Objectives
Table 1. Membrane manufacturer and product
tested.
Though field tests have proven that membranes
Manufacturer
Product
crete, there are problem areas where improve-
NEI
AC Bridge and Deck Seal
ments in test procedures or materials are needed.
Polyguard
665 LT
If a membrane cannot be fully adhered to the deck,
Protecto Wrap
M140A-R
or it somehow becomes damaged during construc-
Royston
10AN Easy Pave ER
Soprema
Sopralene Flam Antirock
tion or is unable to resist splitting when cracks
WR Grace
Bituthene 5000
develop in the underlying deck or bituminous
overlay, moisture and chlorides can leak through
of a membrane to rock puncture.
objectives of this work were to develop laboratory
Water vapor permeance: to determine how eas-
ily water vapor can pass through a membrane.
ing for their ability to resist cracking, blistering,
and puncturing. ASTM lists a number of tests to
For reference and general interest, Appendix A
evaluate various engineering properties of tape,
presents technical data from manufacturers' bro-
rubber, roofing, plastics, and geomembranes. The
chures for each membrane evaluated in this
problem is that there is no group of standards, or
project.
ways to interpret them, that all manufacturers fol-
low when reporting performance data for their
ADHESION
products. As a result it is difficult, if not impos-
sible, to rate one membrane against another based
Lack of adhesion is considered to be the lead-
on manufacturer-supplied data. Our plan was to
ing cause of membrane blistering. This study re-
review these and other literature to develop a set
viewed current testing standards, such as ASTM
of testing standards specific to the above objec-
C794, D903, and D1000, and developed one that
tives.
could be used as the standard by which to evalu-
ate the ability of sheet membranes to adhere to
Approach
concrete.
NETC developed a list of sheet membranes that
have been used on bridge decks in New England.
Procedure
From that list we invited suppliers of membranes
Adhesion was measured by peeling strips of
to participate in this study by providing materials
membrane off mortar. The test consisted of adher-
and by making test samples. (Several suppliers of
ing membranes to carefully prepared mortar sur-
liquid membranes were also interested in partici-
faces, cutting the membrane into strips, and ap-
pating but were not accommodated, because test-
plying a tensile load at a constant rate of extension
ing liquid membranes was not within the scope
until each strip peeled off the mortar a predeter-
of this study.) The intent of this work was to rec-
mined distance. Test specimens were prepared, as
ommend tests to evaluate one membrane against
shown in Figure 1, according to manufacturers'
another. We acknowledged that until a systematic
recommendations. Two sets of test specimens for
field test is conducted, these laboratory tests could
each of the six membrane types were constructed:
not reliably predict expected service life, as labo-
CRREL made three specimens and the membrane
ratory tests do not simulate field conditions and,
supplier made three. Making two sets of samples
therefore, only suggest possible outcomes in the
helped to determine if choice of applicator influ-
field.
enced results. The specimens were prepared as
This project subjected sheet membranes from
follows:
the six manufacturers shown in Table 1 to the fol-
At room temperature, mix one weight of Type I
lowing four tests:
portland cement with two weights of Ottawa
Adhesion: to evaluate the adhesion developed
sand (2030 grade) according to ASTM C305.
Cast mortar into 6- 6- 21-in. (15- 15- 53-
between a membrane and a concrete substrate.
Tensile strength and elongation: to determine
cm) molds and cover with sheet of plastic.
how well a membrane can resist and accom-
After 24 hours, strip molds and cure mortar
modate movement of the concrete deck.
beams in room-temperature limewater for 14
days.
2