NBSIR 74-488
EPA-550/9-74-007
Measurements of Railroad Noise-Line
Operations, Yard Boundaries,
and Retarders
J. M. Path
D. S. Blomquist
J. M. Hemen
M. Tanca
NATIONAL BUREAU OF STANDARDS
December 1974
Joint EPA NBS Study
Approved for public release; distribution unlimited
Applied Acoustics Section
Mechanics Division
Institute for Basic Standards
National Bureau of Standards
Washington, D. C. 20234
-------
NBSIR 74-488
EPA-550/9-74-007
MEASUREMENTS OF RAILROAD NOISE - LINE
OPERATIONS, YARD BOUNDARIES,
AND RETARDERS
J. Ml. Path, D. S. Blomquist, J. M. Heinen and M. Tarica
Applied Acoustics Section
Mechanics Division
Institute for Basic Standards
National Bureau of Standards
Washington, D. C. 20234
December 1974
Final Report
Prepared for
Office of Noise Abatement and Control
U. S. Environmental Protection Agency
Washington, D. C. 20460
< V
t, *f
f '' . .^ff
*<"». U 0«
U. S. DEPARTMENT OF COMMERCE, Frederick B. Dent. Secretary
NATIONAL BUREAU OF STANDARDS. Richard W. Roberts. Director
-------
NBSIR 74-488
EPA-550/9-74-007
MEASUREMENTS OF RAILROAD NOISE - LINE
OPERATIONS, YARD BOUNDARIES,
AND RETARDERS
J. M. Path, D. S. Blomquist, J. M. Heinen and M. Tarica
Applied Acoustics Section
Mechanics Division
Institute for Basic Standards
National Bureau of Standards
Washington, D. C. 20234
December 1974
Final Report
Prepared for
Office of Noise Abatement and Control
U. S. Environmental Protection Agency
Washington, D. C. 20460
<"«. »
U. S. DEPARTMENT OF COMMERCE, Frederick B. Dent, Secretary
NATIONAL BUREAU OF STANDARDS, Richard W. Roberts. Director
-------
Abstract
A field investigation of noise emission from railroad operations vas
conducted. The objectives of the study were the establishment of a data
base on the noise levels associated vith railroad operations, both line
(trains in transit) and yard, and the development of measurement procedures
that could be utilized in regulations applicable to the noise from rail
carrier equipment and facilities. For trains in transit, measurements vere
made as a function of horizontal distance from the tracks [five locations
at 25, 50, 100, 200 and UOO feet] and as a function of microphone height
[three different heights at the 25 and 50 foot microphone locations]. Train
passby data'are presented as the maximum A-veighted sound level observed
during the passby and as Single Event Noise Exposure Levels (both A-weighted
and one-third octave band levels). A-weighted sound level measurements
were made at the boundary of the railyard, at 0.1 second intervals, for
periods of time ranging from 1 to 23 hours over several days. These data
are presented as the energy equivalent sound level and the level exceeded
ten percent of the time. The directionality of retarder noise was also
investigated. Measurements were made of the noise emitted in various direc-
tions during retarder operation.
-------
Table of Contents
1. Introduction 1
2. Field Test Program 1
2.1. Railroad Line Operations . . . . . 2
2.1.1. Field Test Site (Line) 2
2.1.2. Test Procedure (Line) . 6
2.1.3. Test Results (Line) 8
2.2. Railroad Yard Operations . . 11
2.2.1. Field Test Site (Yard) h6
2.2.2. Test Procedure (Yard) U6
2.2.3. Test Results (Yard) 50
3. Conclusions 92
lj. Acknowledgement 93
5. References 93
6. Appendix A. Data Acquisition and Analysis System for
Line Operations 9^
7- Appendix B. Procedures for Calculation of
L and SENEL 97
8. Appendix C. Data Acquisition and Analysis System for
Rail Yard Boundary Measurements 99
9. Appendix D. Data Acquisition and Analysis System for
Retarder Noise Measurements 102
-------
1. Introduction
The U. S. Environmental Protection Agency (EPA) is charged, under Sec-
tion IT of the Noise Control Act of 19T2 (Public Law 92-571*), with the dev- .
elopment of railroad noise emission standards. The law states "After con-
sultation with the Department of Transportation, EPA is required to promul-
gate regulations for surface (rail) carriers engaged in interstate commerce,
including regulations governing noise emission from the operation of equip-
ment and facilities of such carriers."
The lack of data in the public domain on the noise levels associated
with railroad operations necessitated the establishment of a substantial
data base prior to Federal rule making in this area. Through an inter-
agency agreement, EPA requested the assistance of the National Bureau of
Standards (NBS) in the establishment of such a data base. These data, in
conjunction with data from other sources [l, 2] , provide the technical
basis for the proposed EPA interstate rail carrier noise emission regulations.
2. Field Test Program
For the purpose of this report, the broad.range of noises emitted by
railroad operations has been divided into two categories line operations
(trains in transit) and yard operations.
The movement of locomotives and freight/passenger cars over main line
and local branch main line tracks is termed line operations. For trains in
transit, there exists two major .noise contributors the noise from the
locomotive, or road power unit, and the wheel/rail interaction noise which
defines the car-ge,nerated noise levels.
Railroad yard operations, on the other hand, include all operations
which are conducted within the confines of the yard property boundaries,
including the classification of freight cars and services relating to the
performance testing and routine maintenance of cars and locomotives. The
classification process the uncoupling of cars from incoming trains and
recoupling them into outgoing trains bound for various destinations is
the major yard activity. The various noise sources associated with this
operation include: (l) switcher engine noise as incoming cars are pushed
up the hump for weighing, classification and destination determination,
(2) wheel/rail and retarder noise as the speeds of the free-rolling rail
cars which have been pushed over the hump are controlled by retarders
rails which squeeze against the wheels of the moving cars as they are
guided to the outgoing train make-up area and (3) the coupling noise as
the free-rolling rail cars bump into the other (stationary) cars of the
outgoing train.
Figures in brackets indicate the literature references at the end of this
report.
-------
The noise associated with both yard and line operations was investi-
gated during this study with emphasis on the development of measurement
procedures and the establishment of a data base appropriate to railroad
operations. . .
2.1. Railroad Line Operations
The noise levels associated with trains in transit are dependent upon
the physical characteristics of the train, the operating speed, the condi-
tion of the wheels and rails, total weight and length of the train, and the
contour of the track bed. Although these variables can be determined, they
cannot be controlled in a field study such as this; therefore, it is diffi-
cult to correlate the noise levels and frequency spectra of successive train
passbys. This section presents a discussion of the field -test site and
test procedures utilized during the data acquisition phase of the railroad
line operation" study as well as a presentation of .the resultant data.
2.1.1. Field Test Site (Line)
The high speed main line of the Chesapeake and Ohio Railroad, located
adjacent to the Montgomery County Fairgrounds in Gaithersburg,, Maryland (in
close proximity to the National Bureau of Standards) was selected as the
field test site for the line operation noise study. The Fairgrounds pro-
vided a large grass-covered (mowed) open area free of any large reflecting
surfaces. Figure 1 is a contour map of the test site and surrounding area.
The roads interspersed throughout the area are dirb with the exception of
the one adjacent and parallel to the tracks which is paved. The stands
for the baseball field are open style grandstand bleachers. Immediately
south of the tracks is a fairly dense growth of weeds and brush about 2
to 3 feet thick and 6-1 feet high. Behind the brush is a large open area
that drops in elevation until it reaches .Interstate Highway 70-S which is
20 feet below the level of the track bed. '
Microphones were located along a line perpendicular to the tracks as
indicated in Figure 1. The point of intersection of track and the line of
microphones is approximately 520 feet from the nearest point of I-TO-S.
Along the microphone line, the ground elevation decreases as it recedes
from the tracks. The land was surveyed to establish the elevation of the
microphone positions relative to the track bed (see Figure 2). For the
purpose of this measurement, the track bed is defined as the top of the
wooden ties.
At this'location, two types of rails exist continuous welded rail
on the westbound tracks and Jointed rail on the eastbound track. Since
grade could be an important parameter affecting train noise, the track ele-
vation was surveyed 300 feet on either side of the intersection between the
microphone array and the track (see Figure 3). Eastbound trains go up a
slight grade as they pass the microphones while westbound trains go down
the grade.
-------
Figure 1. Field test site for trains in transit study showing microphone
locations, (££), and the location of the mobile instrumentation
van, A. Scale 1 in. = 200 ft.
3
-------
400 FT.
13.5 FT
EASTBQUND
TRACK
-6
WESTBOUND
TRACK
103 FT.
15.3 F
GROUND
LEVEL
Figure 2. Ground elevations at the microphone locations relative to the track bed of
the Chesapeake and Ohio main line at the Gaithersburg test site.
-------
3.22 FT
1 41 F
1 IN =75 FT. (HORIZONTAL)
1 IN. = 2 FT. (VERTICAL)
Figure 3. Track elevation 300 feet on either side of the intersection point of the track
and the line of microphones.
-------
2.1.2. Test Procedure (-Line)
As stated previously, the microphones -were located along a line per-
pendicular to the direction of travel of), the trains. For one series of
measurements', five microphones, located1 .tet distances of 25» 50, 100, 200
and 1*00 feet as measured from the centerline of the westbound track, formed
a horizontal array. Each microphone was mounted on a tripod and positioned
at a height of k feet above ground level. Since the ground along the micro-
phone line was not level (see Figure 2) the, line-of-sight distance between
the microphones and the tracks were slightly different from the nominal
distances cited above. Table 1 shows the angle and line-of-sight distance
for each microphone in the horizontal/array with respect to both the east-
bound and westbound tracks. ,,
Table 1 Angle (0)'and line-of-sight distances (d) for
each microphone in the horizontal array with
respect to the eastbound and westbound tracks.
(See Figure 2). ,' .
Microphone
1
' 2
3
k
j
5
West Track
0
6° 23'.
1° 9'
. 0° Ul'
1° U&«
1° 37 '
d.ft
25.2
50.0
100. 0..
s 200.1
U00.2 &
East Track
0
U'° 10'
0° 5V
- o?-36!
1° kv
£$vSir31^
d,ft
38.6
63.5
113.5-
213.6
* U13.6.
A second series of measurements were also conducted utilizing a verti-
cal, rather than a horizontal, microphone array. For these Measurements
microphones were mounted a't heights of ^, 10 and 15 feet above the ground
at horizontal distances of 25 and '50 feet as measured from the centerline
of the westbound track. /Figure k illustrates the array and shows the micro-
phone heights with respect to the track bed and ground level. The associated
table gives the angle and line-of-sight. distance for each microphone with
respect to the track b'ed. , ^- '
"f.
During both series of measurements, the microphones were connected
through coaxial cables to the tape recording and monitoring equipment housed
in the mobile instrumentation van. The van wa.s located approximately 125
feet from the westbound track and 100 fe"e%"tb'"*Ehe"'/eal5t of the line along
which the microphones were located point A :in front of the stands in
Figure 1. The data from each microphone were "recorded on one channel of a
seven-channel F.'M. tape recorder. « The recorder was manually started and
stopped upon the approach and subsequent departure of each train. Appendix
A contains a detailed discussion; o'f the instrumentation which comprised the.
data, acquisition and analysis system for line operation studies.
-------
T
O #3
13.8 FT.
FASTBOUND
TRACK
WESTBOUND
TRACK
8.8 FT.
6 #2
12 FT.
7 FT.
MICROPHONE
#1
#2
#3
#4
#5
#6
WEST TRACK
B
6°23'
19°24'
28°54'
1°9'
7°58f
13°30f
d, FT.
25.2
26.5
28.6
50.0
50.5
514
EAST TRACK
0
4°10'
12°52'
19°43'
0°54f
6°17'
10°42'
d, FT.
38.6
39.5
40.9
63.5
63.9
646
Figure 4.
TRACK
LEVEL
GROUND
LEVEL
Layout of vertical microphone array showing
angles (0) and line-of-sight distances (d)
for each microphone with respect to the
eastbound and westbound tracks.
-------
Statistical information on each train., such as, total number of cars,
loaded cars, empty cars, total weight, etc., was provided by the Chesapeake
and Ohio Railroad Company (see Tables 2 aijid 3, Section 2.1.3). Train speed
was determined by timing the train with a
marked distance of ^50 feet. Depending oiji the train length, repeated tim-
ings were obtained and the average used to compute the speed.
stop watch as it traversed a
2.1.3. Test Results (Line)
Data were obtained during 23 train passbys. For 12 of the passbys a
horizontal microphone array was utilized which consisted of five microphones
located at 25, 50, 100, 200 and HOO. feet from the centerline of the west-
bound track. All microphones in the horizontal array were mounted on tri- ,
pods at a height of U feet above the ground. Measurements of the remaining
11 train passbys were made utilizing a vertical microphone array. In this
case microphones were located at heights of 4, 10 and 15 feet above the
ground at locations of 25 and 50 feet from the centerline of the westbound
track. It was felt that the results of this data acquisition program would
provide the necessary data base (l) of the noise levels associated with trains
in transit and (2) to allow for the selection of appropriate numbers of micro-
phones as well as location and height specifications to ensure adequate
characterization of train noise.
In addition to the A-weighted sound level, the Single Event Noise Expos-
ure Level ( SENEL) was investigated as a descriptor of train passby noise.
SENEL is mathematically defined as: .
scm-io iog *Mf |i - 10 loglo \i I"" 10 L(t>/1° dtl. , , a)
O J C0 ^L'W-oo J
where p is the time-varying, mean-square-sound-pressure at the point of obser-
vation, L is the corresponding sound level, p is the standard reference pres-
sure (20 micropascals), t is the standard reference time (l second) and t is
the time (in seconds). From a practical standpoint, of course, the integration
is only carried out over a finite time interval which essentially includes all
of the acoustic energy from a given passby.' The SENEL value .is very depend-
ent on the integration time selected; errors as great as 10 dB can occur if
the time is'too short. This is especially critical as the microphone distance
from the train is increased. Considering the train as a line source, this
effect was investigated theoretically and, considering the length and speed
of the train and the microphone distance, the integration time was selected
for each train passby such that in no case was the error due to the finite .
integration time greater than 1 dB at any microphone location.
2/
The procedure by which the SENEL integral was evaluated from the analog
sound pressures is discussed in Appendix B.
-------
Also, SEWEL is easily relatable to the energy equivalent noise level (L ),
which is the level of steady state continuous noise having the same energy
as the actual time varying noise. Among the many scales used for noise and
its effect, L appears to emerge as one of the most important measures of
environmental noise effects on man [3].
The data for the 25, 50, 200 and i*00 foot microphones of the horizontal
microphone array are presented in the following tables and figures. It should
be noted that due to instrumentation failure, data were not obtained at the
100 foot microphone location. On several occasions one or more microphones
were inoperative during the train passby and therefore, data are not available
in these instances either. Table 2 presents information- on the characteris-
tics of the 12 trains which were measured. Data such as the train number
(identification number of the lead locomotive), the direction of travel, num-
ber of locomotives, number of cars and whether the cars were empty or loaded,
the total weight and length of the train and the speed of the train, are
included. The acoustic data are presented in Fig-ores 5-16. Each figure
corresponds to a particular train and is composed of two parts labeled (a)
and (b). The one^third octave band Single Event Noise Exposure Level versus
frequency data for each microphone position are presented in Figures. 5a-l6a,
while Figures 5b-l6b present the A-weighted Single Event Noise Exposure Level
and the maximum A-weighted sound level during the train passby plotted versus
the perpendicular distance from the center of the track on which the train
was running. In the upper right-hand corner of Figures 5b-l6b are shown the
average attenuation with distance (decibel/doubling of distance) of both the
SEWEL and 1^) data. ' \ .
The one-third octave band SENEL spectral data show that, as expected,
train passby noise is characterized by l
-------
those of the other microphones were determined [L(x) - L(50)H . These
differences were computed for the octave band SENEL values over the fre-
quency range from 63 to ^000 Hz. The average, values for each frequency were
plotted against the microphone distances. A straight line was fitted to
the average value data using the method of least squares. The average
values, the range, and the least squares line are presented in Figures 17
and 18. The data in these figures are separated according to direction of
travel east or west which also corresponds to differences in track
type, and grade. It should be noted that the least square lines have been
displaced and forced to go through 0 dB at 50 feet for westbound trains
and at 63 feet for eastbound trains. The average deviations [L(x) - L(50)]
are also plotted versus frequency for each microphone location as shown in
Figures 19 and- 20 for the west, and east bound trains, respectively.
As expected, there is a general tendency for an increased rate of ab-
sorption at higher frequencies. The data also seem to indicate that-des-
tructive interference is occurring in the region of 500 Hz (this phenomena
will be discussed in detail later in this section). Since the data include
both locomotives and cars for which the effective source heights, and
hence the expected rate of attenuation, are different no quantitative con-
clusions can readily be drawn.
In order to determine'the influence of microphone height as a parameter,
a vertical microphone array was utilized to measure the noise.from 11 pass-
ing trains. As stated earlier,, the vertical array consisted of six micro-
phones three at heights of U, 10 and 15 feet above the ground at a dis-
tance of 25 feet from the centerline of the westbound track and three at the
same heights at a 50 foot distance. The data obtained with the U foot:high
microphone located at 50 feet were found to be erroneous; therefore, the1
only tie with the horizontal array was the jt foot high microphone located at
25 feet and this microphone was selected as the reference microphone*
These were 11 train passbys; however, on two occasions east and west
bound trains passed the microphone array simultaneously. These, are noted on
Table 3 which presents data on the characteristics of the trains measured
utilizing the vertical microphone array.
Figures 21 and 22 present the differences in the A-weighted Single
Event Noise. Exposure Level and A-weighted sound levels that existed-between
the reference microphone and the other microphones [L(x,y) -
L(x) is the noise level measured at the microphone location whose hori-
zontal distance from the source is defined within the parenthesis, i.e.,
L(50) is the level measured at the 50 foot.microphone location.
L(x,y) is the noise level measured at the microphone location whose
horizontal distance from the source and height above the ground are .
defined within the parenthesis, i.e., L(25,10) is the level measured
at the 25 foot microphone location for a microphone height of 10 feet.
10
-------
The horizontal distances shown correspond to data for westbound trains
the distances for the eastbound trains were 13-5 feet greater.
The chief conclusion to be drawn from these data is that some care is
required in attempting to predict levels at one vertical height from measure-
ments at some other height. At a horizontal measurement distance of 50 feet,
assuming a 15 foot high locomotive and a 15 foot high microphone, ah acoustic
signal originating from the roof-top exhaust would travel about 8.5 feet
further by undergoing one reflection from the ground than it would travel
in going directly from the exhaust to the microphone. A distance of 8.5
feet corresponds to one-half wavelength for sound at a frequency of about TO
Hz. This is in the frequency range where the maximum sound pressure levels
due to the locomotive engine firing frequency occur. This observation would
suggest that the anomalously low levels at the 15 foot high microphone at a
horizontal measurement distance of 50 feet were due to destructive inter-
ference between the direct signal and that reflected from the ground.
For measurements using a microphone h feet above the ground, assuming
a hard reflecting surface, at. a distance of 100 feet from a locomotive (dis-
tances that have been suggested for regulatory purposes), a 15 foot high
source (i.e., locomotive exhaust) would result in destructive interference
at about 500, 1500, 2500, 3500,..Hz and constructive interference at about
1000, 2000, 3000, UOOO, Hz. The first frequency at which destructive
interference occurs is well above the frequency range associated with the
fundamental firing frequency of the locomotive engine. Thus one would not
expect serious measurement errors .due to interference phenomena. Similarly,
measured sound levels should be reasonably independent of small differences
in microphone height, provided the terrain is reasonably flat.and level.
However, if there were a small valley between the train and the microphone,
destructive interference could occur at frequencies near that of acoustical
radiation associated with the fundamental firing frequency of the locomotive
engine. As an example, assume the ground falls off to about 10 feet below
track level at 50 feet away and then rises to be level with the track at
100 feet away. For exhaust noise from a 15 foot high locomotive, destruc-
tive interference would occur, (for a k foot high microphone) at frequencies
of about 80, 2^0, 1+00» 560,,..Hz, Destructive interference would occur
near these same frequencies if the ground fell off, for example, to 7 feet
below tra,ck level at a distance of 50 feet and then rose to about 5 .feet
above track level at 100 feet.
2.2. Railroad Yard Operations .
The noise levels associated with a railroad yard are dependent upon a
variety of activities within the yard. The primary noise sources/ typically
are the various retarders, the coupling of cars, and the working and idling
locomotives both road and switcher. This section presents a discussion
of the field test site and test procedures utilized during the data acquisi-
tion phase of the railroad yard operation study.as well as a presentation
of the resultant data.
11
-------
Table 2. Characteristics of trains on the main line during measurements
utilizing the horizontal microphone array.
made
Train*
1*103
6607
71*11
1*051*
3823
1*036
1*51*8 .
6970.
1*031
3555
6955
5983
Direction
West
West
West
i
West
West
West
West
West
West
East
East
East
Locomotives
3
2
5
6 ,
2
1*
i*
2
3
3
2
'l
Empty
Cars
9k
0
138
0
20
0
63
32
o -'
27
16
2
Loaded
Cars
. H8
0
0
0
3'
81
13
5
' 77
' 59
8
2
Speed
Ft/sec
62
..112
33
56
30
88
' 1*3
51
87.
1*6'
>* ' :
'36 ..
Weight
Tons
7380
300
1*890
900
1016
6889
1*162
2100
581*0
1*800.
1535
500
Length
Feet
7020
136
6961*
1*08
1390
1*160
3920
1912
3900
1*332
1288
26.0
*The numbers refer to the identification numbers of the lead locomotives,
12
-------
w
w
CO
o
X
w
w
CO
H (0
§H
2! C
U «J
W rt
O o
!3 f^
H
CO 0)
^
Q
53 CQ
2 >o
PQ
I
O
O
§
H
re
2SFT MICROPHONE
© SOFT nI CROPHONE
Gl 200FT 01 CROP HONE
400FT tllCROPHONE
63 125
250 500 1000 2000 4000 8000
FREQUENCY, Hz
Figure. 5a". Single event noise exposure level versus frequency
for train no. 4103.
LOCOMOTIVES
3
EMPTY
CRRS
94
LORDED
CflRS
48
SPEED
FT/SEC)
62
WEIGHT
( TONS )
7380
LENGTH
(FEET)
7020
13
-------
20
110
100
90
o
w ,
o ' 80
w
8
,70
50
50
DISTANCE, ft
100 200
500
10
G> SENEL.-5.0 OB/00
A LttflX. -4-6 OB/00
20 50
DISTANCE, m
100
Figure 5b. Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB re 20 ^Pa and 1 sec
ond) versus distance for train no. 4103.,
-------
.4
W
W
o
On
X
w
w
w
H W
o
2! r-H
E-t TJ
25 (3
U (TJ
>
W (0
O o
2 CN
H
tt (1)
CQ
PQ
1
EH
U
O
sa
E-t
25PT MICROPHONE
O SOFT H1CR0PH0NE
20QFT MICROPHONE
ADOPT MICROPHONE
125
250 500 1000 2000 4000 8000
FREQUENCY, Hz
Figure 6a. Single event noise exposure level versus frequency
for train no. 6607.
LOCOMOTIVES
2
EMPTY
CflRS
0
LORDED
CRRS
0
SPEED
FT/SEC)
112
WEIGHT
(TONS)
300
LENGTH
(FEET)
136
15
-------
CQ
0
O
W
Q
W
20
no
100
90
80
70
60
50
DISTANCE, ft
100 200
500
GSENEL.-3.6 OB/00
A LtlRX. -4.8 OB/DO
10
20 50
DISTANCE, m
100
Figure 6b.
Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB re 20 piPa and 1 sec
ond) versus 'distance for train no. 6607.
16
-------
25FT MICROPHONE
CD SOFT MICROPHONE
CD 200FT MICROPHONE
400FT MICROPHONE
63 125
250 500 1000 2000
FREQUENCY, Hz
4000 8000
Figure la. Single event noise exposure level versus frequency
for train no. 7411.
LOCOMOTIVES
5
EMPTY
CflRS
138
LORDED
CflRS
0
SPEED
FT/SEC)
33
WEIGHT
(TONS)
4890
LENGTH
(FEET)
6964
-------
CQ
W
Q
O
Q
W
E-i
33
U
H
20
110
100
90
80
'70
60
50
DISTANCE, ft
. 100 ,
200
500
10
GSENEL.-4.8 OB/DO
A LMflX, -3.7 OB/00
20 '50
DISTANCE, m
100
Figure 7b. Maximum A-weighted sound level (in dB re 20 j
and A-weighted SENEL (in dB re 20 uPa and 1 sec-
ond) versus distance for train no. 7411.
18
-------
ri
25FT MICROPHONE
CD SOFT MICROPHONE
CD 200FT MICROPHONE
400FT MICROPHONE
63 125
250 500 1000 2000 4000 8000
FREQUENCY, Hz
Figure 8a. Single event noise exposure level versus frequency
for train no. 4054.
LOCOMOTIVES
6
EMPTY
CflRS
0
LORDED
: CflRS
0
SPEED
FT/SEC)
56
WEIGHT
(TONS)
900
LENGTH
(FEET)
408
3-9
-------
Q
2
D
O
W
Q
W
EH
SB
0
20
110
100
90
60
70
,60'
50
DISTANCE, ft
100 .. . 200
500
SENEL.-4.8 OB/00
-5.6 OB/00
10
20 50
DISTANCE, m
100
Figure 8b. Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB re 20 juPa and 1 sec-
ond) versus distance for train no. 4054.
20
-------
w
w
g
CO
o
0*
X
w
w
CO
H W
25 e
flj
w m
Pu
o o
3
m
I
8
H
O 25FT M1CR0PH0NE
SOFT M1CR0PH6NE
CP 200FT niCR0PH0NE
400FT MICROPHONE
63 125 250 500 1000 2000 4000 8000
FREQUENCY, Hz
Figure 9a. Single event noise exposure level versus frequency
for train no. 3823.
LOCOMOTIVES
2
EJ1PTY
CflRS
20
LORDED
CflRS
3
SPEED
FT/SEC)
30
WEIGHT
(TONS)
1016
LENGTH
(FEET)
1390
.21
-------
O
W
a
e>
H
20
no
100
90
80
70
DISTANCE, ft
50 100 200
500
10
SENEI.-4.4
LflflX. -4.9
OB/00
OB/00
20 50
DISTANCE, m
100
Figure 9b.
Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB r,e 20 juPa and 1 sec
ond) versus distance for train no. 3823.
-------
w
W
pa
w
H W
o
-------
Q
O
Q
W
EH
a
20
110
100
90
80
70
60
50
DISTANCE/ ft
100 200
500
10
OSENEL,-4.3 OB/00
. -4.2 OB/00
L L i^
20 50
. .DISTANCE, m
100
Figure lOb.
Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB re 20 ^Pa and 1 sec
ond) versus distance for train no. 4036.
2k
-------
25FT MICROPHONE
(D SOFT MICROPHONE
CD 200FT MICROPHONE
400FT MICROPHONE
63 125 250 500 1000 2000 4000 8000
FREQUENCY, Hz
Figure lla. Single event noise exposure level versus frequency
for train no. 4548.
LOCOMOTIVES
4
EMPTY
CflRS
63
LOflOEO
CflRS
13
SPEED
FT/SEC)
43
WEIGHT
(TONS)
4162
LENGTH
(FEET)
3920
-------
CO
T)
o
W
Q
w
20
110
100
90
80
70
60
SO
DISTANCE, ft
, 100 200
500
10
SENEL.-3.7
LHflX. -4.0
OB/00
08/00
20 50
DISTANCE, m
100
Figure lib.
Maximum A-weighted sound level (in dB re 20 pi Pa)
and A-weighted SENEL (in dB re 20 juPa and 1 sec-
ond) versus distance for train no. 4548.
26
-------
O SOFT niCROPHONE
CD 200FT J1ICR0PHONE
400FT niCRQPHONE
125
250 500 1000 2000 4000 8000
FREQUENCY, Hz
Figure 12a. Single event noise exposure level versus frequency
for train no. 6970.
LOCOMOTIVES
2
EMPTY
CRRS
32
LORDED
CflRS
5
SPEED
[ FT/SEC )
51
WEIGHT
(TONS)
2100
LENGTH
(FEET)
1912
27
-------
w
>
w
j
Q
3
D
O
Q
W
EH
V
O
H
20
110
50
DISTANCE, ft
. . 100 200
500
100
90
80
70
60
SENEL.-4.2
LPIflX. -4.5
OB/00
08/00
10
20 SO
DISTANCE, m
100
Figure 12b.
Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB re- 20 piPa and 1 sec
ond) versus distance for -train no. 6970.
28
-------
130
25FT MICROPHONE
CD SOFT MICROPHONE
400PT MICROPHONE
125
250 500 1000 2000 4000 8000
FREQUENCY, Hz
Figure 13a, Single event noise exposure level versus frequency
for train no. 4031.
LOCOMOTIVES
3
EMPTY
CRRS
0
LORDED
CRRS
77
SPEED
FT/SEC )
87
WEIGHT
(TONS)
5840
LENGTH
(FEET)
3900
29
-------
M
O
%
J
£
M
i-q
Q
O
CO
Q
W
E-i
S3
O
20
no
100
90
80
70
60
50
DISTANCE, ft
100
200
500
SENEL.-4.1
LtlflX, -4.3
OB/00
06/00
10
20 50
DISTANCE, m
100
Figure 13b. Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB re 20 pPa and 1 sec-
ond) versus distance for train no. 4031.
30
-------
i4
£
W
W
H>
o
X
w
w
en
H w
o
EH T3
I8
W rt
s*
O O
2 CN
H
cn a)
Q
CQ
CQ
U
O
Q
a;
H
sc
EH
130
120
O 25FT MICROPHONE
O SOFT MICROPHONE
CD 200FT MICROPHONE
4QOFT MICROPHONE
125 250 500 1000 2000 4000 8000
FREQUENCY, Hg
Figure 14a. Single event noise exposure level versus frequency
for train no. 3555.
LOCOMOTIVES
3
EMPTY
CfllS
27
LOflDEO
CflRS
59
SPEED
FT/SEC)
46
WEIGHT
(TONS)
4800
LENGTH
(FEET)
4332
31
-------
CO
tJ
A
s
w
D
o
Q
W
^
a
o
H
20
110
100
.90
80
70
60
50
.DISTANCE, ft
100
200
500
O SENEL.-3.6
A LflRX. -3.8
OB/00
OB/00
10
20 50
DISTANCE, m
100
Figure 14b.
Maximuin,-A-weigh ted sound level (in dB re 20
and A-weighted SENEL (in dB re 20 /*Pa and 1 sec
ond) versus distance for train no. 3555.
32
-------
25FT MICR0PH0NE
CD SOFT MICROPHONE
CD 200FT MICROPHONE
400FT MICROPHONE
63
Z50 500 1000 2000
FREQUENCY, Hz
. . .
4000 8000
Figure 15^. Single event noise exposure level versus frequency
for train no. $955.
LOCOMOTIVES
2
EMPTY
CflRS
16
LOflOEO
CflRS
8
SPEED
FT/SEC)
49
WEIGHT
(TONS)
X535
LENGTH
(FEET)
1288
33
-------
ffl
Q
Z
8
Q
W
EH
S3
O
20
no
DISTANCE, ft
50 100
200
500
100
90
80
70
60
0SENEL.'-4.1 OB/00
_4>5 OB/00
10
20 50
DISTANCE, m
100
Figure 15b.
Maximum A-weighted sound level (in dB re 20
and A-weighted SENEL (in dB re 20 pPa and 1
ond) versus distance for train no. 6955.
sec-
-------
1-1
g
w
w
CO
O
04
X
w
w
CO
H
O
a
a c
W «J
W fll
cu
H
CO
Q
03
CQ
a
EH
CJ
O
H
35
E-"
130
120
110
100
<» 25PT MICR0FH0NE
O SOFT niCR0PH0NE
m 200PT MICROPHONE
A 400PT MICROPHONE
63 125 250 500 1000 2000* 4000 801
8000
FREQUENCY, Hz
Figure 16a. Single event noise exposure level versus frequency
for train no, 5983,
L0C0n.0TIVES
1
EMPTY
CflRS
2
LOflOEO
CflRS
2
SPEED
FT/SEC)
36
WEIGHT
(TONS)
500
LENGTH
(FEET)
260
35
-------
CQ
Q
O
CO
a
w
20-
110
100
90
80
70
60
10
50
DISTANCE, ft
100 200
500
0 SENEL. -3.5 08/00
-5.2 08/00
20 50
DISTANCE, m
100
Figure 16b.
Maximum A-weighted sound level (in dB re 20 juPa)
and A-weighted SENEL (in dB re 20 fiPa and 1 sec-
ond) versus distance for train no. 5983.
36
-------
CQ
Ti
w
03
O
0<
X
w
w
w
H
O
E-i
a
w
>
w
w
tj
o
&
H
CD
CQ
1
O
O
20
10
0
-10
10
0
10
10
0
10
10
0
'10
10
DISTANCE, ft
50 100
200
63 H?
-2.2 OB/00
125 HZ
-2.2 06/00
250 HZ
-3.3 08/00
I
500 HZ
-5.1 OB/00
20 50
DISTANCE, m
100
500
Figure 17a,
Octave Band SENEL [L(x) - L(50)] versus distance for
west bound trains.
37
-------
o
in
8
W
W
W
H
O
w
w
o
H
Q
PQ
10
0
-10
10
0
-10
10
0
10
10
*
0
10
Figure 17b.
DISTANCE, ft
50 100
200
500
10
1000 HZ
-4-8 08/00
2000 HZ
-*-3 08/00
4000 HZ
-4-° 08/00,
fl-WT
-4.3 OB/00
20 50
DISTANCE, m
100
Octave band SENEL [L(x) - L(50)l versus distance for
west bound trains.
-------
CQ
o
in
w
yj
H
O
2
g
W
>
w
w
2
H
w
t;
o
20
10
0
-10
- 10
w o
w ~IQ
o
X
111
10
i
0
-10
10
0
-10
DISTANCE, ft
50 100
200
500
63 HZ
-2.7 OB/00
125 HZ
-1.8 08/00
250 HZ
-3.2 OB/00
500 HZ
-4.6 OB/00
10
-4-
20 50
DISTANCE, m
100
Figure I8a. Octave band SENEL [L(x)
east bound trains.
39
- L(50)] versus distance for
-------
CQ
TJ
o
in
w
>
a
W
W
H
O
S3
E-i
W
g
w
H
Q
I
u
o
10
0
-10
10
0
10
10
0
10
10
0
10
. DISTANCE, ft
50 100 200
500
1000 HZ
-3.8 08/00
2000 HZ
-3.5 OB/00
4000 HZ
-3.6 05/00
fl-WT
3.7 OB/00
10
20 50
DISTANCE, m
100
Figure 18b. Octave band SENEL [L(x) - L(50)] versus distance for
east bound trains.
-------
i-q
W
>
W
CO
o
CM
X
w
W CO
CO -O
H
o >
> J
W
I
W
^ -»
o x
*^ t»-^
H ,J1
CO
Q
I
u
o
0
-5
-10
-15
63 125 250 500 1000 2000 4000
FREQUENCY, Hz :
Figure 19. Octave band SENEL [L(x) - L(50)]
versus frequency for west bound
trains.
Ui
-------
i-q
D
W
o
w
W OQ
w -o
H
O -
EH O
13 m
W ^-
> ^
W
I
O
^
H
Q
I
O
O
-5
C -10
-15
63 125 250 500 1000 2000 4000
FREQUENCY, Hz
Figure 20. Octave band SENEL [L(x) - L(50)]
versus frequency for east boun4
trains. .
-------
Table 3. Characteristics of trains on the main line during measurements made
utilizing the vertical microphone array.
Train*
3692
69614
61*93
1*100
Iil57**
|_1*108**
6955**
696!+**
9910***
11*56***
9911***
Direction
West
West
East
East
East
West
East
West
East
East
East
Locomotives
1*
2
. I
2
3
3
2
2
It
1
. 1*
Empty
Cars
TU
0
2
313
0
0
u
0
0
3
0
Loaded
Cars
1*6
0
2
0
118
68
29
0
0
0
0
Speed
Ft/sec
39
1*1*
30
ia
1*1*
52
33
1*1*
87
107
88
Weight
Tons
6682
300
500
1*230
7730
1*730
2900
300
236
239
236
Length
Feet
. 6032
136
260
. 61*21+
5868
31*68
1720
136
3l*9
3^9
31*9
*The numbers refer to the identification numbers of the lead locomotives.
**Simultaneous Passby
***Commuter Trains
-------
OQ
if)
CM
O
o
X
0
p
CO
o
w
w
CO
H
O
25
H
W
55
H
CO
Q
W
EH
a
c
-5
It
[]
0
-10
25/10
25/15
50/10
50/15
D WEST BOUND
O ERST BOUND
A WEST/ERST BOUND
O ERST BOUND COMMUTER
MICROPHONE LOCATIONS
horizontal distance/height
Figure 21.
A-weighted SENEL [L(x,y)
microphone locations.
- L(25,4)] for various
-------
in
CN
J
s
w
Q
O
Q
EH
I
H
X
-5
-10
-15
o
25/10
25/15
50/10
50/15
WEST BOUND
EflST BOUND
WEST/EHST BOUND
ERST BOUND COMMUTER
MICROPHONE LOCATIONS
horizontal distance/height
Figure 22. Maximum A-weighted sound level [L(x,y) - L(25,4)] for
various microphone locations.
-------
2.2.1. Field Test Site (Yard)
Rail yard measurements were made at the Norfolk and Western Railroad
Terminal located in'Roanoke, Virginia1. The Roanoke terminal is the eastern
hub of the Norfolk and Western Railway system'and as such is operated on a
2^-hour, T days per week basis. "The following statistics are presented
to provide an indication.of the size and activity of the facility:
There is an average of ^500 cars handled daily through the terminal,
with peak loads near 6500 cars. An average of over 85 trains arrive
and depart Roanoke on a daily basis.
Approximately 2100 cars each day are classified over the dual hump
(master retarders).
The classification yard contains 55 classification tracks with a
capacity of approximately 1950 cars.
The receiving yard contains 20 tracks with a capacity of approximately
2000 cars. . .
The hump computer controls 2 master, 2 intermediate, and 9 group
retarders, and 65 switches.
The terminal'contains 228 miles Of track. . .
A reduced reproduction' of a detailed map of the Roanoke Yards is shown
in Figure 23. Superimposed on the map are the microphone positions at the
yard perimeter (locations Al, A2, -Bl and B2) and within the yard (location
C) which were utilized for rail yard measurements.
2.2.2. Test Procedure (Yard)
Measurements were made at four'locations (designated Al, A2, Bl and B2)
along the boundary of the Roanoke train yard and at one location (designated
C) within the yard. . . .
Microphone positions Al and A2 were selected because of their proximity
to the intermediate and group retarders and the car coupling area respectively.
Figure 2k shows an overview of this area with microphone number Al in the
foreground. The two microphones at location A were mounted on tripods at a
height of 5 feet above the ground. They were located at the edge of an em-
bankment which was approximately 50 to 60 feet above the level.of the track
bed of the nearest track. The line-of-sight distances from the microphones
to the edge of the nearest track were 65 and 8l feet for locations. Al and A2 .
respectively.
At microphone positions Bl and B2 (see Figure 25) the microphones were
also mounted 5 feet above the ground at the edge of: an embankment. At this
1*6
-------
" =400f
PROFILE
Figure 23. Layout of the Norfolk and Western Railroad Terminal, Roanoke, Virginia, showing
microphone locations, designated Al, A2, Bl, B2, and C, for rail yard noise
measurements.
-------
00
Figure 24. Overview of the retarder and car coupling areas of the Roanoke train yard showing
microphone position Al.
-------
Figure 25. Overview of the engine idling area of the Roanoke train yard showing the location of microphone
position Bl.
-------
location the embankment was 35 to 1*5 feet above the level of the nearest
track. The noise levels measured at this location were dominated by the
noise from stationary (idling) and moving locomotives.
In order-to gain"a better understanding of the noise levels and the
directionality patterns of the noise associated with retarders, measure-
ments were also made within the yard interior. This location/is designated
location C. Figure 26. shows the eight microphone positions utilized. Re-
tarder number 1 was selected as the primary source to be studied; therefore,
all positions selected are in relation to this retarder. The noise from
retarder number 2 was also measured and dimensions relative to this retard-
er are also given. Microphone position 1 was established, on a line perpen-
dicular to retarder number 1 and 50 foot from the centerline (both longitu-
inal and lateral) of this retarder. Microphone heights of 5, 10 and 15 feet
above the ground were utilized at microphone position number 1. A line was
then drawn through microphone position number 1 parallel to the long axis of
retarder number 1. Microphone positions 2 and 3 were located along this line
at an angle of 30° and 1*5° respectively, in relation to the line from micro-
phone number 1 perpendicular to retarder number 1. At these two positions,
microphone heights of 5 and 15 feet were used. A final position, number H,
was located at an angle of 75° but as close to retarder number 1 as possible
rather than along the line of microphone positions 1,2 and 3. At location
k a single microphone height of 5 feet, was utilized. Since various loca-
tions and heights were utilized, it was determined that one way to keep
track of the positions would be to designate each with an angle and a height
which would uniquely, define each measurement position. For example, the
microphone at the five foot height at position 1 was designated (0°, 5 ft.).
For each measurement two microphones were.utilized. One of the micro-
phones was always at the reference position location 0°, 5 ft? while
the other microphone was successively placed at the other seven.test positions
as indicated in the table of Figure 26. Figures 27 and 28 show the reference
microphone (0°, 5 feet) and a test microphone (0°, 10 feet) from two differ-
ent perspectives showing the area in and around the retarder locations.
2.2.3. Test Results (Yard) .
A-weighted sound level measurements were made at the boundary of the
rail yard at 0.1 second intervals utilizing a. mini-computer-based digital
data acquisition system (described in Appendix C). Data were taken for per-
iods of time ranging from 1 to 23 hours over a 7 day period. , ..
Data at positions Al and A2 were taken from 1200 hours of the lit5th day .
of 1973 until 1000 hours of the 150th day of 1973. At positions Bl and B2,
data were taken from 1100.hours of the 150th day of 1973 until 1000 hours
of the 151st day. The data resulting from these measurements are presented
in this section in the form of the A-weighted sound levels exceeded ten
percent of the time (L->Q) and-the energy equivalent A-weighted sound levels
(L ), both plotted as functions of time. .
50
-------
MICROPHONE LOCATIONS
SCALE:
. = 25 FT.
LOCATION
#1
#2
#3
#4
8
0°
30°
45°
75°
HEIGHT, FT.
5. 10, 15
5, 15
5. 15
5
Figure 26. Measurement location C. Microphone locations near
retarders. 0 is defined .as the angle between lines
drawn from microphone number 1 and any other microphone
through the intersection of the longitudinal center-
line of retarder number 1 with the perpendicular line
drawn from microphone number 1 to retarder number 1.
51
-------
Figure 27. . Measurement position C. Retarder number 1 is to the left, retarder
number 2 is to the right with the master retarder and hump in the
background. The two microphones in this photograph are the reference
, microphone (0 = 0°, height = 5 feet) and one of the seven test
microphones (9 = 0°, height =10 feet).
52-
-------
Figure 28. Meapurement position C. View of retarder number I. The two
microphones in this photograph are the reference microphone
(6.» 0°, height =5 feet) and one of the seven test micro-
phones (9 = 0°, height = 1,0 feet).
.53
-------
The energy equivalent noise level y L ', for a stated period of time is
the level of a constant, or steady state, noise which has an amount of acous-
tic energy equivalent to that contained in the measured time-varying noise.
L is mathematically defined as: .
L =10
eq
where p is the time-varying, mean-square sound pressure at the point.of
observation, L is the corresponding sound level, p is the standard refer-
ence pressure (20 micropascals) and T is the period of integration. **
The A-wei'ghted L-0 sound levels at locations Al and A2 are presented
in Figures 29a-3^a for days lit 5 - 150. , The hourly equivalent A-weighted
sound levels, i.e., L for each hour, for locations Al and A2 for days
lit5 - 150 are present!! in the complementary Figures 29b - 3^b. Similar
data for measurement locations Bl and B2 are presented in Figures 35a - 36b.
The data points on .these plots represent the cumulative noise level during .
the preceding hour; that is, the data point at 1200 hours represents the
noise which occurred between 1100 and 1200 hours. The lack of data at cer-
tain hours on these plots is due either to'inclement weather or electrical
power failures or power fluctuations which affected the data acquisition
system at the field test site. . .
To provide some indication of the,correlation between specific activity
within the yard.and the L and L sound levels measured at the boundary of
the railyard, retarder activity from 1600 hours on day 1^9 until 0900 hours
on day 150 has been summarized in Table h. Since measurement locations Al
and A2 along the railyard boundary were in the vicinity of the active, retard-
ers, the operations can .easily be compared with the corresponding values of
L,^. and L A-weighted sound levels for these days as shown in Figures 33a,
33&, 3*ta and
The data contained on plots 29a T 36b are compressed into two. summary
plots (Figures 37 and 38) which show the L and L A-weighted sound levels
for the total time period (days 145-151). Note tha£ the data prior to 1100
hours of day 150 was for location A .while data after this time was for loca-
tion B.
It is important at this time, on the basis of these data, to examine
the relationship between LIQ and L and evaluate the appropriateness of the
two measures as descriptors of the noise emanating from railroad yards.
Figures 39 arid Uo show plots of L versus L at microphone positions Al '
2- The procedure by which the L integral was evaluated from the digital
data is discussed in Appendixes.
-------
DAY OF YEAR, 145
o
CM
JH
OQ
*.
J
.£'
W
Q
O
cn
o
iH
^
Q
W
EH
ac
90
80
70
60
50.
O POSITION Al
A POSITION A2
200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 29a. A-weighted L10 sound level versus time.
-------
DAY OF YEAR, 145
Ul
W
O
CO
Q Hi
W 04
O O
M
-------
DAY OF YEAR, 146
(0
PH
O
(N
-------
DAY OF YEAR, 146
CD
W
Q
S3
s
w cu
O o
M tN
is
H
a
x
3
90
80
70
60
50.
O POSITION Al
A POSITION A2
200 400 600
800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 30b. Hourly equivalent A-weighted sound level versus time.
-------
Ul
VO
o
0)
CQ
O
*.
W
w
Q
D
O
Q
m
EH
a
o
90
80
70
60
50,
DAY OF- YEAR, 147
QQ
O P6SITION Al
A POSITION A2
200 4OO 600
800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 3la. A-weighted L,Q sound level versus time.
-------
o
W
O
CO
Q nJ
W PL,
O o
H
-------
DAY OF YEAR, 148
o
CN
w
Q
O
CO
o
rH
^
Q
w
E-i
O
H
90
80
70
60
50,
O PQSITI0N Al
A POSITION A2
200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 32a. A-weighted LIQ sound level versus time.
-------
DAY OF YEAR, 148
i-q
£
w
8
CO
Q id
W 0i
EH i
W
O o
H H
I
Q
X
90
80
70
60
50.
CD PQSITION Al
A. P0SITION A2
200 400
600 800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 32b. Hourly equivalent A-weighted sound level versus time.
-------
DAY OF YEAR, 149
-------
DAY OF YEAR, 149
o
to
Q «0
W &
EH S.
go
M 01
iZ (1)
13
W
a
w
90
80
70
60
SO.
CD PQSITI8N Al
A POSITION A2
200 400 600
800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 33b. Hourly equivalent A-weighted sound level versus time,
-------
ON
o
(N
0)
w
Q
O
o
r-l
A
Q
W
O
I .
90
70
60
SO,
DAY OF YEAR, 150
O POSITION Al
A POSITION A2
200 400 600 . 800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 34a. A-weighted L,Q sound level versus time.
-------
DAY OF YEAR, 150
^
e
PL)
O
CO
Q
W
E-i
O o
M CM
3 (U
W
CD
a
w
D
O
ffi
90
80
70
60
50,
O POSITI6N Al
A POSITION A2
200 400 600
800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 34b. Hourly equivalent A-weighted sound level versus time.
-------
DAY OF YEAR, 150
(0
o
01
0)
ffl
Tt
A
'£
w
o
CO
o
rH
^
Q
W
E-i
O
90
80
70
60
50.
©POSITION Bl
A POSITION B2
200 400 600
800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 35a. A-weighted LIQ sound level versus time.
-------
DAY OF YEAR, 150
oo
W
O
05
Q «J
W CM
^ i
a
0,0
I
O POSITION Bl
POSITION B2
~0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 35b. Hourly equivalent A-weighted sound level versus time.
-------
o
-------
w
>
w
8
CO
Q 03
W
O o
a
>H
£
D
O
90
80
70
60
sa
DAY OF YEAR, 151
O P8SIT16N Bl
A P8SITI8N B2
200 400 600
800 1000 1200 1400 1600 1800 2000 2200 2400
TIME, hrs
Figure 36b. Hourly equivalent A-weighted sound level versus time.
-------
Tatle If. Suryr^ry of retarder activity from 1600 hours on day
ll*9 until 0900 hours on day 150.
Day
Ik?
150
Retarder
Start
161$
1705
1756
181*0
1936
2019
2100
2131
2223
2323
0010
0100
Oll*8
0211*
0310
0333
pi*o6
01*36
051*5
0627
0816
0835
Operations
Finish
1650
. 1736
1808
iaui
195^
20U5
2127
2139
2300
2331
0050
0122
0207
0233
0321
OUl3
01*23
01*5**
0607
0635
0829
0913
Empty
Cars
90
35
31
*
19 .
33
Ik
9
27
2
36
3
20
38
9
101
25
1*9
23
0
9
1*1
Loaded
Cars
0
35
51
«
2k
31
23
9
25
18
19
50
,22
21
25
0
8
0
60
18
21
51
*Unknown
71
-------
10
(0
CM
0)
(-1
DQ
O
w
w
a
D
O
W
o
1-1
^
d
W
O
95
85
75
65
55
95
85
75
65
POSITION 2
55
POSITION 1
1200
145
\
/
I
1200
146
I I
r
Location A
I
I
1200 . ~^,- i200
147 148
HOUR
DAY OF YEAR
1200
149
Location B
1 200
150
1200
151
Figure 37. A-weighted L,Q sound level versus time.
-------
o
r>i
(U
ffl
13
W
a
j
Q
2
D
O
W
Q
«
&H
K
O
g
H
D
a
w
95
85
75
65
55
95
85
75
65
POSITION 2
55
\
POSITION 1
1200
145
I
_l
1200
146
_L
V
I
1200 1200
147 148
HOUR
DAY OF YEAR
II
1200
149
1200
150
1200
1S1
Figure 38. Hourly equivalent A-weighted sound level versus time.
-------
m
(^
o
CVJ
0)
CQ
13
i-q
Q
D
O
CO
Q
13
K
O
80
70
O
X
POSITION fll
OflY 149
©
©
©
©
©
©
©
©
©
©
© ©
© ©
©
©
©
70
80
90
HOURLY EQUIVALENT A-WEIGHTED SOUND LEVEL, dB re 20
Figure 39. Interrelationship between hourly LIQ and L
eq
-------
o
0)
o
^
J
W
Q
O
CO
Q
w
&H
33
O
80
70
D
O
35
POSITION R2
OflY 149
-------
and A2 respectively for measurements made on day 1^9. As can easily "be
seen, the correlation between the two descriptors is not very good in this
case. The probable reason for the large differences as much as 20 dB
between L.. and L is the fact that short duration, high level noises which
do not occur moree£han 10 percent of. the time have a significant influence
on the value of L but absolutely no influence on the value of LIQ. A
review of the.rawe&ata confirmed this to be the case for railyard measurements
taken during this study. For those hours where the differences between L.Q
and L were the largest, the raw data showed that high level noises occurred
for nlirly 10 percent of the time and therefore, the L value tended to be
much higher than the Llf. value (which was not influenced at all by the high
level noises since they did not occur more than 10 percent of the time dur-
ing the hour.of interest). What these data show is that the nature of the
activities within a railyard are such that L.. is a poor descriptor of the
noise in this case. Similar data have been reported [**] for L... and L
data at sites near airports. Differences between the two descriptors were
as much as 20 dB over a major portion of the day at a residential site
under the landing path of Los Angeles International Airport while for a
suburban residential site, comparable L.. n and L values were reported over
a 2k hour day (typical differences on the ordere8f a few decibels or less).
L is a relatively simple descriptor but it should be utilized with caution,
especially in situations where high level sounds occur for short periods of
time.
However, L is not without problems either. It has been previously
reported [l], inn;he case of railroad yard boundary measurements that:
"In general, the 10 minute sample times utilized for this survey [Wyle
survey] were of insufficient duration for accurate measurement of the yard
activities, indicating that due to the random nature of most yard opera-
tions, 2k hour continuous recordings would most likely be required."
In order to investigate this problem further, the data for the. time
period from OOUl hours to 02^0 hpurs of day 150 were selected for investi-
gation as to the variation one could expect in the values of L as a
result of the integration time selected. The results (for measurement posi-
tions Al and A2) are presented in Figures 1*1 - ^5 for integration times of
1, 3, 10, 30 and 60 minutes. Since the A-weighted sound levels were digit-
ally recorded every 0.1 second, this corresponds to 600, 1800, 6000, 18,000
and 36,000 samples, respectively. The values of L plotted in these fig-
ures correspond to the beginning point of the integration time; that is, in
Figure k2 the data plotted at kO minutes represents the L value for.the
period extending from the beginning of the kOth minute toe£he beginning of
the U3rd minute. Since only.a two-hour data sample was used, there are no
L values plotted over the last period of integration of the two-hour period.
On the basis of these data, it would appear that regulation of rail yard
noise emission levels, in terms of L , would require, at a minimum, continu-
ous monitoring for each of several representative hours in a given day. Con-
-------
100
INTEGRATION TIME = 1 min.
40
50 60 70 80
TIME, min
90
100
110
120
Figure 41. Equivalent A-weighted sound level versus time,
-------
100
INTEGRATION TIME = 3 min.
CD
w
>
w
f-3
.
D
O
o a*
W 3.
H
0)
w
a
>
a
w
93
87
80
73
67
Position.Al
60,.
10 20
30 40
50 60 70
TIME; min
80 90 100 110 120
Figure 42. Equivalent A-weighted sound level versus time.
-------
100
> 93
87
Q0i
w a.
£-
35 O
O '67
D
INTEGRATION TIME = 10 min.
10 20 30
40 50 60
TIMEf min
80 90 100 110 12C
Figure 43. Equivalent A-weighted sound level versus time.
-------
00
o
100
A
g 93
W
a
O
en
a
87
S 80
a)
73
> 67
a
60,
INTEGRATION TIME =30 min.
Position Al
Position A2
10 20 30 40
50 60 .70
TIME, min
80 90 100 110 120
Figure 44. Equivalent A-weighted sound level versus time.
-------
100
INTEGRATION TIME = 60 min.
w
93
00
§
D
O
CO
Q
m
SS' 80
W (1)
I
<< 0)
'O
EH
25
W
73
\
Position Al
""- Position A2
67
D
CH
60j
10 20 30
40 50 60 70
TIME, min
80 90 100 110 120
Figure 45. Equivalent A-weighted sound level versus time.
-------
tinuous monitoring for 2i* hours on representative days would, of course., yield
more reliable results. On the other.hand, if it is desired to relate varia-
tions in L with specific variations in yard activity, integration times in
the range 8? 1 to 10 minutes would be preferred.
As discussed previously in Section 2.2.2., measurements were also made
within the rail yard near the active retarders to investigate the characteris-
tic retarder noise levels and directionality both in the horizontal and
vertical planes. A reference microphone was utilized'in conjunction with
seven different test microphones. Each test microphone was located at a
different height/angle combination in relation to the location of the refer-
ence microphone (see Figure 26 for microphone locations with respect to
retarders 1 and 2).
Data were obtained for 58 passes through retarder number 1 and 37 passes
through retarder number 2. The data for each train car passing through the
retarders are presented in the form of (l) the differences in the A-weighted
sound level between the test and the reference microphones and (2) the maxi-
mum A-weighted sound levels at the reference microphone. For selected passes
through retarder number 1, one-third octave band spectral analysis was also
performed and the data are presented here.
The/-differences between the maximum A-weighted sound levels at the test
[L(fr,x)] and reference [(L(0, 5)] microphones are presented graphically-in
Figures U6 and 1*7- In addition, the data are tabulated in Tables 5 and 6*
The level differences are coupled with an identification of the type, of rail
car passing through the retarder at the time of the measurement. No informa-
tion was obtained'as .to whether the numbers and types of cars sampled during
this study were representative of the long-term operational statistics for
this particular rail yard. An indepth study.of retarder squeal (which was
not the intent of this study) would of necessity have to investigate such
facts as car age, condition and type of wheels, car weight, environmental
conditions, speed, etc. . . .;
In summary, the average sound level differences between the test and ref-
erence microphone locations as well as the standard deviations are presented
in Table 7 Note that in the case of position (75,5) the data are shown for
retarder number 1 and number 2 seperately rather than combined. The data were
plotted in this manner since this location is much closer to retarder number 1
than it is to 'retarder number 2. At all other measurement locations, the
microphone is approximately equidistant from the two retarders."
L(0,x) is the noise level measured at the microphone location whose
height above the ground and angular location with respect to the per-
pendicular line drawn from microphone number 1 (see Figure 26) to the
longitudinal centerline of.retarder number 1 are defined within the
parenthesis, i.e^, L(0,5) is the level measured at an angle,'0, of
0° and a microphone height, x, of 5 feet.
82
-------
C*
22
20
tt
^
In 16
o
J 14
l
~ 12
X
.. t
5 10
^ 8
w
W e
T D
M
Q
Z 4
D
0
W
2
w
EH
tc n
u
| -2
i
S -4
>< -6
-8
-10
-
- -L-:-ri^
.
C
f
{
i
(
.
c
. -;_.
^ww ^^^~
«f^*» ^""^T
t
)
I^WKV '1
i©
) C
\
V
)
) <
)
I^HH^ 4*«
)
\
>
~ ~
(
T
) ^
(
(
'"1 !' ' -
r "
)
*?
c
^
_Y
V
(
)
(
(
(
(
(
F i 1.^11
^_
)
5
>o
)
)O __j
)
_ ^J
(
_J
t
\
,. «_^^v
(
C
_c
(
.^K^ - .-
> 1
) (
T^
\rr\
t\-J
.(
C
t
t
(
(
^**
)
)
)
1
)
)
)
^~
D
3
0/10 0/15 30/5 30/15 45/5 45/15 75/5
Figure 46.
MICROPHONE LOCATIONS (angle/height)
Maximum A-^weighted sound level [L(0,x) - L(0,5)] for
various microphone locations, retarder no. 1.
83
-------
24
22
20
ffl
0
18
i "i
? 16
o
^ 14
i
9 12
J '°
M
" 6
Q
1 4
Q 2
w
g 0
H
i -2
MAXIMUM
i
»- i i i
o co en ^
-12
-
1
C
__
c
(
I
1
c
m^mitm *)
)
P" :
^IBB^ «_-w*i
)
)
)
j
J
) (
C
MHMB «H^M
^^^ «H^
(
.^MBMt _ m^ff^m
C
(
) (
(
^MOBB ^ ^^^
(
^HBM> ^>M«
_1HIBM «
)tD .
B^W ^«I«B
)
5
^
^^i* ««
)
r
. c
_J
^^M «^B^
«H^^
(
. (
(
0/10 0/15 30/5 30/15 45/5 45/15
MICROPHONE LOCATIONS (angle/height)
75/5
Figure 47. Maximum A-weighted sound level [L(6,x) - L(0,5)] for
various microphone locations, retarder no. 2.
8U
-------
Table 5. The difference in maximum A-weighted sound level [L(0,x) - L(0,5)] between the
test and reference microphones for 58. passes through retarder number 1.
TYPE CAR
0°,10 Ft TYPE CAR 0°,15 Ft TYPE CAR 30°.5 Ft TYPE CAR .30°,15 Ft
targe Tank
Box
Box
Gondola
lit. »t
11.6
5.0
1.8
Box
Box
Box
Box & Tank
Box
Box
Box
Tank
2.0
-3.0
2.2
-2.0
3.8
7. It
5-8
-5.6
Flat (2)
Box
Box(2)
Box
Box
-8.lt
7-6
-l.it
3.0
-0.2
TYPE CAR
It5°,15 Ft TYPE CAR 75°,5 Ft
Box(2)
Flat
Box(2)
Box(2)
Box(2)
Box
Flat & Tank
Gondola(2)
Gondola(2)
21.6
16.1*
17,2
21. h '
21.6
17.8
19.6
16.1*
13, It
BOX
Tank
Cement
Box(2)
Box(2)
Box
lt.lt
2. It
8. It
2. It
lt.0
7.2
Box(2)
Box
Box.
Box
Box
Box
Box
Box
Gondola
Box
, Large Flat
Coal(2)
Tank(2)
7.2
-7.6
-lt.0
-0.6
-5.2
-11.2
It. 8
10.2
llt.O
-2.8
13.lt
12.lt
7.0
85
-------
Table 6,
TYPE CAR
The difference in maximum A-weighted sound level [L(0,x) -L(0,5)]
between the test and reference microphones for 37 passes through
retarder .number 2. .
0°.1Q Ft TYPE CAR
0°.15 Ft TYPE CAR 30°.5 Ft
Box(2)
Gondola(2)
Box
Tank(2)
TYPE CAR
Box
Box
Box
Box
Tank
Automobile
Automobile
Box
Box
Box(2)
Gondola
TYPE CAR
Box
Grain
Box
Box
3.0
10.6
-l*.l*
10.8
30°, 15 Ft
0.6
-3.2
l*.6
0.1*
-1.2
l*.6
-2.1*
-2.1*
3.0
8.8
1.6..
.75°, 5 Ft
-3.6
-2.2
-5.1*
6.1*.
Box
Coal(2)
Gondola
Box(2)
Box(2)
Box
TYPE CAR
Box(2)
. Box
Box
Box & Cement
Box
Box(2)
Gondola
10.0 Cement 5-6
15. ** Box 1.2
9-0
8.8
5-8
18.6
1*5°, 5 Ft TYPE CAR 1*5°, 15 Ft
18.0 Flat l*.l*
18.8 Grain 5.8
20.8 Box -0.2
16.2
16.0
19.2
18.6
86
-------
Table 7. Summary pf the average differences in A-weighted sound
level between the test and reference microphones and
the corresponding standard deviations.
Retarder
#1 & #2 .
#1 & #2
#1 & #2
#1 & #2
#1 & #2
#1 & #2
#1
#2
Microphone
Position
#1
#1
.#2
#2
#3
#3
#U
#1*
Angle
0°
0
30
30
. 1*5
U5
75
75
Height
10 ft.
15
5
15
5
15
5
5
L(0,x) - L(0,5)
Average
+U.8 dB
+10.0
+1.7
+0.9
+18.3
+U.3
+2.9
-1.2
Std. Dev.
U.I dB
5-1
U. 2
1*.3
2.U
2.6
8.6
5-2
It can easily be seen that the noise radiation characteristic of retarders
exhibit strong directionality in both the horizontal and vertical planes. Much
more detailed mapping of the sound field would be needed to adequately char-
acterize the directions of minimum and maximum radiation.
The maximum A-weighted sound levels for the reference microphone were
tabulated for all passes through the retarders. The tabulated data were
grouped into 5 dB steps for the range of 100 to lUQ dB. Since only, those
cases where the maximum exceeded an A-weighted sound level of 100 dB are pre-
sented, these data should not be construed as being indicative of the average
noise levels associated with retarder operations. The tabulation was per-
formed individually for each retarder. These data are presented in Figures
U8 and U9 for retarders number 1 and 2 respectively. '
To provide an indication of the spectral content of "retarder squeal" a
limited amount of one-third octave band analysis was performed. One event
was randomly selected from the group of events that comprised each 5 dB step
for passes through retarder number 1. These events are labeled A through G.
The one-third octave band sound pressure levels versus frequency measured at
the reference microphone at the time corresponding to the occurrence of the
maximum A-weighted sound level are presented in Figures 50a and 50b for each
of the randomly selected events. The absence of data at certain frequencies
indicate that the sound pressure levels were not above the base line of the
analysis equipment. It should be noted that on curves A and G there is a
single datum point at 63 Hz.
87
-------
00
oo
w
0
w
I
!3
W
£
w
35
30
25
g 20
o
o
o
IS
10
105 110 115 120 125 130 135
I MAXIMUM A-WEIGHTED SOUND LEVEL, dB re 20 /*Pa
140
Figure 43.
Percent of occurrence for maximum A-weighted sound levels
above 100 dB measured at the reference microphone for
passes through retarder number 1.
-------
35
W
O
25
00
v£>
o
o
o
w
o
w
20
15
10
fi
00
105 110 115 120 125 130 135
MAXIMUM A-WEIGHTED SOUND LEVEL, dB re 20 /A?a
140
Figure 49. Percent of occurrence for maximum A-weighted sound levels
above 100 dB measured at the reference microphone for
passes through retarder number 2.
-------
vO
O
o
CM
(1)
P«
Q
o
co
0)
w
I
o
o
8
H
a
EH
140
130
120
110
100
90
80
70
140
130
120
110
100
90
80
70
63
8
0
125
250 1500
IK 2K 4K 63 125
. FREQUENCY, Hz
250
1500
IK
2K
Figure 50a. Third octave band sound pressure level versus frequency at
the time corresponding to the occurrence of the maximum
A-weighted sound level measured at the reference microphone
for passes through retarder number 1.
-------
140
^ 130
120
(U
>-i
CQ
D
to
to
g
D
O
CQ
Q
03
W
U
O
8
H
ffi
110
- 100
w
w 90
80
70
140
130
120
11 0
100
90,
80
70
I II II III
63 125 250 1500
FREQUENCY, Hz
2K
63 125 250 1500 1
FREQUENCY, Hz
CURVE
A
B
C °
D
E
F
G
A-WT, DB
103.0
108.8
112.6
117.2
120.4
125.2
136.6
A-WT
RANGE, DB
100-105
105-110
110-115
115-120
120-125
125-130
135-140
PERCENT
OCCURRENCE
8.6
15.5
17.2
29.3
25.9
1-7
1.7
Figure 50b. Third octave band sound pressure level versus frequency at
the time corresponding to the occurrence of the maximum
A-weighted sound level measured at the reference microphone
for passes through retarder number 1.
-------
3. Conclusions
Based on the data obtained during the conduct of this test program, the
following conclusions can be drawn:
The Single Event Noise Exposure Level (SENEL) value is very dependent
on the integration time selected; errors as great as 10 dB can occur if
the time is too short. This is especially critical as: the microphone
distance from the train is increased.
A general tendency exists for the rate of attenuation to increase with
the number'of cars in the train, reflecting a greater contribution
from wheel/rail noise (high frequency) as opposed to locomotive noise
(low frequency).
If the terrain between the train and the measurement location is not
reasonably flat and level, destructive interference can occur at fre-
quencies near that of the acoustical radiation associated with the
fundamental firing frequency of the locomotive engine.
Caution should be exercised if attempts are made to predict the
noise levels for trains in transit at locations other than the ones
at which measurements were actually taken. This is especially, criti-
cal -for changes in vertical height. . . .
The nature of activities within a railyard are such that LIQ is a
poor descriptor of the noise at the boundary of a railroad yard.
Regulation of railyard noise emission levels, in terms of L ,
would require, at a minimum, continuous monitoring for each of
several representative hours in a given day. Continuous monitor-
ing for 2k hours would be preferable..
The noise radiated from active retarders is highly directional in
both the horizontal and vertical planes, and any attempts to regulate
retarder noise should consider this directionality.
-------
U. Acknowledgement
Appreciation is expressed to officials of the American Association
of Railroads, officials and personnel of the Norfolk and Western Railroad,
and officials and personnel of the Chesapeake and Ohio Railroad, for their
assistance in this measurement program.
The authors also express appreciation to the following members of the
Applied Acoustics Section; John S. Forrer, Charles 0. Shoemaker and
Richard B. Gold for the development of the data acquisition and analysis
system and Daniel M. Corley and Ronald L. Fisher for the associated soft-
ware.
5. References
[l] Swing, J. W. and Pies, D. B. , Assessment of noise environments around
railroad operations, Report WCR 73-5 (Wyle Laboratories, El Segundo,
California, July 1973).
[2] Rickley, E. J. , Quinn, R. V, and Sussan, N. Rf, Noise level measure-
ments of railroads: freight yards and wayside, Report DOT-TSC-OST-
73-**6 (Department of Transportation, Transportation Systems Center,
Cambridge, Massachusetts, May
[3] Public health and welfare criteria for noise, Report 550/9-73-002
(U. S. Environmental Protection Agency, Washington, D. C., July 27,
1973). Available from the Superintendent of Documents, U. S.
Government Printing Office, Washington, D. C. 20^02
Bishop, D. E. and Simpson, M. A., Noise levels inside and outside
various urban environments, Sound and Vibration 8^ (5)» 51-5** (May
[5] American National Standard Specification for Sound Level Meters,
Sl.lj-1971 (American National Standards Institute, New York, New
York, April 1971).
93
-------
6. Appendix A. Data Acquisition and Analysis System
for Line Operations
Figure-A-1 identifies the components of the data acquisition system
utilized for the measurement of noise from trains in transit. To describe
the workings of the system, the following example is cited with the con-
tribution of each component discussed.
Consider a train passing an array of microphones. As the train moves
forward, it causes pressure fluctuations which travel as waves and activate
each microphone's diaphragm into vibration. These vibrations are transduced
into an AC vpltage which can be recorded for analysis at a later time. The
microphone itself was a three-rpart subsystem comprised of a one inch con-
denser microphone cartridge, protecting grid and a microphone preamplifier.
Battery-powered microphone power supplies .were utilized to provide the neces-
sary polarization voltage to the microphones. It was not practical to locate
the tape recorder next to the microphone array, since one wanted to minimize
undesired reflection effects; therefore, long cables carried the signal from
the microphone to the recording facility housed in a mobile instrumentation
van. Once the signal reached the tape recorder there existed a need for
signal conditioning prior to actual recording. A specially designed elec-
tronic system provided the necessary amplification/attenuation capability
and in addition, through a series of panel lights, provided an indication as
to whether or not 'a tape channel had become saturated (i.e., the signal had
exceeded the dynamic range of the recorder) and thus the data were hot
acceptable. The signal from each microphone.was then recorded on one track
of the seven-channel F.. M. tape recorder. Windscreens were placed over the
microphones at -all times.
A single point calibration utilizing a pistonphone which produced a
12k dB sound pressure level (re 20 micropascals) at a frequency of 250 Hz
was used for system calibration in the field. .
Once the data had been recorded, the analog tapes were returned to the
National Bureau of Standards for reduction and analysis. Figure A-2 identifies
the equipment which was utilized1 for analysis purposes. Each tape was played
back a channel at a time through the real-time analyzer. An interface was
necessary to ensure compatibility between the real-time analyzer and the
mini-computer. The time constant for the one-third octave filters was 0.2
second above 2 kHz and below 2 kHz the time constant increased with decreas-
ing frequency to 20 seconds at 20 Hz. The time constant for the A-weighting
network was 2kO milliseconds which corresponds to the requirement for "RMS
Fast" specified in American National Standard Sl.U-1971 [5]. Once all data
had been analyzed in one-third octave bands, the computer stored the data and
dumped it onto digital magnetic tape formatted to be acceptable to the large
NBS computer which was utilized for further analysis and graphical plot gen-
eration.
-------
o
WINDSCREEN NORMAL
PROTECTING F E T MICROPHONE
GRm PREAMPLIFIER
PISTONPHONE
CONDENSER
MICROPHONE
ANALOG
TAPE SYSTEM
©
©
..
AMPLIFIER
MICROPHONE
POWER SUPPLY
Figure A-l. Data acquisition and recording system for noise on main line.
-------
ANALOG
TAPE SYSTEM
0
©
O
O
FORMATING
R.T.A.
CONTROL
REAL TIME
SPECTRUM ANALYZER
INTERFACE
COUPLER
oo
A
NBS
COMPUTER
DIGITAL MAGNETIC
TAPE RECORDER
Figure A-2. Data reduction and analysis system for noise on main line.
O
O O O O
BUFFER STORAGE
MINI - COMPUTER
-------
T. Appendix B. Procedures for Calculation
of L .and SENEL
eq
The equivalent sound level (L ) is the average, on an energy basis,
noise level (usually the A-weighted level) integrated over some specified
amount, of time. The purpose of L is to provide a single number measure of
the time-varying noise for a predetermined time period- Equivalent^ in this
case, means that the numerical value of the fluctuating sound is equivalent
in level to a steady, state sound with the same amount of total energy. L
is defined as:
dt
l
where p is the time-varying, mean-square sound pressure at the point of obser-
vation, L is the corresponding sound level, p is the standard reference pres-
sure (20 micropascals) and T is the period of integration. .
A specialized mini-computer-based digital data acquisition system (des-
cribed in Appendix C) was utilized for measurements of A-weighted sound levels
for rail yard boundary measurements. Data were sampled at 0.1 second inter-
vals. For discrete sampling of the A-weighted sound level for a specified
time period, equation B-l becomes:
n
10
, (B-2)
where L is the instantaneous A-weighted sound level for the ith sample and n is
the number of samples of L in a specified time period.
The Single Event Noise Exposure Level (SENEL) provides a measure .which
quantifies the effect of duration and magnitude for a. single event. In this
case, SENEL is a measure of the individual train passby which time integrates
the level accumulated during this event with reference to a duration of one
second. SENEL is defined as:
£ M2 '&' - 10 log10 [i £ lO1^'/10 dtl ,
L o - L o . J
SOB. - 10 log10 - 10 log10 lO dt , (B-3)
where p is the time-varying, mean-square sound pressure at the point of obser-
vation, L is the corresponding sound level, p is the standard reference pres-
sure (20 micropascals) and t is the standard reference time (l second).
An analog data acquisition system (described in Appendix A) was utilized
for rail line noise measurements. As the train passed the microphone array,
voltages corresponding to the sound pressures at each of the measurement loca-
97
-------
tions were recorded on magnetic tape in'analog form. During analysis, each
analog tape was played back a channel at a time through a one-third octave
band real-time-analyzer interfaced to a mini-computer. One-third octave band
sound pressure levels and A-weighted, sound levels were digitized and stored
on magnetic tape. The analog data were, sampled at 0.3 second intervals. For
temporal sampling of the data, equation B-3 becomes:
n
SENEL = 10 login 7 S 10 V10 At, (B-4)
o 1=1
where L is the instantaneous A-weighted sound level or one-third octave sound
pressure level for the ith sample, At is the time interval between samples,
t is the standard reference time (l second) and n is the number of samples
included in the time interval which essentially includes all of the acoustic
energy from a given passby. That is, from a practical standpoint, the noise
samples must be taken during the time the signal is within a given number
of decibels down from the maximum value. As was pointed out in Section 2.1.3.,
the SENEL value is very dependent on the integration time selected. For this
study, the integration time for each passby was selected to ensure that the
error due to the finite integration time was no greater than 1 dB at any micro-
phone position.
-------
8. Appendix C. Data Acquisition and Analysis System
for Rail Yard Boundary Measurements
To facilitate data acquisition in the field, a specialized mini-computer
based digital data acquisition system was designed and fabricated by NBS. This
system was utilized for measurements of A-weighted sound levels near the rail
yard property line.
The analog portion of the system consisted of.a condenser microphone, a
battery-powered microphone power supply, an amplifier, an A-weighting net-
work, a true r.m.s. detector log converter and a sample and hold amplifier
(see Figure C-l). The dynamic range of the amplifier was BO dB. The r.m.s.
detector had a time constant corresponding to r.m.s. fast response for a
type-I sound level meter as specified in American National Standard Sl.it-
1971[5]« The sample-and-hold circuitry was under.computer control and
maintained the time coherency between the two channels utilized for data
acquisition. A third channel was used for calibration and synchronization.
The digital portion of the system consisted of a three-channel multi-
plexer, an eight-bit analog-to-digital converter (ADC), an asychronous first
in-first out memory (FIFO), a time-of- day clock (the data and time of day
are recorded automatically) and a power fail safe unit to ensure that no
data were lost in the event of a power failure. The system was self-correct-
ing in time of day and channel synchronization when power failed and was
designed so. that no data were lost while the computer was writing data on
the digital tape or writing the analyzed data on an output device. Addition-
ally, a read-only-memory (ROM) was used for the timing of the various func-
tions of the digital section.
The data were sampled and held ten times per second. The aperture time
of the sample and hold circuitry was 20 nanoseconds with a hold drift rate
of one millivolt per second. One millisecond after the data were sampled,
the reference channel was multiplexed to the ADC. The two data channels
were digitized using a ten bit ADC. The output of the ADC was connected to
a first in-first out asynchronous external memory so that, data could be
written on magnetic tape without losing new input data.
Initial calibration and check-out of the system in the field was per-
formed using a program which interrogated the multiplex interface and
printed the internal reference value and the values for channels one and
two on the teletype writer. Additionally a Fortran program was used to
scan the data tapes and print.selected values as a check on the quality of
the data while still in the field. A pistonphone which produced a 12k dB
sound pressure level (re 20 micropascals) at a frequency of 250 Hz was also
used for single point calibration. The digital tapes were returned to the
National Bureau of Standards for reduction and analysis. Figure C-2 identi-
fies the instrumentation which was utilized for analysis purposes.
99
-------
MICROPHONE AMPLIFIER
MICROPHONE AMPLIFIER
o
o
MAGNETIC TAPE
oo
A
WEIGHTING
WEIGHTING
TIMING
ROM
COMPUTER
TRUE
RMS
SAMPLE
HOLD
TRUE
RMS
SAMPLE
HOLD
REFERENCE
»
FIRST IN
FIRST OUT
AOC
Figure C-l. Data acquisition system for yard noise.
-------
o o o o
BUFFER STORAGE
MINI-COMPUTER
w
^
1
D.
'
t^
w
O O
A-
NBS
COMPUTER
k^
c
^v
OMPUTER PLO
TTF
DIGITAL MAGNETIC
TAPE RECORDER
Figure C-2. Data reduction and analysis system for yard noise.
-------
9- Appendix D. Data Acquisition and Analysis System for
Retarder, Noise Measurements
The data from the reference, and test position microphones were recorded
on separate channels of a two-channel, direct record tape recorder. The data
acquisition system is shown in Figure D-l. A single point calibration util-
izing a pistonphone which produced a 12k dB sound pressure level (re 20
micropascals) at a frequency of 250 Hz was used for system calibration in the
field.
Once the data had .been recorded, the analog tapes were returned to the
National Bureau of Standards for reduction and analysis.. Figure D-2 identifies
the instrumentation which was utilized for reduction and analysis purposes.
The time constant for the one-third octave filters was 0.2 second above 2 kHz
and below 2 kHz the time constant increased with decreasing frequency to 20
seconds at 20 Hz. The time constant for the A-weighting network was 2^0 milli-
seconds which corresponds to the requirement for "RMS Fast" as specified in
American National Standard Sl.lHl97l[5].
102
-------
WINDSCREEN
TISTONPHONE
NORMAL
PROTECTING
GRID
*D*-
F.E.T. MICROPHONE
PREAMPLIFIER
CONDENSER
MICROPHONE
ANALOG
TAPE SYSTEM
(TWO CHANNEL)
MICROPHONE
POWER SUPPLY
Figure D-l. Data acquisition and recording system for retarder noise.
-------
ANALOG TAPE
SYSTEM
(2 CHANNEL)
O.
o
O
o
FORMATING
R.T.A.
CONTROL
REAL TIME
SPECTRUM ANALYZER
INTERFACE
COUPLER
o o
A
NBS
COMPUTER
DIGITAL MAGNETIC
TAPE RECORDER
O
o o o o
BUFFER STORAGE
MINI-COMPUTER
Figure D-2. Data reduction and analysis system for retarder noise.
-------
NBS-1UA (REV. 7-73)
U.S. DEPT. OF COMM.
BIBLIOGRAPHIC DATA
SHEET
1. PUBLICATION OR REPORT NO.
NBSIR 74-488
2. Gov't Accession
No.
3. Recipient's Accession No;
4. TITLE AND SUBTITLE
Measurements of Railroad Noise Line Operation, Yard
Boundaries, and Retarders
5. Publication Date
6. Performing Organization Code
7. AUTHOR(S) J. M. Path, D. S. Blcmquist, J. M. Heinen, and
M. Tarica _____
8. Performing Organ. Report No.
NBSIR 74-488
9. PERFORMING ORGANIZATION NAME AND ADDRESS
NATIONAL BUREAU OF STANDARDS
DEPARTMENT OF COMMERCE '
WASHINGTON, D.C. 20234
10. Project/Task/Worlc Unit No.
2130498
11. Contract/Grant No.
12. Sponsoring Organization Name and Complete Address (Street, City, State, ZIP)
Office of Noise Abatement and Control
U. S. Environmental Protection Agency
Washington, D.C. 20460
13. Type of Report & Period
Covered
Final
14. Sponsoring Agency Code
15. SUPPLEMENTARY NOTES
16. ABSTRACT (A 2 00-word or less factual summary almost significant information. If document includes a significant
bibliography or literature survey, mention it-here.) .
.A field investigation of noise emission from railroad operations was
conducted. The objectives of the study were the establishment of a data
base on the noise levels associated with railroad operations, both line
(trains in transit) and yard, and the development of measurement procedures
that could be utilized in regulations applicable to the noise from rail
carrier equipment and facilities. For trains in transit, measurements were
made as a function of horizontal distance from the tracks [five locations
at 25, 50, 100, 200 and U00 feet] and as a function of microphone height
[three different heights at the 25 and 50 foot microphone locations]. .Train
passby data are presented as the maximum A-weighted sound level observed
during the passby and as Single Event Noise Exposure Levels (both A-weighted
and one-third octave band levels). A-weighted sound level measurements
were made at the boundary of the railyard, at-0.1 second intervals,, for
periods of time ranging from 1 to 23 hours over several days. These data
are presented as the energy equivalent sound level and the level exceeded
ten percent of the time. The directionality of retarder noise was also
investigated. Measurements were made of the noise emitted in various direc-
tions during retarder operation.
17. KEY WORDS (six to twelve entries; alphabetical order; capitalize only the first letter of the first key word unless a proper
name; separated by semicolons)
Acoustics; noise measurement; noise (sound); railroad yard; trains
18. AVAILABILITY (XX Unlimited
| ' For Official Distribution. Do Not Release to NTIS
I ! Order From Sup. of Doc;, U.S. Government Printing Office
Washington, D.C. 20402, SD Cat. No. CH ' .
I ! Order From NationalTechnical Information Service (NTIS)
Springfield, Virginia 22151
1?. SECURITY CLASS
(THIS REPORT)
UNCLASSIFIED
20. SECURITY CLASS
(THIS PAGE) '
UNCLASSIFIED
21. NO. OF PAGES
104
22. Price
USCOMM-DC 29042-P74
------- |