18th World Conference on Nondestructive Testing, 16-20 April 2012, Durban, South Africa

Manual Weld Inspection with Ultrasound

- Conventionally or with Phased Arrays?
Dr. (USA) Wolfram A. Karl DEUTSCH, Dipl.-Geol. Stefan KIERSPEL
Karl Deutsch Prf- und Messgertebau GmbH + Co KG
Otto-Hausmann-Ring 101
42115 Wuppertal, Germany
w.deutsch@karldeutsch.de, kierspel@karldeutsch.de , www.karldeutsch.de
Summa ry For several years now portable Phased Array Systems have been available
and used and the question arises, whether Phased Arrays will replace the conventional
manual ultrasonic testing method in the near future. In this paper these two techniques
are compared under the aspect of manual weld testing. The result is that both
techniques offer advantages and disadvantages concerning the complexity of the
inspection, the evaluation, the inspection speed, the type of documentation, the
qualification of the inspector and finally the total cost.

Introduction
In the past years the development of ultrasonic phased array systems has made good
progress. The phased array technique was first applied in fully automated testing
systems, but meanwhile small portable instruments are available due to
miniaturization of electronic components and improved computing pow er, thus
enabling mobile manual inspections similar to conventional ultrasonic testing. The
question is now, whether phased arrays will replace the conventional ultrasonic
testing method and whether this makes sense concerning technical and economic
aspects. For some special applications the phased array technique provides
advantages but the question is still, whether this applies accordingly also for the
classical fields of ultrasonic inspection.
Such a classical field is the manual inspection of welds, wh ich is applied worldwide
in all industries and it always stands under considerable strain to be efficient as far as
possible. New techniques are therefore welcome.
But in spite of that many users face problems to decide, whether in their special scope
of work it makes sense to invest time and money into the new technique or whether it
may be better to remain with the conventional methods. In the following , both
techniques conventional ultrasound vs. phased arrays are compared under the
aspect of weld testing. At the end the advantages and disadvantages of both
techniques are presented side by side, without any judgment. This shall provide
decision-making aids for users to find the right answer for their projects.

Of course it is necessary to mention that the inspection procedures are described as applied in
the practice, e.g. the application of a position encoder when using phased array probes.
Encoders are normally not applied in the case of conventional weld testing.
The principle of the phased array technique is intentionally not described in this paper, because
it is well-known meanwhile.

1. Manual Weld Inspection with Conventional Ultrasound

1.1 Selection of the equipment
At the beginning of each weld inspection the appropriate equipment has to be
selected. Normally digital ultrasonic testing (UT) instruments are used, although
sometimes also older analogue instruments are applied. The system adjustment and
the type of documentation are different, however not the inspection itself. Due to the
fact, that digital instruments are mostly used, only the procedure for digital systems
will be described in the following.
More extensive is the selection of the right probe(s). Most often angle beam probes
are used. It is necessary to select the right sound frequ ency, test angle and transducer
size. According to the standards and specifications sometimes also multiple angles
are required, so that several probes have to be selected and applied consecutively.

1.2 Selection of the inspection range

After selection of the test equipment the inspection range has to be specified. In order to
cover the whole weld cross-section, the weld has to be scanned at least between the half
and full skip distance, see figure 1.
Minimum Scan Length

Fig. 1: Selection of the inspection range

The minimum length of the meandering probe scan depends on the material thickness and
on the angle of sound incidence. It is useful to mark the borders with two lines parallel to
the weld seam.

1.3 Instrument adjustment

Prior to operation the testing system has to be adjusted.
The distance adjustment is carried out using the standard calibration blocks K1 or K2, or
if the material is different concerning the sound velocity, this is done by means of an
appropriate reference block. If the sound velocity is known it ca n be entered directly into
the digital instrument. Modern digital instruments contain a probe database, including the
probe delay. The parameters of the distance adjustment can be stored and loaded when
required.
The sensitivity is adjusted by means of artificial reference reflectors, e.g. side drilled holes or in
case of the DGS method also on the back wall of the reference block. In this case the data for
sound attenuation and the transfer correction have to be entered. If several probes have to be
applied, the sensitivity adjustment has to be done for each of them.

1.4 Inspection procedure

To test the weld the probe is scanned meandering between the borders of the minimum scan
length (figure 2). This enables a complete scan of the weld cross-section and length. In order
to detect also defects which are not oriented longitudinally in the weld, the probe is swivelled
(figure 3). This procedure is rather time consuming and an experienced inspector needs
between 15 and 30 minutes for a weld length of one meter.

Verschiebebereich

Schwenkbereich

Verschiebebereich
parallel

ebene
Reflektoren

Fig. 2: Meandering scan of the weld

voluminse
Fehlstellen

Fig. 3: Swivelling of the probe

1.5 Localization and evaluation of inhomogeneity

After having detected an inhomogeneity in the weld, it has to be evaluated concerning location
and size.
When all parameters, i.e. x-value, material thickness and the angle of sound incidence are
entered correctly, the digital instrument directly provides the defect position relative to the
probe. The defect position then can be marked on the object or entered into the protocol
table in relation to a fixed reference point on the object under inspection.
The defect size is evaluated using the 6 dB drop method (half value length), if the defect is
larger that the sound beam diameter. Otherwise the DGS or DAC methods are applied,
where the defect reflectivity is compared to artificial reflector types. This size estimation is
rather inaccurate. However, due to missing alternatives, it is recognized and used
everywhere.

1.6 Documentation
For documentation the A-scan and the according parameters like echo amplitude, equivalent
reflector size, projection distance and depth can be transferred from the instrument, and if
wanted, also all adjustment parameters. All further parameters, i.e. the defect position in
relation to the object's datum point have to be entered manually into a table or drawing.

Fig. 4: A-scan with evaluation values

2. Manual Weld Inspection with Phased Arrays

2.1 Selection of the equipment
Concerning the instrument selection it has to be ensured that enough parallel an d serial
channels are available for the required inspection and that enough focal law calculations are
possible.
The criteria for the probe selection are frequency, number of elements, pitch, element width
and length. In case of weld inspection the frequency is normally 2 or 4 MHz. Here typically
linear phased array probes are used with 16 or 32 elements. Furthermore it has to be
considered that the right plug type or adaptor is available.
In order to enable an angle beam inspection with shear waves, additionally a plastic wedge is
necessary.

2.2 Specification of the inspection method

At the beginning the operator has to decide, whether the inspection shall be done by
means of a sectorial scan with fixed virtual probe and different angles (figure 5) or a
line scan with fixed angle and moving virtual probe (figure 6). In this case long
probes with 64 or more elements and enough space on the material surface are
necessary.

Fig. 5: Sectorial scan

2.3 Specification of the inspection range

Fig. 6: Linear scan

Also with phased arrays a weld is inspected between the half and full skip distance. In case of
sectorial scan the weld cross-section is covered with a large angle of sound incidence, e.g. 70,
in the bottom and with a small angle, e.g. 40, at the surface, see figure 7. This could cause the
problem, that the weld bevel (and defects such as lack of fusion) might not be hit by an
appropriate ultrasonic testing angle.

Fig. 7: Covering the weld cross-section with a sectorial scan

2.4 Instrument adjustment

The adjustment of a phased array instrument is much more complex compared to a conventional
ultrasonic flaw detector. The length and the angular range have to be adjusted correctly in order to
cover the whole cross-section. The focal law calculation needs to consider that the sound exit point
changes and thus the length of the delay in the wedge. Modern phased array instruments provide
some scan plan aids. However, a certain experience of the operator is required. The time adjustment
can be done using the calibration block K1.
For sensitivity adjustment several calibrations are necessary: balancing of the different element
sensitivities, angle corrected gain (ACG) and, if wanted, the time corrected gain (TCG). For this
purpose more complex test blocks are currently proposed. The defect size is still estimated using the
conventional DGS or DAC method since defect sizing from a sector scan is problematic.

2.5 Definition of the coordinate system

When using the phased array technique for the manual weld inspection normally a position
encoder is connected. This enables the generation of C- and B-scans showing the position of
the defects in relation to the reference point and coordinate system on the object. The
coordinate system has to be well defined and the encoder has to be calibrated carefully.
Theoretically it is of course possible, to use encoders also in case of the conventional
technique. This is state of the art for testing machines. However, it is not common practice for
manual inspections, because additional scanning and imaging hardware / software is required.
2.6 Inspection procedure
After all adjustments and calibrations the inspection can begin. In contrast to meandering
scanning with conventional probes a phased array probe only has to be moved along one line
parallel to the weld seam with an appropriate offset from the weld centre. This enables an
inspection which is 10 to 20 times faster than conventional testing. However, only sections
of 20 to 40 cm can be inspected at a time due to mechanical handling issues and the amount
of data, which has to be stored. This means, that the scan has to be stopped after 20 to 40 cm,
stored and then started again. Larger wall thicknesses might required more than one scan.

2.7 Representation of the inspection results

A valuable advantage compared to the conventional ultrasonic testing is the possibility to
generate top, side and end views of the weld, indicating correct defect positions in relation
to the object coordinate system.
Figure 8 represents a C-scan as top view of the weld, a sectorial scan (S-scan) as crosssectional view and an A-scan. The scans are linked together and can be selected via cursors.
This means, that with a cursor in the C-scan the corresponding S-scan can be selected and
with a cursor in the S-scan the A-scan accordingly.

Fig. 8: Defect marked by cursor in phased array screen (A-scan, S-scan, C-scan)

2.8 Defect localization and size evaluation

The defect localization is relatively easy due to the prese ntation in the above mentioned
views. If a defect is larger than the sound beam the half value width can be determined by
means of the cursors. The vertical and horizontal positions are directly available.
However, if the defect size is smaller than the sound beam diameter, the flaw size still has
to be estimated by means of the DGS or DAC method using a conventional UT probe. A
size measurement in a phased array picture is not yet possible. A good example is shown in
figures 9 and 10, where the images of two side drilled holes with different diameters are
compared.

Fig. 9: S-scan of the 50 mm hole in the K1

Fig.10: S-scan of the 5 mm hole in the K2

Although there is a considerable difference in size of the two holes it is not possible to estimate
the dimensions from the S-scan. It is possible to improve the imaging by means of focusing to
the according depths. A dynamic depth focusing is not yet available on actual portable phased
array instruments. Furthermore there are not yet standards for the phased array technique and
therefore the defect size estimation still has to be done by means of the conventional DGS or
DAC methods.

2.9 Documentation
The documentation is relatively easy by means of the different views, which contain all
position information. But as mentioned under 2.8, the defect size has to be evaluated
conventionally using the DGS or DAC method. This requires additional documentation.
For this paper numerous test samples were investigated which contained natural weld
defects. Conventional ultrasonics, phased array, TOFD (time of flight diffraction) and for
some cases also X-ray testing were compared.

Fig. 11: Weak detection for phased array insonification from one side (top), X-ray without
indication (middle), and good detection with phased array from second side (bottom)

Fig. 12: The defects could also be well detected with TOFD (middle)

Fig. 13: Metallographic result with lack of fusion parallel to surfaces (not good for X-ray)
The influence of the wall thickness, weld bevel angle and respective test angles, optimum
distance of probe to weld, flaw size to be detected, etc, were documented. Still it remains
difficult to give clear recommendations, because of the large range of test applications.

3. Conclusion
The comparison of the two inspection methods shows that both techniques have
advantages and disadvantages. To provide a clearer overview and possibly as decision making aid the aspects are summarized below:
3.1 Inspection with conventional ultrasound advantages
The conventional ultrasonic inspection is a well-established method for decades. Extensive
experiences and empirical values are available, also for difficult object geometries and material
properties. There are experienced inspectors who do not need additional qualification. The
inspection and acceptance criteria are well described in standards. The conventional setup of the
equipment is relatively easy. By swivelling of the probe echoes of inclined oriented defects can
be optimized. The costs for the equipment and the operators training are relatively low.

3.2 Inspection with conventional ultrasound disadvantages

The meandering scanning of a weld is very time consuming. The presentation of defects in
an A-scan is very abstract and the evaluation needs well-experienced operators. The
documentation is difficult because of missing imaging. The documentation of defect
positions has to be done manually. An inspection with more than one angle requires
sequential operations, which increases the inspection time. A later data analysis is not
possible due to missing data recording.

3.3 Inspection with phased arrays advantages

Due to imaging by B-, C-, S- and L-scans the evaluation and documentation is easier. The
inspection speed is higher because of fast scanning in combination with electronic beam
steering. A connected position encoder simplifies the defect localization. The continuous
data recording enables a later interpretation of the results with a PC.

3.4 Inspection with phased arrays disadvantages

The setup for a phased array system is considerably more difficult, which may result in
many errors. The larger phased array probes can cause coupling and space problems. A
careful preparation and probe positioning is required in order to benefit from coordinate
related imaging. Standards for the application of phased arrays and acceptance criteria
are still under development. Special trainings of the operators are necessary and finally
the start-up investment as well as the operating costs (for instrument, probes and wedges)
are higher compared to the application of conventional equipment.

4. References
[1] V. Deutsch, M. Platte, M. Vogt: Ultraschallprfung Grundlagen und industrielle
Anwendungen (Ultrasonic Testing Principles and Industrial Applications, in German
language), 372 pages, Springer Publishing House, 1997.
[2] V. Deutsch, M. Platte, M. Vogt, W. A. K. Deutsch, V. Schuster: Ultrasonic Testing
Compact & Understandable, 77 pages, Castell Publishing House Wuppertal, 2002.

[3] V. Deutsch, M. Vogt,: Ultraschallprfung von Schweiverbindungen (Ultrasonic Testing of

Welds, in German language), 124 pages, DVS Publishing House Dsseldorf, 1995.
[4] prISO/DIS 13588:2010-11 Non-destructive testing of welds - Ultrasonic testing - Use of
(semi-) automated phased array technology
[5] prEN 16018:2009-12 Non destructive testing - Terminology - Terms used in ultrasonic
testing with phased arrays
[6] ASTM E2491:2006 Standard Guide for Evaluating Performance Characteristics of PhasedArray Ultrasonic Examination Instruments and Systems
[7] L. Le Ber, O. Roy, N. Jazayeri: Applications of Phased Array Techniques to NDT of
Industrial Structures, TINDT2008, 2008
[8] L. Le Ber, O. Roy, P. Benoist: Ultrasonic Phased Array Inspection Modelling with CIVA,
Modelling NDT, 2007.
[9] P. Benoist, P. Calmon, S. Leberre, T. Sollier: CIVA, An Integration Software Platform for
the Simulation and Processing of NDT Data, WCNDT 2004.