5.13 TRAIN NOISE

Back

The reaction of people to noise resulting from a train passing by differs from

that produced by automobile and truck traffic. The noise due to the passage

of a train has a definite beginning and ending and a finite duration. On the

other hand, urban traffic noise is more or less continuous. There are fewer

miles of train tracks than miles of highways, so train noise generally affects

fewer people.

Railway noise in the community is often a short-term annoyance and

not a threat for hearing damage. The ambient noise level is restored after the

train has passed. Railway noise may produce a different psychological

response than other noise sources. In fact, train sounds may be somewhat

pleasant to retired railroad workers. As a result of these factors, train noise

is often treated in terms of the community response to the noise of trains

passing (Dept. of Transportation, 1978).

5.13.1 Railroad Car Noise

The standard railroad bed construction in the United States involves a tieand-

ballast construction. The ties are generally made of treated wood, and

the ballast is a crushed rock aggregate placed between the ties and on

drained and graded earth. The main function of the tie is to distribute the

Noise Sources 211

Copyright © 2003 Marcel Dekker, Inc.

load fromthe steel rail section. This type of construction offers better sound

attenuation than elevated structures or open concrete support structures.

The sound generation correlations given in this section apply to the tie-andballast

construction.

There are several contributions to railroad car noise generation,

including (a) wheel/rail interaction, (b) car coupler interaction, and (c)

vibration of structural components of the railroad car. When the railway

and railroad car are properly maintained, these components are difficult to

distinguish. As a general rule, the wheel/rail component is usually the main

source of noise generated by a passing train (Veґ r, 1976).

The four main contributions to rail/wheel noise generation for railroad

cars are (a) noise produced by rail roughness, (b) flat spots on the railroad

car wheels, (c) gaps in the rail joints, and (d) rubbing of the wheel flange and

the rail. The rubbing action of the wheel flange and the supporting rail can

be significant for tracks with sharp curves. An increase in the noise level as

much as 15dBA has been reported (Cann et al., 1974). The high-frequency

‘‘squeal’’ and low-frequency ‘‘howling’’ sound of the railroad car going

around a curve is usually not a major noise problem because the radius of

most tracks is fairly large by design. The correlations in this section do not

consider the effect of track curvature.

Impact noise occurs when a railroad car wheel with a flat spot rolls on

the rail. The flat spots may result from non-uniform service wear or wear

due to hard braking. When there are gaps between the joints of the rails,

impact noise will occur when the railcar wheel moves across the joint. This

noise is particularly noticeable if one rail is slightly higher at the joint than

the adjoining rail.

There are several noise-control procedures that can be used to reduce

the wheel/rail noise. The rails may be ground to provide a smoother and

flatter rail surface, which reduces the noise by 3 to 6dBA. The wheels may

be turned or ground to eliminate flat spots. The wheel/rail noise may be

reduced by as much as 8–10 dBA, depending on the severity of the wheel

wear, by machining or grinding the wheels. Rail joints may be eliminated by

using continuous welding of the rail joints. Noise reductions by as much as

8–10 dBA, depending on the degree of track unevenness, may be achieved by

using continuous rails. Finally, some degree of noise reduction may be

achieved by modifying the railcar support to include vibration damping in

the suspension system (Lipscomb and Taylor, 1978).

The A-weighted sound level due to the passage of one train of cars,

excluding the noise fromthe locomotive, is proportional to the time required

for the train to pass, TL=V, and proportional to the train speed, V, raised to

the third power. The quantity TL is the length of the train cars, not including

the length of the locomotive, as shown in Fig. 5-8. The A-weighted noise

212 Chapter 5

Copyright © 2003 Marcel Dekker, Inc.

level due to one train of cars passing at a distance ro ј 30m (100 ft) is given

by the following expression:

LC ј 10log10рTL=VЮю30log10рVЮю43:5 (5-81)

The total length of the railroad cars TL is in units of meters, and the train

speed V is in units of m/s. The average length of one railroad car is approximately

17.85m (58.6 ft).

5.13.2 Locomotive Noise

Most of the locomotives in the United States are driven by diesel–electric

systems. The diesel engines drive an onboard electric generator that, in turn,

provides electrical energy to the drive-wheel electric motors.

The sources of noise from the diesel–electric locomotive include

(a) diesel engine exhaust noise, (b) cooling fan noise, (c) engine structural

vibration, and (d) traction motor blower. In addition, there is some noise

generated due to wheel/rail interactions and vibration of the structural

components of the locomotive body. The contribution of each of these

noise sources is illustrated in Table 5-27 for a 3000 hp diesel–electric drive

locomotive at full throttle. It is noted that diesel engine exhaust noise and

cooling fan noise are predominant noise sources for the locomotive.

The exhaust system noise may be attenuated by about 6 dBA by using

exhaust-driven turbochargers on the diesel engine. Exhaust silencers may

also be used to reduce the exhaust noise. The installation of a silencer may

present a difficult design problem, because of the limited space on board the

locomotive. The noise from the locomotive under idle conditions is

produced primarily by vibration of structural elements of the locomotive.

Noise Sources 213

FIGURE 5-8 Train length for noise correlations.

Copyright © 2003 Marcel Dekker, Inc.

Under idle conditions, the use of an exhaust silencer will not significantly

influence the overall locomotive noise lvel.

The A-weighted sound level for a stationary locomotive at a distance

of 30m (100 ft) from the locomotive may be correlated by the following

expression (Magrab, 1975):

Lo ј 10 log10рhpЮ ю 57:2 _ _tc (5-82)

The quantity hp is the rating of the engine in horsepower; _tc is 6 dB for a

turbocharged engine and zero otherwise.

The A-weighted sound level due to the passage of NL locomotives with

a speed V at a distance ro from the centerline of the tracks is given by the

following expression:

LL ј Lo ю 10 log10р_ro=2VЮ ю 10 log10 NL (5-83)

All locomotives are equipped with safety devices, such as horns, bells,

or sirens. The sound from these devices can be 10–20 dB higher than the

noise level of the train. The noise from these safety devices is usually considered

as being necessary for the safe operation of the train, and is not

considered when noise reduction procedures are proposed.

5.13.3 Complete Train Noise

The A-weighted sound level for one pass-by of the complete train, railroad

cars plus locomotive, at a distance ro ј 30m (100 ft) is found by combining

the railroad car and locomotive noise levels:

L1 ј 10 log10р10LC=10 ю 10LL=10Ю (5-84)

214 Chapter 5

TABLE 5-27 Noise Contributions for a 3000 hp Diesel–Electric Driven

Locomotive Under Full Throttle Conditionsa

Noise source

LA at 30 m, dBA

(full throttle conditions) Energy fraction, %

Engine exhaust 84 52

Cooling fan 83 41

Engine vibration 66.5 1

Traction motor blower 75 6

Overall sound level 87 100

aThe noise levels are measured at a distance of 30m (100 ft) from the

locomotive.

Copyright © 2003 Marcel Dekker, Inc.

One of the purposes of predicting the train noise is to evaluate the

noise impact on the areas surrounding the track. As discussed in Chapter 6,

one parameter used as an indicator of community response to noise is the

day–night level, LDN. The day–night level is the energy-averaged A-weighted

sound level with an extra (10 dB) emphasis on sound generated at night. The

nighttime noise is usually more annoying than the same noise level occurring

during the daytime. The day–night level due to pass-by of several trains at a

distance ro ј 30m is found from the following expression:

LDNрroЮ ј L1 ю 10 log10 X _ 49:37 (5-85)

The quantity X is the effective number of pass-byes, with the nighttime

traffic weighted 10 times as heavy as the daytime traffic:

X ј Nd ю 10Nn (5-86)

The quantity Nd is the number of pass-byes during the daytime, defined as

the period between 7:00 a.m. and 10:00 p.m., and Nn is the number of passbyes

during the nighttime, defined as the period between 10:00 p.m. and 7:00

a.m.

The day–night sound level at any distance r from the centerline of the

tracks depends on the distance. When the observer is within a distance equal

to one-third of the total length of the train, Tt, the train radiates sound

approximately as a line source:

LDN ј LDNрroЮ _ 10 log10рr=roЮ (for r _ Tt=3Ю (5-87)

The reference distance is ro ј 30 m.

When the observer is located at a distance beyond one-third of the

train length, the train appears more nearly as a point source, and the radiation

approximates a spherical source. The day–night sound level in this case

is given by the following expression:

LDN ј LDNрroЮ _ 10 log10рTt=3roЮ _ 20 log10р3r=TtЮ (for r > Tt=3Ю

(5-88)

The average length of one locomotive is 19.5m (64 ft), and the range of

locomotive lengths is from about 18m (59 ft) to 21m (68.9 ft). The average

length of one railroad car is about 17.85m (58.6 ft).

Example 5-9. A train is made up of two 2000 hp locomotives and 70 railroad

cars. The train engine is not turbocharged. The train passes near the

site of a proposed shopping center at a speed of 25 m/s (56 mph). The train

passes four times during the day and two times during the night. The distance

from the centerline of the tracks to the property line of the future

shopping center is 240m (787 ft). Determine the day–night sound level due

Noise Sources 215

Copyright © 2003 Marcel Dekker, Inc.

to the pass-by of the trains. This information could be used in connection

with a noise impact study of the shopping center site.

The total length of 70 railroad cars is found as follows, using the

average car length:

TL ј р70 cars)(17.85 m/car) ј 1249:5m р4099 ft or 0.776 miles)

The sound level due to one pass-by of the railroad cars is found from Eq.

(5-81):

LC ј 10 log10р1249:5=25Ю ю 30 log10р25Ю ю 43:5

LC ј 17:0 ю 41:9 ю 43:5 ј 102:4 dBA

We note that the time required for the 70 railroad cars to pass by is

(1249:5=25Ю ј 50 sec.

The sound level generated by one stationary locomotive is found from

Eq. (5-82):

Lo ј 10 log10р2000Ю ю 57:2 _ 0 ј 90:2 dBA

The noise level due to two locomotives moving at 25 m/s is given by Eq.

(5-83):

LL ј 90:2 ю 10 log10Ѕр_Юр30Ю=р2Юр25Ю_ ю 10 log10р2Ю

LL ј 90:2 ю 2:8 ю 3:0 ј 96:0 dBA

The combined sound level for the railroad cars and the locomotives at

a distance of 30m from the tracks is found from Eq. (5-84):

L1 ј 10 log10р1010:24 ю 109:60Ю ј 103:3 dBA

The effective number of train pass-byes is found from Eq. (5-86):

X ј Nd ю 10Nn ј 4 ю р10Юр2Ю ј 24

The day–night sound level at a distance of 30m from the tracks is

found from Eq. (5-85):

LDNрroЮ ј 103:3 ю 10 log10р24Ю _ 49:37 ј 103:3 ю 13:8 _ 49:37

ј 67:7 dBA

The total length of the train is as follows:

Tt ј р2 locomotives)(19.5 m/locomotivesЮ ю 1249:5

ј 1288:5m р4227 ftЮ

Then,

1

3 Tt ј р1

3Юр1288:5Ю ј 429:5m > r ј 240m

216 Chapter 5

Copyright © 2003 Marcel Dekker, Inc.

The day–night level for the train noise at a distance of 240 m (787 ft) from

the tracks is calculated from Eq. (5-87) in this case:

LDNј 67:7 _ 10 log10р240=30Ю ј 67:7 _ 9:0 ј 58:7 dBA

This value is almost 4 dBA higher than the upper limit of 55 dBA recommended

by the Environmental Protection Agency (EPA) for environmental

noise.

Навигация