Page 168 - Numerical Analysis and Modelling in Geomechanics

MODELLING OF GROUND WAVES 149
The impact of the hammer onto the pile head was modelled with the
conceptual model, and various values for the parameters , and were considered.
The closest match with site pile head records was achieved with and Typical
ranges of values for the parameters are and The values for stiffness of the dolly
and mass of the anvil are slightly outside these bands because of the special pile
cap required for the SIMBAT tests.
The correlation between field measured and computed pile head displacement-
time and force-time functions was quite close in magnitudes, but the force
duration was overestimated, Figure 5.13.
In stage 2 the force-time function was imposed onto the head of the model of
pile shaft plus springs and dampers. The displacement-time functions were
computed at each of the spring and damper sets on the shaft and toe.
In stage 3, the displacement-time functions were imposed onto the shaft and
toe interface nodes of the FE/IE mesh. The computed surface vibrations at 5.5 m,
10m and 16.5 m from the pile are compared with measured values in Figures
5.14 and 5.15.
Some discrepancies are apparent in the radial values, with measured S-waves
generally larger than those computed. Also, there appears to be an observed early
P-wave component, probably due to an eccentric strike, and a later hammer
bounce trace. However, close agreement is shown in the vertical waves. Overall,
however, the agreement is encouraging, considering the ground variability and
the imperfect impacts.

Impact driving at Flitwick, Bedfordshire
A second calibration exercise is now presented, for the driving of a 12m long
steel H-pile (305×305×89 kg/m), using a 3200 kg drop hammer falling through 1.
0 m. The soil conditions comprised topsoil and soft clays to 2.4 m, loose sand
and gravel at 2.4–4.8 m, then very dense uniform sands to considerable depth.
The water table (WT) was at 2.4 m. Although the static elastic moduli for the
three layers were estimated to be 5 MPa, 20 MPa and 50 MPa respectively, the
dynamic stiffnesses used for the stage 3 computation were identified as 20 MPa
above the WT and 200 MPa below. The three-stage computation was conducted
for the condition of 7 m penetration depth.
A summary of some of the parameters used in the analyses of the hammer
impact, the pile shaft and the pile toe is given in Table 5.2.
Although no pile head signals were available during driving, the ground
surface vibrations were recorded, and comparisons with the computed values are
shown in Figure 5.16, for radial particle velocities at 7 m and 16.5 m distances
from the pile. The observed signals appear to include an extra peak from an early
P-wave arrival, which is accentuated at 16.5 m, and was probably caused by an
eccentric strike. Otherwise close correlation is achieved.