Page 323 - Numerical Analysis and Modelling in Geomechanics

304 ANALYSIS AND DESIGN OF PILE GROUPS

Table 10.6 Comparison of different analyses for 3-pile group under general loading
conditions

process, and would be expected to be different for bored piles and for driven
piles. As discussed by Randolph (1994), for driven piles, the soil modulus may
be expected to be higher in the zone immediately around the pile, while for bored
piles the soil modulus will be reduced. The near-pile E s will tend to influence
strongly the deformation of the single pile, whereas initial values of E s will
affect interaction effects between piles. Typical values for the near-pile soil
modulus for bored and driven piles have been reported by Poulos (1993) and
Poulos (1994), respectively. The discussion below will give some indication on
the assessment of the initial soil modulus.
It is now well understood that the values of E determined from conventional
s
triaxial tests with external measurement of axial strain of the soil sample (which
is highly inaccurate at strains less than about 0.1%) are usually much smaller
(typically one-fourth to one-tenth) than the initial modulus (Jardine et al., 1984).
The most reliable means of obtaining the low strain shear modulus (G ), which is
o
connected with E by the formula E =2G (1+v ), is to carry out in situ shear wave
s
s
0
s
velocity measurements. Mandolini and Viggiani (1997) showed that there is a
substantial agreement between low strain shear moduli derived from cross-hole
data and those backfigured from pile loading tests, with a trend of the latter to fit
the lower limit of the geophysical measurements. If in situ shear wave velocity
measurements are not available, G may be determined in the laboratory using
o
bender elements (Viggiani and Atkinson, 1995).
However, all of these means of measuring shear moduli are expensive and
time-consuming, and are rarely available in the early stages of design. Thus, a
preliminary assessment of initial soil modulus may be obtained from empirical
correlations with the results of conventional in situ and laboratory tests.