The Principles: History

Chapter (2): The Principles Of TOFD
CHAPTER (2): THE PRINCIPLES OF TOFD
2.1 A BRIEF HISTORY OF TOFD

In the 1960s and 7970s,the construction of nuclear power plants around the world was reaching its
highest level. Dozens of plants were being built every year and many had been in service for several
years. The technology often required new metallurgy and the service conditions were not always well
understood. This, coupled with the extreme safety requirements for nuclear projects, meant that vessel
construction was usually thicker than many previous similar projects. In the 1960s and 1970s, the NDT
standards were based on radiography. With construction involving heavier sections than previously,
along with the concerns for weld integrity to assure the safety of the vessels, radiography was not
always the best option. Radiography was generally slow due to very long exposure through thick
metal sections. As well, radiography was not always capable of detecting some flaws due to the loss of
resolution caused by radiation scatter.
Pulse-echo ultrasonic testing soon found a popular niche for itself. It was much more efficient than
cobalt at penetrating 300mm of steel. It also had the added benefit of no elaborate radiation safety
barriers and regulations to worry about. Moreover, it was known to be more sensitive to planar flaws
that might go undetected by radiography.
Nuclear plants are all required to undergo periodic inspections of the components. This includes
assessment of the pressure vessels' welds. When pulse-echo inspections began to locate flaw
indications in the vessels and piping during these periodic inspections, a new demand was placed on
the NDT; how big are these flaws? The need to better assess the size of these flaws became the origin
of the time-of-flight diffraction technique (TOFD).
Dr. Maurice G. Sitk and his staff, at the National NDT Centre, in Harwell, UK, are credited with
developing the TOFD technique in the early 1970s. The impetus for their work was the need to know,
with some accuracy, the vertical extent of flaws detected in the pressure retaining components.
Around the same time, fracture mechanics engineering was developing into an important tool in the
understanding of how metals fail and the importance of f1.aw size and shape in the failure
mechanisms. Fracture mechanics engineers have long knoum that the flaw dimensions as well as the
metal structural properties and service conditions need to be quantified in order to determine if a
component can operate without failing (fracturing). This put a new demand on NDT; in that accurate
flaw sizing was required, as an integral part of the equatiory for determining whether or not a
component could be left in service, or required removal or repair.
Until that time (and often even today) the standard method of flaw sizing, using ultrasonic methods,
relied on the beam boundary methods. Simply stated, this assumes that the echo amplitude of a flaw
has some relationship to the ultrasonic beam dimensions. Options included:
31.
a -20dB drop
a -6dB drop
a -Max. Amp. (another variation on dB drop options)
a -Relative amplitude (e.9., AVG/DGS)
Amplitude drop methods of sizing relate

relative probe motion from a maximum
to a lower amplitude. In some
techniques, the centre of the beam is
used; whereas others require that beam -6dB beam edges
dimensions be determined first.
!
!
!
I
i Top
t1
view Slde Vieu
Figure 2-1 Beam Boundary Sizing basis
Figure 2-1 shows two views of an angle

beam probe and its 6dB sound envelope
in the test medium. If we assume that we
direct the beam at a reflector having
dimensions greater than the beam
dimensions at that distance, a maximum
a)
amplitude signal will result. For an ideal
reflector, movement left or righf along
the length of the reflector, will maintain a
constant soundpath; and amplitude will
remain constant until the edge of the
reflector is approached. \A4ren the
centreline of the probe aligns with the
edge of the reflector one half of the beam
energy is no longer reflected. This
reduces the echo signal to half of what it
was at maximum, i.e. a 6dB drop is
observed. Determining the length of such
an ideal defect is merely a matter of
finding the points where the signal drops
6dB from its maximum and measuring
the distance the probe has been moved l€t
(see Figure 2-2(a)). Apparent length (defect is shorter than beam width)
Figure 2-2Length sizing by the dB drop method
CZ
When the defect dimensions are less than the beam dimensions (see Figure 2-2, b) lateral movement
from a maximum amplitude results in immediate reduction of signal amplitude. But the probe
displacement is more a result of beam dimension than defect dimension. This is more Pronounced as
the ratio of beam width to defect width increases.
Similar conditions exist for vertical extent determinations. These are made by moving the angle beam
towards and away from the reflector. However, in this plane, the amplitude of the reflector changes
for two reasons:
1) portion of beam interacting on the reflector's surface
2) distance amplitude variation due to change in sound path distance
Since most defects will be smaller than the beam, in at least one dimensiory amplitude drop methods
are normally used with knowledge of beam divergence; and corrections (by use of beam spread
overlays) are made for this divergence. The half angle of divergence for a 6dB drop can be estimated
from equation (1,.22). However, this will not accurately correct for refraction differences, from a wedge
to metal, in contact testing methods. Accurate beam shapes are determined empirically from
calibration blocks. The IOW (Institute of Welding) block is most often used for this purpose. The hole
ends and sides are used to determine beam width and beam height respectively.
The "MaxAmp" technique is also an amplitude drop technique and relies on the faceted nature of
some flaws. The probe is moved until the last facet seen just starts its amplitude drop. (MaxAmp is
described in "IJltrasonic Flaw Detection for Technicians- third editioru by ].C.Drury, Silverwing,2004).
In 1958, Krautkramer developed a series of formulae to estimate the size of a defect as an equivalent to
a flat disc reflector. These formulae relate echo height to beam path and target size. With these
relationships, Krautkramer developed curves, relating distance along the horizontal axis to gain in dB,
compared to a back-wall reflector as a proportion of the probe diameter. The original German term for
the system is AVG (Abstand Verstaerkung and Groesse). Translated to English, these words are
Distance, Gain and Size, hence the term DGS. The general DGS diagram is shown in Figure 2-3.
o
!
.=
o
!
gt
o.
E
o
to
o
ul
10,5 I 5t0
RelstiYe dktance l zx lD?
Figure 2-3 Standard DGS diagram tr)
JJ
In the 1970s, Prof. Dr. I.N. Ermolov proposed equivalency equations that allowed targets other than
DSRs (flat-bottom holes) to be used.
Figure 2-4, illustrates the predicted amplitude drops, with increasing distance, for a specific probe
(5MHz 16mm diameter) radiating into steel; and compares the rate of decay with distance for four
different targets.
5elect target
.l
Diffiier [mml
$en.rdo.4 12 nDls 1,ed.,[Ma3,{0] ri .. rBnsi.er 9udi.eAreaisl
20r 06 m':
Fltsl Ueiecf
AtmdEn k(dd 0 0l il/m NomrFal@q a! rHz w&.k€e{Ll 06a m
Diamele{dl lnml u3r:on.P6rlon[Nl ]00.00 mm
.
,lnllnite Clinds 1 ! ?D !{1.1
!1idlh[] lmml
]U
Figure 2-4 Ermolov equivalent target curves
These techniques were the norm throughout the 1970s and engineers had to rely on the size values
provided by NDT technicians using these techniques. When the pressure of fracture mechanics was
made to bear on the sizing provided by these methods, they were proven to be wanting. The
inadequacy of these methods was made clear as a result of a series of famous round-robin trials.
Over the years, many international exercises have been carried ouf comparing TOFD to the
traditional inspection methods, such as radiography and ultrasonic pulse-echo. These include the
PISC trials (Programme for the Inspection of Steel Components) for nuclear applications (PISC I was
held in 1979,PISC II in 1986 and PISC III in 1993), the DDT or Defect Detection Trials held by UKAEA
(United Kingdom Atomic Energy Authority) 1983, as well as trials by the Dutch Welding Institute
(NIL), the Electric Power Research Institute (EPRD and the American Society of Mechanical Engineers
(ASME).
Without exceptiory each of the trials confirmed the ability of TOFD to provide results for volumetric
defects with a greater accuracy than the conventional methods to which it was compared. Only raster
pulse-echo, which has now developed into the phased array technique, had a comparable success rate.
When viewed with respect to the goal of determining the efficacy of sizing, TOFD constantly
outperformed the other options. This is clearly illustrated in Figure 2-5.
34
a ;
t
'6
;
i ?
6
E E
8
:
l0 20 30 10 50 60 ?0 86 90 100 110
oet*l lhroJgh 6tl s'E clclrrn,hed Oy 49 de$rucl.va

etomr{t,o^ lmml
{a}}DTPlcret {s}DDTplde I
Figure 2-5 Amplitude vs. diffraction sizjng methods from DDT plate 1 (left: all sizing techniques,
right: TOFD onlY) tzr
Similarly, reliability of inspection techniques was compared during the NIL/KINT thin plate project.
Figure 2-6 indicates the reliability values derived for different inspection techniques. These were as a
result of examinations of welded plates in a thickness range of 6mm to 15mm, containing some 250
implanted weld defects.
IOFD PE PE X-raY Gamma Slanual

\teander Line r.ry UT
scan Scan
Figure 2-6 NIL/KINT thin plate project results (8)
A11 plates were examined with the techniques shown; and later the results were verified by
destructive testing. From these results, it follows that the Mechanised Pulse-echo Meander Technique
results in the highest probability of detection (PoD) and TOFD delivers the lowest False Call Rate. It
should however, be kept in mind that the Mechanised Pulse-Echo Meander Technique (2 axes
motorised raster scan) requires a total scanning time which is many times that of TOFD.
35
These studies show that although the original goal of TOFD was to provide improved sizing
capabilities, it is now acknowledged to have the added benefit of providing a fast and reliable
detection method.
Mr. Bill Browne presented a paper on the internet in September of 1997o, where he laid out some of
the basics involved, but also made a loud statement on the "mystique" that had built up around TOFD
by that time. He commented on the misunderstandings of TOFD:
"Some of this tack of understanding emanates from the mystique built up by those responsible for its
introduction. For many years scientists promoted the technique as a highly specialised 'sizing' tool - so complex
that it required their specialist knowledge and sophisticqted technology to ffictit;ely apply - and unsubstantiated
claims were made about what the technique could and could not achiet:e.
This may haoe been the case in the aery early stages of eoolution but TOFD has now been around for more than
25 years, its effectioeness has been proaen by performance demonstration (more so than sll other inspection
methodologies combined) and znhateoer mystery once surrounded the technique has been completely dispelled by
repeated applicational success in the hands of routinely qualified personnel using what is now relatiaely
' standar d' equipment. "
2.2 GENERAL PRINCIPLES

Ultrasonic inspection methods using pulse-echo techniques have primarily been based on specular
reflection. This implies a "mirror-like" reflection of sound from the probe, off the flaw and back to the
probe again. In traditional pulse-echo methods attempts are made to provide a beam path to the flaw,
such that the beam strikes the flaw, perpendicular to the plane with the flaw's major dimensions. This
is intended to provide the maximum amplitude resPonse in pulse-echo mode.
But trying to arrange the beam to always impinge on a flaw, such that a perpendicular incidence
occurs, is virtually impossible. This accounts for some of the lack of detection by manual ultrasonic
methods. Even a large, flat reflector could have a 20dB drop by the beam striking it only 5" off the
perpendicular. This could mean that even large flaws could be missed, if "detection" is based on
signal amplitude exceeding a threshold.
But sound. waves have several possible interactions, when they strike a boundary. Reflection is only
one thing that may occur. Refraction, mode conversiory polarisation, attenuation and diffraction are
also possible. Of these boundary interactions, diffraction is the important consideration in TOFD
studies (naturally, since the name, time-of-flight diffraction derives its name from the phenomenon).
Diffraction occurs when an obstacle distorts a wave, e.g., if the wave travels through a gap it may be
diffracted. It is the ability of a wave to spread around corners. Hence, diffraction occurs at the edge of
an obstacle.
The typical examples used to illustrate the effects of diffraction are slits (openings) in obstructions in a
"ripple-tank". The "wave source" generates a pulse that moves towards the slits and the opening
allows a portion of the wavefront to pass through. As the wavefront moves through the slits, the
centre portion of the slits allow the wave to move forward unimpeded. However, the portions that
touch the slits are retarded and bend the wavefront away from the path along the central axis.
36
f, f
i
!
I
g
fi
ir'..' .- .,;i ,. ......,'-r!,,..-.!: :i:i. j.r.r!..,j-,,.E

e*cfrnotlon lh6, erth'fror,on lfrml
(r) DDT PlaE , (a, DDt Plre I
Figure 2-5 Amplitude vs. diffraction sizjng methods from DDT plate 1 (left: all sizing techniques,
right TOFD only) 0)
Similarly, reliability of inspection techniques was compared during the NIL/KINT thin plate project.
Figure 2-6 indicates the reliability values derived for different inspection techniques. These were as a
result of examinations of welded plates in a thickness range of 6mm to 15mm, containing some 250
implanted weld defects.
TOf D I'E PE X-ray Gamma Vlanual

\{eander l-ine r.ry UT
scan scan
Figure 2-5 NIL/KINT thin plate project results (8)
A11 plates were examined with the techniques shown; and later the results were verified by
destructive testing. From these results, it follows that the Mechanised Pulse-echo Meander Technique
results in the highest probability of detection (PoD) and TOFD delivers the lowest False Call Rate. It
should however, be kept in mind that the Mechanised Pulse-Echo Meander Technique (2 axes
motorised raster scan) requires a total scanning time which is many times that of TOFD.
35
Chaoter (2): The Principles Of TOFD
t\/a
These studies show that although the original goal of TOFD was to provide improved sizing
capabilities, it is now acknowledged to have the added benefit of providing a fast and reliable
detection method.
Mr. Bill Browne presented a paper on the internet in September of 1997,2, where he laid out some of
the basics involved, but also made a loud statement on the "mystique" that had built up around. TOFD
by that time. He commented on the misunderstandings of TOFD:
"Some emanntes from the mystique built up by those responsible for its
of this lack of understanding
introduction. For many years scientists promoted the technique as a highly specialised 'sizing' -tool - so complex
qnd unsubstantiated
that it required their specialist knowledge and sophisticated technology to ffictioely apply -
claims were made about what the technique could and could not achieae.
This may hqae been the case in the aery eaily stages of eoolution but TOFD has noro been nround for more than
25 years, its effectioeness has been prooen by performance demonstration (more so than qll other inspection
methodologies combined) and whatez)er mystery once surrounded the technique has been completely dispelled by
repeated applicational success in the hands of routinely qualified personnel using what is now relatioely
' st andar d' e quipment. "
2.2 GENERAL PRINCIPLES

Ultrasonic inspection methods using pulse-echo techniques have primarily been based on specular
reflection. This implies a "mirror-1ike" reflection of sound from the probe, off the flaw and back to the
probe again. In traditional pulse-echo methods attempts are made to provide a beam path to the flaw,
such that the beam strikes the flaw, perpendicular to the plane with the flaw's major dimensions. This
is intended to provide the maximum amplitude response in pulse-echo mode.
But trying to arrange the beam to always impinge on a flaw, such that a perpendicular incidence
occurs, is virtually impossible. This accounts for some of the lack of detection by manual ultrasonic
methods. Even a large, flat reflector could have a 20dB drop by the beam striking it only 5" off the
perpendicular. This could mean that even large flaws could be missed, if "detection" is based on
signal amplitude exceeding a threshold.
But sound waves have several possible interactions, when they strike a boundary. Reflection is only
one thing that may occur. Refractioo mode conversion, polarisation, attenuation and diffraction are
also possible. Of these boundary interactions, diffraction is the important consideration in TOFD
studies (naturally, since the name, time-of-flight diffraction derives its name from the phenomenon).
Diffraction occurs when an obstacle distorts a wave, e.g., if the wave travels through a gap it may be
diffracted. It is the ability of a wave to spread around corners. Hence, diffraction occurs at the edge of
an obstacle.
The typical examples used to illustrate the effects of diffraction are slits (openings) in obstructions in a
"ripple-tank". The "wave source" generates a pulse that moves towards the slits and the opening
allows a portion of the wavefront to pass through. As the wavefront moves through the slits, the
centre portion of the slits allow the wave to move forward unimpeded. However, the portions that
touch the slits are retarded and bend the wavefront away from the path along the central axis.
36
'a. . .:4. .+.-

7: '2.: .?r..
?2.:'
7ya&ou,r@. ./r;: .?.2-:
Pulse source ',.tv
.. / . .:.
=
'-%.
z,
...,.
'.9.
:&4
'2. '...4...;
.-Z- :VZ::::
I
I
I
/@4,;
lui3.
7a..: ?4::
"W: '&-
.7-.:.
74-:
,*:-
I
, '12. 2,-
;:2 .:.2;.
'% :-. &
&&-; ,-..
zi aza a?; .4..
.?ra-z
.&
%&; .tu -:.
ww,a
:
..- ..- %..
E. W.
%,,
W..
wtu..
'.w. k- ;..-
. @ v;,
Difftacted 'WWE
:-WWZ.:
9; 7; 2;.- 7'::'
wavefront :?
-k. '..?.{3 ' t 7 '' ::?'' Diffracted
& ,4
i; , '': a:-: t-', WavefrOnt
% ;l
W i, ,/
,/
l;,
W- '2.t
%
EL
&;.1 :1
W
E;: :i
'& 4l
Single slit :; Lds bbstruction
t* ::
:&-il
Figure 2-7 Dilfrachon effects i n a ripple tank simulaLion
Diffraction has been discussed in ultrasonic basics. The single slit example in Figure 2-7 is also typical
of illustrations used to explain beam divergence. The illustration on the right side of Figure 2-7, with
the model of the long obstructiorL is representative of a planar flaw in the soundpath of a pulse
radiating from a probe.
A photoelastic image of the effect is provided in Figure 2-8. Here a plane-wave pulse has moved from
left to right and the diffraction patteffL off the ends of the notch, is seen as well-formed circles.
Figure 2-8 Photoelastic image of a compression mode pulse diffracting at a flaw
37
ES Of TOFD
The fact that TOFD relies on the detection of diffracted waves, makes the technique significantly
different from pulse-echo ultrasonic methods.
perhaps, not obvious, there is an underlying problem. Unlike a reflected wavefront, the diffracted
wavefront is very weak. Whereas, a large amplitude signal from a reflected wave is generally
(although not always correctly) assumed to have originated from a large flaw, no such assumption can
be made in TOFD, since the amplitude of any diffracted beam will always be relatively weak when
compared to a reflected beam.
Detection and timing of the low amplitude signals, diffracted from the tips of defects, forms a basis for
TOFD. These provide information from which we can derive absolute position and size measurement.
The position and size of the indication in TOFD has little to do with amplitude resPonse.
Before moving on, it is important to note that a diffraction effect can be detected in essentially all
d.irections. This is 4ue to the spherical nature of diffracted waves. Strictly speaking, it is therefore
possible that the principles could apply to the pulse-echo technique. In fac! there is a sizing method
that uses this effect. The Tip-echo technique, popular in japan in the 1980s and documented by Gruber
at the Southwest Research Institute in 1980, uses variation on the backscatter effects of the specular
and tip diffraction echoes. The concept in Figure 2-9 illustrates the signal received from a standard
pulse-echo probe with a transverse mode directed at a surface connected notch.
Figure 2-9 Tip-echo sizing principles for backscatter sizing
Although this is a tip-diffraction sizing method, it is not generally considered TOFD. The TOFD
concepts developed in Harwell are based on forward scatter. This is achieved by using two separate
transducers in a directly opposed tandem configuration. The probes should be well matched with a
short pulse (i.e. broadband), wide beam (i.e. generally small element sizes) of the same refracting
angle. One probe is connected to the transmitter of the ultrasonic instrument and the other to the
receiver.
The transmitted pulse is no different from the transmitted pulse in a pulse-echo setup. It experiences
losses via divergence, absorption and dispersion in the material under test. Further losses occur as a
result of reflections by flaws within its effective envelope. The portions of the pulse-energy diffracted
by the flaws are the very low amplitude signals that TOFD uses to assess the material integrity.
38
The individual A-scans (i.e. the echo responses from a single position in the test piece) are not
generally informative to the operator. OnIy by capturing a series of these responses and processing
them in a useful fashiory can they be used to discriminate flaws from background. Thery with the
proper time measurements, the flaw depth and size can be calculated.
There is still a requirement to ensure that adequate instrument gain is used in the process; however,
because the technique does not rely on threshold amplitude for detection of the relevant signals, it is
not amplitude dependant (for defect size measurement). As a result, TOFD is not as susceptible as
pulse-echo testing to inconsistent surface conditions and couplant conditions.
As noted, the normal TOFD setup uses a pair of closely matched probes. Because ths element size is
relatively small, the beam is divergent (almost spherical instead of planar). This provides extensive
volume coverage, so that the technique is not as dependent on variations in probe position or defect
orientation relative to nominal probe angle, as would be traditional pulse-echo methods.
Although most TOFD is done using compression mode, it is not always necessary and sometimes
shear mode can be advantageous (although potentially difficult to analyse).
It is important to note that there are many ways in which the TOFD technique differs from standard
pulse-echo; however, the most obvious difference is probably the display of the signals. TOFD was
introduced to NDT just as the computer age was dawning. It has arguably been a direct result of the
availability of computer capabilities that has made TOFD, the success it is today. Without the ability to
collect and display the signals used for TOFD analysis, the use of simple A-scans as used in manual
pulse-echo ultrasonic testing would have rendered the new technique nothing more than an
interesting lab experiment. Because so much of the TOFD technique relies on computing capabilities,
further coverage on computing and digitisation will be provided.
2.3 DIFFRACTION AS IT APPLIES TO TOFD

The accepted, basic TOFD setup is achieved by
using two separate transducers in a directly
opposed tandem configuration. The probes
should be well matched with a short pulse and
wide beam. This is usually accomplished with
small diameter probes that are highly damped
(broadband). One is connected to the ultrasonic
instrument as a transmitter and the other as a
receiver.
This results in a characteristic signal pattem

when testing welds, in plate that is not
excessively thick. The setup and the associated
signals are drawn schematically in Figure 2-10.
Figure 2-10 Basic TOFD setup
39
Chapter (2): The PrinciPles Of TOFD
Several aspects of Figure 2-10 must be noted, in order to recognise some of

the essential features of
detected at the receiver'
TOFD. The transmitter produces a single pulse which provides all the signals
of the signals indicated' The figure
Only the soundpath distances determine the different arrival times
assumes a single compression mod,e is present and ignores mode conversions'
Lateral wave is the
the shortest metal
term given to the .o*pr"rsion mode that is just below the test surface and has
not too a portiolr of the
distance between the transmitter and receiver. If the plate thickness is great,
pulse will travel to the far side of the plate and reflect, to produce the signal
identified as the Back-
wall. Flaws within the plate thickness will be seen at points between the lateral wave and back-wall
signals. Figure 2-10 indicates the signals are viewed in the un-rectified
mode (i.e. RF) So as to provide
as having an initial
phur" irrfor-ation. Assuming the lateral wave is a reference phase, it is illustrated
then have a negative phase, with
positive rise. The diffracted ,pp". tip and reflected back-wall signals
,espect to the lateral wave. When diffraction is detected from the lower tip, its
phase will be the same
as the Iateral wave.
Diffraction effects have generally been theoretical discussions in most ultrasonic training
classrooms'
Although similar effects can be represented in ripple-tanks, the complexities of events in a solid are
not posslble to duplicate in a rippie tank. However, the effects are possible to see using
photoelastic
the sequence
visualisation. A series of photos have been made using photoelastic imaging to illustrate
images show four
of events and the wave-modes present in a TOFD setup. The following photoelastic
exposures of a pulse, generated by a TOFD setup interacting with a square notch
in a glass model' The
produce a nominal
probes are 12mm diameter SMHz elements mounted on refracting wedges, which
60o compression mode in the glass sample. Since this is a sub-critical
angle for the compression mode,
Snell,s Law indicates that the conditions will also result in a 30o transverse
mode. Whereas the
schematic, in Figure 2-10, simplified the process by indicating just the main compression mode
interactions, photoelastic imaging reveals the mode conversions to transverse
mode as well'
Descriptions of the events along the pulse-path are included, following the figure
in each step' The
has been optically
notch is 1mm x 3mm and the vertical line is merely the seam where the glass
bonded to prevent acoustic reflections at the bond boundary'
Step 1 of the TOFD Process is seen
in Figure 2-11.. A Pulse is seen Probe
leaving the wedge with 65'L-mode

and 34"S-mode well formed. Arcs
due to the diffraction effects from
the upper and lower tiPs of the
piezo-element are seen for both
modes. Also, a well-defined shear
head wave is forming as the lateral
wave interacts at the near surface, at
glancing incidence.
I
Figure 2-L1 TOFD process steP 1
40
In step 2, see Figure 2-12, tl:re 3rr.rr.

notch face reflects the L-mode and
mode converts a portion to a strong
reflected S-mode. The upper and
lower tip signals have formed before
the shear head wave arrives at the
top of the notch. The initial
compression mode has spread and
reflected off the back-wall to form a
weak reflected L-mode, but the
process also creates a strong mode-
Figure 2-12TOFD process step2
converted S-mode.
By Step 3 of the TOFD process, see Figure 2-13, the lateral wave has reached the receiver wedge
position, at the same time as the arc from the upper tip diffracted signal is about half way from the tip
to the surface. The lower tip signal from the compression mode lags the upper tip signal and in this
image it is obscured in the mode-converted shear that has formed off the upper tip. The tip diffracted
signals are seen to have the shape of circles centred on their respective upper and lower notch
surfaces. The shear head wave is about to interact with the notch and it will diffract a separate pair of
tip signals. Although it is not possible to identify the modes of all these signals on the A-scan, we can
identify the mode in the image by noting the wavelength differences.
Figure 2-13 TOFD process step 3
By step 4, see Figure 2-1.4, lhe process is ready for the receiver to detect the L-mode upper tip
diffracted signal, which is followed shortly thereafter by the lower tip. Faint tip signals can be seen
from the interaction of the shear head wave with the notclr, but these will arrive much later than the
tip signals off the compression mode. At the lower right we can see that the upper tip arc is about to
meet the point on the back-wall where the S-mode is reflecting off the back-wall. At the point that the
S-mode reflects, we also see a strong L-mode-converted signal.
41.
Figure 2-T4TOFD process step 4
In order to allow interpretation of the TOFD signals, we require a sensible means of data presentation.
Capturing several A-scans is necessary and results in:
o Large amounts of data collected
. Phase information requirement
o Detection of small amplitude signals for displa,v
This has best been accomplished using greyscale imaging of the RF signals.
2,4 DATA ViSUALiSATIONJ

Earlier efforts at a TOFD display used isometric projections of the full A-scan waveform. Although
some degree of assessment was possible, the demand for accurate knowledge of the phase information
was not well addressed by the sort of display seen in Figure 2-15.
Figure 2-15 A version of isometric A-scans
The most effective option was to convert the amplitude displacements, on a normal A-scan, with a
small bar of grey shading representing the amplitude at each point along the time base. Each A-scan
then became a line of shading with the zero voltage indicated by a mid-range grey and maximum
positive and negative voltages indicated by the extremes (i.e. black and white). The basics of the
visualisation are seen inEigtre 2-76.
42
I{hite
G
Time
One A-scan is replaced by one grey-coded line ,
Figure 2-15 Basis for the TOFD greyscale
The greyscale display is used for the overall display when a scan is completed; however, it is a
requirement of the TOFD system to store the underlying A-scans that made up the greyscale image.
A single line of greyscale is not very informative. The great strength of this display is seen when
motion is added and an A-scan is capfured and converted to grey-scale at regular intervals along the
scan. The effect of adding motion to this display is seen inFigure 2-17.
Back-wall
wave
Figure 2-17 Accumulating A-scans to build the TOFD display tto)
Probe motion in TOFD always maintains a constant separation of the probes. They must constantly be
aligned facing each other. Since most (but not all) TOFD inspections involve weld inspection it is
normal for probe motion to be made along the length of the weld. However, in some situations the
operator may find it useful to move the probe pair perpendicular to the weld centreline. Of course,
this motion may be limited if the weld cap has not been removed.
Terminology has evolved from relative probe motion and the associated resultant displays. This has
provided a source of some confusion and even some experts are often inconsistent in usage. Figure
43
2-18 illustrates that the probe motion relative to the beam direction defines the scan type. \zVhen the
probes are moved along the weld length, the motion is perpendicular to the beam directiory so the
scan is called a non-parallel scan. When the probes are moved in the same direction as the beam, the
scan type is called a parallel scan.
Non-parallel scan
i
ilE!&rtt
.k*{}r$q*
nri{ qrf**?}
Parallel Scan
Itr;irqr qi I
ri ,(r{rI I
?r*l:s$
t;!x lr&ttu{lt
eC! cx!*tti
Figure 2-18 TOFD scan tyPes (11)
To some users, this has rationalised a difference in the data display naming. For many decades a cross-
sectional display of ultrasonic data having the upper surface and lower surface displayed has been
called a B-scan (the A-scan reserved for the scope display and C-scan for the plan view). Since the
introduction of TOFD, some have named the display formed by the non-parallel scar; as a D-scan
display. This same group reserves the display, formed from the parallel scan, as a B-scan.
This terminology is not consistent with all users. In fact, the de-facto Code on the matter (EN 583-6
Non-destructive testing - Ultrasonic examination - Part 6: Time-of-flight diffraction technique as a
method for defect detection and sizing), uses the term B-scan for both displays.
2.5 ADVANTAGES AND LIMITATIONS OF TOFD

If one were to listen to some of the proponents of TOFD, it would seem that TOFD is the panacea for
inspection problems. This is clearly untrue. It has its advantages and limitations, like any NDT
method. Depending on the applicatiory TOFD may stand as a useful option on its own. In other
situations it is best used with support from other NDT methods or as a support option to other NDT
methods.
44
A brief list of TOFD pros and cons should help the practitioner decide how and when to best use this
NDT tool.
2.5.1, ADVANTAGES
2.5,1.1. REPEATABILITY
real units
TOFD, especially when used with a positioning encoded system, provides measurements in
(e.g., millimetres) that are much more useful to engineers than dB's or equivalent scales of response' A
to TOFD
scan made of a weld with a TOFD setup, by one operator, will essentially be identical
a scan
ideal for
made by another operator, assuming both use similar probes and settings. This makesTOFD
flaw monitoring.
2.5,1,2 ACCURACY
Levels of accuracy attainable by TOFD are quoted to be within + 0.5mm in terms of through-wall
notches, but
extent. This might be a bit optimistic when dealing with real defects instead of machined
a standard deviation of error in vertical sizing of 1mm is perhaps not unrealistic with a good
technique.
Depending on the nature of the defecf length sizing on the order of 1-2mm is also achievable. Position
along the weld and with respect to the weld centreline can usually be established to within 0.5mm;
and angular dispositions can be resolved to within a few degrees, when appropriate scan procedures
are used. This accuracy and reliability makes TOFD a suitable NDT tool for fracture mechanics
assessment; otherwise destructive methods and physical measurement would be required'
2.5.1.,3 DATA STORAGE

As with most computerised NDT systems, TOFD systems now provide digital storage of all relevant
parameters. Scan parameters and acquired scan data can be retrieved and redisplayed at any time'
position information ensures that flaw locations can be reliably identified and results from periodic
inspections can be compared for changes. Digital data storage allows elaborate analysis techniques
including noise-suppression, pattern recognition processes, signal subtraction and extraction and a
variety of other digital signal processes (DSP).
2.5.1..4 DETECTION SPEED
When a TOFD scan can be made using a single non-parallel scan, results are fast (almost immediate)
and limited only by the scanning speed. Concerns with manual UT, for reliability being compromised
by scanning speed does not apply to TOFD. The many round-robin comparisons have shown that
TOFD is fast and efficient, providing some of the highest probability of detection of any of the NDT
methods.
Commercial considerations make speed an important factor. TOFD can advantageously provide both
detection and sizing from the same data, without recalibration and rescanning, which has a significant
bearing on time and cost.
45
Chapter (2): Ihe Principles Of TOFD
2,5,1,5 VOLUME COVERAGE

Most NDT methods have volume coverage and resolution linked to speed. However, TOFD provides
volumetric coverage by linearly scanning wide beam transducers at relatively high speed. Position
and inspection raw data are acquired and displayed in nearly real time. Therefore, unlike radiography
where thicker sections would require longer exposures on slower film to maintain the same
resolution, it is possible to inspect thicker sections, without a reduction in speed or resoluticin when
scanning with TOFD.
The full volume of weld material and HAZs are tested at speeds which cannot be considered by
conventional manual ultrasonic methods, but are accomplished with TOFD. Scan rates of 100mm/s to
150mm/sec are commor1 using even manual probe movement. Thus higher rates of coverage can be
achieved when automated scanning is used. It should be remembered that these rates relate to the
length of weld volumetrically inspected in a single pass of the transducers and not just to the scanning
speed of the probes.
2.5,1.6 SENSITIVITY
Sensitivity of TOFD can be an advantage or disadvantage. It depends on your point of view. TOFD is
generally configured to see ezterything.This can mean everything right dor,r,,n to grain structure. \Atrhen
the test specimen is relatively clean or the material is highly refined, there is no issue with high
sensitivity. Flowever, where the test material contains many major anomalies to be reported or in
coarse material (e.9., some stainless steels) where the grain boundaries are on the order of the size of
the flaws, TOFD sensitivity can be construed as a hindrance and, in certain circumstances, can make
interpretation and evaluation a time consuming ordeal, or not even feasible.
\rVhen the data storage advantage is considered, in light of sensitivity, it might be noted that one of the
features of digital processing is the ability to increase gain via software. That means that small (un-
saturating) signals can be increased, after data collection.
2,5,1.7 EASY DISCRIM/NATION OF DEFECTS AND GEOMETRY
A common problem experienced in manual ultrasonic testing of welds is the issue of the operator's
skill in differentiating between flaw signals and signals originating from surface geometries. \zVhen
TOFD is carried out on a butt weld, where the root and cap re-enforcement are not removed, the
TOFD data display can provide an unambiguous discrimination of flaws from weld re-enforcement
metal.
2,5,1..8 FLAW OR/ENTATION

Because of the omni-directional aspect of the spherical diffracted signals, TOFD is sensitive to
virtually all types of defects regardless of orientation. This is partly attributable to the wide angular
coverage of the divergent beam used. Providing the flaw falls within the effective beam envelope, the
low amplitude signals diffracted from the flaw edges will be captured and displayed in correct
relative position.
46
2.5.1-.9 COUPLING STATUS

TOFD data can be collected by manual or mechanised methods of probe motion. Any manual
ultrasonic operator, doing pulse-echo testing, monitors the A-scan for scan quality. They can recognise
when the coupling is not as effective, by a loss of the grass level. However, in the case of TOFD
scanning, the operator does not monitor the A-scan and when scans are lengthy or mechanised; the
operator has no sense of the coupling condition by simply looking at the probes moving.on the
surface. By observing the data collected for the lateral wave amplitude and the associated grain-noise,
the TOFD display is an effective indicator of how well the probes were coupled. Maintaining coupling
is made somewhat more difficult than with standard manual scarrring, because both the transmitter
and receiver must be well coupled to the test surface.
2.5.1.10 REDUCED OPERATOR RELIANCE

Since TOFD data can be collected and stored to a computer file for later analysis, it is possible to
reduce the reliance of the test on the probe operator. Many applications can be configured by a senior
operator and then the data acquisition assigned to a field team. This might consist of a person that
operates the computer data acquisition unit and another that pushes the probe along the weld.
Sufficient experience and competence is required by this team to ensure that the data collected is
adequate. Final assessment and evaluation can be carried out at a later time by the senior operator.
2.5.2 LIMITATIONS
2.5.2.1 WEAK SIGNALS
Typically the diffracted signals associated with TOFD are 20-30dB lower than those associated with
specular reflections using pulse-echo techniques. This tends to put a strain on the ultrasonic receiver
units and most are operated near their maximum amplification capabilities. Electrical noise is a
common problem with many TOFD systems. Attempts to reduce this noise generally involve the use
of pre-amplifiers near the probe or remote pulser/pre-amp combinations.
2.5.2,2 DEAD ZONES
The most widely accepted limitation to TOFD is the loss of information due to ring-time. This is
especially noticeable at the entry surface. However, a similar zone occurs on the far side (back-wa1l).
[1eu7ns(e) points out that TOFD does suffer from a near surface effect, caused by its inherent lateral
wave; but he considers that this is not a serious problem unless very near surface sizing is called for.
He further points out that very few near surface embedded defects can be considered integrity-critical.
It is also debatable whether the near field characteristics of single-element compression probes and the
inherent dead zone effects of twin probes could provide better resolutiory using conventional
reflectivity methods. Radiography rr.ay (or may not) have some improved near surface sensitivity, but
provides no worthwhile through-wall positional information.
If such defects are of genuine concerry then a combination of TOFD and remote field eddy current or
ultrasonic 'head wave' testing should be performed.
47
2.5.2.3 FLAW CLASSIFICATION LIMITATION

Flaws detected by radiography are generally given some form of identification; e.9., slag, porosity,
lack of fusion. Manual ultrasonic tests attempt to provide similar identification by characterisatiory via
echo-dynamics of the signal, with respect to probe movement. TOFD provides no opportunity for the
echo-dynamic assessments seen in manual methods.
Because of the limitation on the characterisation capabilities of TOFD, a simple but very useful scheme
is recommended. TOFD signals may be grouped into one of three flaw types:
r Point-like
. Threadlike
. Planar
Point-like flaws have neither length nor height (typically these are isolated pores or very small
inclusions). Thread-like flaws have length but a vertical extent cannot be assessed from the TOFD
display. This means that the lower tip signal cannot be discerned. Planar flaws have both vertical
extent and length (meaning a lower tip signal is discernible).
From a fracture mechanics point of view these categories are all that is required to assess the effect
that a flaw could have on the intended use of a component.
48
Chapter (3): Instrumentation Used In TOFD
CHAPTER (3): INSTRUMENTATION USED IN TOFD
An ultrasonic system for TOFD must be capable of providing a means of transmitting and receiving
ultrasonic signals. In this regard it is identical to the instruments used for manual pulse-echo UT. In
additioru a TOFD system must be capable of storing, displaying and analysing ultrasonic signals. As
well, there needs to be consideration for the mechanical aspects, in that it must provide fixed spacing
between the transmitting and receiving probes. Although not essential, it is generally considered
preferable to ensure that probe motion is encoded and that the position of the probe pair is maintained
within prescribed tolerances, with respect to a reference position, such as the weld centreline.
3.1 BACKGROUND OF THE ELECTRONICS

Ultrasonic inspection technology has come
a long way from the early units using
single channels and electron-beam
Cathode Ray Tubes (CRTs) like that seen in
Figure 3-1.
a lh-,
t rH,
U:
iqei*s
&J r&d a*
Jk8rliam
,e$r
In the early stages, industrial ultrasonic

''j t
instruments were entirely analogue based.
In fact most systems still used valves
(called vacuum tubes in North America).
As solid-state electronics became more
ffi
atSa
c$
"reer,tq m
Cer*Eqr
s#
*#*
&*#Ek*
popular, instruments grew smaller.
"
g*€
Eventually a two-sided electronic structure i *€**
was developed; one side of the electronics
being digital and the other, analogue. Figure 3-1 Krautkramer USIP1O circa 1950's(12)
Primarily, it is the control section of the

instrument that is digital. More recently
the out-put data is also in a digital format.
The concept of digital instruments ultimately me;U:ls that the components are operated in a binary
condition: the item is set, either 'on' or'off'. This has opened the way for computer control of the
ultrasonic instrument. Small programmes on EPROM chips (erasable programmable read-only
memory) are now common on portable instruments and by stepping through a variety of
programmes, the instrument can be made to operate under toggled controls and even display the A-
scan as a digitised representation of the analogue output.
As a result of the analogue origins to ultrasonic instruments in the mid 20th century, there are still
some residual anachronisms brought forward into the 21't century. For example, although the A-scan
display is now a reconstructed image on a computer monitor (complete with graticules and voltage
representation of the signal), codes still require assessment of the display's vertical and horizontal
49
linearity. These features were monitored in the analogue days, when the old CRT was used and the
deflector plates could be knocked out of alignment.
In-depth knowledge of NDT electronics is neither required nor expected as part of the certification
programme for a TOFD operator. However, overview coverage of the main concepts is expected and
will be helpful in understanding some of the component functionality'
Topics addressed in this chapter will include information on how the pulser provides the excitation
voltage, how the receiver processes input voltages, motion controls, data acquisition, signal processing
and display formats.
Motion control itself is not an NDT topic. It is introduced here because it is now a common aspect of
advanced NDT systems. E.g., although it may be possible to hand operate a TOFD probe pair; this is
not usually the case in industrial applications. With the requirement for full data recording of
ultrasonic signals in TOFD, it is common to mount probes in some form of a holder and some degree
of mechanisation is used to manipulate the probe and collect the ultrasonic signals during the motion
process.
With the use of the TOFD greyscale display, all TOFD instruments now have some aspects related to
computers and hence digitisation of output. It will therefore be necessary to address some of the basic
concepts of digitisation as part of the training in TOFD.
3.2 PULSERS AND RECEIVERS

With the advances in solid-state electronics, ultrasonic pulser-receivers are another example of
miniaturisation. The 20 kg suitcase-size boxes of the 1950s have been replaced by small packages held
in the hand. The pulser itself can now be found as a single chip component only 9mm x 9mm x 1mm
high. Receivers and processors are added to a printed circuit board, populated by electronic
components, including TCG and gating circuitry, A/D converter and time delay circuitry.
Board size is often based on a standard such as seen in Figure 3-2. The instrument in Figure 3-2 is
relatively small, but an even smaller version can be found.
Figure 3-2 Single charu:rel UT system 170mm x 50mm format board for ROV - subsea application {t)
50
Chapter (3): Instrumentation Uq94la lqIP
Whether or not, the parameters of the pulse to the transducer are digitally controlled or not, the pulse
itself is an analogue signal. Similarly, the ultrasonic vibration that a transducer senses from a
reflection generates a voltage across the transducer that is also an analogue signal.
The quality of the pulser and the receiver has a great effect on the information obtainable in ultrasonic
testing. The following will consider some of the options and their features.
3.2.1 PULSERS
Essentially all that is required to vibrate a piezoelectric Spike pulse
transducer is an alternating voltage. Howevet,

characteristics of the pulse voltage will dictate, how the
element vibrates. This is similar to pushing a Person on
a swing. If the pushes are applied at the natural
frequency of the loaded swing, large amplitudes can be Tone burst pulse
achieved. If no! a rough, low amplitude ride, results' In
ultrasonic testing, maximum amplitude displacement is
not always desirable. When precise timing is needed (as
would be for thickness tests on thin wall material), short Square wave pulse
duration pulses are better. Even a 'ringy' probe can be
made to dampen its vibration with the correct pulse
characteristics.
Figure 3-3 Pulse types
Three pulse shapes are commonly used in ultrasonic
flaw detection units: spiked, bipolar tone burst and
square wave. These are illustrated in Figure 3-3.
3.2.1..1 SPIKE PULSERS
Oscillating pulses are generally formed by a Thyristor

capacitor discharge. \Alhen the capacitor (srvitch)
discharges, the rapidly applied voltage across the DC power
Transducer
transducer causes it to vibrate. The purpose of the supply

damping resistor (inductor) is to increase the rate
of voltage decay. This is the damping available to Charging
resistor
the operator and is used to control the ring-down Cheirging
time. The old thyristor-style pulser is illustrated
in Figure 3-4.
Figure 3-4 Spike pulser
3,2,1..2 TONE BURST PULSERS

Tone burst pulsers allow maximum energy output from transducers by adjusting the frequency of the
voltage applied. This can be done in several ways. A chopped voltage from a waveform generator
allows the selection of pulses of different shapes, frequencies and durations, as in Figure 3-5.
51
#v/ ifu il]l ifu

Sine wave
ilAi\t\l !l\/
j]rt
ill,Al
ilt
Saw-tooth wave
+ +r 1 Bi-directional square wave
-[rfl,.l] ulrlt*Li UIU LIJiI illl

1III]]
Negative square wave
tLrlLILtL t],l'Ul L-t]-Il-l- l Jilt tllt tLit

l- lXt Positive square wave
Change shape Change frequency Change ring time and

pause time between pulses
Figure 3-5 Tone burst pulser waveforms
Tone burst signals usually consist of several cycles. These are preferred for velocity determination
using interferometry. As well, very high frequencies can be derived using tone burst pulsing. Thus it
is used in acoustic microscopy, where frequencies in the gigahertz range are used.
3.2.1.3 SQUARE WAVE PULSERS
Square wave pulsers have become the preferred laboratory style of pulser, both for pulse-echo and
TOFD ultrasonic systems. Similar to the spike pulser, the square wave pulser charges a capacitor,
which discharges across the transducer. Holding the switch closed in the circuit, for a controlled
amount of time, then rapidly restoring the pulse voltage to zero, causes two displacements of the
transducer. The displacements at the transducer are opposite in phase, so by timing the recovery
voltage, a constructive interference can be effected between the original backward moving wave,
reflected off the probe backing and the second impulse from the pulser.
Adjustments of the pulse voltage and the pulse width are possible, thereby making square wave
pulsers a versatile tool to optimise transducer performance. By choosing the best pulse width to obtain
constructive interference, less voltage need be applied to the probe, thereby reducing noise level. By
pulsing at a frequency higher than that for maximum output bandwidth can be increased and lower
frequenry components reduced. Pulsing a transducer at a frequency lower than that for maximum
amplitude, increases mechanical damping and provides a sharper cleaner signal with little ring-on.
These features of the square wave pulser are illustrated in Figure 3-6
52
Bac
In Figure 3-6, the square wave pulse 1?.0m 17.500 l8.st8
shape is shown on the left. In the first

case, the pulse is set to a fairly short
duration, 12.375 nanoseconds (ns) and
the applied voltage is -498 volts. To the
right of the received pulse shape is the
signal of a co-polymer transducer,
nominally 30 ll/Hz, using a glass target. l7-50{J 13 ntl
Signal amplitude is relatively low,

implying that the transducer output is -uidtI
25. tso rs
not maximised. However, an advantage v0lff
to this signal is that none of the low -s47,843 V
frequency components of the transducer

are excited, and the bandwidth is high.
&E
1?Jr0 1t.!90
In the second case in Figure 3-6, the
pulse width has been adjusted to
provide a maximum output from the
transducer at, 25.15 ns. Voltage applied
is -547 volts, up slightly from the first
case (in the first case, voltage was the
maximum possible, for the pulse width Figure 3-6 Pulse width effect on transducer excitation
applied).
Lower frequency components may be added to this signal, compared to the first case, but the
bandwidth is reduced by the greater output near the resonant frequency.
In the third case in Figure 3-6, the pulse has been increased to 51 ns and the voltage is essentially the
same as applied in the second case. Transducer output is reduced and the ring-on is virtually
eliminated. It is therefore possible to increase the damping of the transducer's vibration by increasing
the pulse length beyond the resonant frequency.
The middle example in Figure 3-6 is indicated as having the maximum output using 25ns pulse
duration. This suggests that the natural frequency of the probe is significantly different from the
"nominal" frequency. Maximum amplifude output occurs when the pulse width matches half the
period of the natural frequency. This implies that the natural frequency of the probe used in the
illustration is actually 20MHz instead of 30 MHz. For a true 30 MHz probe, the pulse duration would
be 76.7 ns for maximum amplitude output. At 12.375 ns and 50.699 ns the examples indicate the
effects of pulse durations nearly half and double that for the natural frequenry. Adjusting the pulse
duration outside that range usually results in significant reductions in output amplitudes.
TOFD probes are usually designed to produce a relatively short ring-time in order to minimise the
entry surface dead zone. To achieve reasonably short pulses, the TOFD probe is typically designed
with mechanical damping behind the element to produce pulses with about 80% or more bandwidth
(i.e.1.-2 cycles).
53
When considering pulse duration in a TOFD system, the operator should select the duration best
suited to the element. For a square wave pulse, the optimum duration is half the rycle time. For
example, for a SMHz probe, one cycle has pulse duration of 200 nanoseconds, so the pulse duration of
the pulser to excite the 5MHz element would be 100 nanoseconds. An operator, knowing that
adjustments, over and under the optimum, can affect the frequency output; can use this to their
advantage to reduce the ring-time, in order to reduce the dead zones in the technique.
3.2,1,.4 BANDWIDTH
Reference has been made to bandwidth in previous chapters. The term bandwidth describes a
frequency spectrum of pulses and receivers capable of amplifying them. Although we refer to a
specific frequency for a probe (e.g., its nominal frequency) in fact, unless the probe was being driven
in a continuous mode with a sinusoidal motiory it will not have just one frequenry. Instead it will have
a band of frequencies around the nominal value.
Two general terms are used to describe transducer bandwidth. Broadband transducers have a
relatively wide range of frequencies and narrowband or tuned transducers have a smaller range of
frequencies.
Assessment of the frequency content of a pulse can be done mathematically using a process called the
fast Fourier transform (FFT). This displays the relative distribution of the frequencies compared to the
energy of each frequency. Analysis of the FFT for bandwidth usually uses assessment of 6dB drop
from maximum amplitude. The selection of bandwidth is essential for achieving certain test results;
narrow bandwidth for highly sensitive testing or broadband for high resolution testing. For effective
phase-interference timing, phased array probes are usually considered to have a bandwidth of 80% or
more. Bandwidth is determined by looking at the frequency content of the pulse. It is determined by
the equation
BW =l(f"- ft)/f,l x 100 (3.1)
\rVhere;
fu:Upper frequency at the 6dB drop

fi: Lower frequency at the 6dB drop
f,: Centre frequency
Figure 3-7 illustrates the frequency distribution of a narrowband (left image) and broadband
transducer (right image). The images are also examples of symmetrical curve (left image) and
asymmetrical curve (right image). The term f, inthe image indicates the peak frequency. Note that the
maximum or peak frequency may not be the same as the centre frequency.
54
: Instrumentation Used In TOFD
lfc
t- tc
.-_ 6
-6dt] a
a
l. f,,
=
2 __t_____'1 i____:, -----i----
I
I 1i
"O'-T-x t- -4 5 5 ? e 5 to I :2 't---;---t-t 4 5 a ,
FFIOLE!CYifiH?]
B 'r 12
tgEoal€xgr li4&*
Figure 3-7 Transducer bandwidths - narrowband (left), broadband (right)tr+r
Figure 3-8 illustrates the relative bandwidth of a probe compared to its signal shape. A nominal 5MHz
pulse with about 5 cycles is seen to have a bandwidth of 33%; whereas, a pulse with 1.5 cycles is seen
to have a bandwidth of 80%.
Figure 3-8 Signal shape/ring-time (left) vs. bandwidth (right)
Hedrick Et a1(1s) defines broadband transducers as those having a bandwidth over 15%. Therefore, any
NDT work would use broadband transducers; and the concepts of broadband and narrowband
transducers in NDT are simply relative terms. TOFD probes generally have a bandwidth of 75"/o or
more (typically less than2 cycles).
55
3): Instrumentation Used In TOFD
3.2.2 RECEIVERS
According to Krautkramer(1), pulsers apply voltages of 100
to 1000 volts to the probe. However, received signals are
three to four orders of magnitude smaller (typically on the
order of a few millivolts e.g. 0.001V to 0.01V). This causes
several problems. One is the shock of the pulse voltage
that is transferred to the receiver in the pulse-echo mode
of operation. Another problem is the need to amplify the
relatively small signal from flaws, without amplifying
noise. The latter is further complicated because the
frequency of the received signal may not be the same as
the transmitted pulse envelope (accounting for even
smaller signals from the transducer).
Pulse€cho Conn€ctions
\zVhen switchedfrom pulse-echo to transmit-receive, there
is no longer a physical electric connection between the Figure 3-9 Pulse-echo (PE) versus
two components. Figure 3-9 illustrates this switching. Transmit-Receive (TR) connections
In TOFD systems, the separation of pulser and receiver is not actually required. F{owever, because
most TOFD ultrasonic instruments can be used as pulse-echo instruments, the switching is usually
built into all systems.
To obtain a signal capable of being displayed and subsequently processed, the received signal caused
by the small transducer vibrations must be amplified. The amplification Process is quite involved and
also includes filtering and sometimes subsequent attenuation.
The first stage is the circuit protection that protects the preamplifier from the pulser voltage when in
pulse-echo mode. The preamplifier can use transistor-type amplifiers that provide about 20 to 40dB of
gain. The frequency response of the preamplifier is usually broadband. Some high pass filtering may
be incorporated to improve signal-to-noise ratio by eliminating some radial mode components of the
probe and line interference. Preamplifier bandwidths are usually flat, from about 1 MHz to L5 }dHz;
and this is not operator adjustable.
Following the preamplifier, the signal is passed through a broadband attenuator. This protects
subsequent circuitry from saturation and provides a means of calibrated adjustment of signal height.
Attenuation is usually equipped with coarse (20dB) and fine (1dB) switching. Some portable
instruments have been made that use very large preamplifiers. As a result, even with maximum
attenuation, signals from normal beam inspections of plate could not be reduced to below full screen
height.
It is important to note that, when it comes to signal analysis, saturation is an undesirable condition.
Flaw signals are often required to be assessed for amplitude. F{owever, when an automated process
collects data at a high gain, the signals from some flaws may exceed 100% screen height. Usually this
means that the actual amplitude is unknowrg i.e. the only thing the operator can discern is that the
56
amplitude is greater than99.9'/" screen height. When absolute amplitude, relative to the reference, is
required (e.g., determine that a signal from a flaw is 740% of reference), it may be necessary to reduce
the amplification to avoid saturating signals. As well, many post-processing analysis techniques
cannot be done using saturated signals.
Attenuated signals are passed on to RF amplifiers, which can be linear or logarithmic. Linear
amplifiers are most commonly found on UT instruments. \zVhen using a linear amplifier for the
receiver, amplitude of a signal is proportional to receiver voltage. However, receiver gain control is in
dB increments; therefore signal amplification by 6dB gain doubles the signal height. This limits the
range of useful amplification to about 34dB (34dB raises a 2% FS}i' signal to 100%). \zVhen a
logarithmic amplifier is used, the scale is dB linear, so each increase of 1dB gain is about 1"/" of the
screen height. Expressed another way, the dynamic range of the logarithmic amplifier is 6.3 times
greater than the linear amplifier. Some logarithmic amplifiers can exceed the 100d8 dlmamic range
(i.e. 1dB gain results in something less than 1% FSH).
3.2.3 FTLTERS (BROADBAND AND BANDPASS)

Frequency filtering can be applied to RF amplified signals. Normally, bandpass filters are used to
eliminate noise from higher and lower frequenry sources. These are selectable by the operator and are
labelled to correspond to the centre frequency of the filter. Normally the bandpass filter is set to
correspond to the nominal frequency of the probe. Wideband filters are also available. The amplitude
of the signal as compared to the best bandpass filter does not significantly change, when wideband is
selected, but the signal will often be noisier (Figure 3-10). As a guide, bandpass filters are used in
pairs, with the low pass filter approximately twice the probe nominal frequency and the high pass
filter set to approximately 0.5 times the probe nominal frequenry.
l4ldrDdd FtU
td lllE Fk
-
r:- +, $E F-abr
-
i$ds of .6hc
:-.smffi
Frecluency ilvlHz)
Figure 3-10 Receiver filters
After a signal has been passed from the main receiver amplifiers, it can be further processed.
Subsequent processing includes:
. Video Smoothing
. Gating
r TCG/TVG
57
ter (3): Instrumentation Used In TOFD
3.2.4 VIDEO SMOOTHING

Video smoothing is a process whereby a curve is used to replace the rectified A-scan. This tends to
eliminate the bumps seen in the simple full wave or half wave rectification displays and can have the
added benefit of reducing the digitising frequency and also the amplitude error that result from the
digitisation process. Figure 3-11 illustrates the effects of video smoothing. Since smoothing is applied
to rectified signals, this process is not applicable to TOFD.
tlF Rectified Video Srnoothed
Figure 3-11Video signal smoothing
3.2.5 GATES
Essential to computer imaging of ultrasonic data is the ability to extract information from regions of
time, which can be selected, to monitor for signals in that time. The region being monitored is said to
be"gated". The time along the gated regiory or amplitude within the gate when a signal occurs; or
both the time and the amplitude, can be gated by the system. Alarm or recording thresholds can be set
for signals occurring in the gate. Gates are an essential component in automating inspection systems.
Gate positions are usually facilitated by auxiliary controls. Gate positions on the screen are noted by
extra traces or markers on the A-scan display. Gate controls include start and end adjustments,
threshold settings (amplitude which a signal must.reach, before alarmed or collected) and positive or
negative settings. If positive gating is used, a signal must exceed a set minimum threshold. If negative
gating is used, a signal in the gate must fall below the threshold, before alarming. Signal amplitude
monitoring for flaw detections is typical of positive gating. A coupling monitor using a through-
transmission signal that triggers an alarm (e.g., audio or visual), if the coupling signal has reduced
below a given threshold, is typical of negative gating.
3.2,6 TCG/TVG
In pulse-echo techniques, in order to display the same size defects, occurring at different depths as the
same amplitude on the display, different gain levels are required. M*y systems have the ability to
adjust the amplification with respect to time. When amplification is variable with respect to time, it
provides time corrected gain (TCG). This is also called time variable gain (TVG) or swept gain. By
allowing more amplification to be added as time or distance incteases, signals from reflectors of the
same surface area can be adjusted to the same amplitude, at any distance. Amplifying the response of
more distant echoes, avoids the inconvenience of distance-amplitude-correction curves, allowing an
alarm threshold at a fixed percentage of the screen height to be set across the entire screen.
Since TOFD is a non-amplitude based ultrasonic method, TCG is not used in the process; however,
since the same instrument may be used for both TOFD and pulse-echo (in fact many multi-channel
58
instruments can provide both techniques at the same time) the instrument being used for TOFD may
be equipped with TCG capabilities.
3.3 DATA ACQUISITION AND AUTOMATED SYSTEMS

Laboratory UT instruments are often part of a complex collection of hardware and software that can
be considered a data acquisition system. These can be used for precise material characterisations in
scientific studies or they may be part of industrial production systems. Features of laboratory
instruments and data acquisition systems can be applied to several methods of NDT, so a general
outline of concepts can be considered.
Collecting in{ormation about an object or condition is generally considered data acquisition. This
usually involves collecting information about one parameter with respect to another, e.g., monitoring
temperature against time, or acoustic velocity versus temperature. Data acquisition can be done
simply by an operator recording readings manually. In the time-temperature example, the operator
watches a thermometer (analogue or digital) and records the values of both temperature and time at
various time intervals. Records of a single parameter have little meaning unless they can be related to
something else.
Scientific and engineering applications, today, require very large numbers of readings to be taken with
exacting precision. Several hundreds or thousands of readings over several hours, or even over just a
few seconds, are easily accomplished using computers. \zVhen computers are incorporated into data
acquisition, the process becomes automated. Computers can then be used for not only collecting the
readings, but also sorting and subsequently analysing them.
Advantages of automated systems include:

o Accuracy . Repeatability
. Capability . Safety
r Consistency ' SPeed
o Cost
Accuracy, consistency and repeatability can avoid unnecessary replacement cost, if a test proves "No
deterioration". Conversely, minor changes, detectable only by the accuracy of an automated system,
can be the reason for concern to condemn a component and thereby save millions of dollars by
avoiding costly and untimely catastrophic failure due to an undetected change resulting from a
manual scan. The case for advanced automated NDT systems is clear.
In general automated NDT systems have the following:
o A central control system (usually a computer)
o A sensor and attached apparatus (the NDT equipment)
. A means of moving the sensor or part
o A means of collecting and displaying the output from the NDT apparatus.
59
3.3.1 SYSTEM COMPONENTS

COMPUTER - Central to the whole system is the computer. Computers come in various shapes and
sizes. There are two main types of computers: analogue and digital. Analogue computers are
somewhat archaic now and rarely found in common use. They are hardwired devices using current
flow and switches to address their logic functions. For our purpose, we will be concerned only with
digital computers.
SENSOR - The sensor is part of the NDT equipment. NDT equipment functions are familiar to any
experienced NDT technician. An important feature for an automated system is how !o use the signal
generated by the instrument. In some machines it is possible to provide an input to the NDT
instrument, thereby facilitating computer control of some of the instrument functions. This can allow
for remote control.
MOTION CONTROL - Motion control can be done in a variety of ways and can be as simple as
switching a drive motor on or off. Or it may involve complex closed-loop systems, controlling position
and velocity, based on a feedback monitoring system measuring torques, so as not to break the
inspection tool by over-straining it. For effective data display some form of positional information
must be added to the motion control. This is often accomplished by counting steps on a stepper motor,
simple timing, or most accurately by use of positional encoders.
DATA DISPLAY - The data display output is merely a means of providing a useful record of the test
results. This might be as simple as a voltage output, taken from the NDT instrument and displayed as
a line of varying position with increasing time (a strip chart recording) or it can be a complex set of
data points showing waveforms collected and processed for display with positional information.
The information collected by computer can further be processed to:
- Reduce noise (e.g., signal averaging)

- Enhance pertinent signals (amplitude i,,,.,i,.
colouring or signal processing)
- Correct for geometric characteristics
(e.g., SAFT)
Moaior co*rol
Figure 3-12 illustrates the basics of an inlertace

Ii;ffi,
', "G+
automated system showing the computer as
central to all of the inspection activity. A
PC-style computer addresses motion control
and collects analogue information from the
NDT inspection instrument. The results are
displayed on the computer monitor with the
option to print.
Figure 3-12 An automated scanning rig
60
3.4 MOTTON (AUTOMATED AND SEMI-AUTOMATED SYSTEMS)

3.4.1, MOTION CONTROL
Traditionally, ultrasonic inspections have
primarily been a manual function. With the
introduction of computerised systems
collecting data, probes are more often held
in a fixed guide mechanism. Probe motion
is then accomplished by advancing the
probe in a scan pattern. For simple
inspections this is usually a straight line
parallel to the weld axis. The most basic
form of probe motion uses an operator to
manually push the probe(s) along a
magnetic or mechanically fixed guide strip.
This form of motion would be considered
semi-automatic. These systems are
inherently error prone, and require the
operator to have a greater degree of skill
than is required with motorised systems.
Figure 3-L3 Manually operated scanner
Figure 3-13 is an example of a manually ESLinkScarurer(rrl
operated scanner incorporating a position
encoder. The sc;rnner illustrated is
comprised of a probe holder (holding 4
probes, PA and TOFD) and a spring-loaded
split-ring clamp that incorporates pairs of
wheels that straddle the weld. The ring of
paired wheels holds the probes in a fixed
position relative to the weld centreline.
When the weld is on a flat plate the
technique of using rollers does not ensure
that the probes will straddle the weld
equally. A magnetic guide strip can help
reduce probe wander, by presenting a flat Figure 3-14 Manually operated scanner - magnetic
stiff edge, against which the operator can guide strip used to keep probes tracking equidistant
hold one of the probes. Such a simple from weld(14
option is illustrated in Figure 3-14.
Today, motor control is an increasingly common means of moving the probes along the part. Motor
control systems make inspections significantly more repeatable. The inspection apparatus is simply set
up the same way, positioned in the same place and driven at the same speed. The only variable
61.
): Instrumentation Used In TOFD
remaining ought to be the condition of the part. Use of motorised systems significantly reduces the
opportunlties for operator error. This is because the operator is physically hands off while the
inspection is underwaY.
Furthermore, most motor-controlled systems use a built-in encoder to track the position of the
inspection in real time. This allows an operator to control position with much greater precision than
they could achieve manually. In additioru the operator spends a minimum amount of time'within
dangerous environments, such as near moving parts or in irradiated or hot areas. Finally, motorised
systems provide constant pressure and orientation of the probe. This maintains consistency
throughout the scan. Systems dedicated to pipeline girth weld inspection are an example of the
specialised application where a motorised inspection platform is developed for phased auay
ultrasonic inspections.
Figure 3-15 is an image of a
scanner mounted on a guide-
band clamped to the pipe. The
same guide-band that was used
to position the welding heads
can often be used by the UT
system. Multiple probes and a
couplant irrigation system are
also incorporated into these
systems.
Figure 3-15 Motorised system - Eclipse Scientific TDPipe-Runner{10)
3.4.2 ENCODERS
Although some equipment can be made to collect and display signals based solely on time (i.e' the
operator selects how many A-scans will be collected and displayed each second), such a display is
generally limited to static evaluations. As suclU using this system to collect data requires either very
smooth and constant speeds for scanning, at a set rate; or the assumption that defect sizing will not be
relevant, and therefore it will only be used to locate defects.
In order for it to be possible to measure the length and locatiory from the data collected, instead of
from the marks made by pencil on the test piece, a means of coordinating the travel of the probe to the
test surface is used. These tools are called positional encoders. They can be obtained in several forms,
but their operation essentially does the same thing.
Modern technology has made these systems extremely accurate and extremely small. In additior; they
are rarely, entirely mechanical in nature, but instead have become primarily non-contact systems. We
will consider the two main types: optical and magnetic encoders.
62
r (3): Instrumentation Used In TOFD
3.4.2.1 OPTICALENCODERS
angle and direction of a rotary
A rotary optical encoder is a sensor that uses light to sense the speed,
to provide the same information for linear
shaft. A linear encoder reads a linear strip instead of a disk,
position, so they are inherently free
motion. optical encoders use light instead of contacts to detect
(no contact bounce). Accuracy of an optical
from contact wear and the digital outputs are bounceless
patterns are created using preci5len'digital
encoder is as good as the code wheel. The code wheer
system or a laser, each guided by closed-loop precision
vision
plotters and cut using either a punching
systems.
The light source used for optical encoders is usually a

point soulce light emitting diode (LED)' Most
the right is colrimated into parallel light rays
optical encoders are the transmissive wpe, meaning that
pattern is detected using
that pass through the disk (or strip) pattern. The image of the
a
transistor logic) digital quadrature outputs'

photodetector sensor and converted to TTL (transistor to
There are also reflective type encoders. These bounce
collimated light off a patterned reflective code
one side of the code wheel makes it a
wheel. Fitting all of the electronics of a reflective encoder onto
more comPact design than transmissive types'
encoder, using photodiodes to produce a
Figure 3-16 and Figure 3-17 illustrate the components in an
both distance and direction' In this
quadrature encod.er outpu! which enables this encoder to display
between pulses.
case, direction is determined via the phase difference
disk, LED (light emitting diode)' and a
The key components of an incremental encoder are a glass
which are equally spaced to deflect
photo detector. The transparent disk contains oPaque sections
An optical encoder uses a light
light, while the transparent sections allow light to be passed through.
of the disk' The light that shines
emitting diod,e, whicir shines light through the transparent portions
signal output'
througliis received by the photo detector, which produces an electrical
Electronics
Figure 3-15 Optical encoder sil6uif(rs) Figure 3-1,7 Optical encoder components(1e)
63
:
r (3): Instrumentation Utq4 ]. fQIP
Most incremental encoders have a second set of

pulses, being offset (out of phase) from the first Channel A
set of pulses, and a single pulse that indicates
each time the encoder wheel has made one
complete revolution. If A pulse occurs before B Channel B
pulse, the shaft is turning clockwise, and if B
pulse occurs before pulse A, the shaft is turning
counter clockwise. C pulse occurs once Per Channel C
revolution. Figure 3-18 illustrates the pulse
pattem of a quadrature encoder that provides Figure 3-18 Optical encoder - quadrature pulse
the direction information (with channel C being Patterns
the'reference' pulse).
Automated scanning systems incorporating optical encoders require calibration. This involves moving
the scanner over a specific distance and counting the number of pulses. Then a calibration factor is
used (number of pulses per millimetre).
3.4,2.2 MAGNETIC ENCODERS

A magnetic encoder uses the same principle to determine a position as an optical encoder, but does it
using magnetic fields rather than light. With a magnetic encoder, a large magnetised wheel spins over
a plate of magneto-resistive sensors. ]ust as the disk spins over the mask to let light through in the
predictable patterns of an optical encoder, the wheel causes predictable responses in the sensor, based
on the strength of the magnetic field.
The magnetic response is fed through a signal conditioning electrical circuit. The number of
magnetised pole pairs on the wheel, the number of sensors, and the type of electrical circuit all work
together to determine the resolution of the magnetic encoder.
The key advantage to using magnetism, as the element to produce a signal, is that it is unaffected by
very demanding environments - including dust, moisture, extreme temperatures and shock. Figure
3-19 and Figure 3-20 illustrate the components and assembly of a magnetic encoder.
Figure 3-19 Magnetic encoder components(20) Figure 3-20 Magnetic encoder(21)
64
3.5 TOFD EQUiPMENT PACKACES

In previous chapters, we described general aspects of TOFD instrumentation and some of the
associated software features. This included background information on encoders and digitisation.
Scanning apparatus required for TOFD can be as minimal as manual UT, in that the probes are
connected to the ultrasonic instrument and they are secured with a fixed probe separation'(probe
holder).
Reduced to the bare essentials, TOFD equipment
may be limited to:
r 1 Probe pair
. Probe holder (to maintain probe spacing)
. Pulser-receiver(basic)
o Data acquisition system (computer with data
display)
Figure 3-21 illustrates the basics using a laptop
computer as the acquisition/control computer, a
single T/R pulser-receiver and a probe pair and
basic holder to keep the distance between them
constant.
Figure 3-21 Basic TOFD equipment{zz)
Although it may be adequate to collect data for calibration purposes using a timed data collection,
most weld and corrosion inspection work using TOFD now requires positional information. This
invariably means that some form of positional encoder is used. When probe movement is made over a
fixed object, several options exist for mechanisation. The probes can simply be moved by hand or
alternatively, movement may be facilitated bv motors on the framework.
More extensive systems would include:
. I or more Probe pairs
. Probe holder (to maintain probe spacing)
r Pulser-receiver(s)
o Pre-amp system
. Data acquisition system (computer with data display)
. Encoder(s)
. Irrigation system for couplant
. DSP data analysis software
An example of a TOFD system with more features is illustrated in Figure 3-22.
65
Instrumentation Used In TOFD
Figure 3-22Enhanced TOFD system performing TOFD inspection during TIG weldingt2s)
Figure Z-22 is a view of the probe manipulator attached to a TIG welding machine. The probes are
housed in perspex shoes either side of the we1d. Water cooling is applied to the probes via the clear
plastic tubing. -ouplant is pumped to the shoe via the blue tubing and is then removed via the red
tubing. The wheel encoder to track the probes can be seen in the background. The B-scan image is an
example of the TOFD scan, acquired during trials, after undergoing processing with the University of
Surrey software. The scan is of a metre length of TIG weld, at approximately 75% completion. The
weld contains several defects which have automatically been recognised by the software (labelled in
red).
A typically well-equipped TOFD kit is illustrated in Figure 3-23. The components would normally be
used with a stand-alone TOFD instrument, equipped with a data acquisition computer and display
monitor (built in or separate). The wedges included in the set are designed with irrigation ports so
couplant can be pumped under each probe. This would require further couplant tubing and a source
of liquid couplant. An effective option, to pressurise the couplant to flow, is a manual PumP such as
those used for garden sprayers.
66
Cha r (3): Instrumentation Used In TOFD
il
f!,*
:3:1*
,*.r+
!,r'l
Figure 3-23 TOFD UtilitY kitt2rr
Included in the kit seen in Figure 3-23:

o Scanner with 150-mm frame
o Two 10-MHz (3-mm element diameter)
and 70' operation
o Three sets of wedges with carbide wear pins and irrigation holes for 45",60o,
a Two transducer cables 2.5m long
a Two cable connector adaPters
a spring-loaded, friction-driven, and water proof Mini-wheel encoder
a Brackets for flexible encoder positioning
a Additional 250-mm frame for wider probe centre separation
a 30dB pre-amplifier and cable
between the probe
Note that the pre-amplifier is most often used where there are long lengths of cable
the effects of noise
and receiver. A preamplifier is often placed close to the receiver probe to reduce
and electrical interference. It is used to boost the signal strength along the cable to the instrument
without significantly degrading the signal-to-noise ratio'
67

The Principles: History

Uploaded by

Copyright:

Available Formats

The Principles: History

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Principles: History

Uploaded by

Copyright:

Available Formats

Chapter (2): The Principles Of TOFD

CHAPTER (2): THE PRINCIPLES OF TOFD

2.1 A BRIEF HISTORY OF TOFD

Amplitude drop methods of sizing relate

Figure 2-1 Beam Boundary Sizing basis

Figure 2-1 shows two views of an angle

Figure 2-4 Ermolov equivalent target curves

oet*l lhroJgh 6tl s'E clclrrn,hed Oy 49 de$rucl.va

IOFD PE PE X-raY Gamma Slanual

Figure 2-6 NIL/KINT thin plate project results (8)

2.2 GENERAL PRINCIPLES

ir'..' .- .,;i ,. ......,'-r!,,..-.!: :i:i. j.r.r!..,j-,,.E

(r) DDT PlaE , (a, DDt Plre I

TOf D I'E PE X-ray Gamma Vlanual

Figure 2-5 NIL/KINT thin plate project results (8)

2.2 GENERAL PRINCIPLES

'a. . .:4. .+.-

Figure 2-7 Dilfrachon effects i n a ripple tank simulaLion

Figure 2-8 Photoelastic image of a compression mode pulse diffracting at a flaw

Figure 2-9 Tip-echo sizing principles for backscatter sizing

2.3 DIFFRACTION AS IT APPLIES TO TOFD

This results in a characteristic signal pattem

Figure 2-10 Basic TOFD setup

Several aspects of Figure 2-10 must be noted, in order to recognise some of

as the Iateral wave.

leaving the wedge with 65'L-mode

In step 2, see Figure 2-12, tl:re 3rr.rr.

Figure 2-13 TOFD process step 3

Figure 2-T4TOFD process step 4

2,4 DATA ViSUALiSATIONJ

Figure 2-15 A version of isometric A-scans

Figure 2-18 TOFD scan tyPes (11)

2.5 ADVANTAGES AND LIMITATIONS OF TOFD

2.5.1.,3 DATA STORAGE

2,5,1,5 VOLUME COVERAGE

2,5,1..8 FLAW OR/ENTATION

2.5.1-.9 COUPLING STATUS

2.5.1.10 REDUCED OPERATOR RELIANCE

2.5.2.3 FLAW CLASSIFICATION LIMITATION

CHAPTER (3): INSTRUMENTATION USED IN TOFD

3.1 BACKGROUND OF THE ELECTRONICS

In the early stages, industrial ultrasonic

Primarily, it is the control section of the

3.2 PULSERS AND RECEIVERS

transducer is an alternating voltage. Howevet,

3.2.1..1 SPIKE PULSERS

Oscillating pulses are generally formed by a Thyristor

transducer causes it to vibrate. The purpose of the supply

3,2,1..2 TONE BURST PULSERS

#v/ ifu il]l ifu

+ +r 1 Bi-directional square wave

-[rfl,.l] ulrlt*Li UIU LIJiI illl

tLrlLILtL t],l'Ul L-t]-Il-l- l Jilt tllt tLit

Change shape Change frequency Change ring time and

Figure 3-5 Tone burst pulser waveforms

3.2.1.3 SQUARE WAVE PULSERS

shape is shown on the left. In the first

Signal amplitude is relatively low,