deeper hard limestone layer. A third and
more limited number hit karst or other
anomalous features, and these piles may
never develop design capacity, regardless of
length. There is no geotechnical reason why
these distributions of tip elevation for the
three groups necessarily overlap.
The piles belonging to Group 1 (i.e.,
between Els. -50 ft and -110 ft) are symmetrically
distributed within each pier group.
The mean (or median) of each of these
distributions depends on the depth of the
bearing strata under each respective pier,
but for the five piers shown in the Figure 4,
this elevation is 70 to 80 ft. The histograms
of the Group I piles in Figure 4 are for pier
groups of 10 to 12 piles. Similar histograms
for the larger, 35-pile piers adjacent to the
mid-channel navigation passage exhibit
more convincing normality, which supports
the distribution assumption for the endbearing
pile groups. Standard goodness-offit
tests are consistent with the assumption.
The probability distribution of these piles
is designated, f(ℓb ), in which the subscript
indicates bearing piles.
The piles belonging to Group II have
punched through their target bearing layer.
They are essentially normally distributed
based on standard goodness-of-fit tests.
The middle histogram in the figure shows
all the “punch-through” piles from across
the site, not just those associated with one
pier. While the average depths of the Group
I piles vary from pier to pier depending on
local stratigraphy, the variation of Group II
TECHNICAL
piles appears statistically similar across the
site. The probability distribution of these
piles is designated, f(ℓp ), where the subscript
indicates punch-through piles.
The piles belonging to Group III are
suspected to have penetrated into karst
lesniewski/123RF
Statistical analysis of the historical pile driving records led to an interesting interpretation of
the mechanisms leading to the extreme variability of pile lengths, and potentially, to a basis for
identifying risk in a future bridge project across Tampa Bay
www.piledrivers.org PILEDRIVER | 107
/profile_lesniewski
/www.piledrivers.org
/
