(66.2F), and following that would have 1.2 hours to
peratures below freezing. The mortar, mixed at room
temperature, was cast into 50.8- ↔ 101.6-mm plastic
be completely finished. Cold weather would increase
these times: each 10C drop in temperature would
cylindrical molds and placed into 20, 5, 10, and
20C (68, 23, 14, and 2F) rooms within 30 minutes
double them. Thus, anything that can accelerate setting
of the mix water being added to the cement. Three
often translates into big savings in time, labor, and money
samples from each mix from each temperature were
at the construction site, particularly during the winter.
tested in uniaxial compression according to ASTM C
None of the commercial admixtures appeared to
39 (ASTM 1997b) at 7 and 14 days. Prior to testing,
accelerate setting. This was not unexpected for the Type
the cylinders were stripped from their molds, warmed
A, D, and F admixtures, as they can retard setting (Table
to 5C (41F), and capped with unbonded neoprene held
1). However, the Type C and E admixtures, being ac-
within a steel-retaining cup conforming to ASTM C
celerators, should have reduced the initial and final set
1231 (ASTM 1997d). At 14 days, the remaining three
times by about one hour (Table 1), but they did not.
cylinders from each mix in each coldroom were moved
Instead, both of these admixtures appeared to increase
to the 20C (68F) room for an additional 18 days be-
the initial set by more than one hour. Final set was simi-
fore testing.
larly affected. As a possible explanation for the slow
setting times, past studies have found that accelerators,
when used at higher doses, can prolong setting times
RESULTS AND DISCUSSION
(Korhonen et al. 1997a). Another possible explanation
for these poor showings could be that the room in which
Individual admixtures
the setting times were done was cooler on the days that
Table 3 shows the effect of each admixture on the
the accelerators were tested. Measurements taken later
showed that the room did vary by as much as 5C daily.
properties of fresh mortar. The admixtures' effect on
Since the testing was done over a several-week period,
stiffening is given in terms of setting time. Initial set
between-test temperature differences are likely. Thus,
defines the onset of rigidity--when bleed water disap-
if, as mentioned earlier, each 10C drop doubles set-
pears and finishing can begin--whereas final set ap-
ting time, a 5C drop could have easily increased ini-
proximates the time when finishing becomes impos-
tial set by an hour or more. Time did not allow these
sible (when measurable strength development begins).
For example, a concrete mix containing a mortar frac-
setting tests to be repeated. However, the calcium chlo-
ride was tested when the room was steady at 19C
tion similar to that of the control mortar would be ex-
pected to be workable for up to 3.2 hours at 19C
(66F). This time, the initial set was shortened by half
Table 3. Properties of fresh mortar. Set times and flow were conducted at 19C while freez-
ing points were obtained at 20C. The letters in column 1 correspond to mortars made with
the admixtures of Table 1. Water contents were adjusted so that all mortars had approxi-
mately the same flow.
Initial
Final
Freezing
Mortar
set
set
point
Flow
w/c
designation
(hr)
(hr)
(C)
(%)
ratio
Control
3.2
4.4
1.9
104
0.48
A Low dose*
4.7
6
2.1
102
0.48
A High dose*
8.5
9.6
2.3
105
0.45
C Low dose
4.2
5.75
1.9
105
0.48
C High dose
4.0
5.3
2.7
105
0.47
D Low dose
5.25
6.5
1.9
102
0.46
D High dose
9.4
10.75
1.9
105
0.46
E Low dose
4.25
5.4
2.0
102
0.48
E High dose
5.0
6.5
2.7
104
0.42
F Low dose
4.3
6.0
1.9
104
0.45
F High dose
6.1
7.75
1.9
102
0.39
2% CaCl2†
2.7
4.5
na**
na**
0.48
*Low and high refer to manufacturer's recommendations.
†Maximum dose per cement weight allowed by the American Concrete Institute.
**na means that these tests were not conducted.
3