Geoscience Reference
In-Depth Information
5 CASE HISTORIES
5.1 GRS Walls Constructed with Fine-Grain Soils
One possible advantage of the GRS wall is that the select fill does not have to
comply with tight grading requirements. As such, fine-grain soils may be used in
the construction of a GRS wall to achieve cost savings. The use of a crushed shale
in the construction of a GRS wall supporting the on- and off-ramps of a major
interchange in Western Sydney was reported by Won et al. (1994). The maximum
wall height is about 8m. The soil reinforcement is a high-tenacity polyester strap
known as Paraweb. The select fill was compacted to near-maximum dry density
(as determined by standard Proctor test), and the foundation material was
competent. The wall was instrumented with
.
Load bolts to measure reinforcement tension, noting that only the load
bolts of the lowest level of reinforcement survived the construction
.
Earth pressure cells to measure foundation stress
.
Extensometers at three levels to measure internal displacements
.
Survey points to measure horizontal wall displacements
The monitoring until 1994 showed that a facing panel bulged out by
100 mm although the horizontal displacements of other instrumented panels were
typically less than 50mm. Since then, the lateral movements of certain wall
panels have continued at a slow rate. The as-measured reinforcement tension was
significantly lower than the designed value but also manifested a slow increase
with time. Although the causes of the higher wall deflection are a matter of
debate, compaction of fine-grain soil at a moisture content on the dry side of
optimum will lead to a high matrix suction that may dissipate with time. This
matrix suction can be modeled by an apparent cohesion that reduces with time.
The effects of a reduction of apparent cohesion with time can be studied by
conducting a FLAC analysis of a “fictitious” GRS wall as shown in Fig. 8 .
The dimensions of this wall were chosen to ensure adequate overall
stability. The reinforcement was modeled as elastic and with high interface
parameters to suppress pullout failure. Hence the wall could not have any form of
internal stability. Both the general and select fill were modeled by the Mohr-
Coulomb elastic-plastic model following a nonassociative flow rule (dilatancy
angle
was assumed
for both the general and select fill. The reinforcement was assigned an elastic
axial stiffness of 1000 kN/m. The construction of the wall was modeled in a layer-
by-layer manner, and an apparent cohesion of 60 kPa was assumed in the analysis
to represent the initial matrix suction. As such, the analysis is a total stress
analysis. Dissipation of matrix suction after completion of construction was then
simulated by a progressive reduction in apparent cohesion via time stepping.
¼
0). For the purpose of this exercise, a friction angle of 30
8
Search WWH ::




Custom Search