LOCALIZATION OF DWELL FATIGUE CRACKS IN TI-6242
ALLOY SAMPLES
S. I. Rokhlin, J.-Y. Kirn, B. Xie, V. A. Yakovlev and B. Zoofan
The Ohio State University
Nondestructive Evaluation Program
Edison Joining Technology Center
1248 Arthur E. Adams Dr.
Columbus, Ohio 43221
ABSTRACT. An in-situ ultrasonic guided wave technique is employed for real-time monitoring
of crack initiation and evolution during dwell, cyclic fatigue and creep tests of Ti-6242 alloy
samples. Ultrasonic signals are acquired continuously during the test at different levels of fatigue
load using a high-speed data acquisition system. The initiation time and growth history of
primary and multiple secondary cracks are assessed. Localization of the secondary cracks is
performed by both the in-situ ultrasonic method and an ultrasonic immersion scanning method
which we call "vertical C-scan" (VC scan). The VC scan is developed for imaging small cracks
aligned normal to the fatigue sample axis. The fusion of ultrasonic and microradiographic
images exhibits good agreement in crack location. Joint use of the three techniques provides
location, shape, and size of the secondary cracks.
INTRODUCTION
Due to their high specific strength, high service temperature and good fracture
toughness, titanium alloys are widely employed in aircraft engine components, e.g.
compressor blades and spools. The strength of these alloys is manipulated and
maximized by controlling their microstructure during thermomechanical processing.
However, it has been reported that these alloys show anomalously high primary creep
strains at low temperatures (0.2Tm (K)) and low applied stresses (0.6ay) [1]. The
accumulation of high creep strain results in a significant reduction of fatigue life during
the dwell cycle fatigue testing. Alloys that contain coarse oc/(3 microstructures have been
found to be most susceptible to dwell cycle fatigue [2]. To date the initiation
mechanism of the cold dwell fatigue crack in Ti-6242 alloy has been investigated [1-3],
but the details of the interaction between microstructure, loading history, hydrogen
content and fatigue life are not fully understood.
While the crack initiation site and the failure mode of a fatigue sample could be,
in principle, inferred from fractographs of the fracture surface, the time and conditions
CP657, Review of Quantitative Nondestructive Evaluation Vol. 22, ed. by D. O. Thompson and D. E. Chimenti
© 2003 American Institute of Physics 0-7354-0117-9/03/S20.00
1371
of crack initiation cannot be so determined. A straightforward approach to
understanding the crack initiation mechanism may be direct observation of the
microstructure surrounding a just-initiated dwell fatigue crack. For this purpose, early
detection and accurate localization of the initiated small cracks are of great importance.
In general, the monitoring of surface crack initiation is a difficult task.
Furthermore, in Ti-6242 alloy the dwell fatigue cracks initiate at multiple sites, which
adds complexity to the crack monitoring. In addition, the shape and size of the fatigue
sample allows limited room for access to the cracks.
In this paper, ultrasonic evaluation methods that have been developed to obtain
information on initiation, propagation and location of the dwell fatigue cracks in Ti6242 alloy samples are presented. An in-situ ultrasonic guided wave technique
developed for monitoring of fatigue crack initiation in Al-2024 alloy [4,5] is
implemented during dwell/cyclic fatigue and creep tests of Ti-6242 alloy samples. For
localization of secondary cracks an ultrasonic scanning method (vertical C-scan) is
developed. This technique is applied after mechanical testing in conjunction with
microradiography for precise localization and sizing of the cracks.
SAMPLES AND FATIGUE TESTS
Cylindrical and rectangular standard fatigue samples (ASMT E606-92) were
used in the experiments. Some of the fractured circular samples were obtained from GE
after mechanical tests. The cylindrical samples were 5 mm diameter and 19 mm length
in the gage section. One of the cylindrical samples has been shot-peened. The
rectangular samples were prepared from high microstructure Ti-6242 o/|3 forge and
machined to be flat fatigue specimens with 2 mm thickness, 6 mm width and 12.5 mm
length in the gage section. The samples had no start notch to study the crack initiation
mechanism and failed during the dwell fatigue, regular low cycle fatigue and creep
tests. The test conditions are summarized in Table I.
IN-SITU ULTRASONIC MONITORING OF DWELL FATIGUE CRACK
INITIATION AND PROPAGATION
Experimental Setup
To monitor the initiation and propagation of the dwell fatigue cracks, we have
employed ultrasonic guided waves excited in the fatigue sample. The transducer
assembly is clamped on the fatigue sample so that the ultrasonic reflection signals are
collected without stopping the fatigue test as shown in Fig. 1. The longitudinal wave
transducer with center frequency 5 MHz is used with a specially designed polystyrene
wedge to generate the Lamb wave in the 2 mm thick Ti-6242 alloy sample as shown in
Fig. 1. More precisely, So and AO Lamb waves are excited at a given value of the
frequency-thickness product. A wedge guided wave reflector has been mounted on the
sample (Fig. l(b)) to have a reference signal for interpretation and processing of the
signals acquired.
1372
TABLE I. Samples, mechanical test conditions and fatigue life
TABLE I. Samples, mechanical test conditions and fatigue life
Sample
Type of test
Test condition
Fatigue life
Type offatigue
test
Test
condition
Fatigue
life
Sample
F-1
Dwell
352 cycles
σmax
=869
MPa, R=0
Dwell fatigue
F-l
352 cycles
0^=869
MPa, R=0
Dwell time=2
min
Dwell
time^l min
F-2
Low-cycle
12,095 cycles
σ
max=869 MPa, R=0
F-2
Low-cycle
12,095 cycles
^3x^869 MPa, R=0
fatigue
fatigue
F-3
Creep
744 min
σ =869 MPa
744 min
Creep
F-3
a =869 MPa
R-1
Dwell fatigue
1,654 cycles
σmax=869 MPa, R=0
1,654 cycles
R-l
Dwell fatigue
dmax-869
MPa, R=0
Dwell time=2
min
Dwell time=2 min
17,877 cycles
R-2
Dwell fatigue
σmax=827 MPa, R=0
R-2
Dwell fatigue
17,877 cycles
<W=827
MPa, R=0
Dwell time=1
min
Dwelltime^l min
(a)
(a)
Fatigue load
Computer and data
Pulser/
Computer
and data
acquisition
Pulser/
Receiver *—————*
acquisition
software
Receiver
software
4
Transducer
Fatigue load
Transducer
MTS
controller
Wedge
Wedge
(polystyrene)
(polystyrene)
Ult
r as
on
ic W
ave
ref
l ec
tio
Sample type
Sample
2 mm Xtype
6 mm
2 mm
X 6 mm
rectangular
rectangular
2 mm X 6 mm
2 mm
X 6 mm
rectangular
rectangular
2 mm X 6 mm
2 mm
X 6 mm
rectangular
rectangular
2 mm diameter
2 mm diameter
cylindrical,
shot-peened
cylindrical, shot-peened
2 mm diameter
2 mm
diameter
cylindrical
cylindrical
(b)
(b)
transducer
transducer wedge
wedge
reflector
reflector
M
crack
crack
n
Subsurface Fatigue load
Subsurface
crack Fatigue load
crack
FIGURE 1.
1. (a)
(a) Setup
Setupfor
forin-situ
in-situultrasonic
ultrasonicexperiment
experimentduring
duringfatigue
fatiguetests,
tests.(b)(b)Lamb
Lambwave
wavereflector
reflectorforfor
FIGURE
obtainingreference
referenceultrasonic
ultrasonicsignal
signal
obtaining
Software for
for fatigue
fatigue control
control and
and the
the experimental
experimentalsystem
systemfor
forultrasonic
ultrasonicdata
data
Software
acquisition
have
been
developed.
The
software
controls
a
12
bit,
125
MHz
digitizing
acquisition have been developed. The software controls a 12 bit, 125 MHz digitizing
computer board
board to
toacquire
acquireand
andprocess
processultrasonic
ultrasonicreflection
reflectionsignals
signalsatatdifferent
differentlevels
levelsofof
computer
fatigue
load.
The
data
acquisition
and
the
ultrasonic
pulser/receiver
are
triggered
fatigue load. The data acquisition and the ultrasonic pulser/receiver are triggered bybya a
counter that
that isis controlled
controlled by
by the
the computer,
computer,allowing
allowingthe
thesystem
systemtotobebetriggered
triggeredatat
counter
predetermined loads
loads and
and time
time intervals.
intervals. The
The control/data
control/dataacquisition
acquisitionsystem
systemisisshown
shown
predetermined
schematically in
in Fig.
Fig. 2.2. The
The system
system allows
allows continuous
continuousultrasonic
ultrasonicmonitoring
monitoringofofthe
the
schematically
sample during
during fatigue
fatigueloading.
loading.The
Theacquisition
acquisitionrepetition
repetitionrate
rateofofthe
theultrasonic
ultrasonicsystem
systemisis
sample
selected to
to allow
allow decay
decay of
of the
the ultrasonic
ultrasonic events
events ininthe
thesample.
sample.Using
Usingthe
thereal-time
real-time
selected
computer control
control data
data acquisition
acquisitionsystem,
system,the
theultrasonic
ultrasonicsignals
signalswere
wereacquired
acquired9 9times
times
computer
during aa dwell
dwellperiod
periodand
and20
20times
timesduring
duringan
anunloading-loading
unloading-loadingperiod.
period.InInprinciple,
principle,the
the
during
signals may
may be
be acquired
acquired much
much more
more often;
often;we
welimited
limitedthe
thenumber
numberofofacquisitions
acquisitionstoto
signals
reduce the
the amount
amount of
of stored
stored experimental
experimental data.
data.During
Duringthe
thecyclic
cyclicfatigue
fatiguetest,
test,data
data
reduce
acquisition
was
performed
every
2,000
cycles.
At
each
event
of
data
acquisition
acquisition was performed every 2,000 cycles. At each event of data acquisition 5050
signals were
were obtained
obtained per
per cycle.
cycle.The
Theultrasonic
ultrasonicsignals
signalswere
wereobtained
obtainedevery
every2 2minutes
minutes
signals
during
the
creep
test.
during the creep test.
In-Situ Monitoring
Monitoringof
of Crack
CrackInitiation
Initiationand
andEvolution
Evolution
In-Situ
As an
an example,
example, Figure
Figure 33(a)
illustrates ultrasonic
ultrasonic backscattered
backscattered and
and wedgewedgeAs
(a) illustrates
reflector signals
signals atatdifferent
differentnumbers
numbersofofdwell
dwellfatigue
fatiguecycles.
cycles.The
Thetime
timedomain
domaingates
gatesare
are
reflector
shown on
on the
theultrasonic
ultrasonictrace
traceatatdifferent
differenttimes
times(along
(alongthe
thesample
sampleaxis).
axis).
shown
1373
Load out
Load
out out
Load
h w
MTS
|MTS|
MTS|
load data read
load
data
read
load
data
read
30 kHz
30A/D
kHz
16 bit
30 kHz
16 bit
A/DA/D
16 bit
^
Data
acquisition module
Data acquisition module
1
Data acquisition
f— |———————
RF out
125 MHz
Pulser
RF out
MHz
sample T
»/Receiver
Pulser
MHz
Pulser RFout
12125
bit125
A/D
sample
j T^T
sample
/Receiver
12 12
bitbit
A/D
A/D
/Receiver
Trigger In
MTS
Trigger
In In
Trigger
MTS
module
M:
•
Computer
( omputer
Computer
' -_
- '— _ -T
J———
MTS
MTS
MTS
microprofiler
microprofiler
microprofiler
TTL
TTL
TTL
n_
GUI
GUI
GUI
Counter
Counter
Counter
&& &
Amplifier
Amplifier
_lu^.^^gfm5
RS-232
RS-232
RS-232
t
-— — •
Fatigue
control
module
Fatigue
control
module
Fatigue
control
module
"
FIGURE
Schematic
of
data
acquisition/fatigue
control
system
FIGURE
2. 2.
Schematic
data
acquisition/fatigue
control
system
FIGURE
2. Schematic
of of
data
acquisition/fatigue
control
system
Reference signal
Reference signal
0.2
0.0
Signal amplitude
Signal amplitude
0.2
0.1
0.1
0.0
-0.1
-0.1
-0.2
-0.2
-0.3
12 3 4 5678
910
12 3 4 Gates
5678at different
910 times
Gates
at20different
times
15
25
30
35
40
Time (µ sec)
-0.3
15
20
25
0.20
30
Time (µ sec)
35
40
Ultrasonic reflection am plitude
Crack reflection signal
Crack
0.3 reflection signal
0.3
0.20
reflector
reflector
Ultrasonic reflection am plitude
transducer
transducer
transducer
wedge
wedge
wedge
(a) (a)
=
(b)
(b)
(b)
Major crack
crack leading
leading to failure j
Major
Major crack leading to failure
0.15
0.15
0.15
E
cc
Gate 1]
Gate
Gate 1
Gate22
Gate
0.10Gate 2
8 0.10
150.10
0.05
a 0.05
0.05
5
0.00
0.00
00
0
Gate 33
Gate
Gate 3
Gate 4
4
Gate
\
Gate 4 Crack
Crack retardatior
retardation
\
Secondary crack
crack initiation
initiation
Crack retardation
Secondary
Secondary crack initiation
100
100
100
200
200
300
300
400
400
500
500
Number
of dwell
fatigue
cycle
Number
200 of dwell
300 fatigue
400 cycle 500
600
600
600
FIGURE
Ultrasonic
signals
differentgates
gatescorresponding
correspondingtotodifferent
differentlocations
locationsalong
alongthe
thesample
sample
FIGURE
3. 3.
(a)(a)
Ultrasonic
signals
in in
different
Number of dwell fatigue
cycle
axis. (b) Changes of ultrasonic reflection signal amplitudes in different gates versus number of cycles.
axis,
(b)
Changes
of
ultrasonic
reflection
signal
amplitudes
in
different
gates
versus
number
of
cycles.
FIGURE 3. (a) Ultrasonic signals in different gates corresponding to different locations along the sample
(Sample F-1).
(Sample
F-l). of ultrasonic reflection signal amplitudes in different gates versus number of cycles.
axis.
(b) Changes
(Sample F-1).
Changes
reflectedsignal
signalmaximum
maximumamplitudes
amplitudesinindifferent
differentgates
gateswith
withnumber
numberofof
Changes
ofofthethe
reflected
cycles are summarized in Fig. 3(b). The results imply multiple crack initiation and
cycles of
arethe
summarized
in Fig.maximum
3(b). Theamplitudes
results imply
multiple gates
crackwith
initiation
andof
Changes
reflected signal
in different
number
growth. The signal amplitude change during the initial time period (until about 50
growth.
The
signal amplitude
changeThe
during
the imply
initial multiple
time period
(until
about and
50
cycles
are
summarized
in
Fig.
3(b).
results
crack
initiation
cycles) is due to stabilization of the transducer holder on the sample; there is no
cycles)The
is due
to amplitude
stabilizationchange
of the during
transducer
holder time
on the
sample;
there
is no
growth.
signal
the
initial
period
(until
about
50
evidence supporting crack initiation in this period. Drastic increase in signal amplitude
evidence
supporting
crack initiation
in transducer
this period. holder
Drastic on
increase
in signalthere
amplitude
cycles)
is
due
to
stabilization
of
the
the
sample;
is
no
occurs during the final stage of the fatigue life. The positions of the initiated cracks are
occurs supporting
during the final stage
of the in
fatigue period.
life. The positions
of the initiated
are
evidence
initiation
signalcracks
amplitude
determined fromcrack
the time
delays ofthis
signals in Drastic
differentincrease
gates. Inin particular,
gate 1
determined
from
the
time
delays
of
signals
in
different
gates.
In
particular,
gate
1
occurs
during thetofinal
stage ofcrack
the fatigue
The positions
of the ininitiated
are
corresponds
the primary
leadinglife.
to sample
failure (signals
all othercracks
gates are
corresponds
to
the
primary
crack
leading
to
sample
failure
(signals
in
all
other
gates
are
determined
fromthat
theare
time
delays
of signals
in different
In particular,
gateFor1
from cracks
located
behind
the fracture
surface ingates.
this particular
sample).
from cracks
that
are located
behind
the to
fracture
surface
in(signals
this particular
sample).
For
corresponds
to
the
primary
crack
leading
sample
failure
in
all
other
gates
are
example, the time delay of gate 4 is 4.8 µsec which corresponds to 7.5 mm distance
example,
the
time
delay
of
gate
4
is
4.8
(isec
which
corresponds
to
7.5
mm
distance
from from
cracks
are surface.
located behind the fracture surface in this particular sample). For
thethat
fracture
from the fracture
surface.
alsodelay
important
that4from
Fig.µsec
3 (b)which
the crack
growth history
be inferred.
example, theIt is
time
of gate
is 4.8
corresponds
to 7.5canmm
distance
It
is
also
important
that
from Fig.
3 (b)1the
crack growth
history canat be
inferred.
For
example,
the
signal
evolution
in
gate
indicates
crack
initiation
around
400
from the fracture surface.
Forcycles.
example,
the signal
evolution
in4 shows
gate 1crack
indicates
crackat initiation
at cycles
around(which
400
Likewise,
the
signal
in
gate
initiation
around
170
It is
also important
that
3 (b)
the initiation
crack growth
history
be inferred.
cycles.
Likewise,
the signal
in from
gate 4Fig.
shows
crack
at around
170can
cycles
(which
For example, the signal evolution in gate 1 indicates crack initiation at around 400
cycles. Likewise, the signal in gate 4 shows crack initiation at around 170 cycles (which
1374
is isearlier
earlierthan
thanthe
theprimary
primarycrack
crackinitiation)
initiation)followed
followedby
bycrack
crackretardation
retardationatatabout
about400
400
cycles
while
the
primary
crack
grows
till
failure.
The
retarded
crack
becomes
cycles while the primary crack grows till failure. The retarded crack becomes active
active
again
againafter
after500
500cycles.
cycles.The
Thecrack
crackretardation
retardationmay
maybebeattributed
attributedtotocrack
crackre-nucleation
re-nucleationatat
a aboundary
with
a
neighboring
colony
where
the
crystal
orientation
is
much
boundary with a neighboring colony where the crystal orientation is muchdifferent
different
than
thanthat
thatofofthe
thecolony
colonywhere
wherethe
thecrack
crackoriginally
originallyinitiated
initiated[6].
[6].Using
Using the
the technique
technique
described
describedwewecan
canobtain
obtaininformation
informationononinitiation
initiationtimes
timesfor
fordifferent
different cracks
cracks inin the
the
sample.
Although
interaction
of
the
ultrasonic
wave
with
multiple
cracks
sample. Although interaction of the ultrasonic wave with multiple cracksmay
maycause
cause
shielding
shieldingofofthethesignals,
signals,it itisisbelieved
believedthat
thatcrack
crackinitiation
initiationand
andpropagation
propagation can
can bebe
monitored
successfully
while
cracks
are
small.
monitored successfully while cracks are small.
ULTRASONIC
ULTRASONICVERTICAL
VERTICALC-SCAN
C-SCANTECHNIQUE
TECHNIQUE
Experimental
ExperimentalSetup
Setup
We
Wehave
havealso
alsodeveloped
developeda amethod
methodfor
forultrasonic
ultrasoniclocalization
localizationofofthe
thesecondary
secondary
cracks
in
the
fractured
fatigue
samples
after
mechanical
testing.
Since
cracks in the fractured fatigue samples after mechanical testing. Since cracks
cracks are
are
oriented
orientednearly
nearlyperpendicular
perpendiculartotothe
thesample
sampleaxes
axeswe
we developed
developed aa scanning
scanning method
method
shown
shownschematically
schematicallyininFig.
Fig.4.4.We
Wecall
callthis
thisscanning
scanningmode
mode“VC
"VCscan”.
scan".
The
Theangles
anglesofofthethetransducer
transducerholder
holderare
aredesigned
designed inin such
such aa way
way that
that the
the
ultrasonic
ultrasonicbeam
beaminside
insidethe
thesample
sampleisisreflected
reflectedby
bythe
thecrack
crackand
andthe
thereflected
reflected beam
beamisis
received
receivedbybythethetransducer
transducerononthe
theopposite
oppositeside
sideofofthe
thesample.
sample.The
Theincident
incidentangle
angleisis
o
determined
determinedtotobebe1919°that
thatisislarger
largerthan
thanthe
thefirst
first(longitudinal
(longitudinalwave)
wave)critical
criticalangle
angleso
soasas
o
toto
produce
producea 45
a 45°refracted
refractedshear
shearwave
waveininthe
thesample.
sample.As
Asshown
shownininFig.
Fig.4,4,when
whenthere
thereisisaa
crack
crackorora flaw
a flawininthe
thescanned
scannedcross-section,
cross-section,the
therefracted
refractedshear
shearwave
wavebeam
beamisisreflected
reflected
o
o
, departs
ononthethecrack
cracksurface
surfaceatat4545°,
departsthe
thesample
sampleatat1919°and
andisisreceived
receivedby
bythe
thereceiving
receiving
transducer.
transducer.When
Whenthere
thereisisnonocrack
crackininthe
the sample
sample cross-section,
cross-section, the
the transmitted
transmitted
ultrasonic
ultrasonicbeam
beamtravels
travelsaway
awayfrom
fromthe
thetransducer
transducerdirection
directionand
anddoes
doesnot
not reach
reach the
the
receiver.
receiver.The
Thesystem
systemconsists
consistsofoftwo
twoultrasonic
ultrasonicfocus
focustransducers,
transducers,aatransducer
transducerholder
holder
assembly
assemblyand
andananimmersion
immersionX-Y-Z
X-Y-Zscanning
scanningbridge.
bridge.The
Thesample
sampleisis mounted
mounted on
on aa
mechanical
mechanicalalignment
alignmentdevice.
device.The
Thecenter
centerfrequencies
frequenciesofofthe
thetransducers
transducersare
are25
25MHz
MHzand
and
MHzforforthe
the transmitter
transmitter and
and receiver,
receiver, respectively.
respectively. The
The focal
focal length
length ofof the
the
2020MHz
transducersis is5.08
5.08mm
mm(2”).
(2").The
Thediameter
diameterofofthe
thetransmitter
transmitterisis0.635
0.635mm
mm(0.25”)
(0.25")and
and
transducers
thatofof
receiveris is1.27
1.27mm
mm(0.5”).
(0.5").
that
thethe
receiver
Flat
fatigue
sample
Flat
fatigue
sample
scan
Transmitter
Transmitter
Receiver
Surface
Surface
or or
subsurface
subsurface
cracks
cracks
Refracted
shear wave
45o
crack
FIGURE
Schematic
vertical
C-scan
(VC
scan)
technique
FIGURE
4. 4.
Schematic
of of
thethe
vertical
C-scan
(VC
scan)
technique
1375
19o
The
cross-sections at
at different
different
The ultrasonic
ultrasonic scans
scans are
are performed
performed over
over the
the sample
sample cross-sections
vertical
locations
along
the
sample
producing
a
set
of
vertical
scans.
The
approximate
vertical locations along the sample producing a set of vertical scans. The approximate
diameter
size of
of the
the scanning
scanning is
is 0.1
0.1 mm.
mm.
diameter of
ofthe
the ultrasonic
ultrasonic beam
beam is
is 11 mm.
mm. The
The step
step size
Both
round
and
flat
fatigue
samples
were
evaluated.
Although
the
results are
are
Both round and flat fatigue samples were evaluated. Although the results
shown
for
post
mortem
fatigue
samples,
the
same
scanning
technique
can
be
applied
shown for post mortem fatigue samples, the same scanning technique can be applied
for
for engine
enginecomponent
component inspection
inspection after
after manufacturing
manufacturing or
or in
in service.
service.
Localization
Localization of
of Secondary
Secondary Cracks
Cracks
Figure
obtained from
from the
the round
round sample
sample R-l
R-1
Figure 55 shows
shows two
two ultrasonic
ultrasonic scan
scan images
images obtained
for
illustration.
This
sample
had
a
rough
shot-peened
surface.
The
circle
indicates
the
for illustration. This sample had a rough shot-peened surface. The circle indicates the
sample
boundary.
The
image
of
the
sample
cross-section
at
z=5.98
mm
from
the
sample
sample boundary. The image of the sample cross-section at z=5.98 mm from the sample
top
is observable
observable in
in the
the image
image due
due to
to
top shows
shows no
no indication
indication of
of flaw.
flaw. The
The sample
sample surface
surface is
ultrasonic
is marked
marked in
in the
the image).
image). The
The gray
gray area
area in
in
ultrasonic scattering
scattering on
on the
the rough
rough surface
surface (it
(it is
the
to grain
grain noise.
noise. Since
Since the
the dwell
dwell fatigue
fatigue
the central
central part
part of
of the
the image
image (marked
(marked A)
A) is
is due
due to
crack
of some
some grains
grains aligns
aligns with
with the
the
crack initiates
initiates when
when the
the specific
specific crystal
crystal orientation
orientation of
loading
in studying
studying dwell
dwell fatigue
fatigue crack
crack
loading direction,
direction, grain
grain noise
noise assessment
assessment is
is important
important in
initiation.
mm shows
shows clear
clear indications
indications of
of two
two
initiation. The
The image
image for
for the
the cross-section
cross-section at
at z=6.82
z=6.82 mm
internal
are 0.44
0.44 mm
mm and
and 0.29
0.29 mm.
mm.
internal fatigue
fatigue cracks
cracks (Fig.
(Fig. 5).
5). The
The lengths
lengths of
of the
the cracks
cracks are
Since
is under
under residual
residual compressive
compressive
Since the
the sample
sample was
was shot-peened,
shot-peened, the
the sample
sample surface
surface is
stress
sample volume.
volume. In
In fact,
fact, in
in the
the other
other
stress and
and the
the cracks
cracks were
were initiated
initiated inside
inside the
the sample
samples,
and on
on the
the sample
sample surface
surface as
as shown
shown in
in
samples, the
the fatigue
fatigue cracks
cracks initiated
initiated both
both inside
inside and
the
thefollowing
following figures.
figures.
Figure
round sample
sample R-2
R-2 without
without shotshotFigure66shows
shows an
an ultrasonic
ultrasonic scan
scan image
image for
for the
the round
peening.
there is
is very
very little
little surface
surface scattering,
scattering,
peening. Since
Since this
this sample
sample had
had aa smooth
smooth surface,
surface, there
so
previous sample
sample
soititisismore
more difficult
difficult to
to identify
identify the
the sample boundary. Similarly to the previous
grain noise
noise appears
appears in
in the
the image
image (marked
(marked A). At distance z=7.01 mm from
grain
from the
the fracture
fracture
surface, indication
indication of
of aa surface
surface crack
crack appears.
appears. The crack depth is determined
surface,
determined to
to be
be 1.7
1.7
mmand
and the
the width
width 1.1
1.1 mm.
mm.
mm
Z=5.98 mm
mm
Z=5.98
Z=6.82 mm
z
Scanning
direction
Receiver
Transmitter
A
Rough surface
surface
Rough
scattering
Internal defects
defects
Internal
FIGURE5.5.VC
VCscan
scanimages
imagesofofsample
sampleR-1
R-l atatdifferent
different heights
heightsfrom
from the
the fracture
fracture surface.
surface.
FIGURE
1376
Scanned at
z=7.01 mm
Scanned
at
z=7.01 mm
15X
Fracture surface
Fracture surface1x
Surface crack
7.2 mm
7.2mm
1.5 mm
Surface
Surface
crack
crack
1 mm
A
375 µm
40X
1 mm
15X
1.7 mm
1.7mm
FIGURE 6. Comparison of ultrasonic VC scan image at z=7.01 from the fracture surface with
FIGURE
6. Comparison
microradiographic
images of
for ultrasonic
sample R-2.VC scan image at z=7.Ql from the fracture surface with
microradiographic images for sample R-2.
The microradiograph of the sample is also shown in Fig. 6 for comparison. The
Theofmicroradiograph
samplesurface
is also determined
shown in Fig.
6 for
The
distance
the crack from of
thethe
fracture
from
thecomparison.
microradiography
distance
of
the
crack
from
the
fracture
surface
determined
from
the
microradiography
image is 7.2 mm and the depth 1.5 mm, which is close to the parameters determined
image
is 7.2
mm andimage.
the depth 1.5 mm, which is close to the parameters determined
from the
ultrasonic
from the Figure
ultrasonic
image.
7 compares the in-situ ultrasonic crack monitoring curves, ultrasonic VC
7 compares the in-situ
crack
curves,
ultrasonic
VC in
scans Figure
and microradiographic
imagesultrasonic
for sample
F-3.monitoring
The secondary
cracks
indicated
scans
and
microradiographic
images
for
sample
F-3.
The
secondary
cracks
indicated
in
the microradiography correspond to those identified from the in-situ ultrasonic
the
microradiography
correspond
to
those
identified
from
the
in-situ
ultrasonic
measurements and the ultrasonic VC scan images. One additional crack is found by
measurements
the ultrasonic
VC 6scan
One
additional
crack
is the
found
by is
both ultrasonicand
techniques
at around
mmimages.
from the
fracture
surface
while
crack
both
ultrasonic
at aroundimage.
6 mm from
the fracture
the crack
invisible
in thetechniques
microradiographic
The locations
andsurface
sizes ofwhile
the cracks
thatisare
invisible
the microradiographic
image. The locations
andare
sizes
of the cracks
that are
obtainedinfrom
the VC scan and microradiographic
images
summarized
in Table
II. It
obtained
from
the
VC
scan
and
microradiographic
images
are
summarized
in
Table
should be noted that in the row Crack-3 three closely located cracks were foundII.
in Itthe
should
be noted thathowever
in the row
closelycrack
located
cracks
the
microradiograph;
theyCrack-3
appear three
as a single
in the
VCwere
scanfound
image.inThese
microradiograph;
however
they
appear
as
a
single
crack
in
the
VC
scan
image.
These
cracks are close in vertical cross-section (1.0 mm, 2.0 mm and 2.8 mm from the fracture
cracks
are close in vertical cross-section (1.0 mm, 2.0 mm and 2.8 mm from the fracture
surface). The microradiography has better vertical resolution than VC scan imaging. As
surface). The microradiography has better vertical resolution than VC scan imaging. As
one can infer from Fig. 4, to resolve the two close parallel cracks the z position of the
one
can infer from Fig. 4, to resolve the two close parallel cracks the z position of the
VC scan should be changed by a skip distance equal to the sample thickness. This limits
VC scan should be changed by a skip distance equal to the sample thickness. This limits
vertical resolution of the VC scan. One should also note that the microradiography
vertical resolution of the VC scan. One should also note that the microradiography
underestimates the sizes of the cracks inclined to the sample surface.
underestimates the sizes of the cracks inclined to the sample surface.
SUMMARY
SUMMARY
Nondestructive techniques for obtaining information on initiation, propagation,
Nondestructive techniques for obtaining information on initiation, propagation,
location and size of dwell fatigue cracks in Ti-6242 alloy sample are presented. An inlocation and size of dwell fatigue cracks in Ti-6242 alloy sample are presented. An insitu ultrasonic guided wave technique is developed for real-time monitoring of crack
situ
ultrasonic guided wave technique is developed for real-time monitoring of crack
initiationand
andevolution
evolutionduring
duringfatigue
fatiguetesting.
testing.It Itallows
allowsmeasuring
measuringthetheinitiation
initiationtime
time
initiation
and
growth
history
of
primary
and
multiple
secondary
cracks.
The
secondary
cracks
and growth history of primary and multiple secondary cracks. The secondary cracks areare
localizedand
andsized
sizedusing
usingthe
thevertical
verticalC-scanning
C-scanningtechnique
technique(VC
(VCscan).
scan).The
Thescanned
scanned
localized
ultrasonic
images
were
compared
with
microradiographic
images
and
found
to
ultrasonic images were compared with microradiographic images and found to bebein in
goodagreement.
agreement.Combining
Combiningthe
thethree
threeNDE
NDEtechniques
techniquesenables
enablesone
oneto todetermine
determinethethe
good
initiation
times,
shapes,
orientations
and
sizes
of
the
cracks.
initiation times, shapes, orientations and sizes of the cracks.
1377
0.30
0.28
G
H
Am plitude (V)
0.26
H
0.24
G
0.22
0.20
J
K
0.18
K
0.16
J
0.14
0.12
2
1
3
N
0
200
200
400
400
600
600
4
N
Tim e (m in)
250 µm
Time (min)
FIGURE
FIGURE 7.
7. Comparisons
Comparisons of
of in-situ
in-situ ultrasonic
ultrasonic crack
crack monitoring
monitoring curves,
curves, ultrasonic
ultrasonic VC
VC scan
scan images
images and
and
microradiographic
microradiographic images
images for
for sample
sample F-3.
F-3.
TABLE
TABLE II.
II. Sizes
Sizes of
of the
the secondary
secondary cracks
cracks in
in Sample
Sample F-3
F-3 by
by VC
VC scan
scan and
and microradiography
microradiography
Crac
Crac
kk
11
22
33
44
Ultrasonic
Ultrasonic VC
VC scan
scan
Distance
Length
Depth
Length
Distance
Depth
from
from top
top
1.8
0.4
0.3
mm
0.4 mm
mm
0.3mm
1.8 mm
mm
2.2
mm
0.25
0.2
mm
2.2mm
0.25 mm
mm
0.2mm
2.5
1.0
mm
0.3
mm
1.0mm
2.5 mm
mm
0.3mm
2.5
mm
2.5mm
0.7
mm
0.7mm
Microradiography
Microradiography
Distance
from top
top
Length
Length
Distance from
0.32
mm
0.32 mm
1.6
mm
1.6 mm
2.1 mm
mm
2.1
1.0
1 .0 mm,
mm, 2.0
2.0 mm
mm
&
& 2.5
2.5 mm
mm
2.75 mm
mm
2.75
0.3 mm
0.3mm
0.2 mm
0.2mm
0.2 mm,
mm, 0.7
mm &
0.2
0.7 mm
&
0.3 mm
mm
0.3
0.45 mm
mm
0.45
ACKNOWLEDGEMENTS
ACKNOWLEDGEMENTS
This
by the
Aviation Administration
Administration (FAA)
(FAA) under
under contract
contract
This work
work was
was sponsored
sponsored by
the Federal
Federal Aviation
#97-C-001
#97-C-001 as
as aa part
part of
of the
the project
project “Evaluation
"Evaluation and
and Microstructure-based
Microstructure-based Modeling
Modeling of
of
Cold
the Airworthiness
Airworthiness Assurance
Assurance Center
Center of
of
Cold Dwell
Dwell Fatigue
Fatigue in
in Ti-6242”
Ti-6242" through
through the
Excellence.
Excellence.
REFERENCES
REFERENCES
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
6.
6.
Evans,
Evans, W.
W. J.,
J., Scripta
Scripta Met.,
Met., 21,
21, 1223,
1223, (1987).
(1987).
Evans,
Int. J.
J. Fatigue,
Fatigue, 16,
Evans, W.
W. J.
J. and
and Bache,
Bache, M.
M. R.,
R., Int.
16, 443,
443, (1994)
(1994)
Kassner,
Kassner, M.
M. E.,
E., Kosaka,
Kosaka, Y.
Y. and
and and
and Hall,
Hall, J.
J. A.,
A., Metall.
Metall Trans.,
Trans., 30A,
30A, 2383
2383 (1999).
(1999).
Rokhlin,
Rokhlin, S.
S. I.I. and
and Kim,
Kirn, J.-Y.,
J.-Y., Int.
Int. J.
J. Fatigue,
Fatigue, in
in press
press (2002).
(2002).
Kim,
Kirn, J.-Y.
J.-Y. and
and Rokhlin,
Rokhlin, S.
S. I.,
I., Int.
Int. J.
J. Solids
Solids Struct.,
Struct., 39,
39, 1487
1487 (2002).
(2002).
Arata,
Arata, J.
J. J. M., Kumar, K. S., Curtin, W. A. and Needleman, A., Int. J. Fracture, 111,
163
163 (2001).
(2001).
1378