ECNDT 2006 - Tu.2.5.

The History of AUT

Frits DIJKSTRA, Röntgen Technische Dienst bv, Rotterdam, The Netherlands
Jan DE RAAD, Rotterdam, The Netherlands

Abstract. Automated Ultrasonic Testing (AUT) of welds has a long history.

Around 1950, the first ideas on how to inspect a weld with multiple ultrasonic
transducers and a scanning mechanism arose and were patented. The first AUT
machine for the inspection of pipeline girth welds was successfully tested in the
lab in 1959.
In this paper, the history and backgrounds of mechanised and automated
ultrasonic inspection are illustrated with pipeline girth weld inspection as an
example. The paper not only addresses the technical developments, but also
standardization of the basic zonal concept, development of acceptance criteria for
weld imperfections and improvements in accurate defect sizing which were crucial
for AUT introduction and acceptance. The paper will highlight lessons learned
during this process.
Recent developments in phased array technology, using for example
innovative Inverse Wave field EXtrapolation (RTD IWEX) technology are going
to enable 3D imaging of defects in the real sense. First results show not only
enhanced defect detection capabilities but also reliable determination of shape,
orientation and size with the ultrasonic wavelength as the only limitation. The
potential impact of such a development will be discussed.

1. Introduction

From the very beginning of the history of ultrasonic testing (UT), shortly after the second
world war, it was realized that (manual) UT would have a difficult time competing with
radiography (RT). This had a good reason: RT produces a permanent record, UT doesn’t.
And manual UT was, apart from too slow, also considered too subjective and dependent on
operator’s skills to ever be able to beat (or even equal) radiography.
A few years later, around 1950, a young engineer in RTD realized that something
had to be done about this. He was convinced that UT (if properly applied) should be more
reliable in detecting planar defects than RT [1]. By connecting multiple probes to a
scanning mechanism and using a paper chart recorder, he argued, it should be possible not
only to produce a record like RT does, but also to arrive at a reproducible inspection. In
addition, he realized that the inspection could only be made fast enough if performed in one
scan along the weld, using multiple transducers each directed to a certain portion (zone) of
the weld.
This young engineer was Arie de Sterke, renowned Dutch NDT pioneer. He applied
for a patent on his multi-probe system using a zonal concept in November 1951. The patent
was awarded in 1952 in The Netherlands, Belgium, Germany, France and Great Britain [2].
Fig. 1 shows one of the illustrations of the patent. The first commercial applications,
introduced by RTD in 1956, were on longitudinal welds in pipe mills (fig. 2).

1
Fig. 1: Illustration from the 1952 patent Fig. 2: AUT in a German pipe mill, around 1956

Nowadays, Automated Ultrasonic Testing (AUT) is used all over the world, for
many different applications: nuclear components, long-distance pipelines, plant piping,
pressure vessels, tanks etc., not only on carbon steel welds but (with special transducers)
also on more “difficult” welds (austenitic and Duplex steel). In pipeline construction
industry, AUT has replaced RT almost entirely, particularly in offshore pipe laying.

2. A fifty year’s learning experience in AUT

In 1957, Arie de Sterke came up with the idea to apply his mechanized UT, up to
then successfully used in pipe mills, for pipeline girth welds as well. “If it works for
longitudinal welds, why not for circumferential welds?”. This led to the construction of the
first RTD Rotoscan® in 1959 (Fig. 3).

Fig. 3: First RTD Rotoscan, 1959

2
 The equipment was designed for mobile application and was successfully tested in
the laboratory. The trade name “Rotoscan” was registered, which registration still exists
today.
Between the laboratory tests of the first Rotoscan (1959) and the first commercial
successes (1989) lay thirty years. In the following sections it will be outlined why this took
so long.

3. Technical developments

The 1959 prototype turned out to be far ahead of its time. Neither the technical state
of the art nor pipeline industry were ready for it. This was the reason why the project was
“mothballed” for almost 20 years.
Around 1970, RTD became involved in NDT of nuclear facilities. This included the
inspection of nuclear reactor pressure vessels (Fig. 4), involving wall thicknesses up to 400
mm and the presence of austenitic clad layers. Also the primary piping was part of the
inspection scope, with its wall thicknesses up to 60 mm as well as austenitic welds and
stainless steel clad layers (fig. 5).
For all these inspections, multiple probe systems were used in a zonal concept.
High-end multi-channel ultrasonic equipment was developed, by RTD and others. All these
inspections had to be performed remotely operated, using multi-probe scanning
mechanisms for safety reasons (radiation protection) [3].

Transducers

Transducers

Fig. 4: Inspection of the reactor pressure vessel (1975) Fig 5: Inspection of primary circuit weld (1975)

Between 1970 and 1980, a vast amount of experience was gathered in ultrasonic
NDT of nuclear components. In 1975, RTD published a paper [4] arguing that the
technology now developed could very well be used for pipeline NDT as well. This paper
highlighted the superior detection of planar and crack-like defects, as well as the necessity
to use tandem technique, dependent on wall thickness and weld shape. The paper also
highlighted the specific advantages of mechanized inspection for pipeline girth weld
inspection during offshore pipe laying.

3
4. Field tests

A few years later, in 1978, RTD extensively validated Rotoscan together with two
major potential customers. This project was mainly driven by the rapid developments in
offshore pipe construction using lay barges. The trials used a “second generation” Rotoscan
and included performance evaluation on hot welds - simulating lay barge conditions - with
surface temperatures up to 200°C. The validation tests showed good correlation between
the ultrasonic findings and the intended defects. It was also confirmed that inspection at
elevated temperature was possible, using water as a coolant and coupling medium.
Experience was gathered with the use of stable wedge polymers, to minimize temperature
influences on probe angle and sensitivity.
Unlike the 1959 Rotoscan prototype, which had been based on an unwieldy pipe
cutting manipulator (fig. 3), the “second generation” Rotoscan used a heavy yoke
manipulator intended for offshore use where weight is not a major issue. To reduce
inspection time the manipulator included two identical probe sets, 180º apart, so rotation
over 180° rather than 360° plus some overlap was sufficient to cover the entire weld (fig. 6)
[5].

Transducers
2 sets of
transducers

Fig 6: Concept of second generation Rotoscan Fig. 7: Field trial on Castoro Sei
(photograph by courtesy of SAIPEM)

In the same year, 1978, RTD conducted a field test on the Castoro Sei lay barge,
operated by SAIPEM, see fig. 7. This field test showed the potential capabilities of
Rotoscan – serious defects missed by RT were detected - but also its shortcomings: many
false calls were experienced and a high sensitivity to electrical interference from the lay
barge’s communication systems. The latter could be solved in a relatively short time. The
first shortcoming would take another ten years to solve.
Because high-grade steels started to be applied in Canada for construction of long-
distance cross-country pipelines, there was an increasing need for a high POD and better
sizing to enable the use of Engineering Critical Assessment (ECA) criteria. The Canadian
pipeline industry therefore considered another NDT method with better capabilities than
RT. Being aware of the RTD efforts towards the development of a pipeline AUT system,
Alberta Gas Trunk Line (AGTL, now part of NOVA) contracted RTD in 1978 to

4
demonstrate the capabilities of the second generation Rotoscan, along with standard
radiography, during a Canadian cross-country pipeline construction project (fig. 8).

Fig. 8: Rotoscan during field trial in Canada

During these Canada trials in summer 1978, Rotoscan’s field performance was not
good enough to fulfil all Canadian code requirements. Nevertheless the results were very
promising [6]. Signal to noise ratio was sufficiently high to detect fusion defects with a
height of 1 mm and up. Over 100 welds a day could be inspected. The immediate feed back
to the welders, pioneering as well with mechanized welding, was highly appreciated. RTD
therefore decided to continue further development.
One of the first steps taken after these field tests was to construct a light-weight
scanner to replace the bulky yoke construction. The renowned welding company CRC-
Evans could be convinced that a scanner could best be based on their welding mechanism.
In this way RTD could use the bands already present for welding. This would considerably
speed up the AUT process.
In 1979 the Rotoscan, now equipped with the new scanner on a band, was tried at
several summer- and winter projects of AGTL. For these projects, RT was the mandatory
NDT method. Defects detected with both RT and UT were judged, as usual, against
workmanship criteria, so Rotoscan was used complementarily. Results showed that a
considerable number of relevant planar flaws missed by RT were, as expected, detected by
AUT. Nevertheless, some further refinements were required. Particularly the high rate of
false calls, caused by weld geometry, cap and root, should be reduced. This was considered
feasible by proper measures, and RTD accepted the challenge to continue its development
efforts.
Over a period of ten years, many improvements were introduced and tested in
Canada during numerous field trials. These included winter jobs at temperatures as low as
minus 35°C. This work was strongly supported by the Canadian pipeline industry, in
particular AGTL/NOVA. These improvements included, in addition to continuous
equipment and probe modifications, the selection of low-temperature coupling liquids and
umbilical cables. But the most relevant improvements, better discrimination between root
defects and geometry as well as detection and quantification of porosity, had yet to come.

5. Root inspection improvement

In the 1977/1978 version of Rotoscan, recording was still done by means of a go –

no go recorder. Only signals exceeding a pre-determined threshold were recorded. Such

5
thresholds were based on workmanship criteria and set on simple reference plates with
artificial defects. Discrimination between weld root defects and indications from the root’s
geometry was done by applying a gate that ended at the weld centre line. The idea was that
defects would be in the gate, and geometry would not (fig. 9). In many of the field trials
and evaluations it became clear that this concept had severe shortcomings in case of weld
misalignment.
Fast industrial analogue paper chart recorders had become available in the
meantime. Using these, it was possible to record the full echo amplitude and transit distance
in each gate. But the single gate concept was not an adequate one, false call rate was still
too high. Too often echoes from the weld penetration and cap were interpreted as rejectable
defects.
Part of the false calls were caused by wall thickness variations. RTD therefore
designed UT equipment with gate lengths continuously adjusted on the basis of a
continuous wall thickness measurement. This facility worked well, but was not sufficient to
solve the entire problem.
As a next step, an extra gate was introduced. One short gate, as usual, to record
defects, the other (longer) gate to record also the echo from the weld bead penetration
(geometry). The latter was used as a reference. In both gates, amplitude and transit distance
were recorded. Indications having a shorter transit distance than the geometry echo could
be interpreted as flaw signals (fig. 10).

Fig. 9: Old concept of root defect discrimination Fig. 10: Root defect discrimination using extra gate

Although interpretation of these analogue records was painstaking, this concept

reduced the number of false calls to almost zero. This improvement (late eighties) can be
considered the last step before “Mapping” could be introduced. We just had to wait until
industrial computers were fast enough.

6. Towards smaller zones

During the early trials, welds had been divided in only three zones regardless of
wall thickness or weld preparation. These zones were inspected by individual non-focused
probe elements. Since, at that time, the only way to size defect height was to count the
number of zones a defect occurred in, a higher thickness resolution was needed to enable
use of the ECA-approach. This required the use of more zones and, as a consequence, more
probes with narrower beams [6,7].

6
 A study by AGTL and the Canadian Atomic Energy Commission (AEC) indicated
that focused beams would provide the required through-thickness resolution and that such
probes could be manufactured. It also indicated that beam angles should be perpendicular to
the anticipated orientation of the potential defects, to achieve maximum reflection. As a
consequence, the geometry of the weld preparation would determine the required beam
angles and widths, and tandem technique would have to be used when steep bevels were
applied.

7. Anisotropy of pipeline steels and reference plates

With the introduction of smaller zones, beam angles and beam geometry became
more critical. But there was also another reason for beam precision to become more
important: high-grade pipeline steels (X65 and up) are slightly acoustically anisotropic.
This had already been observed by RTD around 1985. Similar observations were published
by Ginzel et al in 1995 [8]. As a consequence, these fine-grain steels show a non-constant
sound velocity. Shear wave velocity can vary up to 10 %, dependent on beam angle. This
effect can differ per pipe supplier or –batch and depends on chemical composition, pipe
production process (rolling) and grain size. This is the reason that, in pipeline AUT,
reference plates have to be made out of the actual pipe material. Fortunately, once the
velocities in the plate are known, probes can be tailored to the individual pipe material.
With smaller beams, and the use of tandem technique, flat bottom holes became
necessary for sensitivity setting. Their introduction required a special reference plate
geometry. RTD introduced the Z-shaped reference plate (often referred to as “calibration
block”) in 1980 (fig. 11).

Fig. 11: Z-shaped reference plate (1980). Some of the flat bottom holes and notches are clearly visible.

8. First commercial success

By 1989, after numerous improvements and trials, the Canadians considered

Rotoscan mature enough to award RTD with the first commercial AUT job: a contract
where AUT was the sole inspection technique [7]. Trans Canada Pipe Lines (TCPL, now
part of NOVA) was the launching customer, applying an adapted national code with
modified criteria. Only during start-up of the construction project, RT was used
complementary to AUT.

7
 This contract can be considered the break-through in the cross-country pipeline
market for AUT. It would take almost another decade until the same thing would happen in
offshore industry.
By 1993, an AUT standard was issued in Canada, mainly thanks to the persistent
efforts of the companies that now form NOVA [9]. It was also in this period that other
pipeline AUT providers started to enter the market, and the name AUT was introduced.

9. Introduction of Mapping and ToFD

During the experimental years of AUT, an ever-increasing amount of data had to be

recorded on go-no go and analogue paper chart recorders. At some projects, analogue
recorders were in use with 32 channels of information, creating an almost insurmountable
wealth of data. All this information had to be interpreted in the relative short time available.
Being aware that “mapping” would presumably solve the data recording and interpretation
problem, RTD eagerly and impatiently waited for powerful and robust computers to enter
the market.
When ultimately these computers became available in the early nineties, RTD made
great efforts by in-house software development to introduce mapping as a promising
solution. It was anticipated that mapping, very comprehendible for the operator and others
involved, would provide the proper presentation to quickly deal with the vast amount of
data necessary to interpret the condition of a weld.

Fig. 12: Principle of root mapping Fig. 13: Porosity mapping

The idea behind mapping in the root is that it facilitates data interpretation by
“pattern recognition” to differentiate weld flaws from geometrical indications. As on
radiographs taken from welds, mapping allows viewing indications in relation to the
position of the weld root (fig. 12) [9].
Mapping also provided a good solution to record the presence of porosity - a
prerequisite to inspect manually made welds with AUT [6,7,9] - which was almost
impossible until mapping was introduced. Mapping not only shows the presence of porosity
but, to some extent, it is also able to establish its severity. Similar to root mapping, porosity
mapping uses an additional gate covering the entire weld volume (fig. 13).
The implementation of Time of Flight Diffraction (ToFD) technique (fig. 14) was a
logical additional step, after some projects executed by RTD for the American Gas
Association (A.G.A., now PRCI) had demonstrated that the POD of pulse-echo in
combination with ToFD was higher than with the use of pulse-echo alone. In addition,

8
ToFD added defect height sizing capabilities, independent of amplitude and therefore
especially significant for defects with unexpected orientations [10,11].

Transmitter probe Receiver probe

A
B
C
D

A B C D

upper tip backwall echo

lateral wave lower tip

Fig. 14: Time of Flight Diffraction Fig. 15: ToFD image showing weld defect

Having added mapping and ToFD, the Rotoscan data visualization method
introduced by RTD around 1993 (fig. 16, left) was already very similar to the concept used
in present days (fig. 16, right).

Fig. 16: Examples of Rotoscan visualizations of welds with intended defects (left: 1994, right: 2004)

The 1994 display shown in fig. 16 already showed ToFD, pulse-echo (amplitude and transit
distance), root- and porosity mapping, go – no go and coupling check traces. Although a lot
of functionality for easy interpretation and flaw evaluation was added in the years to follow,
this concept of pipeline AUT data visualization first introduced by RTD is still used today,
by most AUT providers, and was described in ASTM Standard E 1961 in 1998.

9
Computerized data storage also enabled the data to be copied to mass storage media such as
CD-ROMs, as a compact, easily accessible and durable way of data archiving by clients.

10. AUT goes offshore

To change from one NDT-method to another in offshore pipe laying requires very
good incentives. Lay barges used to be equipped with RT equipment on mainline and
double-joint stations. NDT operators as well as welders were familiar with X-ray.
But good old X-ray had gradually become too slow for the new, faster GMAW
welding processes. This situation not only required a faster NDT method, but also one that
would be better able to detect inclined planar defects such as lack of fusion (LOF). Shorter
inspection times and a high POD on lack of fusion defects ultimately made a change in
NDT method quite attractive. The advantages outbalanced the drawbacks by far. The most
impressive feature of AUT was its high speed: a few minutes per weld is sufficient, even
for the largest pipe diameters. Because of the “real-time” presentation of results,
interpretation can take place while the weld is being scanned, thus hardly any time is lost in
waiting for results.
The first offshore AUT job was performed by RTD in 1996. Allseas, a pipeline
contractor then located in the Netherlands, decided to apply AUT on a small diameter (3
inches) piggyback line (fig. 17). To increase production speed to 360 welds a day, they
successfully replaced double-wall single-image RT by AUT as a more efficient (faster) and
economic alternative [10]. This offshore job marked the beginning of a successful
introduction period of AUT in offshore pipe laying.

Fig. 17: Offshore inspection of 3” piggyback line (courtesy Allseas)

11. Current situation

Today, pipeline AUT has grown into a very professional way of performing NDT
[12,13]. Many developments, partly covered by patents, support inspection quality and
speed. Examples are:

10
• A job preparation wizard automatically generates a lay-out of the probe system,
including tandem probe combinations, as well as a reference plate drawing.
• Automatic Calibration Control automatically checks equipment performance settings
before each weld is scanned.
• Amplitude-based defect height sizing is performed automatically by the software. The
capabilities of UT to determine defect height are, in general, considered to be limited
and influenced by many factors, and have been subject to long-standing discussions.
Nevertheless, pipeline AUT has achieved a high level of height sizing accuracy. This is
because pulse-echo probe sets (including tandem combinations) can in many cases be
accurately designed for optimal reflection on certain defects with known orientations
and locations, such as lack of side wall fusion and lack of penetration [16,17]. To
enhance sizing accuracy, software algorithms use probe-specific input parameters for
each zone. ToFD forms a safety net for sizing of defects with unexpected orientations.
• Interactions of multiple defects are automatically evaluated and compared with the
code’s acceptance criteria.
• To be able to do the inspection at elevated temperature, temperature-stable probe wedge
materials are used. In addition, wedge temperature is continuously monitored.
• Phased arrays have been introduced to replace multiple probe sets [13,14]. These
phased arrays “emulate” conventional probes in a zonal concept, according to the
requirements of standards such as ASTM E 1961.
• Procedures have been developed to inspect pipelines and risers with internal clad layers
for corrosion protection (CRA – Corrosion Resistant Alloys) and austenitic welds [15].
These procedures are based on the use of dual angle probes generating compression
waves, which were first introduced (by RTD) around 1970 for use in the nuclear
industry.
• Personnel qualification and QA aspects have been developed to a level that quality and
consistency is ensured in job preparation, NDT performance and data interpretation.

12. Legislation

The first decade of commercial AUT on pipeline girth welds was performed on the
basis of existing national or company standards, commonly based on good workmanship.
Existing acceptance criteria for weld imperfections were usually adapted to reflect the
detection capabilities of AUT [6,7]. Practices were mainly derived from API 1104. In all
cases, job-specific qualifications were a necessity. This was not a satisfactory situation.
During this period of AUT pioneering several national and international research
projects were carried out, to provide code-writing organizations with sufficient information
to compile generic standards, based on qualified and validated AUT characteristics. As an
example, a Dutch joint industry round-robin project on wall thicknesses between 6 and 15
mm compared RT and UT techniques. In other (international) projects, performance of
mapping and ToFD facilities in AUT systems were extensively qualified on a large number
of welds with induced defects. Particularly the results of work for PRCI/AGA, with many
international members, created considerable market exposure.
In the same period, other projects were directed towards acceptance criteria
development. All this material, and numerous efforts of others, ultimately led to the issue of
an international standard. ASTM issued the first document to describe a standard practice
of AUT of girth welds based on the zonal concept and including suggestions for acceptance
criteria (based on Workmanship and ECA) in 1998. Zonal division and ECA criteria were
also described in the DNV offshore standard DNV-OS-F101 appendix E, issued in 2000. In
1999, API issued standard 1104 19th edition.

11
 These documents provide means of ensuring a uniform and consistent inspection,
regardless of variation in equipment, service providers and application environments.
Because the very important legislation became available, millions of pipeline girth welds
have been (and will be) inspected with AUT.
Currently, close to 150 AUT systems are in use all over the world. All current
pipeline AUT providers now use the same zonal concept, mapping and ToFD. As a result
of optimized AUT equipment in combination with inspection procedures and appropriate
acceptance criteria, weld repairs are now almost exclusively performed if relevant defects
are present [9,16,17].

13. Lessons learned

This paper shows, that NDT technologies from different industrial disciplines can
support developments for other applications by cross-fertilisation.
It is interesting to observe that the real commercial break-through in AUT on
pipeline girth welds occurred simultaneously with the introduction of other AUT service
providers in this market. This supports the idea that the market only accepts such a method
in a competitive market.
Once the technical developments in AUT had reached a level enabling acceptance
by pipeline industry, other than technical aspects also started to become important.
Therefore, another important lesson is: to develop a piece of equipment accepted by
industry may be a major achievement, but sustained operation of dozens of these units in
complex projects all over the world is just as important. This requires a professional,
project-based organisation capable of maintaining a high level of operational excellence
and flexibility. Personnel training, job preparation, procedure control, logistics, equipment
maintenance, all these aspects are equally important for commercial and technical success.
Looking at the history of pipeline AUT, it becomes evident that many companies were able
to develop equipment and sometimes even qualify for a job, but only a few managed to stay
in business for many years.
Last but not least: the long history of pipeline AUT introduction demonstrates that
technical developments only are not sufficient to introduce such an NDT approach. It takes
the heavy involvement of clients, regulatory bodies and NDT industry to achieve broad
acceptance [18].

14. Future

In recent years, Phased Arrays were introduced in pipeline AUT by RTD and other
pipeline NDT providers. These arrays are used to “emulate” multiple probes in a zonal
concept (fig. 18), according to present legislation (e.g. ASTM E 1961).
For selected jobs, phased arrays are preferred over conventional multi-transducer
systems because they offer additional operational advantages: job preparation is more
efficient, because the same pair of arrays can be used for many weld shapes and wall
thicknesses. Beam geometry optimisation is done by selecting optimum beam parameters in
the software. As a result, phased arrays offer a substantial amount of operational flexibility.
The present family of phased arrays used in pipeline AUT is of the linear type. This
means that they consist of a single array of elements, offering beam steering (angle of
incidence, index point, focus) in one plane. It is envisaged that, in the future, phased arrays
will also be used in other modes, offering even more capabilities. Matrix (two-dimensional)

12
arrays will offer beam steering in two planes, also enabling beam movement in lateral
direction. This is expected to result in even higher POD’s and more accurate sizing.

Fig. 18: Phased array used to “emulate” the zonal discrimination concept

A further possible step is the use of ultrasonic arrays in a completely different mode:
illumination of a weld defect and reconstruction of the signals, received by the array’s
elements, in a computer will offer the possibility of generating a three-dimensional image
of the defect, regardless of its location, shape and orientation. RTD is currently developing
this for AUT, in close co-operation with Delft University of Technology. The imaging
theory used here is based on Inverse Wave Field Extrapolation (IWEX), which was first
developed for seismic exploration of oil and gas fields. Using this for AUT will allow
enhanced detection, accurate sizing and defect characterisation with the ultrasonic wave
length as the only factor limiting accuracy [13,14,19]. After completion of the proof of
principle in 2005, using synthetic signals obtained from finite difference simulations, the
technology is now being experimentally verified on test pieces with reflectors of various
shapes and sizes. An example is shown in fig. 19.

Fig. 19: Array probe on test piece with tilted slit (left) and resulting image (right)

Such developments do not only require extensive technical and scientific work, but
also a lot of validation on welds will be involved to gain confidence in the technique. The
results of such validations will, in turn, provide the necessary knowledge and information
to be able to develop legislation. Only after such a trajectory, which can take several years,
it will be possible to benefit from these developments.

13
15. Conclusions

After many years of persistent pioneering and marketing by RTD, AUT of pipeline
girth welds is now offered by a number of NDT providers as a worldwide service and has
reached a high level of professionalism. The process of development and introduction took
several decades. Pipeline industry recognized the advantages over radiography and showed
this by allowing field trials, performing studies and stimulating legislation.
Primary key factors in the ultimate success of pipeline AUT became its capability to
discriminate between defects and geometry indications, to detect and quantify porosity and
to use ECA criteria. These three factors helped to avoid false calls, thus reducing the
number of unnecessary repairs.
AUT practices, now used by all AUT providers, are still based on the zonal concept
patented by RTD in 1952.
Developments in pipeline AUT are still in progress. Defect detection,
characterisation and sizing will, in the future, be further enhanced by efficient use of
ultrasonic phased arrays.
AUT has been able to demonstrate, during its commercial use in the last two
decades, that is really the viable and better alternative to Radiography a young Dutch
engineer predicted it to be in 1950.

16. References

[1] A. de Sterke, The Examination of Longitudinally Welded Tubes, presented in IIW Commission V in
1957, IIW Doc. 57/58

[2] British patent No. 726.824 “Method for the Inspection of Materials, especially of a particular Zone
thereof, with the aid of ultrasonic waves”, awarded to RTD, Rotterdam in 1952

[3] J.A. de Raad, F.H. Dijkstra, Unique Manipulators for NDE in the Nuclear Industry, Proceedings of the
8th International Conference on NDE in the Nuclear Industry, 17 – 20 November 1986, Kissimmee,
Florida, USA

[4] A. de Sterke, Can Reactor Test Systems be used for efficient Pipeline NDT? British Journal of NDT,
Vol. 17, No. 4, July 1975 (President’s Honour Lecture 1974)

[5] United States Patent No. 4,375,165 “System for Inspecting Welded Joints in Pipelines by means of
Ultrasonic Waves”, awarded to RTD, Rotterdam in 1983

[6] A.G. Glover, D. Hodgkinson, D. Dorling, The application of Mechanized Ultrasonic Inspection and
Alternative Acceptance Criteria to Pipeline Girth Welds, Proceedings of the Pipeline Technology
Conference, Ostend, Belgium, October 1990

[7] B. Gross, J. O’Beirne, B. Delanty, Comparison of Radiographic and Ultrasonic Inspection Methods on
Mechanized Girth Welds, Proceedings of the Pipeline Technology Conference, Ostend, Belgium,
October 1990

[8] E.A. Ginzel, A. Ginzel, R.K. Ginzel, Study of Acoustic Velocity Variations in Line Pipe Steel,
Materials Evaluation, May, 1995

[9] D. Hodgkinson, D.V. Dorling, A.G. Glover, Mechanized Ultrasonic Testing of Pipeline Girth Welds
Produced by Gas Metal Arc Welding, Proceedings of the Pipeline Technology Conference, Ostend,
Belgium, September 1995

14
[10] J.A. de Raad, F.H. Dijkstra, Mechanized UT Now Can Replace RT on Girth Welds During Pipeline
Construction, Proceedings of the Pipeline Technology Conference, Ostend, Belgium, September 1995

[11] H.A.M. van Merriënboer, Field Application of Mechanized US Inspection of Large-Diameter

pipelines, Proceedings of the Pipeline Technology Conference, Ostend, Belgium, September 1995

[12] M.R. Blinde, F.H. Dijkstra, J.A. de Raad, Reliability of AUT Systems for the Inspection of Girth
Welds during Pipeline Construction, Pipes and Pipelines, Vol. 46, No.4, july-August 2001

[13] T. Bouma, N. Pörtzgen, F.H. Dijkstra, Advances in Non-destructive Testing (NDT) of Pipeline Girth
Welds, Proceedings of the Conference “Application and Evaluation of High-Grade Line Pipes in
Hostile Environments, November 2002, Yokohama, Japan

[14] N. Pörtzgen, A. Gisolf, G. Blaquière, Advances in Imaging of NDT Results, Proceedings of the
WCNDT, August-September 2004, Montreal, Canada

[15] F. Kopp et al, Production and Inspection of Steel Catenary Riser Welds, Presented at OTC, May 2003,
Houston, Texas, USA

[16] F.H. Dijkstra, J.A. de Raad, Why develop Acceptance Criteria for Pipeline Girth Weld Defects,
Proceedings European-American Workshop “Determination of Reliability and Validation Methods of
NDE”, Berlin Germany, June 1997

[17] F.H Dijkstra, J. van der Ent, T. Bouma, Defect sizing and ECA Criteria, State of the Art in AUT,
Proceeding of Pipeline Conference, Brugge, Belgium, May 2000

[18] E.A. Ginzel, Mechanised Ultrasonic Inspections of Pipeline Girth Welds, A Brief History, NDT.net,
March 2000, Vol. 5, No. 03

[19] N. Pörtzgen, Practical Results of Ultrasonic Imaging by Inverse Wave Field Extrapolation,
Proceedings of the ECNDT, September 2006, Berlin, Germany

15