International Ultrasonic Testing Online Conference (UT-Online 2021) Nov 2021 at [Link].

net/UTonline2021

Variations on Time-of-Flight Diffraction

More info about this article: [Link]

Ed Ginzel1
Oleg Volf 2
1
Materials Research Institute, Waterloo, Ontario, Canada
e-mail: eginzel@[Link]
2
MSc, PEng, ACCP III, Calgary, Alberta, Canada
e-mail: olegvolf@[Link];

Abstract
The term “Time-of-Flight Diffraction” (TOFD) has been around since the mid-1970s. However, earlier
applications of the concepts have been documented by Ermolov since 1968. Provided the technique makes a
timing measurement and relies primarily on diffracted wave energy, rather than specular reflection, there is no
reason why it should not be considered a TOFD technique. This paper looks at variations on the traditional TOFD
technique using probes located on the same side of the joint being inspected, rather than on opposite sides. Results
of the technique are illustrated for both steel and HDPE joints.

Keywords: Time-of-Flight Diffraction, TOFD, backscatter, forward scatter, ultrasonic

1. Introduction
The now-popular Time-of-Flight Diffraction (TOFD) technique is credited to the work of Silk
and Lidington in 1975 [1]. However, a similar approach had previously been discussed by
Ermolov [2,3,4] that he called техники пропускания (transmission techniques). Early
development of the technique focussed on detecting and sizing flaws in welds. It used two
probes straddling the weld, one operating as a transmitter and the other as a receiver. This
configuration was used to avoid the strong signals often associated with specular reflections
rather than diffraction. However, Charlesworth and Temple [5] note that, in their opinion,
provided a single probe technique makes a timing measurement and relies primarily on
diffracted wave energy, rather than specular reflection, there is no reason why it should not be
included as a TOFD technique. By this approach, it is not necessary to limit TOFD to just a 2-
probe technique relying on forward scatter detection of diffracted waves.

Setting aside the single probe approach, we can consider TOFD to involve the detection of
scattered diffracted energy regardless of the probe orientation. This is illustrated in Figure 1.

Creative Commons Attribution 4.0 International (CC BY 4.0) license

Page 1 of 10
 Rx
Rx
Rx
Rx

Tx

Figure 1 TOFD may have a skew of almost 0° to 180° between Tx and Rx

According to Bray [6], as long as the diffracted energy is received on the same surface as it was
transmitted, we could consider this a TOFD technique.

This paper describes variations on the “traditional” approach where the transmitter and receiver
are placed opposite each other. Placing the transmitter and receiver elements on the same side
of the joint being inspected is not new. Dual TRL probes had long been used to inspect coarse-
grained materials, but they were used to detect specular reflections.
Early in 2021, ASTM E3044 [7] underwent its 5-year review. Provision had already been made
in that standard for one-side TOFD. In the 2021 edition, an annex was added to better describe
an approach for one-side TOFD.

Although not common in North America and most of Europe, the concepts of one-side TOFD
(or same-side TOFD) are now well established in eastern Europe where they have been referred
to as the “quasi-tandem” technique [8,9,10]. By 2004 there was already a Russian national
standard devised for its use [11]. Provisions to direct a beam to a plane on a cylinder involves
some geometrical calculations. Unfortunately, the GOSSTSTANDARD provided no setup
information on the process.

A few years after the GOSSTSTANDARD was published a pair of Moldovan researchers [12]
dug into the optimisation process for beam placement in the chord technique. The
GOOSTSTANDARD considers pipe up to 31mm wall thickness. Isaenko describes how the
technique was OK for wall thickness under 10mm but “problematic” for over 10mm thickness.

In spite of numerous Russian papers describing the process, none of the papers identified has
actually illustrated the data collected and how it can be analysed. To remedy this lack of
evidence, the authors have collected scans using the principles described by Isaenko &
Kalyukin and demonstrate the advantages of one-side TOFD.

Page 2 of 10
2. Data Collection Methods
2.1 AUT Development

Early data results were collected using a technique added to a pipeline inspection system
designed by Eclipse Scientific in 2013. One-side TOFD data was added to the display of the
TrueView system in order to improve detection and sizing of flaws near the outer surface of
girth welds. Experiments were carried out using tandem, side-by-side and even delta TOFD
configurations (see Figure 2).

Figure 2 TrueView Experimental Tandem and Back-scatter TOFD wedges

Because these were used in conjunction with zonal pulse-echo phased-array data acquisition, it
was ideal to compare detection results. Figure 3 illustrates detection of the 3mm diameter flat
bottom hole (FBH) calibration targets in a 24-inch diameter pipe with 24mm wall thickness.
Note the absence of the lateral wave allowing detection of the Fill 6 FBH with a 1.5mm ligament
to the outer surface.

F6
F5
F4
F3
F2
F1
HP

Figure 3 AUT Data Display with strip chart pulse-echo, volumetric pulse-echo and
backscatter TOFD data displayed (arrows indicate responses from zonal targets)

2.2 HDPE Butt Fusion Application

In the development of an annex for ASTM E3044, work was carried out to optimise probe
placement using the geometric concepts described by Isaenko [12].

Page 3 of 10
In Isaenko’s paper, they use rectangular shaped elements and separation measurements are
made to the closest corners of the elements. The authors have used the centre of beam rays in
the test material as assumed from either circular or rectangular elements. These rays are then
projected back to the elements using Snell’s Law to determine the appropriate incident angles.

To facilitate a scan of an HDPE butt fusion joint, a holder was configured for the probes that
provided water coupling and the calculated separation, skew and incident angle of the beam.
The calibration targets used in a previous project [13] were scanned to demonstrate the detection
capabilities. The holder design, targets and scan data are seen in Figure 4. Both the inside and
outside surface notches in Figure 4 are machined 1mm below the nominal surface of the pipe.
The uppermost FBH has a ligament to the outside surface of only 1mm.

a
b

Figure 4 HDPE butt-joint calibration scan a) 2mm diameter FBHs in 6-inch NPS
pipe, b) probe holder design c) B-scan display

The full circumference of a pipe containing a known contamination flaw was scanned from
both sides of the joint. In addition to the known flaw, a significant a section of non-fusion was
detected approximately 90mm long.

Figure 5 illustrates scans made from each side of a 6-inch NPS butt fusion joint.

Page 4 of 10
 Non-fusion Mid-wall
contamination

Non-fusion Mid-wall
contamination

Figure 5 HDPE butt-joint, upper from left side, lower from right side

Different characteristics of the flaws can be seen when approached from opposite sides. This
suggests that scans from both sides of a joint should be carried out if possible.

2.3 Steel Plate

A typical flawed weld sample used for training was used to demonstrate the same-side TOFD
technique on a steel weld. This plate was 16mm thick and contained a single V weld with three
embedded flaws.

Unlike HDPE butt-fusion joints where most flaws occur within 2-3mm either side of the joint
centreline, in a steel V weld there can be significant separation from one bevel surface to the
other. Therefore, although it may be possible to scan from just one side of the weld, there will
be uncertainty as to the depth of the indication. Figure 6 shows scan results from the Left side
(top) and Right side (lower) of the same weld.

The 3 flaws are as follows;

Flaw Type Length Height Offset from Distance Depth
(mm) (mm) Centreline from edge
(mm)
1 Lack of fusion 13 4 1.5 86 12
2 Root Crack 18 5 -1.5 173 11
3 Incomplete 23 4 0 226 12
Penetration

Page 5 of 10
 LoF Root crack IP

.
Figure 6 16mm thick steel plate weld scanned with same-side TOFD at 5 MHz; above is
the scan from Left side and below the scan from the Right side

The red cursor at 9.3µs is approximately the top at the centre of the weld. Flaws on the near-
surface on the near-side of the cap would occur earlier and on the far side of the weld would
occur later.

3. Civa Simulations
In designing some of the application setups, Civa simulations were used to optimise skew, probe
centre separation and refracted angles. These confirm the ability of same-side TOFD to provide
improved near-surface resolution.

Page 6 of 10
 a. HDPE Butt-fusion Joint Confirmed by Civa
The parameters used to design the probe-holder for the HDPE application were determined
using Civa simulations. Figure 7 shows the probe placement and resulting B-scan for the Civa
simulation of the setup used to generate the calibration scan seen in Figure 4. The configuration
used a 67.5° refracted beam and an included skew angle of 50° with the standoff such that the
beam centrelines crossed at the centreline of the joint.

Figure 7 Civa simulation of same-side TOFD on HDPE pipe

b. Steel Plate Confirmed by Civa

Scanning of the steel plate used in Figure 6 was simulated in Civa as well. The plate was 16mm
thick with a single V bevel. Probes simulated were 6mm diameter 5MHz. An included angle
between probes of 40° and a 66° refracted beam in compression mode was used to confirm the
flaw imaging.

Figure 8 Civa simulation of same-side TOFD on 16mm thick steel plate weld

The Civa B-scan indicates a slight time variation between near and far side flaws and strong
shear mode components occurring later in time.

Page 7 of 10
 c. Thin wall Pipe
Using the calibration block design concepts for the HDPE pipe, same-side TOFD scanning of
a steel pipe 115mm diameter 6mm wall thickness was simulated in Civa. Calibration-type
target included a 1x5mm inside surface notch, a 1x5mm outside surface notch and a 2mm and
3mm diameter flat bottom hole placed at the midwall depth. Probes 6mm diameter 5MHz were
used in the simulation and placed with an included angle between probes of 54° with a 75°
refracted compression mode.

Only about 0.25µs separates the peak amplitude response of the inside and outside notches;
however, the total volume of the weld region can be evaluated without concern for a dead zone.
Even with a 10MHz probe, using the traditional TOFD would have the lateral wave dead zone
extend to 3mm depth and the backwall dead zone would be about 0.9mm.

Figure 9 Civa simulation of same-side TOFD on 6mm thick steel pipe calibration targets

4. Conclusions
Same-side TOFD (also known as one-side TOFD, back-scatter TOFD or the Quasi-chord
technique) has been demonstrated to provide useful detection capabilities. With an optimised
skew angle between the transmitter and receiver elements, the lateral wave and backwall signal,
that are present in the traditional TOFD (180° skew) can be eliminated. Eliminating the lateral
wave and backwall signals improves near surface resolution of flaws.

Because the technique directs the beam across the weld and relies on a component of the
diffracted energy returning to the same side of the weld it was transmitted from, the low side of
a mis-matched joint (high-low condition) cannot be assessed. This limitation is the same for
traditional TOFD.

Page 8 of 10
This paper demonstrated the same-side TOFD technique on a limited range of wall thicknesses.
As with traditional TOFD, the beam spread determines the thickness that a single probe position
and refracted angle can be expected to cover. For thick cross sections, multiple positions are
required. This was the rationale for the dual same-side TOFD wedge designed by Eclipse and
illustrated in Figure 2.

Testing thin wall joints with any version of TOFD can be limited by the weld cap size. In
traditional TOFD, the use of high refracted angles is required to provide sufficient standoff
between the nose of the probe and the edge of the weld cap. The same limitation is present for
single-sided TOFD.

Unlike traditional TOFD where a simple solution of the Pythagorean theorem can be derived
to estimate flaw depth and height, the variables involved in same-side TOFD are numerous and
a single approach cannot address the variations. Although for HDPE butt-fusion joints the
estimate of flaws occurring within a short distance of the centreline can be valid, for steel welds
with V bevels, the time difference between the near and far bevel faces can introduce greater
uncertainties in depth of indications.

Civa simulation has been confirmed to be useful in evaluating the variables to optimise the
techniques possible with same-side TOFD.

5. Acknowledgements
We would like to thank Dr. Mark Lozev for locating the references to the original work on
TOFD by Prof. I.N. Ermolov and his translations from Himchenko and Bobrov’s book,
Nondestructive Testing in Chemical and Oil Engineering (1978).
Work on these concepts has been ongoing by the authors for several years and has been
augmented by the simulations afforded by Civa. The authors would like to extend special
thanks to Philippe Dubois and Fabrice Foucher at Extende for their ongoing support with Civa.
We would also like to thank Robert Ginzel, Lincoln Cassie, Mohammad Marvasti and Jonathan
Lesage for allowing us to borrow samples used in this project.
A special note of thanks goes to Paul Holloway who provided us with an equation driven file
to design the probe holders in 3D CAD.

References
1. Silk, M.G., Lidington, B.H., The potential of scattered or diffracted ultrasound in the
determination of crack depth, Non-Destructive Testing, Volume 8, Issue 3, June 1975
2. Ermolov, I.N., Ultrasonic methods for nondestructive testing, University Textbook,
Moscow, Publisher Moscow Federal University, 1968
3. Ermolov, I. N., Korolev, V.D., Ultrasonic bubblers for automated control of products,
Defectoskopia=Nondestructive testing, 1970, No. 3, pages 114-116
4. Ermolov, I.N., Chimchenko, N.V., Sechal, B.A., Ultrasonic testing of electroslag welds,
Annual publications of the Federal Institute of Chemical Machine Building, Moscow,
1973, vol. 57, pages 198-217.

Page 9 of 10
5. Charlesworth, J. P, Temple J. A. G., Engineering Applications of Ultrasonic Time-of
Flight Diffraction, Research Studies Press Ltd., England, Second Edition, 2001
6. Bray, D., Section 3. Ultrasonic Systems for Industrial Nondestructive Evaluation,
Ultrasonic Nondestructive Evaluation Engineering and Biological Material
Characterization, CRC Press, 2003
7. ASTM E3044, Standard Practice for Ultrasonic Testing of Polyethylene Butt Fusion
Joints, American Society for Testing and Materials, Vol. 03.03, 2016
8. Giller, G., Mogilner, L., Khomenko, V, Technologies and Hardware of Ultrasonic
Testing welded Joints of Steel and Polyethylene Pipelines, 15th World Conference on
Nondestructive Testing
[Link] Roma (Italy) 15-21
October 2000,
9. Zakharov, A.V., Advantages of Using Ultrasonic Chord-Type Probes in an Elastic
Wear Plate without a Case for Inspection of Tube Articles’ Welding, Russian Journal
of Nondestructive Testing, Vol. 42, No. 2, 2006, Pleiades Publishing, Inc., 2006
(Original Russian Text published in Defektoskopiya, Vol. 42, No. 2, 2006)
10. Ushakov,V.M., Mikhalev, V.V., Davydov, D.M., Sensitivity of a Flaw Detector
during Ultrasonic Testing by Chord-Type Transducers, Russian Journal of
Nondestructive Testing, Vol. 44, No. 11, 2008, (Original Russian Text published in
Defektoskopiya, Vol. 44, No. 11, 2008).
11. GOSSTANDARD RF. UD-21R, Methodology for Ultrasonic testing of butt ring
Welded joints of steel and polyethylene Gas pipelines (for probes of chord type), Gas
Supervision Department, Gosgortekhnadzor of Russia, 2004
12. Isaenko, F., Kalyukin, G., Calculation and Designing of Universal Transducers for the
Ultrasonic Control of Welded Joint Connections of Polyethylene Pipes,
[Link] , ECNDT, 2010
13. Ginzel. E., Elastomeric Wedges for Pulse-Echo Phased-Array on HDPE Butt Fusion
Joints, [Link]
echo_PA_probes_for_HDPE.pdf, [Link], Issue May, 2020

Page 10 of 10