.
/']~
LA-UR- jD
~/
r:r
Approved for public release;
distribution is unlimited.
QAlamos
NATIONAL
LABORATORY
---EST
.
1943
---
Title: TIME SERIES PREDICTIVE MODELS OF PIEZOELECTRIC
ACTIVE-SENSING FOR SHM APPLICATIONS
Author(s): Gyuhae (NMI) Park, LANL
Eloi
F.
Figueiredo, LANL
Kevin
M.
Farinholt, LANL
Charles
R.
Farrar, LANL
Intended for: SPIE INTERNATIONAL SYMPOSIUM
ON
SMART
STRUCTURES
AND NONDESTRUCTIVE
EVALUATION,SAN DIEGO,
CA, MAR. 7-11,2010.
Los Alamos National Laboratory,
an
affirmative action/equal opportunity employer, is operated by the Los Alamos National Security,
LLC
for the National Nuclear Security Administration
of
the U.S. Department of Energy under contract DE-AC52-06NA25396. By acceptance
of this
article, the publisher recognizes that the U.S. Government retains a nonexclusive, royalty-free license to publish or reproduce the
published form of this contribution, or to allow others to do so, for U.S. Government purposes. Los Alamos National Laboratory requests
that the
publisher identify this article as work performed under the auspices of the U.S. Department of Energy. Los Alamos National
Laboratory strongly supports academic freedom and a researcher's right
to
publish;
as
an
institution, however, the Laboratory does not
endorse the viewpoint of a
publication or guarantee its technical correctness.
Form
B36
(7/06)
Time Series Predictive Models
of
Piezoelectric
Active-Sensing
for
SHM Applications
Gyuhae Park, Eloi F. Figueiredo, Kevin M. Farinholt, Charles
R.
Farrar
The Engineering Institute
Los Alamos National Laboratory
Los Alamos, NM 87545
In this paper, the use
of
time domain data from piezoelectric active-sensing techniques
is
investigated for structural health monitoring (SHM) applications. Piezoelectric
transducers have been increasingly used in
SHM
because
of
their proven
advantages. Especially, the use
of
known and repeatable inputs at high frequency ranges
makes the development
of
SHM
signal processing algorithm easier and more efficient.
However, to date, most
of
these techniques have been based on frequency domain
analyses, such as impedance-based or high-frequency response functions (FRF) -based
SHM
techniques. Even with Lamb wave propagations, most researchers adopt frequency
domain or wavelets analysis for damage-sensitive feature extraction. This process usually
requires excessive averaging to reduce measurement noise and more computational
resources, which is not ideal from both memory and power consumption
standpoints. Therefore in this study, we investigate the use
of
autoregressive models with
exogenous inputs (ARX) with the measured time series data from piezoelectric active-
sensors. The test structures considered in this study include a section
of
CX
-100 wind
turbine blade and a 2 x 2
ft composite plate, where the plate was subjected to a series
of
impact loadings to induce damage in the form
of
fiber delimation. The performance
of
this technique is compared to that
of
traditional autoregressive (AR) models, traditionally
used in low-frequency passive sensing techniques, and that
of
FRF-based analyses, and
its superior capability in
SHM
is demonstrated. This paper outlines the advantages
of
this method over traditional frequency-domain analyses and provides guidelines for using
time-series data from active-sensors for real-world
SHM applications.
UNCLASSIFIED
Time Series Predictive Models
of
Piezoelectric
Active-Sensing
for
SHM
Applications
Gyuhae Park, Eloi Figueiredo, Kevin
M.
Farinholt
,
Charles
R.
Farrar
The Engineering I
nstitute
Los Alamos National Laboratory, USA
2010 SPIE
Smart
Structures
and
Materials/NDE
and
Health
Monitoring
,
March
8-11
, 2010
A
..
LosAIamos
I
!("';o
I School of
Jacobs
Engineeing
Slide
1117
Engineering Institute
UNCLASSIFIED
Outline
• The application of time series predictive models obtained from
piezoelectric active-sensing techniques is presented for
SHM.
•
Time series autoregressive models with exogenous inputs
(ARX) are implemented to extract damage sensitive features,
• Experimental results will be summarized to demonstrate the
capability
of
the proposed method,
,I"J
Q Los Alamos
" - • , ~ r
hap:!1
ea
rt
h2
tech.
files.
wordpreu.coml200811 O/suzlo
"",b<oe·,PB
~
UCS
D I School of
-----e-n-g-in-ee-r-in-g-In-s-ti-tu-te----
Jacobs
Englnqrfng
Slide
1117
UNCLASSIFIED
SHM based
on
Piezoelectric Active-Sensing
• Being increasingly used
• The methods includes electro-mechanical impedance methods, high-
frequency response functions, guided waves.
• Most of these methods have been based on frequency domain
(impedance) or other approaches for damage sensitive feature
extraction.
-
Excessive
averaging
, intensive computation
• Predictive models of time series data obtained from piezoelectric active-
sensing technique have not been extensively used
in
SHM applications.
iI
F
-"·
j
~
l
-
X I
,.Q
~2
0.. 06 0.8
1.0
1.1
1.01
l.6
1.8
2.0 0 10
;0
lO
40
50
':""J
...
1
J0iD
I School
of
LoS
Alamos Engineering
Institute
Jacobs Engineering
S
~
de
3
11
7
UNCLASSIFIED
SHM of Wind Turbine Blades
•
Why
Turbine Blades?
-
Most
expensive
component
to
repair
-
Rotational
imbalance
results
in
additional
damage
-
Huge,
remote
locations,
wireless
desirable
•
We investigated several active-sensing
techniques to detect damage
in
composite
turbine
blades (Light-Marquez et al. 2010)
Slide
4117
2
UNCLASSIFIED
Time Series Models
•
An
autoregressive (AR) model predicts the current time point
in
a
series as a
linear combination of n previous time points.
p
x
="
ax
.
+E
I L.. J
'-J
,
)=
1
• The
AR
representation can
be
extended to an autoregressive with
exogenous inputs (ARX) by including the effect of input to a system
x =
~ax
. +
~f3
.
y
. .
+e
, L.. J ' - J L J
'-J
I
)=1 ) =0
•
An
ARX (p,q) model
is
used to capture the input/output relationship,
utilizing the information associated with a "known" input provided by
a piezoelectric active-sensing system.
• A proper model order selection
is
critically important.
A
Los Alamos
I 0 I School
of
'.
Jacobs Engineeing
Slide Sl17
Engineering Institute
UNCLASSIFIED
SHM using Time Series Models
• Damage sensitive feature extractions
- Residual Errors: Use the time series predictive model estimated
from the baseline condition to predict the response of data
obtained from a
potentially damaged structural condition.
- ARX Parameters
• Have been extensively used passive-sensing SHM
techniques.
• Some advantages:
- Applicable
to
nonlinearity detection
- A well-established topic
(e
.g. speech pattem recognition)
- Algorithms can be
easily embedded into digital signal processers
A
Los Alamos
U
CSD
I School
of
Jacobs Engineering
Slide 6/17
Engineering Institute
3