equals or exceeds that of the control concrete cured
tures yielded good low-temperature strength, but
at 5C for seven days and for 14 days is regarded
set too quickly to allow proper transport, place-
as acceptable. Three admixtures fell below these
acceptance limits: CM-42 (mix 1E_3), Ca(NO2)2+
concrete from developing ice, but did not pro-
WRDA-19 (mix 1F_8), and CM-48 (mix 1E_5).
duce adequate strength.
Figure 1c, corresponding to a curing tempera-
At the end of each task, the research team ex-
ture of 10C, shows that only the KC1 admixture
amined the test results, improved the admixture
(mix 1A_20) developed strengths higher than the
formulations, and planned the next task. The pro-
acceptance limit. However, its 28-day strength fell
cess included ten tasks until satisfactory admix-
below that of the control concrete cured at 5C.
tures were produced. Although the chief selection
The admixture K2CO2 + lignosulfonate (mix 1D_4)
parameter for the admixtures of Phase 1 was their
strength performance at 5C at seven days, other
came close to the acceptance band. The strengths
of concrete made with all other admixtures were
important parameters considered were alkali con-
lower than the acceptance limit. Figure 1d, corre-
tent, strength performance at room temperature,
sponding to a curing temperature of 20C, shows
corrosion potential, workability, slump loss, and
that none of the admixtures provided acceptable
cost.
strength at this temperature.
Two admixtures were selected for more compre-
The admixtures tested in the ten tasks of Phase
hensive evaluation in Phase 2. These showed good
1 were formulated based on experiences found in
strength enhancement at low temperature and did
the literature, preliminary tests conducted by the
not contain significant alkalis or chloride ions.
partner organizations prior to this project, physi-
cal and chemical data available for the chemicals
Phase 2: Laboratory evaluation of
involved, and the researchers' knowledge of ce-
best admixtures from Phase 1
ment chemistry. The admixture KC1 was devel-
oped at CRREL before this CPAR project began.
Objective
KC1 was included in this project as a benchmark
To evaluate the most promising admixtures from
for other admixtures, and to expand the experi-
Phase 1 in terms of the parameters most relevant
mental data on its performance. A U.S. patent on
to their use in concrete. (See Table 6.)
this admixture was granted to the Army on 22
March 1994. The low-temperature strength per-
Table 6. Tasks included in
Phase 2.
formance of KC1 was significantly superior to all
other admixtures tested. However, KC1 contains
Task no.
Title
significant amounts of sodium, and therefore may
pose a durability concern if used with alkali-reac-
1
Set times
tive aggregates.
2
Corrosion
3
Hydration products
The research team met at the end of each task
4
Alkali-silica reaction
to discuss the test results and future directions.
5
Air-void analysis
Some of the admixtures tested in early tasks were
6
Shrinkage
then reformulated, and other new formulations
were included in later tasks. The formulations
were chiefly based on predictions of the freeze-
Experimental approach
ASTM C 494 contains standards for chemical
of individual compounds.
admixtures used at temperatures above freezing.
Because portland cement is a mixture of vari-
There is no specific standard for admixtures be-
ous chemical compounds that have individual
low the freezing point of water. The parameters
chemical properties, and the admixtures were
chosen for the evaluations in Phase 2 were se-
made of more than one chemical compound, the
lected to approximate the standards provided in
combined chemical system for each concrete mix
ASTM C 494 to the extent possible. Two antifreeze
was very complex. The research process involved
admixtures, code-named DPTC and DP, were se-
several cycles of formulation and empirical vali-
lected based on their strength performance at
5C. Daraset did not perform as well as these
dation. Therefore, although some tasks did not
admixtures at 5C, and therefore was not tested
produce satisfactory admixtures, they provided a
foundation for the development of better admix-
further. Daraset performs well at temperatures
tures in later tasks. For example, some admix-
above freezing. PolarSet is a new admixture de-
6