Uploaded by jose.serraschaves

NDT44 - Phased Array UT - Welds 060319 - Textbook (1)

advertisement
Phased Array Ultrasonic Testing (PAUT) - Welds
NDT44
Training and Examination Services
Granta Park, Great Abington
Cambridge CB21 6AL
United Kingdom
Copyright © TWI Ltd
Phased Array Ultrasonic Testing (PAUT) Welds
Contents
Section
Subject
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
History
The Principles of Phasing
Beam Forming
1.1
1.2
1.3
1.4
1.5
Transducer arrays
The ultrasonic beam
Phased array aspects of beam characteristics
Fermat’s principle
Phased array beam focusing and steering
Hardware: Pulsers, Receivers, Motor Control and Encoders
1.6
1.7
1.8
1.9
1.10
1.11
1.12
1.13
1.14
1.15
1.16
1.17
1.18
1.19
1.20
1.21
1.22
1.23
1.24
1.25
1.26
NDT44-60319
NDT Prelims
Advantages of digital control
Pulsers and receivers
Pulsers
Spike pulsers
Tone burst
Square wave pulsers
Receivers
Gates
TCG/TVG
Data acquisition and automated systems
Automated systems in general
System components
Instrument outputs
Scanning displays and scanning equipment
Memory and digitisation aspects
Data processing
Scanning equipment
Limitations of mechanised scanning
Scanning speed
Encoders
Asynchronous versus synchronous systems
1
Copyright © TWI Ltd
Calibration and Scanning Considerations for Weld Inspections using
Phased Array UT
1.27
1.28
1.29
1.30
1.31
Phased array calibrations
Scanning calibration
Phased array technique development
Phased array data analysis
Phased array codes and standards
Industrial Applications of Phased Array UT
1.32
1.33
1.34
1.35
1.36
1.37
1.38
1.39
1.40
Electric resistance welds
Aerospace fuselage fastener cracking/corrosion
Power generation – turbine blade roots
Power generation – heavy nozzles
Petrochemical pipeline construction (PipeWIZARD)
Other applications using phased arrays
ToFD by phased array
Backscatter sizing
Portable phased array
Pipeline Girth Welds
1.41
1.42
1.43
1.44
1.45
Principles of the zonal discrimination technique
Strip charts
Evaluation thresholds
Examples of indication types
Acceptance criteria and sizing
Further Reading and References
Glossary of Terms
Supplementary Information
NDT44-60319
NDT Prelims
2
Copyright © TWI Ltd
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
Date
NDT44-60319
NDT Prelims
Revision
number
Changes made
3
Copyright © TWI Ltd
Standards and Associated Reading
Standards
BS EN ISO 13588
Non-destructive testing: Ultrasonic Testing – Use of automated
phased array technology.
BS EN ISO 18563-1
Non-destructive testing: Characterization and verification of
ultrasonic phased array equipment. Part1: Instruments.
BS EN ISO 18563-2
Non-destructive testing — Characterization and verification of
ultrasonic phased array equipment. Part 2: Probes
BS EN ISO 18563-3
Non-destructive examination of welds: General rules for metallic
materials. Part 3: Combined systems.
BS EN ISO 17640
Non-destructive testing of welds: Ultrasonic
Techniques, testing levels and assessment.
BS EN ISO 17635
Non-destructive examination of welds: General rules for metallic
materials.
BS EN ISO 16827
Non-destructive testing — Ultrasonic testing — Characterization
and sizing of discontinuities.
BS EN ISO 20601
Non-destructive testing of welds - Ultrasonic testing - Use of
automated phased array technology for thin-walled steel
components.
BS EN ISO 19285
Non-destructive testing of welds – Phased Array Ultrasonic
testing (PAUT) - Acceptance levels
BS EN ISO 1330-2
Non Destructive testing: Terminology Part 2: Terms common to
NDT methods.
BS EN ISO 1330-4
Non Destructive testing: Terminology Part 4: Terms common to
ultrasonic testing.
BS EN 16018
Non-destructive testing — Terminology — Terms used in
ultrasonic testing with phased arrays.
BS EN ISO 23279
Non-destructive testing of welds — Ultrasonic testing
Characterization of indications in welds.
ISO 18175
Non-destructive testing: Evaluating performance characteristics
of ultrasonic pulse-echo testing instruments without the use of
electronic instruments.
BS EN ISO 16810
Non-destructive
principles.
BS EN ISO 16811
Non-destructive testing — Ultrasonic testing — Sensitivity and
range setting.
BS EN ISO 19675
Non-destructive testing — Ultrasonic testing — Specification for
a calibration block for phased array testing (PAUT)
ISO 9712
Non-destructive
personnel.
NDT44-60319
NDT Prelims
testing
testing:
4
—
Ultrasonic
Qualification
testing
and
testing
—
–
—
General
certification
of
Copyright © TWI Ltd
ASTM E-317
Standard practice for evaluating performance characteristics of
ultrasonic pulse-echo testing instruments and systems without
the Use of electronic measurement instruments.
ASTM E-1324
Standard Guide for Measuring Some Electronic Characteristics of
Ultrasonic Testing Instruments.
ASTM E-2700
Standard Practice for Contact Ultrasonic Testing of Welds Using
Phased Arrays.
ASTM E-2491
Standard guide for evaluating performance characteristics of
phased-array ultrasonic testing instruments and systems.
ASME
2235
Code
case
Use of ultrasonic examination in lieu of Radiography, Section I,
Section XII.
ASTM E-2192
Standard guide for planar flaw height sizing by ultrasonics.
ASTM E-1065
Standard guide for evaluating characteristics of ultrasonic
search unit.
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
Advances in Phased Array Ultrasonic Technology Applications,
By Olympus NDT, Published by Olympus NDT., 2007, ISBN 0-9735933-2-6
Automated Ultrasonic Testing for Pipeline Girth Welds, A Handbook,
By E.A.Ginzel, Published by Olympus NDT, 2006, ISBN 0-9735933-4-2
NDT44-60319
NDT Prelims
5
Copyright © TWI Ltd
COSHH Regulations Relevant toTWI Training and Examination Services
Introduction
The use of chemicals in NDT is regulated by law under the Control of Substances
Hazardous to Health (COSHH) Regulations 2005. These regulations require the School to
assess and control the risk of health damage from every kind of substance used in
training. Students are also required by the law to co-operate with the School’s risk
management efforts and to comply with the Control Measures adopted.
Hazardous Data Sheets
The School holds Manufacturers Safety Data Sheets for every substance in use. Copies
are readily available for students to read before using any product. The Data Sheets
contain information on:







The trade name of the product for example Magnaglo, Ardrox.
Hazardous ingredients of the products.
The effect of those ingredients on people’s health.
The hazard category of the substance for example irritant, harmful, corrosive or
toxic.
Special precautions for use for example the correct Personal Protective Equipment
(PPE) to wear.
Instructions for first aid.
Advice on disposal.
EH40: Occupational Exposure Limits
What is exposure?
Exposure to a substance take into the body. The exposure routes are by:




Breathing fume, dust, gas or mist.
Skin contact.
Injection into the skin.
Swallowing.
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
NDT44-60319
NDT Prelims
6
Copyright © TWI Ltd
Safety and Environmental Requirements
Phased array UT will require the use of couplant and possibly cleaning agents some of
which may be hazardous to health. Prolonged or repeated contact of such materials with
skin or mucous membranes shall be avoided.
Testing materials shall be used in accordance with manufacturer’s instructions. National
accident prevention, electrical safety, handling of dangerous substances, personal and
environmental protection regulations shall be observed at all times.
Cautions and Warnings
Some of the test samples used on the ultrasonic courses are heavy and can become
slippery when covered in couplant. Care should be taken when handling samples and
suitable PPE, in particular; safety footwear and barrier cream shall be used.
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.
Safe Manual Handling
The initial approach to avoiding the hazards presented by manual handling is to
investigate whether the need for manual handling can be eliminated or reduced.
NDT44-60319
NDT Prelims
7
Copyright © TWI Ltd
NDT Certification Schemes
CSWIP: Certification Scheme for Personnel
Managed by TWI Certification Limited, a TWI Group company
formed in 1993 to separate TWI’s activities in the field of
personnel and company certification thus ensuring continued
compliance with international standards for certification bodies
and is accredited by UKAS to ISO 17024.
TWICL establishes and implements certification schemes, approves training courses and
authorises examination bodies and assessors in a large variety of inspection fields,
including: Non-destructive testing (NDT), welding and plant inspectors, welding
supervisors, welding coordination, plastic welders, underwater inspectors, Integrity
management, general inspection of offshore facilities, cathodic protection, heat
treatment.
TWI Certification Ltd
Granta Park, Great Abington, Cambridge CB21 6AL, UK
Tel: +44 (0) 1223 899000
Fax: +44 (0) 1223 894219
Email: twicertification@twi.co.uk
Website: www.cswip.com
PCN: Personal Certification in Non-destructive Testing
Managed and marketed by the British Institute of Non-Destructive
Testing (BINDT) which owns and operates the PCN certification
scheme, it offers a UKAS accredited certification of competence for NDT
and condition monitoring in a variety of product sectors.
The British Institute of Non-Destructive Testing
1 Spencer Parade, Northampton, NN1 5AA, UK
Tel: +44 (0)1604 259056
Fax: +44 (0)1604 823725
Email: pcn@bindt.org
Website: www.bindt.org/Certification/General_Information
Both schemes offer NDT certification conforming to both EN 3 and ISO 9712;
Qualification and Certification of NDT personnel.
NDT44-60319
NDT Prelims
8
Copyright © TWI Ltd
The PCN Scheme
A summary of the general requirements for qualification and PCN certification of NDT
personnel as described in PCN/GEN Issue 5 revision R.
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.
PSL/8A
PCN documents: Issue status.
PSL/30 Log of pre-certification experience.
PSL/31 Use of PCN and UKAS Logo.
PSL/42 Log of pre-certification on-the-job training.
PSL/44 Vision requirements.
PSL/49 Examination exemptions for holders of certification other than PCN.
PSL/51
Acceptable certification for persons supervising PCN candidates gaining
experience prior to certification.
PSL/57C
Application for certification, experience gained post examination.
PSL/67 Supplementary 56 day waiver.
PSL/70 Request for L2 certificate issue to a L3 holder.
CP9
Requirements for BINDT Authorised Qualifying Bodies.
CP16
Renewal and recertification of PCN Levels 1 and 2 certificates.
CP17
Renewal and recertification of PCN Level 3 certificates.
CP19
Informal access to Authorised Qualifying Bodies by third parties.
CP22
CP25
Marking and grading PCN examinations.
Guidelines for the preparation of NDT procedures and instructions in PCN
examinations.
CP27
Code of Ethics for PCN certificate holders.
NDT44-60319
NDT Prelims
9
Copyright © TWI Ltd
Levels of PCN certification
Level 1 personnel are qualified to carry out NDT operations according to written
instructions under the supervision of appropriately qualified level 2 or level 3 personnel.
Within the scope of the competence defined on the certificate, level 1 personnel may be
authorised by the employer to perform the following in accordance with NDT
instructions:




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 2 personnel have demonstrated competence to perform and supervise non
destructive testing according to established or recognised procedures. Within the scope
of the competence defined on the certificate, level 2 personnel may be authorised by the
employer to:










Select the NDT technique for the test method to be used.
Define the limitations of application of the testing method.
Translate NDT standards and specifications into NDT instructions.
Set up and verify equipment settings.
Perform and supervise tests.
Interpret and evaluate results according to applicable standards,
specifications.
Prepare written NDT instructions.
Carry out and supervise all level 1 duties.
Provide guidance for personnel at or below level 2.
Organise and report the results of non-destructive tests.
codes
or
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.
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 personnel have demonstrated:



A competence to interpret and evaluate test results in terms of existing codes,
standards and specifications.
Possession of the required level of knowledge in applicable materials, fabrication and
product technology sufficient to enable the selection of NDT methods and techniques
and to assist in the establishment of test criteria where none are otherwise available.
A general familiarity with other NDT methods.
NDT44-60319
NDT Prelims
10
Copyright © TWI Ltd
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
ET
PT
MT
RT³
RI
UT
VT
BRS
RPS
Basic knowledge
Level 1 hours
Level 2 hours
Level 3 hours
40
40
40
16
24
24
16
24
32
40
80
72
N/A
56
N/A
40
80
72
16
24
24
16
N/A
N/A
N/A
24
N\A
(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.
NDT44-60319
NDT Prelims
11
Copyright © TWI Ltd
Industrial NDT experience
Industrial NDT experience in the appropriate sector may be acquired either prior to or
following success in the qualification examination.
In the event that the experience is sought following successful examination, the results
of the examination shall remain valid for up to two years.
Documentary evidence (in a form acceptable to the British Institute of NDT, ie on PCN
form PSL/30) of experience satisfying the following requirements shall be confirmed by
the employer and submitted to the BINDT AQB prior to examination or directly to BINDT
prior to the award of PCN certification in the event that experience is gained after
examination.
Table 2: Minimum duration of experience for certification
NDT method
ET
MT
PT
RT
UT
RI
VT
Level 1
3
1
1
3
3
N/A
1
Experience, months
Level 2
9
3
3
9
9
6
3
Level 3
18
12
12
18
18
N/A
12
Work experience in months is based on a nominal 40 hours/week or the legal week of
work. When an individual is working in excess of 40 hours/week, he may be credited
with experience based on the total hours but he shall be required to produce evidence
of this experience.
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
ET
PT
MT
RT
RI
UT
VT
Level 1
40
30
30
40
N/A
40
30
BRS
30
RPS
N/A
Note: All level 1 specific theory papers have 30 questions.
All level 2 specific theory papers have 36 questions.
Level 2
40
40
40
40
40
40
40
N/A
20 plus 4 narrative
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 reexamination 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.
NDT44-60319
NDT Prelims
12
Copyright © TWI Ltd

A candidate who achieves a passing grade of 70% in each of the examination parts
(general, specific or practical) but whose average score is less than the required 80%
may be re-examined a maximum of two times in any or all of the examination parts
in order to achieve an overall average score of 80%, 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.
2.
3.
4.
5.
6.














Aerospace.
Castings.
Welds.
Wrought products and forgings.
Pre and in-service inspection (multi sector).
Railway.
There are many additional supporting documents varying from vision requirements
PSL44 to renewal and recertification (levels 1 and 2 – CP16; level 3 – CP17) and so
on.
The document defines many terms used in certification of NDT personnel (PCN
general section 3).
The certification body (BINDT) meets the requirements of ISO17024 (PCN general
section 5).
BINDT approves authorised qualifying bodies (AQBs) to carry out the examinations
(PCN general section 5).
The document sets out the levels of PCN certification and what each level of
personnel is qualified to do (PCN general section 6). There are three levels of PCN
certification.
Candidates for examination must have successfully completed a BINDT validated
course of training at a BINDT authorised training organisation (PCN general section
7).
Table 1 shows the minimum required duration of training for all levels and methods
plus a section of notes.
Table 2 gives the minimum duration of experience for each level and method.
A candidate is required to have a vision test of colour perception and a near vision
test (Jaeger number 1 or N4.5) (PCN general section a) – the near vision test to be
taken annually.
Examination applications are made directly with the AQB.
PCN levels 1 and 2 initial exams comprise general, specific and practical parts.
Table 3 shows the number of general questions at levels 1 and 2 examinations.
There are 30 specific questions on the level 1 papers.
There are 36 questions on the level 2 specific papers.
NDT44-60319
NDT Prelims
13
Copyright © TWI Ltd


A variety of practical samples are tested depending on the method and sector.
A level 3 exam comprises of a basic exam and a method exam however the basic
exam needs to be passed only once. Table 4 shows the number of basic exam
questions. Table 5 shows the number of level 3 exam questions.
Table 4: Number of basic examination questions
Part
Examination
Number of
questions
A
Materials technology and science, including typical defects
in a wide range of products including castings welds and
wrought products.
30
B
Qualification and certification procedure in accordance with
this document
10
C
15 general questions at level 2 standard for each of four
NDT methods chosen by the candidate, including at least
one volumetric NDT method (UT or RT).
60
Table 5: Main method examination



Part
Subject
Number of
questions
D
Level 3 knowledge relating to the test method applied
30
E
Application of the NDT method in the sector concerned,
including
the
applicable
codes,
standards
and
specifications. This may be an open book examination in
relation to codes, standards and specifications.
20
F
Drafting of one or more NDT procedures in the relevant
sector. The applicable codes, standards and specifications
shall be available to the candidate.
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.
NDT44-60319
NDT Prelims
14
Copyright © TWI Ltd
The CSWIP Scheme
CSWIP certification is a scheme which covers the qualification of NDT inspection staff to
meet the requirements of European and International Standards. The following
document describes how the CSWIP NDT system works.
DOCUMENT No. CSWIP-ISO-NDT-11/93-R: Requirements for the certification of
personnel engaged in non-destructive testing in accordance with the requirements of EN
473 and ISO 9712 (5th Edition October 2008)
The document can be viewed on the CSWIP website at www.cswip.com/pdfs/cswip-isondt.pdf
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.
Summary of the CSWIP scheme







It is managed and administered by TWI Certification Ltd on behalf of its stakeholders.
It meets or exceeds the criteria of EN473 and ISO9712.
There are 11 parts covering various industry and product sectors.
The certification body meets the requirements of ISO17024.
The CSWIP-ISO-NDT document sets out the levels of certification and what each level
of personnel is qualified to do.
There are three levels of CSWIP certification.
Candidates for examination must have successfully completed validated course of
training at an authorised training organisation.
The CSWIP-ISO-NDT-11/93-R document details:












Minimum required duration of training for all levels and methods plus a section of
notes.
Minimum duration of experience for each level and method.
Requirements for candidates to have a vision test of colour perception and a near
vision test (Jaeger Number 1 or N4.5) to be taken annually.
Applications for examination are made directly with the AQB.
Certification available referencing the relevant document parts.
Level 1 and level 2 initial exams which comprise of general, specific and practical
parts.
The number of general questions at level 1 and 2 examinations.
Level 3 exams comprising of a basic exam and a method exam however the basic
exam needs to be passed only once.
A pass is obtained where each part is 70% or over with an average grade of 70% or
over.
A CSWIP certificate is valid for 5 years.
Renewal and recertification requirements.
Level 3 structured credit system for renewal.
NDT44-60319
NDT Prelims
15
Copyright © TWI Ltd
Phased Arrays and Mechanised Ultrasonic Testing
A Training Handbook
E. Ginzel
Copyright 2002-2008
Third Edition
2008
by
E.A.Ginzel
Prometheus Press
Waterloo, Ontario
Canada
Copyright 2002-2007 by E.A.Ginzel
Copyright 2008 by E.A.Ginzel and TWI Ltd.
All rights reserved.
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.
NDT44-60319
NDT Prelims
16
Copyright © TWI Ltd
Preface to the third edition
This is the third edition of this handbook. Originally this book was intended to provide a
supplemental reference for people taking training in phased array inspection of pipeline
girth welds using the zonal discrimination technique. The first edition contained aspects
of background that were medically related and not totally pertinent to the industrial
applications of NDT.
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.
This edition is intended to provide supplemental information to phased array training,
however, it is not geared to address just the girth weld inspection using the zonal
discrimination technique. Instead, it is to provide a more general overview of applied
industrial phased array inspections.
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.
This includes what should be a refresher on the fundamentals of ultrasonic transducer
and sound field theory. Concepts associated with the traditional single-element
ultrasonic testing are shown to have equivalents applicable to phased array applications
in NDT.
NDT using phased arrays is usually associated with mechanisation and digital controls so
a brief coverage on these main aspects is provided.
An essential aspect of phased array systems is the complexity of displaying and
analysing the results of acquiring data in the form of A-scans using many different
angles and delays. Some examples of field data collection and analysis are provided.
To provide a useful method of assessment of students’ understanding, several chapters
contain problems that are either based on basic knowledge covered in the chapter or
practical mathematic solutions using the fundamental equations used in ultrasonic
testing.
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.
NDT44-60319
NDT Prelims
17
Copyright © TWI Ltd
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.
This handbook is intended to improve students’ understanding of phased array aspects
of mechanised ultrasonic inspections. Some background is supplied with regard to the
ancillary aspects of mechanised ultrasonic inspections, data acquisition and computer
imaging.
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 postprocessing enhancements that can be used with phased array systems.
NDT44-60319
NDT Prelims
18
Copyright © TWI Ltd
Section 1
History
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.
It is speculated that the principle of phased-arrays had probably been known
much earlier as they related to submarine warfare. As a result of the military
aspect the technology was kept secret. Around the same time parallel work by
DG Tucker at the Birmingham University in the UK was published.
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 timingcircuitry 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.
Until design considerations were better understood, early versions of medical
phased array probes suffered from artefacts. These are signals that occur due
to portions of the off-axis beam interacting with off-axis features not intended
to be detected in the vicinity of the beam. These resulted from grating lobes
which result from the beams that emanate at predictable angles off-axis to the
main beam.
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.
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.
NDT44-50816
History
1-1
ww.twitraining.com
Copyright © TWI Ltd
In addition to eliminating ghost images, eliminating grating lobes also improves
the signal-to-noise ratio by ensuring that the main beam energy can be
maintained over a wider range of angles thereby assuring better directivity
relative to the background energy.
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.
He accomplished this by heterodyning (using the beat frequency) the RF signal
from each channel to an intermediate frequency, then a phase manipulating
circuit could perform fine delay changes while a coarse delay line held a large
constant delay. In this way the imaging system created multiple-receive foci
without expensive ultra-low noise delay lines.
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.
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.
NDT44-50816
History
1-2
ww.twitraining.com
Copyright © TWI Ltd
Figure 1.1 3-D rendering of a foetal face.
Image courtesy Toshiba Medical (www.medical.toshiba.com).
In late 2000 Siemens developed their 3D imaging system. The following is a
report
from
the
newspaper
the
Telegraph,
in
the
UK
(http://www.telegraph.co.uk.).
NDT44-50816
History
1-3
ww.twitraining.com
Copyright © TWI Ltd
Revolutionary: a new ultrasound scanner can produce 3D
images of an unborn baby
Figure 1.2 Ultrasound scan image and photography.
A photo for the album before baby is born
By Roger Highfield, Science Editor
(Filed: 30/01/2001)
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.
The three-dimensional scanner, launched by Siemens yesterday, uses a
conventional ultrasound transducer to make an image which is rendered from a
two-dimensional image by software performing 100 billion calculations a
second. Dr Rose de Bruyn, from Great Ormond Street Hospital, London, said
“Incredibly detailed images of the unborn foetus can now be readily available at
the patient's bedside using these imaging techniques”.
NDT44-50816
History
1-4
ww.twitraining.com
Copyright © TWI Ltd
Section 2
The Principles of Phasing
2.
The Principles of Phasing
Ultrasonics has its foundations in wave mechanics. There are a number of
phenomena associated with waves. Waves can be: reflected, refracted,
diffracted and polarised. Interference of waves can also occur.
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.
Coherent wave sources:
These are sources which have the same frequency and are in phase with each
other.
NDT44-50816
The Principles of Phasing
2-1
Copyright © TWI Ltd
It is the principles of interference of waves that is the central concept in phased
array ultrasonics. Waves can be added together either constructively or
destructively.
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.
NDT44-50816
The Principles of Phasing
2-2
Copyright © TWI Ltd
Add to a maximum
1
2
Add to a null value
3
4
5
6
7
Figure 2.1 Wave Interference with different wavelengths.
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 then
the opposite moving particle displacements would cancel each other and no
pulse occurs.
NDT44-50816
The Principles of Phasing
2-3
Copyright © TWI Ltd
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).
Figure 2.2 Wave Interference with same wavelengths.
Ultrasonic phased array technology is simply a special application of traditional
single element ultrasonic testing. Strictly speaking it may be thought of as
having its principles based on Huygens’ Principle.
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).
The concept that the wavefront is composed of wavelets (secondary waves) is
fundamental to phased-array ultrasonics. The fact that we still concern
ourselves with the main wavefront means that all the other aspects of
ultrasonic treatment apply too; ie we can still consider the Near Zone
calculations, focal spot calculations and divergences as would be associated
with single element systems.
In traditional (single element) ultrasonics we normally use a compression mode
transducer. This is polarised to cause an expansion when a voltage is applied.
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.
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.
NDT44-50816
The Principles of Phasing
2-4
Copyright © TWI Ltd
-----------
++++++++
Poled ceramic
No applied charge
Peak positive cycle
A single wavefront
forms with
++++++++
----------Peak negative cycle
Expansion condition
displaces particles
along the element
surface
diffraction bending
at the edges
A displacement
point exists at
every domain point
Unstressed mode
Stressed mode
Figure 2.3 Single Piezo-Element operation.
In a phased array probe, the same voltage is applied to the same expansional
deforming of a piezoelectric material. The probe face is not a single element but
instead many small elements all connected via the same sort of wiring as a
single element and all mounted to the same backing.
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.
Plane wave formation
Interference wavelets
Unstressed mode
Stressed mode
Figure 2.4 PA concept with multiple elements and wavelets.
NDT44-50816
The Principles of Phasing
2-5
Copyright © TWI Ltd
The actual result of this concept is seen using photoelastic visualisation. Figure
2.5 is a series of exposures showing the effect of pulsing adjacent elements
simultaneously.
Figure 2.5 Photoelastic verification of constructive interference of wavelets.
From left to right we show incremental increases in the number of elements
used, from 5, to 10, to 20 to 25 elements. Each image is taken at 15mm from
the glass surface (probe immediately above) and the weaker wavelets from the
shear diffraction arcs can be seen between the probe and compression
wavefront.
Figure 2.6 illustrates the traditional single element probe with backing and
electrical contacts.
Figure 2.6 Internals of the traditional single element probe.
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.
However, in preparing the probe, the element starts as a single rectangular unit
potted in the backing and a diamond dicing saw makes cuts about as deep as
the piezo material is thick. That gap is filled with a damping material to stop
cross-talk.
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.
NDT44-50816
The Principles of Phasing
2-6
Copyright © TWI Ltd
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.
Arranging the electrodes on such small elements is problematic so contacts are
made similar to printed circuits and may be on flexible backings. With many
contact points pre-made and aligned along the edge (as in the image on the
right of figure 2.7) a matching multi-pin connector can be fitted to the contact
pins protruding from either face of the element structure.
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.
As more precision is required more elements are added. Miniaturisation of the
process is becoming a specialty in the medial field. Figure 2.8 illustrates two
phased array probes designed in Duke University.
The Center for Emerging Cardiovascular Technologies at Duke University
designed a real-time volumetric scanner for imaging the heart. The team in
collaboration with the Volumetric Medical Imaging Inc. at Durham, North
Carolina produced a 40MHz 1.2µ chip completed in 1994 that was the basis for
the beam-former in the world's first electronically steered matrix-array 3-D
ultrasound imager.
This development uses MEMS (micro-electro-mechanical system) technology to
achieve a 64x64 (4096) array the size of a small coin (about 18mm diameter).
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.
NDT44-50816
The Principles of Phasing
2-7
Copyright © TWI Ltd
Figure 2.8 Small scale large arrays, courtesy Duke University CECT.
NDT44-50816
The Principles of Phasing
2-8
Copyright © TWI Ltd
Section 3
Beam Forming
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
Figure 3.1 Array types.
When the electronics are arranged to simply pulse individual or groups of
elements in some order, the array is said to be sequenced. A linear sequenced
array might simply step through pulsing each element one at a time while the
other elements were receiving or the individual element is used in pulse-echo
mode.
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).
Using standard equipment to pulse the ring-sections sequentially by means of
multiplexing, a series of focal spots corresponding to the parameters of each
ring were obtained.
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.
If several elements are pulsed simultaneously the wave front produced is the
result of interference of the various spherical waves from each element. The
effect is a wavefront similar to one transmitted by a plane element having the
same dimensions as the multi-element array.
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
NDT44-060319
Beam Forming
3-1
Copyright © TWI Ltd
www.twitraining.com
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.
Figure 3.2 Beam steering.
By the same principles, focusing can be achieved by delaying the pulse to the
inner elements. This requires non-linear delays as shown in figure 3.3.
Figure 3.3 Beam focusing.
Arrays using the phase interference resulting from timing of pulses to achieve
beam steering or focusing are called phased arrays. Phased arrays can
accomplish two important features by dynamic changes to the delay times to
the elements;
1
2
NDT44-060319
Beam Forming
Beam steering.
Dynamic focusing.
3-2
Copyright © TWI Ltd
www.twitraining.com
Using precise timing and phase interference effects, we can constantly vary
both incident angle and zone of maximum resolution.
It is of course possible to combine the beam steering and focusing, see figure
3.4.
Figure 3.4 Phased Array steering and focusing combined.
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.
Figure 3.5 The Rho-Theta phased array probe.
Each numbered segment of the illustration represents a separate element in the
array (61 elements are indicated in figure 3.5).
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.
NDT44-060319
Beam Forming
3-3
Copyright © TWI Ltd
www.twitraining.com
Matrix arrays can scan in two dimensions and offer considerably more flexibility,
albeit at a price.
Circular or annular arrays are specific for normal beam inspections, eg billets,
forgings. Rho-theta arrays offer the maximum resolution of spherical focusing
with beam steering but the complexity and size make their construction tedious
and costly.
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.
Before moving on to the specifics of how we might address the use of a phased
array system it will be helpful to recall some of the basics of ultrasonic beam
characterisations.
3.2
The ultrasonic beam
As mentioned earlier, the sound field produced by a phased array probe can
have the same quantitative treatment as the single element versions.
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.
The software calculates the transmission pressure so the traditional –6dB used
for pulse-echo analysis of beams is represented by the –3dB when calculated in
calculated transmission-only format.
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.
This calculates sound pressure variation which is squared to show in dB of
reflection. The Radial plot is taken at the end of the Near Field (179.9mm) and
is indicated to be 3.2mm diameter.
NDT44-060319
Beam Forming
3-4
Copyright © TWI Ltd
www.twitraining.com
Figure 3.6a The sound field, Axial beam.
Figure 3.6b The sound field, Radial beam.
Calculations of the peak and boundary conditions of beams play an important
role in ultrasonic testing in general but they are especially important in planning
the phased-array beam. Equations for the determination of the most common
parameters are found in most basic ultrasonic testing guides.
Much of what can be done with a phased-array system in NDT takes advantage
of the ability to alter some of these parameters the next portion of this text
reviews the main items calculated for ultrasonic beams. These include:




NDT44-060319
Beam Forming
Near zone.
Beam diameter.
Focal zone.
Beam spread and half angle.
3-5
Copyright © TWI Ltd
www.twitraining.com
Focusing considerations such as effects of varying the acoustic velocity and gain
(sensitivity) due to focusing will also be considered.
3.2.1
The near zone
The location of the last maximum is known as the near field distance (N or Y0)
and is the natural focus of the transducer. The far field is the area beyond N
where the sound field pressure gradually drops to zero.
The near field distance is a function of the transducer frequency, element
diameter/energised area of array and the sound velocity of the test material as
shown by Equation 1:
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 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.
3.2.2
Beam diameter
An inspection sensitivity is affected by the beam diameter at the point of
interest. The smaller the beam diameter the greater the intensity of energy that
is reflected by a flaw at a particular position. The -6dB pulse-echo beam
diameter at the focus can be calculated with Equation 2 or 2a. For a flat
transducer we use Equation 2a with SF =1.
Equation 2
DB 6 dB 
1.02Fv
fD
Equation 2a
DB6dB  0.2568DS F
NDT44-060319
Beam Forming
3-6
Copyright © TWI Ltd
www.twitraining.com
Where:
DB-6dB
F
v
f
D
SF
3.2.3
=
=
=
=
=
=
Beam diameter (at the 6dB drop boundary).
Focal length.
Material sound velocity.
Frequency.
Element diameter.
Normalized focal length (see Equation 6).
Focal zone
For a focused probe the starting and ending points of the focal zone are located
where the on-axis pulse-echo signal amplitude drops to -6dB of the amplitude
at the focal point. This was termed the working field when describing the near
zone in Section 3.2.1.
Equation 3 gives the length of the focal zone:
Equarion 3
 
N S SFF22[ [ 22 ] ]
FF
ZZN
1100.5.5
SSFF
Where:
FZ = Focal zone.
N = Near field.
SF = Normalized focal length (see Equation 6).
3.2.4
Beam spread and half angle
All ultrasonic beams diverge. In other words, all transducers have beam spread.
Figure 3.7 gives a view of a sound beam for a flat transducer. In the near field
the beam has a complex shape that narrows. In the far field the beam diverges.
Figure 3.7 Principles of beam spread, from www.ndt.net
NDT44-060319
Beam Forming
3-7
Copyright © TWI Ltd
www.twitraining.com
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
Sin( ) 
2
0.514v
fD
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.
3.2.5
Focusing configurations
Traditional single element transducers are available in three different focusing
configurations:
1 Unfocused (flat).
2 Spherically (spot) focused.
3 Cylindrically (line) focused.
For the traditional single element probes focusing is accomplished by either the
addition of a lens or by curving the element itself.
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.
In an unfocused transducer, this occurs at a distance from the face of the
transducer which is approximately equivalent to the transducer’s near field
length. The last signal maximum occurs at a distance equivalent to the near
field, a transducer, by definition, cannot be acoustically focused at a
distance greater than its near field.
When focusing a transducer, the type of focus (spherical or cylindrical), focal
length and the focal target (spherical point or flat surface) need to be specified.
Based on this information, the radius of curvature of the lens or the transducer
can be calculated. This varies based on above parameters. When tested, the
measured focal length should be stated as being determined from the target
specified.
There are limitations on focal lengths for transducers of particular frequency,
element diameter combinations and target designations. The maximum
practical focal length for a point target focal designation is 0.8*(multiply) Near
Field length.
NDT44-060319
Beam Forming
3-8
Copyright © TWI Ltd
www.twitraining.com
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.
In addition to the limitations on maximum focal lengths, there are limitations on
the minimum focal lengths. With phased array probes the minimum focal length
is normally taken as 10% (0.1) of the natural Near Zone. These limitations are
typically due to the mechanical limitations of the transducer.
3.2.6
Focal length variations due to acoustic velocity differences
The measured focal length of a transducer is dependent on the material in
which it is being measured. This is due to the fact that different materials have
different sound velocities.
F
WP
MP
When specifying a conventional transducer’s focal length it is typically specified
for water. Since most materials have a higher velocity than water, the focal
length is effectively shortened. This effect is caused by refraction (according to
Snell’s Law) and is illustrated in figure 3.8.
Figure 3.8 Focal length alteration due to change in material.
This change in the focal length can be predicted by Equation 5. For example,
given a particular focal length and material path, this equation can be used to
determine the appropriate water path to compensate for the focusing effect in
the test material.
Equation 5
Wp  F  M d (
vm
)
vw
Where:
Wp = Water path.
Md = Material depth.
F
= Focal length in water.
vm = Sound velocity in the test material.
vw = Sound velocity in water.
NDT44-060319
Beam Forming
3-9
Copyright © TWI Ltd
www.twitraining.com
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.
3.2.7
Focusing gain
Focused immersion transducers use an acoustic lens element curvature to
effectively shift the location of the near zone toward the transducer face. The
end result can be a dramatic increase in sensitivity, at the focal point.
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.
Figure 3.9 Increased sensitivity by focusing, courtesy Panametrics.
Note: The amplitude from a small defect cannot exceed the echo amplitude
from a flat plate.
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 pulseecho 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.
NDT44-060319
Beam Forming
3-10
Copyright © TWI Ltd
www.twitraining.com
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.
3.3
Phased array aspects of beam characteristics
The forgoing information should be a simple matter of refresher material
covering the basics of ultrasonic sound field equations.
The concepts covered above have equivalents in phased array sound field
calculations;
eg a 5 MHz linear array phased array probe is constructed of 60 elements each
10mm wide and cut with a centre of 1mm spacing and a gap of 0.1mm.
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)
We can therefore use:
Equation 1
N
D2 f
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.
NDT44-060319
Beam Forming
3-11
Copyright © TWI Ltd
www.twitraining.com
Even then we should look at the very first approximation made for near zone
distance. Equation 1 stated N =D2 f/4v
This equation is derived from the more familiar:
Equation 1b
N
D 2  2
4
Equation 1 was made assuming that the dimension of the element was 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.
2
Wavelets form their individual wavefronts from each element and as indicated
in figures 3.2, 3.3 and 3.4, the wavelets are essentially circular. Due to their
circular shape the pressure is approximately uniform in amplitude as we move
around the wavefront off-axis from the front of the element.
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.
There will be some obliquity factor resulting in a reduction of pressure on the
wavefront as we move towards 90° from the forward direction so at some point
we do not have very large amplitude on which to build our constructive
interference.
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.
Therefore, to steer to large angles, small individual elements are necessary. It
is generally recommended that when designing a phased array probe for a
specific application the designer must determine the maximum angle the beam
is to be steered. Then the individual element width is established so that at that
angle the beam amplitude from the individual element is reduced by no more
than 6 dB.
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.
NDT44-060319
Beam Forming
3-12
Copyright © TWI Ltd
www.twitraining.com
3.4
Fermat’s principle
For an ultrasonic operator, the important feature of phased arrays is the ability
to direct a beam to a desired location. This uses Fermat’s principle, illustrated in
figure 3.10.
Snell Point
X or Scan axis
Interface
Law scan offset
Focal Point (X,Z)
Depth
Angle
Figure 3.10a/b Fermat’s Principle as applied to phased array ultrasonics.
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).
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.
NDT44-060319
Beam Forming
3-13
Copyright © TWI Ltd
www.twitraining.com
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.
To have an operator carry out such a collection of calculations every time they
want to change a firing sequence for an inspection is not practical. Therefore
this is done using a computer algorithm. Calculating the required delays is
simplified by software calculators, an example is shown in figure 3.11.
Figure 3.11 Phased array delay calculator GUI (RD Tech).
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.
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 controlling computer is commonly referred to as the data acquisition unit.
The calculator is also a programme and may reside in the data acquisition unit.
The calculator allows control over the pulser, receivers and the input
parameters that configured the initial focal law can be seen displayed
numerically and graphically as shown in figure 3.12.
NDT44-060319
Beam Forming
3-14
Copyright © TWI Ltd
www.twitraining.com
Figure 3.12 Pulser and receiver settings on the computer monitor.
A very simple schematic representation of the electronic components used in
the phased array instrument setup is seen in figure 3.13.
Figure 3.13 Phased array equipment schematic.
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.
Figure 3.14 Electronic scan patterns.
NDT44-060319
Beam Forming
3-15
Copyright © TWI Ltd
www.twitraining.com
a
b
c
For linear scans, arrays are multiplexed using the same Focal Law.
For sectorial scans, the same elements are used but the Focal Laws are
changed.
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.
3.5
Phased array beam focusing and steering
This is something of a review as we have already covered the principles in the
review of the beam characteristics and associated equations as they related to
single element probes.
3.5.1
Beam focusing

Focusing coefficient (K) is defined as:
𝐾 =
𝐹
𝑁
Where F = Focal distance, N = Near zone length.

Beam dimension (dst) in steering plane at focal length is given by:
d st  
F
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.
3.5.2
Beam steering
Important aspects of beam steering for phased array probes include:








The capability to modify the refracted angle of the beam generated by the
array probe.
Allows for multiple angle inspections, using a single probe.
Applies asymmetrical (eg linear) focal laws.
Can only be performed in steering plane when using 1D (linear)-arrays.
Can generate both L (compression) and SV (shear vertical) waves using a
single probe.
Steering capability is related to the width of an individual element of the
array.
Steering range can be modified using an angled wedge.
Maximum steering angle (see also below) (at –6dB), is given by:
 st  0.5

e
Where:  = the wavelength, e = the individual element width.
The maximum steering angle is limited by the following factors:
NDT44-060319
Beam Forming
3-16
Copyright © TWI Ltd
www.twitraining.com






Element pitch.
Frequency.
Element width.
Wedge angle.
Number of active elements.
Focal range.
Although a pitch of half a wavelength is the mathematical limiting factor for
beam steering, in practise the other variables will cause noise before the half
wavelength angle is achieved. In particular total internal, reflection and surface
wave generation within the wedge will prevent high steering angles.
As a recommendation the maximum steering angle should be restricted to
approximately 15 degrees either side of the beam nominal angle and the probe
should be tested through the programmed range for acceptable noise levels
before a procedure is finalised.
3.5.3
Electronic (linear) scanning
Electronic scanning is the ability to move the acoustic beam along the axis of
the array without any mechanical movement, see figure 3.15.
Figure 3.15 Electronic scanning.
The beam movement is performed by time multiplexing of the active elements
(repeating the focal law stepping through one element at a time using the next
adjacent element from the start element of one focal law as the next start
element).
Scanning extent is limited by:


Number of elements in array.
Number of channels in the acquisition system.
By multiplexing of the active elements in different order or by alternating the
number of active elements in the aperture the lateral resolution of a linear scan
can be increased by adding points between the existing ones in figure 3.15.
This function is called double resolution and it is available on some phased
array instruments.
NDT44-060319
Beam Forming
3-17
Copyright © TWI Ltd
www.twitraining.com
3.5.4
Sectorial scanning
When all aspects of the focal law are held fixed except for the delays to alter
angles so that a range of angles are covered by the beam, the scan pattern is
called a sectorial or S-scan (sometimes also called an azimuthal scan).
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.
Figure 3.16 Sectorial scanning courtesy Olympus NDT.
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.
3.5.5
Compound Scan
Compound scan combines multi angle Sectorial scanning with Electronic
scanning and usually offers more scanning coverage and therefore reduces
scanning time and increases productivity. The beam movement is performed by
time multiplexing of the active elements with different focal laws so that a
range of angles is produced.
3.5.6
Combined beam processing
The phased-array technique allows for almost any combination of processing
capabilities:


NDT44-060319
Beam Forming
Focusing + steering.
Linear scanning + steering.
3-18
Copyright © TWI Ltd
www.twitraining.com
3.5.7
Beam apodization
Beam apodization is an electronic feature that is able to reduce side lobes by
applying lower voltage to the outer elements of the active aperture.
3.5.8
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:






For coarse grained materials lower frequencies are used.
For thin materials and high resolution higher frequencies are selected.
For long path lengths to the areas of interest larger element sizes (aperture
sizes) are selected.
For specific angles and when required that only the shear mode be
generated, a refracting wedge of a specified angle is used.
For vertically oriented flaws a tandem configuration of transmitter and
receiver may be appropriate.
For very accurate defect sizing or high sensitivities a larger aperture is
required.
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).
Total length of a linear array may be a compromise between pitch and steering
capability so as to achieve volume coverage of a weld using angled beams in a
linear scan.
These aspects of phased array probe selection will be looked at again in the
applications section.
NDT44-060319
Beam Forming
3-19
Copyright © TWI Ltd
www.twitraining.com
Section 4
Hardware: Pulsers, Receivers,
Motor Control and Encoders
4
Hardware: Pulsers, Receivers, Motor Control and Encoders
The preceding materials provided an overview of the phasing principles used in
phased array ultrasonic probes. Although it may be possible to hand operate a
phased array probe (as is done in some medical applications) this is not usually
the case in industrial applications.
Note: Recently several manufacturers have introduced portable phased array
ultrasonic instruments. This allows a phased array probe to be used like a fixed
wedge single element probe but some manufacturers have also incorporated
ability to take input from encoders to mechanise the scan and produce B and C
scans.
Probes can be mounted in some form of a holder and some degree of
mechanisation is used to manipulate the probe and collect the ultrasonic signals
during the motion process.
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).
As solid-state electronics became more popular, instruments grew smaller and
eventually developed with a 2-sided electronic structure; one side of the
electronics being digital and the other analogue. Primarily it is the control
section of the instrument that is digital.
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.
Small programmes on EPROM chips are now common on portable instruments
and by stepping through a variety of programmes the instrument can be made
to operate under toggled controls and even display the A-scan as a digitised
representation of the analogue output.
4.1
Advantages of digital control
Some aspects of UT are not practical or possible to make digital; input power
supply and the transmitted and received ultrasound are always analogue.
However, many input controls and some outputs are feasible as digital signals.
Digital controlled ultrasonic instruments have many advantages of the older
analogue units:







Accuracy, time or clock based instead of deflector plates.
Repeatability, exact settings can be recalled.
Storage of settings to memory, all parameters stored.
Speed of setup, simply recall stored parameters.
Signal processing.
Display options, eg projection scans, tomographic presentation of data.
Data recall for reporting.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-1
Copyright © TWI Ltd
www.twitraining.com
There are some limitations to digital instruments which must be recognised:


Digitisation frequency, maximum probe frequency is limited by digitising
frequency.
Digital filtering.
Both will limit the higher frequency response for the instrument.
4.2
Pulsers and receivers
Whether or not the parameters of the pulse to the transducer are digitally
controlled or not the pulse itself is an analogue signal. Similarly, the ultrasonic
vibration that a transducer senses from a reflection generates a voltage across
the transducer that is also analogue.
Normal ultrasonic instruments have a single pulser and receiver. Phased array
systems have a series of pulsers and receivers. Usually sold in multiples of 16
or 32, phased array systems will always have a separate pulser and receiver
connection for every element on an array. Limits on the total number that can
be used in a single focal law may apply.
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.
The operator must select a probe that is suitable to the electronics’ capabilities.
Selecting a phased array probe with 128 elements would not be useable on a
system with only 64 channels (pulser-receivers). However, a pair of 60 element
probes could be used on a 128 channels system. This would leave 8 channels
unused.
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.
Phased array pulser-receivers are an amazing example of miniaturisation. The
pulser-receiver is built on a printed circuit board populated by electronic
components and usually includes TCG and gating circuitry, A/D converter and
time delay circuitry; all in a package that is no bigger than a person’s finger.
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-2
Copyright © TWI Ltd
www.twitraining.com
In ultrasonic testing, maximum amplitude displacement is not always desirable.
When precise timing is needed, as would be for thickness tests on thin wall
material, short duration pulses are best. Even a ringy probe can be made to
dampen its vibration with the correct pulse characteristics.
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.
Spike pulse
Tone burst pulse
Square wave pulse
Figure 4.1 Pulse types.
4.4
Spike pulsers
Figure 4.2 shows the components in a spike pulser.
Thyristor
(switch)
DC power
supply
Transducer
Charging
resistor
Tuned circuit (with
damping resistor)
Charging
capacitor
+
Figure 4.2 The spike pulser.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-3
Copyright © TWI Ltd
www.twitraining.com
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.
4.5
Tone burst
Tone burst pulsers allow maximum energy output from transducers by
adjusting the frequency of the voltage applied. This can be done in several
ways. A chopped voltage from a waveform generator allows selection of pulses
of different shapes, frequencies and durations, as in figure 4.3.
Sine wave
Saw-tooth wave
Bi-directional square wave
Negative square wave
Positive square wave
Change shape
Change frequency
Change ring time and
pause time between pulses
Figure 4.3 Tune burst pulser waveforms.
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.
4.6
Square wave pulsers
Square wave pulsers have become the preferred laboratory style of pulsers.
Similar to the spike pulser, the square wave pulser charges a capacitor which
discharges across the transducer. By holding the switch closed in the circuit for
a controlled amount of time, then rapidly restoring the pulse voltage to zero
causes two displacements of the transducer.
The displacements at the transducer are opposite in phase so by timing the
recovery voltage a constructive interference can be effected between the
original backward moving wave reflected off the probe backing and the second
impulse from the pulser.
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-4
Copyright © TWI Ltd
www.twitraining.com
Figure 4.4 Maximum excitation
When the second pulse is timed to fire on the same phase constructive
interference occurs and maximum excitation is achieved (fig. 4.4).
Figure 4.5 Minimum excitation
When the second pulse is timed to fire on the opposite phase destructive
interference occurs and minimum excitation is achieved (fig. 4.5).
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-5
Copyright © TWI Ltd
www.twitraining.com
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 4.6.
Figure 4.6 Pulse width and how it affects transducer excitation.
In figure 4.6 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.
To the right of the received pulse shape is the signal of a co-polymer
transducer, nominally 30 MHz, using a glass target. Signal amplitude is
relatively low implying that the transducer output is not maximized. However,
an advantage to this signal is that none of the low frequency components of the
transducer are excited and the bandwidth is high.
In the second case in figure 4.6, 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).
Lower frequency components may be added to this signal compared to the first
case but the bandwidth is reduced by the greater output near the resonant
frequency.
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.
Significant output increase can be achieved using a bi-polar square wave
pulser. This provides a voltage that is first negative going (or positive going)
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-6
Copyright © TWI Ltd
www.twitraining.com
and held to a maximum for a time equivalent to half the natural period of the
piezo-element.
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 4.7 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 4.8 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).
The combination of tuneable pulse-width and phasing interference can make for
an increased particle displacement in the test material as much as 3-4 times
that of a single element used with an un-tuned pulser.
Figure 4.7 Single element with spike pulse at 700 volts.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-7
Copyright © TWI Ltd
www.twitraining.com
Figure 4.8 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 4.9 illustrates
this switching.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-8
Copyright © TWI Ltd
www.twitraining.com
PRF & sweep
generator
pulser
PRF & sweep
generator
pulser
receiver
Pulse-Echo connections
receiver
Transmit-Receive connections
Figure 4.9 P-E versus TR connections.
To obtain a signal capable of being displayed and subsequently processed the
received signal caused by the small transducer vibrations must be amplified.
The amplification process is quite involved and also includes filtering and
sometimes attenuation.
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.
Some high pass filtering may be incorporated to improve signal-to-noise ratio
by eliminating some radial mode components of the probe and line interference.
Preamplifier bandwidths are usually flat from about 1-15 MHz and this is not
operator adjustable.
Following the preamplifier, the signal is passed through a broadband
attenuator. This protects subsequent circuitry from saturation and it provides a
means of calibrated adjustment of signal height. Attenuation is usually
equipped with coarse (20dB) and fine (1dB) switching.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-9
Copyright © TWI Ltd
www.twitraining.com
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.
Attenuated signals are passed on to RF amplifiers which can be linear or
logarithmic. Linear amplifiers are those most commonly found on UT
instruments. When using a linear amplifier for the receiver amplitude of a signal
is proportional to receiver voltage.
However, receiver gain control is in dB increments, therefore signal
amplification by 6dB gain doubles the signal height. This limits the range of
useful amplification to about 34 dB (34 dB raises a 2% FSH signal to 100%).
When a logarithmic amplifier is used, the scale is dB linear so each increase of
1dB gain is 1% of the screen height. Expressed another way, the dynamic
range of this logarithmic amplifier is 6.3 times greater than the linear. Some
logarithmic amplifiers can exceed 100 dB dynamic range (ie. 1dB gain results in
something less than 1% FSH).
Frequency filtering can be applied to RF amplified signals. Normally bandpass
filters are used to eliminate noise from higher and lower frequency sources.
These are selectable by the operator and are labelled to correspond to the
centre frequency of the filter.
Normally the bandpass filter is set to correspond to the nominal frequency of
the probe. Wideband filters are also available. Amplitude of signal compared to
the best bandpass filter does not significantly change when wideband is
selected but the signal will often be noisier, see figure 4.10.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-10
Copyright © TWI Ltd
www.twitraining.com
1-5MHz filter
3-7MHz filter
7-15MHz filter
Wideband
filter
Relative
Amplitude
2
6
4
8
10
Frequency (MHz)
Spectrum of a nominal 5MHz transducer
Figure 4.10 Receiver filter.
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.
Gate positions are usually facilitated by auxiliary controls. Gate positions on the
screen are noted by extra traces or markers on the A-scan display.
Gate controls include start and end adjustments, threshold setting (amplitude
at which a signal must reach before alarmed or collected) and positive or
negative settings. If positive gating is used a signal must exceed a set minimum
threshold. If negative gating is used a signal in the gate must fall below the
threshold before alarming.
Typical of positive gating is signal amplitude monitoring for flaw detections.
Typical of negative gating is a coupling monitor using a through transmission
signal that alarms a gate of the coupling reduced and reduces the signal below
a given threshold.
Figure 4.11 shows a digital A-scan display with three gates available. Data
collection options for gated regions may include time, amplitude and waveform.
When time or amplitude is selected a threshold is set by positioning the vertical
level (amplitude) of the gate. When Waveform information is selected there is
no amplitude threshold and the entire waveform over a specified time interval is
collected.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-11
Copyright © TWI Ltd
www.twitraining.com
Time Gate set at 30% FSH threshold
Amplitude Gate set at 25% FSH threshold
Waveform Gate (threshold does not apply)
Figure 4.11 Gating display.
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.
Amplifying the response of more distant echoes avoids the inconvenience of
distance-amplitude-correction curves and allows an alarm threshold at a fixed
percentage of the screen height to be set across the entire screen.
Figure 4.12 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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-12
Copyright © TWI Ltd
www.twitraining.com
TCG points
Figure 4.12 Gating display.
4.10
Data acquisition and automated systems
Laboratory UT instruments are often part of a complex collection of hardware
and software that can be considered a data acquisition system. These can be
used for precise material characterisations in scientific studies or they may be
part of industrial production systems.
Many features of laboratory instruments and data acquisition systems are best
detailed by explaining automated inspection systems. These concepts can be
applied to several methods of NDT so a general outline of concepts will be
covered first.
Collecting information about an object or condition is generally considered data
acquisition. This usually involves collecting information about one parameter
with respect to another, eg monitoring temperature against time.
Data acquisition can be done simply by an operator recording readings
manually. In the temperature example the operator would watch a
thermometer (analogue or digital) and record the values of both temperature
and time at various time intervals. (Records of a single parameter would have
little meaning unless they can be related to something else).
Scientific and engineering applications today require very large numbers of
readings to be taken and these with exacting precision. Several hundreds or
thousands of readings over several hours are easily accomplished using
computers.
When computers are incorporated into a data acquisition process the process
becomes automated. Computers can then be used for not only collecting the
readings but also sorting it and subsequently analysing it.
Advantages of automated systems include:





Speed.
Consistency.
Accuracy.
Repeatability.
Safety.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-13
Copyright © TWI Ltd
www.twitraining.com


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.
4.10.3 Position recording accuracy
Manual operations are considered good if an operator can hold tolerances and
make position measurements in the range of 0.5-1.0mm but automated
systems routinely record locations to micron (10-6m) accuracy.
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.
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.
Accuracy, consistency and repeatability can avoid unnecessary replacement
cost if a test proves no deterioration. Conversely, minor changes detectable
only by the accuracy of an automated system could be the reason for concern
to condemn a component and thereby save millions of dollars by avoiding costly
and untimely catastrophic failure due to an undetected change resulting from a
manual scan.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-14
Copyright © TWI Ltd
www.twitraining.com
The case for advanced automated NDT systems is clear.
4.11
Automated systems in general
In general automated NDT systems have the following:




4.12
A
A
A
A
central control system (usually a computer).
sensor and attached apparatus (the NDT equipment).
means of moving the sensor or part.
means of collecting and displaying the output from the NDT apparatus.
System components
Central to the whole system is the computer. Computers come in various
shapes and sizes. There are two main types of computers, analogue computers
and digital computers.
Analogue computers are somewhat archaic now and rarely found in common
use. They are hardwired devices utilizing current flow and switches to address
their logic functions. For our purposes we will be concerned only with digital
computers.
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.
In some machines it is possible to provide an input to the NDT instrument,
thereby facilitating computer control of some of the instrument functions. This
can allow for remote control.
Motion control is of various sorts and can be as simple as switching a drive
motor on or off. It may involve complex closed loop systems controlling position
and velocity based on a feedback monitoring system measuring torques so as
not to break the inspection tool by over-straining it.
For effective data display some form of positional information must be added to
the motion control. This is often accomplished by counting steps on a stepper
motor, simple timing or most accurately by use of positional encoders.
The data display output is merely a means of providing a useful record of the
test results. This might be as simple as a voltage output taken from the NDT
instrument and displayed as a line of varying position with increasing time (a
strip chart recording) or it can be a complex set of data points showing
waveforms collected and processed for display with positional information.
The information collected by computer can be further processed to:



Reduce noise (eg signal averaging).
Enhance pertinent signals (amplitude colouring or signal processing).
Correct for geometric characteristics (eg SAFT).
Figure 4.13 illustrates the basics of an automated system showing the computer
as central to all the inspection activity. A PC style computer addresses motion
control and collects analogue information from the NDT inspection instrument.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-15
Copyright © TWI Ltd
www.twitraining.com
The results are displayed on the computer monitor with the option to print to a
printer.
Printer
PC with
A/D board and
motor control
Motion Control
interface
Inspection Instrument
Probe
Test Piece
Turntable
Y slide
Figure 4.13 An automated scanning rig.
4.13
Instrument outputs
Physical properties measured in NDT can include; temperature, pH, pressure,
distance, velocity, mass or optical, acoustic and electrical energy. The sensors
used convert these properties to an electrical quantity; voltage, current or
resistance.
The sensors, as a result of changing the physical property to an electrical
quantity are also termed transducers (transducing energy from one form to
another).
Electrical data can be considered a signal or waveform. This is usually a voltage
varying with time. Signals can be either analogue or digital.
Analogue signals are continuous and can change an arbitrary amount in a small
time interval. Computers prefer digital signals. These are discrete values in
specified constant time intervals.
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 4.14.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-16
Copyright © TWI Ltd
www.twitraining.com
A
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 4.14 Analogue to digital shaping.
Figure 4.14 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.
Sampling along the horizontal axis is time-based and a single sample is taken
at each time interval. This value is the peak or average and the closest whole
value that the interval corresponds to in the given time interval is the value
assigned to that point.
Frame C shows the converted digital representation of the analogue signal. The
amplitude axis is left for reference and the amplitude of each sample is
indicated at the bottom of each bar.
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.
The process of changing an analogue signal to the computer friendly digital
signal is called digitisation. The electronic device that accomplishes this is called
an analogue-to-digital converter (ADC) and the associated electronics to
accomplish this conversion is usually incorporated on a printed circuit board
inserted as a card in the computer so the hardware is often called an A to D
card.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-17
Copyright © TWI Ltd
www.twitraining.com
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).
Most people are aware of the reverse process in entertainment devices whereby
music information on a CD or DVD diskette is played back on a set of speakers
(the sound wave output of the speakers is always analogue).
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).
A MUX is used to select which of the analogue signals will be converted at any
given time. Figure 4.15 shows a block diagram of how a variety of analogue and
digital inputs and outputs might be arranged on a computer.
Figure 4.15 Multiplexed inputs and outputs via ADCs and DACs.
Computers used in data acquisition come in a variety of sizes, formats and price
ranges. At one time three groupings applied; microcomputers, minicomputers
and mainframes. These were roughly based on size of memory.
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.
Integral to any automated data acquisition system is the data acquisition
software. Data acquisition software is used to collect data, analyse the data and
display the results. Without the ability to analyse and display the results of data
collection the millions of bits of data that can be collected would be
unintelligible to the average operator.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-18
Copyright © TWI Ltd
www.twitraining.com
Processed data can be output to monitors or printers in the form of tables,
graphs or even be made to duplicate stripchart or oscilloscope (A-scan)
presentations.
Several varieties of specialised software exist in NDT inspection systems. These
are often used in conjunction with specialised instruments. The data acquisition
software collects all aspects of the signals and records all instrument
parameters. Some software also addresses motion control and positional
information.
4.14
Scanning displays and scanning equipment
4.14.1 A-scans
Scan terminology has become slightly more complicated since Robert
McMaster’s NDT handbook was published in 1959. The instantaneous display of
echo amplitude along a time-base is still called an A-scan. This is the image
afforded by all UT scope instruments.
Vertical displacements may be bi-directional (RF display) or mono-directional
(rectified) and the horizontal trace represents elapsed time or distance of
propagation, see figure 4.16.
probe
RF
flaw
Rectified
Scan
Display
Figure 4.16 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.
The amount of time displayed is determined by the length of gate used to
collect information. Intensity or colour may be used to indicate amplitude and
phase (if RF signals are collected).
These concepts are shown in figure 4.17. 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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-19
Copyright © TWI Ltd
www.twitraining.com
Note: The traditional definition of a B-scan referred to a cross-sectional
representation of a test specimen and was independent of probe motion with
respect to beam direction, although most B-scans were made using a zero
degree compression mode and a top and bottom of the part could be
represented by the entry signal and the backwall signal.
This traditional concept is more difficult to imagine when the beam is angulated
but if we maintain the principles of time or distance on one axis and probe
displacement on the other it is still a reasonable treatment of the term B-scan.
Main Bang
probe motion
Time
Far wall interface
flaw
Start
Position
End
Display
Scan
Figure 4.17 Formation of an amplitude B-scan.
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 4.18.
probe motion (raster or zig-zag)
Y
}
Gated region
X
Y
Display
X
Scan
Figure 4.18 Formation of an amplitude C-scan.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-20
Copyright © TWI Ltd
www.twitraining.com
Amplitude may be indicated by intensity (as with shades of grey) or colour
where the screen height is divided into ranges each assigned a different colour,
eg 0-20% blue, 21-50% green, 51-80% orange, 81-100% red.
Alternatively, for a C-scan it is also possible to monitor the position of a signal
in time and assign a colour or greyscale to soundpath distances (or depths).
4.14.4 Other letters for scan presentation types
These were the traditional A, B and C-scans. However variations have been
added to the basics and some of these have resulted in increased use of the
alphabet.
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 4.19.
25mm
0
1
actual flaw
position
2
3
6
5
4
peak amplitude
probe position
Interface signal
flaw signal
1 2 3 4
5
6
opposite wall signal
Figure 4.19 Flaw and probe positions offset on angle scans.
In the above illustration the probe has moved to position 3 (approximately
10mm travel) before the signal from the flaw was peaked. The actual position of
the flaw is some distance ahead of the probe.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-21
Copyright © TWI Ltd
www.twitraining.com
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.
In the example this would be 10(Sin45°) = 10(0.707) or the off-set to the
actual flaw position is 7mm from the encoded position. It is obvious then that
off-set is dependent on the soundpath, deeper flaws being off-set more.
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.
In the B-scans and C-scans just described, we collected amplitude information
in a single plane. If we were to combine the raster scan used for the C-scan
with the point by point A-scan capture used in the B-scan, we can obtain all the
possible information for the volume inspected.
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.
4.14.5 Depth encoded C-scan
If a plan view was plotted with indications exceeding a threshold, typically 5%
FSH, represented as a different colour for each 20% increment of depth, we
could have a depth distribution of flaws. This has been termed a depth-encoded
C-scan, see figure 4.20.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-22
Copyright © TWI Ltd
www.twitraining.com
Figure 4.20 Depth encoded C-scan.
Figure 4.18 is a depth encoded C-scan of a ceramic disc. A 0.1 microsecond
gate was used to monitor the position of small reflectors in a grid pattern.
Lighter shades of grey indicate targets further away and darker shades indicate
targets are closer.
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 4.21) and that must be corrected
for angle and assigned an equivalent depth. This is termed by some a D-scan.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-23
Copyright © TWI Ltd
www.twitraining.com
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.
X-Y Projection (C-Scan)
Y-Z Projection (E-Scan)
X-Z Projection (D-Scan)
Figure 4.21 Project scan nomenclature.
4.14.6 E & P-scans
Other terminology has been used to identify these displays. Depending on the
view and sometimes depending on the software manufacturer, these have also
been called E-scans (for end view) and P-scans (for projection view).
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-24
Copyright © TWI Ltd
www.twitraining.com
A typical S-scan presentation is illustrated in figure 4.22 where the beam is
swept from -5° to +30° and the amplitude responses are colour-coded for the
three side drilled holes. The vertical scale and horizontal scale can be read
directly for depth and horizontal position.
Figure 4.22 S-scan display.
4.15
Memory and digitisation aspects
In the description of analogue to digital conversion it was noted that typically
an 8 bit ADC is used thereby providing 28 or 256 levels of vertical (resolution).
The sampling rate of the ADC will dictate the time interval along the A-scan that
is captured and digitised. Flash A to D converter boards are also available in a
variety of speeds typically 20-100 MHz but slower and faster varieties are also
available.
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.
Figure 4.23 shows an analogue signal from a 10 MHz probe. Digitising at 100
MHz allows reasonable reproduction but at 20 MHz the original analogue trace is
just barely recognizable (the dashed line of the analogue trace is supplied as
reference, only the dots would appear on the scope).
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-25
Copyright © TWI Ltd
www.twitraining.com
10 MHz analogue signal
0.1 s per division
10 MHz digital signal
ADC at 100 MHz
i.e. 10 samples per div.
0.1 s per division
10 MHz digital signal
ADC at 20 MHz
i.e. 2 samples per div.
0.1 s per division
Figure 4.23 Digitising effects on oscilloscope waveform quality.
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.
Temporal or distance resolution is solely a function of ADC rate, amplitude
resolution is a function of both ADC rate and number of levels of sampling for
example, number of bits. For UT data acquisition systems 8 bit sampling is
presently the most common.
An important aspect of digitised amplitude is the effect on dynamic range.
Historically the most common bit rate has been 8 bit digitisation. Accuracy of
amplitude assessment is based on the number of divisions of sampling in the
vertical direction.
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-26
Copyright © TWI Ltd
www.twitraining.com
Computer-based ultrasonic systems have been increasing the bit-rates used
and the higher-end units are sometimes using 12-bit digitisation. The product of
the bit-rate is the number of samples that the vertical (amplitude) range can be
divided into.
For example:



8 bit = 28 = 256 ie 256 intervals of vertical sampling (PipeWIZARD).
10 bit = 210 = 1024 Intervals of vertical sampling (Omniscan).
12 bit = 212 = 4096 Intervals of vertical sampling (Tomo3).
0-256
-128 to 0
0 to + 127
This can be illustrated graphically figure 4.24 shows an RF waveform and a
rectified waveform presented on a graph with a colour code for amplitude on
either side.
8 Bit Rectified Signal (Volumetrics)
8 Bit RF Signal (TOFD)
Figure 4.24 8-Bit digitising effects on dynamic range.
Signal amplitudes are usually stated in dB and the concept of dB is simply a
ratio from dB = 20 log10 h1/h2 where h1 and h2 are the relative amplitudes of
two signals.
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.
Note: Half the amplitude is -6dB so reducing the number of points by a factor of
2 reduces the dB dynamic range by 6 (ie. 48-6=42) and the smallest vertical
screen interval is 0.8%.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-27
Copyright © TWI Ltd
www.twitraining.com
When we use higher Bit-rates there is an interesting bi-product!
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 10bit 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 4.25 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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-28
Copyright © TWI Ltd
www.twitraining.com
Figure 4.25 9-Bit digitising gate display to increase dynamic range.
Digitising an A-scan is the first step in constructing a B-scan. Each sample must
be saved to computer memory, therefore larger scan lengths and larger time of
the gated period, require more memory than small scans and gated times.
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.
We must also consider that 100mm time equivalent is traversed by the
longitudinal wave to cover just the 50mm thickness, hence:
(50 x 2)  5.9 = 16.9 µs.
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).
If an A-scan is collected at 0.5mm intervals, the data generated would be:
270 x 1 x 2 x 100 = 54000 bytes (54 kB)




The first parameter is 270 points per A-scan.
The second is 1 byte per sample point on each A-scan.
The third is the number of A-scans per mm (2).
100 is the length of the scan in mm.
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
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-29
Copyright © TWI Ltd
www.twitraining.com
above conditions, 5.4 MBytes would be generated. Caution should be taken to
collect only that amount of information that is absolutely necessary.
In a multi-channel system where several B-scans (or D-scans) and a ToFD scan
are collected the file size of even a simple linear scan parallel to the weld axis
(with no rastering) could quickly result in file sizes of several tens of Mega
Bytes (MB).
4.16
Data processing
One of the added features of data acquisition systems is the ability to perform
subsequent processing of the stored signals. Since the advent of digital storage,
several techniques have been derived to enhance the information collected. This
process is generally termed digital signal processing
(DSP).
Effectiveness of DSP relies on the quality of the captured signal. Quality
determining factors include:





How well transducer and data acquisition system are matched.
Sampling period.
Signal quantisation level.
Calibration.
Material attenuation.
Any unwanted disturbance in the useful frequency band that is introduced to
the signal is considered noise. Noise may have several sources; the transducer
itself, instrumentation, spurious waves from scatter, geometry and mode
conversions, as well as surrounding electrical noise.
Defects may originate in areas where geometric configurations form stress
raisers or entrapments for chemicals that can lead to corrosion, cracking or
both. The defect occurring in this area may be corrupted or completely masked
by the surrounding conditions. Conversely, geometries may be misinterpreted
as defects.
B-scans, C-scans or other imaging displays allow defect detection by illustrating
the big picture, where subtle trends are noticeable that might not be evident in
the static A-scan display.
In spite of the improved notice-ability afforded by imaging, spurious signals
from noise sources may still mask defects. Various techniques have been
developed to enhance pertinent information to suppress the masking effects of
noise.
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-30
Copyright © TWI Ltd
www.twitraining.com
Digital Signal Processing (DSP)
Two Dimensional
One Dimensional
Filtering
Spectrum analysis
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 4.26 illustrates such a process.
indication
Raw data
Original
Data
Filtering
DSP
Filter process
1 MHz filter
indication
Processed
Data
Enhanced signal
Figure 4.26 Filtering by DSP.
A somewhat simpler form of processing is signal averaging. Signal averaging
allows a flaw signal to be drawn out of the background noise by the principle
that a flaw signal is coherent but noise is not.
A coherent repetitive signal added to itself n times will increase by a factor of n,
whereas noise added to itself n times will increase by the square root of n. After
n iterations the signal to noise ratio of the averaged waveform is improved by
n.
Two dimensional DSP techniques are used to enhance spatial information. As
such, two dimensional DSP is applied to B-scan and C-scan images. It may be
noted that B-scans and C-scans contain no more information than the A-scans
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-31
Copyright © TWI Ltd
www.twitraining.com
used to generate them. However, they provide spatial relationships not possible
to determine from isolated A-scans.
Amplitude averaging is sometimes used on C-scan displays. This tends to
smooth edges and eliminate isolated peaks. The average of a grid (3x3, 5x5,
7x7) is placed at the centre of the grid. The average value may be linear or
weighted. An example of image filtering is shown in figure 4.27.
A tandem configuration was used to scan 3 flat bottom holes. Amplitude
samples in the gated region were taken every 0.5mm on a 0.5mm grid pattern
raster scan. The upper image in figure 4.25 shows the raw image.
Below that is an image formed by a 5x5 non-weighted averaging. Showing the
scan backlash offsets being corrected. A 9x9 convolution filter smoothes the
image further and more closely represents the round shapes that form the FBH
targets.
Figure 4.27 2-dimensional DSP (matrix averaging).
B-scan DSP enhancements are also performed. Of the processing methods used
with B-scans, synthetic aperture focusing technique (SAFT) is the best known.
Transit-time for the ultrasonic beam to travel to and from a point is a hyperbolic
function of the probe position and target depth.
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.
When no defect is present the interference is destructive and the signal is small.
This SAFT processing may be performed in either two dimensions or three
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-32
Copyright © TWI Ltd
www.twitraining.com
dimensions, however,
processing time.
three
dimensional
SAFTing
requires
considerable
An example of the improved signal-to-noise ratio and lateral resolution of SAFT
processing are shown in figure 4.26.
In figure 4.28 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.
Before SAFT processing
After SAFT processing
Figure 4.28 DSP by SAFT.
A variation of SAFT is the so-called ALOK (German: Amplituden und Laufzent
Orts Korwen). The expected travel time hyperbolic curves are used to improve
signal-to-noise ratio of defects however, no synthetic focusing occurs.
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-33
Copyright © TWI Ltd
www.twitraining.com
Correcte
d S-scan
Linear
view
Figure 4.29 DSP using angular correction for S-scan.
4.17
Scanning equipment
Scanning apparatus is required for positional information. Knowing a reflector
exists in a test piece is of little use unless its position can be determined.
Position will be crucial in ascertaining if the reflector is a flaw or geometry; if
the reflector is determined to be a flaw and it occurs in a weld, position will
assist in evaluation and characterisation. Although simple ruler measurements
from surface references are often used for manual scanning, indexing devices
are usually used in mechanised scanning.
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-34
Copyright © TWI Ltd
www.twitraining.com
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.
4.18
Limitations of mechanised scanning
Not all aspects of inspection need be mechanised. There will always be cases
where manual techniques are more cost effective although given unlimited
funding all manual scanning could be mechanised to some degree.
It should be noted that mechanical limitations might apply to mechanised
systems. The most common of the limitations would be scanning speed. Even
when a computer and ultrasonic systems can produce and collect the data at
high rates of travel speed there may be mechanical impediments to moving the
probe(s) at the maximum speed that can be computer collected.
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.
An example of technology advances may be seen in the pipeline girth weld
inspections. Older systems in the 1980s were based on the pulser PRF and were
hard pressed to scan a weld having 6 weld zones at more than 40mm/s.
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 (1015MB) at twice the speed.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-35
Copyright © TWI Ltd
www.twitraining.com
Figure 4.30 illustrates several of the options for scanning equipment. Whether
contact, immersion or gap techniques are used the principles involved are the
same.
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 4.30 Scanning equipment with positional encoders, all images courtesy
RD-Tech.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-36
Copyright © TWI Ltd
www.twitraining.com
4.19
Scanning speed
ADC rates have already been discussed and if large gated times are required for
B-scans, the ability of the computer CPU to process the information may
present a speed limitation. If the computer has not had enough time between
samples blank lines on the B-scan result indicating missed data points.
Similarly, missed data points can result when generating C-scans at too high a
travel speed.
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.
Response times of the recording devices such as strip-chart recorders may
require several pulse signals to ensure the true maximum amplitude is
indicated. Therefore a probe must be in the vicinity of a reflector for a time
sufficient for the recording equipment to respond. This will be in part
determined by the size of the beam and by the size of the calibration or
minimum target.
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.
Empirically established scanning speeds may be found or specification or code
can stipulate maximum speeds based on probe or beam size and PRF.
An example of specification dictated speed states...scanning velocity Vc shall be
determined by:
 PRF 
Vc  Wc

 3 
Where Wc is the narrowest -6dB width at the appropriate operating distance of
the transducer determined by design requirements and PRF is the effective
pulse repetition frequency for each transducer. This example requires 3 firings
within the 6 dB beam width.
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.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-37
Copyright © TWI Ltd
www.twitraining.com
Improved technology makes PRF considerations irrelevant. With the digital
control of data acquisition systems the entire process from firing the element(s)
to collecting, displaying and saving the received signals is all computer
controlled. There is still a master clock and the computer sequences all
activities off this clock.
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.
There is no independent pulser, firing away, oblivious to the data acquisition
system. Instead, the sequence of all events for the scan is queued off the pulse
from the encoder indicating each 1mm of scanner advance.
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:
1
The computer is indicated to record the encoded positions (calibrated in
millimetres, or perhaps inches, as referenced from a specified origin).
2 The computer is told to load the first phased array focal law.
3 Fire the transmitters in the correct order and delay.
4 Arrange the receivers to receive the pulse from the transmitted signal.
5 Apply the correct receiver gains to the applicable channels.
6 Collect the time information from the time gate.
7 Collect the amplitude information from the amplitude gate.
8 Store the amplitude and time to memory.
9 Repeat the above steps for all channels (changing to store waveforms
instead of amplitude and time where applicable).
10 Print one line of displayed data to the monitor.
11 Wait for the next 1mm increment pulse from the encoder and begin again.
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 4.31.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-38
Copyright © TWI Ltd
www.twitraining.com
Figure 4.31 Missing data points.
The equation:
 PRF 
Vc  Wc

 3 
Does not apply for such encoder-triggered data acquisition systems.
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.
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.
Reflective type encoders bounce collimated light off a patterned reflective code
wheel. Fitting all of the electronics of a reflective encoder onto one side of the
code wheel makes it a more compact design than transmissive types.
Figure 4.32 illustrates the components in an encoder using photodiodes to
produce a quadrature encoder output which enables this encoder to display
both distance and direction.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-39
Copyright © TWI Ltd
www.twitraining.com
In this case direction is determined via phase difference between pulses.
Figure 4.32 Optical encoder components.
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.
Figure 4.33 illustrates the pulse pattern of a quadrature encoder that provides
the direction information (with channel C being the reference pulse).
Figure 4.33 Optical encoder quadrature pulse patterns.
Automated scanning systems incorporating optical encoders require calibration.
This involves moving the scanner over a specific distance and counting the
number of pulses. Then a calibration factor is used (number of pulses per
millimetre).
Other positional indicating devices include potentiometers and resolvers.
4.21
Asynchronous versus synchronous systems
Most ultrasonic technicians having used a traditional mono-element probe in
pulse-echo mode are familiar with the concept of PRF (pulse repetition
frequency). This is the rate that the ultrasonic instrument pulses the probe.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-40
Copyright © TWI Ltd
www.twitraining.com
In a multi-channel system where several probes may be used PRF controlled
systems require high clock speeds to ensure that all the channels are fired in
the allowed sample interval.
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 4.34.
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.
Asynchronous multi-element
scanning
Synchronous multi-element
scanning system
Figure 4.34 Asynchronous and synchronous systems.
When using a phased-array system for the ultrasonic pulses it is essential that
the system be synchronous. If we consider a phased-array system, we may
think of each UT channel as a focal law. Then the encoder-pulse triggering the
events must fire all the foal laws prior to the next encoder pulse.
Data acquisition by the asynchronous systems often uses computer algorithms
to select the maximum, minimum or average values of the gated information
received by the ultrasonic instrument between encoder pulses. This value is
then transferred to the computer for data display (and to memory).
Synchronous systems have only a single firing (unless averaging is used) for
each channel so the single gated value (per channel or focal law) is transferred
to memory for display.
A rule of thumb is often used in ultrasonic data acquisition systems. This
requires that at least three firings of the ultrasonic pulse for each channel is had
over a distance equal to the 6dB dimension of the beam.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-41
Copyright © TWI Ltd
www.twitraining.com
In fact, some Codes or Standards actually state this in an equation format:
V
Wc * prf
3
This was addressed in section 4.19 when considering scanning speeds. As noted
there, the equivalent is maintained for synchronous systems by having three
samples within the 6dB beam width. Therefore for synchronous or fire on
position systems one sample every 1mm will achieve this if the beam width is
3mm or greater.
NDT44-060319
Hardware: Pulsers, Receivers, Motor Control and Encoders
4-42
Copyright © TWI Ltd
www.twitraining.com
Section 5
Calibration and Scanning Considerations for Weld
Inspections using Phased Array UT
5
Calibration and Scanning Considerations for Weld Inspections
using Phased Array UT
Although phased array systems are used for many applications in NDT, one of
the most diverse applications is weld inspection. Traditional mono-element-type
inspections of welds are typically carried out using two or three different
probes.
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.
Parameters of the inspection such as probe dimensions, beam angles, nominal
frequency of the probe and scan patterns are typically given a range of
allowable values in the specifications or regulating standards.
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.
5.1
Phased array calibrations
Issues with calibrating phased-array systems primarily relate to the large
number of variables to deal with (compared to single element probes with fixed
angle wedges).
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.
When an electronic or sectorial scan is used, variations between the electronics
of each pulser, receiver and variations between probe elements may result in
small gain variations from one focal law to the next. The efficiency of
generation varies with angle and declines away from the natural angle of the
wedge.
When a delay line or refracting wedge is used, variations in path distances
within the wedge will result in some focal laws requiring more or less amplifier
gain. A method of compensating for gain variations so as to normalize the set of
focal laws in an electronic or S-scan is required.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-1
Copyright © TWI Ltd
www.twitraining.com
When a phased array probe is used on a delay line or refracting wedge,
calculations for beam steering and projection displays rely on the Fermat
principle.
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!
Figure 5.1 Probe-wedge data entry menus.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-2
Copyright © TWI Ltd
www.twitraining.com
Figure 5.2 Probe-wedge reference values.
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.
5.1.1
Calibrating range and delays
Range on the ultrasonic instrument is set using the A-scan. This applies for the
phased-array instrument too. For an E-scan with a 0° focal law, with or without
a wedge, the range is set the same as it would be for a single element zerodegree probe. This involves simply setting the responses from a flat plate (such
as the 25mm thickness of the IIW block).
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-3
Copyright © TWI Ltd
www.twitraining.com
Figure 5.3 Setting range for an E-scan with 0° Focal Laws.
Similar range and delay adjustments to compensate for the variations of
distance travelled in a wedge can be made for S-scans and scans using either
angled or delay-line wedges but these require a radius such as is available on
the V1 or V2 blocks. The configurations are shown in figure 5.4.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-4
Copyright © TWI Ltd
www.twitraining.com
Del ay adju stme nt movement for S-sca n with refra cting wedge
Delay a dju stme ntmoveme nt fo r ele ctronic sca n wit hrefrac ting wed ge
Maximum
amplitude from
radius
corresponds to
Focal Law that
aligns with
centre of 100mm
radius.
This alignment
with the line on
the IIW block
indicates the exit
point from the
wedge of the
beam for that
focal law.
Del ay adju stme nt movement for 0° with de la yl ine
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-5
Copyright © TWI Ltd
www.twitraining.com
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.
Similarly, confirmation of a single angle of refraction is also not a useful
function for phased arrays. As will be seen later in this chapter most analyses
are made using the display software.
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.
5.1.2
Compensating for attenuation
Setting sensitivity is a standard function used in many NDT systems. Its main
functions are to ensure a minimum or agreed-upon level of gain or detection
and a means of providing a repeatable inspection.
It is perhaps risky, referring to a minimum detection level. This implies that we
will always find a certain sized flaw. However, when using ultrasonic methods
the amplifier gain is not the only parameter that determines if a particular flaw
is detected or not.
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.
To overcome some of the issues involved in these uncertainties it is normal to
set sensitivity on a fixed simple target. These are invariably symmetric and
easily machined shapes when inspecting welds.
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.
All of these targets provide a repeatable reference upon which to set an echo
response to reference amplitude and allow for the attenuation effects of the
material.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-6
Copyright © TWI Ltd
www.twitraining.com
However, when dealing with phased-array techniques there are other
parameters to consider in addition to the material attenuation effects. Again,
these primarily deal with the wedge variations.
One method of wedge-attenuation compensation applies to assessments of
wedge-attenuation for E-scans where 1D linear array probes are used. To
compensate for the attenuation effects the phased-array system is configured
for the focal laws to be used.
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.
Alternatively, this process may be computerized so that a dynamic assessment
of sensitivity adjustment is calculated by the computer. A dynamic assessment
would simply require the operator to move the probe back and forth over the
SDH ensuring that all the focal laws used have the SDH target move through
the beam.
Wedge attenuation corrections would then be calculated by the phased-array
system to ensure that the amplitude of the SDH detected by each focal law
would be adjusted to the same amplitude. Figure 5.5 illustrates how this may
be accomplished for 0° and angled E-scans.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-7
Copyright © TWI Ltd
www.twitraining.com
Figure 5.5 Attenuation compensation for E-scans.
Assessment of wedge-attenuation compensation requires a constant steel path
to ensure that only the effects of wedge variations are assessed.
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.
A recommended target for S-scans is a radius similar to that of the 100mm
radius of the V1 block or the 25mm or 50mm radii in the V2 blocks.
Use of the radius for S-scan configurations also provides correction for echotransmittance effects intrinsic in angle variation.
Figure 5.6 Attenuation compensation for S-scans.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-8
Copyright © TWI Ltd
www.twitraining.com
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).
Figure 5.7 Automatic gain addition for attenuation effects.
If appropriate compensation cannot be achieved, for example if the angular
range is so large that the signal amplitude cannot effectively be compensated,
then the range must be reduced until it is possible to compensate.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-9
Copyright © TWI Ltd
www.twitraining.com
5.1.3
Instrument linearity
The individual pulser and receiver components of phased-array ultrasonic
instruments operate essentially the same as any single channel ultrasonic
instrument. Conformance to linearity requirements as described in several
national or international standards (such as ASTM E-317) may be carried out.
However, due to the digital-control nature of all phased-array instruments and
the fact that multiple pulsers and receivers are used, it is required that phased
array instruments be assessed for linearity differently than traditional singlechannel units. This involves assessing the phased array instrument A-scan
display.
First adjust the time-base of the A-scan to a suitable range to display the pulseecho 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.
Variable impedance plugs
Figure 5.8 Phased-array linearity block.
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 nonsaturating signals of interest for display height and amplitude control linearity
assessments.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-10
Copyright © TWI Ltd
www.twitraining.com
5.1.4
Display height linearity
With the phased array instrument connected to a probe (shear or longitudinal)
and coupled to any block that will produce 2 signals as shown in figure 5.9,
adjust the probe so that the amplitudes of the two signals are at 80% and 40%
of the display screen height.
If the phased-array instrument has provision to address a single element probe
in pulse-echo mode then the two flat bottom holes with adjustable acoustic
impedance inserts in the custom linearity block shown in figure 5.8 provides
such signals.
Figure 5.9 Display height linearity verification.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-11
Copyright © TWI Ltd
www.twitraining.com
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.
5.1.5
Amplitude control linearity
Phased-array instruments are rated by the number of pulser-receivers and total
elements possible to address. A 16/64 phased-array instrument has 16 pulsers
and receivers that are used to address up to 64 elements.
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.
Using this probe, configure the phased-array ultrasonic instrument to have an
electronic raster scan. Each focal law will consist of one element and the scan
will start at element number 1 and end at the element number that corresponds
to the number of pulsers in the phased-array instrument.
Couple the probe to a suitable surface to obtain a pulse-echo response from
each focal law. The backwall echo from the 25mm thickness of the IIW block or
the backwall from the 20mm thickness of the custom linearity block illustrated
in figure 5.8 provides a suitable target option.
Alternatively, immersion testing can be used. Immersion configurations tend to
be less susceptible to coupling problems from one end of the probe to the
other.
Select channel 1 of the pulser-receivers of the phased-array instrument. Using
the A-scan display, monitor the response from the selected target. Adjust the
gain to bring the signal to 40% screen height. This is illustrated in figure 5.10.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-12
Copyright © TWI Ltd
www.twitraining.com
Figure 5.10 Channel 1 backwall echo set to 40%.
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%.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-13
Copyright © TWI Ltd
www.twitraining.com
5.1.6
Time-base linearity (horizontal linearity)
Timebase assessment of the display is an anachronistic throwback to analogue
Cathode Ray Tubes where the magnetic deflector plates used to display the
timebase were often unstable or easily knocked out of alignment with a bump.
In spite of this anachronism, the function is still carried out on digital displays.
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).
To assess the timebase, configure the phased array instrument to display an Ascan presentation. Select any compression wave probe or 0° compression wave
focal law and configure the phased-array instrument to display a range suitable
to obtain at least 10 multiple back reflections from a block of a known
thickness. The 25mm wall thickness of the IIW block is a convenient option for
this test.
Set the phased-array instrument analogue-to-digital conversion rate to at least
80MHz. With the probe coupled to the block and the A-scan displaying 10
clearly defined multiples as illustrated in figure 5.11, the display software is
used to assess the interval between adjacent backwall signals.
Figure 5.11 A-scan display for horizontal linearity.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-14
Copyright © TWI Ltd
www.twitraining.com
Acoustic velocity of the test block should be predetermined using the methods
similar to those described in ASTM E-494. These are entered into the display
software and the display configured to read out in distance (thickness or true
depth).
Using the reference and measurement cursors determine the interval between
each multiple and record the interval of the first 10 multiples. Acceptable
linearity may be established by an error tolerance based on the analogue-todigital conversion rate converted to a distance equivalent. For example at
100MHz each sample of the timebase is 10ns.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-15
Copyright © TWI Ltd
www.twitraining.com
Table 5.1 Linearity verification report form.
LINEARITY VERIFICATION REPORT FORM
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
Allowed
Range
100
Small Actual
(%)
Ind.
Height
dB
Allowed range
47-53
40
+1
44-46
90
42-48
40
+2
48-52
80
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
40
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
Time-base Linearity (for 25mm IIW blocks)
Multiple
1
2
3
4
5
6
7
8
9
10
Thickness
25
50
75
100
125
150
175
200
225
250
±0.5
mm
±0.5m
m
±0.5m
m
±0.5m
m
±0.5m
m
±0.5m
m
±0.5m
m
±0.5m
m
±0.5m
m
±0.5m
m
Measured
interval
Allowed
deviation
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-16
Copyright © TWI Ltd
www.twitraining.com
5.1.7
Active element assessment
This assessment is used to determine that all elements of the phased array
probe are active and of uniform acoustic energy.
During normal operation of a phased array probe, each of the elements is
addressed by a separate pulser and receiver. Phasing principles assume that
each element contributes identically to its adjacent element.
To ensure that a uniform beam is constructed a method must be used that
ensures the electronic performance of the phased-array instrument is identical
from element to element and any differences are attributable to the probe
itself.
To ensure that any variation of element performance is due only to probe
construction, a single pulser-receiver channel is selected to address each
element.
To assess each element individually the phased array probe to be tested is
connected to the phased-array ultrasonic instrument and any delay line or
refracting wedge removed from the probe.
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).
Alternatively an immersion method using a fixed water path may be used and
the water-steel interface signal monitored instead of the steel wall thickness.
Configure an electronic scan consisting of one element that is stepped along
one element at a time for the total number of elements in the array. This should
ensure that the pulser-receiver number 1 is used in each focal law or if the
channel is selectable it should be the same channel used for each element.
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).
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-17
Copyright © TWI Ltd
www.twitraining.com
If a pre-packaged program is available for checking element activity, this can
be used as an alternative.
Table 5.2 Probe element activity chart: Enter receiver gain for 80% FSH.
Element
1
2
3
4
5
6
7
8
9
1
0
1
1
1
2
1
3
1
4
1
5
1
6
Gain
Active (√)
Inactive (x)
Data collected is used to assess probe uniformity and functionality. Comparison
to previous assessments is made using the same instrument settings (including
gain) that were saved to file.
The receiver gain to provide an 80% response should be within a range of ±3dB
of any previous assessments and within ±3dB of each other.
The total number of inactive elements and number of adjacent inactive
elements in a probe should be agreed upon and identified in a written
procedure.
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.
The number of inactive elements allowed should be based on performance of
other capabilities such as focusing and steering limits of the focal laws being
used.
No simple rule for the number of inactive elements can be made for all phasedarray probes. Typically, if more than 25% of the elements in a probe are
inactive, sensitivity and steering capabilities may be compromised.
Similarly, the number of adjacent elements allowed to be inactive should be
determined by the steering and electronic raster resolution required by the
application.
Stability of coupling is essential for the comparison assessment. If using a
contact method and the assessment of elements produces signals outside the
±3dB range the coupling should be checked and the test, run again.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-18
Copyright © TWI Ltd
www.twitraining.com
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.
Cable continuity adapters can be made that allow the multi-strand connectors
to be tested independently. These adaptors can be connected to the phased
array instrument directly to verify that all output channels are active or they
can be connected to the probe-end of the cable to indicate the continuity of the
individual co-axial connectors in the inter-connecting cable.
Figure 5.12 illustrates an example of a display used to identify inactive channels
in a phased array instrument or cable.
Each horizontal
line of colour
represents a
separate element
(channel).
Missing lines (no
colour) indicate
inactive elements
(3 inactive
Figure 5-12 B-scan display element activity.
5.1.8
Beam characterisation (profile)
Beam profiling is a common procedure with most ultrasonic applications. The
size of the beam as it moves along its axis is a critical part of detection
capabilities and sizing methods.
Either immersion or contact probe applications can be addressed using these
procedures. Assessments of contact probes may suffer from variability if proper
precautions are not taken to ensure constant coupling conditions.
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 beamprofile assessments with no or little mechanical motion.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-19
Copyright © TWI Ltd
www.twitraining.com
Where mechanical motion is used it should be encoded to more accurately
relate signal time and amplitude to distance moved.
Linear-array probes have an active plane and an inactive or passive plane.
Assessment of the beam in the active plane should be made by use of an
electronic scan sequence for probes with sufficient number of elements to
electronically advance the beam past the targets of interest.
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 sidedrilled 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.
Figure 5.13 Profiling E-scan of side drilled holes.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-20
Copyright © TWI Ltd
www.twitraining.com
The profile is accomplished by data collection of the entire waveform over the
range of interest. A display with amplitude as a colour or greyscale is
constructed.
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.
Figure 5.14 is a B-scan display of the electronic scan represented in figure 5.13.
Depth is in the vertical axis and electronic-scan distance is represented along
the horizontal axis.
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.
Off-axis lobe
effects
Figure 5.15 Angle corrected B-scan display of beam profile of SDH.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-21
Copyright © TWI Ltd
www.twitraining.com
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.
Waveform collection of signals using a combination of electronic scanning in the
active plane and encoded mechanical motion in the passive plane provides data
that can be projection-corrected to provide beam dimensions in the passive
plane.
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.
Figure 5.16 Passive axis beam profiling.
Figure 5.17a illustrates an alternative to the stepped intervals shown in figure
5.16. In this case a through hole is arranged perpendicular to the required
refracted angle so as to provide a continuous transition of path length to the
target.
Figure 5.17a Combined active and passive axis beam profiling.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-22
Copyright © TWI Ltd
www.twitraining.com
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 5.17b Combined active and passive axis beam profiling projected C-scan
option.
5.1.9
Determining phased-array beam-steering limits
It has been noted that in the design of a phased array probe there are factors
that limit the extent to which the beam can be steered away from the
perpendicular. These were given in section 3.5.2 where it was explained that:
Steering capability is related to the width of an individual element of the array:
Maximum steering angle (at –6 dB), given by
 st  0.5
Where:

e
= the wavelength, e = the individual element width.
Manufacturers will normally provide guidance on the limits based on this
principle and incorporate some conservatism. It is therefore possible to
calculate a theoretical limit based on nominal frequency and manufacturer
provided information on the element dimensions.
However, several parameters can affect the theoretical calculations. These are
primarily related to the nominal frequency of the probe. Some of the
parameters affecting actual frequency include; pulse length, damping, use of a
delay-line or refracting wedge and variations in manufacturing processes on
thickness lapping and matching layers.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-23
Copyright © TWI Ltd
www.twitraining.com
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.
Steering capability will usually be based on a comparison of signal to noise
ratios at varying angular displacements. Beam steering capability will also be
affected by project requirements of the beam.
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.
The following recommended method uses a series of SDHs and a phased array
system configured for its intended use; ie with or without a refracting wedge or
delay-line, unfocused or a defined focal distance and the specific test material
to be tested.
A block is prepared with a series of side drilled holes in the material to be used
for the application at the distance or distances to be used in the application.
The side-drilled-hole pattern should be as illustrated in figure 5.18 or 5.19.
When no focusing is used or when focusing is at a fixed soundpath distance the
holes would be as indicated in figure 5.18. These are at 5° intervals at a 25mm
and 50mm distance from a centre where the probe is located.
Figure 5.18 Beam steering assessment block – constant sound path.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-24
Copyright © TWI Ltd
www.twitraining.com
Vertical pla ne 50mm from exit point
85°
80°
75°
70°
65°
60°
55°
5°10°20°30° 40° 50° 55° 60°
Horizontal plane 25mm from exit point
50°
45°
40°
35°
30°
Block dimensions 150mm x 100mm x 25mm
Figure 5-19 Beam steering assessment block – fixed plane sound path.
Similar assessments are possible for different applications. When a set of focal
laws is arranged to provide resolution in a plane instead of a sound path
distance, the plane of interest may be used to assess the steering limits of the
beam.
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.
When a probe is used in direct contact with a verification block such as
illustrated in figure 5.19 the lack of symmetry either side of the centreline
prevents both positive and negative sweep angles being assessed
simultaneously.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-25
Copyright © TWI Ltd
www.twitraining.com
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 timingdelay 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.
Evaluation of steering limits is made using the dB difference between the
maximum and minimum signal amplitudes between two adjacent side drilled
holes.
For example; when a phased array probe is configured to sweep + /- 45° on a
block such as illustrated in figure 5.18, the highest of the pair of the SDH’s
which achieves a 6dB separation could be considered the maximum steering
capability of the probe configuration.
The 6dB separation may not always be adequate. Acceptable limits of steering
may be indicated by the maximum and minimum angles that can achieve a prespecified separation between adjacent holes. Depending on the application a
6dB or 20 dB (or some other value) may be specified as the required
separation.
Steering capabilities could be part of a system specification; eg a phased array
system is required to achieve a minimum steering capability for 5° resolution of
2mm diameter side drilled holes of plus and minus 20° from a nominal midangle.
Conversely, a system may be limited to S-scans not exceeding the angles
assessed to achieve a specified signal separation, eg -20dB between 2mm
diameter SDHs separated by 5°.
An alternative assessment may use a single SDH at a specified depth or sound
path distance. Displaying the A-scan for the maximum and minimum angles
used would assess the steering capability by observing the S/N ratio at the
peaked response. This method could have the steering limit pre-defined as S/N
ratio being achieved.
In all cases, when assessing steering limits, caution must be taken to observe
grating lobe signals. Grating lobe signals may be present at significant angular
separation from the main or intended signal.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-26
Copyright © TWI Ltd
www.twitraining.com
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.
5.2
Scanning calibration
In section 5.1.2 we touched on setting delays and sensitivity. Delay
adjustments are required in phased-array inspections to ensure that the
analysis software correctly identifies the interface between the probe or probewedge and the test piece. The sensitivity addressed in that section related to
the effects of attenuation within a wedge.
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  V0ed
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).
Sensitivity corrections described in section 5.1.2 were made using a single
target at a fixed metal soundpath (either a SDH at a fixed depth for E-scans or
a radius for S-scans). This corrected for not only the variations in the wedge
paths for the different focal laws but also the echo-transmittance effects
whereby higher refracted angles are less efficient at transmitting and receiving
signals.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-27
Copyright © TWI Ltd
www.twitraining.com
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.
Presently there are no options for addressing AVG/DGS curves using phased
array. However, either of the DAC or TCG options can be adapted to phased
array inspections.
Use of a DAC is possible on a phased array unit; however, its suitability is
debatable. A DAC is used by the operator as a drawn or calculated curve that
may or may not be linked to a gate system such that signals breaking the DAC
set off an alarm.
Use of a DAC would be limited to applications where the system is configured
with a single focal law and the operator is using the probe as if it was a single
element probe.
Figure 5.20 shows DAC being constructed on an A-scan display for a 35° shear
mode focal law.
Figure 5.20 DAC for a phased-array focal law.
For most phased array applications, analysis is made using projected scans (Bscans, 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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-28
Copyright © TWI Ltd
www.twitraining.com
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.
Figure 5.21 TCG for a phased-array focal law.
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.
However, it should be remembered that if the effects of delay and wedge
attenuation have been addressed as part of the initial setup, then only the
material attenuation and beam divergence is reducing the echo response from
the target.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-29
Copyright © TWI Ltd
www.twitraining.com
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).
Figure 5.22 Calibrating using a SDHs for phased-array focal laws.
5.3
Phased-array technique development
Having established methods for quality verification in section 5.1 and for setting
scanning sensitivities for typical weld volume inspections in section 5.2, the
next step is to plan for how the phased array beams can be used to provide an
effective weld inspection.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-30
Copyright © TWI Ltd
www.twitraining.com
Transverse flaws are generally rare in most welding processes; however, if
there is a requirement for their detection, then options such as matrix arrays
with beam skewing capabilities and multiple probe configurations or multiple
pass techniques with probes skewed to direct the beams essentially parallel to
the weld axis can be added.
An operator preparing for a weld inspection using phased array equipment and
techniques considers the essential parameters in much the same way as they
would for a single element manual scan. Each of the parameters has an
equivalent in manual scanning.
5.3.1
Parameter
Angles
Phased Array
Single or range is
available
Single Element
Only one angle per
probe
Apertures
Variable in Active
direction fixed in
passive
A single probe size is
selected
Frequencies
A single frequency per
probe
A single frequency per
probe
Display
S-scan/E-scan
(others)
Only A-scan unless
mechanised (but
never S-scan)
Manual or Mechanised
May be either
May be either
Manual raster scanning
Scanning techniques can be developed to use the phased array probe as a
hand-held manually manoeuvred item operated in the same raster fashion as
an operator would use for a tradition weld inspection with a simple A-scan and
single element probe.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-31
Copyright © TWI Ltd
www.twitraining.com
Bottom and
top markers
True
depth
scale
Indication
Stand
off
scale
Figure 5.23 Monitoring an S-scan display – manual scanning.
5.3.2
Encoded linear scanning
Although it is possible to perform phased array weld inspections using the
manual raster motion typical of single element techniques, the greatest
advantages of phased array weld inspection are had using various degrees of
mechanisation. Mechanisation of weld inspections need not be as complex as
two or three axes of motion control using motorised actuation.
The simplest form of mechanisation would involve connecting an encoder to the
phased-array probe and moving it along the weld. This simple linear scan as it
is called has proven to be the most popular option for phased-array weld
inspections.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-32
Copyright © TWI Ltd
www.twitraining.com
The phased-array instrument then must be configured to collect the data with
suitable parameter inputs to ensure that the flaws located are rotated correctly
ie skew is 180° different between the two probes.
Single and dual phased-array mounting with encoders is illustrated in figure
5.24.
Encoder
Irrigation
connection
Dual probe
carriage
Figure 5.24 Phased array probes and encoders.
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 rescans due to missed data and poor coupling.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-33
Copyright © TWI Ltd
www.twitraining.com
Encoder and probe moved
along guide-strip
Guide-strip held in
place
with magnet
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.
The scanning apparatus can be a simple single or double probe holder or may
be equipped to hold several phased array probes in a single fixture for more
complex scanning, see figure 5.26.
Large diameter multi-probe
pipe scanner
Small diameter pipe scanner
Figure 5.26 Motorised phased array scanners.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-34
Copyright © TWI Ltd
www.twitraining.com
5.3.3
Modelling the weld inspection – scan plans
A requirement of nearly all national or international Standards is the inclusion of
a description of the volume coverage of the test piece. This is variously called
the scan plan, the scanning technique or the procedure, depending on the
terminology used in the specific industrial venue.
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.
With the complexity of three dimensional issues when addressing nozzles and
the many possible skip options when dealing with the sweep of angles available
with S-scans, many users have found it convenient to develop computer
assisted drawings.
Several options have developed:




Spread-sheet based.
Simple ray tracing.
Complex ray tracing.
Finite element modelling.
In order to streamline the process of zonal discrimination, a modelling tool with
a firmware feedback was developed. All phased array focusing is based on the
Fermat model whereby a minimum arrival time along a given path is used to
calculate the focal law delays.
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.
This displays a table of focal law parameters and a graphic representation of the
centre of beam rays is provided indicating where the beam is directed and
focused, see figure 5.27.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-35
Copyright © TWI Ltd
www.twitraining.com
Figure 5.27 Spreadsheet-based beam modelling.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-36
Copyright © TWI Ltd
www.twitraining.com
Figure 5.28 Simple raytrace beam modelling.
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.
Phase FX
Advanced Focal Law calculator
Figure 5.28 Phased-array beam forming models.
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 Ascans 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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-37
Copyright © TWI Ltd
www.twitraining.com
The weld modelled is an austenitic weld on a carbon-steel to stainless clad
surface (as seen in the overlay of the B-scans).
In the bottom right of the image is seen an A-scan representation of a flaw that
exists in the weld metal with the noisy scattering effects of the grain structure
occurring prior to the flaw signal. (This figure is from a presentation by L. Le
Ber and uses the CIVA Simulation software from CEA).
Figure 5.29 Phased-array beam predicted models by Finite Element.
5.4
Phased-array data analysis
Data acquisition and setup software menus provide the operator with the
options to control the focal laws and ensure that the data collected is suitable
and in accordance with the technique designed for the application.
Although the preferred acquisition uses encoded acquisition methods most
instruments have provision for timed-acquisitions (ie a specified number of
samples collected per second).
Timed acquisitions may be adequate for analysis of a specific location but in
order to make meaningful length measurements from acquired data, the
operator will need real length units for analysis.
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 reentered for repeating the test in subsequent inspections.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-38
Copyright © TWI Ltd
www.twitraining.com
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.
TWI
OmniScan Report
Setup TA_T5AW134L60S_P01A_01 for weld inspection technique
development for pipe sample UT-328
Report
Date
Report
Version
Setup Filename
2004/07/26 1.0R2
Date of
Inspection Save Mode
Inspection Version
TA_T5AW134L60S_P01A_01.ops 2004/07/26 1.0R2
AScan
OmniScan OmniScan Module Type
Type
Serial #
Module
Serial #
Calibration Data
Due
Filename
OmniScan
MX
112375-0
2005 / 02 / 60_P01A_##
28
112356-0
OMNI-M-PA128-UT
Contractor
Eclipse
Technician
Robert, Larry and
Tim
Customer
Test
Project
Weld Technique
Development
Site
ESP Lab
Part Number
Flawtech UT328
Probe characterization
Probe
Model
Probe
Serial #
Wedge
Model
Wedge
Angle
Probe
Aperture
5L64-A2
506X001
Custom
36.6º
9.60 mm
Probe
Max
Frequency Frequency
5.00 MHz
N/A
Lower
Higher
Center
Bandwidth
Frequency Frequency Frequency (MHz)
Bandwidth
(%)
-6 dB
N/A
N/A
N/A
N/A
N/A
-20 dB
N/A
N/A
N/A
N/A
N/A
Date
Time
Procedure
Calibration Characterization
Block
Gain
N/A
N/A
N/A
N/A
N/A
00000Setup
Beam Delay Start
(Half
Path)
Range
(Half
Path)
9.79 us
250.66 mm 60
-0.17 mm
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
PRF
5-39
Type
Averaging
Factor
PA
1
Copyright © TWI Ltd
www.twitraining.com
Scale Type Scale
Factor
Video
Filter
Pretrig
Rectification Band Pass
Filter
Compression 49
On
0.00 µs
FW
Voltage
Gain
Mode
Wave Type Sound
Velocity
Pulse
Width
80 V
12.00 dB
PE
(Pulse
Echo)
User
Defined
3197.2 m/s
100.00 ns
Scan Offset
Index
Offset
Skew
0.00 mm
0.00 mm
0.0º
Gate
Start
Width
Threshold
Synchro
I
0.02 mm
5.12 mm
20.00 %
Pulse
A
24.65 mm
16.31 mm
60.00 %
Pulse
B
202.31 mm 19.58 mm
44.00 %
Pulse
TCG Point
Number
Position
(half
path)
Gain
1
0.00 mm
0.0 dB
2
36.00 mm
25.5 dB
3
65.00 mm
32.5 dB
4
91.00 mm
40.0 dB
5
123.00 mm 46.0 dB
6
142.00 mm 50.0 dB
5 MHz (3.1 7.5 MHz)
Scan area
Scan
Start
Scan
Length
Scan
Resolution
0.00 mm
360.00 mm 1.00 mm
Synchro
Max Scan
Speed
Clock
N/A
Index
Start
Index
Length
Index
Resolution
0.00 mm
37.86 mm
37.86 mm
Calculator
Used
Element
Qty
16
First
Element
1
Last
Element
Resolution Wave
Type
64
Start Angle Stop Angle Angle
Resolution
60.0º
60.0º
User
Defined
1
Focus
Depth
1.0º
Material
Velocity
3197.2 m/s
Scan Type
150.00 mm
Linear
Figure 5.30 Setup parameter summary.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-40
Copyright © TWI Ltd
www.twitraining.com
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.
5.4.1
File saving and naming protocol
Once the scan has been completed and the operator satisfied that the data is of
a good quality for analysis it is saved as a file. This simple function is perhaps
not as well attended as it should be.
Many projects can involve scanning of hundreds or even thousands of metres of
weld. A single weld may be 20-30m long and done in 0.5 to 1m lengths.
Keeping track of the welds by means of some sensible file-naming system will
be absolutely critical to the success of the analysis and reporting!
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.
No single file-naming protocol can be demanded or provided for all applications;
however, whatever system is used it should be recoded as part of the reporting
process.
5.4.2
Data analysis displays - options
With the ability to collect the echo responses from so many angles and
standoffs simultaneously, phased-array data display can be a challenging
matter.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-41
Copyright © TWI Ltd
www.twitraining.com
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.
Options for the views available may include:





Top/side/end – uncorrected versus volume corrected.
Polar views.
Scans.
Projections (volume between cursors or single plane).
Merged data from all or selected focal laws.
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:





Overlays (weld profile, flaws).
Threshold adjustments – electronic gain (soft gain) and colour palettes.
Smoothing of pixels.
Selection of timebase scales (eg half-path, true depth, time).
Cursors (various types including markers and reference and measurement).
Orienting ones perspective when so much data is collected is critical for proper
analysis. The old R/D Tech Cube was developed as a handy tool for this
purpose.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-42
Copyright © TWI Ltd
www.twitraining.com
Figure 5.31 The R/D tech cube.
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.
53° slice
61° slice
flaw
Root
geometry
Figure 5.32 Top-side-end view of a pipe butt weld.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-43
Copyright © TWI Ltd
www.twitraining.com
A block cursor on the S-scan is seen and it is positioned so that it passes
through the red (highest amplitude) indication on the S-scan. The software
indicates that this maximum occurs at the focal law at 53°.
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 (volumecorrected 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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-44
Copyright © TWI Ltd
www.twitraining.com
Figure 5.33 Confirmation of flaw position with 2 S-scans.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-45
Copyright © TWI Ltd
www.twitraining.com
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.
Figure 5.35 Merged S-scans with corrections and smoothing.
The process of merging has divided the inspection volume into small volumes
made of the three different scan dimensions:

Scan increments X A-scan time (distance) interval X angle (or standoff)
increment.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-46
Copyright © TWI Ltd
www.twitraining.com
A properly designed technique for weld inspection ensures that at least 2 beams
pass through every voxel. This ensures that every volume of the weld is
interrogated by at least 2 beams.
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.
With the entire inspection volume reduced a 3-dimensional grid it is then
possible to make slices through the grid in each of the orthogonal planes.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-47
Copyright © TWI Ltd
www.twitraining.com
End Projection
Projections may consist of a single
slice or a range of slices projected
orthogonally
through the surface
Sid e
Pro je
c
tio n
Depth Projection
Figure 5.36 Merged S-scans with projections.
By limiting the projections to either a single plane or just a small range of one
of the planes it is much easier to identify details of the indications of concern.
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).
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-48
Copyright © TWI Ltd
www.twitraining.com
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 510mm 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.
An A-scan is provided in the display shown in figure 5.37, however, this is not a
true A-scan as extracted from the raw data. Instead, this is a composite of
maximum amplitudes for all the voxels that make up the reference cursor
position (red cursor in the End view on the upper right pane).
It can be seen that this pseudo-Ascan has a jagged shape (not characteristic of
a 100 MHz digitisation rate of the timebase) and the peaks are indicated as
occurring at 20mm, just over 40mm and at 60mm.
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.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-49
Copyright © TWI Ltd
www.twitraining.com
Figure 5.37 Projections limited to Flaw region.
5.5
Phased-array codes and standards
Codes and Standards in NDT are generally of two sorts; those that regulate
aspects of the equipment used for a particular test method and those that set
out the steps or requirements to carry out the application of a specific test
method.
Examples of the equipment regulatory documents include:




BS EN ISO 18563-1: Non-destructive testing: Characterization and
verification of ultrasonic phased array equipment. Part1: Instruments.
BS EN 16392-2: Non-destructive testing: Characterization and verification of
ultrasonic phased array equipment. Part 2: Probes.
BS EN ISO 18563-3: Non-destructive examination of welds: General rules
for metallic materials. Part 3: Combined systems.
ASTM E-2491: Standard guide for evaluating performance characteristics of
phased-array ultrasonic testing instruments and systems.
Examples of standards regulating the implementation of the test method
include:


BS EN ISO 13588: Non-destructive testing: Ultrasonic Testing – Use of
automated phased array technology.
ASTM E-2700: Standard Practice for Contact Ultrasonic Testing of Welds
Using Phased Arrays.
In the second group we could also include standards such as:
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-50
Copyright © TWI Ltd
www.twitraining.com


ASME Code case 2235: Use of ultrasonic examination in lieu of Radiography,
Section I, Section XII.
ASTM E-2192 Standard Guide for Planar Flaw Height Sizing by Ultrasonics.
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:




Requirements for limits on the deviation of the measured exit point
compared to the manufacturer’s indicated exit point on a wedge.
Limitations for a maximum deviation of measured angle compared to the
manufacturer’s indicated angle on the wedge.
Limitation of the area of the probe.
Failure to address the fact that multiple pulser-receivers are used.
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.
To address the new technologies some Codes have incorporated wording that
allow for deviation from the traditional application of manual ultrasonic testing.
For example:

AWS D1.1: Structural Welding Code (Steel) states, “Variations in testing
procedure, equipment and acceptance standards not included in Part F of
Section 6 may be used upon agreement with the Engineer. Such variations
include other thicknesses, weld geometries, transducer sizes, frequencies,
couplant, painted surfaces, testing techniques, etc.”
AWS D1.1 makes it clear that other options may be used provided that these
are agreeable to the contracting parties and the Engineer approves.
Another code that is often associated with ultrasonic examination of welds is the
Boiler and Pressure Vessel Code by ASME (American Society for Mechanical
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-51
Copyright © TWI Ltd
www.twitraining.com
Engineers). This has long recognised phased-array systems as an option to
ultrasonic examinations.
ASME Section V has recognised phased-array UT as an option to the single
element manual techniques since the December 1992 addendum. At that time it
was added as a non-mandatory appendix and identified as one of the
computerised imaging techniques (CITs).
Since then ASME has been developing mandatory appendices to provide more
explicit instructions on the proper use and calibration of phased array systems
for weld inspections.
A more general concern for the assessment of the instrumentation was
requested outside the pressure vessel industry. This led to the development of
an ASTM (American Society for Testing and Materials) standard:

ASMT E-2491: Standard Guide for Evaluating Performance Characteristics of
Phased-Array Ultrasonic Examination Instruments and Systems.
Materials in section 5.1 of this book are based on this document.
NDT44-060319
Calibration and Scanning Considerations for
Weld Inspections using Phased Array UT
5-52
Copyright © TWI Ltd
www.twitraining.com
Section 6
Industrial Applications of Phased Array UT
6
Industrial Applications of Phased Array UT
Phased array ultrasonic instruments may have had a slow start getting into
industrial applications but now several examples exist where users have found
the advantages of phased arrays superior to that of conventional UT. Several
examples are provided in this section.
6.1
Electric resistance welds
Shaped arrays replace several single probes arranged either side of the weld.
Old technology requires that an operator must try to use friction guide rollers to
keep the weld centreline between the probes. This is not reliable and the area
of interest often wanders outside of the beam coverage.
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
Figure 6.1 Conventional ERW UT.
NDT44-50816
Industrial Applications of Phased Array UT
6-1
Copyright © TWI Ltd
www.twitraining.com
Figure 6.2 Phased array ERW UT.
6.2
Aerospace fuselage fastener cracking/corrosion
The Autoscan 1701A Fastener Hole Inspection System was designed to detect
small cracks on the faying surface of wing skin fastener holes with the fasteners
installed.
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
NDT44-50816
Industrial Applications of Phased Array UT
Faying Surface Corner
Cracks
6-2
Copyright © TWI Ltd
www.twitraining.com
Figure 6.3 Phased array for aerospace fastener holes.
6.3
Power generation - turbine blade roots
Turbine components present a special concern for access and complex shapes
to inspect. Phased arrays have become a popular solution to providing coverage
of a variety of locations from the single point of access.
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.
NDT44-50816
Industrial Applications of Phased Array UT
6-3
Copyright © TWI Ltd
www.twitraining.com
6.4
Power generation – heavy nozzles
The curvature and limited standoff access on heavy nozzles provide another
application for phased arrays to solve the inspection problems. Figure 6.5
shows how a sectorial scan allows the necessary coverage of the critical inner
radius.
Figure 6.5 Phased array sectorial scanning of nozzle inner radius.
6.5
Petrochemical pipeline construction (pipe wizard)
Pipeline girth weld inspection by UT has rapidly been taking over from
radiography. Dividing the weld into vertical divisions and using separate beams
for each zone has provided the basis for the zone discrimination technique.
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).
NDT44-50816
Industrial Applications of Phased Array UT
6-4
Copyright © TWI Ltd
www.twitraining.com
Figure 6.6 Pipeline girth weld inspection – phased array versus multiprobe.
Figure 6.7 Phased array GUI for girth weld focal laws.
6.6
Other applications using phased arrays
The preceding examples of phased arrays show specific industry applications
where phased arrays have improved the inspection techniques by improving
coverage from a restrictive access, increased inspection speed and the potential
for detection and sizing.
Phased array technology easily incorporates all aspects of traditional ultrasonics
and can be made to essentially duplicate the performance of a single element
probe. Two examples are ToFD and backscatter sizing.
NDT44-50816
Industrial Applications of Phased Array UT
6-5
Copyright © TWI Ltd
www.twitraining.com
6.7
ToFD by phased array
Figure 6.8 shows ToFD scans from two different projects. The left side uses a
single element ToFD probe 6mm diameter 10MHz and the plate is 45mm thick.
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 6.8 Conventional versus phased array ToFD.
6.8
Backscatter sizing
Flaw sizing has always been a critical aspect of NDT. In recent years it has been
recognised that tip diffraction techniques afford the best options for ultrasonic
sizing.
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).
NDT44-50816
Industrial Applications of Phased Array UT
6-6
Copyright © TWI Ltd
www.twitraining.com
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.
Reflected wave

Diffracted wave D2
Incident wave
Diffracted wave D1
Flaw of Size D
D1
D2
Example A-scan with diffracted signal separation
Figure 6.10 Tip signal origins.
The delta time between D1 and D2 does allow for some estimate of sizing via
 2D 
t  
sin 
 c 
NDT44-50816
Industrial Applications of Phased Array UT
6-7
Copyright © TWI Ltd
www.twitraining.com
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.
The principles of the depth and orientation technique can be illustrated by the
probe positions as related to the echo dynamic peak positions in figure 6.11.
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.
NDT44-50816
Industrial Applications of Phased Array UT
6-8
Copyright © TWI Ltd
www.twitraining.com
Probe motion
parallel to notc h axis
FL25
FL1
EDM notch
FL1 = focal law 1
FL25 = foc al law 25
Figure 6.12 Schematic of phased array probe on notch blocks.
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.
Since an electronic raster is also occurring where the waveforms are collected,
the stacked A-scans (B-scans) at each point along the scan length provide the
echo dynamics of the flaw in the beam.
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 lower left is a volume corrected side view, B-scan composed of a composite
of the total number of focal laws in the vertical and each encoded data point
along the scan axis in the horizontal axis.
The A-scan in the lower right is the display of the angled cursor on the upper
right.
NDT44-50816
Industrial Applications of Phased Array UT
6-9
Copyright © TWI Ltd
www.twitraining.com
Figure 6.12 Tomographic projections.
An example of the sizing and orientation assessment afforded by the technique
is shown in figure 6.13.
NDT44-50816
Industrial Applications of Phased Array UT
6-10
Copyright © TWI Ltd
www.twitraining.com
4mm notch indicated by
cursors as 4.2mm and
44.7° inclination
Figure 6.13 Sizing and orientation determination by phased array backscatter.
6.9
Portable phased array
Numerous other single element applications can be duplicated and usually
improved using phased array techniques. Through the introduction of the
portable phased array instrument many limitations of field mechanisation are
reduced while preserving all the features of display of the larger lab-type units.
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.
OmniScan MX
(Olympus NDT)
Porta-pal
(Amdata)
Focus
(AGR/TechnologyDesig
n)
FocusLT
(Olympus NDT)
PhasorSX
(GE NDT)
X32
(Harfang
Microelectronics)
Pocket PA
(M2M)
NDT44-50816
Industrial Applications of Phased Array UT
MultiX
(M2M)
6-11
Rapidscan
(NDT Solutions)
Copyright © TWI Ltd
www.twitraining.com
SAPHIR (IntelligeNDT)
Figure 6.14 A sampling of phased array units.
NDT44-50816
Industrial Applications of Phased Array UT
6-12
Copyright © TWI Ltd
www.twitraining.com
Section 7
Pipeline Girth Welds
7
Pipeline Girth Welds
In section 6.5 the application of phased arrays to pipeline girth welds was
noted. This particular application of phased arrays has played a significant role
in an industry that has had a long and staid history in NDT.
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.
Thanks to the forward looking ideas of J. de Sterke in RTD Rotterdam in 1959
and the advances during the 1970s made by the committee responsible for the
British Standard PD 6493 in which the concepts of fitness-for-purpose were
codified in 1980, a more effective and more reliable alternative to radiography
was developed.
De Sterke envisioned an ultrasonic weld inspection that used three probes and
made multi-passes around the weld to test the weld integrity. This effectively
divided the weld into vertical intervals and became the basis for the zonal
discrimination technique.
The ultrasonic zones were easily incorporated into the concepts of fitness-forpurpose 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 early 1990’s Canadian pipeline construction projects were using
automated ultrasonic testing (AUT) systems using the concepts of zones with
acceptance criteria based on Engineering Critical Assessment (initially derived in
PD 6493).
These systems were now approved to replace radiography on all large
diameter pipeline construction.
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.
NDT44-50816
Pipeline Girth Welds
7-1
Copyright © TWI Ltd
www.twitraining.com
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.
PipeWIZARD (Olympus NDT)
RTD with Technology Design Hardware
NDT Brasil using TD Focus
GE Weldstar
Figure 7.1 Some phased array systems using zonal discrimination.
NDT44-50816
Pipeline Girth Welds
7-2
Copyright © TWI Ltd
www.twitraining.com
7.1
Principles of the zonal discrimination technique
As noted, the phased array systems applied to this niche industry have all been
adapted from principles developed using multi-probe systems. The methodology
was first contemplated in the 1960’s and continued on into the 1970’s using
standard contact probes.
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 concept of zonal discrimination can be summarised as follows:
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.
This is shown in figure 7.2.
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.
An operator evaluates the chart results and makes a decision as to weld
acceptability based on the length of signals exceeding a threshold as set out in
specifications and regulating codes.
NDT44-50816
Pipeline Girth Welds
7-3
Copyright © TWI Ltd
www.twitraining.com
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.
Misfire is used here to denote a condition unique to CRC’s GMAW processes with
an internal welding head. It occurs when the arc is unsuccessfully initiated and
no metal is deposited as a result. It is considered different from incomplete
penetration (IP) that might be the term more appropriately used when the
vertical land is not fused (this is coined lack of cross penetration (LCP) in the
CRC slang).
Zone Identification
Zones
6
Discontinuity
2nd Fill & Cap
5
Solidification Crack
1st & 2nd Fill
1st Fill
Non Fusion1st Fill
4
Hot Pass (Upper)
3
Hot Pass (Lower)
2
Land for Cross Penetration
1
Root
RootRoot
Lack of Cross Penetration
Root Bead Porosity
Fill
Porosity
Burn Thru
Misfire
Hi-Lo
Figure 7.3 Common weld flaws in GMAW welds.
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.
NDT44-50816
Pipeline Girth Welds
7-4
Copyright © TWI Ltd
www.twitraining.com
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).
Figure 7.4 Spots sizes on ultrasonic zones.
Spot size is a result of beam focusing. Three practical options are available for
focusing using contact probes:
1
2
3
Lens focusing: Flat element with curved plastic plates mounted in front,
figure 7.5.
Curved element focusing: The element is curved and an insert in the gap,
figure 7.6.
Phased array focusing, figure 7.7.
NDT44-50816
Pipeline Girth Welds
7-5
Copyright © TWI Ltd
www.twitraining.com
Lens material 1
Wedge material 2
Figure 7.5 Lens focusing.
Curved element
Gap material (same as wedge)
Wedge material
Figure 7.6 Curved element focusing.
Figure 7.7 Phased array focusing.
NDT44-50816
Pipeline Girth Welds
7-6
Copyright © TWI Ltd
www.twitraining.com
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.
Of paramount importance in the zonal discrimination technique is the alignment
of the beam. This is accomplished using targets in a project-specific section of
pipe.
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.
Figure 7.8 shows a model of a calibration block as it might be mounted in a
dummy to hold it in place as the scanner moves the probes past the targets.
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
notch
OD notch
OD notch
OD notch
Trans OD
notch
Figure 7.8 CAD model and actual calibration block.
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.
NDT44-50816
Pipeline Girth Welds
7-7
Copyright © TWI Ltd
www.twitraining.com
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.
If the combination of weld geometry and probe size is inadequate it may
require multiple phased array probes or a redesign of the probe using more
elements, a larger element height and/or gap.
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
Target (FBH or notc h)
Beam path
Gated tim e
Figure 7.9 Typical phased array focusing to zone targets.
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.
NDT44-50816
Pipeline Girth Welds
7-8
Copyright © TWI Ltd
www.twitraining.com
In addition to the zonal targets volumetric targets, transverse targets and
sometimes ToFD targets are added. Volumetric targets are usually 1.5mm
diameter FBHs milled at 45° and having the FBH end at the centreline as
illustrated in figure 7.10.
Figure 7.10 Volumetric targets.
7.2
Strip charts
Of the several concepts derived by Arie de Sterke the stripchart display turned
out to be one of the most important. That is not to say that de Sterke invented
it, just that he took advantage of the old technology available at the time for
multi-channel data acquisition.
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.
With the development of the multi-channel strip-chart format accompanied by
the volumetric B-scan and ToFD displays, evaluation of the AUT data has been
made a fast and reliable process.
The strip-chart portion of the display is intuitive. Channels are arranged as if
the weld was folded out from the centreline. The outer edges are the uppermost
(outside surface) zones and the two inner zones (middle of the chart region)
are always the inside surface of the pipe.
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.
NDT44-50816
Pipeline Girth Welds
7-9
Copyright © TWI Ltd
www.twitraining.com
Figure 7.11 Symmetry of strip-chart presentation.
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 Bscans 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.
Some UT companies have overcome this limitation by introducing layered
views. By toggling the views they can, for example, view all the strip-charts and
ToFD on one view then all the volumetric B-scans and ToFD on the other view.
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.
Evaluation is a step-by-step process. It is eventually the review, following the
interpretation of the indications noted, to determine whether they meet the
specified acceptance criteria.
NDT44-50816
Pipeline Girth Welds
7-10
Copyright © TWI Ltd
www.twitraining.com
It starts with an assessment of the calibration to ensure that the equipment is
ready to perform correctly on the welds. A mental checklist of requirements
would be carried out upon calibration scan review:








Assess for correct and complete header information (file/project
identification).
Assess for correct amplitude from each zonal target (eg 70-99%).
Ensure overtrace is -6dB to about -14dB zone separation (as applicable).
Ensure adequate lateral separation of target signals.
Ensure targets occur at the correct time gate position.
Ensure that correct colour-change thresholds exist for amplitude/time gates.
Ensure that mapping channels’ targets are set for 80% on volumetric
targets.
ToFD should be well balanced with lateral wave at 50-90% screen height.
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
(volumetric
s)
ToFD
Figure 7.12 Typical items to correct on a calibration scan.
Having assessed the calibration is acceptable the process of weld inspection
may begin. The scanner would be taken from the dummy and placed so that
the scanner zero reference mark aligns with the zero start position on the girth
weld and the scan would commence.
NDT44-50816
Pipeline Girth Welds
7-11
Copyright © TWI Ltd
www.twitraining.com
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.
Figure 7.13 Close-up of calibration strip-chart (using 2 root channels).
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:
NDT44-50816
Pipeline Girth Welds
7-12
Copyright © TWI Ltd
www.twitraining.com




Confirm that the header is correct including weld identification.
Ensure coupling check is OK. There should be no areas more than minimum
flaw lengths missed and a policy should exist to be able to use ToFD and
mapping to assist in areas where the coupling channel indicated signal loss.
Remember, in pitch-catch coupling checks not all losses of coupling signals
are true indications of loss of coupling.
Check the arrival times of root geometry indications to confirm that the
guide band position is correct.
Check for lines that would indicate missing data (scan was too fast). Some
may be tolerated but too many (for example >1-2% of scan length and no
more than 2 or 3 adjacent lines missing) could prevent assessment of the
weld to the acceptance criteria and risk missing flaws.
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.
When the scan of the weld is deemed to be of acceptable quality the process of
interpretation begins.
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
2
3
Flaws.
Geometry.
False indications (over-trace).
Flaws include, lack of fusion (sidewall), centreline cracks, incomplete
penetration, lack of cross-penetration, inter-run non-fusion (cold lap), porosity
and copper inclusions to mention a few. Only flaws are evaluated against the
acceptance criteria.
Geometry conditions could include mismatch (high-low), weld cap surface, root
surface or the pipe surface near the weld edges.
NDT44-50816
Pipeline Girth Welds
7-13
Copyright © TWI Ltd
www.twitraining.com
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.
Geometric and false indications must be identified correctly so as not to
compare the signal responses from these non-relevant indications to the
acceptance criteria. Since there are three display formats the operator begins
with an assessment of one format and then moves on to the others.
Most acceptance criteria are based on amplitude responses so the strip-chart
data is the first step in the analysis and then the operator moves on to the
volumetric and uses the ToFD as an aid in characterisation and sometimes as
an aid to sizing.
7.3
Evaluation thresholds
The first step is to review the strip-charts for any signals in excess of the
Evaluation Threshold. This is the amplitude a flaw signal must be in order to be
considered for evaluation against the allowed length criteria. It should be noted
that this level can vary.
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.
When equivalent reflector parameters are used it is seen that these Evaluation
thresholds are very similar (the main difference is merely the degree of
overtrace apparent due to increased off axis sensitivity when using 2mm
diameter FBH sensitivity).
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.
NDT44-50816
Pipeline Girth Welds
7-14
Copyright © TWI Ltd
www.twitraining.com
7.4
Examples of indication types
An example of each of the three flaw types is provided here. Operator
experience and logical assessment of the information available in the three
displays (strip-charts, B-scans and ToFD) can be used to construct a reasonable
assessment of the flaw or other feature of concern.
Geometry signals
Figure 7.14 Cap geometry due to dual torch Vee 32mm wall J-bevel.
False signal indications
NDT44-50816
Pipeline Girth Welds
7-15
Copyright © TWI Ltd
www.twitraining.com
Figure 7.15 False signal due to overtrace.
Flaw signal indications
Figure 7.16 Flaw indication of side wall lack of fusion.
7.5
Acceptance criteria and sizing
Once the operator decides that an indication is a flaw they must then determine
if the flaw is acceptable or not. Depending on the code worked to this may
mean a single step or a two-step process.
Most traditional acceptance criteria are based on good workmanship or what a
good welder can consistently weld. This has nothing to do with the effect of the
defect on the serviceability of the structure.
NDT44-50816
Pipeline Girth Welds
7-16
Copyright © TWI Ltd
www.twitraining.com
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.
These are generally based on some form of apportioning of height based on
amplitudes of a flaw detected in adjacent zones. Where feasible, the operator
may be able to use ToFD to improve the vertical size estimate.
Sizing, in all NDT methods, must be considered statistically, particularly when it
involves sizing in one dimension when the effects on the test are actually
derived from two dimensions as it is the case for ultrasonic tests.
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.
Standard deviation is the preferred scientific method of error assessment and
unlike simple average error preferred by certain users in the Gulf of Mexico;
standard deviation provides a realistic indication of precision of the sizing. To
simply consider overall accuracy provides no indication of the precision of the
individual assessments of sizing.
NDT44-50816
Pipeline Girth Welds
7-17
Copyright © TWI Ltd
www.twitraining.com
Accuracy is the degree of veracity while precision is the degree of
reproducibility. This is comparable to shooting at a target. Accuracy describes
the how close the shots are to the target centre.
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.
Figure 7.17 Accuracy versus precision.
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.
Estimations for standard deviation of error in sizing was a consideration when
the engineers first put BS PD 6493 together. In the 1980s they estimated SD =
2mm for automated UT size estimates of flaw height.
NDT44-50816
Pipeline Girth Welds
7-18
Copyright © TWI Ltd
www.twitraining.com
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.
A table of typical workmanship-style acceptance is seen in figure 7.18 and
typical ECA-based acceptance criteria are seen in figure 7.19.
Flaw
Cracks
Maximum allowed length
0
Linear surface (LS) indications
25mm aggregated in any 300mm or 8% of
weld length for welds with a total length of
less than 300mm
Linear buried (LB) indications
50mm aggregated in any 300mm or 8% of
the weld length for welds with a total
length of less than 300mm
Volumetric cluster (VC) indications
Maximum dimension exceeding 13mm
Volumetric individual (VI) indications
Maximum dimension of an individual flaw
exceeding 6mm in both width and length
Volumetric root (VR) indications
Maximum dimension exceeding 6mm or a
aggregate length exceeding 13mm in any
continuous 300mm of weld
Any accumulation of relevant
indications (AR) over the evaluation
level
50mm in any continuous 300mm length of
weld or 8% of the weld length for welds
less than 300mm total length
Figure 7.18 Workmanship-style acceptance criteria.
Figure 7.19 ECA-style acceptance criteria.
NDT44-50816
Pipeline Girth Welds
7-19
Copyright © TWI Ltd
www.twitraining.com
Section 8
Other Reading and References
8
Other Reading and References
Materials have been drawn from many sources to compile this handbook.
Websites are listed with credits to the authors where they are used in the body
of the book. These and other websites are also listed here.
Two other main sources of information drawn on were:
Phased Array presentations
by Dr. Michael Moles.
Ultrasonic Inspection 2 - Training for Non destructive Testing
by E. A. Ginzel.
Websites
History of UT (medical)
Dr. Woo on his website
http://www.ob-ultrasound.net
Characteristic Parameters of Ultrasonic Phased-array Probes and Equipment.
H. Wüstenberg, A. Erhard, G. Schenk
BAM - Berlin
http://www.ndt.net/article/v04n04/wuesten/wuesten.htm
Phased array technology concepts, probes and applications
Jerome Poguet
http://www.ndt.net/article/v07n05/poguet/poguet.htm
Phased Array Probes
http://www.imasonic.com/
http://www.vermon.com/
Phased Array Equipment
http://www.rd-tech.com
Texts
Introduction to Phased Array Ultrasonic Technology Applications,
R/D Tech Inc., Published by R/D Tech Inc., 2004, ISBN 0-9735933-0-X.
by
Advances in Phased Array Ultrasonic Technology Applications,
Olympus NDT, Published by Olympus NDT., 2007, ISBN 0-9735933-2-6.
by
Automated Ultrasonic Testing for Pipeline Girth Welds, A Handbook, by
E.A.Ginzel, Published by Olympus NDT, 2006, ISBN 0-9735933-4-2.
Understanding Ultrasound Physics: Fundamentals and Exam Review by Sydney
K. Edelman, Published by: Educational Sonographic Professional Inc.; 2nd
edition (June 1994) ISBN: 0962644439.
NDT44-50816
Other Reading and References
8-1
Copyright © TWI Ltd
www.twitraining.com
Essentials of Ultrasound Physics
by James A. Zagzebski, Publisher: Mosby; 1st edition (January 15, 1996) ISBN:
0815198523.
Data Acquisition Techniques Using Personal Computers, by
H. Austerlitz, Academic Press, 1991 ISBN 0-12-068370-9.
Ultrasonic Instruments and Devices: Reference for Modern Instrumentation,
Techniques, and Technology, by
Emmanuel Papadakis (Editor), Publisher: Academic Press; (January 2000)
ISBN: 0125319517.
Automated Ultrasonic Inspection of Welds, IIW Sub-Commission VC, The
International Institute of Welding 1989
Journal articles
Ultrasonic phased-arrays for non destructive testing,
McNab, A, Campbell, M.J., NDT International, 1987, vol.20, no.6, pp.333-337.
Current Applications and future trends in phased array technology, X.E.Gros,
N.B.Cameron, and M. King, Insight, November 2002, vol. 44, no. 10, pp673678.
NDT44-50816
Other Reading and References
8-2
Copyright © TWI Ltd
www.twitraining.com
Glossary of Terms
Glossary of Terms
The terms found in this glossary are the result of contributions from several of
the industry manufacturers and users.
Angle corrected gain
Can also be called ACG. This is compensation for the variation in signal
amplitudes received from a constant soundpath during S-scan calibration. The
compensation is typically performed electronically at multiple angles. Note that
there are technical limits to ACG, ie beyond a certain angular range
compensation is not possible.
Beam Apodization
Beam apodization is an electronic feature that is able to reduce side lobes by
applying lower voltage to the outside elements.
Beam steering
The ability of a phased array system to electronically sweep the beam through a
range of incident angles without probe movement.
Compound scan
Compound scan is a combination of electronic and sectorial scans that provides
multiplexing across different elements and sweeping through a defined range of
angles.
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.
Double resolution
Double resolution is a function available on electronic scans for increasing
lateral resolution.
Dynamic depth focusing (DDF)
A programmable, real-time array response on reception by modifying the delay
line, the gain and the excitation of each element as a function of time (see
figure below). DDF replaces multiple focal laws for the same focal range by the
convolution of the emitted beam with separate focused beams at the receiving
stage. In other words, DDF dynamically changes the focal distance as the signal
returns to the phased array probe. DDF significantly increases the depth of field
and Signal to Noise Ratio.
NDT44-060319
Glossary of Terms
9-1
Copyright © TWI Ltd
www.twitraining.com
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.
E-scans are equivalent to conventional ultrasonic probe performing a raster
scan. For angle beam scans, the Focal Law typically compensates for the
change in wedge thickness.
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.
Focusing (types): Depth focusing
Is one of several focusing options available to phased array probes. Focusing
may be at a specific plane, such as (True) Depth, Projection (vertical plane) or
some other Focal Plane focus.
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.
NDT44-060319
Glossary of Terms
9-2
Copyright © TWI Ltd
www.twitraining.com
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.
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.
The phased array system is computer-controlled with software typically user
friendly so that the operator can simply programme in the required inspection
parameters. Usually, a wedge is used to optimise inspection angles and
minimize wear. Phased arrays are particularly useful for regions with limited
access, rapid inspection of components such as welds, imaging and storing data
and sizing cracks by tip diffraction.
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.

Circular array probes: Elements on a cylinder, for tube inspection from
the inside without a mirror.

Convex array probe: A curved array probe designed typically for
inspection from the inside of tubes.

Concave array probe: A curved array probe designed typically for
inspection from the outside of tubes.

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.

Linear array probes: Probes made using a set of elements juxtaposed and
aligned along a linear axis. They enable a beam to be rastered, focused,
swept and steered along a single azimuthal plane (active axis) only.
NDT44-060319
Glossary of Terms
9-3
Copyright © TWI Ltd
www.twitraining.com

Matrix array probes: These probes have
dimensions using different elements. Matrix
checkerboard format though other designs
ultrasonic beam steering etcetera in multiple
an active area divided in two
array probes are typically in a
are used. These probes allow
planes.

Sectorial array probe: An annular array probe in which the annular rings
are subdivided into multiple elements.

Sparse matrix array: A matrix array containing less than 100% elements
such that effective gaps occur between elements. Sparse arrays are
typically used in larger arrays where instrumentation and array costs are
significant.

TRL PA: Transmitter-receiver longitudinal wave phased array probe. A dual
array probe generating longitudinal waves primarily used for austenitic
metal inspections.
Refracted steering angle
Deviation relative to the natural refracted angle.
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.
Terminology for array probes:

Grating lobe: Undesirable lobes of ultrasonic energy caused by the regular,
periodic spacing of array elements.

Active aperture: The size of the group of active acoustic elements (A).

Axial resolution: Axial resolution is the minimum distance along the
acoustic axis – for the same angle – for which two adjacent defects located
at different depths are clearly displayed by amplitude decay of more than 6
dB from peak-to-valley.

Cross-coupling: Also called cross-talk. An undesirable condition where
array elements are activated, electrically or acoustically, by adjacent
elements.

Element width: In a rectangular element, the acoustic element’s short
dimension (e).

Element length: In a rectangular element, the acoustic element’s long
(W), (L), (H) dimension, see passive aperture.
NDT44-060319
Glossary of Terms
9-4
Copyright © TWI Ltd
www.twitraining.com

Element pitch: The distance between the centres of two adjacent array
elements (p).

Elevation: The same as passive aperture.

Elevation focus: Focusing a transducer in the passive direction by either
applying a lens or by shaping the ceramic.

Lateral resolution: The minimum distance between adjacent defects
located at the same depths which produce amplitudes clearly separated by
at least 6dB from peak to valley.

Passive aperture: The dimension of an array element’s length.

Saw cut: Also called kerf (gap). The space between adjacent elements (g).

Virtual probe: A group of individual array elements, pulsed simultaneously
or at phasing intervals to generate a larger acoustic aperture.
e
W
p
g
A
Wedge parameters for phased array probes:

Coupling: Method for keeping the wedge to test material interface wet or
coupled. Can also be called irrigation.

Contoured wedges: These are wedges machined to match the contour of
the component, for example, curved to match a pipe circumference.

Wedges: Can be contoured in more than one dimension, see pictures for
examples.

Dual array wedge: Wedge made to accommodate two phased array
crystals. These are generally used in a pitch-catch application. They
normally have a barrier dividing the wedge in half.

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.

Natural refracted angle: The (unsteered) angle of the refracted beam into
a given material.
NDT44-060319
Glossary of Terms
9-5
Copyright © TWI Ltd
www.twitraining.com

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.

Refracted steering angle: The refracted angle (+/-) excursions the wedge
is designed to produce in the test material beyond the nominal refracted
angle produced from the wedge angle.

Roof angle: In a complex angle wedge, the roof angle is the minor angle,
while incident angle is the major angle.

Squint angle: Rotation of the phased array transducer on the wedge.

Wedge angle: The incident angle of a wedge as referenced to the normal
longitudinal axis.

Wedge velocity: The longitudinal wave speed of the wedge material.
Schematic
dimensions
of
wedge
and

(x1): Distance from back end of wedge to projected centre of first
element:(See picture below.)

(x2): Distance from projected centre of first element to diffuser or the thick
end of the wedge. The diffuser is usually serrated to diffuse unwanted
internal echoes.
NDT44-060319
Glossary of Terms
9-6
Copyright © TWI Ltd
www.twitraining.com
Schematic of various wedge types
NDT44-060319
Glossary of Terms
9-7
Copyright © TWI Ltd
www.twitraining.com
S
Supplementary Information
Download