Experiments on Large Structures - Dynamic Behavior of Concrete and Seismic Engineering - page 205

carrying the measured signal is grounded at both extremities. Both groundings are
unlikely to have the same potential, and the potential difference causes currents to
flow within the loop. The signal will therefore be modulated by the potential
difference, which will induce noise in the measuring system. We can avoid forming
such a loop by grounding only one end of the signal carrying wires.
Tribo-electricity effects are due to the static electricity produced by friction
between two different surfaces. Thus, the bending and folding of a cable, a shock or
pressure can create an electrical charge between conducting wires and the outer
shield, or between the conducting wires themselves, which generates a voltage
between both ends of the wire. For this reason, we recommend use of low-noise
wires. These are generally tested along their whole length by manufacturers, to
guarantee low noise levels. When adding connectors at the ends of these wires, we
should take care not to create leaking spots . Using carbon tetrachloride as a solvent
and xylene as a cleaning agent is also helpful.
6.2. Dynamic loads
The development of reliable testing methods and high-frequency analog-to-
digital data recording devices has facilitated the dynamic testing of large civil
engineering structures. Two main dynamic loading methods are used.
The first test involves applying a harmonic load, via an unbalanced mass, as an
exciter (the principle of which is illustrated in Figure 6.3). The harmonic forced
response is measured and the process is repeated over a given frequency range. The
amplitude of the harmonic force applied to the structure is proportional to the
squared excitation frequency. The main drawback of this method is that it is difficult
to generate an appreciable harmonic force at low frequencies, and impossible to
obtain the static response of the tested structure. The testing method proceeds with
rather low increments of the excitation frequency to obtain a sufficient frequency
resolution, which involves protracted experiment times.
Despite this, the advantages of the method are appreciable, which explains why
it is still used so often. To begin with, the whole power available for excitation
concentrates at only one frequency. It also enables structure linearity to be tested by
the successive introduction of different amplitude harmonic forces for the same
excitation frequency: this is easily achieved by varying the mass or its eccentricity.