Structural Health Monitoring Using Smart Sensors - ideals ...

Figure 7.9. Details of the joint (Gao, 2005).
Figure 7.10. Magnetic shaker (Gao, 2005).
member can be easily removed or replaced to simulate damage without dissembling the
entire structure. A detailed picture of the joint is shown in Figure 7.9. As can be seen, a
truss member can be installed/removed by tightening/loosening the collars at the two ends
of a member.
The truss is excited vertically by a Ling Dynamic Systems permanent magnetic V408
shaker (see Figure 7.10) that can generate a maximum force of 20 lbs. with a dynamic
performance ranging from 5 to 9,000 Hz. A band-limited white noise is sent from the
computer to the shaker to excite the truss structure up to 100 Hz. The shaker is connected
to the bottom of the outer panel using a small steel rod. A PCB piezotronics load cell
(model 208B02) is installed between the steel rod and the bottom of the joint to measure
the input to the structure. This load cell has a sensitivity of 50 mV/lb, a frequency range of
0.001 to 36,000 Hz, and a measurement range of 100 lbs. in both compression and tension.
In this report, input force measurement is not used for damage detection. However, the
load cell is installed to confirm that the input force is approximately a band-limited white
noise.
Two Imote2s are firmly attached on a node of the truss by hot glue to measure
structural responses at this point in three directions (see Figure 7.11). Two Imote2s are
installed next to each other; corresponding signals from the two Imote2s are expected to
be identical if the truss node behaves as a rigid body without rotational motion. Even
though rotational behavior of the truss node is not zero, translational motion is expected to
be of much larger magnitude. The two signals should, therefore, show good agreement.
Differences in the two signals are attributed primarily to observation noise specific to each
110