Experiments on Large Structures - Dynamic Behavior of Concrete and Seismic Engineering - page 223

exciter and a generator set capable of supplying the necessary power (about 8 kW).
Nevertheless, the accuracy of the results enables the validation of the numerical
models, taking interactions between the different sub-structures into account, and
thus facilitates validation of the other modeling techniques used to estimate the
seismic response of hydraulic works.
From this perspective an ambitious project was undertaken, in collaboration with
Hydro-Quebec, aimed at studying the dynamic interaction phenomena taking place
between the dam, the foundation and the reservoir ([PRO 97], [PAU 02]). These
sub-structures interact during an earthquake and play quite an important role in the
dissipation of energy. Moreover, in Nordic countries, the ice cover represents
another sub-structure liable to influence the dynamic response of the system.
The 84 m high Outardes 3 density-dam was chosen for the test campaign. It is
located in the northern part of Quebec and its reservoir is covered with ice almost 6
months a year. Very little experimental data on the dynamic behavior of structures
with such ice coverings exists, so the test campaign was also geared towards
assessing the effect of ice on the dam's response. Forced vibration measurements
were carried out in July and February, with daily temperatures ranging from -10 to
-15°C in the latter case. The ice cover was measured and found to reach a thickness
of up to 1.5 m along the upper bank.
The unbalanced mass exciter used for the tests can generate a force up to 90 KN
within a 0 to 20 Hz range. Figure 6.18 illustrates the instruments used for the forced
vibration tests. As the aim is to identify as many resonance values as possible within
a given frequency range, it is sometimes necessary to change the exciter's location.
When placed at the center of the structure, the latter will tend to favor symmetric
modes. Although modifying the orientation of the stress in the horizontal plane is
possible, it is sometimes necessary to move the excitement point to excite certain
anti-symmetric modes as well.
The exciter positions used for the Outardes-3 tests are indicated by double
arrows in Figure 6.18, and the accelerometers locations are identified by cubes. The
latter were placed on the crown, in various locations in the inspection galleries and
on the lower bank in order to model a three-dimensional representation of the
resonance shapes.
To assess the reservoir-dam interaction, the hydrodynamic pressures generated
by the motion of the dam and wave propagation inside the reservoir were measured.
These pressures were collected by hydrophones mounted at regular intervals on a
cable. The upper part of Figure 6.18 (summer tests) shows the cable network from