Environmental Engineering Reference
In-Depth Information
plane, and all of the frequencies are flapping frequencies. A data acquisition
system, type NI PXI-4472, was used and the sampling frequency was 1,000 Hz.
The start-up wind velocity of the wind turbine is 3 m/s; the rated wind velocity
and rated rotating speed are 8 m/s and 400 rad/s, respectively. The test is con-
ducted in the wind tunnel laboratory at the Harbin Institute of Technology, China.
The vibration responses of blades under different wind velocities were measured to
determine the relations between the frequency and rotating speed. Six cases of
wind velocity were used in this test: 0, 4, 5, 6, 7, and 8 m/s. The duration under
each wind velocity is approximately 3 min. When the wind velocity is 0, the
vibration signals are measured by rapping the blades.
Structural damage was simulated by abrasing laminations from the surface of
one blade. Figure
9.9
b shows the artificial damage on the blade. Damage levels
were controlled by different abrasion depths and areas. It should be noted that this
damage cannot be quantified.
9.2.4.2 Damage Detection Results Based on Experimental Data
Modal frequencies of the first three modes corresponding to different wind speeds
and damage levels are shown in Fig.
9.10
, which shows that both rotation and
damage change the modal frequencies of the blades. The first damage level has
slight effect on the first mode. Anyway, in some cases, it is difficult to distinguish
whether the decrease in frequencies is due to the decrease in rotational speed or
structural damage according to data in Fig.
9.10
.
After using the PCA algorithm, Fig.
9.11
shows the damage index under dif-
ferent damaged levels for the experimental data. X-axis indicates number of the
wind speed samples. From Fig.
9.11
, the effects of rotation have largely been
removed. The damage indices for the damaged cases are separated from the
undamaged ones, but the different levels of damage are not clearly detected.
9.3 Fatigue Damage Detection Based on High Spatial
Resolution DPP-BOTDA
9.3.1 Principles of DPP-BOTDA
The schematic of DPP-BOTDA sensing system is shown in Fig.
9.12
. BOTDA
system employs a stimulated Brillouin scattering (SBS) technique [
48
]. Two
counterpropagating laser beams, i.e., a pump pulse and a CW probe wave, are
injected from both ends of the sensing fiber. The two laser beams have certain
frequency differences which are near the Brillouin frequency of sensing fiber. Thus
acoustic wave can be excited by the interaction of these two laser beams. The
pump pulse is backscattered by the phonons, and part of its energy is transferred to
the CW. The power gain of the CW, i.e., Brillouin gain signal, is measured at the
output end of the probe light. The relation between Brillouin gain of the CW and
