of wind turbine blade modeling exist for levels of detail ranging
from low-fidelity beam models to geometrically accurate highfidelity,
finite-element models. The decision depends on what type
of analysis is needed and the availability of resources. If desired, the
precise geometry of the blade airfoils, placement of materials, and
internal structural geometry can be represented in a high-fidelity,
finite-element model as shown in Figure 2. These types of models
predict a wide range of phenomena including detailed stress contours
and local buckling. Low-dimension, finite-element models,
however, such as beam models, are suitable for other purposes.
This type of model uses representative cross section properties
and is useful for calculating, for example, deflections and natural
frequencies, but does not capture the detailed local behavior of the
high-fidelity model. Modal tests provide data to evaluate the chosen
model form and the parameters that comprise the model.
Test-Analysis Correlation. Once tests have been conducted,
models can be analyzed using the conditions (e.g., boundary conditions)
of the test to make a final set of predictions to assess the
credibility of the model. If the model predicts the test observations
within a pre-determined adequacy criterion, then the model is
considered valid for the purpose of the analysis. In many cases,
however, a model does not adequately predict all aspects of the test
to the predetermined adequacy criterion required for validation.
Then the model is calibrated or updated by modifying the model
form and/or model parameters (material properties and geometric
properties) to best represent the test observations.
Updating model parameters enables one to improve the model
so that it agrees with the test data; however, it must be done in a
physically meaningful manner. Parameters with well known values
are typically held constant; one example is the total mass of the
blade, because it can be accurately measured. On the other hand,
material and geometric properties typically have some uncertainty.
These are the parameters one would consider varying to calibrate
the model. We will later review results from our study to also use
static-load-deflection test data for model calibration along with
traditionally used modal test data.
Blade Testing and Experimental Uncertainty
In this section, we describe the free boundary condition modal
tests and static tests that were performed to provide calibration data
for the BSDS blade structural model development. A modal test of
the BSDS blade was conducted using a free boundary condition,
which is shown in Figure 3. The support conditions were designed
to minimize their effect on the modal parameters by optimal placement
and low stiffness at the two support locations using bungee
cords. Experimental quantification of the uncertainty in the measured
modal parameters is discussed later in this section.
The static test setup for the BSDS blade is shown in Figure 4.
This test was conducted at the National Wind Technology Center
(NWTC) in Golden, CO. The whiffle-tree apparatus is visible above
the blade and provides the upward vertical load at three locations
while the blade is constrained at the root.
From this test, deflection data were obtained as a function of
the measured load input, which provides a means to calibrate
the stiffness properties of the blade model. However, note that
this test provided no information regarding the properties of the
blade section outboard of the outer loading position, because this
portion of the blade is not stressed in this loading arrangement.
This is important to consider when calibrating structural models
based on static tests. The uncertainties in the root boundary condition
and the measured loads/deflections were not quantified in
these tests.
We now turn our attention to characterization of uncertainty in
experimental modal tests. Proper pretest design and test technique
are critical for the validation of blade models. In Reference 3, we
presented an experimental study for quantifying the uncertainty
in the modal parameters for the BSDS blade. In that study, we
considered test-setup uncertainty, measurement uncertainty, and
data analysis uncertainty. Bias errors in the test setup were found
to be the largest sources of uncertainty. The principal sources of
bias error were due to the support conditions (boundary conditions)
and instrumentation (mass-loading and cable damping).
of wind turbine blade modeling exist for levels of detail ranging
from low-fidelity beam models to geometrically accurate highfidelity,
finite-element models. The decision depends on what type
of analysis is needed and the availability of resources. If desired, the
precise geometry of the blade airfoils, placement of materials, and
internal structural geometry can be represented in a high-fidelity,
finite-element model as shown in Figure 2. These types of models
predict a wide range of phenomena including detailed stress contours
and local buckling. Low-dimension, finite-element models,
however, such as beam models, are suitable for other purposes.
This type of model uses representative cross section properties
and is useful for calculating, for example, deflections and natural
frequencies, but does not capture the detailed local behavior of the
high-fidelity model. Modal tests provide data to evaluate the chosen
model form and the parameters that comprise the model.
Test-Analysis Correlation. Once tests have been conducted,
models can be analyzed using the conditions (e.g., boundary conditions)
of the test to make a final set of predictions to assess the
credibility of the model. If the model predicts the test observations
within a pre-determined adequacy criterion, then the model is
considered valid for the purpose of the analysis. In many cases,
however, a model does not adequately predict all aspects of the test
to the predetermined adequacy criterion required for validation.
Then the model is calibrated or updated by modifying the model
form and/or model parameters (material properties and geometric
properties) to best represent the test observations.
Updating model parameters enables one to improve the model
so that it agrees with the test data; however, it must be done in a
physically meaningful manner. Parameters with well known values
are typically held constant; one example is the total mass of the
blade, because it can be accurately measured. On the other hand,
material and geometric properties typically have some uncertainty.
These are the parameters one would consider varying to calibrate
the model. We will later review results from our study to also use
static-load-deflection test data for model calibration along with
traditionally used modal test data.
Blade Testing and Experimental Uncertainty
In this section, we describe the free boundary condition modal
tests and static tests that were performed to provide calibration data
for the BSDS blade structural model development. A modal test of
the BSDS blade was conducted using a free boundary condition,
which is shown in Figure 3. The support conditions were designed
to minimize their effect on the modal parameters by optimal placement
and low stiffness at the two support locations using bungee
cords. Experimental quantification of the uncertainty in the measured
modal parameters is discussed later in this section.
The static test setup for the BSDS blade is shown in Figure 4.
This test was conducted at the National Wind Technology Center
(NWTC) in Golden, CO. The whiffle-tree apparatus is visible above
the blade and provides the upward vertical load at three locations
while the blade is constrained at the root.
From this test, deflection data were obtained as a function of
the measured load input, which provides a means to calibrate
the stiffness properties of the blade model. However, note that
this test provided no information regarding the properties of the
blade section outboard of the outer loading position, because this
portion of the blade is not stressed in this loading arrangement.
This is important to consider when calibrating structural models
based on static tests. The uncertainties in the root boundary condition
and the measured loads/deflections were not quantified in
these tests.
We now turn our attention to characterization of uncertainty in
experimental modal tests. Proper pretest design and test technique
are critical for the validation of blade models. In Reference 3, we
presented an experimental study for quantifying the uncertainty
in the modal parameters for the BSDS blade. In that study, we
considered test-setup uncertainty, measurement uncertainty, and
data analysis uncertainty. Bias errors in the test setup were found
to be the largest sources of uncertainty. The principal sources of
bias error were due to the support conditions (boundary conditions)
and instrumentation (mass-loading and cable damping).
การแปล กรุณารอสักครู่..