Acoustical Noise Control 71diurnal variations; observations on selected days of the
week capture especially noisy events varying from day
to day or occurring at certain times during the week.
All of these analyzers, with the exception of the
Brüel and Kjaer 2143, will capture only the sound pres-
sure levels over time. The 2143 will also capture the
spectrum of the noise as well. If an analyzer such as the
2143 is not available, it is advisable to make a number
of measurements of the spectrum as well as the
time-stamped level record.
The data collected from the site survey should be
combined with projections of the levels anticipated in the
sound room and what will be tolerated in adjacent spaces.
4.2.2 Transmission Loss
Once the noise load is known and the desired NC is
determined, attention must be given to the design of
systems that will provide enough isolation to achieve
the goal. Transmission loss is the loss that occurs when
a sound goes through a partition or barrier. A higher TL
number means more loss, i.e., less acoustic energy gets
through. If the desired NC or noise limit is known, and
the noise load is known, a designer must then design
barriers or partitions that have appropriate TL to meet
the design goal.
(4-1)4.2.3 Sound Barriers
The purpose of a sound barrier is to attenuate sound. To
be effective, the barrier must deal with airborne as well
as structure-borne noise. Each barrier acts as a
diaphragm, vibrating under the influence of the sound
impinging upon it. As the barrier vibrates, some of the
energy is absorbed, and some is reradiated. The simplest
type of barrier is the limp panel or a barrier without any
structural stiffness. Approached theoretically, a limp
panel should give a transmission loss increase of 6 dB
for each doubling of its mass. In the real world, this
figure turns out to be nearer 4.4 dB for each doubling of
mass. The empirical mass law deduced from real-world
measurements can be expressed as
(4-2)where,
TL is the transmission loss in decibels,
M is the surface density of the barrier in pounds per
square foot.
Transmission loss also varies with frequency, even
though Eq. 4-1 has no frequency term in it. With a few
reasonable assumptions, the following expression can
be derived, which does include frequency:^8(4-3)
where,
f is the frequency in hertz.Fig. 4-7 is plotted from the empirical mass law stated
in Eq. 4-3, which is applicable to any surface density
and any frequency, as long as the mass law is operating
free from other effects.From Fig. 4-7 several general conclusions can be
drawn. One is that at any particular frequency, the
heavier the barrier, the higher the transmission loss. A
concrete wall 12 in (30 cm) thick with a surface density
of 150 lb/ft^2 (732 kg/m^2 ) gives a higher transmission
loss than a 3 in (6 mm) glass plate with a surface
density of 3 lb/ft^2 (14.6 kg/m^2 ). Another conclusion is
that for a given barrier the higher the frequency, the
higher the transmission loss.
The straight lines of Fig. 4-7 give only a partial
picture since barrier effects other than limp mass domi-
nate. Fig. 4-8 shows four different regions in the
frequency domain of a barrier. At extremely low
frequencies, stiffness of the barrier dominates. At some-
what higher frequencies, resonance effects control as
the barrier vibrates like a diaphragm. Above a critical
frequency, a coincidence effect controls the transmis-
sion loss of the barrier. The mass law is an important
effect in determining barrier performance, but reso-
nance and coincidence cause significant deviations.TL 10Pincident
Ptransmitted= log©¹§·---------------------------TL=214.5logM+ 3Figure 4-7. The empirical mass law based on real-world
measurements of transmission loss. Surface density is the
weight of the wall corresponding to a 1 ft^2 wall surface.TL=14.5log Mf 16–