Structural%20Damage%20Identification

1
Structural Damage Identification

• LOW FREQUENCY TECHNIQUES

Patrick Guillaume
2
Damage Detection Philosophy

• Rytter (1993,) introduced a damage state
  classification
• system which has been widely accepted by the
  community
• dealing with damage detection and SHM. Following
  these
• lines, the damage state is described by answering
  the following
• questions (Sohn et al. (2003))
• (1) Is there damage in the system? (existence)
• (2) Where is the damage in the structure?
  (location)
• (3) What kind of damage is present? (type)
• (4) How severe is the damage? (extent)
• (5) How much useful life remains? (prognosis)

3
The Monitoring Process (1/2)

• 1. Operational Evaluation
• Operational evaluation answers questions related
  to the
• damage detection system implementation, such as
  economic
• issues, possible failure modes, operational and
  environmental
• conditions and data acquisition related
  limitations.
• 2. Data Acquisition, Fusion and Cleansing
• Data acquisition is concerned with the quantities
  to be measured,
• the type and quantity of sensors to be used, the
  locations
• where these sensors are to be placed, sensor
  resolution,
• bandwidth, and hardware.
• Data fusion, as a discipline of SHM, is the
  ability to integrate
• data acquired from the various sensors in the
  measurement
• chain.
• Data cleansing is the process of selecting
  significant data
• from the multitude of information, i.e., the
  determination of
• which data is necessary (or useful) in the
  feature selection
• process.

4
The Monitoring Process (2/2)

• 3. Feature Extraction and Information
  Condensation
• Feature extraction is the process of identifying
  damage sensitive properties,
• which allow one to distinguish between the
  damaged and undamaged
• structural states.
• Information condensation becomes increasingly
  advantageous
• and necessary as the quantity of data increases,
  particularly
• if comparisons are to be made between sets of
  data
• obtained over the life cycle of a system.
• 4. Statistical Model Development for Feature
  Discrimination
• An important issue in the development of
  statistical models
• is to establish the model features sensitivity to
  damage
• and to predict false damage identification.

5
Damage in Composite Materials

• The use of fibre reinforced plastics (FRP) as an
  alternative
• to conventional materials, such as metallic
  alloys, is undergoing
• increasing growth, especially in the
  aeronautical, naval
• and automotive industries, because of their
  excellent mechanical
• properties, low density and easy of shaping.
• Composite materials possess specific strengths
  and Youngs
• moduli many times greater than those of the most
  widely used
• metallic materials, such as steel, aluminum and
  titanium.
• However, the extreme sensitivity of composite
  materials to
• impact loads constitutes a hindrance to their
  utilization. In
• aeronautical structures, for example, the
  components may have
• to undergo (i) low energy impacts caused by
  dropped tools or
• mishandling during assembly and maintenance, (ii)
  medium
• energy impacts caused in-service by foreign
  objects such as
• stones or birds and (iii) high energy impacts
  caused by military
• projectiles (Matthews (1999), Silva (2001) and
  Carvalho
• (2003)).

6
Some Damage Detection Techniques

• Static Stiffness Variations
• Linear stiffness in function of the number of
  fatigue cycles

7
Some Damage Detection Techniques

• Natural Frequencies and FRFs
• The development of modal analysis techniques for
  damage
• detection and SHM arose from the observation that
  changes
• in the structural properties have consequences on
  the natural
• frequencies. Nevertheless, the relatively low
  sensitivity of
• natural frequency to damage requires high levels
  of damage
• and measurements made with high accuracy in order
  to
• achieve reliable results. Moreover, the capacity
  to locate
• damage is somewhat limited, as natural
  frequencies are global
• parameters and modes can only be associated with
  local
• responses at high frequencies.

8
Some Damage Detection Techniques

• Natural Frequencies and FRFs
• Messina et al. (1992) proposed the damage
  location assurance
• criterion (DLAC) in location j, which is a
  correlation
• similar to the modal assurance criterion (MAC) of
  Allemang
• and Brown (1982), and is given by
• the experimental frequency shift vector
• the analytical frequency shift vector
• A zero value indicates no correlation and a unity
  value indicates
• perfect correlation between the vectors involved
  in the DLAC
• relationship. Damage location and dimension is
  identified
• by maximizing this objective function.

9
Some Damage Detection Techniques

• Natural Frequencies and FRFs
• Zang et al. (2003a) presented criteria's to
  correlate
• measured frequency responses from multiple
  sensors and proposed
• using them as indicators for structural damage
  detection.
• One criterion is the global shape correlation
  (GSC)
• function, which is sensitive to mode shape
  differences but
• not to relative scales, being defined as
• is a column of FRF baseline data
• is a column of the current measured FRF data

10
Some Damage Detection Techniques

• Mode Shape Changes (MAC)
• The MAC value (Modal Assurance Criteria),
• which is probably the most common way of
  establishing a
• correlation between experimental and analytical
  models, is
• defined by Allemang and Brown (1982) as
• West (1984) uses the MAC to determine the level
  of correlation
• between modes from the test of an undamaged Space
  Shuttle Orbiter
• body flap and the modes from the test of the flap
  after loading.

11
Some Damage Detection Techniques

• Mode Shape Changes (COMAC)
• Although the MAC can provide a good indication of
  the disparity
• between two sets of data, it does not show
  explicitly where in the
• structure is the source of discrepancy. The
  co-ordinate MAC
• (COMAC) has been developed from the original MAC.
  It is
• the reverse of the MAC in that it measures the
  correlation at
• each degree-of-freedom (DOF) averaged over the
  set of correlated
• mode pairs. The COMAC identifies the co-ordinates
• at which two sets of mode shapes do not agree,
  and is
• defined as (Lieven and Ewins (1988))

12
Some Damage Detection Techniques

• Mode Shape Curvature
• As an alternative to the use of mode shapes as
  damage features, mode shape
• curvature (i.e., second order derivative) can be
  used to obtain spatial information
• about vibration changes. It has been reported in
  literature that absolute changes in
• mode shape curvature can be a good indicator of
  structural damage (Pandey et al.,
• 1991).

13
Some Damage Detection Techniques

• Changes in Modal Flexibility
• Another class of damage identification methods
  makes use of the dynamically
• measured flexibility matrix to estimate changes
  in the static behaviour of the structure.
• Typically, damage is assessed by comparing the
  measured flexibility matrix, computed
• on a basis of the reference modal data, to the
  measured flexibility matrix
• computed on a basis of the damaged condition.

14
Some Damage Detection Techniques

• Changes in Strain Energy
• Instead of using mode shape curvature directly,
  derived quantities, such as strain
• energy, can be chosen as damage features. The
  Changes in Strain Energy (CSE)
• method localizes structural damage as a decrease
  in modal strain energy between
• 2 structural DOFs (Stubbs et al., 1992). For a
  linear elastic beam structure, the
• strain energy can be computed on a basis of the
  mode shape curvature. In that
• case, the damage index for element i centered
  around DOF i, can be written as

15
Some Damage Detection Techniques

• Sensitivity-based Approach
• The last method is based on the interpretation of
  changes in natural frequency
• by means of the sensitivity of the natural
  frequencies of a reference condition to
• local mass changes in an experimental DOF or
  local stiffness changes between two
• adjacent DOFs. The method requires both the
  natural frequencies and normalized
• mode shape estimates of the structure in its
  reference condition as well as the
• corresponding natural frequency estimates of the
  damaged condition.

16
Some Damage Detection Techniques

• Sensitivity-based Approach
• Combined approach

17
Some Damage Detection Techniques

• Sensitivity-based Approach
• Iterative Weighted Pseudo Inverse

18
Experimental Validation
19
Experimental Validation
20
Experimental Validation

• MAC

21
Experimental Validation

• COMAC

22
Experimental Validation

• Mode Shape Curvature

23
Experimental Validation

• Strain Energy

24
Experimental Validation

• Modal Flexibility

25
Experimental Validation

• Sensitivity-based Approach

26
Experimental Validation

• Sensitivity-based Approach (combined approach)

27
Experimental Validation

• Sensitivity-based Approach (combined approach)
• Iterative Weighted Pseudo Inverse

28
Experimental Validation
29
Experimental Validation
30
Experimental Validation
31
Experimental Validation
32
Experimental Validation