102 Chapter 5
5.2.1.5 Interpreting Test Results
As mentioned at the beginning of this chapter, the acous-
tical treatment industry is rife with misinformation. Test
results, sadly, are no exception. Information in manufac-
turer literature or on their Web sites is fine for evaluating
materials on a cursory basis. This information should
eventually be verified, preferably with an independent
laboratory test report. If manufacturers cannot supply test
reports, any absorption data reported in their literature or
on their Web sites should be treated as suspect.
When absorption data is evaluated, the source of the
data should be understood, both in terms of which stan-
dard method was used and which independent test labo-
ratory was used. Again, the test reports can help clear up
any confusion. Close attention should be paid to subtle
variations in test results, such as a manufacturer who
tested the standard-minimum area of material in lieu of
the standard-recommended area of material for an
ASTM C423 test. If two materials are otherwise similar,
a variation in sample size could explain some of the vari-
ation in measured absorption.
Additionally, there are reproducibility issues with the
reverberation chamber method. Saha has reported that
the absorption coefficients measured in different labora-
tories vary widely, even when all other factors—e.g.,
personnel, material sample, test equipment, etc.—are
kept constant.^13 Cox and D’Antonio have found absorp-
tion coefficient variations between laboratories to be as
high as 0.40.^2
Finally, it is worth noting that Sabine absorption coef-
ficients will often exceed 1.00. This is a source of great
confusion since theory states that absorption can only
vary between 0.00 (complete reflection) and 1.00
(complete absorption). However, the 0 to 1 rule only
applies to, for example, normal absorption coefficients,
which are calculated using the measurement of direct
versus reflected sound intensity. Sabine absorption coef-
ficients, remember, are calculated using differences in
decay rate and by dividing the measured absorption by
the sample area. In theory, this should still keep the
Sabine absorption coefficients below 1.0. However, edge
and diffraction effects are present and are frequently
cited (along with some nominal hand-waving) to explain
away values greater than 1.0. Edge and diffraction
effects are true and valid explanations,^14 but can be
confusing in their own right. For example, samples are
often tested with the edges covered—i.e., not exposed to
sound. Absorption coefficients greater than 1.0 resulting
from such a test can therefore be attributed mainly to
diffraction effects, which is the process where sound that
would not normally be incident on a sample is bent
towards the sample and absorbed. The confusion arises
when these test results are utilized in applications where
the edges of the sample will be exposed to sound.
A better explanation might be simply that Sabine
absorption coefficients are not percentages. The vari-
ables in the calculation of the Sabine absorption coeffi-
cient are rate of decay and test sample area. A change in
the former divided by the latter is basically what is
being determined, which does not strictly conform to the
definition of a percentage. Based on this explanation, an
DSAB value greater than 1.0 simply indicates a higher
absorption than a value lower than 1.0, all other factors
being equal. For example, a material with a Sabine
absorption coefficient of 1.05 at 500 Hz will absorb
more sound at 500 Hz than the same area of a material
having a Sabine absorption coefficient of 0.90, provided
that both materials were tested in the same manner.
Regardless of the validity of Sabine absorption coef-
ficients greater than 1.0, they are usually rounded down
to 0.99 for the purposes of predictive calculations. This
rounding down is especially important if, for example,
equations other than the Sabine equation are used to
determine reverberation time. Of course, there has been
ample debate about this rounding. For example, techni-
cally it is not rounding but scaling that is being done. As
Saha has pointed out, why only scale the numbers
greater than 1.0—what’s to be done, if anything, with the
other values?^135.2.2 Porous AbsorbersPorous absorbers are the most familiar and commonly
available kind. They include natural fibers (e.g., cotton
and wood), mineral fibers (e.g., glass fiber and mineral
wool), foams, fabrics, carpets, soft plasters, acoustical
tile, and so on. The sound wave causes the air particles to
vibrate down in the depths of porous materials, and fric-
tional losses convert some of the sound energy to heat
energy. The amount of loss is a function of the density or
how tightly packed the fibers are. If the fibers are loosely
packed, there is little frictional loss. If the fibers are
compressed into a dense board, there is little penetration
and more reflection from the surface, resulting in less
absorption.
Mainly because there is a veritable plethora of extant
information with which to work, the Owens Corning 700
Series of semi-rigid glass fiber boards will be discussed
in the next section to not only highlight one of the more
popular choices for porous absorber, but also to illustrate
various trends—such as absorption dependence on thick-
ness and density—that are not uncommon with porous
absorbers in general.