No. FAA-RD-71-22
STODY OF VISIBLE EXHAUST SMOKE
FROH9 AIRCRAFT JET ENGINES
John Stockham and Howard Betz
III Research Institute
10 West 35th Street, Chicago, Illinois 60616
I ONE 1971
Availability is unlimited. Document may be released to the
National Technical Information Service, Springfield, Virginia
22151, for sale to the public.
Prepared for
OF TRANSPORTATION
FEDERAL HIATIB3 ADMINISTRATION
Systems Research & Development Service
D. C., 20580
-------
The contents of this report reflect the views of the
contractor, which is responsible for the facts and
the accuracy of the data presented herein, and
do not necessarily reflect the official views or
policy of the FAA or Department of Transportation.
This report does not constitute a standard,
specification, or regulation.
-------
TECHNICAL REPORT STANDARD TITLE
1. Report No. 2. Government Accession No.
FAA-RD-71-22
4. Till* and Subtitle
Study of Visible Exhaust Smoke From
Aircraft Jet Engines
7. Author'.)
John Stockham and Howard Betz
9. Performing Orgonizotion Home and Address
IIT Research Institute
10 West 35th Street
Chicago, Illinois 60616
12. Sponsoring Agency Name and Address
Federal Aviation Administration
Systems Research & Development Service
Washington, B.C. 20590
& National Air Pollution Control Administratio;
3. Recipient's Catalog No.
5. Report Date
June 1971
6. Performing Organization Code
8. Performing Organization Report No.
FAA-NA-71-2U
10. Work Unit No.
11. Conlroct or Grrmt No.(502- 306- ' . !
DOT-FA69WA-2208
13. Type o( "Report and Period Cohered
Final Report
14. Sponsoring Agency Code
1
15. Supplementary Notes
NONE
16.
t xhe objective of this study was to relate the visibility of
inflight jet exhaust to the SAE smoke number. A method based on photo-
graphic photometry was developed for measuring the optical density of
smoke plumes. This method was related to visibility and to the smoke
number through transmissometer measurements and visibility theory. A
portable transmissometer, capable of operating over a wide range of
optical path lengths and under varying ambient light conditions was
fabricated for use on this study. The mathematical expression relatir.~
the transmission measurements to the smoke number was derived. Liminal
visibility requirements of smoke trails, developed from light scat-
tering theory, correlated with actual visual observations and the trar.s-
raissometer and photometry measurements. Test results, with the engines
investigated, indicate that SAE smoke numbers below 23 were associated
with invisible exhaust plumes. Samples of the exhaust smoke showed the
particles to be composed of lacy agglomerates. At the nozzle, the
geometric median particle diameter was 0.052 \j.m. At a distance of
10 nozzle diameters the geometric median particle diameter was 0.13 utr.
at cruise condition.
17. Key Words
Pollution
Smoke
Jet Aircraft
18. Distribution Statement
Availability is unlimited. Document
may be released to the National
Technical Information Service,
Springfield, Virginia 22151, for
sale to the public.
19. Security Clo»»if. (of this report)
I Unclassified
20. Security Clossif. (of this poge)
Unclassified
21. No. of Poges
75
22. Price
Form DOT f 1700.7 (8-69)
-------
PREFACE
This report was prepared by the IIT Research Institute for the
Federal Aviation Administration and the National Air Pollution
Control Administration. The work effort was o part of a progrem
of the Aircraft Division, Systems Research and Development
Service, Federal Aviation Administration, and the Division of
Motor Vehicle Research and Development, National Air Pollution
Control Administration.
The work was administered under the direction of Mr. G. R. Slushei
who served as project manager. Provision of the facilities,
conduct of the tests end collection and reduction of the dc'ta for
the SAE smoke numbers were furnished by the Propulsion Section,
Aircraft Branch, Test and Evaluation Division, National Aviation
Facilities Experimental Center, Atlantic City, New Jersey.
111
-------
TABLF OF CONTENTS
Page
INTRODUCTION 1
(1.) Purpose 1
( 2.) Background 1
DISCUSSION 3
(3.) Modeling Studies 3
(3.1) Light-Scattering Theory 3
(3.2) Visibility Requirements 3
(3.3) Theory Relating the Smoke Number
and Optical Transmission of Jet
Exhaust Flumes 9
(3.3.1) Reflectivity of Soot
Deposition Filter Paper 11
(3.3.2) Optical Transmission of
Smoke Plumes 12
(3.3.3) Relating DR to DT and
Operating Parameters 13
(3.3.U) Relating Smoke Number to
Filter Density IS
(U.) Static Tests of Engines and Aircraft 16
(U.I) J-57-P-37A Engine Mounted in the
Test Cell 16
(U.2) Light Transmission of J-57 Engine
Exhaust for Various Path Lengths 17
(U.3) JT-12-6 Engine Mounted in Wind Tunnel 20
(U.U) Tied-Dovm Aircraft Tests 20
(U.U.I) F-100 Single-Engine
Aircraft 20
(U.U.2) Convair 880 Aircraft 2U
(U.U.3) Multiple Engines of Convair
880 Aircraft 2U
(U.U.U) Lockheed Jetstar Aircraft 30
(U.5) Discussion and Summary of Results
of Static Engine and Aircraft Tests 30
(5.) In-Flight Observations 32
(5.1) Flights Evaluated at NAFEC 32
(5.1.1) F-100 and F-105 Aircraft
In-Flight Tests 32
(5.1.2) Convair 880 Aircraft
In-Flight Tests 39
(5.1.3) Lockheed Jetstar Aircraft
In-Flight Tests 39
-------
TABLE OF CONTENTS (Continued)
Page
(5.2) Observations of Aircraft In-Flight
at Chicago O'Hare Airport 39
(5.3) Discussion and Summary of In-Flight
Tests 39
CONCLUSIONS US
REFERENCES U6
APPENDIX A Transmissometer (4 pages) 1-1
APPENDIX B Photographic Instrumentation (7 pages) 2-1
APPENDIX C Sampling Jet Exhaust for Particular
Matter (U pages) 3-1
APPENDIX D SAE Standard Method for Smoke
Measurements (3 pages). U-l
VI
-------
LIST OF ILLUSTRATIONS
Figure . Page
1 Visibility of Line Targets 8
2 Electron Photomicrographs of Jet Engine
Exhaust Particles 10
3 Smoke Instrumentation 23
4 Microdensitometer Trace of Number 1 Engine
of Convair 880 Aircraft 26
5 Number 1 Engine of the Convair 880 Aircraft 27
6 Microdensitometer Traces from Tied-Down
Convair 880 Aircraft 28
7 Plot of Smoke Number vs Optical
Transmission 33
8 NAFEC/Atlantic City Airport 35
9 Photographs of F-100 and Convair 880 Aircraft
In Flight 36
10 Photographs of Jet Exhaust Trails During
Takeoff at O'Hare Airport U3
1.1 Transmissometer 1-2
1.2 Optical Schematic 1-3
2.1 Typical H & D Curve 2-3
2.2 Illustration of Use of Photographic
Photometry 2-6
3.1 Exhaust Particulate Sampling Apparatus 3-2
3.2 IITRI Smoke Sampler in the Engine Test Cell 3-3
U.I F-100 Aircraft with Sampling Probes U-2
U.2 General Electric Smoke Sampling Console U-3
Vll
-------
LIST OF TABLES
Table Page
1 Relationship Between Ringelmann Number,
Light Transmission, and Particle Concentra-
tion of Turbojet Engine Exhaust 5
2 Particle Size Distribution of Exhaust from
J-57 Engine 7
3 Smoke Numbers and Transmission Data of Jet
Smoke - J-57 Engine 18
4 Light Transmission as a Function of Path Length 19
5 Smoke Numbers and Transmission Data of Jet
Smoke-JT-12-6 Engine Mounted in Wind Tunnel 21
6 Smoke Quality from a Tied-Down F--100
Aircraft 22
7 Smoke Quality from the Number 1 Engine of a
Tied-Down CV-880 Aircraft 25
8 CV-880 Aircraft Tied-Down Multiple Engine
Observation 29
9 Smoke Numbers from the Number 4 Engine of a
Tied-Down Jetstar Aircraft 31
10 Summary of Smoke Number and Transmission
for Static Tests 34
11 Observations on the F-100 Aircraft In-Flight
at NAFEC 37
. 12 Observations on the F-105 Aircraft In-Flight
, at NAFEC 38
13 Observations on the Convair 880 Aircraft
In-Flight at NAFEC 40
14 Photographic Measurement of Jet Smoke Trails
from Aircraft In-Flight at O'Hare Airport and
Visual Estimates of Transmission 41
3.1 Size Distribution and Concentration of Exhaust
Particles - Royco Particle Counter-J-57-P-37A 3-5
3.2 Size Distribution and Concentration of Exhaust
Particles - Royco Particle Counter JT12-6 3-6
viii
-------
INTRODUCTION
(1.) Purpose
The purpose of this research program was to develop
techniques for the objective determination of visibility of
jet exhaust smoke and no relate these techniques 1:0 the smoke
number obtained for engines mounted in test cells. l^se of the
established relationships should enable the prediction of in-
flight visibility of smoke trails from test cell data. A
theoretical model of smoke trail visibility using the physical
and optical properties of the smoke particles was included in
the research effort.
(2.) Background
Smoke trails from aircraft turbine engines are a
symbol of air pollution. Elimination of these smoke trails
would erase a source of pollution ccmp.iairr.s and improve visi-
bility in and around jetports. Engine smoke is currently
determined by a filter stain technique described by The
Society of Automotive Engineers (Reference 1). The results are
expressed as a numerical value known as the SAE smoke number.
This technique is applied to engine test cell operations. A
need exists for an objective measure of the visibility of smoke
trails of aircraft in-flight, especially at take-off and
approach power conditions, and to relate this measurement to
the SAE smoke number. The only currently accepted method for
specifying the visual quality of smoke, trails from aircraft
in-flight is the Ringelmann method (Reference 2). This is a
subjective measurement. Results are highly dependent: upon the
conditions under which the measurements are performed ^Refer-
ence 3).
Two techniques were developed to accomplish the
objectives of this program. The first involved a transmisso-
meter to measure the light a r.r. er.ua tier ci r.he ?.mckef and the
second was photographic photometry. A special two-path trans-
missometer was developed for measuring the optical attenuation
of smoke from test cell engines and engines of tied-down air-
craft. The transmissometer is described in Appendix A.
Photographic photometry was developed f~-r eva.";u.-ji:ing the visi-
bility of smoke from aircraft in-flight and f'ro-n tied-down
aircraft. This procedure is described in Appendix 3.
-------
The following measurements were made during this
program:
(1) Photographic photometry measurements and Ringelmann
observations of in-flight jet aircraft exhaust
were madp at: Chicago's CTHare Airport and at the
National Aviation Facilities Experimental Center
(NAFEC), Atlantic City. New Jersey. Aircraft
studied included a variety of commercial jets and
the F-100, F-105, CV-880, and the Lockheed Jetstar.
(2) Photographic photometry and transmissometer mea-
surements, Ringelmann observations, and the filter
stain technique were a.1.1 used to evaluate smoke
emissions from tied-down aircraft at NAFEC.
Aircraft used were the F-10CL CV-880,. and the
Jetstar. NAFEC personnel obtained the SAE smoke
numbers.
(3) Transmissometer measurements and SAE smoke numbers
were made on test; engines. One: a J-57 engine.
was mounted on a static, sea-level, open-air test
stand; the other, a JT-12 engine was mounted as
a flight installation in a test wind tunnel.
Samples of exhaust smoke were obtained during
these tests using a specia?uly constructed sampling
rig described in Appendix C.
Quantitative data from the instrumental techniques were
correlated with the subjective Ringelmann observations and visi-
bility theory. The correlation established the relationship
between the smoke number, the optical attenuation,, photographic
photometry, and the visual appearence of jet smoke.
-------
DISCUSSION
(3.) Modeling Studies
(3.1) Light-Scattering Theory
The visibility of a smoke plume is a function of
the optical and physical properties of the particles comprising
the plume, the width of the plume, and the contrast between the
plume and its background.
Where a smoke particle is small in relation to the
wavelength of incident light, the total attenuation of light, i.e.,
extinction, E, is given by tne expression:
E = 24^ A
999 999
(n + n k) + 4(rr - n k^ -1)
where,
2ur
tt =1T
X = wavelength of light, urn
r = particle radius, pm
n, k = functions of the refractive index, r\, of the particle.
The refractive index, n.. of a particle is a complex number when
the particle both absorbs and scatters the incident light. In
these instances, the refractive index is equal to n(l - ik),
where n is the coefficient that defines the scattered light
and nk the coefficient that defines the absorbed light. The
term i is the operator V-l. The complex refractive index of
carbon, the major particulate component; of turbojet engine
exhaust (Reference 4), for incident light of wavelength 0.490 um
is 1.59 (1-1.051). Thus, n = 1.59 and k = 1.05.
The extinction of light is due to two separate
and distinct phenomena; light scattering, S, and light absorp-
tion, A. The phenomena are defined as follows:
/ : 9 2
Q 4 . 2. , .
s = 8 a n.._^_A
3 r]2 + 2 - (2)
A = E - S
-------
Substituting the best available information into
Equations (1) and (2), the following values for E, S, and A
result:
E = 0.9
S = 0.1
A = 0.8
Thus, the extinction of light by turbojet engine smoke
is due to 117o scattering of light and 89% absorption of light.
The attenuation of light by a smoke cloud, where the
cloud is so diffuse that complications due to particle inter-
actions are not encountered, is given by the expression:
where,
I = the intensity of the incident light
I = the intensity of the transmitted light
C = number concentration of smoke particles
E = extinction coefficient
r = particle radius
L = depth of cloud intervening between I and I.
The ratio I/I is the transmission, T, of light through the
smoke . °
Thus ,
2.303 log T = -nr2CLE (4)
Equation (4) illustrates the dependence of light
t.ransmittance through a smoke plume on the length of the viewing
path. Thus, the position of the observer in relation to the
smoke plume is important to plume visibility.
If the path length, the particle size, and the extinc-
tion coefficient are known, it is possible to calculate the
number of particles per unit volume of gas that will produce
various levels of light attenuation. The data in Table 1 is an
example. The path. length chosen is the diameter of the J-57
-------
TABLE 1
RELATIONSHIP BETWEEN RINGELMANN NUMBER,
LIGHT TRANSMISSION, AND PARTICLE
CONCENTRATION OF TURBOJET ENGINE EXHAUST
Percent
Transmission
98
95
90
80
60
40
20
Ringelmann
Number
_ _
1/4
1/2
1
2
3
4
Particle Concentration
at Engine Nozzle
Millions of Particles/cm3 (1>
17
50 1
100 i
t
(1) Particle concentration data is calculated from equation
4 using a path length of 56 cm and a particle size radius of
0.026 urn. The path length is the nozzle diameter of the J-57
engine and the particle size is the geometric median particle
radius of the exhaust particles at the nozzle under cruise
conditions, Table 2.
-------
turbojet engine exhaust nozzle, 56 cm. The geometric median
particle diameter of exhaust smoke at the nozzle of the J-57
engine is 0.052 wm, Table 2. Thegdata in Table 1 show that
particle concentrations of 1 x 10 particles/cm^ corresoond to
transmission values of 90% or an equivalent Ringelmann number
of 1/2.
(3.2) Visibility Requirements
The visibility requirements for jet smoke trails is
similar to those of line targets such as wires, poles, and
antennas. Two basic factors are involved in line target visi-
bility. The target must be wide enough to be resolved by the
unaided eye and must have sufficient contrast from the back-
ground to be distinguished. Tests by Douglas (Reference 5) show
flagpoles and radio towers can be resolved if the angle sub-
tended at the eye of the viewer is greater than 3 or 4 sec . of
arc. Assuming a nondif fusing smoke track and no attenuation by
the intervening atmosphere, a 2 f t . smoke track could be seen
at a distance of 20 miles. Obviously, the angle subtended is
not a factor limiting the visibility of jet smoke trails.
The relationship between angle subtended and contrast is shown
in Figure 1. These, data (Reference 6) were developed in the
IITRI laboratory under ideal conditions; a black line target
(contrast = -1) subtending only 0.6 sec. of arc was visible.
Contrast, C , between a smoke plume of luminance, B ,
viewed against an extended background of luminance B, , is given
by the expression:
n
r = P
- B,
From this equation, a totally light absorbing line
target has a contrast of -1 and a black smoke plume will have
a contrast ranging from 0 to -1 depending on the amount of
background light transmitted through the plurce. For a plume
that scatters a negligible amount of light such as a carbon
particle smoke (Reference 7)
Cp = T - 1 (6)
where T is the transmission.
-------
TABLE 2
PARTICLE SIZE DISTRIBUTIONS OF EXHAUST
FROM THE J-57 ENGINE (1)
Sampler Location,
Engine Diameters
from Engine Nozzle.
0
i
2-1/2
10
. Engine
Thrust
Setting
Approach
75% Norm.
Cruise
Approach
75% Norm.
Cruise
Approach
75% Norm.
Cruise
Particle Size Distribution
Geometric
Median Dia.
um dg
0.053
0.052
0.084
0.084
0.076
0.096
0.13
Geometric
Standard Devia
tiOn °n;
1.63
1.46
1.33
1.40
1.51
1.38
1.40
i
(1) Data obtained from electron microscope samples collected
by the IITRI sampler described in Appendix C.
-------
2.0
0.4
0.2
0.0
Visible Region
Visible Region
Invisible
Region
I I I
-101234 56
Brightness Contrast
8
FIGURE 1. VISIBILITY OF LINE TARGETS.
8
-------
Middleton (Reference 8) discusses the liminal contrast
for visibility of a variety of target shapes, sizes, and
luminance levels. For targets and luminance levels that approach
conditions under which jet engine exhaust plumes are usually
viewed the liminal contrast ranges from about 1 to 57,. The
visual range of objects in natural light is usually based on
the assumption that the liminal contrast is 2?0. On this basis,
the liminal contrast value of 27> was selected for jet exhaust
trails. Substituting 27, into Equation (6) indicates that jet
smoke with transmission values of 98% or higher will be invisi-
ble.
In conclusion, the modeling studies show that carbon
smokes attenuate light mainly by absorption. This factor
permits the use of a simple expression to relate contrast and
transmission measurements to liminal visibility. This rela-
tionship indicates that jet smoke must transmit more than 987,
of the incident light to be invisible. Observing the J-57
engine at the nozzle and perpendicular to the exhaust flow,
theory predicts that a smoke with a 987, transmission will have
a particle concentration of 1.7 x 10 particles/cm^.
The mass concentration of the smoke particles can be
computed from number concentration data .if the particle density
is known. Because of the lacy agglommerated structure of turbo-
jet exhaust particles, as shown in Figure 2. handbook densities
cannot be used. If the value of 0.04 g/cm3 reported bv Horveth
and Charlson (Reference 9) for carbon floes is used, the mass
concentration of the smoke producing a transmitted value of
T
987o is 0.05 mg/m . Values found in the literature for exhaust
o
particulates are 36 mg/m for a PW-JT8D turbofan engine (Refer-
O
ence 10) and 0.5 and 27.0 mg/m for an unstated turbojet engine
at approach and takeoff power. (Reference 11)
(3.3) Theory Relating the Smoke Number and Optical
Transmission of Jet Exhaust Plumes
It is the purpose of this theoretical analysis to
relate the SAE smoke number, the.mass flow of an engine, and
the dimensions of the nozzle, to the optical attenuation of the
smoke plume. The first relationship considered is the effect
of particle concentration in the exhaust gas on the staining
of the filter paper.
-------
at engine nozzle
.* *
2-1/2 nozzle diameters from the engine nozzle
^
-------
(3.3.1) Reflectivity of Soot Deposit on
Filter Paper
The reflectivity of soot deposit on filter, paper
may be represented by:
i
lR-V° (7>
where,
I = incident light flux
IR = reflected light flux
m = total mass of gas through the filter
Q = number of particles per unit mass of gas
K, = specific attenuation coefficient
a = area of filter paper
b = a number expressing the exponential
retention of tne filter.
Equation (7) can be restated as:
b
R = _R = 10 1 a (8)
where,
R = absolute reflectivity of the stained spot.
Taking the logarithm of Equation (8) gives:
Iog10£ - DR = Kl Ab (9)
\a / \ /
where DR is the optical reflection density of the soot deposit,
The validity of this equation may be tested as follows:
Taking the logarithm of Equation (9) we have:
log DD = b(log m + log Q - log a) + log K, (10)
K J-
In a given smoke sampling system, the area, a, of the filter
and K-, are constants.
11
-------
Equation (10) shows that if Q is constant at a given
power setting and different masses of gas are filtered, a
straight line will result when log DR is plotted against log
m. The slope of this line will be b. If the power setting is
changed, resulting in a different value of Q, another straight
line with slope b will result, providing Q remains constant as
m is varied. Data for a variety of power settings, therefore,
yield a family of parallel straight lines. These lines are
offset because of the different particle concentrations, Q,
produced at the different power settings. Plots of this type
(Reference 12) give a series of parallel straight lines. Data
presented in Reference 13 were replotted and resulted in
parallel straight lines foj_- approach, cruise and takeoff power.
(The line for idle was poorly defined by the data points and
appeared to have a slightly different slope. This is not
surprising since the smoke numbers for this low power setting
are less accurate.) The importance of the relationship is that
the slope of the lines is dependent on the filtering character-
istics of the filter. Thus, the filter must be specified.
Changes in filter specifications will be difficult to accommo-
date.
(3.3.2) Optical Transmission of Smoke Plume
The transmission of light through a smoke cloud
may be represented by the expression:
IT = IQ 10"C L K2 (11)
where,
I = incident flux
IT = transmitted flux
Ko = specific absorption coefficient
L = optical path length
C = concentration of particles/unit volume of gas
in the plume.
Equation (11) may be rewritten as:
T = -I. = 10"C L K2 (12)
I
12
-------
vhere,
T = transmission
or:
Iog10 1 = DT = C L K2 (13)
where D is the optical transmission density of the plume.
Equations (12) and (13) are universally accepted expressions
and need no test for validity.
(3.3.3) Relating DR to D and Operating Parameters
Neither Equation (9) nor Equation (13) are parti-
cularly useful for the study of engine exhaust because they
contain the quantities Q and C. Neither of these expressions
are measured directly by the filter stain technique or by the
transmissometer. However, the reflection density of the filter
stain can be related to the transmission density of the plume
at the engine nozzle through the use of engine parameters and
measured values. Thus, theory may be tested by experimentation
If each unit mass of gas and unit volume of gas
issuing from the nozzle of an engine at fixed operating condi-
tion contains the same number of particles (this is not always
strictly true, for at times the smoke issues in a series of
puffs) Equation (14) can be written:
N = MQ (14)
where,
M = mass of gas per unit time passing through the
engine
N = number of particles per unit time in the exhaust.
The particle concentration/unit volume, C, is:
(15)
where,
V = gas velocity (or particle velocity)
A = area of the nozzle.
13
-------
Substituting this value for C in Equation (13) gives
DT =
1
L K
AV
(16)
Rearranging the above equation to solve for Q gives:
Q - DT (-43L)
1 MLK2
(17)
Substituting this value for Q into Equation (9) gives:
b
n
°
AV
n = K
DR Kl T a MLK2
(18)
Equation (18) relates the filter stain reflection density to
the transmission density of the plume through the mass flow
M, the area A of the nozzle, and the velocity V of the gas
flow at the nozzle. This equation may be simplified if the
relationship between velocity and mass flow is substituted.
reasonable assumption for velocity in subsonic flow is:
V = K,
M
(19)
where K-, is the reciprocal of the gas density,
for V in Equation (18) gives:
DR = K
D
T a L K
Substituting
(20)
Equation (20) establishes the relationship between the filter
stain density and the transmission density of the plume. The
optical path length L is the diameter of the nozzle. The
optical density is also measured at the nozzle. The mass flow
M of the engine has been eliminated by the use of Equation (19)
14
-------
Taking the log of Equation (20) gives:
log DR = b (log DT - log L + log m - log a + log K-j
- log K2) + log Kx (21)
A simple means for testing the validity of Equation (21) would
be to plot the reflection density of the filter stain DR against
the transmission density DT of the plume on log-log paper for
a series of values of Q resulting from running the engine at
different power settings. The quantities m and a are held
constant in the smoke sampling system as the power setting is
varied. All other parameters in the equation are constant
including b which is related to the retention of the filter
material. If the equation is valid a straight line with slope
b should result. By substituting into Equations (20) and (21)
known values of all parameters and measured values of plume
transmission and corresponding values of filter stain reflection
density it should be possible to evaluate the constants and
predict the optical transmission from the filter stain for any
engine in terms of DD and the nozzle diameter L.
K
(3.3.4) Relating Smoke Number to Filter Density
The smoke number is defined as:
R
SN = (1 - -5^)100 (22)
where,
R = diffuse reflectance of the smoke spot
R = diffuse reflectance of clean filter paper.
Restating Equation (21), we have:
100 (23)
Rs 100-SN
or,
log1Q ^ = DR = 2 - Iog10 (100-SN) (24)
15
-------
where,
DR = reflective density of the smoke stains.
^Equation (24) gives the relationship between smoke
number "SN and reflection density, DR> Hence by using this
value and Equations (20) or (21) it should be possible to pre-
dict the transmission of the smoke plume as a function of SN
and nozzle diameter L.
(4.) Static Tests of Engines and Aircraft
Static engine and aircraft tests were performed at
NAFEC by both IITRI and FAA personnel. A number of different
engines and aircraft were studied. These were: 1) a J-57-P-
37A Pratt and Whitney Engine mounted on a Static, Sea Level,
Open Air Test Stand; 2) a JT-12-6 engine mounted as a flight
installation in a test wind tunnel; and 3) several aircraft
tied-down on the National Guard run-up area at the NAFEC/Atlantic
City Airport. These aircraft were an F-100, a CV-880,and a
Lockheed Jetstar.
Various types of instrumentation were applied on each
of the tests so that correlations between the results were
possible.
(4.1) J-57-P-37A Engine Mounted in the Test Cell
The J-57 engine instrumentation consisted of an
IITRI transmissometer to obtain smoke transmission data, the
G.E. smoke sampler (Appendix D) to obtain smoke numbers, and
the IITRI smoke sampler to obtain particle size data and smoke
numbers.
Table 3 summarizes the data obtained. The trans-
mission value reported is the one-way transmission as deter-
mined by the transmissometer and was obtained by taking the
square root of the two-way measured-transmission value. The
SAE smoke number is calculated as described in Appendix D.
The other smoke numbers reported are variants of the SAE smoke
numbers. They were obtained by substituting a Millipore filter
for the standard Whatman filter or using a white backing rather
than a black backing when measuring the reflectance of the
filter deposits. For all power settings, no visual detection
of smoke could be made by looking across the exhaust at 90° to
16
-------
the jet nozzle. However, the irregular dark background of the
test cell was not conducive to visible observations. Smoke
could be seen only by looking along the plume path. It appeared
as puffs with low obscuration values, even at the higher power
settings.
Data in Table 3 show that smoke numbers collected
through the IITRI probe are nearly identical to those collected
through the NAFEC probe when both are located at the nozzle.
Since the NAFEC probe was designed for isokinetic sampling, it
is concluded that isokinetic sampling is not required for smoke
number determination. Since both probes gave similar results
at the nozzle, the differences between the smoke numbers obtained
as the IITRI probe was moved downstream from the nozzle reflect
the effect of plume dilution and expansion.
(4.2) Light Transmission of J-57 Engine Exhaust
for Various Path Lengths
Visual observations of smoke plumes with both
tied-down aircraft and aircraft in-flight indicated that the
visibility of the smoke was dependent upon the viewing geometry.
It was observed that a plume may be invisible when viewed at
right angles to the plume axis, but visible as the plume was
observed at other angles. The phenomena is due to the increase
in the attenuation caused by the longer optical path through
the plume. To illustrate this phenomena under controlled con-
ditions, experiments with the J-57 test cell engine were con-
ducted. The transmissometer was used to measure the light
attenuated by the plume at four angles, 90°, 45°, 30°, and 19.5°,to
the plume axis. These correspondto 1, 1.4, 2 and 3 --times the
normal (90°) path length through the plume. The engine was
operated at a series of selected power settings for each angle.
The results of these measurements are summarized in Table 4.
A distinct decrease in transmission as the path length
through the smoke is increased is noted. If the visibility
threshold is taken to be 98% transmission (see page 9 ) then
a smoke which is invisible when viewed at 90° may become clearly
visible when viewed at other angles. The results are not pre-
cise, because at the smaller angles (30° and 19.5°) it was neces-
sary to place the corner reflector approximately 30 and 50 ft.
downstream from the nozzle of the engine to protect it from the
blast. The plume had become much wider at this point. Never-
theless, the principle of increased absorption with path length
is clearly demonstrated.
17
-------
TABLE 3
Test Number
IITRI NAFEC
~
I
8
9
2
3
4
5
10
11
12
13
J4
(1)
(2)
(3)
(4)
(5)
(6)
31
33
35
37
23
25
27
29
39
41
43
45
47
Distance in
Nozzle^Dia.
From Engine
To IITRI Probe
nnd Transmisso.
0
0
0
0
2-1/2
2-1/2
2-1/2
2-1/2
10
1C
10
10
10
SMOKE
NUMBERS AND TRANSMISSION DATA OF JET
SMOKE
J-57 ENGINE MOUNTED IN TEST CELL
Smoke Numbers (l>
G. E. Smoke Sampler
Whatman Filters^2)
Engine
Gas Flow
Lbs/Sec
53
' 94
114
145
53
94
120
148
53
94
114
145
168
Engine
Thrust V
Setting E
Idle
Approach
757. Norm,
Cruise
Idle
Approach
757. Norm,
Cruise
Idle
Approach
757, Norm
Cruise
Takeoff
NAFEC
Ihite
tackinj
21
36
. 49
55
19
35
. 50
58
21
35
. 48
55
58
Probe
Black (t
> Backing^
12
27
. 40
47
11
26
42
51
14
27
40
48
51
IITRI
, , White
4;Backin
16
34
48
54
19
33
42
--
10
22
34
39
--
Probe
Black
£ Backing
25
41
47
11
24
34
--
6
15
. 25
31
--
Millipore Filters^
NAFEC
White
Backing
10
26
38
49
10
24
43
54
10
24
43
48
-.-
Probe
Black
j Backi ng
10
25
37
48
9
24
42
53
10
23
42
46
--
IITRI
White
Backing
6
21
41
47
8
20
31
--
4
13
21
28
--
Probe
Black
j Backing
6
18
40
46
8
18
30
--
4
12
20
27
--
IITRI Sampler
Millipore Fllcer8(3) IITRI
IITRI Probe
White ,,,.
Backing13-'
9
68
94
>100
6
30
59
58
0
9
26
40
40
Transmissi
Percent
Transmiss
99.3
98.4
96.3
95.6
99.0
97.3
95.8
96.0
99.5
98.1
97.0
95.0
89.4(6)
The NAFEC probe was always located at the engine nozzle. The IITRI probe was located at the distances from the engine nozzle given
in column 3. All G. E. smoke sampler data were collected and analyzed by NAFEC personnel.
Whatman filter data were obtained by interpolating reflectance readings to a filtration density of 0.3 ft /in .
Millipore fiiLei: date were obtained by interpolating reflectance readings to & filtration density of 0.0565 ft /in .
The SAE smoke number.
3 2
Because the IITRI sampler filtered diluted exhaust it was necessary to extrapolate the data to a filtration density of 0.0565 ft /in
at thrust settings other than idle.
This value appears to be erroneous. The value of 94.27, reported in Table 4 appears more reasonable.
-------
TABLE 4
LIGHT TRANSMISSION AS A FUNCTION OF PATH LENGTH
J-57 Engine in Test Cell
: Engine
Thrust
Setting
Idle
Approach
75% Norm.
Cruise
. Takeoff
j
Percent Transmission
Relative
1
98.2
97.0
95.3
94.3
94.2
Optical
1.4
99.0
97.0
94.5
92.4
91.6
Path Through
2
98.9
96.4
92.0
89.0
88.0
Plume ^
3
97.8
93.0
85.8
82.2
79.6
'(1) Relative paths correspond to angles of 90
and 19.5° to the plume axis.
45° 30
19
-------
(4.3) JT-12-6 Engine Mounted in Wind Tunnel
The JT-]2-6 engine instrumentation consisted of
the IITRI transmissometer, GE smoke sampler, and IITRI smoke
sampler. No visual or photographic observations of smoke could
be made. The transmissometer was located at a distance of
7-1/2 nozzle diameters downstream from the engine nozzle and
viewed the smoke plume through windows located on opposite sides
of the tunnel. The IITRI smoke sampling probe was also placed
7-1/2 nozzle diameters from the engine. It was connected to
the dilution apparatus by a heated sampling line. This was
necessary because personnel safety considerations prevented
the placement of the dilution apparatus just outside the wind
tunnel at the point where the probe was placed.
Table 5 summarizes the data obtained on the JT-12-6
engine. The JT-12-6 engine gave transmission values exceeding
98% at all power settings. Based on the modeling study, this
engine should not produce a visible smoke. The SAE smoke
numbers were less than 25 at all power settings.
(4.4) Tied-Down Aircraft Tests
(4.4.1) F-100 Single-Engine Aircraft
An F-100 aircraft, operated by the New Jersey
National Guard, was obtained for study. This aircraft has a
J-57-P21A engine, similar to the engine tested in the test cell.
The plane was tied-down at the National Guard run-up area at
the NAFEC/Atlantic City Airport. The exhaust was analyzed with
the IITRI transmissometer, the GE smoke sampler, by photographic
photometry, and by visual observations. A photograph of the
instrumentation is shown in Figure 3. The test results are
summarized in Table 6.
The smoke numbers of the F-100 engine are similar
to those obtained with the test cell engine. The smoke number
at cruise is higher than at takeoff. This inconsistency is
due to the movement of the aircraft at takeoff power. The
nose wheel depresses and the tail of the aircraft rises at
takeoff power; thus, the location of the sampling probe in
relation to the engine nozzle is disturbed. The data show a
a distinct correlation between transmission and photographic
20
-------
TABLE 5
SMOKE NUMBERS AND TRANSMISSION DATA OF JET SMOKE
JT-12-6 ENGINE MOUNTED IN WIND TUNNEL
Smoke Numb
ers' '
Test Numt.
IITRI
WT- 1 , 2
WT-4,5,7
WT-6,8,9
,er
TIA"FEC
77,78
80,81
82,83
Engine
Cas Flow
Lbs/Scc
31
46
56
Engine
Thrust
Setting
Approach
Cruise
Takeoff
NAFEC 1
White
Backing
6
29
'IJ
C.
Whatman Filters*2'
Probe
Black ,
Backing^ '
21
25
IITRI
1 Backing
5
18
E. Smoke
Probe
Black
Backing
t,
10
12
Sampler
Ml
lllipore Filters*3'
NAFECTrobe
White
Backing
2
21
30
Black
Backing
3
?0
. 29
1 1 TB I 1
White
Backing
1
R
11
probe
Black
Backing
2
7
10
IITRI Sampler
Milllpore Filtei
IITRI Probe
White ,,.
Backing13''
J
n
17
r* ' IITRI (6)
Transmissometer v"'
Percent
Transmission
99.1
98.3 i
98.1 '
(1) The NAFEC probe w«» always located at the 12 o'clock position at the engine nozzle. The IITRI probe was always In the center
of the wind tunnel 7-1/2 nozzle diameters from the engine. All C. E. smoke sampler data were collected and analyzed by NAFEC personnel.
(2) Sec note 2, Table 3.
(3) See note 3, Table 3.
(ft) The SAE smoke number.
(5) See note 5. Table 3.
(6) Transmlssomett-r located 7-1/2 nozzle diameters from engine nozzle. Path length was the diameter of the wind tunnel at this location.
-------
TABLE 6
SMOKE QUALITY FROM A TIED-DOWN F-100 AIRCRAFT
J-57-P21A ENGINE
Engine
Thrust
Setting
Idle
Approach
Cruise
Takeoff
Engine
Gas
Flow
Lbs/Sec
53 .
94
137
159
IITRI
Transmissometer.. ^
70 Transmission^ '
Photometry ,,v
70 Transmission^ '
98 100
97
95
94
100
<100
91-94
Smoke Number
Whatman Filter
Black ^W01116
Backing^ 'Backing
9 15
23 31
54 62
48<5> 58<5>
Visual Transmission
Across Plume. % (3)
At Well Behind
Nozzle Aircraft
Clear Clear
Clear Clear
Clear 90-95^
Clear 90-95^
CD
The transmissometer viewed the exhaust at the nozzle and at right angles to the direction
of plume travel. The photographs were taken similarly by lying on the ground and photo-
graphing the exhaust against a sky background.
SAE smoke number.
Viewed against the normal background of the sky horizon.
Smoke is not continuous but billowy.
(5) Low smoke numbers are attributed to the movement of the aircraft at takeoff power.
(2)
(3)
-------
-
"
.«
fete :,,;,
:w&*
' S'«**fe ,5 jf^
i^t.«
FIGURE 3. SMOKE INSTRUMENTATION.
-------
photometry. Also, smoke with transmission values higher than
9770 and smoke number values less than 23 are invisible. Smokes
with transmissions less than 97% and smoke numbers above 23
are visible.
(4.4.2) Convair 880 Aircraft
The left outboard engine of a CV-880 aircraft,
operated by the FAA, was studied with the same methods used for
the F-100 aircraft. The engine was a CJ805-3B. Results
are summarized in Table 7. The data show excellent agreement
between the transmission values obtained by the transmissometer
and by photographic photometry. Plumes with transmission values
of 98% and SAE smoke number of 23 or less were not visible.
Conditions were ideal for photographic photometry.
The plume was well above the horizon, and the sky density below
the plume was essentially the same as above, For these reasons,
a takeoff microdensitometer trace is shown in Figure 4. The
values presented in Table 7, as wel-l as the other tables re-
porting photometry data, are for the maximum decrease in optical
density. For the illustrated example, this value is in the
center of the plume. A positive print of the negative from
which the microdensitometer trace was made is shown in Figure 5.
(4.4.3) Multiple Engines of Convair 880 Aircraft
Multiple engines of the Convair 880 aircraft were
tested using the IITRI transmissometer, photographic photometry
and visual estimates of obscuration. These results are summar-
ized in Table 8. In most cases, the transmission measured
photographically is less than that measured with the transmisso-
meter. This is apparently due to ground interference with the
plume. The height of the transmissometer was approximately that
.of the engines. The photographs, however, show the plume down
to the horizon line which is well below the line of sight of
the transmissomrter. The microdensitometer traces, Figure 6,
show a continuous increase in the density down to the horizon
line. Since the obscuration values given in the tables are
the maximum obscuration values (at the ground line) they will
be higher than the transmissometer values because the trans-
missometer was fixed at the approximate average exhaust height
with the engines. At the time the test was made the influence
of engine tilt and ground dirt was not fully recognized.
24
-------
TABLE 7
SMOKE QUALITY FROM THE NUMBER 1 ENGINE OF A TIED-DOWN CV-880 AIRCRAFT
CJ805-3B ENGINE
Engine
Thrust
Setting
Idle
Approach
Cruise
Takeoff
Engine
Gas
Flow
Lbs/Sec
42
124
147
163
IITRI
Transmissometer
Reading m
% Transmission^ '
98
89
87
87
Photographic
Photometry ,~^
7o Transmission^- '
96
87
85
86
Smoke Number
Whatman Filter
Black m White
Backing^ ' Backing
23 32
60 67
' 69 73
65<6) 70(6)
Visual Transmission
Across Plume. %
At Well Behind
Nozzle Aircraft
100 100
90-95 80-90^3^
90-95 75-85(4)
90-95 75-85(5)
to
(1) See note 1, Table 6.
(2) SAE smoke number.
(3) Smoke is not continuous but billowy. The smoke puffs have a transmission of 80%.
Integrating the clear sky and the puffs yield an estimated visual transmission of
80-90%. Viewing the smoke from in front of the engine, the transmission varies
from 40-60%.
(4) Integrated puffs and background sky give transmission values of 75-85%. When
viewed from in front of the engine the transmission is 40%.
(5) Smoke is more uniform in composition, not so puffy.
(6) Low smoke numbers are attributed to movement of the aircraft at takeoff power.
-------
Decreasing Density
_ . r-r j- -
- Ground Surface
FIGURE 4 . MICRODENSITOMETER TRACE OF
NUMBER 1 ENGINE OF CONVAIR 880
AIRCRAFT - TAKEOFF POWER.
26
-------
N
-J
-S ,
akaflJJL... f'-- .'.. _ Jifete.,.
,, ,,..>,,,,«,,.?--.. l^;*>l?'r*wS*T*?>
_;
FIGURE 5. NUMBER 1 ENGINE OP CONVAIR 880 AIRCRAFT. TAKEOFF POWER.
-------
Idle Power
Approach Power
Cruise Power
Take-Off Power
FIGURE 6. HICRODENSITOMETER TRACES FROM TIED-DOWN
CONVAIR 880 AIRCRAFT - ENGINES 1, 2, and 3
OPERATING.
28
-------
TABLE JB
CV-880 AIRCRAFT TIED-DOVfN MUI.TIPLE KNGINE OBSERVATION
Engine :
Number in>.-.v
Operation^ '
1
1
1
1
1-2
1-2
: 1-2
;
1-2
; 1-2-3
1-2-3
1-2-3
1-2-3
1-2-3-4
1-2-3-4
1-2-3-4
1-2-3-4
f Engine
Thrust
Setting
Idle
Approach
Cruise
Takeoff
Idle
Approach
Cruise
Takeoff
Idle
Approach
Cruise
Takeoff
Idle
Approach
Cruise
Takeoff
IITRI Transmissometer |
% Transmission (2) I
99
91
89
89
98
81
74
71
95
65
52
43
94
61 1
i
38 !
30 __j
Photographic i
Photometry
7<> Transmission
96
87-94
85
86
91
63
49
49
95
49
38
37
94
44
28
38 0)
Visual Transmissions
7o Transmission
100
80-95
75-85
75-85
100
i
100
85
80-85
75-80
100
70-75
70-75
70-75
(1) The number 1 engine is the left outboard engine.
(2) Transmissometer located past the tail of the aircraft at the nozzle as in previous test
(3) It is believed that the power was being reduced when the photograph was taken.
-------
All values obtained with either technique show a similar varia-
tion with power setting and number of engines.
The results of the multiengine study were largely
nullified because of ground dirt becoming airborne during the
tests. The aircraft was positioned at the run-up area with
engines 3 and 4 exhausting over the ground rather than concrete.
The blown ground dirt and the downward tilt of the plume is the
reason why the microdensitometer traces continue to increase
in density down to the horizon. However, the effect of com-
bining smoke plumes should be predictable from consideration of
increased path length.
(4.4.A) Lockheed Jetstar Aircraft
A Jetstar, owned by the FAA, was observed in the
same manner as the Convair 880 aircraft. This aircraft's engine
is similar to the JT-12-6 engine studied in the wind tunnel.
No smoke was visible at any angle of viewing and no smoke
could be detected photographically.
No transmission data were obtained with the 1ITR1
transmissometer because of an amplifier malfunction. Repairs
could not be completed in time to make the tests. Smoke numbers
for the Number 4 engine were measured with the GE smoke sampler
and the results are shown in Table 9.
Results show that probe location must be studied
if representative smoke numbers are to be obtained. Also, a
single probe location may not be a representative location for
all power settings. For example, the 3 o'clock probe position
gave smoke at takeoff averaging about 3070 lower than the 6 o'clock
position. At approach, the 3 o clock position gave smoke numbers
25-30% higher than the 6 o'clock position. The SAF. smoke
numbers were below 23 at the 3 o'clock position for all thrust
settings. The SAE smoke number at the 6 o'clock position
reached 35 at takeoff thrust.
(4.5) Discussion and Summary of Results of
Static Engine and Aircraft Tests
.This discussion will show the empirical rela-
tionship between the visibility, luminance, and smoke number.
Table 10 is a collection of data taken from Tables 33 5, 6, 7
and 9, arranged to show more clearly the comparison of smoke
30
-------
TABLE 9
SMOKE NUMBERS FROM THE NUMBER 4 ENGINE
OF A TIED-DOWN JETSTAR AIRCRAFT
I Engine
' Thrust
Setting
Approach
Cruise
Takeoff
i
Smoke Number 1
Probe at
6 O'clock
Backing
Black(l) White
8.6
30.0
i 35.0
1
12.0
37.0
42.5
Probe at 3 0
'Clock
Backing
Black(l) White
10.7 15
23.0 30
23.0 31
.5
.9
i
.7
(1) SAE Smoke Number
31
-------
number, transmission and visibility. Figure 7 is a plot of the
measured transmission against the smoke number. These values
of transmission were measured at right angles at the nozzle on
a single engine. As discussed on page 17 , we assume that smoke
with transmission values greater than 98% will be invisible.
In Figure 7 a line at 98% transmission is shown to divide the
plot into two parts. Above this line, the exhaust is invisible;
and below the line the smoke is visible. It will be seen that
SAE smoke numbers less than 23 are invisible.
A comparison between transmission measurements made
with the transmissometer and with photographic photometry,
Table 10, indicates that both techniques lead to similar values
within experimental error. Hence, photographic photometry is
a valid measure of smoke plume transmission. Photographic
photometry should therefore result in true transmission values
when applied to aircraft in flight,
A comparison of Columns 4 and 5 with Column 6 of Table
10 show a reasonable agreement between the transmission
measurements and visual estimates of transmission.
(5.) In Flight Observations
(5.1) Flights Evaluated at NAFEC
A number of different aircraft were studied
in-flight at NAFEC/Atlantic City Airport, Figure 8. Photographs
and visual estimates of the obscuring power of the smoke were
made. The aircraft were observed during practice landing
operations on runway 13-31. The observers were stationed at
the junction of taxiways G and F. Aircraft maneuvers consisted
of touch and go landings and takeoff or low altitude fly-bys
at about 500 feet.
(5.1.1) F-100 and F-105 Aircraft In-Flight Tests
An F-100 aircraft made a number of low altitude
fly-bys at several power settings with and without afterburners.
A typical photograph is shown in Figure 9. Results of the.
visual observations and the photographic photometry measurements
are given in Table 11. Two passes by an F-105 aircraft are
reported in Table 12.
32
-------
c
o
H
w
CO
s
u
H
g
o
u
0>
PL,
100
oo
96
94
92
90
88
86
tfl
J.U. i
IE!
im
4-i
tfri-
rtrr
33
-B
ritf
0 10 20 30 40 50 60 70 80
SAE Smoke Number
OJ57 in test cell
XJT-12 in wind tunnel
QF-100 aircraft tied-down
VCV-880 aircraft #1 engine tied-down
(All Data Taken from Table 10)
FIGURE 7.
PLOT OF SMOKE NUMBER vs
OPTICAL TRANSMISSION.
33
-------
SUMMARY OF SMOKE NUMBER
AND TRANSMISSION FOR STATIC TESTS
I
Engine > Engine
Thrust
Setting
SAE m
Smoke *> '
Number
Transmission,
IITRI
Transmis-
someter
J-57 on Idle 16 i 99.3
Test Stand Approach 27 98-4
75° Norm. 41 96.3
'
Cruise
i Takeoff
JT-12 in ' Approach
' Cruise
Tunnel
F-100
Tied-Down
CV-880
Tied-Down
#1 Engine
Jet-Star
#4 Engine
Takeoff
Idle
Approach
Cruise
Takeoff
Idle
Approach
Cruise
Takeoff
Approach
49 95.6
51 94.2(2)
4 99.1
21 98.3
25 98.1
9 98
23 97
54
48
23
60
69
65
9.8(3)
Cruise 26.5
95
94
98
89
87
87
__
--
Takeoff 29
i I
(1) Smoke number for black backing and Whatman #4
(2) No data was obtained for takeoff power at the
Photu.
Trans-
mission
..
--
--
--
100
100
<100
91-94
96
87
85
86
100
100
100
filters.
time the
% |
Visual i
Estim. i
Trans . '
--
-.
--
--
--
--
100
100
90-95
90-95
100
80-90
75-85
75-85
100
100 i
100
i
other
(3)
test data was obtained (Table 3) due to the high noise level and
vibration causing a malfunction of the voltmeter. The value shown
here is from Table 4.
Smoke numbers for Jetstar are averages from black backing values
at the two probe positions. See Table 9.
34
-------
SCALl
FIG. 8 NAFEC/ATLANTIC CITY AIRPORT, ATLANTIC CITY, NEW JERSEY
35
-------
'
»
.
F-100
.
v- .^r.ss
-
.
CV-880
-
CV-880
CV-880
FIGURE 9. PHOTOGRAPHS OF F-100 AND CONVAIR 880
AIRCRAFT IN FLIGHT.
36
-------
OBSERVATIONS ON THE F-100 AIRCRAFT IN FLIGHT AT NAFEC
(l\
Pass v '
Number
1
2
3
4
5
5
6
7
.8
: 8
9
9
: 10
10
11
11
12
12
: 13
13
14
14
Thrust
Setting
Takeoff(5)
Takeoff
Military
Military
Cruise
Cruise
Military
Military;.
Military
Military
Military
Military
Military
Military
Military
Military
Cruise
Cruise
Cruise
Cruise
Cruise
Cruise
Viewing*- '
Angle
90°
90°
90°
90°
90°
70° (6)
90°
90°
45° (7)
90°
30° (7)
90°
20- 30° (7)
90°
30° (7)
90°
20° (7)
90°
10- 15° (7)
90°
10° (6)
90°
Percent Light Transmission Through Plume
Photographic (3)
100
100
94
93
96
92
96
95
89
95
86
96
89
94
90
95
87
95
80
95
90
95
Visual (4)
100
100
90
90
90
90
85-90
85-90
70-80
70-80
50
80
50
80
50
80
60
85
60
85
60
85
(1) No.'s 1-7 on one film and 8-14 on other film taken on different days.
(2) Viewing angle - angle made by line of sight and plume axis.
(3) Transmission measured by photographic photometry (Appendix B).
(4) Transmission as visually determined at time of flight.
(5) With afterburner. All other data are without afterburner.
(6) Aircraft flying away from observer, after landing exercise.
(7) Aircraft approaching observer during landing or fly-by exercise.
37
-------
TABLE 12
UbbEKVATIONS ON THE F-105 AIRCRAFT IN FLIGHT AT NAFEC
L
! . . .
Pass Estimated
Number ! Thrust
i Setting
I
1
1 i Military
2 j Military
i -------
Viewing Percent Light Transmission
Angle Through Plume
Photographic(l)
90° 96
90° 95
Visual(2)
85-90
85-90
i
(1) Transmission measured by photographic photometry (Appendix B)
(2) Transmission is visually determined at time of flight.
33
-------
(5.1.2) Convair 880 Aircraft In-Flight Tests
Similar tests were made with the FAA Convair 880
aircraft. The data is summarized in Table 13. The data pre-
sent an outstanding example of how smoke density increases as
the smoke path is increased due to the position of the observer
in relationship to the smoke track. When viewed at right angles
visual transmission values of about 80-9070 are typical. When
viewed at 45°, visual transmission is on the order of 50% or
less. Figure 9 shows some photographs of the CV-880 aircraft
in-flight. As seen from the photographs the air during this
series of flights seemed excessively turbulent.
(5.1.3) Lockheed Jetstar Aircraft In-Flight Tests
No smoke was visible to the eye and no smoke was
detectable from photographs of the Jetstar. This aircraft's
plume appeared clear under all conditions of viewing.
(5.2) Observations of Aircraft In-Flight at
Chicago's O'Hare Airport
Table 14 summarizes the results of photographs
of commercial aircraft in-flight taken at O'Hare Airport on
September 25, October 23, and November 26, 1969. Most of
the photographs were taken of aircraft during takeoff but s.
few, as noted, were taken of landing aircraft. Typical photo-
graphs are shown in Figure 10. The location of the observer
relative to the plume path was far from ideal. In most instances
the observer and camera were located at an undesirably long
distance from the flight path.
(5.3) Discussion and Summary of In-Flight Tests
The flights observed at NAFEC show that the visual
estimates of transmission are slightly lower (5-1070 lower) than
the photographic measurement. This discrepancy did not appear
in the tied-down tests previously described and is probably due
to the subjecture effect of a small plume viewed against a large
expanse of light sky. Photographic transmission measurements
made at 90° to the plume axis are always higher than those made
at lesser angles. A similar trend may be seen in the visual
estimates. Cere was taken to evaluate visually the change in
transmission with the viewing angle at NAFEC. The importance
of this factor was not fully recognized in the earlier tests
at O'Hare Airport.
39
-------
TABLE 13
OBSERVATIONS ON THE
CONVAIR 880
AIRCRAIT IN FLIGHT AT NAFEC
Pass
Number
Thrust
Setting
1 Takeoff
i
1 Takeoff
1 ' Takeoff
2 ; Approach
2
2
2
3
3
3
3
3
4
Approach
Approach
Approach
Approach
Approach
Takeoff
Takeoff
Takeoff
Approach
(1) Transmission
Viewing
Angle
90°
45° (5)
0°(5)
45° (4)
90°
5- 10° (5)
0° (5)
5-10° (4)
10- 15° (4)
90°
45° (5)
20° (5)
0-5° (4)
Percent Light Transmission
Through Plume
Photographic (1)
70 & 82(3)
45
52 & 70(3)
62 & 83(3)
81
44
-70
65 & o 7
70
70
44
52
44
Visual (2) |
50-70
50-70
I 50-70
50
85-90
i
t
I
80
60-80
80
20
30
50
measured by Photographic Photometry (Appendix B)
(2) Visual estimate of transmission made at timr
(3) Two distinct
of flight.
trails appear on photograph-transmission given
for each trail.
Aircraft approaching observer during landing or fly-by
exercise.
'(5) Aircraft flying away from observer after landing exercise.
40
-------
PHOTOGRAPHIC MEASUREMENT OF JET SMOKE
TRAILS
FROM AIRCRAFT IN FLIGHT AT
O'HARE AIRPORT AND VISUAL ESTIMATES OF TRANSMISSION
Type and
Operational
Mode of
Aircraft(l)
727
727
DC9
727
727
727
: 727*
; 727
707
737
! 727
: 727
i
i
; 707
1 - -.
I 2 engine
Distance
Behind
Nozzle(2)
at nozzle
25 ft.
100 ft.
at nozzle
75-100 ft.
100-200 ft.
at nozzle
50 ft.
150 ft.
at nozzle
100 ft.
at nozzle
at nozzle
at nozzle
at nozzle
at nozzle
at nozzle
at nozzle
at nozzle
100 ft.
at nozzle
100 ft.
200 ft.
at nozzle
I Number
! of
Plumes (3)
Percent Light Transmission
Through Plume
Photographic (4) Visual(5)
1 67
1 ; 69
1 i 83
3 83,80,84
25
2 76,82
1
1
. 1
1
1
' 1
; 1
1
1
1
1
1
2
1
1
1
1
1
1
76
58
65
73
50
73
68
70
76
74
86
100
74,83
68
73
86
50
40
35
10
30
35
70
45
35
30
55
86
89
77
35
41
-------
Type and
Operational
Mode of
Aircraft(l)
2 engine
2 engine
2 engine*
DC9
DC9
727*
i
707*
727
i
i
i
' Distance
Behind
Nozzle(2)
at nozzle
at nozzle
at nozzle
at nozzle
at nozzle
at nozzle
j at nozzle
[ at nozzle
!
Number
of
Plume (3)
1
1
1
1
1
1
1
1
Percent Light Transmission '
Through Plume
Photographic (4) Visual (5)
93
76
73
76
75
63
93
73
60
20
25
20
20
10
85
15-25
(1) Asterisk identifies those aircraft landing at O'Hare. All other
aircraft were observed during takeoff. All aircraft are jet
aircraft.
(2) Estimated distance behind engine nozzle that the photographic
transmission measurements were performed.
(3) Number of smoke plumes at point where photographic transmission
measurements were performed.
(4) Transmission measured by photographic photometry (Appendix B).
The values are for a 90° viewing angle.
(5) Transmission as visually determined at time of light. The
effect of optical path length was not fully appreciated at the
time these measurements were performed. While the visual measure-
ment was made "on mark" from the camera operator in effect the
visual observer was mentally measuring the smoke density up until
'the time the photographs was taken. The visual transmission values
therefore are not instantaneous but, intregated over a short
period of time.
42
-------
..
V
,
,-
727 Jet - Ringelmenn Number 4-1/4 707 Jet - Ringelmarm Number 3/U
September 25, 1969. September 25, 1969.
FIGURE 10. PHOTOGRAPHS OF JET EXHAUST
TRAILS DURING TAKOFF AT O'HARE
AIRPORT.
43
-------
At O'Hare, a much greater difference exists between the
photographic and visual estimates. The Ringelmann reader
always gave a reading equivalent to 0.75 to 1 Ringelmann
numbers higher than the photographic method. These measure-
ments were made with no previous experience on aircraft smokes
and it appears that the observer tends to exaggerate the black-
ness of the trail against the expanse of sky. Also, the visual
measurements reported were made on command from the camera
operator at the time the photographs were taken. Because the
smoke observer was mentally measuring the smoke density up
until the time the photographs were taken the value reported
is an integrated rather than an instantaneous reading. The
photographic results essentially represent a plume viewing
angle of 90°. However, the visual observations show the
influence of viewing the smoke through greater optical path
lengths prior to taking the photograph. For later measurements,
the observer became more critical and was aware of great dif-
ferences for various viewing angles on the same smoke trail.
There appears to be no reason for demoting the photographic
values in Table 14 because they agree with transmissometer data
obtained under controlled conditions. Hence, the apparent
difference in transmission as reported by the visual observer
at NAFEC and O'Hare is due to the increased experience of the
observer. It should be noted that the photographic results
were not obtained until several weeks after the observations
were made, so they could not influence the observer.
44
-------
CONCLUSIONS
It is concluded that:
1. Consideretions of turbojet exhaust visibility
indicate that siuoke plumes with transmission values greater
than 9S7o will be invisible.
2. Correlations between the SAE smoke TV.;trb«*r and
transmissoinettn- datp indicate thpt for the engine* investigated,
smoke numbers less than 23 ere associated with invisible plumes.
3. Photographic photometry provides a means for
evaluating the optical density of smoke trails from in-flight
turbojet aircraft.
^. A transmissometer, of the type designed for use on
this program, provides a means for evaluating the optical density
of turbojet engines mounted in test cells and on tied-down
aircraft. Transmission data is less sensitive than the SAE
method of smoke measurement.
5. A Ringelmann smoke observer tends to overestimate
the obscuration properties of jet smoke. This is believed due
to several factors. Among these factors are the relative
crudeness of the Ringelmann scale, the wide field of vision,
the accentuated visibility of line targets, end the increase
in the optical path length through smoke, when the plune is
viewed et angles other than 90°.
6. Attempts to evaluate the effect of multiple plumes
were nullified by ground dirt stirred up by the engine blest.
However, the effect should be predictable based on theoretical
considerations, as indicated by angular measurements performed
on a single engine exhaust plume.
7. Jet engine smoke is composed of lacy agglomerates.
The geometric median agglomerated particle size of the exhaust
from the J-57 engine operated et cruise power is 0.052 urn. The
geometric standard deviation is 1.^6. At 10 nozzle diameters
from the nozzle the geometric median size of the agglomerates is
0.13 urn.
-------
REFERENCES
Aerospace Recommended Practice, ARP 1179, Smoke Measurement
Techniques, Society of Automotive Engineers Inc.,
2 Pennsylvania Plaza, New York, New York, 10001, Issued May U,
Bureau of Mines Information Circular 7718. The Ringelmann
Smoke Chart, U.S. Department of Interior, August
3. Marks, L., Mech. Engr . , J59, 681, 1937.
4. Sawyer, R. , Astronautics and Aeronautics, 8, 62, 1970.
5. Douglas, D., National Bureau of Standards Memo, May 20, 1954.
6. Langer, G., "The Development of Chemical Smoke Tracking
Aids," IITRI Final Report C086, 1957.
7. Conner, W., Hodkinson, J., "Optical Properties and Visual
Effect of Smoke Stack Plume,'1' PHS Publ. 999-AP-30, 1967.
8. Middleton, W., "Vision Through the Atmosphere," University
of Toronto Press, Toronto, Canada, 1952.
9. Horvath, H., Charlson, R., Amer . Ind. Hyg . Assoc. J., 30,
500, 1969. ~~
10. Sallee, P. G., "A Status Report on Jet Exhaust Emissions,"
Air Transport Assoc. of America, 1000 Connecticut Ave . ,
Washington, D.C. 20036, Dec. 11, 1967.
11. Dix. D., Bastress, E., "Nature and Control of Aircraft
Engine Exhaust Emissions," Report #1134-1, Northern
Research and Engineering Corp., 219 Vassar St., Cambridge,
. Mass. 02139, Nov. 1968. Report prepared for National Air
Pollution Control Administration, Contract Number PH22-68-27.
12. Shaf fernocker , W. and Stanforth, G. M., Smoke Measurement
Techniques , SAE Air Transportation Meeting, Paper number
680346, April 29, 1968, New York.
13. Slusher, G., "Review of Investigations on Aircraft Pollution,"
FAA Program Review, NAFEC May 1970.
14. Mack, J. E. and Martin, M. J., "Tine Photographic Process"
McGraw-Hill, pp. 202-204.
15. Hardy, Arthur, G., andPerrin, F. H., "The Principles of
Optics," McGraw-Hill Inc., 1932, pp. 212-213.
46
-------
APPENDIX A, TRANSMISSOMETER
A transmissometer, Figure 1.1, was designed and con-
structed to measure the optical transmission of smoke plumes
from jet engines under varying ambient light conditions. This
transmissometer is readily portable and can operate over optical
path lengths ranging from a few feet to several hundred feet
without impairing its effectiveness.
A schematic optical diagram of the instrument is shown
in Figure 1.2. Light from a ribbon filament lamp is focused by
the condensing lens on the aperture plate and collimated by the
achromatic lens. A rotating chopper, located between the con-
densing lens and the aperture plate, modulates the collimated
beam. The beam is directed across the exhaust plume of the
engine under test. On the opposite side of the plume, the beam
is intercepted by a cube-corner retroreflector which directs
the light back along the same path to the collimating lens.
Half of the returned beam.is directed to the SGD-100 silicon
detector by the half-transmitting mirror. Baffles minimize the
amount of light reaching the detector from scattering within
the instrument. An infrared absorbing and visible transmitting
filter is used to approximate visual response.
Light can reach the detector in four ways:
(1) Light from the main beam is reflected by the
corner reflector. This tranverses the smoke plume
twice and is attenuated by absorption and scattering
in the plume. The signal generated in this manner is
the signal of interest.
(2) Light may reach the detector from ambient light
either by scattering or by directly entering the col-
lection optics.
(3) Light scattered in the direction of the collecting
optics by particulate matter in the plume may reach
the detector.
(4) Light scattered within the instrument by lens
elements, the walls of the tubes, and dust on the opti-
cal surfaces may reach the detectors.
A synchronous lock-in voltmeter operating at the chop-
ping frequency is used for signal detection. Since the ambient
light is not modulated by the chopper, the ambient light gives
no signal.
1-1
-------
-.'1 - - .fM»^KMgr .'jjif^asi'SB^JSWBBfiJF
: T -'
1
1
trti nrrtiiti .. i ^^N&>^
;
'
I " - ----..
-J
*-.-.,
:
,
RETROREFLECTOR
TRANSMITTER-RECEIVER
FIGURE 1.1 TRAHSMISSOMETER,
-------
Light Trap
If-
Half
Transmitting
Aperture /Mirror
10-200 ft.-
Jet Plume
Condensing
System
ibbon Filament
Collimator
Detector
SGD 100A
Cube
Corner
Reflector
Alternate
Chopper
Position
Chopper
Position
FIGURE 1.2 OPTICAL SCHEMATIC (NOT TO SCALE).
-------
Chopped light from the beam, which is scattered back
into the optics by the aerosol, as in Item 3 above, will appear
as a real signal. To minimize this signal, the angular field of
view is only 36 minutes of arc. By placing the chopper at the
corner reflector, as shown in Figure 1,2 as the alternate chopper
position, the signal contribution due to backscatter can be
eliminated. By alternating the chopper position, the effect of
scattering can be evaluated. In actual practice, no change in
signal due to backscatter was measured, and for convenience, the
chopper was used at the source position for all measurements.
Light scattered from within the instrument will give
rise to an unwanted signal. However, by careful baffling and
inserting a light trap above the half-transmitting mirror, error
from this source was very small. At distances up to 50 feet,
this signal was less than 0,17, of the signal obtained from a
clear path. For distances of 150 feet or more, where the un-
wanted signal approached 170, the offset-voltage-adjust on the
voltmeter was used to zero out this unwanted signal. Thus,
the effect of internally scattered light can be completely
removed from the true signal.
Errors due to refraction by the plume are also eliminated
by the instrument. The size of the projected uniform beam is
larger than the retroreflector; hence, the illumination of the
retroreflector remains unchanged by refraction in the plume.
The return beam will retrace the same path as the projected
beam so that its intensity is not affected by refractive index
gradients.
The system is used to measure the transmission of smoke
plumes in the following manner: the transmitter-receiver and the
cube-corner retroreflector are set-up on opposite sides of
the smoke plume. The light and chopper are turned on the
beam centered on the retroreflector. The intensity is adjusted
by a variac to give a reasonable signal. This signal is the
signal corresponding to 100% transmission. The system is
allowed to stabilize, and the detected voltage is recorded on a
strip chart recorder. The engine under test is then started,
and the signals resulting from various power settings are re-
corded. The two path transmission is calculated by taking the
ratio of the signal to the signal obtained with 1007=. transmission.
The square root of this ratio is the single path transmission
through the plume.
1-4
-------
APPENDIX B, PHOTOGRAPHIC INSTRUMENTATION
Photographic photometry was chosen as an objective
technique ror quantitatively measuring, the apparent transmission
of smoke plumes from aircraft in flight. If the luminance of
the smoke trail and the luminance of the background upon which
the smoke trail appears can be measured, then the relative
luminance will give a measure of the apparent transmission of
the plume. If the plume is composed of light absorbing particles
that scatter light poorly then the value obtained is a true
measure of the transmittance of the smoke. In actual practice.
it is necessary to use the luminance of the area adjacent to
the smoke plume rather than the area behind. Unless the sky
is composed of small broken clouds, use of the adjacent area
presents no difficulty.
While a direct reading photometer (such as small spot
brightness meter or photographic exposure meter) could be used,
no permanent record is available. Also, since the phenomena is
transient, it is difficult to obtain reliable readings. To
overcome this difficulty, it was decided to photograph the
smoke trails and use photographic photometry to recover the
luminance values from the photograph.
A 35 mm single lens reflex camera with 50 mm. 125 mm,
and 250 mm lenses, was used to photograph the smoke trails.
The longer focal-length lenses gave sufficiently large images
to permit easy measurement of the density of the smoke plumes.
The 125 mm proved most useful for aircraft in-flight. For
tied-down aircraft, the 50 mm proved best. For measuring the
density., a David W. Mann #1140 microdensitometer was used.
The technique of photographic photometry is as follows .
In addition to photographing the subject whose luminance is to
be measured, a transparent gray-step wedge is also photographed.
When a 35 mm camera is used, the gray-step wedge is simply
photographed on a separate frame from the subject. The film is
then developed. Reasonable care is taken to ensure uniform
development. (If a camera other than a 35 mm is used and the
gray-step wedge is photographed on a separate film, care must
be taken so that is receives exactly the same development time,
temperature, and agitation.) In order that the film will exhibit
identical response to the gray-step wedge and the subject, the
gray transparent step wedge is photographed x^ith the sky as
illumination.
-------
A convenient step wedge for calibration of the film
is the Eastman Kodak Step Tablet #3, having 21 density steps
ranging from 0-05 density units to 3.5 density units. The
range of the step wedge is such that when the proper exposure
is given, some steps are underexposed and others are over-
exposed.
The density of each step of the original step tablet
is measured on the microdensitometer, as are the corresponding
steps on the negative.
Density is defined as:
D - log10 !> , log10 I (2.1.
where, I = the light incident on the negative,
T = the fraction of the light transmitted by the negative,
I = the light transmitted.
If the density of the negative step wedge is plotted
against the logarithm of the exposure the characteristic H & D
curve of the photographic emulsion results (Reference 14) .
(The exposure is the product of time and illuminance of the
film and is commonly given in meter-candle-seconds.) It is
not necessary to know the absolute value of exposure, but only
the exposure ratio between the various steps of gray. From
Equation (2.1), it is readily seen that the logarithm of the
transmission ratio, that is, the density difference between
steps of the original gray scale, is proportional to the nega-
tive logarithm of the exposure ratio for the corresponding steps
on the negative. Consequently, it is convenient to plot the
data for the H & D curve as shown in Figure 2.1, where the
density steps of the negative are plotted against the corres-
ponding density steps of the original gray scale.-'
* The density values are measured by a densitometer depend
upon the geometry of the collection optics and the light scat-
tering of the sample as well as upon the absorption of the
sample. That is, the density measured for a scattering sample
by an instrument accepting a large solid angle will differ from
that measured by an instrument accepting a small solid angle.
The first type of instrument measures "diffuse density" and the
second "specular density." A sample which does not scatter, but
2-2
-------
Exposure Ratio
0.01
0.1
1.0
00
0)
c
o
I-t
x:
a
CO
S-i
60
o
4J
O
JT
CL
-------
attenuates only by absorption would give the same measurement
regardless of the instrument, A photographic step wedge does
scatter some light and the density will vary with the instrument
used. Therefore the density as measured by the densitometer may
be different than the effective density as observed by the camera
and an additional step is required before the measured density
differences may be expressed as log exposure ratios. The measured
density may be converted to effective density by experiment as
follows. A series of carefully times camera exposures are made.-
of the step wedge. These various timed exposures provide a
series of known exposure ratios with which the measured densities
Df the original step wedge may be compared. In general it is
found that the density values obtained by different instrument.'-
nay be expressed by D = g D , where D represents the density
o Q. 5
obtained with a specular instrument (limited acceptance angle)
and D, represents the density as measured by a more diffuse
instrument (large acceptance angle) and g is a constant called
Calliers coefficient (Reference 15,) Calliers coefficient will
oe a constant depending upon the characteristics of the two
particular instruments. Such an experiment was performed on the
present program and it was found that density values as measured
by the Mann densitometer corresponded exactly to that of the
camera optics. Therefore, Calliers constant was unity and no
correction was required.
If the corresponding densities of smoke plume and background
on a negative are measured by a microdensitometer and compared
to the H & D curve for that negative, the exposure ratio and
hence, the transmission of the plume can be determined.
It is characteristic of photographic film that a
portion of the H & D curve is a straight-line, thus the density
difference divided by the slope of the straight-line gives the
logarithm of the ratio directly. (A properly exposed negative
takes advantage of this straight-line portion to correctly
reproduce the relative luminances of the scene.) The slope of
the straight-line portion of the H & D curve (commonly called
"gamma," y) is controlled by the development and is constant
over the entire length of the film. Hence, a calibration step
wedge need be photographed only once on the film.. In practice,
it is usually photographed at the beginning and end of the film
to check on uniformity of development.
-------
Figure 2.2 illustrates the steps in determining the
luminance ratio of a smoke cloud to its background. Figure
2.2a represents a negative obtained by photographing the smoke
cloud; Figure 2.2b represents the negative obtained by photo-
graphing the gray wedge; and Figure 2.2c represents an H & D
plot of Figure 2.2b. The density of the steps in the gray-step
negative are plotted against the density of the original step
wedge. Suppose one wishes to measure the smoke luminance ratio
for 2 points, a and b , in Figure 2.2a.The density of each
of the points is measured and the corresponding density values
plotted on the H & D curve as shown in Figure 2.2a. The corres-
ponding original gray step wedge density difference is read from
the abscissa of the curve. The density difference is equal to
the negative log-,Q of the luminance ratio.
Example: The density of point a is 1.30 (see ordinate),
The density of point b is 0.99. These two points correspond
to the original wedge density values a', b1 of 0.75 and 1.10
(see dotted lines on Figure 2.2c.) Therefore:
, . luminance at b , , n n -,c n TC
-log,n : = 1.10 - 0.75 = 0.35
i-U luminance at a
or,
luminance at b = Transmission
°10 luminance at a to!0
of Smoke = -0.35
Hence, Transmission of Smoke = antilog (-0.35) = 0.45 = 45%
Because the points a and b both lie on the straight
line portion of the H & D curve, an alternate calculation is
possible. The density difference between a and b on the
negative is £D = 0.31. The corresponding density difference on
the original gray scale is:
AD' = AD x - = 0.31 x - =0.35
Y 0.885
Hence, = antilog (-0.35) = 4570 transmission.
3
-------
000
c
-*
'x .
DO/0000000000000000
54321
£00000000000000000000 0 0 0
a. Negative of Scene b. Negative of Gray Stepwedge
£ 2.00
CO
&0
cu
25
«« 1,30
o
>, 1.00,
c 0.99
0)
Q
0
Gray Step Levels
(1-2-3-4-5)
a'0.75
.885
2.00 1.00 0
Density of Original Gray Scale
c. H & D Curve of Negative
FIGURE 2.2 ILLUSTRATION OF USE OF
PHOTOGRAPHIC PHOTOMETRY,
2-6
-------
Note that only the luminance ratio between the smoke
and the background, as seen from the observer's viewpoint, is
measured. If a highly scattering medium exists in the inter-
vening atmosphere between the observer and the plume, the obser
ved luminance ratio will be reduced over the value at the plume
That is, contrast attenuation by the intervening atmosphere
may render the interpretation inaccurate. This effect operates
regardless of the type of remote observation and a Ringelmann
reader is subject to the same difficulties. Hence, for relia-
bility, observations must be made at close range and under clear
air conditions.
2-7
-------
APPENDIX C, SAMPLING JET
EXHAUST FOR PARTICULATE MATTER
A special exhaust particulate sampling apparatus was
used to provide a Dilute sample of the exhaust gas for particle
size determination. The apparatus is shown in Figure 3,1.
Exhaust is drawn into the sampler by a syphon and diluted almost
immediately with a large volume of cool air. A tangential
inlet on the diluter facilitates mixing of the exhaust and diluent
air. The cooled, diluted exhaust flows down the dilution cham-
ber and exits at the bottom. Near the exit,samples of the
diluted exhaust are collected on membrane filters and electron
microscope grids. Flows to the apparatus are controlled by
flowmeters. The syphon was calibrated in the laboratory prior
to testing. However, the ram air effect of the exhaust in-
validated the calibration and a thermoanemometer was added near
the dilution chamber exit to monitor the total volume of air
flowing in the apparatus. Knowing this value and the amount
of syphon and dilution air, the volume of jet exhaust drawn
into the apparatus was calculated. Critical orifices regulated
the sample flow rate through the filters. Membrane filters
were used. The instrument is shown in use at the J-57 test
cell in Figure 3.2. The apparatus, with the exception of the
flowmeters and pump, was placed at various distances behind the
engine test fixture; the sampling probe was always located at
the center of the exhaust plume. The probe was 3/8 inches in
diameter.
The important features of the apparatus are enumerated
below:
1. Agglomeration problems are circumvented by quickly
diluting the exhaust.
2. Condensation problems are avoided by large volumes
of cool diluent air.
3. Heated sample lines are not necessary.
4. Sample line deposition problems are mitigated by
the use of short lines, Cleaning is facilitated.
The apparatus was used to obtain size and structure
data on the exhaust particulates. Reflectance measurements
were made on the membrane filters and a smoke number based on
a filtration density of 0.0565 ft.3 of exhaust per in.2 of
filter was calculated. A white background was used during the
reflectance measurement. The results, therefore, are not the
SAE smoke number.
J-l
-------
Syphon
X
Exhaust
Sample
fr
t
Tangential
Inlet
Dilution
Chamber
Royco Particle
Counter Sample
Inlet
Thermo-
couple
Jet Engine Room
Dilut
Air
To SAE
Smoke Sampler
Syphon
Air
i TJ
Flowmeters
37^f Hg
-^ ' Manometer
Clean Filtered Air
To Royco
Counter
Cross-Section AA'
Thermo Anemometer
Restricted Orifices 900 cc/min
Solenoid Valves
Control Room
FIGURE 3.1 EXHAUST PARTICULATE SAMPLING APPARATUS.
3-2
-------
;
" "
FIGURE 3.2 IITRI SMOKE SAMPLER IN THE ENGINE
TEST CELL
3-3
-------
The diluted jet exhaust was also analyzed by a Royco
Model 245 light-scattering particle counter. A 25 ft, sample
line was used to connect the sampling apparatus to the Royco.
The lengthy sample line was needed because the Royco is sensi-
tive to noise and vibration; therefore, it was located in the
control room. The Royco provides particle size distributions
for aerosols whose particles range from 0.3|j to about 10(j . As
shown in Table 2, the vast majority of smoke particles are
below 0.3u. Also, the concentration of particles, even in the
diluted exhaust, exceeded the counting capacity of the Royco
at all power settings, except idle. The Royco data for the
J-57 engine are reported in Table 3.1.
The IITRI sampler was also used to sample the exhaust
from a JT12-6 engine mounted in the NAFEC wind tunnel. The
probe was placed in the center of the wind tunnel 7-1/2 nozzle
diameters from the engine. Due to personnel safety considera-
tions the dilution chamber was located in the control room and
connected to the probe with a heated sampling line. Royco
particle size and concentration data for the JT12-6 engine are
reported in Table 3.2.
A General Electric smoke console was used by NAFEC
personnel to collect samples from the IITRI sampler probe for
smoke number determination. A lengthy, heated sample line,
1/4 inches in diameter, was connected to the top of the syphon
for this purpose. When the smoke number was being determined,
air flows to the syphon and diluter were shut off. The line
to General Electric smoke console was plugged when the console
was not being used.
3-4
-------
TABLE 3.1
SIZE DISTRIBUTION AND CONCENTRATION OF EXHAUST
AS REPORTED BY THE ROYCO PARTICLE COUNTER- J- 57- P-
Test
IITRI
6
7
8
9
2
3
4
5
10
11
12
13
14
Number ;
NAFEC
31
33
35
37
23
25
27
29
39
41
43
45
47
Sampler Location^
Engine Diameters
From Engine Nozzle
0
0
0
0
2-1/2
2-1/2
2-1/2
2-1/2
10
i
Engine
Thrust
Setting
Idle
Approach
75% Norm.
Cruise
Idle
Approach
75% Norm.
Cruise
Idle
10 Approach
10
10
10
75% Norm.
Cruise
Takeoff
0.3-0.6u
TMC(l)
TMC
TMC
TMC
' 465
TMC
TMC
TMC
385
TMC
TMC
TMC
TMC
PARTICLES
37A ENGINE
Thousands of Particles/Ft
of Undiluted
m 0.6-1. 5um
62.2
TMC
TMC
TMC
260
TMC
TMC
TMC
193
TMC
TMC
TMC
TMC
3
*
Engine Exhaust
1.5-3.0um
20.7
TMC
TMC
TMC
103
TMC
TMC
TMC
43.1
201
455
TMC
TMC
3-5um
6.9
61.7
TMC
TMC
2.5
13.5
23.4
TMC
7.2
13.4
36.2
84.4
249
>5um
1.9
6.2
29.8
86.3
1.5
4.5
4.5
38.0
4.8
2.2
2.5
6.5
16.6
Total
_-
--
--
--
832
--
--
--
633
--
--
^ M
(1) Too many particles to count, exceeds about 500,000 particles/ft.
-------
TABLE 3.2
SIZE DISTRIBUTION AND CONCENTRATION OF EXHAUST PARTICLES
AS REPORTED BY THE ROYCO PARTICLE COUNTER-JT12-6 ENGINE
IITRI NAFEC
V?T- 1
WT- 7
WT-9
77
Sampler Location,. j Engine
Engine Diameters Thrust
From Engine Nozzle j Setting
7-1/2 | Approach
7-1/2 j Cruise
7-1/2 Takeoff
Thousands of Particles/Ft.3
of Undiluted Engine Exhaust
0.3-0.6um 0.6-1.5pm 1.5-3.0|am 3- 5pm >5u.ic Tot
270 111 16.3 1.2 0.7 39
TMC(l) TMC 106 35.0 1.4
TMC TMC TMC 90.0 0.3
. x)
(1) Too many particles to count, exceeds about 500,000 particles/ft. .
-------
APPENDIX D, SAE STANDARD
METHOD FOR SMOKE MEASUREMENT
The Society of Automotive Engineers' Committee E-31
has developed a method for measuring and specifying turbine
engine smoke. Their efforts are documented in the Aerospace
Recommended Practice 1179 (Reference 1) . Much of the testing
required for the development of the method was conducted at
NAFEC. The procedure used at NAFEC (Reference 13) involves
filtering different volumes of exhaust gases and evaluating
the optical density of the resulting stain. Volumes of gas,
0.1 ft3, 0.2 ft3, 0.4 ft3, and 0.7 ft3, collected in the plane
of the engine nozzle were accurately measured with a wet test
meter. The graduated series of stains obtained were evaluated
by measuring the optical diffuse reflectance with a reflecto-
meter. Whatman No, 4 filters are specified, but membrane filters
were also used during this study to obtain comparative data.
The smoke number associated with each flow is calculated from
the following equation:
R
Smoke Number = SN = (1 -- -) 100 (4-1)
where, R = the diffuse reflectance of the smoke deposit on
the filter,
R = the diffuse reflectance of the clean filter.
Since R and R are functions of the material used to back the
f ilterssduring"7analysis , the backing material must be specified.
ARP No. 1179 specifies a black backing as the standard but a
white backing was also used during this study. The reportable
SAE smoke number is the smoke number corresponding to a gas
3 2
volume of 0.3 ft /in of Whatman No. 4 filter paper evaluated
on a black backing. It is obtained by plotting the smoke
numbers against gas volume per unit area of filter surface
3 2
(log-log plot) and interpolating the curve obtained to 0.3 ft /in
The probe for collecting exhaust gas is shown in Figure
4 .1 , installed on an F- 100 aircraft. The smoke console built
by General Electric and used to filter the smoke sample and
measure the gas volumes, is shown in Figure 4.2.
4-1
-------
.
-.
-- *n-9yt
- .,,
&-_««.
\\
-
\
' ' I I
.
I
.
a
:^. -:
.
FIGURE
F-100 AIRCRAFT WITH SAMPLING PROBES.
------- |