932
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
Figure 2. Cross-section of South dike at start of test, showing settled
geometry and indicating positions of reference monitoring.
and subsequent erosion of the downstream slope. The earlier
tests on piping had a similar configuration, but with a more
sound clay dike (van Beek et al 2011). Those tests failed to
piping at reservoir levels ranging from 1.75m to 2.3m. In order
to make piping less likely this time, in each test dike a piping
prevention measure has been placed. In the West dike, piping is
controlled by a controllable drainage tube at 3.7m from the
downstream toe, while in the East dike a coarse sand filter has
been placed as a rectangular box 0.5m wide, 0.5m deep around
3.5m from the downstream toe. The grain size of this filter
varies from 1 to 2 mm, the grain size distribution is such that
excessive loss of particles through this filter is prevented
(Burenkova 1993).
2.2
South dike
The South dike was built on a 4.5m thick composition of soft
peat and clay. After construction, it was 4m high, 50m long at
crest level, with a crest width of 3m and side slopes of 1:1.5
(V:H). The core was made of sand, with a 0.5m thick clay layer.
Figure 2 shows a cross-section of the dike at the start of the test,
i.e. after consolidation resulting in a settlement of 0.99m.
The designed failure modes of this dike were slope stability
with a deep sliding plane through the subsoil with a minimum
deformation of 20cm and rupture of the clay cover by high pore
pressures inside the sand core as a result of saturating this core
with water.
3 INSTRUMENTATION
For the instrumentation a clear distinction is made between the
reference monitoring and the instruments of the participating
companies. The reference monitoring was required (and
sufficient) to closely monitor the course of the tests, while the
other instruments were validated and the measurements were
used to make updated predictions of the failures.
A total of nine companies participated with their instruments
– some in all tests, others in only one or two. Each of these
companies were invited to use their own measurements to give
an initial prediction of the failure mode and the conditions at
which failure would occur, and to update this prediction at least
every 24 hours.
Three companies providing dike safety information systems
participated in all three tests. These companies had access to the
data of the monitoring systems being validated through a central
data base. The data of the reference monitoring was not
disclosed during the tests.
3.1
West dike
The reference monitoring was primarily carried out with pore
pressure meters: two to record the water levels in the upstream
and downstream reservoirs, four lines of 17 meters each at the
interface between the lower sand and the well-compacted clay
layer at 0.9m, 2.5m, 4.3m and 11.2m from the downstream toe
and a 3 by 3 grid of pore pressure meters at the bottom of the
sand core: right behind the small clay dike and at 1.8m and
6.0m downstream, respectively. In addition, visual inspections
were carried out at regular intervals, an HD camera facing the
downstream slope taking one frame every five seconds was
used, rainfall data was recorded and the upstream and
downstream discharges were measured.
The following instrumentation was installed by the seven
companies participating in this test:
- glass fibre optics woven into geotextile, measuring
temperature and strain approximately every metre in eight lines
parallel to the toe of the dike, five at the sand/clay interface and
three in the downstream slope;
- a Fast Ground Based Synthetic Aperture Radar system,
measuring a two-dimensional displacement field of the
downstream slope every five seconds;
- two vertical tubes, installed at the upstream crest line,
measuring temperature and strain profiles over depth employing
glass fibre optics;
- a thermic infrared camera facing the downstream slope,
with a resolution of 640x480 pixels and an accuracy of 0.05 K;
- a ground penetrating radar system at 100 MHz, operated by
moving it across the crest of the dike;
- two controllable drainage tubes with measurements of pore
pressure, temperature and discharge, located close to the
sand/clay interface at 3.7m from the downstream toe (lower
tube) and right behind the small clay dike at the bottom of the
sand core (upper tube);
- six pore pressure meters at the sand/clay interface, three at
0.5m from the downstream toe and three at 2.2m.
3.2
East dike
The reference monitoring at the East dike was almost identical
to the West dike, but with four lines of 16 instead of 17 pore
pressure meters at the sand/clay interface.
The six companies participating in this test installed the
following:
- glass fibre optics woven into geotextile, measuring
temperature and strain approximately every metre in eight lines
parallel to the toe of the dike, five at the sand/clay interface and
three in the downstream slope;
- two vertical tubes, installed at the upstream crest line,
measuring temperature and strain profiles over depth employing
glass fibre optics;
- an electric resistivity system employing two rows of 14
electrodes on the downstream slope;
- a thermic infrared camera facing the downstream slope,
with a resolution of 640x480 pixels and an accuracy of 0.05 K;
- a ground penetrating radar system at 100 MHz, operated by
moving it across the crest of the dike;
- ten pore pressure meters at the sand/clay interface, five at
0.7m from the downstream toe and five at 2.2m.
3.3
South dike
The reference monitoring at the South dike consisted of 34 pore
pressure meters and six automatic inclinometers. Twentysix
pore pressure meters were installed in two cross-sections each
13m from the centre line, as indicated in Figure 2, six pore
pressure meters were installed in six water tanks on top of the
crest and the remaining two were installed in the basin on the
non-failing side of the dike and in the ditch which was
excavated during the test to reduce the overall stability. The
inclinometers were distributed along the centre line and both
instrumented cross-sections.
The seven companies participating in this test installed the
following:
- glass fibre optics woven into geotextile, measuring
temperature and strain approximately every metre in three
parallel lines along the whole length of the dike, on ground level
and on two higher levels;
- a system of six extremely accurate inclination instruments,
each mounted on top of a 5.6m steel rod placed on the slope of
the dike (three on the side of the failure, three on the other side);
