Civil Engineering Reference
In-Depth Information
Solution.
Since the zone of liquefaction extends from a depth of 1.2 to 6.7 m, the thick-
ness of the liquefiable sand layer
H
2
is equal to 5.5 m. By entering Fig. 7.6 with
H
2
5.5
m and intersecting the
a
max
0.2
g
curve, the minimum thickness of the surface layer
H
1
needed to prevent surface damage is 3 m. Since the surface layer of unliquefiable soil is
only 1.2 m thick, there will be liquefaction-induced ground damage.
Some appropriate solutions would be as follows: (1) At ground surface, add a fill layer
that is at least 1.8 m thick, (2) densify the sand and hence improve the liquefaction resis-
tance of the upper portion of the liquefiable layer, or (3) use a deep foundation supported
by soil below the zone of liquefaction.
7.4 VOLUMETRIC COMPRESSION
7.4.1
Main Factors Causing Volumetric Compression
Volumetric compression is also known as soil densification. This type of settlement is due to
earthquake-induced ground shaking that causes the soil particles to compress together.
Noncemented cohesionless soils, such as dry and loose sands or gravels, are susceptible to this
type of settlement. Volumetric compression can result in a large amount of ground surface set-
tlement. For example, Grantz et al. (1964) describe an interesting case of ground vibrations
from the 1964 Alaskan earthquake that caused 0.8 m (2.6 ft) of alluvium settlement.
Silver and Seed (1971) state that the earthquake-induced settlement of dry cohesionless
soil depends on three main factors:
1.
Relative density D
r
of the soil:
The looser the soil, the more susceptible it is to volu-
metric compression. Those cohesionless soils that have the lowest relative densities will
be most susceptible to soil densification. Often the standard penetration test is used to
assess the density condition of the soil.
2.
Maximum shear strain
max
induced by the design earthquake:
The larger the shear
strain induced by the earthquake, the greater the tendency for a loose cohesionless soil
to compress. The amount of shear strain will depend on the peak ground acceleration
a
max
. A higher value of
a
max
will lead to a greater shear strain of the soil.
3.
Number of shear strain cycles:
The more the cycles of shear strain, the greater the ten-
dency for the loose soil structure to compress. For example, it is often observed that the
longer a loose sand is vibrated, the greater the settlement. The number of shear strain
cycles can be related to the earthquake magnitude. As indicated in Table 2.2, the higher
the earthquake magnitude, the longer the duration of ground shaking.
In summary, the three main factors that govern the settlement of loose and dry cohe-
sionless soil are the relative density, amount of shear strain, and number of shear strain
cycles. These three factors can be accounted for by using the standard penetration test, peak
ground acceleration, and earthquake magnitude.
7.4.2
Simple Settlement Chart
Figure 7.7 presents a simple chart that can be used to estimate the settlement of dry sand
(Krinitzsky et al. 1993). The figure uses the standard penetration test
N
value and the peak
ground acceleration
a
p
to calculate the earthquake-induced volumetric strain (i.e.,
H
/
H,
expressed as a percentage). Figure 7.7 accounts for two of the three main factors causing
