1145
Technical Committee 106 /
Comité technique 106
corresponded to the air entry value. Assuming a normal
distribution 98 % of the water contents were between 41 % and
14 %. These values corresponded to the average settled water
content and the water content below which asymptotic suctions
de
su
s the following null and alternative
hy thesis were tested:
int have a different
po
oint have the same
population mean to the field capacity mean
med that these values better
rep
sults of the hypothesis testing are included along the
bottom.
igure 3. Dam 1 beach relationship
tion of saturated values.
Th
%. This suggests that during
active deposition sufficient moisture is available to replenish
ng dormancies is this moisture expended
an
it) and greater than 1 (i.e.
above the liquid limit). These categories are plotted in Figure 5
relative to the final
levation at the end of test work.
Figur
lthough it was able to
support a man and prevent the auger hole collapsing.
Baseline sampling on Dam 2 indicated that it had not gained
significant strength, with only 30 % of the samples at the head
veloped.
With depth the mean water content remained constant at 27
%. The variance on the other hand was 50 %
2
for the top 400
mm, 35 %
2
from 400 mm to 1000 mm decreasing to 20 %
2
at
1500 mm and then remaining constant. Large dispersion in the
upper layer is due to this being the freshly deposited layer.
During deposition water seeps into the underlying layers and is
then drawn up during evaporative drying. This process is
reflected in the variance below the freshly deposited layer. The
decreasing dispersion with depth reflects the decreasing
influence of evaporation. The constant variance below 1500 mm
ggests this is the limit of evaporative influence.
To explore the controlling effect of field capacity on the
degree of moisture los
po
H
0
: Water contents at each sampling po
pulation mean to the field capacity mean
H
1
: Water contents at each sampling p
The two-tailed t-test with unequal variances was used to test
the hypothesis. The variances were assumed to be different as
the field samples were taken at various stages of drying whereas
field capacity has a narrower variance. Field capacity values
predicted by the
modified Kovács method (Aubertin et al, 2003)
for
each dam were used. It was assu
resented the grind differences.
Figure 3 and 4 illustrate the distributions of water content at
each position along the Dam 1 and 2 beaches for the entire
study. Re
F
Figure 4. Dam 2 beach relationship
On Dam 1 the alternative hypothesis is accepted for the outer
250 m although the degree of confidence decreases from being
high at the head of the beach. Past 250 m the hypothesis is
rejected as the probability is less than 0.05. On Dam 2 the
alternative hypothesis is accepted for the outer 100 m albeit
with less confidence and rejected for the remainder of the
beach.
These results suggest that during active deposition the steady
state water contents are controlled by field capacity with only
partial suctions developing. Seepage into the beach during
deposition rises to replenish deficits preventing suctions greater
than field capacity developing. Closer to the pool the
distributions were observed to be in equilibrium with the
phreatic surface due to the greater por
is observation was more pronounced on Dam 1 than on Dam
2, presumably due to the fact that phreatic surfaces become
more depressed along longer beaches.
Prior to test work on Dam 1 the test section was left dormant
for 6 months during high evaporative conditions over spring and
summer. Water contents obtained during the baseline sampling
showed extensive drying had taken place. The mean water
content for the upper 1000 mm was 21 %. Based on the quartile
ranges 75 % of the values were below the air entry value of 27
%. And 25 % of the values had water contents indicative of
large suctions being below 15
deficits. Only after lo
d air dry conditions reached.
2.5
Strength Gain
The impact of this limitation on drying to strength gain
during active deposition was investigated by calculating the
liquidity indices based on the average Atterberg limits. The
liquidity indices were then divided into three categories: less
than 0 (i.e. above the plastic limit), between 0 and 1 (i.e.
between the plastic and liquid lim
and 6 respectively to sampling position
e
e 5. Dam 1 distribution of liquidity indices
Figure 6. Dam 2 distribution of liquidity indices
On Dam 1 it is apparent that only the outer section reached a
state of high shear strength (Bovis 2003), with 48% of the
samples having a liquidity index less than 1 and 34% less than
0. However, 75 % of the water contents at this position were
lower than the maximum liquid limit. After 50 m the proportion
of liquidity indices less than 1 was on average 15% for all
sampling points, being slightly higher at 50 m and decreasing
towards the pool. Thus the majority of the interior is prone to
fail under shear (Holtz & Kovacs 1981) a
