Professional Documents
Culture Documents
Inspection of Welds
NDT44
Preliminary Pages
Contents
Standards and Associated Reading
COSHH, H&S, Cautions and Warnings
Certification Schemes
Handbook Front Cover, Copyright Warning
Preface of the Third Edition
Introduction
1 History
2 The Principles of Phasing
3 Beam Forming
3.1 Transducer arrays
3.2 The ultrasonic beam
3.3 Phased array aspects of beam characteristics
3.4 Fermat’s principle
3.5 Phased array beam focusing and steering
4 Hardware: Pulsers, Receivers, Motor Control and Encoders
4.1 Advantages of digital control
4.2 Pulsers and receivers
4.3 Pulsers
4.4 Spike pulsers
4.5 Tone burst
4.6 Square wave pulsers
4.7 Receivers
4.8 Gates
4.9 TCG/TVG
4.10 Data acquisition and automated systems
4.11 Automated systems in general
4.12 System components
4.13 Instrument outputs
4.14 Scanning displays and scanning equipment
4.15 Memory and digitisation aspects
4.16 Data processing
4.17 Scanning equipment
4.18 Limitations of mechanised scanning
4.19 Scanning speed
4.20 Encoders
4.21 Asynchronous versus synchronous systems
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Preface
These notes are provided as training reference material and to meet the study
requirements for examination on the NDT course to which they relate.
They do not form an authoritative document, nor should they be used as a reference
for NDT inspection or used as the basis for decision making on NDT matters. The
standards listed are correct at time of printing and should be consulted for technical
matters.
Note: These training notes are not subject to amendment after issue.
Revision Record
Revision
Date Changes made
number
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
CEN prEN 583-6 Time of Flight Diffraction Technique: A method for defect
detection and sizing.
ASTM E-2192 Standard Guide for Planar Flaw Height Sizing by Ultrasonic
Testing.
Associated reading
Introduction to Phased Array Ultrasonic Technology Applications,
By R/D Tech Inc., Published by R/D Tech Inc., 2004, ISBN 0-9735933-0-X
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Many thousands of substances are used at work but only about 500 substances
have Workplace Exposure Limits (WELs). Until 2005 it had been normal for HSE to
publish a new edition of EH40 or at least an amendment each year. However with
increasing use of the website facilities the HSE no longer publishes a revised
hardcopy edition or amendment.
The web based list which became applicable from 1st October 2007 can now be
found at www.hse.gov.uk/coshh/table1.pdf
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Manual Handling
Manual handling is the operation of supporting or transporting loads by hand. This
includes the lifting, putting down, pushing, pulling, carrying or moving of a load.
Note: In the Manual Handling Operations Regulations 1992 a load includes a person
or animal. More than one-third of all over-three-day injuries reported each year to the
HSE and to local authorities are the result of manual handling.
Hazards
Injuries can be caused by:
Incorrect posture when lifting, even if the weight is within the capacity of the
person.
Excessively heavy or awkwardly shaped components causing strain on back,
arms or hands.
Handling of components with sharp edges for example those coming straight
from machining operations.
Components that are difficult to handle due to a coating of fluid or dust for
example released during machining.
Frequent repetitive movements for example bending or twisting can also result in
injury even if the movement is well within the capacity of the person in a straight lift.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Both schemes offer NDT certification conforming to both EN 3 and ISO 9712;
Qualification and Certification of NDT personnel.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
PCN certification is a scheme which covers the qualification of NDT inspection staff
to meet the requirements of European and International Standards. Typically a
standard or procedure will call for the inspector to be certified in accordance with
EN473 and/or PCN requirements. The PCN general document describes how the
PCN system works.
The points below cover extracts from this document which are major items, the full
document can be viewed on the BINDT website: www.bindt.org.
References
PCN Documents
PSL/4 Examination availability.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Set up equipment.
Carry out the test.
Record and classify the results in terms of written criteria.
Report the results.
Level 1 personnel have not demonstrated competence in the choice of test method
or technique to be used, nor for the assessment, characterisation or interpretation of
test results.
Level 3 personnel are qualified to direct any NDT operation for which they are
certificated and may be authorised by the employer to:
Assume full responsibility for a test facility or examination centre and staff.
Establish, review for editorial and technical correctness and validate NDT
instructions and procedures.
Interpret codes, standards, specifications and procedures.
Designate the particular test methods, techniques and procedures to be used.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Within the scope and limitations of any certification held carry out all level 1 and
level 2 duties.
Provide guidance and supervision at all levels.
Level 3 certificated personnel may be authorised to carry out, manage and supervise
PCN qualification examinations on behalf of the British Institute of NDT.
Where level 3 duties require the individual to apply routine NDT by a method or
methods within a particular product or industry sector, the British Institute of NDT
strongly recommends that industry demand that this person should hold and
maintain a level 2 certification in the applicable methods and sectors.
Training
Table 1: Minimum required durations of training
NDT method Level 1 hours Level 2 hours Level 3 hours
ET 40 40 40
PT 16 24 24
MT 16 24 32
RT³ 40 80 72
RI N/A 56 N/A
UT 40 80 72
VT 16 24 24
BRS 16 N/A N/A
RPS N/A 24 N\A
Basic knowledge (direct access to Level 3 examination 80
Parts A-C)
Note: Direct access to level 2 requires the total number of hours shown in table 1 for
levels 1 and 2. Direct access to level 3 requires the total number of hours shown in
table 1 for levels 1-3. Up to one third of the total specified in this table may take the
form of OTJ training documented using form PSL/42 provided it is verifiable and
covered practical application of the syllabus detailed in CEN ISO/TR 25107:2006.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
In the event that the experience is sought following successful examination, the
results of the examination shall remain valid for up to two years.
Direct access to level 2 requires the total number of hours shown in table 2 for
levels 1 and 2. Direct access to level 3 requires the total number of hours shown in
table 2 for levels 1-3.
Qualification examination
Table 3: Numbers of general questions
NDT Method Level 1 Level 2
ET 40 40
PT 30 40
MT 30 40
RT 40 40
RI N/A 40
UT 40 40
VT 30 40
BRS 30 N/A
RPS N/A 20 plus 4 narrative
Note: All level 1 specific theory papers have 30 questions.
All level 2 specific theory papers have 36 questions.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Re-examination
A candidate who fails to obtain the pass grade for any examination part (general,
specific or practical) may be re-examined twice in the failed part(s), provided the
re-examination takes place not sooner than one month, unless further training
acceptable to BINDT is satisfactorily completed, nor later than twelve months
after the original examination.
A candidate who fails all permitted re-examinations shall apply for and take the
initial examination according to the procedure established for new candidates.
A candidate whose examination results have not been accepted for reason of
fraud or unethical behaviour shall wait at least twelve months before re-applying
for examination.
Summary
The PCN scheme is managed and administered by the British Institute of NDT
(BINDT) on behalf of its stakeholders.
It meets or exceeds the criteria of EN473.
There are six appendices covering various industry and product sectors:
1. Aerospace.
2. Castings.
3. Welds.
4. Wrought products and forgings.
5. Pre and in-service inspection (multi sector).
6. Railway.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
A pass is obtained where each part is 70% or over with an average grade of 80%
or over.
A PCN certificate is valid for 5 years.
Renewal and recertification requirements are covered in CP16 for levels 1 and 2
and CP17 for level 3.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
The scheme is conducted in essentially the same format as the PCN scheme
previously described. One notable difference between the schemes is CSWIP does
not offer Aerospace qualifications.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
A Training Handbook
E. Ginzel
Copyright 2002-2008
Third Edition
2008
by
E.A.Ginzel
Prometheus Press
Waterloo, Ontario
Canada
Copyright Warning
All materials produced for teaching this course of study, including
all lectures and any supplementary materials are protected by
copyright. You are permitted to use these materials only for your
personal study and research. Use of the materials for any other
purposes, including sale of your personal lecture notes, without
express permission of the copyright owner, may infringe copyright.
The copyright owners may take action against you for infringement.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
The second edition of this book reduced the medical origins content and added a
large section on the basic ultrasonic aspects relating to the theory of the method as
they apply to phased array ultrasonic testing. It corrected several of the errors found
in the first edition and added some examples of field applications.
Ultrasonic phasing fundamentals are provided but these are much simplified
compared to the technical authoritative references provided by the Olympus
Publications. As with the previous editions, some coverage is provided on the
construction and considerations for the use of phased array probes.
NDT using phased arrays is usually associated with mechanisation and digital
controls so a brief coverage on these main aspects is provided.
Although the solutions to the problems can be found by reading the contents of the
book, they may not always be as simple as direct repetition of wording. Answers may
sometimes require clever deductions from the information presented for the multiple
questions and the simple equations found in the book may be used to derive
answers to the numeric problems.
www.twitraining.com
Rev 3 January 2010
NDT Prelims
Copyright TWI Ltd 2013
Introduction
Advances in mechanisation capabilities afforded by the rapid growth of the computer
industry have been wide spread. NDT is no exception to the variety of industries that
have benefited from these advances.
The book is structured with a history and then we review the basics of waves to build
up the background on phasing. The principles of wave mechanics and diffraction are
the same for single element probes as well as phased array probes. Therefore the
review of principles for mono-element probes leads to an extrapolation of these
principles to the applications of phased array ultrasonics.
Phased array UT systems are now nearly always computerised and often used in
configurations that hold the probe (or probes) in a carrier. The carrier is almost
always encoded and is often motorised (although manual operation of encoded
carriers is also common).
This makes it necessary for the student to have some background on the apparatus
used with the phased array probe and data acquisition system. Having covered the
principles of phasing and data processing (albeit at a rudimentary level) examples
from industry are used to illustrate the adaptability of phased array techniques and
some of the post-processing enhancements that can be used with phased array
systems.
www.twitraining.com
Section 1
History
Rev 3 January 2010
History
Copyright TWI Ltd 2013
1 History
Phased Array Ultrasonic Testing has its origins in medical ultrasound. An
excellent background on all aspects of ultrasonic testing history is provided
by Dr. Woo on his website at http://www.ob-ultrasound.net
This small section on history draws liberally from Dr. Woo’s information. The
concept of multiple elements in a single housing was not that recent. Tom
Brown at Kelvin and Hughes filed an application for patents of an annular
dynamically focused transducer system as early as 1959 but this was not
based on phasing.
Not until the late 1960s did the timing circuits for the phasing of ultrasonic
pulses become published. In 1968 Jan C Somer published a paper on
electronic sector scanning for ultrasonic medical diagnosis in the Journal:
Ultrasonics.
Prototypes of phased-array systems had all the elements used for each
pulse. Variable time delays were introduced between the elements in both
transmission and reception modes so that the beam was steered in a
particular direction. As improvements to electronics became available so did
the timing-circuitry evolve.
In 1976 Thurstone and von Ramm at the Duke University published a more
advanced version of the electronically steered arrays. Their array generated
ten different receive focal laws. It combined beam steering (previously
developed) with a dynamic focusing in the receive mode. At the time, this
was considered a significant innovation in design.
Grating lobes are unique to phased array transducers and are caused by
the regular periodic spacing of the small array elements. When the energy
of these lobes is reflected by off-axis structures and detected by the
transducer, the signals produced are artefactual and they are considered
ghost images that interfere with evaluation of the main image.
1-1
www.twitraining.com
Rev 3 January 2010
History
Copyright TWI Ltd 2013
The origin of these grating lobes was found to be the inability of the system
to obtain the necessary phase interference for beam steering. It was
overcome by ensuring that each individual element had been cut to a half
wavelength width. Smaller element width ensures that the individual lobes
produced at each element increases the angle of divergence to greater than
90 degrees.
Early medical systems suffered from electrical noise being introduced into
the receiver section. This noise was associated with the multiplexer or
switching network delay. This noise was added when the system changed
delays.
To avoid this electrical delay switching noise required expensive, low noise
delay lines and multiplexers. Another timing consideration was to achieve
effective dynamic focusing. This requires that the delay resolution on every
channel be a small fraction of the ultrasound carrier period. To construct this
would require delay lines containing many hundreds of taps which made
manufacturing both complex and expensive.
Therefore, in the old days analogue systems with high bandwidth and
having large numbers of channels (addressing the large numbers of
elements) made manufacturing prohibitive.
Not until the improvements in electronics in the latter 1970s did this problem
get overcome. In 1979, Samuel Maslak, then at Hewlett Packard (HP),
patented a method for dynamic delays without changing delay taps.
The rapid increase in computing speed and data flow has allowed significant
refinements to phased-array technology. Today, software programmes allow
probe optimised design, beam prediction calculations, precise beam
placement calculations and ultra-fine resolution imaging and pulsed-doppler
methods for assessing fluid-flow velocities.
Medical ultrasound still leads the way in R&D advances but NDT is no
longer lagging so far behind as it had in the previous decades. Medical
development has made some fascinating advances including Doppler and
high contrast resolution imaging.
1-2
www.twitraining.com
Rev 3 January 2010
History
Copyright TWI Ltd 2013
NDT does not yet have much use for the Doppler features of medical
ultrasonic assessments but high temporal resolution and contrast
resolutions are being taken advantage of.
There will be limits to the ability of our NDT systems to have the startling
imaging associated with medical systems. Medical ultrasound has the
advantage of the test materials having acoustic velocities about one quarter
that we are accustomed to in NDT.
This has the effect of significantly shortening the wavelength used which
has a direct bearing on the resolution possible (not to mention that only
compression modes are generated in medical applications).
For an example of phased array high resolution imaging see figure 1.1, it
shows a 3-D rendering of a foetal face.
1-3
www.twitraining.com
Rev 3 January 2010
History
Copyright TWI Ltd 2013
An ultrasound scanner that provides clear and detailed pictures from the
womb will allow parents to see what their unborn child looks like and enable
doctors to detect abnormalities much earlier.
1-4
www.twitraining.com
Section 2
Reflection of waves:
This is the process whereby a wave meeting a boundary between two
media is bounced back and remains in the first medium, eg light striking a
mirror.
Refraction of waves:
This occurs when a wave travels from one medium into another. It is bent or
refracted at the boundary. The wave changes direction and undergoes a
slight change in wavelength.
Diffraction:
Occurs when an obstacle distorts a wave eg if the wave travels through a
gap it may be diffracted. It is the ability of a wave to spread round corners.
Diffraction can occur at the edge of an obstacle.
Polarization:
This occurs when the vibrations of transverse waves are confined to one
plane only.
Interference:
Interference occurs when two or more similar waves meet. There are two
types of interference; Constructive and Destructive.
Constructive interference:
This occurs when the waves are in phase when they meet. In phase
means their crests and troughs coincide. The resultant wave will have
amplitude equal to the sum of the individual waves producing it.
Destructive interference:
This occurs when the waves are out of phase. If the crest of one wave
coincides with the trough of another total destructive interference will
result. In general the amplitude of the resultant wave is obtained by
subtracting the individual amplitudes.
2-1
www.twitraining.com
Rev 3 January 2010
The Principles of Phasing
Copyright TWI Ltd 2013
The result of adding two waves depends on the value of the amplitude and
phase of the wave at the point in which the waves are added. The following
illustrates some examples of how two waves interfere.
Figure 2.1 shows two waves of the same amplitude but having different
wavelengths (frequencies) approaching each other. In frame 1 the separate
wave pulses are seen as black lines.
In frame 2 the original wave pulses are faint coloured lines (orange for the
pulse moving from left to right and blue for the pulse moving from right to
left) and the resultant pulse-shape of the two waves as they pass each other
is seen as the black line.
The main feature in frame 2 shows the positive peaks of the two pulses both
occurring at the same point. In frame 3 we see a well defined location where
the positive and negative peaks of the two pulses occur together.
Simply adding the positive amplitude to the negative amplitude at that point
results in zero amplitude because the magnitude of the values is equal but
the sign is opposite.
Frames 4 to 6 show the progress of the two pulses at increasing times until
at frame 7 the pulses have passed beyond each other and the return to the
original sinusoidal shapes. The addition of the amplitudes is carried out for
the entire waveform at every point in time.
2-2
www.twitraining.com
Rev 3 January 2010
The Principles of Phasing
Copyright TWI Ltd 2013
Add to a maximum
1 2
3 4
5 6
The wavelength of the pulses made in a phased array transducer would all
be the same. Since the displacement of an element of a piezo-material is
proportional to the applied voltage, the amplitudes of the pulses made by
adjacent element would all have the same amplitude too.
Therefore the only variable from one element to the next would be the point
at which the waves meet (ie the phase delay). The point where the adjacent
wave pulse has a maximum displacement that coincides with its neighbour’s
maximum displacement will provide a constructive interference and if the
maximum displacement of one pulse meets the minimum of its neighbour
2-3
www.twitraining.com
Rev 3 January 2010
The Principles of Phasing
Copyright TWI Ltd 2013
then the opposite moving particle displacements would cancel each other
and no pulse occurs.
These concepts are illustrated in figure 2.2. When in phase the amplitudes
of the waves add and no change is seen in the frequency. When 180° out of
phase the waves cancel each other and no particle displacement occurs as
illustrated by the flat line in the 4th frame of the upper image.
1 2 3 4 5 6
Destructive Interference: Two pulses moving towards each other with opposite
phase but the same frequency and amplitudes.
1 2 3 4 5
Constructive Interference: Two pulses moving towards each other with identical
phase, frequency and amplitudes (note the doubling of amplitude at the point the
two pulses coincide).
Defined, the Huygens Principle states; “every point on a wavefront may itself
be regarded as a source of secondary waves. Therefore, if the position of a
wavefront at any instant is known, a simple construction enables its position
to be drawn at any subsequent time” (Christian Huygens 1629-1695).
Typically we apply a voltage pulse to the element via the wires attached to
either side of the element. The pulse is of a short duration and causes the
element to expand as the maximum voltage is reached and then return to its
rest position when the voltage is removed ie the voltage drops back to zero.
2-4
www.twitraining.com
Rev 3 January 2010
The Principles of Phasing
Copyright TWI Ltd 2013
The entire flat surface of the element moves out and back making a
displacement in the surrounding medium and a plane wavefront is initiated,
see figure 2.3.
-----------
++++++++
++++++++ -----------
A displacement
point exists at Unstressed mode
every domain point
Stressed mode
When all of the elements are hit with the same voltage spike all expand in
unison and the effect is the same as hitting a single element of the same
size.
Figure 2.4 illustrates the multi element construction of a phased array probe
where several elements are hit with the voltage pulse at the same time.
Each individual element would, on its own, radiate a curved wavefront due
to the small size of the element. When all are hit at the same time the effect
is to form a single large wavefront having the same dimensions and
characteristics of a single element of the same dimension as the combined
multi-elements of the phased array probe.
2-5
www.twitraining.com
Rev 3 January 2010
The Principles of Phasing
Copyright TWI Ltd 2013
Interference wavelets
Unstressed m ode
Stressed mode
Figure 2.6 illustrates the traditional single element probe with backing and
electrical contacts.
2-6
www.twitraining.com
Rev 3 January 2010
The Principles of Phasing
Copyright TWI Ltd 2013
Figure 2.7 on the left side, illustrates the layout of a single element of the
linear array version of a phased array probe. This is essentially identical to
the single element except for the electrical contacts of which there are as
many as there are elements cut along the length of the probe.
The gap and element width are calculated to provide optimum performance
based on frequency of the element (piezo material thickness). A multi-
contact connector, conductively fused to the prepared piezo element
sections, provides electrical contact to the outer face of the element.
For most applications the element width is about a half wavelength. The
element length, perpendicular to the scan plane, is typically 10-15mm and
focal lengths of the lenses are 5-10mm depending on frequency.
Backing
Piezoelectric
element
Electrode
connections
Quarter-Wave
Matching layers
Lens material
Figure 2.7 Components of the phased array probe, courtesy Philips Medical
Systems.
In the simple compression mode, with all elements fired at the same time,
the rise and fall of the probe face composed of the multi element array is for
all purposes seen the same as a single solid element experiencing the same
deformation.
2-7
www.twitraining.com
Rev 3 January 2010
The Principles of Phasing
Copyright TWI Ltd 2013
The image on the right in figure 2.8 is 3mm x 10mm x 2 microns in size. It is
a 20MHz cardiac catheter transducer. The image on the left is their first
3.5MHz 2D array using the MEMS techniques.
Figure 2.8 Small scale large arrays, courtesy Duke University CECT.
2-8
www.twitraining.com
Section 3
Beam Forming
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
3 Beam Forming
3.1 Transducer arrays
Multiple piezoelectric elements arranged in patterns in a common housing
are called arrays, these are usually linear, matrix or annular in shape, see
figure 3.1.
10 element linear array 4x4 matrix planar array 4 element annular array
This can be used where resolution of a small element is needed but there is
no room for motion or the surface cannot be coupled easily, thereby
allowing no relative probe motion.
In the late 1970’s McElroy and Briers fashioned a probe of concentric rings.
Each ring was composed of cylindrical sections to reduce inter-acoustic
coupling (cross-talk).
The annular array multiple focal spots do not improve resolution over fixed
units but it does permit beam focusing at various distances in the material
being tested.
Elements in an array are usually small and flat. The wavefront off an
individual element is therefore somewhat omni-directional in both
transmission and reception.
3-1
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Phased arrays may use the same sort of multi-element configuration as the
sequenced array. However, the timing of the pulse to each element is a
variable.
If we were to delay the pulse to each successive element by some time less
than half the period of the emitted signal, the wavefront resulting from the
interference is an incident beam with an angle of incidence controllable by
electronics.
Figure 3.2 shows the effect of a linear array pulsed with increasing delay to
each element.
3-2
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
1 Beam steering.
2 Dynamic focusing.
Using precise timing and phase interference effects, we can constantly vary
both incident angle and zone of maximum resolution.
Arrays described above included linear, matrix or annular arrays. To this list
we could add a special version called sectorial annual. This latter version
allows both spherical focusing and angular beam steering, such a probe is
also called a rho-theta probe, see figure 3.5.
3-3
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Y=6.5
47 46
48 45
30 29
49 44
31 28
17 16
50 32 18 15 27 43
8 7
51 33 19 9 3 2 6 14 26 42
1
52 34 20 10 4 5 13 25 41 61
11 12
53 35 21 24 40 60
22 23
36 39
54 59
37 38
55 58
56 57
Linear arrays are the most common type and can perform scanning control
in one plane only. Linear arrays typically minimize the number of elements
required and thereby keep the cost down.
Matrix arrays can scan in two dimensions and offer considerably more
flexibility, albeit at a price.
For the main part, the remaining portions of these notes will use the most
common form of phased array probe (the linear array) to describe the
various aspects of phased array technology.
Figures 3.6a/b show the amplitude profile of a beam axis from a 12mm
diameter 7.5MHz flat probe with respect to distance from the front face of
the probe.
3-4
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
The beam is modelled transmitting into water (V=1500m/s) and the axial plot
shows the beam from 0-540mm (3 near fields). It should be noted that the
3dB calculation has been activated.
3-5
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Near zone.
Beam diameter.
Focal zone.
Beam spread and half angle.
Equation 1
D2 f
N
4v
Equation 1a
D2
N
4
Where:
N = Near field distance.
D = Element diameter (active aperture).
f = Frequency.
v = Material sound velocity.
= Wavelength.
Most work in ultrasonic NDT is done near or just beyond the Near Field. The
range forward or back of the near zone is also of interest and the working
3-6
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
field can be defined by the distances where the maximum pressure drops to
half (-6dB in pulse-echo). The working field, beginning and end of the focal
zone and the Probe diameter are shown in figure 3.6 to extend from 119-
362mm.
Equation 2
1.02Fv
DB 6 dB
fD
Equation 2a
DB 6 dB 0.2568DS F
Where:
DB-6dB = Beam diameter (at the 6dB drop boundary).
F = Focal length.
v = Material sound velocity.
f = Frequency.
D = Element diameter.
SF = Normalized focal length (see Equation 6).
Equarion 3
1 02.5S
FZ N SF 2 [ [
F
]
3-7
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Where:
FZ = Focal zone.
N = Near field.
SF = Normalized focal length (see Equation 6).
For flat circular transducers like that used in the calculations as shown in
figure 3.7, the -6dB pulse-echo beam spread angle is given by Equation 4:
Equation 4
0.514v
Sin( )
2 fD
3-8
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Where:
/2 = Half angle spread between -6dB points.
0.514 is a constant used for the -6dB envelope (for the -20dB envelope the
constant is changed to 0.87).
v = Material sound velocity.
f = Frequency.
D = Element diameter.
It can be seen from this equation that beam spread from a transducer can
be reduced by selecting a transducer with a higher frequency or a larger
element diameter or both.
1 Unfocused (flat).
2 Spherically (spot) focused.
3 Cylindrically (line) focused.
By definition, the focal length of a transducer is the distance from the face of
the transducer to the point in the sound field where the signal with the
maximum amplitude is located.
Transducers with focal lengths beyond these maximums but less than the
near field are called weakly focused. In other words, there may not be an
advantage to a focused transducer over that of a flat transducer.
3-9
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
MP
WP
F
Equation 5
vm
Wp F M d ( )
vw
3-10
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Where:
Wp = Water path.
Md = Material depth.
F = Focal length in water.
vm = Sound velocity in the test material.
vw = Sound velocity in water.
In addition, the curvature of surface of the test piece can affect focusing.
Depending on whether the entry surface is concave or convex, the sound
beam may converge more rapidly than it would in a flat sample or it may
spread and actually defocus.
Figure 3.9 illustrates the relative increase in signal amplitude from small
defects due to focusing where SF is the normalized focal length and is given
by Equation 6.
Note: The amplitude from a small defect cannot exceed the echo amplitude
from a flat plate.
3-11
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Equation 6
SF =F/N
Where:
SF = Normalized focal length.
F = Focal length.
N = Near field.
For example, the chart can be used to determine the increase in on-axis
pulse-echo sensitivity of a 2.25MHz, 25mm element diameter transducer
that is focused at 100mm. The near field length of this transducer is 234mm
and the normalized focal length is (100/234) = 0.42.
From the chart it can be seen that this will result in an increase in sensitivity
of approximately 21dB. Focusing gain in (dB) for cylindrical focused probes
can be estimated as being 3/4 of the gain for spherical focuses.
The concepts covered above have equivalents in phased array sound field
calculations;
What will the near zone be for a focal law that forms a pulse into steel at
zero degrees (ie direct flat contact) with 10 adjacent elements fired
simultaneously?
Dr. I.N. Ermolov assures us that the near zone calculations for a rectangular
probe can use the same equations as a circular disc transducer if the ratio
of length to width does not exceed 2:1. Therefore we can use the applicable
linear dimension for our estimations.
This is effectively a 10mm x 10mm 5MHz probe in contact with steel (Vel. =
5900m/s)
3-12
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Equation 1
D2 f
N
4v or
Equation 1a
D2
N
4
N = 102 x (5x106/(4x5.9x106))
= 21.2mm
This is far too close for our purposes if we want to focus at 50mm so we
could use 20 elements and push the near zone out to 84.8mm. There the
focal spot would be about 5.4mm diameter (as per equation 2).
Then we could use the focusing delays to pull the focal spot back to 50mm
where we would have a focal spot in steel of about 2.7mm (using
equation 6).
There are limits to the extents to which we can carry the comparisons
between single element and phased array probe elements. As Ermolov
pointed out, the approximation of near zone treatment is limited to
rectangular dimensions not exceeding 2:1.
Even then we should look at the very first approximation made for near zone
distance. Equation 1 stated N =D2 f/4v
Equation 1b
D 2 2
N
4
Equation 1 was made assuming that the dimension of the element was
2
much greater than the wavelength of the sound pulse so the term was
dropped. When the individual element is on the order of 1mm this is very
nearly the same size as the wavelength and the off axis effects cannot be
ignored.
3-13
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
In fact some idea of the degree of this can be determined from the
estimation of the near zone of a single element. Inserting values typical of a
phased array probe in equation 1b (eg for a 1mm element dimension
radiating into steel at 5900m/s with a 7.5MHz nominal frequency) the near
zone is less than the wavelength of the pulse. For this example the
wavelength is 0.79mm and the near zone would be 0.12mm instead of the
estimation of 0.32 using Equation 1a.
As we use more and more elements to form our beam the energy is
concentrated on the axis of the beam formed. The rate at which this occurs
is greater for large elements than for small elements (at the same
frequency).
This limits the amount that a phased array can be steered off axis. The off-
axis amplitude available from the individual elements decreases at the same
rate as the individual elements making up the array.
These details are useful to an operator in a general way and the details of
the calculations are left to the probe manufacturer. Typically a manufacturer
may provide a recommended steering range. The operator that fails to
follow the manufacturer’s advice does so at the risk of poor quality
resolution and annoying signals that result from aberrant interference
patterns at the higher steering angles. Refer to 3.5.2.
3-14
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
In 1650 Pierre de Fermat formulated a principle for light paths which we now
use for sound paths in phased array technology. It states that a ray
travelling from one point to another will follow a path such that, compared
with nearby paths, the time required is a minimum.
In phased array ultrasonics, the operator must first determine the point
where the focal depth is to occur, the inspection angle(s) and/or couplant (or
wedge material), plus how many and which elements are to be fired. The
operator also must know details on the array and wedge (if used).
3-15
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
The central ray following Snell’s Law locates the array with respect to the
point of focus and then calculations are made for each raypath from the
elements used to the point of focus, as illustrated in figure 3.10.
Delay times between elements fired are then calculated for each element to
ensure that the time to the point of focus is the minimum, compensating for
velocities in the coupling and test materials.
The calculator produces a file called a focal law which defines the elements
to be fired, time delays and voltages, for both the transmitter and receiver
functions.
3-16
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
This is usually an ASCII file and can be edited and e-mailed as required.
The file is then typically transferred to the computer programme that controls
the pulser-receiver hardware that drives the phased array probe.
The focal laws can be very complex and when several are grouped together
electronic scanning is accomplished ie the movement of the beam without
physical movement of the probe. Electronic scanning options are linear,
sectorial and depth focused as shown in figure 3.14.
3-17
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
a For linear scans, arrays are multiplexed using the same Focal Law.
b For sectorial scans, the same elements are used but the Focal Laws are
changed.
c For dynamic depth focusing (DDF), both the transmitter and receiver
Focal Laws can be changed to optimise pressure and response at a
specific depth.
When combined with a motorised scanner that moves the phased array
probe in a specified path the entire volume of the test piece can be
interrogated.
F
d st
A
Where A is the active aperture, F is the focal length and is the wavelength.
In this case the A is equivalent to the diameter D for a single crystal.
3-18
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
st 0.5
e
Element pitch.
Frequency.
Element width.
Wedge angle.
Number of active elements.
Focal range.
3-19
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
Figure 3.16 illustrates the effect of sectorial scanning. This pattern can be
considered to be similar to the old spinning-head probes used in the
production of medical B-scans where a small angular window is left open for
the probe to transmit and receive through.
3-20
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
The block imaged has a series of side drilled holes that are drilled at
increasing depths as the beam sweeps from left to right. The side-drilled
holes in the block imaged in figure 3.16 are overlaid with a scaled-down
transparency to illustrate the sweep motion.
It will be noted that the strong (red) horizontal signal at the bottom of the
scan image does not match the bottom surface when overlain on the block.
This is merely a projection distortion and the overlay is for illustrative rather
than measurement purposes.
Focusing + steering.
Linear scanning + steering.
3.5.6 Probe performance and selection
Basic considerations for phased array probe selection are identical to those
considered when selecting a single element probe. The material tested its
geometry (shape and thickness) the location and orientation of possible
flaws to be detected are the main considerations. For example:
3-21
www.twitraining.com
Rev 3 January 2010
Beam Forming
Copyright TWI Ltd 2013
In addition to these basic items that are common to both single-element and
phased array probes, the pitch and total number of elements in a probe are
also factors to consider. This makes some aspects of phased array probe
selection a customised process.
The total steering angles required will dictate the pitch of the elements and
this may be limited by the frequency used. The total number of elements
used to provide an effective focus or penetrating capability may be limited
by the instrumentation (more pulsers and receivers are required to address
more elements at a single firing of a focal law).
These aspects of phased array probe selection will be looked at again in the
applications section.
3-22
www.twitraining.com
Section 4
The first systems described in the medical history presented by Dr. Woo
were from the 1960s. At that time industrial ultrasonic instruments were
entirely analogue based. In fact most systems then still used valves (called
vacuum tubes in North America).
The concept of digital ultimately means that the components are operated in
a binary condition; the item is set either on or off. This opened the way for
computer control of the ultrasonic instrument.
4-1
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Both will limit the higher frequency response for the instrument.
For example, a 32/128 phased array unit would have 32 pulsers and 32
receivers and be capable of multiplexing up to 128 channels or elements. In
any single firing a maximum of 32 elements can be used in a single focal
law. It should be noted that not all instruments define their capability in the
same way. Some 32/128 systems must receive on the same 32 elements
that are used for transmission.
4-2
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
These unused channels could be used for single element probes, eg ToFD
pairs, special tandem or transverse configurations or even just spare
channels for the phased array functions in case one of the other channels
malfunctioned.
Quality of the pulser and the receiver has a great effect on the information
obtainable in ultrasonic testing. Subsequently we will consider some of the
options and their features.
4.3 Pulsers
Essentially all that is required to vibrate a piezoelectric transducer is an
alternating voltage. However, characteristics of the pulse voltage will dictate
how the element vibrates. This is analogous to pushing a person on a
swing. If pushes are applied at the natural frequency of the loaded swing,
large amplitudes can be achieved. If not, a rough and low amplitude ride
results.
Three pulse shapes are commonly used in ultrasonic flaw detection units;
spiked, bipolar tone burst and square wave. These are illustrated in figure
4.1.
4-3
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Spike pulse
Thyristor
(switch)
DC power Transducer
supply
Charging
resistor
When the capacitor discharges, the rapidly applied voltage across the
transducer causes it to vibrate. The purpose of the damping resistor
(inductor) is used to increase the rate of voltage decay. This is the damping
available to the operator and is used to control the ring-down time.
4-4
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Sine wave
Saw-tooth wave
Tone burst signals usually consist of several cycles. These are preferred for
velocity determination using interferometry. As well, since very high
frequencies can be derived using tone burst pulsing it is used in acoustic
microscopy where frequencies in the giga hertz range are used.
Adjustments of pulse voltage and pulse width are possible thereby making
square wave pulsers a most versatile tool to optimise transducer
performance. By choosing the best pulse width to obtain constructive
interference, less voltage need be applied to the probe thereby reducing
noise level. By pulsing at a frequency higher than that for maximum output,
bandwidth can be increased and lower frequency components reduced.
Pulsing a transducer at a frequency lower than that for maximum amplitude
increases mechanical damping and provides a sharper cleaner signal with
little ring-on. These features of the square wave pulser are illustrated in
figure 4.4.
4-5
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
In figure 4.4 the square wave pulse shape is shown on the left. In the first
case the pulse is set to a fairly short duration, 12.375 nanoseconds (ns) and
the applied voltage is -498 volts.
In the second case in figure 4.4, the pulse width has been adjusted to
provide a maximum output from the transducer at 25.15ns. Voltage applied
is -547 volts, up slightly from the first case (in the first case voltage was the
maximum possible for the pulse width applied).
4-6
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
In the third case the pulse has been increased to 51ns and the voltage is
essentially the same as applied in the second case. Transducer output is
reduced and the ring-on is virtually eliminated. It is therefore possible to
increase the damping of the transducer’s vibration by increasing the pulse
length beyond the resonant frequency.
Then reversing the voltage and allowing it to swing back through zero volts
to the same voltage maximum but the opposite sign and then brining the
voltage back to zero after holding to the maximum for another half-cycle
time. This would be comparable of a single cycle in the bi-directional square
wave pulser in figure 4.3.
When displayed visually the effects of pulse width and pulse tuning are
more dramatic. Figure 4.5 is an image of a pulse from a 12.5mm diameter
7.5MHz probe with a spherical radius of curvature of 150mm.
The pulse used to obtain this image was from a Spike pulser (no pulse-
length tuning possible) and the voltage applied to the element to obtain the
visible image was 700V.
Figure 4.6 however has a similar intensity at the focal spot (light intensity is
proportional to particle displacement) but uses much less than half the
applied voltage (180V).
4-7
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Figure 4.6 Phased array (16 elements) with tuned pulse at 180 volts.
4.7 Receivers
According to Krautkramer (Ultrasonic Testing of Materials text), pulsers
apply voltages of 100-1000 volts to the probe. However, received signals
are three to four orders of magnitude smaller (a few milli-volts to a few volts
(0.001-1V). This causes a couple of problems. One is the shock of the pulse
voltage that is transferred to the receiver in pulse-echo mode of operation.
The other problem is the need to amplify the relatively small signal from
flaws without amplifying noise. The latter is further complicated because the
frequency of the received signal may not be the same as the transmitted
pulse envelope (accounting for even smaller signals from the transducer).
When switched from pulse-echo to transmit-receive, there is no longer a
physical electric connection between the two components. Figure 4.7
illustrates this switching.
4-8
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
First stage is the circuit protection that protects the preamplifier from the
pulser voltage when in pulse-echo mode. The preamplifier can use
transistor type amplifiers that provide about 20-40 dB of gain, frequency
response of the preamplifier is usually broadband.
4-9
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Some portable instruments have been made that use very large
preamplifiers. As a result, even with maximum attenuation, signals from
normal beam inspections of plate could not be reduced to below full screen
height.
4-10
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Relative
Amplitude
2 4 6 8 10
Frequency (MHz)
4.8 Gates
Essential to computer imaging of ultrasonic data is the ability to extract
information from regions of time that can be selected to monitor for signals
in that time. The region being monitored is said to be gated.
Time along the gated region or amplitude within the gate, when a signal
occurs or both time and amplitude can be gated. Alarm or recording
thresholds can be set for signals occurring in the gate. Gates are an
essential component in automating inspection systems.
Figure 4.9 shows a digital A-scan display with three gates available. Data
collection options for gated regions may include time, amplitude and
waveform.
4-11
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
4.9 TCG/TVG
When amplification is variable with respect to time it provides time corrected
gain (TCG), this is also called time variable gain (TVG) or swept gain. By
allowing more amplification to be added as time or distance increases,
signals from reflectors of the same surface area can be adjusted to the
same amplitude at any distance.
Figure 4.10 illustrates the responses of a side drilled hole being set to the
same amplitude using TCG. The TCG amplification points are seen on a
trace at the bottom of the A-scan.
4-12
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
TCG points
4-13
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Speed.
Consistency.
Accuracy.
Repeatability.
Safety.
Cost.
Capability.
4.10.1 Speed
Automated systems can be arranged to inspect many thousands of parts
per day. For example, in the automotive industry, ECT on bolts can be
performed at 25000 parts/hr. Manual operations would be no match for such
rates.
4.10.2 Consistency
Automated systems should operate under as controlled a condition as is
possible thereby strict objective limits can be selected which would not be
possible where the somewhat more subjective human eye is concerned. For
example, amplitude response measurements of many reflectors in a test
piece.
4.10.4 Repeatability
Due to the control on conditions and the precision of measurement,
inspection results are very repeatable. For example during periodic in-
service eddy current inspections of tubing, location of indications is typically
within 1mm of the test from 3 years previous.
4.10.5 Safety
One of the biggest advantages in using automated systems comes from a
safety point of view. Today remote inspection systems are used in
numerous forms of hostile environments.
4-14
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
For example nuclear - inside reactors and components where testing could
be done manually if no fields existed. Other hostile conditions for human
operators include: extremely low or high temperatures, deep water and
caustic atmospheres.
4.10.6 Cost
Although automated equipment can have some large associated setup
costs, overall inspection cost can be reduced as a result of automated
systems in spite of the large sums spent on the system. For example
increased speed over manual operation saves man hours.
4.10.7 Capability
Some inspections are only possible using automated equipment. For
example in-service volumetric inspections of fuel channels in CANDU
nuclear reactors.
The sensor is part of the NDT equipment. NDT equipment functions are
familiar to any experienced NDT technician. An important feature for an
automated system is how to utilise the signal generated by the instrument.
4-15
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
4-16
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Printer
PC with
A/D board and
motor control
Probe
Test Piece
Turntable
Y slide
If the digital signal amplitude intervals are small and the time intervals are
also small the resultant digital waveform can closely approximate the
analogue waveform. The difference in signal quality between analogue and
digital is demonstrated in figure 4.12.
4-17
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
B
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
C
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0 7 9 10 11 12 8 5 7 7 6 7 11 14 9 5 7 7 5 6 6 4 6 9 9 5 4 8 6 4 6 10 4 3 7 13 5 3 7 5 5
Figure 4.12 shows the steps in converting the continuous analogue signal to
a digital signal by a computer.
Frame A shows the raw input as continuously varying amplitude.
Frame B shows how the conversion must assign an off-set that will ensure
the maximum negative displacement is above the lowest values. The
vertical division of the signals shows from minimum to maximum there are
16 levels. Vertical divisions are multiples of 2 with 256 divisions being
common.
The number of bits that each bar indicates is easily converted to binary code
and read by the computer, eg the bar indicating 3 vertical bits would be read
as 0011, 4 bits is 0100, 7 bits is 0111.
4-18
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
The digital signal that results from this conversion is composed of digital
values of a known range termed the scale factor and these values are
separated by a fixed time interval termed the sampling interval.
The reverse process is also used, ie. converting digital signals to analogue
signals. The device that accomplishes this is called a digital-to analogue
converter (DAC).
Test set-ups may include both digital and analogue equipment. Where
several analogue input or output channels are used in a test system they
are often added to an ADC or DAC through a multiplexer (MUX).
Figure 4.13 Multiplexed inputs and outputs via ADCs and DACs.
4-19
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Today, the differences are not so clearly defined. Except for situations
where the amount of data to be collected is large and must be processed
quickly in addition to performing many other functions (multitasking), most
inspection systems can be automated with some form of a personal
computer.
4-20
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
probe
RF
flaw
Rectified
Scan
Display
Figure 4.14 A-scan display types.
4.14.2 B-scans
When motion is added to the display several options exist. When time is
displayed along one axis and probe position, as moved over the test
surface, along the other, a B-scan is generated.
These concepts are shown in figure 4.15. When a hard copy is made it may
be visualised as a series of A-scans stood on end and stacked one beside
the other.
4-21
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Main Bang
probe motion
Time flaw
Scan Display
4.14.3 C-scans
When maximum amplitude is collected in the gated region and a raster scan
performed, a C-scan results. In this case the probe position is plotted along
both plot axes. This is effectively in a plan view, see figure 4.16.
}
Gated region
Display
X
Y
Scan
4-22
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
When the reference co-ordinates are the surface of the test piece and a
normal beam is used for inspection, nomenclature is straightforward. When
angle beam inspections are used the loss of orthogonal relationship
between beam and test surface can cause some confusion.
Collecting a B-scan for a beam inclined 45° to the test surface is not
significantly different from the normal beam but the operator may have
difficulty relating position of probe to indication as they no longer coincide,
see figure 4.17.
0 25mm
1 2 3 4 5 6
Interface signal
flaw signal
4-23
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
For the example above, if the refracted angle is 45° and the soundpath to
the peaked signal is 10mm, the off-set can be calculated by multiplying the
soundpath by the sine of the refracted angle.
Some data acquisition programmes can correct the display but this requires
every point along each A-scan be multiplied by the appropriate factor prior
to display and the data is then shifted horizontally.
When two probes are used in the T-R mode it is still possible to collect
signals over a period of time and plot the results against probe movement.
This is the principle of ToFD presentations.
We have amplitude information for the gated time for each point on a grid
and since time is equivalent to distance we have amplitude information for
every point in the volume inspected. Although this is very memory intensive
(as we will soon see), it provides sufficient information that tomographic
visualisation techniques can be applied.
4-24
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
A vertical slice was taken through the position at 9.1mm and a horizontal
slice was taken at 8.9 mm and these cross sections are displayed to the left
and above the C-scan image. The distances to the lower 4 targets in the
vertical line are clearly seen to vary in the vertical cross section while they
are fairly consistent in the horizontal.
If a weld was inspected with an angle beam and this data collected by a
suitable data acquisition system, a slice along the weld axis could be
extracted from the data. Examination of the parameters would show there is
only one point in time on each applicable X-Y co-ordinate (see figure 4.19)
and that must be corrected for angle and assigned an equivalent depth. This
is termed by some a D-scan.
4-25
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
This is the preferred terminology for some users of ToFD when the scan
direction moves parallel to the weld axis (note such terminology requires a
weld axis as a reference). The same users of the D-scan terminology then
reserve the term B-scan for the scan made with the ToFD probes moving
perpendicular to the weld axis.
4.14.7 S-scans
Phased array provides a new opportunity for scan presentations due to the
potential for a dynamic nature of the beam. In addition to the standard
presentations the variable angle afforded by the phased array provides us
with the Sectorial or S-scan.
4-26
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
If a 100 MHz ADC is used, sampling occurs every 0.01µs. In pulse-echo this
provides a resolution in steel of 0.016mm (shear) and 0.03 mm (long). The
temporal resolution also dictates the quality of signal reproduced from the
analogue.
4-27
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
A minimum sampling rate of four times the nominal frequency of the probe
used is recommended. This will ensure the digitised amplitude will be within
3 dB of the analogue value.
Five times the nominal probe frequency is required if the digitised sample is
to be within less than 1 dB of the analogue signal amplitude. For example,
for a 10 MHz probe, an ADC rate of at least 50 MHz is recommended for
amplitude critical work.
It will be seen by the operators that the quality of the recorded signal at a
higher ADC rate is much closer to the original (analogue) and makes for
improved signal characterisation. When the signal is rectified and filtered
before digitisation, it may be possible to reduce the ADC rate.
The term bit rate is derived from binary treatment of data whereby there are
8 bits to a byte in computer terminology. Here a bit is one of two options, ie
the values 0 or 1. When the binary value (or 2) is raised to the power of
eight (8) it is considered 8-bit. If it was raised to the power of 10 it would be
10-Bit.
4-28
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
For example:
0-256
For a rectified signal in an 8Bit ADC unit the voltage bias places the zero
point at the bottom and shifts all points positive so dynamic range is 1/256
or 20 log10 (1/256) = -48dB. The smallest % interval on the screen is
(1/256) x100=0.39%.
For the RF signal the same 8 Bit ADC has no bias and signals are positive
and negative. The dynamic range is determined from the zero point to the
maximum displacement (128). 1/128 or 20 log10 (1/128) = 42dB.
4-29
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Large amplitude signals that are 100% or greater can, of course, not be
assigned a real value. They are simply considered saturating. This applies
to analogue or digital displays.
When amplitude is an important factor and its absolute value is required this
means that a re-scan is required to assess the actual amplitude with respect
to the reference level. In an 8-Bit digitised rectified signal once the signal
has reached the 256 level it is saturated. A signal greater than 256 levels
may be 101% or it may be >500% with respect to the full scale display.
A 10-bit digitisation rate would then have the vertical range of any signal
divided into 1024 equal intervals. This would allow us to collect signals at a
lower receiver gain and electronically add gain after the data had been
collected.
1024 amplitude levels have four times the resolution of the 8-bit systems.
That means we could calibrate at a reference level of 20% (instead of the
typical 80% on an 8-bit system) and collect all the A-scans at lower
amplitude.
Signals on our new 10-bit display reaching 25% screen height would have
been 100% on the 8-bit display. The likelihood of troublesome saturating
signals using the 10-bit digitisation would therefore be greatly reduced.
Using the same assessment of dynamic range as for the 8-bit system, the
10-bit system is seen to have a dynamic range of 60dB for rectified signals.
There are two methods that this extra dynamic range can be used. The first
uses the standard display whereby the 100% level on the display is the
maximum displacement of the signal amplitude. Then setting the reference
level to a lower point on the amplitude scale allows a direct reading off the
scale for the amplitude.
The second method is shown in figure 4.23 where the gates are used to
measure the amplitude and even though the display no longer shows an
increase in signal level, the operator can read the measure amplitude as a
digital numeric output.
In the example the digitisation is a 9-bit and the gate output of amplitude
shows that the signal on the extreme left is indicated as having a 200%
amplitude and occurs at 26.55mm.
The next signal has a separate gate (green) and the signal also saturates
the display but not the gate level. That signal is indicated as having
amplitude of 176% and occurring at 39.03mm.
4-30
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
For a simple B-scan using a 4 MHz contact normal beam probe on a 50mm
thick plate we would like to gate the entire thickness for display. We would
use the recommended minimum ADC rate of 16 MHz.
At 16 MHz ADC 16 samples are made each µs, so for the gated time of
16.9µs, 270 samples will be recorded for each A-scan. At each point 8 bits
of amplitude information are collected (8 bits = 1 byte). If our 0° B-scan is to
be collected across a weld and include heat affected zones, a 100mm travel
should suffice (ie. 50mm either side to the weld centreline).
4-31
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
To generate the full volume scans for depth encoded C-scans and D, E and
/or P scans would require several such scans to be made in a single
process. Even if a small square 100 x 100mm was scanned with a 1mm
raster step with the above conditions, 5.4 MBytes would be generated.
Caution should be taken to collect only that amount of information that is
absolutely necessary.
4-32
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Digital signal processing can be generally grouped into two categories; one
dimensional and two dimensional. One dimensional processing is applied to
the captured waveform and may be either filtering or spectrum analysis.
Two dimensional processing is concerned with enhancing spatial structures
of the image.
One dimensional DSP has been touched on briefly. Use of fast Fourier
transforms signal averaging for increased signal to noise ratio. When noise
is known to originate at a higher or lower frequency than the pertinent UT
signal a bandpass filtering process can be applied. This selectively removes
spurious components from the A-scan. Figure 4.24 illustrates such a
process.
4-33
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
indication
indication
Processed
Data Enhanced signal
4-34
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
When the equation of this hyperbola is known, A-scan signals can be shifted
in time and added together. When a defect is present constructive
interference of the waveforms form a large signal.
4-35
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
In figure 4.26 three 1mm notches were scanned. The image on the left is the
raw data as it would appear in a B-scan and the image on the right is the
result of SAFT corrections to improve lateral resolution.
A similar processing can be done on phased array scan results. Figure 4.27
illustrates a before and after condition where an S-scan was used to collect
the raw data. DSP algorithms are used to correct for the angular
displacement that would occur.
Data would be collected over a linear time based display (lower linear
uncorrected image) and then the fact that the probe was not moved is used
to calculate the flaw locations on the arc generated by the S-scan focal
laws.
The correction places the sound origin at a single point and displays the
targets (only 4 were processed in the upper image) as they would appear
relative to the probe position.
4-36
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Correcte
d S-scan
Linear
view
When parts are moved past a probe the relative position is rarely recorded
precisely. Tube inspection stations are often equipped with strip-chart
recorders. Feed-speed and position of the indication on the chart can be
used to locate the indication.
In pipe mills audio alarms and paint sprayer markers alert the operator to
when and where an echo breaks threshold. The spray maker is located
down stream of the probes and its operation is delayed from the time of the
alarm based on the travel speed of the pipe past the probes.
4-37
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
When probe movement is made over a fixed object, several options exist for
mechanisation. In order to fix some sort of reference position, a probe
holder and associated framework is required.
Position may be provided by some form of encoder and the probe moved by
hand. Alternatively, movement may be facilitated by motors on the
framework and again, encoders may provide positional information.
On long scan gantries the gantry support may set up vibrations and shake
the probe so that coupling path or coupling quality is reduced or the scanner
may simply shake some components loose. The risk of damage by
something as simple as a small speck of weld spatter may be greater at
higher scan speeds than at lower speeds.
Today, the phased array systems can scan a 12 zone weld and collect full
waveform scans for ToFD, 6 thickness channels, through transmission
coupling channels and 8 full waveform B-scans (or D-scans if you prefer).
This can now be done at more than double the speed of the older systems
(now about 80-100mm/s). This is all the more impressive when the file size
is considered.
File sizes of the older systems were on the order of 100kB and provided
only amplitude and time information, any projection scans (B-scans) were
only images so were given the term mappings as they did not preserve the
waveform and no ToFD was being provided.
The phased array system typically collects over 1,000 times more data (10-
15MB) at twice the speed.
4-38
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Above is a magnetic crawler with probe-holder motion left and right as the
magnetic wheels move the entire carriage forward or back (application on a
pressure vessel).
Above left is a lab scanning rig with X, Y, Z & motion. Above right is a
phased array pipeline girth weld inspection rig (placed on pipe).
Figure 4.28 Scanning equipment with positional encoders, all images courtesy
RD-Tech.
4-39
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
However, computer CPU is not the only limiting factor. Even if computers
are not used for data acquisition, another limiting factor is the pulse
repetition frequency.
Static calibration may indicate a gain setting to achieve the required signal
amplitude but when a dynamic run is made over the calibration at too high a
speed, the amplitude recorded will be something less than that for the static
calibration.
PRF
Vc Wc
3
In a system where many probes are sequenced via a multiplexer the PRF is
divided amongst the total number of probes. Although many units have
PRF’s of 2kHz, when 10 probes are used in the system, the effective PRF at
each probe is only 200 Hz.
4-40
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
For example in pipeline girth weld inspections using a pair of phased array
probes many functions are carried out during the scan. The weld is divided
vertically into zones with a beam directed at each zone (coverage is
symmetric either side of the weld).
A simple strip-chart style of data displays amplitude and time in a gate for
each zone as well as several channels dedicated to collecting the full pulse-
echo waveforms and ToFD full waveforms.
As the scanner is advanced by the motor (controlled via a motor control unit
and a communication link via the motor control card on the controlling
computer) it causes the encoder to turn and the pulses generated indicate a
specified number of pulses per unit distance.
As the encoder indicates the start of the 1mm interval all functions in the
sequence commence:
4-41
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Many more small checks and functions are carried out but the overall effect
this tries to convey is that many functions are occurring based on the
initiating pulse from the encoder. Some systems have scan speeds around
100mm/sec. This means that the sequence of events that are required to
carry out all the steps in a single millimetre are repeated 100 times each
second.
There is still a limit to what a computer can do in a short period of time and if
the scanning speed is too fast to complete all the functions required in that
1mm interval then all the information for that step is lost. This is seen as a
black line on the display. On the C-scan the same effect (missing data seen
as black lines or spots) can be seen when scanning speed is too high as
shown in Figure 4.29.
The equation:
PRF
Vc Wc
3
Instead, the fact that each firing of the pulser is dictated by the encoded
distance-interval means that the operator needs to determine the dimension
of the 6dB beam width (Wc) and ensure that the encoded sample interval is
less than 1/3 that distance to conform to the intend to the specification
requirement.
4.20 Encoders
A rotary optical encoder is a sensor that uses light to sense the speed,
angle and direction of a rotary shaft. A linear encoder reads a linear strip
instead of a disk to provide the same information for linear motion.
4-42
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Optical encoders use light instead of contacts to detect position, so they are
inherently free from contact wear and the digital outputs are bounceless (no
contact bounce). Accuracy of an optical encoder is as good as the code
wheel.
The code wheel patterns are created using precision digital plotters and cut
using either a punching system or a laser each guided by closed loop
precision vision systems.
The light source used for encoders is usually a point source LED rather than
a conventional LED or filament. Most optical encoders are transmissive type
meaning that the light is collimated light into parallel light rays and passes
through the disk (or strip) pattern.
The image of the pattern is detected using a phased array monolithic sensor
and converted to TTL digital quadrature outputs.
Most incremental encoders have a second set of pulses that is offset (out of
phase) from the first set of pulses and a single pulse that indicates each
time the encoder wheel has made one complete revolution.
If the A pulse occurs before the B pulse, the shaft is turning clockwise and if
the B pulse occurs before the A pulse, the shaft is turning counterclockwise.
The C pulse occurs once per revolution.
4-43
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
When the encoder position pulses are interlaced with the ultrasonic pulses
in such a system the ultrasonic PRF and position pulses are said to be
asynchronous. This is illustrated in the upper portion of Figure 4.32.
Synchronising the UT pulsing with the position pulses ensures all channels
are fired in the sample interval. The only limit is the computers put-through
rate.
4-44
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
4-45
www.twitraining.com
Rev 3 January 2010
Hardware: Pulsers, Receivers, Motor Control and Encoders
Copyright TWI Ltd 2013
Wc * prf
V
3
4-46
www.twitraining.com
Section 5
These usually include a normal beam for checking the test piece for
possible laminations that may interfere with the angle beam paths, plus one
or more fixed angles directed at the weld to detect welding or service flaws.
Phased array inspections are not exactly new in the NDT industry; however,
the rules regulating their use have not been well maintained. It would seem
that it has been assumed that phased-array applications could simply use
the existing guidelines for manual ultrasonic testing. Even the simplest
requirements of regulating codes are often not possible to conform to when
using the superior phased array technology.
This chapter deals with some of the more general aspects of weld
inspection. As issues with differences between the traditional and phased
array requirements arise they will be pointed out and commented on.
The operator may select a probe with a specific frequency using the same
sort of rationale as would be used for single element applications. However,
after that the probe selection and information required to use it becomes
much more complex.
5-1
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
This requires that the operator identify the position (in 3-dimensional space)
of the probe elements. This ensures that the path lengths to the wedge-steel
interface are accurately known.
Verification that the coordinates used by the operator provide correct depth
calculations are necessary. This ensures that the display software correctly
positions indications detected.
Parameter settings for the probe and wedge are usually set using a user-
entry menu field such as illustrated in figure 5.1.
Note: The requirement for the total number of elements and the pitch
dimension in the probe entry portion!
Include in the information for the wedge are the wedge angle, velocity and
the distance above the test piece and the distance from some reference on
the wedge (back of wedge or nose of wedge) is available to the software
calculating the Fermat equations. The reference positions are shown in
figure 5.2.
5-2
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Compensation for attenuation variations and delay times may be made one
focal law at a time or software can be configured to make the
compensations dynamically using software wizards.
The operator would need to enter the appropriate information about the
probe and focal laws used for the selected probe including the velocity of
the part (in this case the compression wave velocity of the IIW block would
be approximately 5900m/s).
Without a wizard the distance to the first backwall echo would need to be set
for each focal law. For the 25mm thickness direction using IIW block this
would be 25mm. If a wizard was available the delays to correct travel time,
in µs to mm soundpath (also called half-path) then all the focal laws could
be configured in a single step.
This would involve identifying the depths to the first and second backwall
echoes for all the focal laws in the E-scan as indicated in figure 5.3.
5-3
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
In all cases in figure 5.4 the probe is moved back and forth and one of the
focal laws is seen to peak when its beam makes a perpendicular incidence
on the radius.
Since the material velocity and the radius are assumed known values, this
provides a time in steel that can be calculated and subtracted from the total
time from the clock start.
The remaining time is the wedge time and this is subtracted from the total
time to provide the zero depth or entry surface.
5-4
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Maximum
amplitude from
Del ay adju stme nt m ovem ent for S-sca n with refra cting w edge radius corresponds
to Focal Law that
aligns with centre
of 100mm radius.
Delay a dju stme ntmov eme nt fo r ele c tronic sca n wit h refrac ting wed ge
Figure 5.4 Setting range for S-scans and angled E-scan Focal Laws.
This process of moving the phased-array probe back and forth on the V1
block is identical to the process used for a single element probe when
determining the exit point of the probe/wedge.
Since all phased-array systems can allow the user to monitor the responses
of a single focal law using a single A-scan, this process can also be used to
illustrate the migration of the exit point series of focal laws being used.
Although the operator could mark the exit point on the probe and use that to
confirm a specific angle (using the same procedure as used for single
element probes) this is not usually a function performed for phased array
systems.
5-5
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Since there is no single exit point none is marked on the phased array
wedge so verification of probe exit point is one of the items that do not fit
with the old technology.
With the old manual methods the operator relied on knowing the exact angle
and exact distance travelled so as to manually plot on paper the position of
any indications. Any errors in actual angle could result in errors of flaw
placement.
With phased array systems the plotting is done using the computing
capability of the software. Its accuracy is confirmed using the targets in a
known calibration/reference block.
Flaw orientation with respect to beam, flaw size, beam area at the flaw,
frequency response of the flaw with respect to the frequency content of the
pulse and several other parameters can play a part in the detection of a
flaw.
Flat bottom holes (FBHs) and side drilled holes (SDHs) are easiest to
manufacture and most adaptable to general conditions. Surface notches are
also used but very angular dependent.
The notches, FBHs and SDHs all provide options to assess the sensitivity of
the beam at various soundpaths. The distance-gain-amplitude (DGS or in
German AVG) method is another option and was designed to avoid the
need for costly calibration blocks.
5-6
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
The probe is acoustically coupled to the block with a side drilled hole at a
known depth. The 1.5mm diameter SDH in the IIW block is a convenient
target for this purpose.
Select the A-scan display for the first focal law configured and move the
probe forward and backward to locate the maximum amplitude signal from
the SDH. Adjust the response from the SDH to set amplitude (for example
80% FSH) and save the parameters in the focal law file.
Repeat the process of maximizing the signal from the SDH and setting it to
80% FSH for each focal law and saving the set-up file after each focal law is
completed.
5-7
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
For S-scans where 1D linear array probes are used, a single SDH results in
a changing steel path for each angle. This makes it unsuitable for this task.
Use of the radius for S-scan configurations also provides correction for
echo-transmittance effects intrinsic in angle variation.
5-8
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Optimising the response from each focal law is best accomplished using a
wizard that automatically adjusts the gain to each focal law so the same
amplitude is achieved from the same target at the same soundpath. Figure
5.7 illustrates the display that could be used for this purpose.
The operator sets the amplitude level that they want to achieve (in this case
80%) and then moving the probe back and forth as indicated in figure 5.6
the peak amplitude from each focal law is seen (as the wavy line) to
approach the target level.
The focal laws (indicated along the horizontal axis) provide peak amplitude
and then the software adds the appropriate gain to each individual focal law
to adjust its amplitude level to the target level (for example 80% in this
case).
This is only likely to occur if the operator elects to attempt making a set of
focal laws that attempts to steer the beam by more than is recommended by
the probe designer.
5-9
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
First adjust the time-base of the A-scan to a suitable range to display the
pulse-echo signals selected for the linearity verifications. A linearity block
similar to that described in ASTM E-317 is selected to provide signals to
assess linearity aspects of the instrument. Such a block is shown in figure
5.8 with a single element probe mounted on it.
Pulser parameters are selected for the frequency and bandpass filter to
optimize the response from the pulse-echo focal law (or a single element
probe may be used) for the linearity verifications. The receiver gain is set to
display non-saturating signals of interest for display height and amplitude
control linearity assessments.
5-10
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Increase the gain using the receiver gain adjustment to obtain 100% of full
screen height of the larger response. The height of the lower response is
recorded at this gain setting as a percentage of full screen height.
Note: For 8 bit digitization systems this value should be 99% as 100% would
provide a saturation signal.
The height of the higher response is then reduced in 10% steps to 10% of
full screen height and the height of the second response is recorded for
each step.
Return the larger signal to 80% to ensure that the smaller signal has not
drifted from its original 40% level due to coupling variation. Repeat the test if
variation of the second signal is greater than 41% or less than 39% FSH.
5-11
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
For an acceptable tolerance, the responses from the two reflectors should
bear a 2-1 relationship to within +/-3% of full screen height throughout the
range 10-100% (99% if 100% is saturation) of full screen height. The results
are recorded on an instrument linearity form.
A 32/128 has 32 pulsers and receivers and can address up to 128 elements.
Each of the pulser-receiver components is checked to determine the
linearity of the instrument amplification capabilities.
To check the receiver amplifiers select a flat (normal incidence) linear array
phased-array probe having at least as many elements as the phased-array
ultrasonic instrument has pulsers.
5-12
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Add gain to the receiver in the increments of 1dB, then 2dB, then 4dB and
then 6dB. Remove the gain added after each increment to ensure that the
signal has returned to 40% display height. Record the actual height of the
signal as a percentage of the display height.
Next, adjust the signal to 100% display height, remove 6dB gain and record
the actual height of the signal as a percentage of the display height.
Signal amplitudes should fall within a range of ±3% of the display height
required in the allowed height range of table 5.1.
Repeat the sequence of gain adjustments from the 40% starting level for all
other pulser-receiver channels. For instruments having 10 or 12 bit
amplitude digitization and configured to read amplitudes in a gated region to
amplitudes greater than can be seen on the display, a larger range of check
points can be used.
For these instruments the gated output instead of the A-scan display would
be verified for linearity. Note: An example of amplitudes greater than 100%
display height is seen in figure 5.11 where gate A% indicates a 200% signal
and gate B% indicates 176%.
5-13
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
The same function can be checked on a phased array instrument and the
display need only be checked using a single channel (since the display
applies to all channels equally).
5-14
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
For steel at 5900m/s each sample along the timebase (10ns) in pulse-echo
mode represents 30µm. A tolerance of ± 3 timing samples should be
achievable by most analogue-to-digital systems. Some allowance should be
made for velocity determination error (~1%). Typically the errors on the
multiples should not exceed ± 0.5mm for a steel plate.
A sample recording table for the linearity checks is indicated in table 5.1.
5-15
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Location: Date:
Operator: Signature:
Instrument: Couplant:
Pulser voltage (V): Pulse Duration (ns): Receiver (band): Receiver smoothing:
Digitization Frequency (MHz): Averaging:
Display Height Linearity Amplitude Control Linearity
Large (%) Small Small Actual Ind. dB Allowed range
Allowed (%) Height
Range
100 47-53 40 +1 44-46
90 42-48 40 +2 48-52
80 40 40 40 +4 62-66
70 32-38 40 +6 77-83
60 27-33 100 -6 47-53
50 22-28-
40 17-23
30 12-18
20 7-13
10 2-13
Amplitude Control Linearity Channel Results: (note any channels that do not fall in the allowed range)
Channel (add more if required for 32 or 64 pulser-receiver units)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
5-16
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
The probe is then acoustically coupled the probe to the 25mm thickness of
an IIW block with a uniform layer of couplant. This may be accomplished by
a contact-gap technique such that the probe-to-block interface is under
water (to ensure uniform coupling).
Set the pulser parameters to optimize the response for the nominal
frequency of the probe array and establish a pulse-echo response from the
block backwall or water path to 80% display height for each element in the
probe.
Observe the A-scan display for each element in the array and record the
receiver gain required to achieve the 80% signal amplitude for each
element. Results may be recorded on a table similar to that in table 5.2.
Note and record any elements that do not provide a backwall or waterpath
signal (inactive elements).
5-17
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Table 5.2 Probe element activity chart: Enter receiver gain for 80% FSH.
Element
Gain
Active (√)
Inactive (x)
This number may be different for baseline and in-service verifications. Some
phased array probes may have several hundred elements and even new
phased-array probes may be found to have inactive elements as a result of
manufacturing difficulties ensuring the electrical connections to elements
with dimensions on the order of a fraction of a millimetre.
No simple rule for the number of inactive elements can be made for all
phased-array probes. Typically, if more than 25% of the elements in a probe
are inactive, sensitivity and steering capabilities may be compromised.
If still outside the acceptable range the probe should be removed from
service and corrected prior to further use. The test using a fixed water path
to a water/steel interface will reduce coupling variations.
5-18
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Prior to removing the probe from service the cable used for the test should
be exchanged with another cable, when possible, to verify that the inactive
elements are not due to a bad cable.
For a single focal law where the beam is fixed and the probe is used in an
immersion setup, the ball-target or hydrophone options described in the
single element characterisation methods such as ASTM E-1065 may be
used.
For phased array probes used where several focal laws are generated to
produce sectorial or electronic scanning, it may be possible to make beam-
profile assessments with no or little mechanical motion.
5-19
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
For phased array probes using focal laws with a large portion of the
available elements to form the beam the number of remaining elements for
the electronic raster may be too small to allow the beam to pass over the
target. In this case it will be necessary to have encoded mechanical motion
and assess each focal law along the active plane separately.
Profile assessments for contact probes can be made using SDHs. The side-
drilled holes are arranged at various depths in a flaw-free sample of the
same material in which focal laws have been programmed for.
Using the linear scan feature of the phased-array system the beam is
passed over the targets at the various depths of interest. The electronic
scan is illustrated schematically in figure 5.13.
5-20
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Time or equivalent distance in the test material is presented along one axis
and distance displaced along the other axis, ie a typical B-scan is used as
illustrated in figure 5.14.
Figure 5.14 B-scan display for beam profiling of side drilled holes.
A similar data display can be made for an electronic scan using a phased-
array probe mounted on a wedge. This would use a simple orthogonal
representation of time versus displacement or it can be angle corrected as
illustrated in figure 5.15.
5-21
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Resolution along the displacement axis will be a function of the step size of
the electronic scan or if the scan uses an encoded mechanical fixture the
resolution will be dependent on the encoder step-size used for sampling.
Resolution along the beam axis will be a function of the digitization rate and
the intervals between the target paths. For highly focused beams it may be
desirable to have small differences between the sound paths to the target
paths (for example 1mm or 2mm).
Beam profiling in the passive plane can also be made. The passive plane in
a linear-array probe is perpendicular to the active plane and refers to the
plane in which no beam steering is possible by phasing effects. Beam
profiling in the passive direction will always require mechanical scanning.
Figure 5.16 illustrates a method for beam assessment in the passive plane.
This technique uses a corner reflection from an end-drilled hole at depths
established by a series of steps.
5-22
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
By presenting the acquired data as a projected C-scan, the beam size can
be assessed based on either colour or greyscale indicating amplitude drop.
The projected C-scan option is schematically represented in figure 5.17b.
Figure 5.7b Combined active and passive axis beam profiling projected C-scan
option.
st 0.5
e
5-23
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
For specific cases and when required to demonstrate the system limits,
users may want to determine the limits for a particular combination of probe
and wedge. Limits on steering are not simply determined by seeing how
many degrees can be detected using the IIW block inserts.
Applications where focusing is necessary may not achieve the same limits
as applications where the beam is not focused as well as steered. Steering
capability may be specific to a sound path distance, aperture and material.
5-24
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
85°
80°
75°
70°
65°
60° 5°10°20°30° 40° 50° 55° 60°
55°
Horizontal p lane 25m m from exit point
50°
45°
40°
35°
30°
Block dim ensions 150m m x 100mm x 25m m
Figure 5-19 Beam steering assessment block – fixed plane sound path.
The block used for assessment would be arranged with side drilled holes in
the plane of interest. Such a plane-specific block is illustrated in figure 5.19
where a series of holes is made in a vertical and horizontal plane at a
specified distance from the nominal exit point. Side drilled holes may be
arranged in other planes (angles) of interest.
For assessments using the block in figure 5.18 place the probe so that the
centre of beam ray enters the block at the indicated centreline. When the
probe is used without a delay line or refracting wedge, the midpoint of the
focal law element array is aligned with the centreline.
For focal laws using only a portion of the total available elements the
midpoint of the element aperture shall be aligned with the centreline. When
delay lines, refracting wedges or immersion methods are used corrections
will be required to compensate for movement of the apparent exit point
along the block entry surface.
5-25
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
To assess the sweep limit in the two directions when using this style of block
requires that the probe be assessed in one direction first and them rotated
180° and the opposite sweep assessed.
Angular steps between A-scan samples will have an effect on the measured
sweep limits. A maximum of 1° between S-scan samples is recommended
for steering assessment however, angular steps are limited by the system
timing-delay capabilities between pulses and element pitch characteristics.
Most of the targets illustrated in figures 5.18 and 5.19 are separated by 5°,
however, greater or lesser intervals may be used depending on the required
resolution. This will of course mean that these blocks need to be custom
made based on the tolerances to be achieved.
5-26
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
In addition to strong grating lobes that may be present 20-30° off axis from
the main (intended) signal the grating lobes may also be present in a
separate mode.
For example the intended focal law was to form a 55° shear-wave pulse but
it also produces a strong compression mode at 80°. This could produce
misleading signals from geometric sources where performing the actual
inspection and the operator might call for repairs where no defects are
present.
When preparing for a weld inspection there is also the effect of material
attenuation and beam divergence to compensate for. For manual ultrasonic
techniques these are typically addressed using Distance Amplitude
Correction (DAC) curves or Time Corrected Gain (TCG) or by the theoretical
AVG or DGS curves.
A similar function can be used for phased array inspections as well. The
attenuation and divergence effects can be approximated by an exponential
curve that can be represented by
Vd V0ed
Where Vd is the signal height measured at some distance d in a material
with an attenuation factor and V0 is the initial signal (voltage). This
assumes that the material is isotropic (the same attenuation in all
directions).
Once the compensations are made for the wedge and echo-transmittance,
this leaves only the exponential decay effects of attenuation to address in
the phased array inspection setup for sensitivity.
5-27
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
This reduces the process to the same as that used for single element
probes where a series of identical targets (such as FBHs and SDHs) at
different soundpaths can be used to compensate for material attenuation
effects.
Figure 5.20 shows DAC being constructed on an A-scan display for a 35°
shear mode focal law.
For most phased array applications, analysis is made using projected scans
(B-scans, C-scans and S-scans) and the actual A-scan is not seen unless
some aspect of the signal is required for tip sizing characterisation.
The projection scans are assembled using colour palettes with colour
indicating amplitude so a DAC on an A-scan is not useful for analysis.
5-28
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Instead, in order that the operator is able to apply the analysis techniques to
the projection scans, all the signals of a specific colour should relate to
amplitudes that are already corrected for the soundpaths travelled.
This is accomplished by TCG whereby the signals later in time are amplified
with increasing gain as travel time increases.
TCGs are set by locating the same size target at increasing distances and
bringing the signal response to a constant level. When completed a series of
SDHs of the same diameter has the same amplitude regardless of the
distance to the SDH. This is illustrated in figure 5.21.
Herein lie what might be considered a problem for phased array systems. It
would seem that a separate TCG is required for every focal law! This would
be very time consuming to construct.
That would imply that for systems that have incorporated a separate step of
delay and sensitivity compensation for the wedge and echo-transmittance,
only a single TCG is required and it can be applied to all the focal laws.
5-29
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Caution is required where the operator views the A-scan with the TCG
overlaid! If the timebase is in units of True-depth (ie corrected for angle) and
the operator scrolls through the A-scans they will see an apparent shifting of
the TCG. However, if the display is made to indicate half-path the same
TCG will apply to all focal laws.
Figure 5.22 illustrates how a series of focal laws striking SDHs in a typical
calibration block produces the same sort of amplitude drop pattern with
increasing time (soundpath distance).
For the purposes of this section we will consider the most likely flaw types to
be oriented parallel to the weld centreline. This will therefore concentrate
efforts at providing effective beam coverage of the weld volume and heat
affected zone (HAZ) for flaws such as lack of fusion, incomplete penetration,
slag, porosity and centreline cracking.
5-30
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
However, with a phased array probe the operator is more likely to monitor
the S-scan and use a display that has a top and bottom of test surface
indicated on the screen so as to assist in locating the origins of signals.
Although E-scans could be used with manual raster scanning it would seem
to be redundant. It is more likely that the S-scan would be used with manual
raster operation of the phased array probe.
Figure 5.23 illustrates an S-scan display with adjustable markers that can be
used to indicate the bottom and top of the test piece. With a scale in mm
that indicates a distance from the probe the operator can estimate the depth
and approximate position of the source of the signal to make a judgement
as to its origin.
5-31
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Bottom and
top markers
True
depth
scale
Indication
Stand
off
scale
Even this simple option of a linear scan has its degrees of enhancement.
Standoff control from the weld when the weld cap is not removed is
sometimes adequate with just the operator sliding the probe as close to the
weld as possible.
The variations in standoff that result are relatively small in the order of 3-
4mm. This can be improved upon using a straight-edge guide. Magnetic
strips work well when inspecting steel components.
Most Standards require welds be inspected from both sides of the weld
(when possible). With a single phased-array probe this requires two scans.
However, some phased-array systems can be arranged to address two
phased array probes simultaneously.
5-32
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Encoder
Irrigation
connection
Dual probe
carriage
When used with a magnetic guide strip (as in figure 5.25) the operator has
good control on the standoff and flaw positioning accuracy is significantly
improved.
Having a solid guide strip to move against also tends to improve the
coupling, prevent stuttering motion (which can cause missing data due to
excessive speed) and generally speeds the entire data collection process by
avoiding re-scans due to missed data and poor coupling.
5-33
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Figure 5.25 Use of phased array probes and encoders with guide strip.
Motorised motion control with solid mountings can add a further degree of
mechanisation to the process. Where high production inspection of a
uniformly shaped part is required (such as pipe girth welds) positioning rings
can be used to facilitate rapid mounting and dismounting of the scanning
apparatus.
5-34
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Using the single element fixed angle beams the process was relatively
simple and often a pencil sketch was adequate to indicate the probe
movement to achieve coverage of the weld and HAZ.
Spread-sheet based.
Simple ray tracing.
Complex ray tracing.
Finite element modelling.
The girth weld inspection software was then designed to allow for the
operator to design the weld by entering the appropriate values to define the
bevel geometry and this included defining the number of zones desired.
5-35
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Even simple weld inspections using phased array probes require some
planning to ensure adequate coverage when trying to minimise probe
movement to a single line scan.
S-scan inspections of simple butt welds can often be done using a single
probe standoff. However, in order that each volume region has at least two
sound beams at different angles pass through it, this usually requires at
least 2 sets of S-scans and a probe with sufficient elements that can provide
starting elements in the S scan focal laws that are sufficiently spaced.
Optimisation is made using a very simple raytrace model indicating the weld
bevel with weld cap allowance, heat affected zone and probe/wedge
dimensions.
The image indicates a 50mm thick plate with a pair of phased array probes
on each side, each with 2 sets of S scans (45°-70°). In figure 5.28 coverage
from a single S-scan per side is seen to be inadequate to achieve full
volume coverage.
5-36
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
More elaborate modelling options linked to the phased array focal law
setups are provided by the manufacturers.
ONDT (Olympus NDT) has developed the Advanced Focal Law calculator
and Harfang Microelectronics provides their Phase FX software for
modelling the beam forming delays and resultant focalisation.
They are ideal (and absolutely necessary) when the array is a 2D array with
steering in more than one plane, see figure 5.29.
By far, the most detailed (and complex and time-consuming) options are
provided by finite element calculations of the beam. Some programmes can
now make good predictions of not only the beam path and shape but the A-
scans that can be expected for specific conditions.
Figure 5.29 illustrates the phased array probe making four S-scans at four
separate positions moving towards a weld.
5-37
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
During setup of the scanning parameters the views required by the operator
are limited. A-scans are used for amplitude verifications and setting TCGs
and ranges. The essential (and sundry) parameters used in the actual
inspection can often be summarised as part of a reporting package.
This is useful to compare to the technique description (to ensure that the
technique was followed correctly) and also provides values that could be re-
entered for repeating the test in subsequent inspections.
5-38
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Although the layout may change from one manufacturer to another or from
one inspection company to another, if the system is not equipped with a
report summary feature, the recorded items will generally be similar.
Figure 5.30 illustrates a typical setup summary. Note that the report form will
also usually indicate the set-up filename.
All the parameters used can (and should) be stored to a specific file. This
will allow the operator to recall all the essential parameters for subsequent
examinations.
Probe characterization
00000Setup
5-39
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Scan area
Calculator
Used First Last Resolution Wave Material
Element Qty Element Element Type Velocity
16 1 64 1 User Defined 3197.2 m/s
Start Angle Stop Angle Angle Focus Depth Scan Type
Resolution
60.0º 60.0º 1.0º 150.00 mm Linear
5-40
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
For the actual data acquisition, the operator usually sets up a view that
permits assessment of scan quality. B-scans (or D-scans) are useful here as
they provide a full display of the waveform for each data point along the
scan path (usually every 1mm as indicated in the table in figure 5.30 where
under Scan Area the Resolution is indicated as 1mm).
This allows the operator to assess if coupling is good and if the scanning
speed is suitable (ie not too fast so that data is being missed). Another
check on scan quality may be required to ensure that the probe(s) are not
drifting too far from the weld.
This might occur when there is a single probe and no guide band or no
fixture is used to hold a steady standoff distance. This can be accomplished
using weld geometry features such as the cap or root and monitoring the
amount of distance variation they have from the start of the scan where the
probe is assumed to be correctly positioned.
At the end of a project the results are often required to be handed in with the
raw data on CDs or DVDs or similar storage devices. The paperwork that
links the reports to the raw data must be traceable and accurate.
The operator needs to decide how best to arrange the information in order
to make a sensible and expeditious decision on the status of the weld.
Although it is possible to look at each A-scan individually this is not usually
an efficient use of time.
5-41
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
As well, single A-scans do not allow the operator to see the overall situation
and often it is the gradual deviation from a pattern that the operator sees
that is cause for concern and further investigation.
In addition to the views available the operator usually has some control on
the enhancements to the views. Depending on the software, these may
include:
This foldout diagram illustrated the top-side-end views and the sort of data
display that might be seen on each when a weld was scanned with a
phased array probe.
Figure 5.31 illustrates the ultrasonic data displays that might be associated
with the views on the left and the probe orientation on the right. Note that
the colour coding of the scales remains consistent from view to view.
The top shows the scan distance along the weld (always a shade of blue)
and the Index (standoff relative to the weld centreline) always a shade of
green.
The depth may be in time, halfpath or true depth (ie corrected for angle) and
may also include TOFD. Scales relating to the ultrasonic path are always
shades of red.
5-42
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
A typical starting analysis display would depend on the details of the scan
and component. The example in figure 5-32 is a pipe butt weld scanned
from the inside surface.
One of the added features for enhancement is the weld overlay. This allows
the operator to visualise where the beam is relative to the weld and typically
the weld centreline is used as the reference.
In order for the standoff to be referenced to the weld centreline the operator
must have correctly set the index offset in the setup so that the display
correctly indicates the distance from the weld.
Root geometry
Figure 5.32 has been made with two views of the same region.
On the left the red vertical cursor is seen aligned with the red dot (indication)
that appears on the top and side views. With the views linked this brings the
volume-corrected S-scan from that scan position onto the upper right pane.
5-43
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Using the weld overlay the operator can see that the flaw occurs 6mm
before the weld centreline and that the fainter line of indications running the
entire length of the scan as seen on the Top view (upper left pane) is due to
the root geometry.
The weld overlay is made with a mirror image shape reflected at the cap
surface. Therefore this also allows us to see that the indication has been
detected on the end of the second half-skip.
Since the views are linked, when the black cursor in the End view (volume-
corrected S-scan) is moved to a different angular slice (61°) where no flaw is
seen on the S-scan, the Top and Side views update and indicate that there
are no flaws at any position along the scan at that angle.
In both views the A-scan is included and we can see the individual A-scan
that is responsible for the line of colour where the black cursor was placed.
Using the same weld and same flaw position, we can see how the scan from
the opposite side of the weld provides useful information that helps to
confirm this indication as a flaw and to establish its position 6mm from the
weld centreline as being accurate.
Figure 5.33 uses the volume corrected S-scans from the two directions (ie
90° and 270° beam skew) to verify the flaw position.
The presentations on the right side of Figure 5.33 show the indication
occurring after the line on the Side view that corresponds to the root
geometry.
5-44
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
By simple modelling it can be seen that there are portions of the weld that
are not well addressed by a simple pair of S-scan from either side of the
weld, even with the skip from the OD surface.
Being from the inside surface of the pipe the inspection is configured to
provide a full skip for coverage of the root and the cap region is covered by
the half skip.
The overlay illustrates how the cap region is missed. In fact, this weld was
inspected using a pair of S-scans from each side of the weld as indicated by
the modelling in figure 5.34.
Pipe ID
Root
Pipe OD
Figure 5.34 Pipe butt weld using 4 S-scans.
5-45
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
When the full collection of A-scans is used from all four S-scans and these
are merged, the operator can get a full volume analysis. With a few other
enhancements such as correction for skip and smoothing to reduce the
pixelisation, analysis becomes a bit easier.
Figure 5.35 is a photomontage made using a single End view from the four
merged S-scans with a sketch of the two phased array probes overlain to
align with the S-scan beams in the.
A series of these S-scans is collected (every 1mm) along the pipe inside
surface to build up the total merged volume of the inspected weld and HAZ.
On the initial slice the weld overlay is seen and because the software was
used to make correction for the skip the mirror image of the weld seen in
figure 5.33 is now just the weld bevel outline for just the 20mm wall
thickness.
The process of merging has divided the inspection volume into small
volumes made of the three different scan dimensions:
The small volumes that result are called voxels which is the volume image
equivalent of the computer image term pixel (which is a contraction of the
words pixel and element).
When the software allows merging of data and the concepts of voxelisation
are used, the entire inspected volume can be considered to be divided into
these small volume packets.
5-46
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
With multiple S-scans and skip paths, the actual number of beams passing
through a volume point may in fact be four or more.
By using the calculated centre of beam axis for every focal law and
incorporating the geometric corrections for skips, the software can then
determine which beams passed through each voxel.
It must then determine the amplitudes at every voxel from every beam and
assign the maximum amplitude that occurred in that voxel from all the
beams.
In the merged data only one amplitude value (the maximum amplitude) is
indicated. This value is used to reconstruct the volume with amplitudes
assigned to each voxel.
The three orthogonal projections used for merged Top-Side-End views are
illustrated diagrammatically in figure 5.36.
Each sample position along the scan (usually in 1mm increments) can be
rebuilt with the maximum amplitude identified for each standoff and depth
position. By linking the cursors the software can project either a single slice
or project several slices from one of the orthogonal planes.
5-47
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
End Projection
Projections may consist of a
single slice or a range of slices
projected orthogonally
through the surface
Depth Projection
tio nc
Pro je
Sid e
Figure 5.37 is a projection display of the same weld as was used in figure
5.33. The flaw located in this case has been highlighted by small projections
limiting the merged data display to just the region a few millimetres around
the flaw.
The flaw as seen from the Top view (upper left) extends along the length of
the scan from 382-405mm. These limits are used to restrict the End view
(upper right).
5-48
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
Since the end view now shows all the end views from 382-405mm (a total of
23mm) this provides the maximum amplitudes seen for 23 sample positions
along the scan.
The Top view (upper Left) also indicates that this indication is at
approximately +6mm from the weld centreline. By adjusting the standoff
projection cursors (these are the red and blue vertical lines in the End view
in the upper right pane) the Side view (lower left pane) shows the flaw
projected from only 5-10mm from the weld centreline. This shows the length
and height of the indication.
The reason that the Top view shows the indication is due to the fact that the
depth of the projection has been limited to the bottom 5mm. Looking at the
purple bar under the Top view (upper left) we see the two black markers on
that bar and the values at the end of that bar indicate 15mm (on the left
side) and 20.7mm (on the right side).
This indicates that the Top View is a projection of the region 15 to 20.7mm
down from the inspection surface. This corresponds to the region just up to
and including the outside pipe surface.
These occur at the half skip, full skip and skip and a half positions and
correspond to the signals from the focal laws that detected the flaw at the
cap (20mm and 60mm True Depths) and the root geometry from the full
skip.
5-49
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
5-50
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
In the second group we could also include the instructional standards such
as:
Standards such as those listed above are the attempts of industry to ensure
a minimum level of functionality or quality. When a new method arrives on
the scene industry often lags the research and development associated with
that new method.
This, for example, was the case with TOFD where it took over 20 years from
its introduction in 1976 to establish the first Standards. Phased array has
been available to NDT users since at least the early 1990s and suffer a
similar problem in industry, ie a lack of Standards.
Part of the problem seems to stem from the fact that it was simply assumed
that phased-array testing was just another form of pulse-echo ultrasonic
testing.
To some extent this is true; however, the code and standards writers had
put wording in the existing codes for manual single element ultrasonic
testing that made the Standards incompatible with the application of
phased-array technologies.
For example:
These and several other issues make even the latest editions of some
Codes and Standards unsuitable to use with phased array technology.
However, when we look at the intent of the Standards as a tool by which to
provide a good quality inspection, it is obvious to many that phased-array
inspections can usually provide a superior inspection compared to single
element probes.
5-51
www.twitraining.com
Rev 3 January 2010
Calibration and Scanning Considerations for Weld Inspections using Phased Array UT
Copyright TWI Ltd 2013
For example:
5-52
www.twitraining.com
Section 6
Using phased array probes and electronic sweeping scanning ensures that
seam tracking problems do not allow the ERW seam to wander outside of
the beam coverage area.
Figure 6.1 shows the conventional setup and figure 6.2 shows the phased
array coverage.
70 45
6-1
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
This design uses a phased array probe with both detection and centring
functions. It was designed to be lightweight, handheld, fast and easy to
operate (less than a minute per fastener) and required blind-test validation.
Figure 6.3 shows the fastener flaw location and on the right the rho-theta
probe schematically positioned over the fastener. Over 500 elements make
up the probe that is about 4cm in diameter.
In the lower image of figure 6.3 is shown how the system is applied to the
underside of the wing.
Faying surface corner
6-2
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
Figure 6.4 shows the angle coverage on profiles on the left and the
mechanised system with a blade image inserted over that.
Figure 6.4 Turbine blade root inspection by phased array beam steering.
6-3
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
Traditional multi-probe systems (as in figure 6.6 right side) are now replaced
with smaller phased array systems (figure 6.6 left side). Mass reduction of
the scanning head from about 15-20kg with 24 or more probes to a 5kg
mass with just 2 phased array probes do the same functions.
Custom GUIs provide a fast and easy method of configuring focal laws for
standard weld profiles. The operator need only enter the wall thickness and
define the number of zones and an initial setup is generated to transfer to
the phased array control software (see figure 6.7).
6-4
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
Figure 6.6 Pipeline girth weld inspection – phased array versus multiprobe.
Figure 6.7 Phased array GUI for girth weld focal laws.
6-5
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
On the right is a ToFD scan using a 7.5MHz phased array probe using 12
elements (equivalent to a 12mm element) on a 32mm wall thickness. The
only differences are those relating to signal duration.
The PA results are seen to have lower frequency content (note the longer
duration of lateral wave). Whereas an operator is limited to the results
obtainable from a single element ToFD pair, a phased array operator can
optimise conditions to some extent. Adjusting the number of elements to
vary the beam coverage and adjusting the angle by a couple degrees up or
down to improve near surface resolution.
As well, the phased array operator has the ability to use a focused beam to
improve sizing resolution. Phased array features such as multiple beam
angles and beam divergence characteristics can be carried out
simultaneously in a single scan whereas a single element process would
require a large scanning array or multiple scans of the same test specimen.
Figure 6.9 illustrates the principles of the backscatter sizing technique for a
surface-breaking flaw. Note that the same principles can be used for both
single element and bi-modal dual element probe techniques.
The examples in figure 6.9 indicate a simple shear mode with the two
conditions of the probe positioned to peak on the flaw tip in the first half skip
(upper) or second half skip (lower).
6-6
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
Figure 6.9 Simple tip diffracted signal using pulse-echo shear wave.
For a planar flaw the general condition for the origin of tip diffracted signals
can be described as in figure 6.10.
Flaw of Size D
D1 D2
The delta time between D1 and D2 does allow for some estimate of sizing
via
2D
t sin
c
6-7
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
Where D is the distance between the defect extremities and is the angle
that the incident beam makes with the line perpendicular to the planar
defect.
The analysis that allows sizing and orientation determination relies on probe
motion and is somewhat dependent on coupling variation as the analysis in
part uses amplitude peaks.
X3
X2
X1
F1 F2
F3
Echo Amplitude
Envelopes of F1, F2 & F3
Echoes
F1
F2
F3
Figure 6.11 Probe positions and soundpaths recorded for peak amplitudes of tip
echoes.
Figure 6.12 illustrates the scan pattern used. Several focal laws are
arranged to sweep across the notch as the scanner advances the probe
parallel to the notch axis.
6-8
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
Probe motion
parallel to notc h axis
FL25 FL1
EDM notch
The advantage of a phased array raster scan in a contact test exists in both
time (faster) and uniformity of coupling thereby making the amplitude trace
less variable than it might be if the probe was a single element being
pushed and pulled as well as sliding sideways as would be the case for a
mechanical raster scan.
When completed, the scan data can be presented in a plan view integrating
the maximum amplitudes over the entire area (a traditional C-scan). This
identifies where the flaws are located along the length of the scan.
Using sufficient gain the tip diffracted signals can be identified. Due to the
echo dynamics available in the raster scan, the patterns that result allow the
first steps in flaw analysis.
Figure 6.12 shows the combination of A, B and C-scans that allow the
operator to view the tip diffraction signals for analysis.
The image in the top left is a C-scan (plan view). In the top right is the
volume corrected end view, B-scan of a single position along the notch with
each focal law A-scan shown at 45°.
The A-scan in the lower right is the display of the angled cursor on the upper
right.
6-9
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
6-10
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
There is even one unit (Omniscan MX) that also includes a portable eddy
current module and associated displays.
Figure 6.14 illustrates some of the phased array systems now available on
the market for NDT applications.
6-11
www.twitraining.com
Rev 3 January 2010
Industrial Applications of Phased Array UT
Copyright TWI Ltd 2013
SAPHIR
(IntelligeNDT)
6-12
www.twitraining.com
Section 7
For over half a century, production pipeline girth weld inspections were the
exclusive territory of radiography. In the late 1970’s this old boys’ club of
pipeline radiographers was given notice by an innovative adaptation of
ultrasonic inspection.
The ultrasonic zones were easily incorporated into the concepts of fitness-
for-purpose where the aspect ratio of the flaw is critical in determining the
ability of the weld to tolerate a flaw of a specified size.
The simple three-probe equipment of the 1960’s soon grew to systems with
20 probes or more and the simple three zones increased in number as the
zones decreased in size.
By the end of the 1990’s the Pipeline Research Council International (PRCI)
had approved a project that established proof that phased-array AUT could
provide the same or better results than the multi-probe systems using single
elements.
Although phased array systems have not completely replaced the old
technology of multi-probe units, they have effectively become the equipment
of preference. By far, the most common phased-array system used on
pipeline construction is now the Olympus PipeWIZARD (PWZ). As of 2008
there were nearly 100 PWZ systems worldwide.
7-1
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Some of the companies that started with multi-probe units have made efforts
to convert to phased-array systems. RTD made an initial attempt with a
hardware design made by Technology Design.
Shaw Pipeline Services Limited, after many years of effort with Krautkramer,
failed to come up with a reliable unit of their own. However Krautkramer (via
GENDT) with an association with UT Quality developed a version they
called WeldStar that has seen some success.
Although not common, a few other systems have been developed that
combine the phased-array features with the zonal discrimination techniques.
One of these is NDT Brasil (using a TD Focus system).
Figure 7.1 illustrates some of the phased-array systems now using the zonal
discrimination technique.
7-2
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
A significant refinement was made in the latter part of the 1980’s with the
development of internally focused contact probes. This kept the beam
concentrated at the area of interest near the weld fusion line. This
advantage of focusing improved zonal discrimination improved the sizing
estimates. It had the side effect of reducing the false-calls of repair due to
operators mistaking geometry signals as defects.
The weld is divided into zones typically 1-3mm high and beam angles are
selected to optimise response off the fusion face of the weld bevel.
Figure 7.2 AUT weld zones and beam paths for three weld bevel shapes
commonly found in pipeline girth welds.
The array of probes is moved around the girth weld by a motorised carrier
that moves along a track typically used by the welding apparatus. Ultrasonic
signals received by the instruments are monitored using electronic gates.
Both amplitude of signal and time of flight of the signal in the gated region
are monitored.
The region gated is from just before the theoretical weld bevel preparation to
just after the weld centreline. Gated output time and amplitude as well as
waveforms are digitised and displayed in a chart format.
7-3
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
ID surface notches and 2-3mm diameter flat bottom holes typically provide
targets on which signal amplitude and travel time in a gate are set. These
targets are machined in a project specific section of pipe and arranged
along the theoretical weld bevel profile. So-called volumetric channels are
added to improve detection and identification of small and off-axis flaws
such as porosity. Most systems now incorporate some form of ToFD as part
of the configuration.
Figure 7.3 illustrates many of the flaws and geometric conditions that an
AUT system will be capable of detecting and identifying. These are shown
on the CRC bevel and typical of the flaws that can form with the exception of
the misfire.
Fill
Porosity
A true separation of weld passes identical to the fixed ultrasonic zones is not
possible. In the welding process, especially where the weld is made with the
joint vertically oriented, the weld puddle does not deposit an equal thickness
of metal in all positions around the girth. Gravity pulls the weld puddle down.
On the top the deposits tend to be thinner and on the bottom deposits
thicker.
7-4
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Variations in weld pass thickness will be less pronounced when the pipe can
be rotated under the arc or when the weld is oriented in the horizontal
orientation as would be the case in an offshore J-lay. The welding pass and
the ultrasonic zones do not always align there is sometimes a bit of
confusion due to the fact that the AUT zone is usually identified by a weld-
pass.
For example in figure 7.3 we see the irregular outlines representing the hot
pass and three fill passes, the final fill pass being the cap pass. However,
the AUT would have two hot pass zones and two fill zones.
The welding process in figure 7.3 uses an internal root pass that is intended
to burn through and consume the vertical land at the cross penetration point.
Therefore, although there is an AUT zone identified as the LCP there is no
separate welding pass that corresponds to an LCP deposit.
Since the focal points of the ultrasonic beams need not always align with the
weld passes the ultrasonic responses from a single flaw are often seen
crossing from one zone into the next (illustrated in figure 7.4 where a single
flaw in the fill two welding pass is seen by AUT zones fill1, fill2 and fill 3).
Spot size is a result of beam focusing. Three practical options are available
for focusing using contact probes:
1 Lens focusing: Flat element with curved plastic plates mounted in front,
figure 7.5.
2 Curved element focusing: The element is curved and an insert in the
gap, figure 7.6.
3 Phased array focusing, figure 7.7.
7-5
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Lens material 1
Wedge material 2
Curved element
Gap material (same as wedge)
Wedge material
7-6
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
The ability to construct many different focused angle beams from a single
phased-array probe provides a significant advantage over the single
element focused probes. In all cases (single element and phased array) the
limits of focusing are similar. These are described in section 3.2.
The targets are FBHs and surface notches. The FBHs are centred in each
zone and the flat bottom of the hole is milled or electric-discharge machined
to align with the theoretical position of the weld bevel.
Project-specific means that the pipe used for the calibration block is from the
same pipe mill used for the construction project. This is to address the
potential for extreme velocity variations that have been quantified in steels
from different pipe mills.
The targets are seen positioned using a transparent effect for the solid CAD
model. To the right is an actual calibration block in its dummy.
TOFD OD notch
notch
OD notch Trans OD
notch
Being project specific material means that the setups used on the calibration
will duplicate beam angles in the tested pipe. However, for proper
construction of wedges and phased-array focal laws, the acoustic velocities
of the material are determined prior to the job.
This involves using an SH shearwave probe and assessing the fast wave of
the birefringent pair seen during the assessment.
7-7
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Figure 7.9 shows how the beams are calculated to align with a path that
attempts to provide a direct perpendicular incidence on the theoretical fusion
face or some form of reflection path is used with a Tandem probe (focal law)
configured to ensure optimum received signals. In a multi-probe system this
would mean that a separate probe (or tandem probe-pair) was required for
each zone.
The phased array probe need only be concerned that the wall thickness and
angle combinations allow the selected probe/wedge to provide adequate
length that the exit points for all the soundpaths can be addressed by a
single probe position.
Figure 7.9 also illustrates the reason for the Zed shape of some calibration
blocks.
When the method of making the FBH targets is by an end-mill cutter the
depth to which a 2mm or 3mm FBH can be milled without risk of breaking
the tool-bit is about 20-25mm.
Squaring off the calibration block 20mm from the theoretical weld centreline
minimises the cutting depth for the affected targets.
12mm
Some users like to add a surface notch as an extra zone for the cap. This
OD notch should be relatively small (1mm deep by 5mm long) so as to
provide a similar sensitivity to the FBHs.
7-8
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
In fact, it was others who later took the strip-chart data display and made it
the ergonomic tool of preference that it has become today.
The symmetry of the left side and right side of the charts is the symmetry
along the weld axis that separates one pipe from the other and a side of the
weld can be labelled (usually upstream and downstream). The basic
symmetry is illustrated in figure 7.11.
7-9
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
The B-scan and ToFD enhancements are arranged with similar symmetry in
mind. Some displays place the ToFD in the centre of all the strip-charts and
B-scans while others set the ToFD off to one side.
B-scans are always made in pairs so are treated similarly to the amplitude-
time strip-charts, with upstream and downstream presentations symmetric
about the weld centreline.
Variation also exists with the organisation of data. Most companies use a
large (21 inch) computer monitor for evaluation of on-screen data. When
large numbers of zones are required, as is the case for thick sections, the
width of each channel on the monitor may be too small for the operator to
make a reasonable assessment of the data.
The operator will use the information on the charts displayed on the monitor
to assess the quality of the calibrations and the quality of the welds including
the scanning quality.
7-10
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Figure 7.12 illustrates some of the quality aspects that might be noted for
correction prior to moving on to the weld scan. This image also illustrates
the other aspects of the data acquisition including ToFD, B-scan and
coupling status displays.
Coupling status
B-scans
(volumetrics) ToFD
7-11
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
A closer look at the details of the strip chart will provide a better
understanding of how the information is displayed. Figure 7.13 is an
expanded view of the responses in the first 4 zone channels in a calibration
scan.
The axial position shown is from 0-70mm (as indicated by the scale on the
left, note the reference cursor across the top of the chart indicating 5.4mm.
A scale with 10 divisions has been overlaid in the strip chart region (lower
right) of the four strips where signals are seen. Displacement of the black
line from the left side of the strip chart indicates amplitude and values can
be seen represented by the amount that the black line moves towards the
right edge of the strip chart.
Similarly, the displacement of the colour bar from the midpoint of the strip
chart can be compared to the inserted scale. R1U, R2U and LU are all close
to the midpoint but HU is noticeably further shifted to the right
(approximately 60% across the chart for HU).
R1U, R2U and LU time-gate lengths were indicated by the operator as each
being 10mm long and HU is only 8mm long. Therefore, in R1U, R2U and LU
each division on the scale represents 1mm and the colour bars can be seen
to be less than 1 division from the centre position.
However, since the time-gate length in HU is only 8mm each division on the
inserted scale represents 1.25mm. The displacement of the colour bar to the
right of centre in HU is about 1 division so approximately 1.2mm inward of
the theoretical fusion line.
Since the zonal targets in the calibration block are assumed to represent the
theoretical fusion line therefore the calibrated position for HU is off the ideal
by 12% of the gated distance.
7-12
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Having made a good calibration scan the operator can move on to scanning
a weld. Even before an operator begins evaluating scan data to determine
the acceptability of a weld they must determine if the scan was of an
acceptable quality.
Again, something of a mental checklist is gone through after each scan. The
items reviewed for general chart quality would include:
If there are quality issues with the weld scan the problems should be
corrected and the scan redone. This would be comparable to a radiograph
with missing identification, processing marks in the weld or too much or too
little density. In radiography a re-shoot would be required; in AUT it requires
a rescan.
Ultrasonic data collected by the AUT systems are of three sources, strip-
charts, volumetric projections (B-scans or mapping) and ToFD. The
operator’s review and assessment of each of these display types is
required.
In general signals displayed on the monitor can be grouped into three types:
1 Flaws.
2 Geometry.
3 False indications (over-trace).
7-13
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
False indications are a result of an odd condition associated with the fixed
positions of the beam having some vertical extent in the linear scan used in
AUT. These are indications of zonal channels seeing flaws that occur off-
axis and are actually in a different zone than the one detecting them.
The original work carried out in Canada used a 2mm diameter FBH
maximised to 80% as the Reference Level or Reference sensitivity level. All
flaw signals greater than or equal to half that amplitude (ie 40%) were
evaluated to the allowed length criteria.
Therefore, the Evaluation Threshold was 40%. Later, some used a 3mm
diameter FBH to establish the Reference Level at 80% and used a 20%
Evaluation Threshold, ie flaw signals at or greater than 20% FSH were
compared to the allowed flaw length criteria.
Note above that not all signals recorded in an AUT scan need be signals
that originate from flaws. The operator must decide what signals are flaws
and only compare those signals to the allowed length criteria.
When the interpretation process begins on the strip-charts, the operator first
looks for signals over the evaluation threshold. It is only those signals that
will be compared against the acceptance criteria.
Amplitude is only one of the two items gated in the strip-charts. The other
item is time and it is the time information that the operator uses to aid in
identification of flaws.
7-14
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Figure 7.14 Cap geometry due to dual torch Vee 32mm wall J-bevel.
7-15
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
7-16
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Nearly all such criteria have been derived from radiographic findings and
have their origins dating back to the 1950’s. When converted to ultrasonic
methods workmanship acceptance criteria have the operator merely assess
the length of any flaw indication that exceeds the reference level or eg 50%
of the reference level.
Applying this in the field usually means the operator measures the flaw
length between the points where the indication exceeds 40% screen height.
Zonal discrimination, we noted, was given its big advantage when the
fracture mechanics engineers developing BS PD 6493 derived the fitness-
for-purpose concepts and required that the flaw height and length be
compared to calculated critical flaw sizes.
When acceptance criteria are based on these ECA concepts the operator
must also estimate the vertical extent of the flaw. Length estimations using
the amplitude limits are often used in this case too.
Length limits are typically assessed from the 20% or 40% amplitude points
at the ends of the indications. Although this method of length sizing is not
extremely accurate (typically within +/-3mm of actual length) it is generally
adequate for most applications.
Sizing of the vertical extent of a flaw is, however, much more critical from
the fracture mechanics engineers perspective. A variety of methods are
used by various users of the zonal discrimination technique.
When the amplitude of a signal is used as the basis for sizing it must be
remembered that the amplitude of a signal is derived from the reflecting
area of a flaw. The flaw length, width, orientation, surface roughness and
backing material (air, silica, water) all have an effect on amplitude.
After many years of assessment on the various sizing efforts used in the
zonal techniques, the standard deviation of sizing error is now being
reduced to approximately 1mm.
7-17
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
Shots that strike closer to the centre are considered more accurate. The
closer a system's measurements are to the centre or accepted value, the
more accurate the system is considered to be. Therefore, if a large number
of shots are fired and were equally distributed away from the centre the
system would be considered accurate.
However, if all the shots are grouped closely together but off to one side of
the centre in a cluster, the system is considered to be precise. The
measurements are precise, though not necessarily accurate. Still, it is not
possible to reliably achieve accuracy in individual measurements without
precision.
If the shots are not grouped close to one another, they cannot all be close to
the centre. Their average position might be an accurate estimation of the
centre but the individual shots are inaccurate. These concepts are indicated
in figure 7.17.
The difference between the mean (average) and the reference (true) value
is the system bias. Generally in AUT it is considered conservatively
preferable to have a slight bias to oversizing. Assessment of the precision is
made using the standard deviation which estimates how far off the mean
value most of the estimates are.
Standard deviation is the root mean square (RMS) deviation of values from
their arithmetic mean. For example, in the population {4, 8}, the mean is 6
and the deviations from mean are {−2, 2}. Those deviations squared are {4,
4} the average of which (the variance) is 4. Therefore, the standard
deviation is the square root of 4, ie 2.
7-18
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
In just over two decades this value was reduced to about half that value.
The error in sizing is an important factor in ECA based acceptance criteria
because it is added to the size estimated by the AUT system when
calculating the allowed flaw length.
7-19
www.twitraining.com
Rev 3 January 2010
Pipeline Girth Welds
Copyright TWI Ltd 2013
7-20
www.twitraining.com
Section 8
Websites
History of UT (medical)
Dr. Woo on his website
http://www.ob-ultrasound.net
Texts
Introduction to Phased Array Ultrasonic Technology Applications, by
R/D Tech Inc., Published by R/D Tech Inc., 2004, ISBN 0-9735933-0-X.
8-1
www.twitraining.com
Rev 3 January 2010
Other Reading and References
Copyright TWI Ltd 2013
Journal articles
Ultrasonic phased-arrays for non destructive testing,
McNab, A, Campbell, M.J., NDT International, 1987, vol.20, no.6, pp.333-
337.
8-2
www.twitraining.com
Glossary of Terms
Rev 3 January 2010
Glossary of Terms
Copyright TWI Ltd 2013
Glossary of Terms
The terms found in this glossary are the result of contributions from several
of the industry manufacturers and users.
Beam steering
The ability of a phased array system to electronically sweep the beam
through a range of incident angles without probe movement.
Dead elements
Elements in an array that are no longer active. Dead elements may
influence the construction of the ultrasonic beam.
Depth focusing
Focusing over a larger range than conventional transducers by electronically
adjusting the beam.
www.twitraining.com
Rev 3 January 2010
Glossary of Terms
Copyright TWI Ltd 2013
Electronic scan
Also termed an E-scan, swept index point or electronic raster scanning.
Note: In some industries, an E-scan is referred to as a linear scan. The
same focal law is multiplexed across a group of active elements; E-scans
are performed at a constant angle and along the phased array probe length.
Focal law
Strictly, a mathematical formula used for firing the phased array instrument.
More generally, a file containing the entire set of hardware and software
parameters for phased array operation which defines the elements to be
fired, time delays, voltages, for both the transmitter and receiver functions.
The other option is to focus at a fixed sound path. In all cases, the ability to
focus is a function of the near zone of the beam and can only occur for
sound paths less than the near field of the unfocused aperture.
Linear scan
In some industries called a one-line scan. A single pass mechanical scan
which is parallel to the weld or region to be inspected.
Multiple group
Also called multi-group. Firing and displaying multiple scan patterns during a
single scan, eg combinations of S-scans, E-scans, ToFD using different
parameter sets or exit positions and possibly with more than one array.
www.twitraining.com
Rev 3 January 2010
Glossary of Terms
Copyright TWI Ltd 2013
Phased array
The phased array technique is a process wherein UT data is generated by
constructive phasal interference formed by multiple elements controlled by
accurate time delayed pulses.
The arrays can perform beam sweeping through an angular range (S-
scans), beam scanning at fixed angle (E-scans), beam focusing, lateral
scanning and a variety of other scans depending on the array and
programming. Each element consists of an individually wired transducer,
with appropriate pulsers, multiplexers, A/D converters and the elements are
acoustically isolated from each other.
Probe types:
Annular array probes: Phased array probes that have the transducers
configured as a set of concentric rings. Annular array probes allow the
beam to be focused to different depths along an axis. The surface area
of the rings is in most cases constant which implies a different width for
each ring.
Daisy array probe: Daisy array probes are effectively a linear array
curved into a circle such that the ultrasound is emitted along the axis of
the circle/cylinder. This type of array can be used with a mirror to inspect
from the inside of tubes.
www.twitraining.com
Rev 3 January 2010
Glossary of Terms
Copyright TWI Ltd 2013
Matrix array probes: These probes have an active area divided in two
dimensions using different elements. Matrix array probes are typically in
a checkerboard format though other designs are used. These probes
allow ultrasonic beam steering etcetera in multiple planes.
Sectorial scan
Also termed an S-scan, swept angle scan or azimuthal scan. This may refer
to either the beam movement or the data display. As a data display, it is a
2D view of all A-scans from a specific set of elements corrected for delay
and refracted angle. When used to refer to the beam movement, it refers to
the set of focal laws that sweeps through a defined range of angles using
the same set of elements.
Skew
The ability of the phased array probe to deliberately skew the beam away
from the main axis. This capability requires a matrix array or 2D array.
Active aperture: The size of the group of active acoustic elements (A).
www.twitraining.com
Rev 3 January 2010
Glossary of Terms
Copyright TWI Ltd 2013
Element pitch: The distance between the centres of two adjacent array
elements (p).
Saw cut: Also called kerf (gap). The space between adjacent elements
(g).
p g
Coupling: Method for keeping the wedge to test material interface wet
or coupled. Can also be called irrigation.
www.twitraining.com
Rev 3 January 2010
Glossary of Terms
Copyright TWI Ltd 2013
Height of first element: The linear dimension from the contact face of
the wedge vertically to the centre of the first element, see pictures.
Lateral array wedge: Array wedge used with the array mounted 90
degrees to the normal array wedge mounting. The resulting beam is
fixed in the incident plane while it is steered as to skew the beam.
Normal array wedge: Array wedge used with array mounted so that
when the array is steered, the resulting beam is steered to vary the
incident angle.
Roof angle: In a complex angle wedge, the roof angle is the minor
angle, while incident angle is the major angle.
www.twitraining.com
Rev 3 January 2010
Glossary of Terms
Copyright TWI Ltd 2013
www.twitraining.com
Rev 3 January 2010
Manual and Encoded Phased Array
Copyright TWI Ltd 2013
S Supplementary Information
www.twitraining.com
Contact us for information on our wide range of courses at:
TWI Training & Examination Services TWI Technology Centre (North East)
Granta Park Aurora Court
Great Abington Barton Road
Cambridge Riverside Park
CB21 6AL, UK Middlesbrough
TS2 1RY, UK
Tel: +44 (0)1223 899500 Tel: +44 (0)1642 216320
Fax: +44 (0)1223 891630 Fax: +44 (0)1642 252218
Email: trainexam@twi.co.uk Email: twinorth@twi.co.uk
Website: www.twitraining.com Website: www.twitraining.com
TWI Technology (S.E. Asia) Sdn Bhd TWI Middle East FZ - LLC
No.8 Jalan TSB 10 Knowledge Village
Sg Buloh Industrial Park Block 14, 3rd Floor
47000 Sg Buloh Office 306/307
Selangor Darul Ehsan PO Box 502931
Malaysia Dubai, UAE
Tel: +603 6157 3526/28 Tel: +971 4 364 30 10/11/12
Fax: +603 6157 2378 Fax: +971 4 367 84 35
Email: inquiry@twisea.com Email: deedar.shah@twime.com
Website: www.twitraining.com Website: www.twitraining.com