Appendix B A Review of State-of-the-Art Techniques For Real-Time Damage Assessment of Bridges

Download as pdf or txt
Download as pdf or txt
You are on page 1of 25

ETL 1110-2-551

31 Aug 98

the acoustic emission technique was suggested in


the 1970s to monitor a military bridge (Pollock
and Smith 1972) and to detect discontinuities in
steel highway bridges (Hutton and Skorpik 1975),
it was only successful in a limited number of
instances (Fisher and Wood 1988). The reason for
this may be multi-fold since NDE is an interdisciplinary technology. It involves mechanical,
electromagnetic, acoustic, and optical techniques
to evaluate the integrity of a structure. When it is
used for civil engineering applications, such as
bridge damage assessment, the discontinuities to
be detected may be large in dimension but complex in structure. Moreover, the inspections are
subjected to environmental influence such as
weather and noise. Accessibility is also more of a
challenge when performing field NDE on inservice bridges as opposed to inspecting items
mass produced in a factory. In addition, there is
less of a tendency for bridge features to be
standardized relative to mass produced factory
items. Consequently, standard NDE procedures
for bridges typically do not exist but must be
uniquely developed for each application. Nevertheless, with the combined strength of both the
NDE tools and the civil engineering professionals,
progress is being achieved in damage assessment
of bridges.

Appendix B
A Review of State-of-the-Art
Techniques for Real-Time Damage
Assessment of Bridges

B-1. Introduction
a. It was reported that, as of November 1991,
35 percent of approximately 590,000 bridges in
the United States were considered structurally
deficient or functionally obsolete (Bagdasarian
1994). Many bridges have become deficient due
to increased age and larger than expected service
loads. In addition, some highway and railroad
bridges ranging from 50 to more than 100 years
old are still performing their intended function in
spite of excessive use (Scalzi 1988). AASHTO
has developed a Manual for Condition Evaluation of Bridges (1994), which provides for
uniformity in the procedures and policies for
determining the physical condition, maintenance
needs, and load capacity of highway bridges.
Recent bridge collapse or near collapse has
focused the need to develop extensive nondestructive evaluation (NDE) techniques for real-time
structural damage assessment to guarantee the
safety of our nationwide transportation system.
Real-time NDE techniques can immediately
provide information such as size, shape, location,
and orientation of discontinuities as part of the
structural damage assessment.

d. Appendix B provides a brief review of


NDE fundamentals and discusses their application
towards evaluating fracture critical members
(FCMs) on bridge structures. The principles and
general applications of each technique are
emphasized along with newly reported bridge
field testing applications. Detailed knowledge
pertaining to the physical bases and instrument
operation procedures for each technique can be
found in Krautkrmer and Krautkrmer (1977)
and Bray and McBride (1992).

b. NDE techniques for material inspection


have been well-known for many years (Krautkrmer and Kratkrmer 1977; Lord 1980; Lew
1988; Bray and McBride 1992). They include
liquid penetrant, eddy current, radiography, magnetic particle, ultrasonic, acoustic emission, and
dynamic property measurement methods. Among
those, ultrasonic and acoustic emission technologies have become the most popular and frequently
used to perform real-time inspection. The
dynamic property measurement method has also
been used to evaluate the integrity of structures.

B-2. Some New NDE Techniques in RealTime Structural Damage Assessment


a. Eddy current method.
(1) Principle.

c. Applications of dynamic property measurement techniques in civil engineering (e.g., buildings, bridges, and dams) are rather rare. Although

(a) The eddy current method is based on the


fundamental work of Farady, Oersted, and
B-1

ETL 1110-2-551
31 Aug 98

Maxwell (Libby 1971). An electrical current


flowing in a wire generates an electromagnetic
field about the wire. The electro-magnetic field
becomes concentrated when the wire is wound in
the form of a coil. Such coils are used for the
eddy current testing of materials. When an
energized coil is placed near the surface of
metallic material, eddy currents are induced in the
material. Coil current must be alternating (ac),
since relative motion between the field and
conductor is required to generate or induce
electricity. Current induced in the metal flows in
a direction opposite to the current in the coil.
Material properties as well as discontinuities, such
as cracks or voids, will affect the magnitude and
phase of the induced current. Thus, with the aid
of suitable instrumentation, eddy currents can
assess the material conditions. Typical test coil
arrangements are shown in Figure B-1.

(2) Applications.
(a) As early as the late 1930s and early 1940s,
investigators began the application of eddy current
techniques to materials evaluation problems.
Commercial instruments became available during
this time and were used extensively during World
War II.
(b) With the rapid development of electronics
and computer science, presently the eddy current
technology can be utilized to assess selected
material properties as well as locate discrete
discontinuities in metallic structures such as
aircraft (Hugemaier 1991), coolant channel
assemblies of nuclear reactor (Bhole et al. 1993),
large diameter pipeline steels (Nestleroth 1993),
etc. A list of selected applications is given in
Table B-1.

(b) The eddy current inspection method has


numerous favorable characteristics that make it
the proper choice for many inspection tasks.
Primary among these advantages is that
mechanical contact is not required between the
eddy current transducers and the test articles.
Eddy current penetration depth and, consequently,
inspection depth can be controlled by adjusting
the frequency of energizing current. The method
has high sensitivity to small discontinuities.
Dimensional measurements and electrical
conductivity measurements can also be made.
Instrumentation for eddy current testing is
relatively low cost for most applications. The
equipment can be automated for high-speed
testing with lightweight portable instruments.

(c) The establishment of accept/reject criteria


for discontinuities is a matter of specifying how
many discontinuities are acceptable, what size and
how close together they are, if they can be allowed
in engineering components of each class, material
thickness, type of material, type of weld, size of
structure, and service condition. Acceptance
criteria are highly item oriented.
(3) Inspection of bridges.
(a) Although not commonly used in inspecting
steel bridges, eddy current method was reported to
have been used to inspect steel bridges in Japan
(Kishi and Ohtsu 1988). The onsite eddy current
system used for steel highway bridges has advantages such as, clear reproduction of results, high
inspection speed, no coating removal required,
and feasiblity for noncontact applications.
However, there are still some problems in using
eddy current methods to inspect welded members
in steel bridges. One problem is the presence of
noise in the electric current during field testing.
Electronic noise signals can hamper the
inspectors ability to identify discontinuities.
Another concern is the significant disturbance
from the spatial distribution of welded

(c) Eddy current techniques are limited in that


only electrically conductive materials can be
tested. There is limited material penetration with
high-frequency energy, and discontinuity
indications are largely qualitative. The many
material, geometric, and electronic parameters
affecting test results often complicate data
interpretation. Thus, considerable care must be
exercised in selecting eddy current techniques and
in evaluating inspection results to avoid
interpretation errors.

B-2

ETL 1110-2-551
31 Aug 98

Figure B-1. Arrangements of eddy current test probes and test objects for (a) probe on one side of test
object, (b) probe encircling test object, and (c) excitation and pick-up on opposite sides of test object

B-3

ETL 1110-2-551
31 Aug 98

compare to measured dynamic response with


either the dynamic response predicted by an
analytical model or the previously measured
dynamic response.

Table B-1
Typical Applications of Eddy Current Inspection
Material Property Determinations
Heat treatment evaluation
Hardness measurements
Fire damage determinations
Impurity content measurement

(b) Theoretically, any discontinuities, cracks


and other variations in structural properties will
alter the vibration characteristics of the structures.
Changes in vibration measurement may be used
for structural damage assessment and monitoring
bridge integrity. However, difficulties in
modeling the restraint from supports as well as
other modeling difficulties can preclude the
comparison of modeled versus measured dynamic
response to identify the subtle effects of cracks. In
addition, environmental effects such as thermal
expansion or debris collecting at expansion joints
can also overshadow subtle changes in dynamic
response due to the presence of cracks when
comparing dynamic response to previous
responses.

Discontinuity Detection
Sheet metal
Foil
Wire
Bar
Tube Testing
Bolt inspection
Weld inspection
Ball bearings tests

reinforcement, which makes the separation


between signal and noise also difficult to interpret.
These problems are circumvented by selfcompensating the permeability difference between
weld metal and base metal and by minimizing
noise signals due to reinforcement distribution.

(c) Some fundamental modes of vibration for a


very simple structure are illustrated in Figure B-2,
which shows a fundamental flexural (a) and
longitudinal (b) mode of vibration of a rectangular
bar. For the flexural mode, the particle excursions
in the material are vertical through the bar length;
conversely, the particle excursions for the
longitudinal mode are in the direction of the bar
length. The natural vibration frequencies for these
two modes can be written respectively as (for a
square bar):

(b) Eddy current signals from different types of


discontinuities in steel girders are shown to be
remarkably different, which may give useful information pertaining to structural damage assessment. Unfortunately, detailed information on
measurements and interpretations of these signals
have not been reported. Additional information
on the eddy current method can be found in Bray
and Stanley (1989) and AWS (1980).

fflex. =

b. Vibration dynamic method.

20.2h
l

E
'

(B-1)

(1) Principle.
(a) Certain properties of structures can be
evaluated using vibration dynamic techniques. It
is well-known that a structure possesses certain
natural vibration frequencies and mode shapes. If
friction is taken into account, the vibrations that
are already excited will decrease gradually. This
is called a damping vibration. The dynamic
vibration tests are carried out by applying a known
forced vibration to the structure and observing its
vibration response. The fundamental concept is to

flong =

h
2l

E
'

where
f = natural vibration frequency (Hz)
h = height or depth of the bar (in.)

B-4

(B-2)

ETL 1110-2-551
31 Aug 98

Figure B-2. Fundamental mode of vibration of a rectangular bar

B-5

ETL 1110-2-551
31 Aug 98

small changes in eigenvalues (7 2) and in


vibration mode shapes { 1}, making the equation
for the perturbed form

l = length
' = density

[(K  K) (72  (72)) M]{1  1} = 0

If the natural vibration frequency is measured, the


Youngs Modulus of the bar can be obtained from
Equation B-3, which suggests that Youngs
Modulus E is closely related to the natural
vibration frequency. Youngs Modulus is one of
the important parameters for bar-like members
such as beams in bridges. Youngs Modulus for a
square bare is calculated by
E =

0.00245f 2l 4'
h2

Equation B-5 clearly shows the relation of the


stiffness change with the changes of the natural
frequency and mode shape.
(2) Applications.
(a) A partial list of physical quantities that may
be measured by these techniques is given in
Table B-2. The experimental results of the
dynamic Youngs Modulus by the vibration
technique can be found in (Wolfenden et al.
1989).

(B-3)

(d) Real solids are never perfect; therefore,


some of their mechanical energy is always
converted into heat. This will lead to a vibration
damping. Measurements of the vibration
damping may provide reliable information for
structural damage assessment.

Table B-2
Physical Quantities Typically Measured by Resonance
Vibration
Length, width, thickness diameter
Modulus of elasticity
Shear Modulus
Poissons Ratio
Density
Modulus of rupture
Discontinuities or other inhomogeneities

(e) Mode shape is related to the structure


properties. For the ideal solid bar mentioned
above, the mode shape is very narrow with a high
peak. But for more complicated structures, such
as bridges, the mode shape depends on the
mechanical properties and the geometric form of
the structure.

(b) Resonance of a structure is reached when


the frequency of an applied vibration force
produced by piezoelectric transducer or
electromagnetic vibration matches the natural
frequency of vibration of the structure. Damage
which plastically deforms a member would also be
expected to change the stiffness and the frequency
of resonance. A loss of stiffness may be
detectable as decrease in the observed resonant
frequency of the member.

(f) Hearn and Testa (1993) present a general


equation of free vibration motion for an undamped
elastic structure.
[K 72 M] {1} = o

(B-5)

(B-4)

where
K = stiffness matrix

(c) Many types of discontinuities and defects


have been reported to have been detected in
structures using the vibration dynamic method.
This capability is usually carried out by the
vibration interrogation of an undamaged structure
or assembly at sufficient frequency levels and
modes to establish trend data. By comparing the
vibration scan of an identical part with this known

M = mass matrix
7 = resonant frequency
1 = the vibration mode shape
A perturbation of the structure is considered in
which a small change in stiffness [ K] produces
B-6

ETL 1110-2-551
31 Aug 98

reference, or by comparing the vibration scan for a


given structure with time, one can detect whether
there has been a change. The relative magnitude
of change can be an indication of damage.
Vibration dynamic techniques have been
developed specifically to detect discontinuities
and defects in various materials.

(c) The equipment brought to the bridge site


operated solely on batteries and included four
accelerometers, a 53.38 N (12-lb) impact hammer,
and a seven-channel frequency modulated tape
recorder. The bridge was instrumented on three
trips: first, for impact testing along the bridge
centerline; then, for ambient vibration on a windy
day; and lastly, for impact testing along one of the
two main girders.

(d) A successful technique relevant to the


vibration testing method was developed by West
(1982 and 1986) for the space shuttle orbiter body
flap test specimen. Several damage sites that were
not detected by conventional NDE techniques
were correctly identified by this technique.
Wests technique can locate damage in certain
types of structures reasonably well, but it is unable
to determine the extent of the damage.

(d) The impact signal and two vertical


acceleration signals were stored during the
inspection on three channels of the tape recorder
for further analysis in the laboratory. For center
line testing, the hammer and accelerometers were
located at the center of nine cross stringers located
5.64 m (18.5 ft) apart. For the torsion test, the
impact hammer and accelerometers were located
on two main girders at the ends of the nine cross
members joining the two main girders. A fourchannel spectrum analyzer was used to perform
the data analysis.

(e) Two areas that have received considerable


attention in recent years are civil engineering
structures and offshore oil platforms (Yao 1982
and Yoa et al. 1982). The goal of the work was to
define a method of assessing the structural
integrity of buildings after their exposure to
overload.

(e) An average of three or four impact tests


were used to arrive at a frequency response
function of a particular accelerometer subjected to
an impact load at another or the same location.
These were then combined in the polyreference
technique to obtain a best fit of the sum of
complex exponential to the inverse Fourier
transform of these frequency response functions.
The natural frequencies of the bridge for the
vertical and torsional vibrations are calculated by
a similar method given by Hearn and Testa
(1993). The experimentally obtained resonance
frequencies are consistent with the theoretical
results. The study indicates that the vibration
dynamic method can be used to assess bridges.
However, no further structural damage
assessments using this method have been reported
by the same investigators.

(3) Inspection of bridges.


(a) Civil engineering structures such as
bridges, buildings, and dams have many natural
frequencies below 100 Hz (Billing 1984). A highpowered hammer has been used to excite the
resonant vibrations of these structures. Partly
based on this information, Beliveau and Huston
used an impact hammer as an exciting source to
test a full-scale pedestrian bridge (Beliveau 1987
and Beliveau and Hutson 1988).
(b) The bridge tested is located on a bicycle
path over Route VT 127 North of Burlington,
Vermont. It has a span of 54.86 m (180 ft)
between abutments and a treated timer deck width
of 3.40 m (11 ft 2 in.) between center lines of the
two main 0.91-m (36-in.) girders and cable-stay
system. The cables have a diameter of 34.93 mm
(1-3/8 in.), and the bottom flanges of the girders
are 0.56 m (1 in. by 10 in.) plate.

(f) The dynamic method has also been used


for safety inspection of prestressed bridges in
Austria (Flesch et al. 1988). The basic concept is
the same: damage of the structure will lead to
deviations of the dynamic parameters from the

B-7

ETL 1110-2-551
31 Aug 98

virgin state. These deviations can be used in a


global manner to assess damage. In this report,
emphasis was put on the aspects of theoretical
model and software.

(j) The signals received by sensors in time


domain from the vibrating bridge are transformed
into the frequency domain by the monitoring
system software. The frequency spectra are
imported into the DADiSP program, where they
are analyzed to determine the natural frequency
peaks for monitoring purposes. Figure B-4 shows
the natural frequencies and mode shapes of signals
from sensors at different locations. Spectrum
clean-up techniques are designed to enhance the
appearance of the frequency spectrum. Cleaning a
spectrum enables one to identify natural
frequencies with less difficulty, thus allowing for
the determination of changes in the natural
frequencies indicating structural deterioration of
the bridge.

(g) Recently, Bagdasarian (1994) reported new


progress on assessing the structural integrity of
bridges through vibration monitoring. Recent
studies at the University of Connecticut have
shown that monitoring a bridges dynamic
characteristics is feasible. Laboratory testing on a
bridge model developed a bridge signature
comprising the bridges natural frequencies and
mode shapes. Changes in the signature
correspond to changes in the models structural
stiffness. Therefore, these components of the
signature (the natural frequencies and mode
shapes) could be used to evaluate the structural
condition of the bridge.

c. Ultrasonic testing method.


(1) Principle.

(h) A prototype monitoring system was developed by Vibra-Metrics, Inc., for placement on an
actual Connecticut bridge. The monitoring system
consists of sixteen accelerometers, two cluster
boxes, and a sentry unit that houses a computer.

(a) Ultrasonic waves are simply vibration


waves with a frequency higher than the hearing
range of the normal human ear (i.e., 20 kHz). The
pioneer work in ultrasonic testing was
accomplished by Langevin of France, who
experimented with ultrasonic submarine detection
methods during World War II. Now, most
practical ultrasonic discontinuity detection is
carried out with frequencies from 200 kHz to
20 MHz. Ultrasonic waves with 50 MHz or

(i) The accelerometers act as sensors for


detecting the bridges vibrations. They are
magnetically attached to the bridge girders and
positioned throughout the floor plan of the bridge
as shown in Figure B-3.

Figure B-3. Bridge floor plan with accelerometer locations

B-8

ETL 1110-2-551
31 Aug 98

Figure B-4. Typical worksheet

higher frequencies are sometimes used in material


property investigations.

pulse on the CR tube screen is from the exciting


pulse, and the second pulse measures the transit
time in the material sample. When a discontinuity
exists in the material, the pulse due to the
discontinuity reflection can be found, as shown in
Figure B-5(b). The intensity of the flow echo is
directly related to the discontinuity properties, and
the position of the discontinuity echo reflects its
locations. Variance in the discontinuity properties
of fatigue cracks in bridge beams using ultrasonic
methods has been reported by Hearn and Cavallin
(1997). In most cases, one ultrasonic transducer
(or probe) is used in the inspection. This method
is also called A-scan presentation, furnishing a
one-dimensional description for a given test point
Figure B-6(a).

(b) Ultrasonic inspection is accomplished by


using electronically controlled pulses introduced
into a material through a transducer. The ultrasonic energy then travels within the material,
finally reaching an outer surface where the ultrasonic waves are received by the same or another
ultrasonic transducer. Materials with discontinuities are diagnosed from the characteristic of
the received ultrasonic energy. In this method, the
wave intensity and the transit time are measured
(e.g., on the screen of a CR tube as shown in
Figure B-5). Perfect material without defects is
assumed in Figure B-5(a), where the first

B-9

ETL 1110-2-551
31 Aug 98

Figure B-5. Intensity transit-time or pulse transit-time method, (a) with sound transmission, (b) with
reflection

(c) In the so-called B-scan method, the


location (depth) of a discontinuity in a specimen is
represented by echo position (usually along the
vertical direction) and the amplitude of the
discontinuity echo by the brightness, as shown in
Figure B-6(b). In the case of two-dimensional
(area) scanning of a test piece (e.g., a plate), the
test results can be presented by means of a C-scan
Figure B-6(c). This method furnishes a top view
of the test piece from the scanned surface with
plotted flaw projection points.
(d) Some of the advantages of ultrasonic methods are as follows: discontinuities can be detected
in metallic and nonmetallic materials; discontinuity distance may be measured from the material
surface; discontinuities can be located in very
thick materials; only single-surface accessibility is
required; both internal and surface discontinuities
may be detected; discontinuity imaging is
possible, and material properties can be measured.

The ultrasonic technique has rapid testing


capabilities, and portable instrumentation is
available for field testing. Equipment for
automatically recording inspection results is
available, and the inspection costs are relatively
low.
(e) Conversely, there are some disadvantages
of ultrasonic testing: there may be difficulties
incoupling energy to rough surface; it may be
impractical to inspect complex shapes; flaw
imaging is complex; and special scanning systems
may be required for inspecting large surfaces.
(2) Applications.
(a) Computer-controlled multifunctional ultrasonic instruments for detecting discontinuities in
materials have been highly developed (Wooh and
Daniel 1994). These techniques can be used to
detect cracks, voids, and other abrupt
B-10

ETL 1110-2-551
31 Aug 98

Figure B-6. A-, B-, and C-scan presentation and scanning method

discontinuity or lack of homogeneity in metallic,


nonmetallic, and composite materials
(Krautkrmer and Krautkrmer 1977, Bray and
McBride 1992, Wooh and Daniel 1994,
Thavasimuthu et al. 1993, Cruby and Colbrook
1992, and Berger 1992). On-line weld monitoring
using ultrasonics is also well developed (Stares
et al. 1990; Bull et al. 1995; and Prikhodko and
Fedorishin 1993).
(b) Major limitations of these techniques
involve attenuation characteristics of certain
materials, access problems, very tight cracks, and
complex geometric configurations. Even with
these limitations, ultrasonic discontinuity
evaluation technology is very effective when
properly applied. Some typical applications are
listed in Table B-3.
Table B-3
Some Sources of Acoustic Emissions
Raw materials
Weldments
Castings, forging
Pipe
Seamless tubes
Railroad wheels, rails, and axles

Adhesive bonds
Aircraft
Spacecraft
Nuclear reactors
Ships
Bridges

(c) Examples of applications which are related


to the structural inspection of bridges include the
following:
 Normal-beam inspection of bars and
plates. One of the most common tests performed
with ultrasound is the inspection of structural
plates and bars for interior discontinuities and
corrosion using a normal-beam and longitudinalwave probe. These tests are applied to both plain
material and fabricated members.
A typical A-scan, digitized RF display of a
normal-beam inspection is shown in Figure B-7(a). The test sample is an aluminum bar
76 mm thick. With a sampling rate of 20 MHz
and an expansion of 1, the time base is 36 s. The
echo occurring at 15.7 s (point A) past the main
bang indicates a discontinuity at approximately
49 mm below the probe. The back echo appears at
24 s (point B) beyond the main bang.
The horizontal bar above the discontinuity echo
(point A) indicates a gate starting at 17.8 s and
2.95 s in length. One use of a gate is to select a
signal for frequency analysis as shown in Figure B-7(b). The output shows the peak power
B-11

ETL 1110-2-551
31 Aug 98

Figure B-7(a). Typical digitized RF display of test of aluminum bar 76 mm (3 in.) thick. Time base is
3.6 E-6x/div.

spectrum frequency to be 5 MHz, which agrees


with the frequency of the probe. Other information shown in the figure describes additional
parameters used in signal analysis (Bray and
Stanley 1984).
 Welded joints. The following types of
manufacturing defects can be detected in welded
butt joints using ultrasonics: slag inclusions,
pores, lack of fusion (cold shuts), lack of
penetration, and cracks. Figure B-8(a) shows the
ultrasonic testing of a welded joint with normal
longitudinal probes, and Figure B-8(b) illustrates
testing of a welded joint with transverse probes.
Figure B-9 shows testing of fillet welds.

allowable discontinuity size in the components is


highly restricted during fabrication. However,
during shipping, handling, and erection of bridge
components, new cracks may grow. Therefore, it
is suggested that bridge components should be
inspected both during and after construction.
(b) In Japan, ultrasonic testing is utilized for
nondestructive in-process evaluation. As a rule,
automatic ultrasonic testing inspection is
performed for welded chord members. Ultrasonic
testing of in-service bridges is time-consuming.
To date, no real-time structural damage
assessment of bridges is reported.
d. Acoustic emission method.

(3) Inspection of bridges.


(a) The ultrasonic techniques are used to
inspect bridge components such as beams, girders,
and chords before a bridge is completed. The

(1) Principle.
(a) Acoustic emission (AE), sometimes called
stress wave emission, is a transient mechanical
B-12

ETL 1110-2-551
31 Aug 98

Figure B-7(b). Power spectrum for ultrasonic signal

Figure B-8(a). Testing of welded joint with normal


probes

vibration generated by the rapid release of energy


from localized sources within materials. Stress or
some other stimulus is required to release or
generate emissions. Emission energy levels can
range from the motion of a few dislocations in
metals to that required to cause catastrophic
cracking of structures. Some stimuli causing
acoustic emissions are given in Table B-4.

Figure B-8(b). Test of welded joint with zigzag


transverse waves

(b) Acoustic emission signals cover a wide


range of energy levels and frequencies but are
usually considered to be of two basic types: burst
and continuous. The term burst is a qualitative
description of emission signals corresponding to
B-13

ETL 1110-2-551
31 Aug 98

Figure B-9. Testing fillet welds; (a) joint not welded through, (b) joint welded through (K-joint)

B-14

ETL 1110-2-551
31 Aug 98
Table 4
Some Sources of Acoustic Emissions
Crack initiation and growth
Dislocation movements
Twinning
Phase changes
Fracture of brittle inclusions or surface films
Fiber breakage, crazing, and delaminations in
composites
Chemical activity

individual emission events. The term continuous


emission is a qualitative description for an apparently sustained signal level from rapidly occurring
emission events. Emission frequencies range from
below to well-above the audible range for humans.
But most practical AE monitoring is accomplished
in the kilohertz or low-megahertz range.
(c) Although emission is characterized as burst
or continuous, signals of either type may propagate in any of the standard ultrasonic modes (i.e.,
shear, longitudinal, or surface waves). Furthermore, a single emission event can generate waves
having more than one propagation mode.

Figure B-10. Generalized count rate versus stress

(d) A wide range of transducer types has been


used to sense acoustic emission from materials,
structures, and industrial equipment. The types of
AE sensors include accelerometers, piezoelectric
transducers, capacitive transducers, optical/laser
sensors, microphones, strain gauges, magneticstrictive sensors, etc.
(e) The most widely used method of quantifying AE signals is the ringdown counting technique, which measures the characteristics of the
emitted signal as its amplitude decays. For a typical sinusoidal AE pulse, an amplitude threshold is
established for the acceptance of signals, and the
number of signals exceeding this threshold is
automatically counted by the instrumentation
system. Signals crossing the threshold are usually
plotted as a function of load, stress, time, or other
parameters. They may be plotted as the count rate
versus stress, or the plot may be of the total or
cumulative count versus the selected parameter.
The data presentation techniques are illustrated in
Figures B-10 and B-11. The significant parameters used in characterizing acoustic emission

Figure B-11. Generalized cumulative count versus


stress

B-15

ETL 1110-2-551
31 Aug 98

events are peak amplitude, frequency, duration of


signal above the selected threshold, number of
counts per event, energy, and rise time. Some of
these parameters are indicated in Figure B-12.
These parameters are related to the various AE
sources in different ways. For example, the AE
count rate associated with plastic deformation at
the tip of a crack may be expressed as (Muravin
et al. 1993)
N =

2
(ChKImax/)Y)

dl/dn

(f) For engineering applications, it is more


convenient to find out the relationship of AE
count rate with the stress intensity factor range
K. For fatigue fracture in materials, Morton et
al. (1974) and Bassim (1987) present the
following equation:
N = A ( K) m

(B-6)

(B-7)

where A and n = experimental constants. This


equation is similar to the well-known Paris law of
fatigue-crack propagation (Paris et al. 1963):

where
dl/dn = C ( K) m

(B-8)

C = a proportionality factor
where
h = thickness of the specimen
dl/dn = crack-growth rate
Kimax = maximum value of the stress intensity
factor

C and m = experimental constants

)Y = yield point
dl/dn = growth rate of the crack length for
loading cycle

(2) Applications.
(a) The AE method has wide applications. It
can be used to monitor changing material

Figure B-12. Parameters used to characterize emission events

B-16

ETL 1110-2-551
31 Aug 98

conditions in real time and to determine the


location of the emission centers as well. Typical
applications include on-board or onsite monitoring
of aircraft, pressure vessels, tank welds, bridges,
and civil engineering structures. In addition,
corrosion and bearings in pumps and other
rotating machinery such as hydraulic valves can be
monitored. Simulated acoustic emission
techniques are also useful for monitoring types of
composite materials. Some recently reported
applications can be found in Ramsamooj (1994),
Yuyama et al. (1994), Glaser and Nelson (1992),
and Fang and Berkouits (1994).
(b) The advantages of AE are rooted in the
basic characteristic where the active defect emits a
signal that will find a path to the monitoring
sensor location. Since it is a passive technique, no
equipment is required to excite a pulse. Further,
the received signals may be recorded for remote or
delayed analysis and for storage. The requirements for equipment mounted on the monitored
structure may be rather small. Other advantages
are that AE techniques are highly sensitive to
crack growth, and locations of growing cracks can
be determined.
(c) Additional advantages are the ability to
monitor an entire system at the same time. With
remote monitoring, the technique can be used in
hostile environments. The item being tested
usually can remain in operation during the
process, and the entire volume of materials and
structures can be inspected at a reasonable cost. It
is also suitable for long-term in-service
monitoring.
(d) Disadvantages of the technique include the
requirement of stress or other stimuli to generate
the acoustic emission event. Therefore, stabilized
cracks cannot be detected with emission techniques. The size of cracks or other defects cannot
be precisely determined. Some materials and
certain tempers of other materials are not very
emissive and are unsuitable for monitoring.
Electrical interference and ambient noise must be
filtered out of emission signals. Also, the multiple
number of travel paths from the source to the

sensor in complex structures can make signal


identification difficult.
(3) Inspection of bridges.
(a) AE techniques were used in the 1970s to
monitor some bridges (Pollock and Smith 1972;
Hutton and Skorpik 1975). In the 1980s,
extensive studies were carried out to use acoustic
emission techniques for bridge inspections (Fisher
and Wood 1988; Hopwood and Prine 1987; and
Green 1988). It was reported that (Hopwood and
Prine 1987) an experimental AE device, the
Acoustic Emission Weld Monitor (AEWM), has
been field tested on six bridges during the study.
The device was also used to test three other
bridges under separate contracts from state highway agencies. The AEWM was evaluated to
determine if it could detect fatigue-crack growth
on in-service steel bridges. The device rejects
high background noise rates typical of bridges and
detects and locates AE activity from known
defects such as cracks and subsurface discontinuities. The AEWM functioned properly in
every field test situation to which it was applied.
The AEWM has demonstrated capability to
perform AE tests on in-service bridges. It may
also be used to detect hidden discontinuities or
assist in making repair decisions concerning
detected discontinuities. The AEWM and AE
testing have been demonstrated to have the
potential for low-cost inspection of critical bridge
members.
(b) Some newly accomplished inspections of
bridges by the AE techniques are also reported
(Hariri 1990, Azmi 1990, Vannoy and Azmi 1991,
and Gong et al. 1992). A comprehensive examination of characteristics of acoustic emission
signals generated from steel beams (rolled and
welded sections) and other steels used in highway
bridge structures was made. The effective frequency range for monitoring highway bridges was
established. In these inspections, the thickness
and surface conditions of bridge components are
varied. Crack lengths were measured at the time
of data collection, and various acoustic emission
parameters were plotted versus the stress intensity

B-17

ETL 1110-2-551
31 Aug 98

factor of the specimens. It was discovered that


acoustic emission signal characteristics for the
steel types used in highway bridges are similar
although, the signals vary according to the thickness of the material. It was also discovered that
the corrosion surface enhances the intensity of the
signal, and paint layers do not have a significant
effect on the attenuation of the AE signals.
(c) Recent inspection of steel railroad bridges
by the AE method is presented by Gong et al.
(1992) in which they demonstrate successes using
AE to find new cracks, to identify active cracks, to
validate the effectiveness of repairs, and to provide damage assessments to assist with repair
prioritization.
(d) As given in Equation B-7, the stress
intensity factor range K for fatigue fracture is
highly related to the acoustic count rate N. The
typical relationships among the acoustic count

rate, the crack-growth rate, and K, based on


laboratory tests of typical bridge steel, are shown
in Figure B-13. It can be seen that the count rate
increases as a crack grows, and for an active
fatigue crack, the count rate presents an increasing
slope under constant cyclic loading. This result
indicates that a positive slope of the count rate
over time during field testing may be used as a
means of judging crack severity when compared to
similar slopes obtained from laboratory tests on
the same material. Table B-5 and Figure B-13
show a method of classifying fatigue cracks
inbridge steel into five safety index levels based
on the range of structure K determined by AE
monitoring. This approach has been effectively
used in interpreting K levels obtained from field
monitoring and applying them to the prioritization
of crack repair programs.
(e) The Monac multi-channel field AEmonitoring system is currently being used to

Figure B-13. Typical relationships among the crack safety index, crack-growth rate, count rate, and K for
bridge steels

B-18

ETL 1110-2-551
31 Aug 98
Table B-5
Fatigue-Crack Characterization for Bridge Steels
Range of K

Crack Safety Index

Crack Description

0  K < 10
10  K < 20
20  K < 30
30  K < 40
40  K

1
2
3
4
5

Minor defect
Slow crack growth
Requires repair
Dangerous
Imminent failure

monitor 36 bridges under normal loading conditions. The system consists of surveillance units
and an IBM PC control unit as shown in the
Figure B-14 block diagram. A line driver and
receiver are located next to each acoustic transducer to ensure the fidelity of weak acoustic
signals after transmission over long cables.
The monitoring distance can be up to 460 m
(1,509 ft). All bridges tested utilized standard
piezoelectric transducers with resonant
frequencies of 200 kHz.

e. Optical fiber method.


(1) Principle.

(f) In total, 353 locations have been monitored


on 36 railroad bridges over three years of testing.
Each location was monitored continuously from 4
to 10 days, during which time about 40 trains
passed over each bridge. Table B-6 gives the
results of this monitoring which located 116 active
cracks; among them 14 cracks had a safety index
of 3 and only one crack had a safety index of 4. No
crack had a safety index greater than 4.
(g) Cracks were often found at the webs of
floor beams near the upper corners of the connection angle with a stringer. In most cases, such
cracks had initiated because of the bending
moment on the stringer end. AE monitoring
indicated that two such cracks had a safety index
of 2. Many fatigue cracks were also found on the
top flanges of stringers.
(h) It was found that welds were always the
most crack-sensitive areas, possibly due to
residual stresses and stress concentrations. Therefore, current recommendations for bridge maintenance and repair favor bolting and riveting
rather than welding (Fisher et al. 1990).

(a) Like an elastic waveguide, a fiber can


guide high frequency electromagnetic waves
(optical waves) (Kao 1988 and Katsuyana and
Matsumura 1989). In a perfect symmetric and
homogeneous fiber waveguide, the waveforms of
guided modes propagate undisturbed along the
waveguide axis. However, a deformation or
inhomogeneity in fiber geometry may cause
coupling among different modes, resulting in
power transfer. In particular, the guided modes
may be coupled to radiation modes, which are not
confined. The resulting power transfer represents
an attenuation. Similarly, any deformation of
material attaching the fiber can also give rise to
the mode coupling and guided wave attenuation.
Therefore, the received optical wave in fiber can
be intensity modulated by the deformation of the
outside material. The guided wave can also be
phase modulated due to the change of its
geometric form which is caused by the strain or
deformation on the outside material.
(b) An optical fiber embedded in structures
will deform together with the structure. The light
passing through the optical fiber can be modulated
either in intensity or in phase. This effect has
been implemented in a form called Smart Strain.
By analyzing the changes of light intensity or
phase transmitted by embedded fiber, dangerous
strain levels in the structure as well as failure of
material may be detected. A system of such
optical fiber sensors embedded in a structure
could act as the nervous system of the structure.

B-19

ETL 1110-2-551
31 Aug 98

Figure B-14. Block diagram of the Monac acoustic monitoring system. RMS = root mean square

(c) A typical experimental arrangement for an


optical fiber sensor is shown in Figure B-15
(Brennan 1988). The output of a laser diode was
used in the input into the optical system which is
configured as a polariscope. The circularly polarized light output from the combination of polarized laser output and quarterwave retarder is
injected into the optical fiber. Strain induced by
beam deformation in the fiber will produce a
change in the polarization state at the output. This
is measured by the second polarizer acting as an
analyzer and the photo-diode. The retardation
between the two principle polarizationaxes is seen
as an intensity variation as measured by the
photo-diode.
(d) Figure B-16 illustrates another example of
optic fiber sensor (Sharma et al. 1981). This is
similar to the two-arm interferometer. The fiber
numbered 5 in Figure B-16, is subjected to the

external influence, but the fiber numbered 4 is isolated. Any change of the light phase in fiber
number 5 will result in the interference fringe displacement. The advantages of optical fiber
include its small size, light weight, faster data
speed, immunity from electromagnetic induction,
positioning, and lower cost.
(2) Applications. Optical fiber sensors have
been widely investigated (Asawa et al. 1982;
Marvin and Ives 1984; Lapp et al. 1988; Maria
et al. 1989). A special conference on fiber optic
smart structures and skins was held with more
than thirty papers presented at the International
Society for Optical Engineering (1988). Following is a brief introduction of some applications of
the optical fiber sensors.
 Mapping strain field. Optical fiber
sensors can be used to provide evaluation of the
B-20

ETL 1110-2-551
31 Aug 98
Table B-6
Results of Railroad Bridge Monitoring

Bridge
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

Bridge Type

Number of
Monitoring
Locations

Number of
Active
Cracks

TTS
TPGV
TPG
DPG/TPG
TPG
TTS/TT
TT/TPG/TTS
DPG/PYT
TPG/TT
DPG
TT/DT/DPG
TPG
DPG/BM/DPGV
DPG/DT
DT/TT
DPG/TPG
DT
TT
TT
TT
DT
TT
TT
DPG
TT
DPG
DPGV
DPGV
DPG
DPGV
DT
DPGV
DPGV
DPGV
DPG
DPG

8
4
6
7
6
8
8
7
5
5
8
6
8
8
8
6
15
13
134
12
5
6
5
6
8
12
2
2
3
2
6
2
3
3
2
4

2
1
3
2
0
3
1
1
0
1
3
4
1
3
6
4
3
2
46
6
0
0
2
3
2
4
1
2
1
0
1
0
2
2
1
3

Crack Safety Index


1

1
1
1
2

1
1
1
1
3
3
1
1
5
1
2
1
19
4

1
2
1
3
1
23
2

1
4

2
3
2
2
1
1

2
1
1

1
1
1
1

1
1
1
2

Bridge types: TTS = through-truss swing, TPG = through-plate girder, PYT = pony truss, DT = deck truss, TPGV = throughplate girder viaduct, TT = through-truss, DPG = deck-plate girder, and BM = beam.

state of strain in a structure, i.e., to map the strain


field in real time (Measures et al. 1988). The
reconstruction of the strain field may be achieved
by using the information obtained from the field
along a finite number of distinct paths (Figure B-17(a)). This is similar to the reconstruction
problem in topography, where the inverse transform is used to obtain the scalar field distribution.
The reconstruction of the strain field may also be
made by using point fiber optic sensors
(Figure B-17(b)).

 Flight control and damage assessment.


The fiber optic system composed of a network of
embedded sensors may be used to measure the
location and extent of damage that may occur
during flight, as well as structural integrity prior to
take off (Udd 1988). This system could be used in
combination with the flight control system to
ensure that the aircraft readjusts into a safe flight
envelope. The sensors could also be used to
measure such parameters as engine temperature,
shock position, structural loading, and temperature
and pressure distributions augmenting the flight
B-21

ETL 1110-2-551
31 Aug 98

Figure B-15. Optical arrangement

control system. This concept could be applied to


measuring the location and extent of damage to an
in-service bridge.

Figure B-16. Illustration of the two-arm


interferometer

 Vehicle detection and vehicle health


monitoring. The vehicle detection and vehicle
health monitoring system requires the installation
of 55-cm-long sensing elements on roadways
(Tardy et al. 1989). To prevent sensor damage,
cable elements are laid into grooves made in the
roadway. The laying process consists of placing
epoxy resin at the bottom of the groove, installing
the sensor cable, then filling the groove with
elastomer. Different groove depths have been
made. A reflective coating on the end face of the
sensor fiber allows the polarization effect to be
interrogated by a single fiber with Y coupler.
When a vehicle passes over the sensor, a signature
of characteristics is developed. The first fringe

B-22

ETL 1110-2-551
31 Aug 98

Figure B-17. Reconstruction of the strain field: (a) integrating fiber optic sensors traversing a measured
field, (b) using point fiber optic sensors

shift corresponds to the increased loading created


by the vehicle. A continuous signal indicates a
quasistatic pressure, while the second fringe shift
corresponds to the escape of the vehicle from the
sensor. The weight and speed of a moving vehicle
may be identified from the fringe number.
 Concrete structure testing. In one
example, 14-cm-long fiber-optic sensing elements
have been embedded in concrete test pieces to
determine their response to external pressure
(Tardy et al. 1989). Before placing the sensors in
concrete, the ribbons outside the sensor are
covered with epoxy resin and sand as shown in
Figure B-18. These new sensors have been tested
in a simple compressive test using a polarimetric
apparatus. First readings indicate a sensitivity loss
and fringe visibility reduction at the output end of
the fiber. The high value of the hardening stress is
due to the shrinking phenomenon of concrete.
Sensor desensitization is accomplished by loading

the test-piece in a perpendicular plane to that of


the ribbons. The observation sensor response is
shown in Figure B-19, where the stress axis origin
corresponds to the hardening pre-load. The fringe
numbers, which are related to the phase change of
the propagating light in the optical fiber versus
pressure, are in agreement with the theoretical
values. If fiber optic sensors are properly
embedded in critical parts of bridges, they may be
used for structural damage monitoring and
structure integrity assessment.
B-3. Summary
a. A review of state-of-the-art techniques for
real-time damage assessment of bridge structures
is provided. Of these methods reviewed, the
vibration dynamic and AE techniques have been
used in the real-time structural damage assessment
of in-service bridges on public roads. These two
methods have different physical bases.

B-23

ETL 1110-2-551
31 Aug 98

Figure B-18. Sensing structure

b. Every bridge has its own natural vibration


(or resonance) frequencies, which are related to
the materials, structural geometry, and integrity of
the bridge. If some components of the bridge are
damaged, the resonance frequencies and mode
shapes will change. The bridge signature can be
used to evaluate the bridge integrity. If cracking
occurs on a bridge, acoustic emission signals are
emitted. AE techniques have been used for
fatigue-crack detection of bridges. The greatest
advantage of AE monitoring over the other NDE
methods is its ability to detect active cracks and to
classify the severity of crack damage.

real-time bridge damage assessment is still under


investigation.

c. Although there are no published papers on


the use of optic-fibers to detect bridge damage,
optic fiber smart sensors have great potential for
monitoring damage incurred in bridge members.
All of the methods discussed are on the cuttingedge of technology for use in assessing bridge
damage. Experimentation of these methods in

(2) The vibration dynamic method can be


effectively used to make real-time damage
assessment of bridge integrity. Each bridge has a
special signature including resonance
frequencies and mode shapes. Some lower
resonance frequencies are simulated easier, since
they have large vibration amplitude and are

(1) Each critical component of a bridge should


be inspected by applying the most appropriate
NDE technology (e.g., visual, eddy current, liquid
penetrant, magnetic particle, ultrasonic testing, or
radiographic inspection) to discover any defects in
the components. Members in tension, localized
areas around stress concentrations, and areas
where a three-dimensional state of stress or high
constraint occurs should be considered in the
inspection plan.

B-24

ETL 1110-2-551
31 Aug 98

Figure B-19. Experimental response from sensor embedded in concrete

separated from other resonance frequencies.


Computer simulations or bridge models may be
useful in determining the best location to place
transducers to sense the vibration of signals.
(3) AE techniques have engaged the interest of
many scientists and engineers in real-time bridge
damage assessment. AE signals are developed
during crack growth and can be monitored on a
real-time sources to the sensor, the signal
identification can be difficult. A combination of
AE with the vibration or ultrasonic technique may
be necessary to completely evaluate a bridge.
(4) The optic fiber method has been widely
used to inspect aeronautical facilities. It has not
been extensively used for nondestructive
evaluation of civil structures such as bridges and
buildings. Several advantages (e.g., its small size,

faster data speed, and low cost) of the optic fiber


method are attractive for bridge inspection.
Application of this method to a bridge would
require tightly binding optic fibers to the critical
bridge members and monitoring the change of
optical intensity or phase at the output end.
d. A new application of fiber optics is in
crack imaging. It is reported that fiber optical
equipment can be used to provide clear and high
resolution images of remote cracks contained
within critical members (Wilson 1983). This
equipment is reported to be easy to handle and
operate. Fiber optic techniques have traditionally
been applied to the inspection of machined parts
on mechanical assemblies, but may also be
applicable to the real-time inspection of small
areas on critical structural members of a bridge.

B-25

You might also like