BACKGROUND

Prior to 2018, the technique used

for ultrasonic testing in CSA W59
was drawn from the U.S. code AWS
D1.1, which used an approach
developed in the late 1960s. The
technique is easily recognized
using large search units and a
fixed estimate of attenuation which
requires only a single side drilled
PHASED ARRAY AND hole to set reference sensitivity.
As sound travels it naturally loses

CHANGES TO ULTRASONIC power just a flashlight beam los-

es intensity over distance. Being

TESTING OF STRUCTURAL able to equally gauge indications

both near and far from the probe

WELDS IN CANADA
is critical in evaluating the work-
manship of a weld. In order to
establish a fixed estimate of sound
attenuation, the technique placed
Most inspection codes worldwide have made strides to very strict limits on the frequency,
introduce newer technologies in the past few decades. size and shape of the transducer.
Phased array ultrasonic testing (PAUT) is one major Additionally, specific inspection
advancement which is now common practice. This article angles were prescribed for weld
joints of different thicknesses,
attempts to distill the technology down to a few main largely regardless of weld geometry.
talking points, highlight the introduction of PAUT into CSA The technique, paired with very
W59-18, and discuss the implications on inspection results. stringent acceptance criteria, has
the formed the basis of ultrasonic
weld inspection of structural welds
in North America for decades.
THE PLOT TWIST one to act as an independent transmitter and receiver
or as a group. By varying the timing of the pulse to
After that nice introduction, you knew there was a each element, the resulting wave front can be steered
“but…” and here it is. Any technique that prescriptive through a range of angles.
is basically immovable when it comes to evolving
technologies. Something as simple as calibrating for Three beams of different angles are shown in Figure 1
actual attenuation to provide the ability to use more below. The blue lines represent the relative time
suitably-sized probes, long used in the ASME world, delay issued to each individual element.
wasn’t possible. As well, angle-specific inspection
procedures effectively ruled out phased array and its In the center image, all the beams are fired at the
“lots of angles at once" approach. The demand for same time, thus all the delays are identical and no
improved probability of detection, fewer false calls, beam steering is performed. This essentially gener-
and better repeatability are things that manufacturers, ates a planar wave front simulating a conventional
construction companies, and inspection firms all strive UT transducer. After refracting at the interface
for. Quality and efficiency are constantly evolving between the wedge and the steel, the resulting
and improving. Eventually, the code had to change beam angle is 55°, which is controlled only by the
for the times. cut angle of the wedge.

INTRODUCTION OF ANNEX X On the left, the elements near the top of the transducer
are fired first (represented by shorter blue delay
In 2018, Annex X was added to CSA W59 to allow lines), while the elements at the bottom are fired last
manual (e.g. hand scanning) phased array and con- (longer blue lines). The resulting composite wave
ventional UT using a modified sensitivity calibration. front steers slightly downwards, generating a 45°
This is referred to as the TCG technique in CSA W59-18, refracted angle in the steel plate shown.
as opposed to the long standing fixed attenuation
technique. The TCG technique is designed to provide Similarly, by reversing the delay laws, the beam can
the same quality levels as fixed attenuation, and the be steered upwards resulting in a 70° refracted angle
code permits that either technique may be used. The as seen in the last image.
fixed attenuation technique remains part of the code
for now as it has been for decades. Typically, beams are directed at 1° increments
throughout a range of angles. Each set of delays
Annex X was designed to be used essentially as a used to produce a beam of a certain angle is referred
standalone set of instructions to reduce the amount to as a focal law.
of cross-referencing to Clause 8 as possible.
By compiling a series of focal laws similar to those
OK, BUT WHAT ABOUT PHASED ARRAY? above over a wide angle range (e.g. 45° to 70°, in 1°
increments), the result is a swept angle (sectorial, or
Phased array ultrasonic testing simply divides a
S-scan) which provides coverage over a much larger
transducer into many small elements, allowing each
area than conventional UT as shown in Figure 2.

Fig. 1: Phased array focal laws generating beams at three different angles.
 Fig. 2: Phased array angular sweep.

A WINTER ANALOGY: BECAUSE WE’RE THE PHASED ARRAY VIEW

CANADIAN
The view most associated with PAUT is the S-scan.
Conventional ultrasound is like cleaning off just one The S-scan is a composite of all the A-scans defined
bare strip on your icy windshield, then driving down in the focal laws. It is the most used view in PAUT
the highway squinting, moving your head side to weld inspection but must be used in combination
side trying to tell if the oncoming lights are two with the individual A-scans at each angle to properly
little cars or one really large oncoming semi-trailer. evaluate indications.

Fig. 5: Phased array S-scan view. Fig. 6: Individual A-scans of all

focal laws.

Fig. 3: View through a conventional UT windshield. The S-scan above shows an indication from a simple
side-drilled hole in a calibration block. The hole was
PHASED ARRAY IS LIKE CLEANING OFF THE round, but the indication is not, and instead looks like
WHOLE WINDSHIELD. a red and yellow banana. Because sound spreads out
as it propagates, a line drawn from the entry point
along a beam angle that looks like it “just misses” the
hole will in fact still reflect a little. This beam spread
is the primary reason why indications on the S-scan
appear elongated.

As well, the beam reflects off the nearside faces of

discontinuities and only a portion of the reflected
beam may return to the wedge and be detected.
This means that the images seen in the S-scan may
not exactly resemble the actual shapes of the defects,
but are used rather for detection and characterization
rather than forming a precise image.
Fig. 4: View through a phased array windshield
 Fig. 7: Actual phased array weld scan plan and data image.

The tightness of the beam and the

ability to keep the smearing effect
to a minimum is aided by use of a
higher number of elements (e.g.
32 versus 16) which improves
beam steering. However, “more
is not always better”, as using
a transducer with simply more
elements will create a wider beam,
which on thinner components can
complicate matters as the entire Fig. 8: Scan plan example.
thickness of the specimen is flood-
ed with sound and differentiation
between the top and bottom determine the position and depth the probe(s) are positioned at set
surfaces becomes difficult. Match- of the defect. In many cases, a distances from the weld centerline
ing the transducer and focal law slight oscillation will result in a for the length of the scan. Prior to
parameters to the weld geometry higher indication level. The data testing, a scan plan is conceived in
is key. produced during manual scanning which the technician evaluates the
are typically screen images showing beam coverage for the weld and
MANUAL VS. AUTOMATED an S-scan with a weld overlay and a determines whether one or more
TESTING corresponding A-scan at the angle stand-off positions are required for
representing the highest indication full volumetric inspection.
In CSA W59-18, manual phased rating. Phased array as well as
array has been implemented in conventional ultrasound are both Since the probes in automated
Annex X. The term manual testing commonly used for manual testing. testing are fixed in place, they
implies that the technician is cannot be rastered or oscillated.
manipulating the probe by hand. Automated testing requires one or Thus a proper scan plan showing
This is the typical “weld scrub” more probes positioned in a fix- probe placement and beam cov-
used in conventional ultrasonic ture which is moved laterally along erage is critical for an accurate and
inspections of structural steel the length of the weld. The data is repeatable inspection. At this time,
weldments. In manual testing, recorded during acquisition. The automated phased array testing
the probe can be freely oscillated term “automated” is somewhat of has not been implemented into
and rastered (moved back and a misnomer, as the scanning sys- CSA W59, although it remains (as
forth, left and right) to peak the tem may be moved manually or by it has for years) an option in Clause
signal in the A-scan and accurately the aid of motors. In either case, 8.2.12 requiring engineer approval.
 Fig. 9: Various PAUT views clockwise from top left: S-scan, End view, Top view, A-scan.

With automated testing, it is

possible to display the data in
several different views which can
provide much more detailed
record of the weld integrity than
static images from manual testing.

CALIBRATION
The sensitivity of the equipment
must be set prior to inspection.
As sound attenuates as it travels,
signals are amplified as they travel
to produce equal sensitivity across Fig. 10: Phased array probe on a side-drilled hole calibration block.
the entire sound path. A typical
calibration block for phased array A-scan. On analog-era machines, Corrected Gain), which is simply a
will include side-drilled holes or the DAC was physically drawn on mathematical correction applied to
notch reflectors at various distances the CRT display using a wax marker. a DAC in order to bring all signals
from the entry surface. The height of indications along the up to the same level. In this manner,
sound path are compared to the indications from reference reflectors
A plot of attenuation is called a height of the DAC. at various distances are all normal-
DAC (Distance Amplitude Curve), ized at the same screen height mak-
which the equipment will automati- More common in the present ing detection and evaluation easier.
cally display as an overlay on the day is the use of a TCG (Time
 Fig. 11: DAC view of sample indication. Fig. 12: TCG view of same indication.

ACCEPTANCE CRITERIA TYPES CODE COMPLIANCE AND PHASED ARRAY

TESTING
There are two main approaches to acceptance criteria.
The names of each are not immediately descriptive, Phased array has been accepted worldwide in many
but can be simplifi ed as follows: codes and standards. In North America, CSA W59,
AWS D1.1 and ASME Section V all include sections
Workmanship (or, “how loud is that sound?”) on phased array testing.
Evaluating the severity of a defect on the amount of
sound it refl ects is referred to as workmanship criteria. In 2018, CSA W59 adopted the use of manual PAUT
This is the simplest approach and the most commonly alongside the ability to use conventional probes other
used. There is a threshold level above which indications than the traditional “snail” and “big red brick” which
are evaluated. Indications below the threshold are been a staple in structural steel weld inspections for
ignored. Usually the indication level and the defect nearly 50 years. Either technique can be used: the
length are the main parts in evaluating for acceptance original “Fixed Attenuation” technique from Clause
or rejection. 8 with the large probes or the “TCG Technique”
in Annex X which permits both manual PAUT and
Workmanship criteria may be applied to both conven- conventional UT. Use of the TCG technique for W59
tional UT and PAUT. The amount of sound that a defect does not require the engineer’s approval, but shall be
refl ects is of course highly dependent upon defect used only by agreement between the contractor and
type, location, orientation, size, and beam angle. the welding inspection organization.

Fracture Mechanics (or, “tell me everything!”) CSA W59-13 and earlier… CSA W59-18
Encoded inspections on equipment such as critical
service pressure vessels may require more detailed Clause 8.2 Clause 8.2
information on the weld fl aws than simply “how
loud” and “how long”. Measuring the distance from
the surface, the through-wall height, the length,
distance from other fl aws and proper characterization
requires considerably more time and expertise than Annex X
a workmanship approach. A fracture mechanics
Fig. 13: Typical search units
approach usually also requires a demonstration for
compatible CSA W59-13
procedure qualification.
and earlier.

Fig. 14: Typical search units

compatible with CSA W59-18.
The TCG technique in Annex X permits the inspector ultrasound. The integration of PAUT through W59
to choose the inspection angle best suited to the Annex X and the resulting multitude of inspection
weld. This is the natural course of action for most angles will provide a much clearer picture of the weld
inspection codes worldwide, and of course provides integrity. But this does not equate to failing more
a natural fit for PAUT where multiple angles are used welds. Whatever was in the welds before will still be
at once. there, but more modern inspection techniques like
the TCG technique and PAUT will help with probability
Adoption of automated PAUT, where encoded digital of detection as it will with reducing false calls.
records can be stored, analyzed and archived, will
eventually be adopted in W59. However, acceptance Anyone involved on the manufacturing, construction,
levels for automated UT should be distinct from those or inspection side of the industry can attest that
of manual UT. the human factor is the major variable in inspection
results. Phased array requires a full understanding of
Inspections of pressure vessels and piping in accor- ultrasonic testing, but results are less ambiguous than
dance with ASME Section VIII and B31 are quite conventional ultrasound where only the A-scan view
commonly performed using phased array and have is used. The effects on probability of detection and
been used with those codes for many years. failure rates vary significantly due to the human factor.
Training, experience, and qualification are key.
PHASED ARRAY AND PROBABILITY OF
DETECTION  Paul Holloway, P.Eng. MASc
President, Holloway NDT & Engineering Inc.
Phased array is no doubt a more comprehensive tool
for ultrasonic inspection of welds. You can see more
at once with PAUT than you can with conventional