Scribd
Upload
84 views13 pages

Introduction To Structural Health Monitoring and Feature Extraction

This document provides an introduction to structural health monitoring and feature extraction. It discusses the motivation for structural health monitoring, including moving from time-based to condition-based maintenance. It also defines damage, classifies damage identification methods, and outlines the structural health monitoring process which includes operational evaluation, data acquisition, feature extraction, and statistical model development. Feature extraction is described as transforming data into damage-related information.

Uploaded by

Hussam Suhail Saleem
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
84 views13 pages

Introduction To Structural Health Monitoring and Feature Extraction

This document provides an introduction to structural health monitoring and feature extraction. It discusses the motivation for structural health monitoring, including moving from time-based to condition-based maintenance. It also defines damage, classifies damage identification methods, and outlines the structural health monitoring process which includes operational evaluation, data acquisition, feature extraction, and statistical model development. Feature extraction is described as transforming data into damage-related information.

Uploaded by

Hussam Suhail Saleem
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

LA-UR-07-3231

Unclassified

Introduction to Structural Health Monitoring and Feature Extraction


Charles R. Farrar Michael D. Todd Presented at Engineering Institute Workshop July 25th, 2006

Engineering Institute

LA-UR-07-3231

Unclassified

Structural Health Monitoring

Engineering Institute

LA-UR-07-3231

Unclassified

Motivation for Structural Health Monitoring


Move from time-based maintenance to condition-based maintenance Combat asset readiness N New b i business models d l
Manufacturers of large capital investment hardware can charge by the amount of life used instead of a time-based lease.

The potential for economic and life-safety benefits coupled with need to integrate diverse technologies makes Structural Health Monitoring a Grand Challenge problem for aerospace, civil and mechanical engineers i th 21st century i il d h i l i in the t
Engineering Institute

LA-UR-07-3231

Unclassified

Definition of Damage
Damage is defined as changes to the material and/or geometric properties of a structural that adversely affect its performance.

All materials used in engineering systems have some i h t h inherent t initial flaws. Under environmental and operational loading flaws will grow and coalesce to produce component level failure. Further loading causes systemsystem level failure. Must consider the length and time scales associated with damage evolution. d l i
Engineering Institute

LA-UR-07-3231

Unclassified

Classifying Damage Identification Methods


Damage Identification Microscopic flaw/damage identification
Used to develop material failure models

Incipient, macroscopic, material/component level damage


Non-destructive evaluation (local, off-line inspection) Wave propagation based structural health monitoring (more global, Wave-propagation-based global on-line)

Component damage/failure system level damage


Structural health monitoring g Condition monitoring (applied to rotating machinery) Health and usage monitoring systems (HUMS, Rotor craft) Statistical process control (monitors system processes where damage can be one cause of loss of process control)

Damage Prognosis Adds prediction of remaining life capability to SHM

Engineering Institute

LA-UR-07-3231

Unclassified

Structural Monitoring

SHM
183 6 R 21 20 19 18 17 16 15 14 12 Ch. 4 Ch. 11 16 Floor
th

Ch. 2 Ch. 10 Roof Ch. 3 57 10 57 10

Ch. 5

177 3

11 10 9 8 7 6 Ch. 8 Ch. 13 3RD Floor 2 G B Basement Ch. 1 Ch. 9 Ch. 6 Ch. 12 9th Floor Ch. 7

Structural monitoring is not used to make decisions regarding system operation and maintenance
Engineering Institute

5 4 3

LA-UR-07-3231

Unclassified

SHM is a Problem in Pattern Recognition

Engineering Institute

LA-UR-07-3231

Unclassified

The Structural Health Monitoring Process


1. Operational evaluation
Defines the damage to be detected and begins to answer questions regarding implementation issues for a structural health monitoring system.

2. Data acquisition
Defines the sensing hardware and the data to be used in the feature extraction process.

3. Feature extraction
The process of identifying damage-related damage related information from measured data.

4. Statistical model development for feature discrimination


Classifies feature distributions into damaged or undamaged category.

Data Cleansing Data Normalization Data Fusion Data Compression (implemented by software and/or ft d/ hardware)

Engineering Institute

LA-UR-07-3231

Unclassified

Introduction to Features
What is a Feature? A feature is some characteristic of the measured response that is extracted via signal processing parameter estimation or some processing, other signal inspection technique Feature extraction transforms data into information It is desirable to have examples of the features from both p damaged and undamaged structures Primary Characteristics of features Sensitivity - Feature should ideally be very sensitive to damage and completely insensitive to everything else (rarely occurs) Dimensionality - Want the feature to have the lowest dimension possible Computational Requirements - Features should be computable p q p with minimal assumptions and CPU cycles

Want to use the simplest feature possible that can distinguish between the damaged and undamaged system t
Engineering Institute

LA-UR-07-3231

Unclassified

Feature Extraction
Approaches to identifying damage-sensitive features.
P t experience Past i Component and system testing Numerical analysis to simulate damaged system response

Features types.
Absolute (derived from single data source, e.g. modal frequency) Relative (derived from multiple data sources, e.g. mode shape)

Damage sensitive features fall into three categories.


Waveform or image comparison g p Model parameters Residual errors between measured and predicted response.

Engineering Institute

LA-UR-07-3231

Unclassified

Feature Extraction
Want many samples of low-dimension feature vectors. The need for low-dimension feature vectors often necessitates some form of information condensation (e.g. compression of accel.-time histories into modal properties). Apply data fusion techniques to extract features from multiple and possibly heterogeneous sources (estimation mode shapes). Quantify features sensitivity to damage feature s damage. Ideally, the features should change monotonically with g damage level. Identify and quantify sources of feature variability. Incorporate feedback from data acquisition and statistical model development portions of the process process.
Engineering Institute

LA-UR-07-3231

Unclassified

Data Normalization
Signal 1 1000 500 Data Point # = 26980 Time Period = 601.17 sec t = 0.02228 sec

Strain

0 500 1000 1000 500 0 100 200 300 Signal 2 400 500 600 700

Strain

0 500 1000 1000 500 0 100 200 300 Signal 3 400 500 600 700

Strain

0 500 1000 0 100 200 300 400 Time (Sec) 500 600 700

7.65 20.35 7.50 First Frequ uency (Hz) 14.7 M ean Frequency Value (Top West Outdoor - Top East Outdoor ) 18.3

7.35

7.45 1.9 -0.2 3 0.6 0

1.1 7.20 -5.1 7.05 -1

0.5

9:15 11:30 13:12 15:13 17:52 20:09 21:20 23:29 1:21 3:19 5:19 7:03 9:22

Engineering Institute

LA-UR-07-3231

Unclassified

Features are Used to Answer the Following:


1 Is the system damaged?
Group classification problem for supervised learning Identification of outliers for unsupervised learning p g

2 Where is the damage located?


Group classification or regression analysis problem for supervised learning Identification of outliers for unsupervised learning

3 What type of damage is present?


Can only be answered in a supervised learning mode G Group classification l ifi ti

4 What is the extent of damage?


Can only be answered in a supervised learning mode G Group classification or regression analysis l ifi ti i l i

5 What is the remaining useful life of the structure? (Prognosis)


Can only be answered in a supervised learning mode R Regression analysis i l i Engineering Institute

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy