Environmental Engineering Reference
In-Depth Information
due to its high signal-to-noise ratio (SNR) and high accuracy. Usually, BOTDA
employs two counterpropagating lights, i.e., a pump pulse and a CW probe wave,
to induce stimulated Brillouin scattering (SBS) and its spatial resolution is
determined by the pump pulse width [
48
]. Shorter pulse width means higher
spatial resolution but lower accuracy. Shorter pulse will get weaker Brillouin
signal, broaden the Brillouin gain spectrum (BGS), and decrease the SNR, which
results in a 1 m spatial resolution limitation for the distributed optic sensors. In
2008, Li W and Bao X et al. proposed the differential pulse-width pair (DPP)
technique in BOTDA achieving high-spatial resolution by using a small difference
in the pulse pair based on the mechanism of prepumping the acoustic wave [
49
].
The spatial resolution of the DPP-BOTDA system is determined by the differential
pulse, i.e., the pulse-width difference of the pulse pair, rather than the original
pulses [
49
,
50
]. Other local sensing techniques, such as PZT-based techniques,
have also been investigated frequently. With these techniques, an array of sensors
can capture the stress waves passively generated by active damage to identify the
type or location of the damage [
9
-
13
]. On the other hand, sensor arrays are also
used for receiving stress waves generated by the actuator arrays to actively detect
the damage in the path of wave propagation [
14
,
15
].
In addition, global vibration-based monitoring methods are also employed to
detect structural damage to blades, which might occur far from the location of the
sensors. The global vibration-based methods can identify damage inside the blade
without having to map over the surface of the entire blade with a sensor. Gross
et al. verified the damage detection method using modal response data from a wind
turbine blade [
16
]. They loosened the bolts at the root of one of the three blades to
simulate a damaged blade and acquired the structural response data using accel-
erometers mounted on the blade. Three damage detection techniques making use
of the strain energy, modal flexibility matrix, and differences in mode shapes have
been used. Ghoshal et al. tested four vibration-based damage detection techniques
on a section of a fiberglass blade. The structural damage was simulated by a steel
plate (additional mass) clamped to the blade. A periodic signal with a frequency
bandwidth of 100-500 Hz was used to excite the blade and the frequency response
function, transmittance function, operational deflection shape, and resonant com-
parison were compared between the damaged and healthy structure [
17
]. Kraemer
and Fritzen presented a three-step concept for the structural health monitoring of
offshore wind energy plants [
18
]. The basic idea is also based on the global
approach and involves vibration measurements, the stochastic subspace fault
detection method, and the multivariate autoregressive model. Whelan et al. pro-
posed an integrated monitoring of wind plant systems, which consisted of a
wireless network and a large array of multiaxis accelerometers to evaluate the
modal properties of the system as well as individual members [
19
]. Dolinski and
Krawczuk described a vibration-based method for damage localization in a
composite wind turbine blade based on a one-dimensional continuous wavelet
analysis applied to mode shapes. They simulated a series of different localizations
and sizes of damage. The results were obtained from both numerical simulation
and a scale model [
20
]. Frankenstein et al. also employed the global damage
