Table 8. Permeance test results compared to manufacturers' data,
sorted by lowest to highest permeance as published by the manufac-
turers.
Modified cup
method permeance
Wet cup
Wt. change
Membrane
(perm) or
(perm) or
per hour*
(gr/hrft2in. Hg)
(gr/hr ft2 in. Hg)
(gr/hr ft2 in. Hg)
manufacturer
Protecto Wrap
0.026
0.003
0.021
0.049
Soprema
0.012
0.0036
0.011
0.02
Royston
0.052
0.05
0.050
NEI
0.034
0.08
0.025
0.000044
Polyguard
0.030
0.1
0.026
0.095
W.R. Grace
0.027
1
0.021
*The permeance values shown in the top row are those found in the manufacturers' data
at the beginning of our testing. The italicized permeance values shown in the second row were
provided later by the manufacturer.
calculating the results. We also had difficulty
A blister can originate in a smaller void con-
in properly maintaining test conditions through-
taining water than one containing dry air.
out the test. There may also be problems
Bridge inspectors should be alerted to the fact
with unit conversions, making it difficult to com-
that a membrane may not be adequately in-
pare results from different laboratories and manu-
stalled if it has unadhered areas larger than a
facturers.
quarter coin. This is the smallest void size that
At the beginning of these tests, we wondered if
can originate a blister.
blisters would be less likely to develop if a mem-
Once a blister initiates, even though it may not
brane can pass water vapor and air but not allow
immediately be evident, it will eventually be-
liquid water to pass. However, these water vapor
come a problem.
transmission tests do not allow us to determine if
The Protecto Wrap and W.R. Grace membranes
that is possible. These membranes are effective
showed the least tendency to contain weak ad-
water vapor retarders and are not likely to act as a
hesion spots.
ventilating layer.
The Protecto Wrap membrane's bond was the
Neither procedure (modified-cup or wet-cup)
one least affected by membrane applicator.
tests the membrane in the condition of use (i.e.,
sandwiched between two permeable materials
Tensile strength and elongation
that are capable of puncturing the membrane).
Membranes are most prone to splitting during
Such a method would require much more devel-
cold weather when cracks in the concrete deck
opment time and numerous trials before common
are at their widest.
acceptance would be likely.
While strength and elongation both describe
tensile properties, the latter is considered most
appropriate to judge a membrane's ability to
CONCLUSIONS
span a crack.
Adhesion
The Soprema membrane had the best elonga-
Poor adhesion is the primary cause of blister-
tion at room temperature.
ing.
The Royston membrane was the least stretch-
A membrane does not have to be perfectly ad-
able at all temperatures.
At 4F, all membranes, except for Royston's,
hered to the deck to avoid blistering.
High bond strength matters less than the conti-
were considered comparable in their ability to
nuity of the bond.
span an active crack.
20
Previous Page
