 3 4 3 4 r:::. ( )2
D
_T. 1 (
T5 - 4 ~()3NX
£.
( ) 7
L
£.
p /p C f5
3 5 NX NX
N N
A (, ( A + K ) (
P5 DX ( P 8 P 9
( )4
N ~ ~
~
A + A DEMAND
P3 C P
+
+ NX
P4 G (5) A DEMAND
D
REF
A CONTROLLED TO OBTAIN ( P - P ) / P CONST,
D 3 4 4
( £. T
,543 ,457 0+ 1 3 (( (( £.
A CONTROLLED SO A (A + K) ::::: e ( 1- T / T ) 2 (T ) 3 (P ) 4f 5A 6
P PP 3 4 4 3 T
P Y-1 P (
T =(-1...)'YiI T ~(2-) 7
4 P P 5 P
4 5
(0 ADJUST
(1 .00062
(2 .502
(3 -.084
(4 -,207
(5 1,175
(6 ,457
(7 .284
( .543
8
( .457
9
FIG URE 56.. CONTROL SYSTEM BLOCK DIAGRAM
86
-------
The variable geometry system is constrained to operate within the boundaries
of the NOx correlation parameter shown in Table XV and containing the equivalence
ratio function rIT/ rIp'
A study was made including the tentative selection of parameters and an error
analysis for full speed and idle conditions assuming typical accuracies for parameter
sensing. Both analog and digital computation mechanization were considered.
The block diagram of Figure 56 illustrates the mechanization. The results
of the two digital analyses are shown in Figures 57 and 58.
Total error in predicted NOx level is not significantly affected by the difference
in operating speed, but is potentially reduced by 50 percent using digital, rather than
analog computing techniques.
R U ~~
Lm!CX
[nnnR MJI\L YS IS. PRH1I',F!Y !\fIEI', cmlTflnL
F lJ L L [, PEE [)
OICH I',L cm~PUT.J.I,TIm:s ~~ECIIMJI7. 1/\ TIm!
, EF1f\nn SOllRCE ~!Or~H!/\L TOLEfH\~!CE (%)
T3 15~5 1
P 3 (,t) . U
T5 1H6~1 ?
P5 36 1.5
F 1 .3
A(TOTI\L) H 1.2
C m:pL.: n: F.XP . ~HH] (;2 . 1
cor~PUTl:: EXP . 5~1:2 .1
C m~p UTE F. X P - . 11 n {I . 1
COr~PlJTE EXP -.2\14 .1
cnr~PUTE F.XP 1 .17~ . . 1
COl~PlJTF. EXP .£157 .1
cor~PIITE EXP .29 H1
CO~PlJTE F.XP .543 .1
CCHPIJTE EXP .457 .1
~lllL TIPL1ER 1 .1
WE!GIIT
.936509
.193561
.1~J52~4
3.r.5093E-~2
1.1~H193
.428193
.UU59r~1
.4UU694
.611257
. [HH1!:\ 1 5
1.5404:2E-07
.U904kJ3
1.84927E-02
.CJ1U037
6.3 5523E-02
4.71972
TOTAL ERROR (%)=
ROO T 6 lJ~!' S [.J lJ ARE E H R C R ( % ) =
I\ssur~H!G 0.5 % lJ~JCERTI\H~TY HJ THE CCRf\ELI\TIO~!
FlHJCT ION. Mm 2 o,{. ACCUf\I\CY OF DEL T A-P /p
CONTROL. THE EXPECTEO RSS EHRon IN MOX LEVEL IS
17.42~7 %
OmJE
EmWR (%)
.9365E<)
.154U4L
. 2 H~ 4 C 9
(1.57639E-0:2
.330:20
.513032
B.0591iJ1E-kJ2
4.0U6,)4E-<12
6.11:257F.-02
O.0IiJ015E-02
1.54U42E-OS
fJ .9 f14 C 3 E - ~ 2
. 1fJ 4 92 7
0.10837E-C2
6.35523E-03
.471971
----------
3 . 3 ~14 5 7
1.2(;991
FIGURE 57. ERROR ANALYSIS - FULL SPEED DIGITAL
87
-------
RUN
LONOX
ERROR ANALYSIS, PRI~ARY AREA CONTROL
IDLE srEED
,rHG IT AL COMPL)T ATIO~!S r.'ECI-IMJIZ IA TION
ERRon SOUHCE ~!orHNAL TOLERANCE (%)
T3 1450 1
P3 23 2.4
T~ 136~ 2
P5 17 :3
F ~5 1
A(TOTAL) 5 1.2 '
CO~PUTE EXP .0~062 .1
C m~ P U H: EX P , . 5 0 2 . 1
cor,~PUTE EXP -. \1I:J4 .1
CDrH'I:,lT l-:X;J ~.:-!,)~f .1
co~~rUTE EXP 1.175 .1
C[)~~PLlTE EXP .457 .1
cm~PUTE EXP .29 1 V1
Cm'PlJTE [XP .5trJ .1
CO~PUTE EXP .457 ~1
r'l1L T IPLIEr-1 1 .1
\'IE IGH T
.U5lJ5U2
.1C1537
.19f'1533
5.5254.7E-~2
.996~k18
.3[J73[J4
.762[152
1. 12049
.523523
.5422V12
. 6903 U 1
.62347
1.67V124E-02
1 .4 V1 [1EJ 5 F. -37
.157071
tl . 2 [} \1 1 4
. TOTAL ERROR (%)-
RonT P,LJ~1 Sr:JLJARE EfHWR (%)-
AS6IJr~HJG n. 5 ~J Ilnr:ERT/\HITY H' THE COnRFU\TIml
Flmr:TIor', N!D;:' % I\CCLJF1ACY OF DELTA-pip
cm:TROL, THE EXPECTEO FISH F.f1F1OR H! ~IOX LEVEL IS
1C.2l:56 % '
FIGURE 58. ERROR ANAL YSIS - IDLE- SPEED DIGITAL
88
ERROR(%)
.flS[J502
.43569
.J81~67
.165764
.996\130
'. tl6486
7.62052F.-i>,2
.112849
5.23523E-~2
5.422(7,2E-'\12
6.911381 E-02
.062347
.167024
1 . 4 ~Hj USE - 21 8
1.57iJ71E-?2
.42Ur.14
----------
tf .33'/73
1.5975
-------
5
CONCLUSIONS
1. The JIC combustor concept has demonstrated the ability to easily meet
the 1975-76 Federal automotive emissions requirements when rig tested over a
simulated Federal Driving Cycle test mode and using a pseudo-variable geometry
arrangement.
2. For a practical automotive gas turbine power plant the JIC combustor
would require a variable geometry system to maintain the emissions below require-
ments over the complete operational envelope.
3. Sufficient design data are now available to be able to make an initial trial
scaling of the JIC-type combustor to any particular regenerative engine cycle con-,
, ditions with some degree of confidence within the size ranges tested.
4. The available operating range of the JIC-3 combustor could be considerably
extended with configurational modifications in the area of combustor diameter, length
and number of primary ports.
5. The general performance parameters of the JIC-3 combustor such as outlet
temperature pattern factor, smoke and wall temperatures, etc. have to date been
acceptable and would likely not be a problem with any other JIC-type combustor design.
6. The control of a variable-geometry JIC type combustor is likely to be a
difficult problem although not outside current capabilities. Improvements in the low
emission operating range could alleviate the problem considerably.
7. Various variable geometry port designs are available for consideration in
a JIC-type combustor system, and both mechanical and fluidic devices appear attractive
at this time.
89
-------
....
6
RECOMMENDATIONS
1. Because of the very low emission levels at the normal combustor operating
conditions, changes to the level or degree of homogeneity of the incoming fuel-air
mixture are very important. In consequence, further work investigating improved
atomization and mixing arrangements should prove valuable in improving lean blow-out.
It is recommended that such investigations would include the testing of the following
atomization systems:
. Pressure swirl atomization simplex and/or duplex
. Splash plate impingement atomization
.Contra-flow injection
2. One of the conclusions that can be drawn from the work performed to date
is that the number of primary ports can be reduced to four without any major effect
on the emissions. Such a reduction is eminently desirable from a control point of
view since it minimizes the number of variable area devices required. However,
improvements in stability through increases in primary zone recirculation rate could
still be achieved by changes in the port shape. It is recommended that further inves-
tigations into stability improvements include primary ports with cross-sectional shapes
chosen to maximize surface area for a given cross section. An investigation into the
possibility of improving the stability through changes in port or jet shape could lead to
important changes on future combustor designs.
3. The results obtained on the fluidic valve indicate that with the present
configuration the control mass flow rates required for modulation are too expensive
to be practical in terms of the external pumping horsepower required. However, in
view of the reduction in mechanical complexity of a variable geometry combustor sys-
tem that is promised by a fluidic device of this nature, it would seem profitable to
continue these investigations. The control flow to supply flow ratio could be reduced
by increasing the diameter of the swirl chamber, for example. It should also be
possible to have the swirl chamber wrapped around the combustor body with a mani-
fold annular exhaust to each port. Design data on such arrangements are not presently
available, and it would seem worthwhile to pursue this approach as a separate future
investigation.
91
-------
4. The primary zone model tests were directed towards broadening the
available operating range of the JIC type combustor by improving the lean stability
with configurational modifications. Some very crude attempts at piloting the reaction
were performed during the A2 testing of the Class A-Mod JIC combustor with encourag-
ing results. It would seem worthwhile, therefore, to .consider a similar effort to the
primary zone model tests in investigating the operation of a piloted combustor concept
to be operated in conjunction w:ith the torch ignitor. Such a system could obviate the
requirement for a variable geometry combustor.
5. It can be seen that the solution of the variable geometry control problem
in any future JIC type combustor system will not be simple, by virtue of the fact that
the operational NOx level of the combustor at any operating point must be synthesized
from a variety of independently monitored parameters. The complexity of the control
system could be gr~atly 'simplified if a "NOx sensor" were available. This sensor would
continuously monitor the NOx level from the combustor and provide the error signal
for the control on a conventional closed loop system. Although such a device is
apparently not currently available, it promises such control advantages that its
development should .be actively pursued.
92
-------
CHI.85
CO
C02
Dp
NO
N02
NOx
PI
PIN
PINLET
Tl
TIN
TINLET
T2
. ~
wa
Wa ,
wf }
Wf
CJ.
~P
~p/p
DoP / q
DoT
NOMENCLA TURE
Arbitrary hydrocarbon composition taken as representative
~JP~ .
Carbon monoxide
Carbon dioxide
Diameter of the primary zone
Nitric oxide
Nitrogen dioxide
Oxides of nitrogen
Combustor inlet pressure
Combustor inlet temperature
Combustor outlet temperature
Combustor mass air flow rate
Combustor fuel flow rate
Port angle
Pressure drop
Pressure drop as a percentage of the inlet
Ratio of combustor pressure drop to local cooling annulus
dynamic head
Temperature rise or increase over combustor
-------
NOMENCLATURE (Cont)
"qJ
Equivalence ratio (actual fuel to air ratio, divided by the
stoichiometric fuel to air ratio)
U prim
Up
primary zone equivalence ratio
UT
overall combustor equivalence ratio
------- |