Chapter (6): TOFD Technique Development

CHAPTER (5): TOFD TECHNIQUE DEVELOPMENT

The principles of TOFD chapter described the basic layout of probes which are arranged on either side
of a weld (the same principles can be used to carry out a form of corrosion tests where there rn-ay not
be any weld to use as a reference). The image in Figure 2-10 shows the basic setup. However, aspects
of the diffraction process along with the limitations caused by ring-time, the electronic equipment and
the part tested need to be considered in greater depth when developing a TOFD technique for a
specific application.

In this chapter, we consider some of the essential parameters when configuring the setup. ln addition,
we also consider variations on probe layouts.

6.1. PROBE SELECTION AND PLACEMENT

TOFD is usually applied to inspect the entire volume of a weld. In many cases, this can be done with a
single pair of probes. The TOFD operator needs to then position the TOFD probe pair(s) in a way that
ensures the area of concern is suitably ensonified. This requires consideration of several factors, some
of which are interrelated. Design of the TOFD technique is a compromise of all the parameters
considered.

Factors considered include (but are not limited to):

o Material tested
o Geometry of component
. Instrumentation available (single or multi-channel)
. Probes (size, frequency, angle (s), number of pairs)
o Detectionrequirements
. Sizing requirements
o Inspection speed and accurary requirements
Applications involving complex geometries, such as T, K and Y joints, can be inspected using TOFD.
However, the calculations involved will often require dedicated algorithms to determine indication
positions relative to the test surface. These calculations are not standard and generally require
customised software. Such calculations and applications are outside the scope of this handbook. For
the purpose of this handbook only the more common configurations of basic butt welds will be
considered.

Initially probe selection is based on the material tested. TOFD considerations for frequency are
identical to pulse-echo. Coarse-grained materials and very thick materials will require lower
frequencies to overcome the attenuation effect of scatter.

The probe size is then considered. It is also based, at least partially, on the material tested. To some
extent larger probe dimensions produce a higher intensity pulse over a greater distance. But the need
for divergence is better addressed using smaller probe dimensions.

99
Chapter (6): TOFD Technique Development

The probe refracting angle is selected based on the geometry of the component tested. Very thick
sections will require small refracted angles to ensure the back-wall can be detected. When the weld
cap is not removed, it will present a restriction that may require a higher angle of refraction to ensure
that the near surface is adequately addressed. Because of the large divergence in TOFD probes, the
actual angles used may be off by as much as 5 degrees from the nominal and no significant
deterioration of the technique will result.
In some cases, the thickness of the component tested is sufficiently large, so that no single probe pair
can be expected to cover the entire thickness. Guidance on all of these items can be found in the
several Codes and Standards now available for TOFD. A typical table found in the standards is
reproduced here as Table 6-1. In this table we indicate the minimum number of zones that must be
used to address the thickness of welds tested by a TOFD setup.

Table 5-l Recommended TOFD setu t on wall thickness

516 ol t;
orffora<45o

819 of t;
or f for rx <45"
1.112 of t
5112 of t

11,11.2 of t;
orffora<45o
In Table 6-L,lhe column identified as "Number of TOFD Setups", refers to the number of TOFD probe
pairs used

Having selected the appropriate probe pair parameters for the application, the operator then needs to
position the probes to provide appropriate volume coverage. The effects of energy redistributiory
upon diffractioru should also be taken into consideration.
The tabulated recommendations, for PCS values, indicate that the beam crossing points are generally
set to two thirds through the test piece (for a single probe pair TOFD setup). For many applications
lhe 2/s guide is adequate; however, it is useful to confirm that the coverage on the far-wall will be
adequate. \A/hen the TOFD inspection is carried out on a single V weld, the root area on the far-wall

100
 r (6): TOFD Tech

pair' However, for a double V weld' the width of the

can be adequately addressed by a single probe
with a single probe pair'
weld cap on the far surface may not be adequately covered
pulse-echo beam divergence
As a guide to volume coverage, we can initially use the 24dB arrangement be used to
accepted that a symmetrical
calculations. When designing a TOFD setup, it is
L,,e probe is transmitting' Since the probes used
indicate the beams. rruJ in Jpite of the faci that only
and angle, the reciprocity concept of transmitter
are to be "matched" with respect to frequency, size
and receiver can be assumed (i.e., probes b" interchanged with no apparent effect on the signal)'
"rn
a simple experiment can be set up' Using a
To verify the beam divergence and associated detectiory
small slit, 1.5mm to 2mm high in a 25mm plate
(or a small siae-arinea hole in a thicker section)' the
TOFD pair is placed symmetrically with respect to the
slit on the opposite side' By moving the pair
fixed) until the slit is no longer detected' the exit
perpendicular to the ,ht 1i."., a parallel ,"ull *ith PCS
dropped detection points can be compared' In
point of the transmitter positions for the peaked and
d.ivergences, at the angles corresponding to the 20-
most cases, the offset wlli be close to the calculated
for a ToFD probe' the standard formula
24dB dropfor that probe. when calculating the divergences,
material' The standard formula
is used; with the medium used for calculations being the wedge
adequate provision is made' for the angled
provides the half-angle of divergence. In order that
the probe element centre at the half angles' until
incidence from the *""dg", the rays are drawn from
theinterfacewiththetestmaterials(e.g.,steeloraluminium).TheruSnell,sLawisapplied.Allthree
of beam and the centre ray' These are illustrated
rays are usually used, i.e., the front of beam, the back
by the shaded regions and the centre of
in FigUre 6-1, where the front and back of beams are indicated
beam is indicated bY the red lines'

detecl small no
Figure getrerated usinli ESB€.l$'fool

Figure 5-1 TOFD far surface detection verification

indicate the limits of detection from the

In the illustrated case, an offset of approximately 7mm would
for nominal 60 degree refraction'
root centre for the setup of a 5mm diameter TMHzprobe,
for setting the probe separations' The
In TOFD, the exit points of the wedges are used as the reference
probe Centre separation is abbreviated PCS and used as the parameter of measure for
probe

placement. As noied, although only one of the probes

in the pair is used in transmit mode' the centre
for both the transmitter and receiver
ray that corresponds to the nominal refracting angle is drawn
rays' This is used to determine the Beam
probes. The crossing point is set based on these centre

101
 Chapter (6): TOFD Technique Development

Intersectiory referred to in Table 6-1. In Figure 6-1, the front of beam edge is not actually seen, as it
glances parallel to the test surface. From this illustration, it can be expected that a lateral wave will be
detected. For a probe with a smaller divergence (such as would occur for a 10MHz probe) the front of
beam could be assessed by the detection of a side-drilled hole placed near the test surface. This is
indicated in Figure 6-2. The lack of response from the side-drilled hole, when the probes are centred
over the weld centreline, would indicate that the probe setup would not be suitable for flaw detections
near the surface.

EiusB.ddnqEhTd tr

Figure 6-2 TOFD volume detection verification

Beam assessments, to establish coverage, are one of the functions of the calibration blocks and
machined targets described above (in addition to setting sensitivity).

Relying on drawings of divergence boundaries to verify coverage is not a guarantee of flaw detection.
Further consideration must be made for the pressure variation with respect to angle. Optimum
diffraction energy is obtained when the included angle at the centrelines is approximately 120 degrees.
This is based on diffraction theory for a vertical slit diffractor. EN 583-6 suggests that a working region
of -35 degrees to +45 degrees from this value may still provide useful signals. This produces an
incident angle range of about 8o to 38o with the vertical slit. However, at the 38 degree incidence,
nearly all compression energy is lost, as it coincides with the critical angle. Therefore, examinations
that rely on angles in this range (35-40 degrees) should be avoided.

In addition to the pressure drop associated with

the transmitted beam, there is a further
consideration for the beam pressure distribution.
Figure 6-3 illustrates a pulse emitted from a
probe striking a vertically oriented crack. This
figure illustrates how a specular reflection occurs
for the bulk of the beam. As it radiates off the
crack, it has a directivity that includes lobes. The
figure also illustrates the rings representing the
diffraction off the crack tips. These diffraction
Figure 6-3 Reflection & diffraction directivity tzs)
rings are also subject to directivity effects.

702
Chapter (6): TOFD Technique Development

A similar directivity occurs for the diffracted waves. Tronsm itter

Figure 6-4 is reproduced from Charlesworth and Temple
to illustrate the effect of incident angle on the relative
amplitude of the diffracted signals in a TOFD setup.
Incident angle theta is plotted as the horizontal axis and
corresponds to the refracted angle emitted and received by
the TOFD probes. From the curve, we see that a refracted
5i gnqt
anglg of about 65o incident on the upper and lower tips of
a vertical planar flaw, will provide the maximum signal
amplitude (pressure) at the receiver. The plot also
o
indicates that, for the upper tip of a planar flaw, the signal !

amplitude drops to a minimum of about 13dB lower for a )

!
E
0o incidence (this would be equivalent to a normal beam
placed directly over the upper tip). For increasing incident
=a
E
6 lt I
co
f
0
I
angles over 65", the amplitude will eventually drop to 6
o I a
zero, at about an 85" incident angle. The amplitude E I
response from the lower tip is more complicated. At about
a 35' incident angle, the flaw face at the lower tip will
provide the conditions for the first critical angle; and the
signal amplitude drops to zero. Upon further decreasing
the incident angle, some rise in the lower tip amplitude I Ocgrc.g
can be seerL to a maximum at about 18'. Further decreases
in incident angle result in amplitude decreases to zero, at Figure 5-4 TOFD Tip signal amplitudes
an incidence angle of zero degrees. vs. Angle of Incidence o)

The dashed line in Figure 6-4 indicates the resolution, theoretically achieved for a 5MHz probe on
steel. The fact that amplitude drops off, for a particular range of angles within TOFD beam coverage,
is further evidence of the need for multiple TOFD zones (setups) as component thickness increases.

The compromise requirements for TOFD configurations should now start to be apparent. In order to
obtain optimum probe separation for maximum volume ensonificatiory a wide separation is used.
Optimum probe separatiory for resolving a separation between the upper and lower tips of a flaw, is
when the total distance travelled is a maximum and occurs with a minimum probe separation. \rVhen
considering the optimum diffraction pressure for detection of diffracted signals from flaws, a PCS,
that provides a120" included angle of the probe beam axes at the flaw tip, is required.

The recommendations in Table 6-1 indicate the beam intersections occurring at 213 the thickness of the
zones. This results in an included angle of nearly 120 degrees. Because of the compromises that are
required to obtain suitable coverage, resolution and detection amplitudes, the exact angle and PCS in
Table 6-1 should be considered as guidance and the optimum combination of parameters is to be
established based on the specifics of the application.

As noted previously, the precise angle used is not normally critical and deviation of 5 degrees is
usually tolerable.

103
 Chapter (6): TOFD Technique Development

Under some conditions, where Table 6-1 indicates that one probe setup may be adequate to fulfil
inspection requirements, the operator may still decide that more than one probe pair and more than
one PCS may be required, in order to obtain improved coverage and resolution. Depending on the
electronics and probe holders available, this may require more than one scan.

6.2 BASIC PROBE ARRANGEMENTS

When a simple plate is welded, the probe parameters can be selected to provide the required volume
coverage in a single pass non-parallel scan. With guidance from Table 6-1 and consideration for the
weld cap in a single V weld, typical setups can be modelled. Beam spread is considered essential in
the single zone, so as to ensure adequate near surface coverage in proximity to the lateral wave. As
weld thickness increases, the beam spread and suitable refracted angles are required to provide both
the required volume coverage and the incident angles at the region of interesf ensuring adequate
amplitude responses from upper and lower tip echoes.

6.2.1 SINGLE TOFD ZONE

3
Frequency 10
Angle 70
PCS 71.

Crossing depth 13mm

Max. angle 90"
Min. ansle 50'
tiSUrr ren.rutcd u!inq tiBcrml ool

Figure 6-5 Single V 20mm wall technique

6.2.2 TWO TOFD ZONE

Figure 6-5 Single V 50mm wall technique

104
Chapter (6): TOFD Technique Development

Diameter Freouencv Angle PCS Crossine depth Max- ansle in zone Min. ansle in zone
J 5 70 90 16mm 90' 60'
6 5 60 138 40mm 70' 44"

6.2.3 THREE TOFD ZONE

IiEiure generated (rsing ESBeamTool

Figure 6-7 Double V 100mm wall technique

Diameter Freorrencv Anole PCS Crossins depth Max. angle in zone Min. angle in zone
J 5 90 16mm 90' 60"
6 5 60 773 5Omm 730 51'
6 5 45 200 100 570

Note: It was decided that the zone spacing would not be made equal for all three vertical extents.
Instead, a near surface limit of 25mm depth was designed. This left a thickness of 75mm to address.
\zVhen divided in half, each of the lower zones is 37.5mm high. The 2/3 crossing point in the zone from
25mm to 62.5mm is about 50mm from the surface. With the 6mm diameter SMHz probe selected for
the lowermost zone, the beam spread is getting close to the limit needed for coverage of the far
surface. Guidance from Table 6-1 indicates that we can direct the crossing point at the full depth when
using a 45o refracted angle. The curves, in Figure 6-4, indicate that this beam angle will suffer from
weak lower tip signals if detected below the centre of beam. However, the sensitivity to the upper tip
diffractions should still be adequate to provide flaw detections. If flaws that are located in this region
require vertical extent sizing, it may require extra TOFD scans or the pulse-echo sizing technique
applied.

105
Chapter (6): TOFD Technique Development

6.3 ALTERNAT]VEPROBEARRANGEMENTS
For various reasons, the simple TOFD setup, with a matched probe on either side of the weld, may not
be possible or adequate in all cases. The following are some examples of other considerations when
setting up a TOFD inspection.

6,3,1, OFFSET SCANS

In the previous examples of TOFD setups, the probe placement was symmetrical at about the weld
centreline. In some cases, this results in areas not being included in the inspection volume, especially
on the far surface. A related problem occurs due to the locus of equal time and the far surface dead
zone. When a specified minimum flaw size on the far surface is to be detected, it may require that the
region of interest be close to the beam crossing centreline.
Figure 6-8 illustrates a 25mm double V weld with a single zone TOFD setup (3mm diameter 10MHz
probe). The PCS has the 70" beams crossing at 213 depth. But, the beam coverage is seen to be
inadequate to detect a relatively large 2mm deep toe-crack on the far surface. The image also
illustrates the calculated dead zone (0.9mm) and the locus of equal time. Any flaws in the weld cap
deeper than the 25mm plate thickness, less the far surface dead zone, would not be resolved from the
back-wall signal (i.e., signals originating in the cap, deeper than 24.1mm would likely not be seen).
Even relatively deep flaws (like the 2mm toe-crack) will be missed, when scanning with the
symmetrical setup in Figure 6-8, due to the locus of equal time. The pink arc at the bottom of the
image in Figure 6-8 indicates the start of the back-wall ringing. The toe crack occurs at a time later
than the back-wall arrival time, so it willbe masked by the ringing.

Locus of equal time

Figure gener.rterl using f:SBeamTool {2mm Toe crack

Figure 5-8 Double V 25mm wall with 2mm toe-crack on opposite wall, not detected

The solution to this problem is the offset scan. By offsetting the probe pairs, from the centreline of the
weld and making 2 scans (or using 2 pairs of the same probes), the toe crack will be detected on the
offset that places the crack above the locus of equal time, as illustrated in Figure 6-9.

1.06
Chapter (6): TOFD Technique Development

Figure 6-9 Offset scan with 2mm toe-crack on opposite wall detected

6.3.2 SA\{E-SlDE TOFD

In the description of the principles of TOFD, in section 2.3, the TOFD technique was defined as being
two probes arranged in opposition. There is a possible variation of that configuration called one-sided
TOFD. This is a tandem arrangement of the TOFD probes on the same side of the weld. Under certain
conditions this can have an advantage, in that access can be limited to a single side for some geometry.
Placing a pair of probes facing a weld, with some spacing between them, it is possible to collect back-
diffracted signals. A noticeable difference with one-sided TOFD is the lack of reference signals as with
the traditional TOFD. With both probes facing the same direction, there is no lateral wave and no
reflected back-wall. Refracted angles need not be identical and the display is not as straight forward to
interpret for depth assessment.

107
 Chapter (6): TOFD Technique Development

n
J-:EC**.-
1---:H-' ..:-=€EffiE'
- ,a:i--:::*E*}=i*
.l :::,-'=$#ffii
1- ^". -
-,.;,ffiF*&:-:= ,1
ljl::ffi"--^*
li
-. 5=*-
."A- "
- -r-E*€*
:;{
1 I

"1_5- *Gfl:-: * '.-

*-.ffi
l=
f,rri
t'; ?il 1v
I

tr

"t,.l5lj#S.tl:: *r.a-!'f,- ,..,,..,,.,,'..,.,r1 .;,S,::i": , t,;l;:i;*r.- -;+

'....

t
__/
I

Figure 5-10 Same-side TOFD layout and responses

Figure 6-10 illustrates the setup and the sort of B-scan display that results from the same-side TOFD
configuration. Note that the root and cap signals are not possible to reference with respect to the
lateral wave and back-wall signals. A mathematically calculated top and bottom would be required to
assess the true depths of the indications.

6.3.3 SPECIAL PROBES FOR NEAR SURFACE FLAWS

The presence of the lateral wave ring-time has long been considered an impediment to TOFD. Lateral
wave removal is available on most systems, but not always used by the technician. Therefore, when a
scan is made and no abrupt changes are seen in the lateral wave, the assumption is made that no near
surface flaws exist. In some cases, this assumption is not adequate for the codes. Extra scans are then
required to investigate the upper region of the component.
Three options are generally considered feasible for ultrasonic tests; high angle shear wave with a dual
element probe, a full skip pulse-echo scan or a high angle compression wave in pulse-echo mode. A
simple pulse-echo scarg limited to monitoring the upper surface of a weld, is useful if it is feasible to
restrict the region of interest to just a few millimetres on either side of the toe of the weld. This option
becomes less attractive for very thick sections. For example, when a weld is made in a plate 100mm
thick, the sound path for even a 45" beam is 280mm to the top surface. The resulting beam spread at
that distance will make the detection relatively imprecise. An effective option in some cases is to use a
high angle dual element probe designed for use in transverse mode. A popular option to the dual
element probe is to use a high angle compression mode probe in pulse-echo. In some cases, users have

108
Chapter (6): TOFD Technique Development

even configured the hardware to fire the TOFD probes, first in transmit-receive mode for the TOFD
data, and then in pulse-echo mode to generate the so-called creeping wave (which is nothing more
than the compression mode glancing along the near surface).

6.3.4 PA TOFD
With adequate electronics built into the system, phased array probes can also be operated in a TOFD
mode.

The probe arrangement is identical to the single element configurations, with a probe either side of the
weld. An appropriate number of elements are then selected to pulse and receive in each probe. The
number of elements and the refracted angles are selectable from the software menu in the phased
array system. The resultant signals are no different from those seen using single element probes with
similar parameters. Figure 6-11 illustrates a phased array setup for TOFD, and Figure 6-12 illustrates
the B-scan collected.

[ig{rc g.n.nred cnE asB.i6!Iea

Figure 6-L1 PA TOFD layout with TMHz probe and 4 elements at 0.6mm pitch

Figure 5-12 Sample B-scan from PA TOFD setup (not noticeably different from mono-element)

Where multiple zones are used, there can be a limitation for phased array TOFD. A single phased
probe pair may be able to configure two zones, but it will be limited to the single frequency of the
probe. Applications with phased array systems often take advantage of the combined pulse-echo and

109
TOFD potential of the phased array, but may augment the phased array TOFD with a mono-element
TOFD, using a dedicated small diameter high frequency probe pair for the uppermost TOFD zone.

6.3.5 IMMERSION AND GAP TESTING OPTIONS AND ISSUES

Replacing the refracting wedge, in a TOFD configuration with water, is perhaps not common, but it is
feasible. With water or similar liquid, as the refracting material, issues of intermittent coupling are
eliminated.
Immersion testing implies the test piece is immersed in the coupling liquid. The probe may be
partially or completely immersed, as well. The transducer uses the couplant as a delay line. Being
fluid, the angle and time in the refracting medium are infinitely adjustable.
The biggest advantage to coupling by immersion is uniformity. Contact coupling always has
fluctuations in proximity and amount of couplant under the shoe. Immersion testing does not have
this variable contributing to amplitude variation.

Incident angle is easily changed in immersion testing and is not limited to discrete angles, as is the
case for contact tests. Since no direct contact to the part is made, no wear occurs and contouring of the
probe is not necessary.

Probes used for immersion testing are relatively straight forward. Except for waterproofing of
electrical connections, construction of the basic unit is the same as for contact probes. Since no wedge
is used, the probe housings need not incorporate wedge adapters; and the elements can be mounted in
a housing with little or no protective face. In fact, PVDF (polymer) probes are constructed with gold
electrodes exposed to the water (providing very high frequency pulses).

Focussing of the beam is simplified in immersion techniques. Cast synthetic resins and moulded
ceramic or polymer elements allow straightforward focussing.

When the immersion fluid is restricted to a small volume in front of the probe, the immersion
technique is more accurately called gap testing. Devices are designed to provide a fixed gap between
the probe-face and test part. The gap is filled with water (or similar couplant). Couplant can
constantly be fed into the gap by a pump or the test may be configured to prevent water loss as the
part is moved past the probes. Plate and tubular products are often tested using these devices.
When the couplant is trapped in a small cavity and provides a fixed water path, the device is
sometimes termed a bubbler. Depending on surface conditions and thickness, the gap may be several
microns or several centimetres. Relative movement, between the probe and part coupled with surface
tensiory will ensure some water is always lost, so a reservoir must be drawn from. A single probe with
gravity fed water-flow can be used in manual scanning or the same principle can be employed for
automated systems with multiple probes. Several gap testing methods are illustrated in Figure 6-13.

110
Chapter (6): TOFD Technique Development

Tx

Figure 6-13 Gap testing options

Of the illustrated options for gap testing in Figure 6-1.3, the one indicating the probes with wedges in a
holder being fed with irrigation (lower left) is essentially the same as most contact setups where water
is pumped through irrigation channels in the wedges. When the gap is used, but is not sufficiently
large so as to avoid sound bouncing between the wedge and the test piece, a gap multiple occurs. This
appears as a double (or treble) lateral wave. Use of gap testing with wedges and just carbide wear tips
should avoid any gap, i.e., carbide tips should be adjusted to ensure the tips are flush with the wedge
surface. In these applications, the purpose of the carbide wear tip is to reduce wear, not to provide a
gap. The result of the gap when carbides are used is seen in Figure 6-14.
rrr, 6510 8$50 SsC 86gt S?8$ S?50 88{n
tttlltl

36-

38*

?iu
-

ia*

Figure 6-l4Lateral wave multiple due to excessive gap

11.1.
Chapter (6): TOFD Technique Development

6.3,6 TOFD ON CURVED SURFACES

TOFD inspections on longitudinal seams of cylindrical components are similar to any butt weld
inspection by TOFD, with the exception that the curvature changes the elevation of the probes. This
results in extending the ring-time effec! thereby limiting the detection capabilities of the inspection.
As well, because the exit points of the probes, on a curved surface, are not at the same level (elevation)
as the midpoint on the test surface, the calculations of depth require corrections.

Some of the considerations are illustrated in Figure 6-15. A 200mm diameter cylinder is modelled with
a 25mm wall thickness and double V weld. 45' TOFD probes are used and positio:red to cross at
approximately 66% through thickness. The PCS would normally be measured along the cylinder
surface. For the crossing depth required, the PCS is 53.75mm. If we compare the resulting angle at the
crossing point, it is not 90' (as would be the case for 45'beams on a flat ptate). Instead the angle made
to the tangent at the midpoint is 64.3", so the intersection angle at the 66% through thickness is very
close to the recommended value for optimum diffraction signal amplitude.

'1
- tigure generated using ESBeamTool

Figure 6-L5 TOFD on cylinder long seam outside surface

When the PCS on the curved surface is compared to the chord (the straight line between the exit
points), it can be seen that the direct path is shorter than the PCS (52mm compared to 53.75mm). As
well, in Figure 6-15, a straight line tangential to the ID and OD has been drawn. Compared to a flat
plate 25mm thick, it can be seen that the probe exit points are lower than the midpoint. Another
difference exists between the flat and curved situation; in that, to achieve the same crossing point at
66% thickness and 64.3" on a flat plate, the PCS would be much greater (about 68.6mm).

71.2
Chapter (6): TOFD Technique Development

64.3" L 64.3" L
PCS
64.57 mm

.3428mm-34.28mm I

_l
9o
60
qb
3!e
3

gEn$nled usins
Fitiure cEn$nled
Fi.iure using lSBerm] ool

Figure 6-16 TOFD on plate to achieve same crossing and angle as in Figure 6-15

Clearly, the geometric calculations used on flat plate cannot apply on curved surfaces. Calculated
values will require geometric calculations based on the circle and the chord made by the line
connecting the exit points and the arc portion between the exit points. We will consider the more
general case illustrated in Figure 6-17.

,'. :.

Figure 5-17 General case for TOFD calculations on curved surface

Normally there are several knovrn parameters to start with and some parameters that must be
selected. \vVhen dealing with a hollow cylinder, parameters known would be the outer diameter and
thickness.

The initial choice of beam crossing will be some percentage of the wall thickness (e.9., the two thirds
crossing point). An estimate of the PCS based on an approximation from a flat plate condition might

113
 Chapter (6): TOFD Technique Development

be used. But the PCS chosen will be measured along the surface so it represents the arc length on the
cylinder. This is indicated as length (s) in Figure 6-17.
The equation for the arc length along a circle is:

Arc length : zrn(*)

,36O', $.1)
Where;
a: Central angle of the arc in degrees
R: Radius of the arc
z: Constant (approx. 3.74759)
The value for the arc length (s) is assumed known, so the arc angle (a) can be established:

Arc angle :360(s)/2rR (6.2)

Next, the chord length(2l) is determined. This is derived from the equation:

t = ZRSd (;) (6.3)

Where;
a: Central angle of the arc in degrees
R: Radius of the arc

With the chord length identified, the next parameter to assess is the saggita. This is the term given to
the distance from the circle to the chord and is effectively the height of the arc. The equation for the
saggita (h) is:

h=R*\tRr-P (6.4)

The calculated values provide sufficient information to use trigonometry in establishing the remaining
parameters. Looking at Figure 6-77, the ends of the chord indicate a PCS condition identical to the flat
plate, with the intended depth of the crossing point being the wall thickness percentage less the
saggita height.

However, a further consideration must be made. When the probe is moved around the circle, its
nominal refracted angle, relative to top dead centre, increases. The effective refracted angle is the
nominal angle plus half the arc angle (a).
Using numeric values as an example, a 300mm diameter pipe, with a 30mm wall thickness, is to be
TOFD tested along the long seam. If this is a plate and the plan is to use a single TOFD zone with the
66% crossing of the beams, then the PCS is 70mm.

The 70mm arc length along the pipe surface represents a 26.7o arc on the 300mm diameter pipe. The
resultant chord length is 69.7mm. The chord offset from the circle (the arc height or saggita) is 4.1mm.
The included angle of the beams at the crossing point is about 134".

The chord end points are located at points that define the arc angle, 26.7". Adding half the arc angle to
the nominal refracted angle of 45o provides an effective refracted angle of 58.4..

774
Chapter (6): TOFD Technique Development

The extended near surface dead zone can be quite deep into the tested component. A model illustrates
the added effect of geometry. Figure 6-18 illustrates a 10MHz TOFD pair on a 300mm diameter pipe
with a 30mm wall. Three near surface notches, one mid-wall target and one far surface notch are used
to indicate the ring depths. All targets are 10mm long. The near surface notches are 5mm, 10mm and
15mm deep. The mid-wall and far surface targets are 3mm high.

The resultant B-scan is seen in Figure 6-18. A horizontal cursor is placed at 20.075ts, where the.,end of
the near surface ring-time is indicated. No evidence of the 5mm deep, near surface notch is seen. Yet,
had this been a flat plate, the predicted dead zone would be only about 5.5mm. The second notch is
indicated by a weak signal that is still within the ringing signal. By the time the scanning gets to the
15mm deep notch the signal appears to occur just at the end of the near surface ring-time.

Figure 5-18 Extended near side dead zone due to cylindrical shape

The phase inversion associated with a mid-wall flaw having the same LSmm depth is seen next to the
15mm deep surface flaw. The far surface notch is easily detected at the end of the scan. However, since
the curvature of the pipe puts the toes of the weld lower than the thickness of the pipe at the
centreline, the far surface dead zones, in areas of concerry make detection of flaws in these regions a
potential problem. The differences, between calculated and observed dead zones and the regions on
the far surface where curvature will limit detection, are indicated in Figure 6-19.

-<"-.---_
-\ .-)
pronoun""i'o-'"t <qon"" ouo to curvature
fiig!,re geilerJted !5ing FSBeamTool \
Figure 6-19 Calculated as compared to observed near side dead zone

115
Chapter (6): TOFD Technique Development

Similar calculations are performed for TOFD, when done from the inside surface of a hollow cylinder.
Unlike the OD access condition, when scanning from the ID, the lateral wave is not as extended.
However, as a result of a slightly longer path where the lateral wave is forced to follow the inner
surface curvature, the lateral wave is generally weaker than for the flat plate condition.

A model of the inside curvature accessis illustrated in Figure 6-20. As with the outside model, the
near surface notches are 5mm, 10mm and 15mm deep. Compared to the notch diffractions, the'lateral
wave amplitude is larger. This is the opposite of the condition for the outside access on the curved
surface.

!
\,;1.=:i-.;:'r."=+,:" l, .,:,,,.:,,..,.'. :,r-i -. I

Figure 6-20 TOFD scan from inside curvature

6.3.7 TOFD ON TAPERS AND T.SECTIONS

Performing TOFD, on surfaces where the probes are not at equal elevations, presents more difficulties
in obtaining useful signals. Taper and tee connections are two examples of configurations that could
result in the transmitter and receiver being at different elevations.
Large pressure vessels or some flange to pipe connections can have welds joining sections of unequal
thickness. Weld procedures usually require that a gradual transition be made in the parent metal to
avoid abrupt thickness changes across the weld. Typically a 4:1 ratio is used to taper from the thicker
section to the thinner.

When the taper is accessible only from the tapered surface, there are limitations similar to those
experienced in TOFD inspections for cylindrical long seams that are inspected from the inside
curvatures. This is illustrated in Figure 6-21.

1.16
Chapter (6): TOFD Technique Development

figure lienerated

Figure 6-21 TOFD scan from tapered surface

The taper transitions the section from 30mm, down to 20mm, wall thickness. At a 4:7 taper this makes
the slope 14". The probe angles selected are different from one another. On the right, the probe is
mounted on a 60" refracting wedge and on the left on a74" refracting wedge. The angle difference is
selected to provide a beam angle, at the crossing point, that is approximately equal on both sides. Due
to the cap restriction on the right side the crossing point is slightly off-centre in the weld. Using a
higher angle beam on the right side can improve the central position of the beam crossing, but the
wedge is an off-the-shelf item.

\zVhen access permits, it is more often preferred to test tapered surfaces from the surface, where both
probes are at the same elevatiory as in Figure 6-22. In the case where there is access from the same
elevation, the only difference from the flat plate configuration is slightly reduced detection on the
lower side of the weld, at the taper, on the far surface. Computation for assessing depths will be the
same as those for the flat plate conditions.

Figure 6-22TOFD scan from opposite tapered surface

T-sections might be considered an extreme case of the inside curvature access. Probes, placed opposite
each other, with one probe on the horizontal arm and one probe on the vertical arm can provide tip
diffracted signals as well.
One can imagine the angle of the taper increasing up to 90". This is the form of the geometry in a T-
section. The amount of sound pressure that moves around the sharp corner at the bend, is
significantly reduced compared to the flat plate.

777
Chapter (6): TOFD Technique Development

Figure 5-23 Immersion TOFD setup for T-section

Figure 6-23 illustrates an immersion setup for a TOFD inspection on a thick T-section. 67rnm diameter
5MHz probes are arranged to produce 70o refracted angles with impact points iust before the weld.
The red lines indicate the transverse mode that forms and the green lines indicate the intended
compression modes. Two flaws are modelled in the setup. One is Smm higtu 7mm long and
perpendicular to the lower plate surface near the edge of the weld. The other flaw is 5mm long, 3mm
high and arranged to be 5o off the parallel of the lower plate near the weld root. With sufficient gairy
the lateral wave and the two flaws can be detected, as indicated inFigure 6-24.

>.!; - .ln ^

-.1--;l -.r, ---

- .-;,-- ,
--
----l::= =:::::!:ii;-i=
"=;+=:-=-;r .: -- -]i=:s
--" .:r:.* -'
-: -;-
-l-.T8, .
::1. -€::j]- -
t_tt-t:i-::; ;li;ry r:
-+ail#- ;
-:-- - :-
:::--'

sgs::i:::::::gs
si:i:i:r:ii$*:!
:: :ir :::::14::lii
=i::::

Figure 6-248-scan of T-section

In Figure 6-24 the cursor is placed at the edge of the vertically oriented flaw. No back-wall signal can
be detected and the flaw indications have a strange double echo effect. The double echo derives from
the shear head wave that is initiated by the compression mode glancing along the surface. This shear
head wave arrives at the flaw slightly after the compression mode and its interaction with the flaw
provides a further mode conversion back to compression mode.
The purpose of this model is to illustrate that detections can be made using an opposing TOFD setup
on T-sections. Details of sizing and depth estimates are not simple trigonometric solutions and

118
 ter (6): TOFD Techni

therefore beyond the scope of this handbook. Further reading on T-section TOFD can be had in the
British NDT ]ournal, Insight tzel.

6.3.8 FOCUSSED TOFD

The ability of phased array systems to steer and focus beams allows the unique capability of
focussed
TOFD. This is a concept normally used after a general detection using unfocussed probes' The
as to more
purpose of focussing the beam in TOFD is to improve the resolution at different depths so
accurately size the vertical extent of a flaw. Focussed TOFD is not a common technique,
but it has
been discussed for some years. A paper by Brillon et a1(30) at the 17h World Conference on
NDT
described how a phased array system could be configured to divide a weld into vertical zones. With
of
the ability of the phased array unit to direct the beam at defined angles and from separate groups
elements, TOFD configurations with varying angles and varying PCS can be made, as seen in Figure
6-25.
['ulser

1 I

Fi gure 6-25 Multt- zone f ocussed TOFD(3o)

With beams focussed near the centreline, Brillon illustrated how flaws, in even a relatively thin 20mm
thick weld, can be sized better (and detected) using focussed beams. Figure 6-26 indicates the four
beams used with the associated B-scans to the right. Upper left is 40o, uPper right is 50". Lower left
is
60. and lower right is 70o. Upper tip identification of the two flaws (labelled 1 and 2 with arrows) are
seen to improve or deteriorate relative to the angle used'

1.19
Chapter (6): TOFD Technique Development

Figure 6-25 40",50o,60o and 70o TOFD (sectorial TOFD) (30)

6.4 MODELLING THE WELD INSPECTION _ SCAN PLANS

A requirement of nearly all national and international ultrasonic inspection Standards is the inclusion
of a description of the volume coverage of the test piece. This is variously called the "scan plan", the
scanning technique, written instructions or the procedure, depending on the terminology used in the
specific industrial venue. With single element, fixed angle beams in pulse-echo, the process was
relatively simple and often a'pencil sketch' was adequate to indicate the probe movement to achieve
coverage of the weld andHAZ.

Because developing a TOFD technique is often a result of making compromises between the many
variables, scan plans for TOFD often provide a means of visualising the optimisation process. With the
ability of software to adjust the parameters (as illustrated in Figure 6-5, Figure 6-6, Figtxe 6-7 and
others in this chapter), software technique development tools are becoming more popular and quickly
replacing the simple line drawings that were popular with the manual UT technique.
For basic (flat plate butt weld) TOFD configurations, the setups calculated by the technique
development software is generally adequate. However, in some cases the standard will also require
that the technique be validated or demonstrated on a block with defined targets.

6.5 SUMMARY OF TOFD PARAMETERS

Any inspection using the TOFD technique would generally have a requirement to produce a written
document detailing the parameters of the inspection for the application. Some codes/specifications are
sufficiently well detailed that they provide adequate information to address all the parameters
required by addressing that document.
Details of the requirements for the procedure and report are addressed in the standards. Flowever, at
a minimum, the technique will need to include all essential parameters. These are the physical settings

1.20
Chapter (6): TOFD Technique Development

that, if changed, would result in changing the test results. Essential parameters for TOFD would
include the following items:
. examination levels e.g., whether or not reference blocks are required
. specification of reference blocks, if required
. requirements for access and surface conditions and temperature
o acceptance criteria
. equipment details (manufacturer and model)
. probe size, frequency, nominal angle, bandwidth
. setup of probes (PCS)
. scanning mechanisms
. couPlant
. volume inspected
. scan increment
r sensitivity methods calibration block materials and targets
o gating (time window)
. filters, averaging pulse-duration, pulser voltage

727
Chapter (7): TOFD Sensitivity Settings

CHAPTER (7): TOFD SENSITIVITY SETTINGS

7.1, TOFDSENSITIVITYREQUIREMENTS
In the American Society of Mechanical Engineers Boiler and Pressure Vessel Code (ASME BPVC),
TOFD is referred to as one of the "non-distance amplitude techniques". ASME specifically recognises
TOFD as an ultrasonic technique that does not use a threshold based on signal amplitude.

Indeed, TOFD is not an amplitude-based ultrasonic inspection method. Instead, the small amplitude
displacement patterns seen on the B-scans are used to identify and characterise flaws. F{owever, this
still requires that sufficient gain be applied to ensure that the subtle variations in phase displacements
on the A-scans can be discerned from the background noise.

In addition to verifying the instrument parameters for general use, a separate "calibration" is done to
ensure both uniformity and a minimum level of sensitivity of an inspection. This is usually considered
a form of "sensitivity calibration.

Distance-Amplitude techniques use a reference target to set detection sensitivity. To this corrections
(extra gain) may be added to compensate for attenuation effects. Targets used as reference and
additional gains can be recorded and are usually set in Specifications, Standards or Codes.
Although TOFD does not have the same requirement for distance amplitude corrections, TOFD does
require some means of assuring a minimum sensitivity to ensure indications are seen over the
background noise level. In addition to the detection aspect of setting sensitivity, there is also the need
of a formula to provide repeatability of the sensitivity used in subsequent inspections. To this end
there are three methods by which TOFD may be configured for sensitivity calibration:

1) Reference signal amplitude

2) Background noise level
3) A set amount of gain over a repeatable target
Guidance on TOFD sensitivity can be found in several regulatory documents. ISO 10863 is a good
example of the options presented for sensitivity settings in TOFD.

"For all examinqtion leaels the sensitioity shull be set on the test object. The amplitude of the lateral usqoe shall
be between 40 'h and 80 % full screen heights (FSH). ln cases where the use of the lateral wqoe is not
appropriate (e.9., surface conditions, use of steep beam-angles), the sensitioity shall be set such that the
amplitude of the back-wall signal is between 18dB and 30dB aboae FSH.IMen the use of neither a lateral waae,
nor a bnck-wall signal is appropriate, sensitioity should be set such that the material grain noise is somewhere
o/o
between 5"/o and 10 FSH." Q1)
The British Standard BS 7706 identifies a fourth option; a representative flaw sample. Flowever, no
two flaws are identical; so other than considering such a flaw as "typical" of the sort being sought

723
 Chapter (7): TOFD Sensitivity Settings

there is little more than a coarse indication of the gain required for that particular flaw, when using
this method.

In fact the BS 7706 document identifies two specific targets and does not actually identify a
"reference" signal per se. The four B57706 methods are:
1) Representative flaw sample
2) Diffracted signals from slits
3) Reflected signals from cross-drilled holes
4) Grain noise
The target signals (slits and side-drilled holes) can be grouped together and considered as any
machined target.

Artificial reflectors (and diffractors) do not respond in the same way as real flaws. However, when
using TOFD, there is generally no direct correlation between the amplitude of the signal and the
severity of the flaw. Thus, these machined targets are merely a convenient method of providing an
easily reproducible target that allows others to produce similar sensitivities. The same can be said for
setting responses off targets for the pulse-echo methods. In addition to providing reproducible
sensitivities, machined targets also provide a demonstration of the distribution of the ultrasonic field
and a means of demonstrating inspection resolution. \zVhen configuring a scan there is usually a
requirement to record the sensitivity determination method and settings. Typically, when calibrating
sensitivity for TOFD, a screen capture or a recorded dynamic scan is made to augment this
requirement.

7.1.1. REFERENCE SIGNAL AMPLITUDE

By far, the most common method of sensitivity setting is the use of the lateral wave amplitude. Unless
an operator has the advantage of working with a logarithmic amplifier, setups must avoid
amplifications that result in saturating lateral wave signals. For any post processing of data using
lateral wave straightening, equalisation and subtraction, the lateral wave must not exceed 100% of the
digitisation capability of the instrument.
As noted above, guidance in the standards suggests setting the lateral wave at between 40% and 80%
of the screen height. This represents a variation of 6dB, which in pulse-echo testing would be an
unheard of tolerance. Tolerance of 6dB on any TOFD setup is not the intent of the standards. Instead,
the standard is considering the effect of beam spread and the selected refracting angle. Table 6-1
indicates a range of angles for different thicknesses. \zVhen angles near 70" are used, the beam spread
provides a large portion of the pressure as the glancing lateral wave. As the refracted angle decreases,
the off-axis pressure also decreases. Therefore, even though the sensitivity to subsurface targets or
grain structure may be equal for a given amplifier setting, the proportion of pressure at the surface,
moving as a lateral wave, is not equal at all angles. TOFD setups (on flat plate) using refracted angles
between about 65o to 75o would therefore provide relatively strong lateral wave signals that should be
set to about 80% screen height. Refracted angles in the range of approximately 55" to 65" will produce
relatively weaker lateral wave signals. These would be set to a lower value screen height, typically
around 40"/o.

1.24
Chapter (7): TOFD Sensitivity Settings

For thick sections, when using a refracted angle near 45o, the amount of sound pressure available at
the surface of a lateral wave is very small compared to the range from 60" and higher. At these angles,
trying to increase gain to achieve a lateral wave signal will result in excessive sensitivity on the beam
axis. Instead of using the lateral wave as the reference signal, it is the back-wall signal that is used
(assuming the flat plate geometry). As with the lateral wave guidance providing a range of
amplitudes, guidance on use of the back-wall when using a refracted beam near 45" suggests.setting
the back-wall signal to 100% screen height and then adding anywhere from 18dB to 30dB extra.

7.1.2 DIFFRACTED OR REFLECTED SIGNALS FROM MACHINED TARGETS

The common targets used in ultrasonic testing are slits, notches, side-drilled holes, thiough holes
(i.e.

through the thickness) or flat-bottom holes. Of these, the through holes and flat-bottom holes are not
generally used in TOFD as they are highly susceptible to probe skew.

Slits may be surface breaking or subsurface, but are generally considered planar (having length and
height). The width is, however, small compared. to the wavelength. Slit width should typically be less
than half a wavelength. With wavelengths in steels of 3mm to 0.6mm (for frequencies 2-10MHz), mill
cutting is not a practical option for fabrication of the targets for higher frequency techniques. As a
result, slits are normally made by electro-discharge machining (EDM). EDM slits are made with
copper electrodes in the shape of the desired target. Copper electrodes can be made with thicknesses
on the order of 200-300 microns (0.2-0.3mm), making the half wavelength recommendation reasonably
achievable.

Notches may provide a similar planar shape, with respect to length and heigh! but are generally
wider. Square or 'V'notches can be made by mill cutting if the depths of cut are not too deep.
In order that a true diffracted signal is obtained, the width dimension of the notch must be small
(again less than half a wavelength); otherwise the signal will be of a reflective nature and will be
relatively strong compared to the diffracted signals from the narrower "slits" or'V' notches.
As with pulse-echo ultrasonic testing, the blocks containing these targets are considered calibration or
reference blocks. These should be constructed within a close tolerance of the thickness of the
inspection task in hand and of a similar material.
Side-drilled holes provide a similarly useful target for setting sensitivity; however, as with notches,
the signal from the top of the target may more accurately be considered a reflection rather than
diffraction. In order to discern the bottom diffraction signal from a side-drilled hole, the diameter
should be at least 2 wavelengths.
Calibrating gain settings with slits, notches or side-drilled holes is accomplished by positioning the
probe pair symmetrically on either side of the targef such that a diffracted/reflected signal is obtained
from the notch tip.
One way of ensuring that the signal is a diffraction signal is to arrange the notch or slit to be surface
breaking on the same side as the probe pair. The diffracted signal is then set to peak-to-peak
amplitude, typically 80% un-rectified A-scan display, and the image and settings recorded.

125
Chaoter (7): TOFD Sensitivitv Settinss

Some of the target layouts that can be used are illustrated below. Not all of the guidance of the
standards is reasonable to achieve. EN 583-6G2) suggests one design with the side-drilled holes and
slits to prevent the upper tip ring signal. It recommends that the block be adequately long to permit
the necessary probe movement over the reference diffractors and 5 such targets, with the diameter of
the SDH (side-drilled hole) at least two wavelengths. Figure 7-1 illustrates the scale required for a
block 30mm thick.
If a nominal SMHz probe is used, the wavelength of the compression mode in steel is 1.2mm. 3mm
diameter SDHs are a common larget size and might be typical of that used in the reference block. EN
583-6 indicates placing the targets (assumed to mean the lower diffraction tip) at 1,0"/"';25"/,, 50"/",75"/"
and 90"/, depths. So as to avoid interference of the slit, when placing the probes centred over a target,
the slits and targets must be separated by at least the PCS pulse, the diameter of the probes. If all the
targets are made in a single block 30mm thick by 30mm wide, the block needs to be about 2Tlmrnlong
if the refracted angle used is 60". The block length is longer for thicker wall and for higher angle
probes (requiring greater PCS). The block can be made shorter by using a wider block (e.9., 60mm)
and drilling blind holes from either side of the block.

F,xun, i.n.d!,J (.ii{ E!3.irIxn

Figure 7-L Side-drilled holes with saw cuts to surface

The ASME Boiler and Pressure Vessel

Code (Section V Article 4), describes the
TOFD reference blocks required for
verifying sensitivity and zone coverage.
The two zone TOFD block is shown in
Figure 7-2. lnstructions require that the
collected data, at a minimum,
demonstrates that the upper target in the
lower zone is detected in the upper data
display and the lower target in the upper
one is detected in the lower zone data
6EI,]ERAL i]OlES:
display. ia) ir eirals ihe ihlal.?r! cf iin lpier :a.a
tbl It eoriir the thi[!:esi ol lhe lorler ;:c]e.

Figure 7-2 ASME section V, article 4 two zone TOFD

$|ss( (::)

126
Chapter (7): TOFD Sensitivity Settings

A useful design was seen in the older Euro norm, EN TS 14751. The design also made provision for
five targets, but they were a combination of side-drilled holes and a surface notch. The SDH at the
near surface was to be a smaller diameter (2mm), always within 3mm of the surface. Whereas the
lower SDHs were 3mm diameter, located at 25"/", 50% and 75% depths. To permit independent
detections, the SDHs were drilled to increasing depths from the end and the notch was positioned at
some distance away from the longest drilled SDH. The general layout is illustrated in Figure 7-3

Figure 7-3 Suggested TOFD reference block design from EN TS 14751G4)

The ISO 10863 document suggests anther reference block design to be used to assess the depth
resolution capability of a TOFD setup. This is illustrated in Figure 7-4. Note the small EDM notch is to
be 0.2mm wide and the surface breaking notches are to have pointed lower tips with an included
angle of 60'.

Figure 7-4 TOFD resolution reference block as per ISO 10863 (31)

727
 Chapter (7): TOFD Sensitivity Settings

Special applications with complex geometries generally combine sensitivity and detection verification
assessments. Such a design is illustrated in Figure 7-5 where the design was used to verify detection of
cracking in a turbine rotor shaft.

.:.- ,- -r,--:.- : r .

Figure 7-5 TOFD calibration blocks - custom application {:s)

Using the response from a target (either a reflected response from a notch or side-drilled hole)
provides some advantages when configuring a TOFD scan:
1) Repeatable inspection sensitivities used in different inspections can be achieved
2) A theoretical "estimate" may be extrapolated for detectability of a flaw of a size comparable to
the target assuring that a "minimum" detection level has been achieved
3) Volume coverage may be assessed by using targets at different depths and offsets
When using side-drilled holes, as the reference target, the maximised response set to 80% peak-to-
peak may not provide adequate scanning sensitivity. Thus, it may be necessary to add a fixed amount
of gain over that level (15-30dB could be required depending on the material and hole diameter).

7.1.3 SENSITIVII'Y GAIN SETTINC USING GRAIN NOISE

Not all test objects and geometric configurations lend themselves to simple use of calibration targets.
As well, there may be cases where there is no pre-judgement of allowable flaw size and it is required
to merely investigate if any flaws are present. To this end, the maximum detection level possible is
limited by the material grain structure. In fact, the object's grain structure becomes the limiting
sensitivity factor in all cases.
By using gain settings that are sensitive to acoustic scatter at grain boundaries, it is assumed that any
flarvs present would need to be significantly larger than the responses from the grain boundaries. For
fine grained materials this is a safe assumption. However, for austenitic materials the grain
boundaries may be of a similar order of magnitude as flaws deemed critical to service. Equally

728
Chapter (7): TOFD Sensitivity Settings

problematig the grain size may be approximately the same dimension as the wavelength, so
differentiating flaws and grains is not feasible.
Setting the gain level is accomplishedby a comparison of the A-scan time base before and after the
lateral wave.
Placing the probe pair on the test object, at the calculated probe separation, the operator displays the
lateral wave and compression back-wall reflection (if appropriate). The pulse duration is optimised to
provide the optimal pulse shape with maximum signal-to-noise ratio. Gain is then adjusted, by
observing the grass level after the lateral wave, so that acoustic grain-scatter is set to about 10% screen
height. Amplitude of the electronic noise, prior to the arrival of the lateral wave, should be at least 6dB
below the amplitude of the grain noise (i.e. 5%).
Note: this technique operates on the assumption that flaws of concern will provide greater amplitude
diffractions than the grain boundaries. This may not always be the case and care should be taken in
abnormally coarse structures, such as austenitic materials and electro-slag or heavy-wall submerged
arc welding. Figure 7-6 illustrates the contrasting B-scans that result when a fine grain and coarse-
grain structure weld are inspected. Typically, austenitic weld grain sizes can vary from about 30pm to
120prm. The recommended frequency for testing these welds is from 2MHz to 4MHz. Even with the
compromise of lower frequency for inspections, the TOFD display may still result in unacceptably
poor signal to noise ratios. Figure 7-6 indicates an acceptable level of grain noise on the left, with the
region just after the lateral wave as having a signal to noise ratio greater than 6dB; but the image on
the right contains grain structure noise comparable to the lateral wave amplitude (approximately 80%)
and a signal to noise ratio less than 2dB.

Figure 7-6 Grain structure seen in TOFD (36)

729
Chapter (7): TOFD Sensitivity Settings

7.2 QUALIFICATION AND DEMONSTRATION BLOCKS

TOFD reference blocks are typically used for assessing the sensitivity level or confirming and
demonstrating resolution of the combined probe frequency and PCS.
A separate function can be provided by blocks with targets. So-called demonstration blocks (or
qualification blocks) can be designed to mimic the conditions of the inspection. These would bb made
of the same material (identical) as the component tested and may contain machined or natural flaws
or a combination of both.
The function of the demonstration block is to essentially show that the technique being applied will
detect the flaws of concern (no technique can guarantee that all flaws wilt be detected all the time) and
provide the required coverage. When the demonstration block contains a weld, the restrictions
presented by the weld cap and root excess metal (if not removed) will be identical to the actual part
tested. This differs from the simple sensitivity reference blocks made from flat plates. \Alhere
restrictions limit the coverage on the far surface, or the resolution needed on the near surface, the
demonstration block can be used to develop auxiliary techniques to ensure detection of any flaws not
adequately detected by the initial TOFD setup.

Demonstrating the efficacy of an NDT technique has been part of standards requirements in the
nuclear industry for many years. With the introduction of the ASME Code Case 2235 (in the year
2000), ultrasonic testing has been accepted as a viable option to radiographic testing. However, in
order to qualify the setups used on ASME BPVC Section VIII pressure vessels/ any UT system must
not only be computerised, it must also demonstrate the capability of detection on a qualification block.
Requirements for the qualification of the test as outlined in ASME Code Case 2235G7) arel
"The procedure shall hsrse been demonstrated to perform acceptably on a qualification block(s). The qualification
block(s) shall be prepared by welding or hot isostatic pressing (HIP) and shall contain q minimum of three flaws,
oriented to simulate flaws parallel to the production weld's fusion line as follows:

(L) one surface flaw on the side of the block representing the oessel OD surface

(2) one surface flaw on the side of the block representing the aessel lD surface

(3) one subsurface flaw

@) lfthe block cqn be flipped during LIT examination, then one flaw may represent both the lD and OD
surfaces. Thus only two flaws may be required"

An example of a combined qualification and calibration block used to comply with ASME Code Case
2235 is seen in Figare 7-7.

130
 : TOFD Sensiti

The qualification block in Figute 7-7

contains more than the minimum
targets. It illustrates a 102mm thick
weld section for a clad vessel' Side-
drilled holes can be used for sensitivity
checks and near and far surface
notches are included in order to
establish the capability of the TOFD
inspection system to resolve the near
surface targets. Embedded flaws are
difficult to manufacture in a weld' One
method that has Proven useful is to
weld the comPonent to a Particular
level and then make an EDM notch at
the bevel edge. This simulates side
wall non-fusion. Two such targets
were placed in the design drawn in !lss( tsa)
FigxeT-7. Figure 7-7 Combined calibration and qualification

Effective use of the demonstration block is made

when used as a dynamic calibration tool' A frame
equipped with steel run-off sutfaces, fitted to the
neight of the demonstration block, can provide a
.ontin ous surface similar to the tested
component. By arranging for a short scan' that
*orr", the probes past all the targets, a record of
the responses can be saved to file.

Figure 7-8 illustrates a frame with thin steel plates

on either side of a calibration block' The block
contains surface and subsurface targets' The near
surface targets are seen as a V pattern of 5 slots'
The scanner moves the probes from one run-off Figure 7-8 Frame holder and dynamic
plate, over the calibration block and onto the calibration/demonstration blocks(16)
tpposite run-off plate. Surface slots are made to
inJicate the positions of the weld centreline, the
weld edge and the outer edge of the heat affected
zone. As part of the dynamic demonstration scan'
detection of the weld edge and HAZ slots indicates
whether offset scans are required or not'

131