Environmental Engineering Reference
In-Depth Information
Comparison of NExT Test Data from a Rotating VAWT
to Simulated Data
An interesting comparison was performed between NExT data processing results and
simulated data and using the known modal parameters. For this activity, a simulation code
(VAWT-SDS), was used to compute the time domain response of the DOE/Sandia 34-m
VAWT test bed (Figure 2-2) during rotation in turbulent wind [Dohrmann and Veers 1989].
The structural model (including rotational effects) used in VAWT-SDS was available to cal-
culate the analytical modal frequencies and analytical modal damping. VAWT-SDS was
used to generate analytical data, which were then input to NExT. The results were compared
to the known frequencies and damping information to test the capabilities of NExT.
Simulated data were generated for the 34-m testbed using a 30 rpm rotation rate and 20
mph (8.9 m/s) turbulent winds with 15 percent turbulence intensity. Stiffness proportional
damping, sufficient to produce a damping ratio of 0.2 percent at 1.4 Hz was added to the
model. Time histories of 2,048 points for eight strain gauge outputs were generated using a
step size of 0.04 s. Sensor noise was simulated by adding a white-noise signal to each simu-
lated time history. The standard deviation of this additive signal was 2 percent of the standard
deviation of each time history. The analytical modal frequencies and damping ratios were
also calculated by extracting the complex eigenvalues from the structural matrices used in the
VAWT-SDS code. VAWT-SDS used the Newmark-Beta numerical integration scheme, and
the approximations inherent in this procedure produced period elongations. The frequency
shifts created by numerical integration were calculated and a correction was added to the
analytical values [James
et al
. 1996].
NExT was used to estimate modal frequencies and damping ratios from the simulated
data so that the NExT results and the analytical values could be compared. Table 11-2 shows
the results of the comparison. With NExT, it was possible to correctly extract the frequen-
cies and damping from the data with the exception of the first two modes at 1.27 Hz and
1.35 Hz. These two modes are very closely spaced, making it difficult to obtain accurate
Table 11-2.
Comparison of NExT With Simulated Results
[James
et al
. 1993]
Mode
Frequency (Hz)
Damping (%)
Simulated
NExT
Simulated
NExT
1st Flatwise Antisymmetric
1.27
1.31
0.2
0.4
1st Flatwise Symmetric
1.35
1.32
0.2
0.3
1st Blade Edgewise
1.59
1.59
0.3
0.3
1st Tower In-Plane
2.02
2.01
0.3
0.4
2nd Flatwise Symmetric
2.43
2.44
0.4
0.5
2nd Flatwise Antisymmetric
2.50
2.50
0.4
0.4
1st Tower Out-of-Plane
2.80
2.80
0.3
0.5
2nd Rotor Twist
3.39
3.39
0.5
0.6
2nd Tower In-Plane
3.46
3.45
0.5
0.4
3rd Flatwise Antisymmetric
3.65
3.63
0.5
0.4
3rd Flatwise Symmetric
3.73
3.73
0.6
0.4
2nd Blade Edgewise
3.88
3.87
0.5
0.3
Search WWH ::
Custom Search